page_content
stringlengths 39
150k
|
|---|
p. 802). As Ai, Peterson, Tice, Bolling, and Koenig (2004) observed, ‘hope seems to be more motivational and
emotional, whereas optimism is more strictly attitudinal and expectational’ (p. 437). Consistent with this key
conceptual distinction, hope by virtue of its agency subtrait is more closely linked than optimism to general
self-efficacy ( Bryant & Cvengros, 2004 ).
MEASURES REVIEWED HERE
Because hope and optimism may be best conceptualized as distinct but related constructs, we review these
measures separately and then discuss global and more domain-specific measures of these constructs. In each
case, we review evidence supporting the reliability and construct validity of the relevant measurement
instruments.
Although the instruments most commonly used to assess hope and optimism in current research were origi-
nally developed during the 1990s, these measurement scales continue today to dominate the field of personality,
social psychology, and behavioral medicine. The persistence and popularity of these instruments in contempo-
rary psychosocial research attests to the breadth and depth of their theoretical and psychometric foundations.
Yet, despite the continued dominance of these basic measurement tools, researchers have also developed new
measures of hope and optimism that reflect emerging conceptual advances and refinements, as well as specific
applications concerning future expectations in relation to particular situational or medical conditions. Because
these later developments in measurement build directly on earlier work, we begin by first reviewing the domi-
nant measurement tools for assessing hope and optimism, and then cover more recent advances in assessment of
these important constructs.
Global Measures of Hope
1.Snyder Hope Scale ( Snyder et al., 1991 ) & Children’s Hope Scale (Snyder et al., 1997)
2.Herth Hope Scale & Herth Hope Index ( Herth, 1991, 1992 )
3.Beck Hopelessness Scale ( Beck, 1993 )
4.Hunter Opinions and Personal Expectations Scale ( Nunn, Lewin, Walton, & Carr 1996 )
5.Integrative Hope Scale ( Schrank, Woppmann Mag, Sibitz, & Lauber 2010)
Domain-Specific Measures of Hope
1.Snyder State Hope Scale ( Snyder, Hoza, & Pelham 1996 )
2.Work Hope Scale (Juntenen & Wettersten, 2006)
Global Measures of Optimism
1.Life Orientation Test ( Scheier, Carver, & Bridges 1994 )
2.Generalized Expectancy for Success Scale ( Hale, Fiedler, & Cochran 1992 )
3.Personal and Social Optimism Questionnaire ( Schweizer & Koch, 2001 )
4.Positive and Negative Expectancies Questionnaire ( Olason & Roger, 2001 )
Domain-Specific Measures of Optimism
1.Cancer Patient Optimism Scale (Radwin et al., 2005)
2.HIV Treatment Optimism Scale (Van de Ven et al., 2010)
OVER VIEW OF THE MEASURES
Although several theorists have developed conceptual models of hope, by far ‘the predominant perspective on
hope in the research literature is Snyder’s cognitive conceptualization (e.g., Snyder et al., 1991, Snyder, Sympson,
Michael, & Cheavens; 2001 )’ (Kwon, 2002 , p. 208). Within Snyder’s framework, hope is defined as ‘a cognitive set
that is based on a reciprocally derived sense of successful (a) agency (goal-directed determination) and (b) path-
ways (planning of ways to meet goals)’ ( Snyder et al., 1991 , p. 571). From the perspective of this two-factor frame-
work, agency is a feeling of efficacy that one can work toward a goal and pathways refer to the development of
plans (or ‘ways’) to achieve desired goals (see also Snyder, 1989, 1994, 1995, 2000, 2002 ). Snyder’s model thus48 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
conceptualizes hope as a cognitive construct. This comprises individuals’ ability to work to personally important
goals ( Snyder, 1994; Snyder, Rand, & Sigmon 2002; Weis & Speridakos, 2012 ).
To assess hope, Snyder et al. (1991) constructed the Adult Hope Scale (AHS), consisting of four statements
designed to reflect agency, four statements designed to reflect pathways, and four unscored ‘filler’ items, with
which respondents indicate their extent of agreement. The reliability and validity of the SHS is backed by much
research. In addition, the SHS is associated with multiple meaningful psychological and physical variables.
A large body of evidence supports the reliability and validity of the SHS as a measure of dispositional hope, and
connects hope to a host of meaningful psychological and physical variables (see Snyder, 2002 ;Snyder et al.,
2001). Indeed, some suggest the SHS has the most construct and external validity of any measure of hope.
Empirical research supports Snyder et al.’s (1991) bidimensional model of hope. Factor analyses of the SHS
have revealed two distinct factors reflecting agency and pathways that are positively correlated, with the typical
magnitude of correlation being about .40 ( Snyder et al., 2001 ). Consistent with the notion of correlated factors,
confirmatory factor analyses have shown that responses to the SHS are accurately represented in terms of an
overarching higher-order hope construct defined by the agency and pathways subtraits ( Babyak, Snyder, &
Yoshinoba 1993; Bryant & Cvengros, 2004; Rand, 2009 ). Thus, researchers often use SHS total score as a unitary
measure of hope (e.g., Snyder et al., 1991 ).
Another body of research on the construct of hope is that of Dufault and Martocchio (1985) , as constructed by
Herth (1991) in the field of nursing research. They define hope as ‘a multidimensional dynamic life force charac-
terized by a confident yet uncertain expectation of achieving a future good which is realistically possible and
personally significant’ ( Dufault & Martocchio, 1985 , p. 380). Hope is assumed to include two spheres /C0generalized
hope, ‘an intangible umbrella that protects hoping persons by casting a positive glow on life’ ( Dufault &
Martocchio, 1985 , p. 380), and particularized hope, which concerns a specific outcome or hope object /C0which
include affective, behavioral, cognitive, affiliative, temporal, and contextual components.
To operationalize the constructs in Dufault and Martocchio’s (1985) model, Herth (1991) constructed the Herth
Hope Scale (HHS), consisting of 30 statements designed to tap three factors: Temporality and Future (the perceived
likelihood of attaining the desired outcome), Positive Readiness and Expectancy (feelings of confidence), and
Interconnectedness (the awareness of interdependence between self and others). With the HHS, respondents indi-
cate degree of agreement with each statement. Despite the underlying multidimensional model, researchers often
treat the HHS as measuring a single global construct. Although there are other existing models of hope, including
those of Beck (1993) ,Nunn and colleagues (1996) ,Hinds (1984) ,Nowotny (1989) ,Post-White, Ceronsky, and
Kreitzer (1996) , and Staats and Stassen (1985) , the conceptual models underlying Snyder et al.’s (1991) and Herth’s
(1991) instruments have been used most often in the empirical literature. More recently, interest has shifted toward
the development of context-dependent measures /C0like the Snyder State Hope Scale ( Snyder et al., 1996 ) and the
Work Hope Scale (Juntunen & Wettersten, 2006) to assess more transitory states of hope.
Optimism, in comparison, reflects a generalized positive expectancy for future outcomes. Although numerous
models of optimism have been developed (e.g., Chang, 2001 ;Colligan, Offord, Malinchoc, Schulman, & Seligman
1994;Dember, Martin, Hummer, Howe, & Melton 1989 ;Gillham, 2000 ;Levy, 1985 ;Malinchoc, Offord, & Colligan
1995), the most dominant theoretical orientation has undoubtedly been Scheier and Carver’s (1985) . In their origi-
nal conceptualization, Scheier and Carver (1985) defined optimism as a stable predisposition to ‘believe that good
rather than bad things will happen’ (p. 219). Thus, as opposed to hope, optimism was originally conceptualized
as a unitary trait representing a bipolar continuum, with optimism and pessimism.
To assess dispositional optimism, Scheier and Carver (1985) constructed the Life Orientation Test (LOT) con-
sisting of four statements designed to reflect optimism, four statements designed to reflect pessimism, and four
unscored ‘filler’ items, with which respondents indicate their extent of agreement. A large body of evidence sup-
ports the reliability and validity of the LOT as a measure of dispositional optimism, and connects optimism to a
variety of important psychological and physical outcomes (see Scheier & Carver, 1992 ). Although early critics
(e.g., Smith, Pope, Rhodewalt, & Poulton 1989 ) argued that the LOT is better conceived as a measure of neuroti-
cism rather than optimism, later evidence has supported the discriminant validity of the LOT as distinct from
general negative affectivity (e.g., Bryant & Baxter, 1997 ;Mroczek, Sprio, Aldwin, Ozer, & Bosse 1993 ;Scheier
et al., 1994 ). Indeed, far more evidence supports the construct validity and cross-sample generalizability of the
LOT than exists for any other existing measure of optimism.
Subsequent conceptual and empirical work led to the refinement of the original LOT. In particular, Scheier
et al. (1994) identified two LOT items that seem explicitly to reflect positive reinterpretation as a coping strategy,
rather than the expectation of positive outcomes per se. Because including these two problematic items in the
LOT might inflate the relationship between optimism and positive reinterpretation ( Affleck & Tennen, 1996 ),49 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Scheier et al. (1994) proposed a ‘minor modification’ (p. 1063) of the LOT, in which these two items and one pes-
simism item (‘Things never work out the way I want them to’) are omitted, and a new optimism item (‘Overall, I
expect more good things to happen to me than bad’) is added, yielding three items assessing optimism and three
items assessing pessimism. Supporting the use of this revised instrument, which is known as the LOT-R, Scheier
et al. (1994) reported relatively stable test /C0retest reliability and further suggested that the original and revised
versions of the instrument are essentially comparable, based on high correlations (‘in the .90s,’ p. 1073) between
total scores on the LOT and LOT-R.
Although Scheier and Carver (1985) originally conceptualized optimism as a global unidimensional trait, other
research indicates that dispositional optimism consists of two separate, but negatively correlated, subtraits reflect-
ing positively-framed optimism and negatively-framed pessimism. To explain discrepancies in outcome expec-
tancies within individuals, for instance, Dember et al. (1989) suggested optimism was bidimensional, consisting
of dispositional levels of both optimism and pessimism. According to Dember and colleagues (1989) , rejection of
pessimism is not equivalent to the endorsement of optimism; nor is the rejection of optimism equivalent to the
endorsement of pessimism. Empirical research suggests this two-factor model is preferable to a one-factor model,
although researchers disagree as to the extent of the correlation between optimism and pessimism.
Using confirmatory factor analysis to investigate the responses of 389 college students to the LOT, for example,
Chang, D’Zurilla, and Maydeu-Olivares (1994) found that a model consisting of correlated dimensions of opti-
mism and pessimism ( r52.54) fit better than did a one-factor ‘total score’ model. Indeed, Scheier and Carver’s
(1985) factor analyses of their own LOT data revealed two correlated ( r52.64) factors, Optimism and
Pessimism, which together fit their data significantly better than a one-factor model. In addition, Bryant and
Cvenrgos (2004) found that a two-factor model with correlated dimensions of Optimism and Pessimism
(r52.63) explained the responses of 351 college students to the LOT better than did a one-factor model.
Additional research with children ( Fischer & Leitenberg, 1986 ), college students ( Bailey, Eng, Frisch, & Snyder
2007), and older adults ( Mroczek et al., 1993; Robinson-Whelen, Kim, MacCallum, & Kiecolt-Glaser 1997 ) sup-
ports a bidimensional, as opposed to unidimensional, model of dispositional optimism. There is also evidence to
support the discriminant validity of this bidimensional structure in predicting mood ( Marshall, Wortman,
Kusulas, Hervig, & Vickers 1992 ) and physical health ( Robinson-Whelen et al., 1997 ).
Several researchers have directly examined the discriminant validity of hope and optimism as assessed by the
SHS and the LOT, respectively. For example, evidence indicates that optimism is more influential on the use of
positive reappraisal as a coping strategy than does hope, whereas hope has a stronger influence on level of gen-
eral self-efficacy than does optimism ( Bryant & Cvengros, 2004 ). Other work has found that hope uniquely influ-
ences college students’ grade expectancies, whereas optimism does not ( Rand, 2009 ); and that hope is a stronger
predictor of life satisfaction than is optimism ( Bailey et al., 2007 ). Moreover, hope and optimism each demon-
strate unique effects in predicting measures of life quality, including mastery, purpose in life, self-acceptance,
social integration, positive affect, and life satisfaction ( Gallagher & Lopez, 2009 ), as well as positive mental and
physical health ( Magaletta & Oliver, 1999 ) and level of depression in reaction to traumatic brain injury ( Peleg
et al., 2009 ).
Comparing the psychometric properties of measures of hope and optimism /C0the Revised Generalized
Expectancy for Success Scale (Hale et al., 1994), the LOT-R, the Hope Scale, and the Hunter Opinions and
Personal Expectations Scale ( Nunn et al., 1996 )/C0Steed (2002) concluded that ‘the LOT and HS are the scales of
choice when assessing hope and/or optimism’ (p. 446). Just as with hope, however, more recent additions to the
optimism measurement literature has focused on context. For example, the Cancer Patient Optimism Scale
(Radwin et al., 2005) and the HIV Treatment Optimism Scale (Van de Ven et al., 2000) have extended general
principles of the LOT-R to assess optimism in specific populations. Below, we describe and evaluate several
important measures of hope and optimism.
Adult Hope Scale (AHS): Children’s Hope Scale (CHS)
(Snyder et al., 1991).
Variable
Snyder et al. (1991) defined hope as ‘a cognitive set that is based on a reciprocally derived sense of successful
(a) agency (goal-directed determination) and (b) pathways (planning of ways to meet goals)’ (p. 571).50 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Description
The original SHS consisted of 12 statements to which respondents indicate their degree of agreement. Eight
items contain hope-related statements and four are fillers. In line with their definition of hope, the AHS contains
four items that assess agency and four items that assess pathways. The four filler items reflect other, non-hope
related constructs. Item response format is on a four point scale from 1 definitely false to 4 definitely true and
total hope scores can range from eight to 32 (Snyder et al., 1991). The AHS has also been normed for use with
children (also the Children’s Hope Scale (CHS) contains six hope statements reworded to be more understandable
for children with a six point Likert-type response scale from 1 none of the time to 6 all of the time and total scores
can range from six to 36) (Snyder et al., 1997).
Sample
The original SHS was validated with six separate samples ( N54126) composed of both introductory psychol-
ogy students and two clinical samples. The overall mean hope scores ranged from 22.60 ( SD54.35) to 25.64
(SD52.93) with an overall average of 24.70 ( SD53.33). Although males and females did not significantly differ
in hope scores, those in the clinical samples reported lower overall hope than those in the college samples
(Snyder et al., 1991). The CHS was validated with a sample of children aged 9 to 14 years ( N5372). Mean hope
scores were similar at pretest ( M525.41, SD54.99) and posttest ( M527.03, SD54.51) (Snyder et al., 1997).
Reliability
Internal Consistency
Cronbach alpha coefficients reported for the original measure range from .74 to .84 among the eight sample
groups (mean 5.77) (Snyder et al., 1991). Alpha coefficients for the Children’s Hope Scale range from .74 to .81
(Snyder et al., 1997).
Test/C0Retest
Test/C0retest reliability for the original measure is acceptable at three weeks ( r5.85), eight weeks ( r5.73), and
10 weeks ( r5.76;r5.82) (Snyder et al., 1991). Test /C0retest reliability for the children’s measure at one month was
r5.71 (Snyder et al., 1997). A recent meta-analysis ( Hellman, Pittman, & Munoz 2013 ) of the past two decades of
research using the SNH reported strong test /C0retest reliability coefficients that did not vary significantly across
different types of samples.
Validity
Convergent/Concurrent
SHS scores correlate positively with measures of optimism (Life Orientation Test r5.60; General Expectancies
for Success Scale r5.55), control (Burger /C0Cooper Life Experiences Survey r5.54; Problem Solving Inventory
r5.62), esteem (Rosenberg Self-esteem Scale r5.53) (Snyder et al., 1991). Also CHS scores correlate positively
with children’s assessment of competencies in scholastics ( r5.59), social acceptance ( r5.43), athletics ( r5.34),
physical appearance ( r5.46), and behavioral conduct ( r5.41) (Snyder et al. 1997).
Divergent/Discriminant
SHS scores correlate negatively with measures of hopelessness and depression (Beck Hopelessness Scale
r52.51; Beck Depression Inventory r52.42) and general measures of psychological problems (MMPI rs52.30 to
2.60) (Snyder et al., 1991). SHS scores do not correlate with the unrelated Self-Consciousness Scale ( rs5.06,2.03).
Also, CHS scores do not correlate with intelligence ( r5.03) (Snyder et al., 1997).
Construct/Factor Analytic
For the CHS, a principal components analysis with varimax rotation yielded a two-dimensional solution
accounting for approximately 58% of the total variance in line with Snyder and colleagues’ (1997) two-
dimensional definition of hope (agency and pathways).
Criterion/Predictive
SHS scores are predictive of high school ( r5.17) and college GPA ( r5.13) as well as expected grades ( r5.32)
and number of goals held by participants ( r5.24) (Snyder et al., 1991). CHS scores correlate positively with Iowa
Test of Basic Skills percentile scores ( r5.50) (Snyder et al., 1997).51 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Location
SHS
Snyder, C.R., Harris, C., Anderson, J.R., Holleran, S.A., Irving, L.M., Sigmon, S.T., et al. (1991). The will and
the ways: Development and validation of an individual-differences measure of hope. Journal of Personality and
Social Psychology, 60 , 570/C0585.
CHS
Snyder, C.R., Hoza, B., Pelham, W.E., Rapoff, M., Ware, L., Danovsky, M., Highberger, L., Ribinstein, H., &
Stahl, K.J. (1997). The development and validation of the Children’s Hope Scale. Journal of Pediatric Psychology,
22, 399/C0421.
Results and Comments
As the SHS is intended to assess the agency and pathways to success domains of hope, it is especially suited
to assess positive achievement cognitions. However, this measure does not seem to assess overall positive feel-ings for the future and as such, might be better utilized as a cognitive measures of confidence in personal abilityor efficacy.
ADULT HOPE SCALE-LIKE ITEMS
Directions: Read each item carefully. Using the scale
shown below, please select the number that best
describes YOU and put that number in the blank
provided.
15Definitely False 2 5Mostly False 3 5Mostly True
45Definitely True
____1. I can problem solve. (Pathways)
____2. I go for my goals. (Agency)
____3. I am exhausted a lot. (Filler)
____4. There are always solutions to issues I face.
(Pathways)
____5. I am not good at public speaking. (Filler)____6. I am a go-getter. (Pathways)
____7. I am anxious about things. (Filler)
____8. I am better than most of my friends at getting to
my goals. (Pathways)
____9. I am prepared for what lies ahead. (Agency)
____10. I am a successful person. (Agency)
Note: Copyright r 1991 American Psychological
Association. No portion of the Adult Hope Scale may bereproduced by any means without permission in writingfrom the copyright owner. These items are similar to theitems in the Adult Hope Scale.
CHILDREN’S HOPE SCALE
Directions: The six sentences below describe how children think about themselves and how they do things in
general. Read each sentence carefully. For each sentence, please think about how you are in most situations. Place a
check inside the circle that describes YOU the best. For example, place a check ( ü) in the circle ( ○) above ‘None of
the time,’ if this describes you. Or, if you are this way ‘All of the time,’ check this circle. Please answer every questionby putting a check in one of the circles. There are no right or wrong answers.
1. I think I am doing pretty well.
○○ ○○○○
None ofthe timeA little of
the timeSome of
the timeA lot of
the timeMost of
the timeAll of
the time
2. I can think of many ways to get the things in life that are most important to me.○○ ○○○○None ofthe timeA little of
the timeSome of
the timeA lot of
the timeMost of
the timeAll of
the time
3. I am doing just as well as other kids my age.
○○ ○○○○
None ofthe timeA little of
the timeSome of
the timeA lot of
the timeMost of
the timeAll of
the time52 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
4. When 1 have a problem, I can come up with lots of ways to solve it.
○○ ○○○○
None of
the timeA little of
the timeSome of
the timeA lot of
the timeMost of
the timeAll of
the time
5. I think the things I have done in the past will help me in the future.
○○ ○○○○
None of
the timeA little of
the timeSome of
the timeA lot of
the timeMost of
the timeAll of
the time
6. Even when others want to quit, I know that I can find ways to solve the problem.
○○ ○○○○
None of
the timeA little of
the timeSome of
the timeA lot of
the timeMost of
the timeAll of
the time
Notes : When administered to children, this scale is not labeled ‘The Children’s Hope Scale,’ but is called
‘Questions About Your Goals.’
The total Children’s Hope Scale score is achieved by adding the responses to the six items, with ‘None of the
time’51; ‘A little of the time’ 52; ‘Some of the time’ 53; ‘A lot of the time’ 54; ‘Most of the time’ 55; and, ‘All of
the time’ 56. The three odd-numbered items tap agency, and the three even-numbered items tap pathways.
Copyright r1997 Oxford University Press. No portion of the Children’s Hope Scale may be reproduced by any
means without permission in writing from the copyright owner.
Reproduced with permission.
Herth Hope Scale (HHS): Herth Hope Index (HHI)
(Herth, 1991, 1992 ).
Variable
The HHS and HHI were based on the Dufault and Martocchio (1985) definition of hope as ‘a dynamic life
force characterized by a confident yet uncertain expectation of achieving good, which, to the hoping person, is
realistically possible and personality significant’ (p. 380).
Description
The original HHS consisted of 30 statements to which respondents indicate their degree of agreement. Ten
items contain statements of temporality and future, ten items contain statements of positive readiness and expec-
tancy, and ten items contain statements of interconnectedness. Item response format is on a four-point scale from
0 never applies to 3 applies and total hope scores can range from 0 to 90 ( Herth, 1991 ). The HHI is an abbreviated
form (12-item) of the HHS used in clinical settings. Item response format is on a four point scale from 1 strongly
disagree to 4 strongly agree and total scores can range from 12 to 48 ( Herth, 1992 ).
The HHS has also been translated into a Spanish-language version ( Arnau, Martinez, Guzman, Herth, &
Konishi 2010 ).
Sample
The original HHS was validated with a sample of well adults ( N5185), a sample of elderly in the community
(N540), and a sample of bereaved elderly ( N5300). Hope scores for the well adults was relatively high ( M580,
range 60-90), slightly lower among elderly in the community ( M572,SD56.31), and lowest among bereaved
elderly ( M554,SD55.60) ( Herth, 1991 ). The HHI was validated with a sample of varyingly ill patients
(N5172). Average scores fell slightly above the scale midpoint ( M532.39, SD59.61) ( Herth, 1992 ). The Spanish
HHS was validated with a sample of 315 Latino students ( Arnau et al., 2010 ).53 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Reliability
Internal Consistency
Cronbach alpha coefficients reported for the origi nal HHS were .94 for the elderly and .95 for bereaved
elderly ( Herth, 1991 ). Similarly, the alpha coefficient for the HHI overall was .97, as well as .96 for adults
with chronic illnesses, .94 for adults with acute illnesses and .98 for adults with terminal illness
(Herth, 1992 ).
Test/C0Retest
Test/C0retest reliability for the original measure was reported at three weeks among healthy adults ( r5.90),
elderly ( r5.89), and bereaved elderly ( r5.91) ( Herth, 1991 ). Test/C0retest reliability for the HHI was reported at
two weeks ( r5.91) ( Herth, 1992 ).
Validity
Convergent/Concurrent
The HHI was positively correlated with the HSS ( r5.92), spiritual well-being ( r5.84), and the Nowotny Hope
Scale ( r5.81) ( Herth, 1992 ).
Divergent/Discriminant
HSS scores correlate negatively with depression (Beck Hopelessness Scale) among healthy adults ( r52.74)
and the elderly ( r52.69) ( Herth, 1991 ).
Construct/Factor Analytic
Factor analysis using combined data ( N5300) supported the three-factor structure (temporality and future,
positive readiness and expectancy, and interconnectedness) of the HHS among the bereaved elderly sample
(Herth, 1991 ). A separate maximum-likelihood factor analysis with varimax rotation supported the three-factor
HHI structure ( Herth, 1992 ).
Criterion/Predictive
HHS scores positively predict coping response ( r5.80) ( Herth, 1991 ).
No criterion/predictive data are currently available for the HHI.
Location
Original Measure (HHS)
Herth, K. (1991). Development and refinement of an instrument to measure hope . Scholarly Inquiry for Nursing
Practice: An International Journal, 5 ,3 9/C051.
Revised Index (HHI)
Herth, K. (1992). Abbreviated instrument to measure hope: Development and psychometric evaluation. Journal
of Advanced Nursing, 17 , 1251/C01259.
Spanish HHS
Arnau, R.C., Martinez, P., Guzman, I.N., Herth, K., & Konishi, C.Y. (2010). A Spanish-language version of
the Hearth Hope Scale: Development and psychometric evaluation. Educational and Psychological Measurement ,
70, 808/C0824.
Results and Comments
In addition to demonstrating strong reliability and validity, the HHS/HHI is especially thorough in assessing
both positive cognitions and emotions. As such, the HHS/HHI has been used extensively in health-related
fields.54 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
HERTH HOPE SCALE
Directions: Please indicate the extent to which the
following statements are true using a scale from 0 never
applies to 3 always applies.
____1. I am looking forward to the future.
____2. I have plans for the future.
____3. I feel scared about my future.*
____4. I feel time heals.
____5. I have hope even when plans go astray.
____6. I have goals for the next 3 /C06 months.
____7. I have coped well in the past.
____8. I can see a light even in a tunnel.
____9. I have plans for today and next week.
____10. I believe that each day has potential.
____11. I have inner positive energy.
____12. I keep going even when I hurt.
____13. I believe that good is always possible.
____14. I feel overwhelmed and trapped.*
____15. I just know there is hope.
____16. I am immobilized by fears and doubts.*
____17. I see the positive in most situations.____18. I am committed to finding my way.
____19. I believe my outlook affects my life.
____20. I can’t bring about positive change.*
____21. I sense the presence of loved ones.
____22. I have deep inner strength.
____23. I have a faith that gives me comfort.
____24. I feel at a loss, nowhere to turn.*
____25. I have support from those close to me.
____26. I can seek and receive help.
____27. I know my life has meaning and purpose.
____28. I feel all alone.*
____29. I feel loved and needed.
____30. I can recall happy times.
Notes . Copyright r1991 Kaye A. Herth. No portion of
the Herth Hope Index or the Herth Hope Scale may be
reproduced by any means without permission in writing
from the copyright owner, who may be contacted at
kaye.herth@mnsu.edu .
*Item reverse scored.
Reproduced with permission.
HERTH HOPE INDEX
Directions: Please indicate your agreement to the fol-
lowing statements using a scale from 1 strongly dis-
agree, 2 disagree, 3 agree, 4 strongly agree.
____1. I have specific possible short, intermediate, or
long range goals.
____2. I have a positive outlook on life.
____3. I believe that each day has potential.
____4. I am scared about the future.*
____5. I see a light at the end of the tunnel.
____6. I have a sense of direction.
____7. Life has value and worth.____8. I am able to recall happy/joyful times.
____9. I feel all alone.*
____10. I have a faith that comforts me.
____11. I have deep inner strength.
____12. I am able to give and receive caring/love.
Notes : Copyright r1992 John Wiley and Sons. No por-
tion of the Herth Hope Index or the Herth Hope Scale
may be reproduced by any means without permission in
writing from the copyright owner.
*Item reverse scored.
Reproduced with permission.
Beck Hopelessness Scale (BHS)
(Beck, 1993 ).
Variable
Beck et al. (1993) defined hopelessness as a lack of hope and optimism; a system of negative expectancies con-
cerning one’s self and future circumstances.55 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Description
The BHS consists of 20 statements to which respondents indicate whether or not that statement is true or false.
Of these, nine items are keyed false and 11 are keyed true. Possible total scores range from 0 to 20 with higher
scores suggesting higher levels of hopelessness ( Beck, 1993 ).
Sample
The BHS was validated with a clinical sample ( N5294) recently hospitalized for suicide attempts ( Beck, 1993 ).
Reliability
Internal Consistency
A Cronbach alpha coefficient was reported (.93) and item-scale correlations ranged from.36 to .76 ( Beck, 1993 ).
Test/C0Retest
No test /C0retest reliability evidence is currently available for the BHS.
Validity
Convergent/Concurrent
BHS scores correlate positively with pessimism (Stuart Future Test) ( r5.60) and depression (Beck Depression
Inventory) ( r5.63) ( Beck, 1993 ).
Divergent/Discriminant
No divergent validity evidence is currently available for the BHS.
Construct/Factor Analytic
A principal components analysis with varimax rotation suggested three BHS dimensions (relating to: feelings
about the future, loss of motivation, and future expectations) ( Beck, 1993 ).
Criterion/Predictive
BHS scores are predictive of clinician-rated measures of hopelessness ( r5.74) and attempted suicide rate
(r5.62).
Location
Beck, A.T. (1993). Beck Hopelessness Scale (BHS). San Antonio, TX: Psychological Corporation.
Results and Comments
The BHS is a classic measure for good reason. It is highly reliable and versatile given the generality of
its items. This scale is especially well-suited for clinical populations but commonly used with general adult
samples as well. As such, the BHS has been utilized in the validation of almost all other measures of hope and
optimism.
BHS-LIKE ITEMS
Directions: Please indicate whether the following
statements are true or false
____1. I have hope when I think about my future.*
____2. I often feel like giving up.
____3. Bad things will always turn around.*
____4. I can’t picture what the future will be like.
____5. I feel like I can accomplish anything.*
____6. I think I will do well for myself in the future.*
____7. The future doesn’t look good.____8. I expect the best things will happen to me.*
____9. Nothing goes right for me.
____10. I am ready for the challenges that lie ahead*.
Notes :C o p y r i g h t r1993 Pearson. No portion of the Beck
Hopelessness Scale may be reproduced by any means
without permission in writing from the copyright owner.
These items are similar to the items in the Beck
Hopelessness Scale.
*Item reverse scored.56 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Hunter Opinions and Personal Expectancies Scale (HOPES)
(Nunn et al., 1996 ).
Variable
Nunn et al. (1996) defined hope as, ‘that construction of, and response to, the perceived future, in which the
desirable is subjectively assessed to be probable’ (p. 532). Accordingly, hope has three domains including tempo-
rality (e.g., future oriented), desirability (desirable outcome), and expectancy (probability of that outcome
occurring).
Description
HOPES consists of 20 statements to which respondents indicate their degree of agreement. Ten of these items
reflect the hope subscale and 10 reflect the despair subscale. Item response-format is on a five-point scale from 0
‘describes me not at all’ to 4 ‘describes me extremely well,’ and total scores can range from 0 to 80 (hope score 5
hope subscale minus despair subscale plus 40).
Sample
The HOPES was developed using three separate samples ( N5307) including a medical school sample
(N5211), an adolescent male sample ( N52157), a hospital staff sample ( N5318), and a post-earthquake com-
munity sample ( N51089). Overall mean hope scores were high ( M558.72, SD510.89).
Reliability
Internal Consistency
Cronbach alpha coefficients for each sample have been reported ( αs5.80 to .92) ( Nunn et al., 1996 ).
Test/C0Retest
A stability coefficient (r 5.71) after 15 months has been reported ( Nunn et al., 1996 ).
Validity
Convergent/Concurrent
HOPES scores correlate positively with a measure of extraversion ( r5.39) ( Nunn et al., 1996 ).
Discriminant Validity
HOPES scores correlate negatively with trait anxiety ( r52.64), neuroticism ( r52.46), locus of control
(r52.46), and depression (Beck Depression Scale) ( rs52.51 to2.67) ( Nunn et al., 1996 ).
Construct/Factor Analytic
A principal components analysis yielded five dimensions when combining the HOPES with EPI neuroticism,
Spielberger’s Trait Anxiety Scale, Beck Depression Scale, and Rotter’s Locus of Control Scale, suggesting the
uniqueness of the HOPES instrument ( Nunn et al., 1996 ).
Criterion/Predictive
Individuals with the lowest HOPES scores also exhibit lower general health scores than those with the highest
HOPES scores. Similarly, individuals with the lowest HOPES scores are affected more by catastrophic events
(e.g., earthquakes) than those with the highest HOPES scores ( Nunn et al., 1996 ).
Location
Nunn K.P., Lewin T.J., Walton J.M., & Carr V.J. (1996). The construction and characteristics of an instrument to
measure personal hopefulness. Psychological Medicine, 26 , 531/C0545.
Results and Comments
Of all the measures of hope reviewed, the HOPES was validated in a uniquely iterative process and as such, is
perhaps the most thoroughly developed of the hope scales. HOPES is also unique in that is was validated with a
wide range of sample populations, making it extremely versatile.57 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
HOPES-LIKE ITEMS
Directions: Please read each statement below and
indicate how well the statement describes you in general
(i.e. Most of the time) by choosing one of the alternatives
from the five point scale (0 /C04) and writing its number
in the space on the left.
4 extremely well; 3 very well; 2 moderately well; 1
NOT very well 0 NOT at all.
____1. I am excited about the future.
____2. I will not be satisfied with my life.*
____3. Life has meaning to me.
____4. The future is uncertain.*
____5. No one cares about my future.*____6. I can handle the challenges I face.
____7. My life has value.
____8. I am so tired that I can’t do the things I want
to do.*
____9. People can’t expect much of the future.*
____10. My future will be productive.
Notes : Copyright r 1996 Cambridge University
Permissions .No portion of the HOPES may be repro-
duced by any means without permission in writing from
the copyright owner.
These items are similar to the items in the HOPES.
*Item reverse scored.
Integrative Hope Scale (IHS)
(Schrank et al., 2010 ).
Variable
Schrank et al. (2010) defined hope as a ‘ ...concept include[ing] a reality reference, in that the desired out-
comes or goals are subjectively perceived as being possible, and it allows for hope to arise both from a negative
as well as a positive starting point, i.e. as a desire for the improvement of an undesirable or an already satisfac-
tory situation’ (p. 418).
Description
The IHS is an integration of the Miller Hope Scale, the HHI, and the SHS discussed above. Originally consist-
ing of 60 items, the IHS was reduced to 23 items using factor analysis. In line with its composite scales, the IHS
contains items reflecting trust and confidence, lack of perspective, positive future orientation, and social relations
and personal values. Item response format is on a six point Likert-type forced-choice scale ranging from 1
‘strongly disagree’ to 6 ‘strongly agree’ (total scores range from 23 to 138) ( Schrank et al., 2010 ).
Sample
The IHS was validated with a sample ( N5489) from the Austrian general population. The overall mean hope
score was 93.78 ( SD512.83) (trust and confidence subdomain M527.81, SD54.03), (lack of perspective subdo-
main M515.10, SD55.39), (positive future orientation subdomain M520.01, SD52.90), (social relations and
personal value subdomain M519.06, SD53.33) ( Schrank et al., 2010 ).
Reliability
Internal Consistency
The Cronbach alpha coefficient reported was .92. Subscale alpha coefficients were also high (trust and confi-
dence alpha of .85; lack of perspective alpha of .85; positive future orientation alpha of .80) ( Schrank et al., 2010 ).
Validity
Convergent/Concurrent
IHS scores correlate positively with the SHS ( rs5.62 to .39), the HHI ( rs5.81 to .64), and the MHS ( rs5.92
to .73) ( Schrank et al., 2010 ).
Divergent/Discriminant
IHS scores correlate negatively with a measure of depression ( r52.68). ( Schrank et al., 2010 ).58 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Construct/Factor Analytic
Principle axis factor analysis using oblique rotation ( N5484) was used to assess the factor structure of the
IHS. The four-factor solution (labeled: trust and confidence, lack of perspective, positive future orientation, and
social relations and personal values) suggested in the definition of hope was supported ( Schrank et al., 2010 ).
Criterion/Predictive
IHS scores positively predict future quality of life ( r5.56). ( Schrank et al., 2010 ).
Location
Schrank, B., Woppmann Mag, A., Sibitz, I., & Lauber, C (2010). Development and validation of an integrative
scale to assess hope. Health Expectations, 14, 417/C0428.
Results and Comments
The IHS provides an excellent alternative to any of its composite measures. Those interested in harnessing the
power of the SHS, the HHI, and the Miller Hope Scale will find the IHS both ‘user friendly’ and psychometrically
sound. Schrank et al. (2010) noted that:
‘All three pre-existing scales have been used in research in a variety of fields and among healthy as well as diseased populations.
Hence, the new scale, being based on these instruments, can be assumed to be equally applicable in healthy as well as ill people while
at the same time having the advantage of most comprehensively covering the concept of hope, being concise, and psychometrically
robust according to our preliminary validation’ (p. 426).
INTEGRATIVE HOPE SCALE
Please indicate the extent to which you agree or
disagree with the following statements from 1, strongly
disagree, to 6, strongly agree.
1.I have deep inner strength.
2.I believe that each day has potential.
3.I have a sense of direction.
4.Even when others get discouraged, I know I can
find a way to solve the problem.
5.I feel my life has value and worth.
6.I can see possibilities in the midst of difficulties.
7.My past experiences have prepared me well for my
future.
8.I’ve been pretty successful in life.
9.I have a faith that gives me comfort.
10.It is hard for me to keep up my interest in activities
I used to enjoy.
11.It seems as though all my support has been
withdrawn.12.I am bothered by troubles that prevent my planning
for the future.
13.I am hopeless about some parts of my life.
14.I feel trapped, pinned down.
15.I find myself becoming uninvolved with most things
in life.
16.There are things I want to do in life.
17.I look forward to doing things I enjoy.
18.I make plans for my own future.
19.I intend to make the most of life
20.I feel loved.
21.I have someone who shares my concerns.
22.I am needed by others.
23.I am valued for what I am.
Notes : Copyright r2010 John Wiley and Sons. No por-
tion of the IHS may be reproduced by any means with-
out permission in writing from the copyright owner.
Reproduced with permission.
Snyder State Hope Scale (SSHS)
(Snyder et al., 1996 ).
Variable
Snyder et al. (1996) defined state hope as a temporally varying type of goal-directed thinking reflecting feel-
ings of optimism toward the future.59 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Description
The SSHS consists of six statements to which respondents indicate their degree of agreement. In line with the dis-
positional focus, the SSHS contains items assessing agency (goal-directly determination) and pathways (planning of
ways to meet goals) simply reworded to reflect the immediate situation. Item response format is an 8-point scale
ranging from 1 definitely false to 8 definitely true and total scores can range from zero to 48 ( Snyder et al., 1996 ).
Sample
The SSHS was validated with a sample ( N5444) composed of introductory psychology students over the
course of 30 days. The overall mean hope scores were similar for men ( M537.24) and women ( M537.06) with
an overall average of 37.15 ( SD56.33) ( Snyder et al., 1996 ).
Reliability
Internal Consistency
Cronbach alpha coefficients range from .82 to .95 (median α5.93) ( Snyder et al., 1996 ).
Test/C0Retest
Test/C0retest reliability for the original measure has been reported at three weeks ( r5.85), eight weeks ( r5.73),
and 10 weeks ( r5.76;r5.82) ( Snyder et al., 1996 ).
Validity
Convergent/Concurrent
SSHS scores correlate positively with dispositional optimism (SHS r5.79) and high dispositional hope is pre-
dictive of high state hope on a daily basis. State hope scores also correlate with state self-esteem ( r5.68) and state
positive affect ( r5.65). State Hope Scale scores also correlate with a daily report of their evaluations of that day
(r5.51) ( Snyder et al., 1996 ).
Divergent/Discriminant
SSHS scores correlate inversely with state negative affect ( r52.47) ( Snyder et al., 1996 ).
Construct/Factor Analytic
A principal components analysis with oblique rotation yielded a two-component solution in line with Snyder
and colleagues’ (1996) two-dimensional definition of hope (agency and pathways), respectively.
Criterion/Predictive
SHS scores are predictive of the number of correct responses on a complex verbal learning task ( r5.27)
(Snyder et al., 1996 ).
Location
Snyder, C.R., Sympson, S.C., Ybasco, F.C., Borders, T.F., Babyak, M.A., & Higgins, R.L. (1996). Development
and validation of the State Hope Scale. Journal of Personality and Social Psychology, 70 , 321/C0335.
Results and Comments
The SSHS assesses agency and pathways to success in the moment. As such, it is a valuable alternative to mea-
suring dispositional hope and especially well-suited to assess cognitions that one can achieve a goal in the
moment. It is also brief and potentially limited in its ability to assess how individuals feel if faced with multiple
problems at the same time, for example.
SSHS-LIKE ITEMS
Directions: Read each item carefully. Using the scale
shown below, please select the number that best
describes how you think about yourself right now and
put that number in the blank provided. Please take afew moments to focus on yourself and what is going
on in your life at this moment. Once you have this
‘here and now’ set, go ahead and answer each item
according to the following scale: 1 5Definitely False;60 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
25Mostly False; 3 5Somewhat False; 4 5Slightly False;
55Slightly True; 6 5Somewhat True; 7 5Mostly True;
85Definitely True.
____1. I can get myself out of problems.
____2. Right now I am going after my goals.
____3. I am overcoming problems I face right now.
____4. Currently, I am successful.
____5. I have so many ways to achieve the goals I have
right now.
____6. I feel like I am doing a good job right now of
doing what I set out to do.Notes : When administering the measure, it is labeled the
Goals Scale. The even-numbered items are agency, andthe odd-numbered items are pathways. Subscale scoresfor agency or pathways are derived by adding the threeeven- and odd-numbered items, and the total StateHope Scale score is the sum of all six items.Copyright r1996 American Psychological Association.
No portion of the State Hope Scale may be reproduced by
any means without permission in writing from the copy-
right owner.These items are similar to the items in the Snyder StateHope Scale.
Work Hope Scale (WHS)
(Juntenen & Wettersten, 2006).
Variable
Juntenen and Wettersten (2006) relied heavily on the Snyder and colleagues’ (1991) conceptualization of hope
as goals, pathways to achieve those goals, and motivation or agency to achieve those goals. Since their scalefocuses on occupational hope, Juntenen and Wettersten (2006) focused on work-related goals, pathways, andagency.
Description
The WHS consists of 24 statements to which respondents indicate their degree of agreement. In line with the
SSHS, the WHS contains items assessing goals, pathways, and agency. Item response format is on a 7-point
Likert-type scale ranging from 1 ‘strongly disagree’ to 7 ‘strongly agree’ and total scores ranging from 24 to 168
(Juntunen & Wettersten, 2006).
Sample
The WHS was validated with a sample ( N5224) composed of introductory psychology students, a community
sample, and a sample of low income women. The overall mean hope scores was 132.09 ( SD522.10) (Juntunen &
Wettersten, 2006).
Reliability
Internal Consistency
A Cronbach alpha coefficient was found to be .90 (Juntunen & Wettersten, 2006).
Test/C0Retest
Test/C0retest reliability for the WHS was reported after two weeks ( r5.90) (Juntunen & Wettersten, 2006).
Validity
Convergent/Concurrent
WHS scores correlate positively with measures of vocational identity ( r5.65), career-decision self-efficacy
(r5.75), agency ( r5.95), pathway ( r5.93), and goals ( r5.95) hope subscales (Juntunen & Wettersten, 2006).
WHS scores correlate positively with a measure of optimism (LOT-R r 5.53), significantly lower than the correla-
tions between WHS scores and vocational identity.
Divergent/Discriminant
No divergent/discriminant evidence is currently available for the WHS.61 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Construct/Factor Analytic
A confirmatory factor analysis as well as an exploratory factor analysis using maximum-likelihood, with obli-
que rotation suggested that the three-factor model was not optimal. Rather, the WHS is better conceptulaized as
a one-factor model (Juntunen & Wettersten (2006).
Criterion/Predictive
WHS scores are significantly lower among the economically disadvantaged and recipients of welfare as com-
pared with those who attend or have attended college (Juntunen & Wettersten, 2006).
Location
Juntunen, C.L., & Wettersten, K.B. (2006). Work hope: Development and initial validation of a measure. Journal
of Counseling Psychology, 53 ,9 4/C0106.
Results and Comments
Hope in specific domains, with the exception of clinical populations, has been neglected until recent years.
The WHS is especially worthwhile because it acknowledges that hope in general (and in clinical populations)
may be different than hope experienced in specific domains like the workplace. As such, the WHS is suited for
researchers interested in work-related feelings of hope.
WHS-LIKE ITEMS
Directions: Please indicate the extent of your agree-
ment with the following statements from 1 strongly
disagree to 7 strongly agree.
____1. I have plans to succeed in my profession.
____2. I know I won’t be able to find a good job.*
____3. Success at work comes easy for me.
____4. Work is personally rewarding for me.
____5. I don’t think I have the ability to do well at
my job.*
____6. I know how to find jobs that I like.
____7. My professional future looks good.____8. Things will work out well for me.
____9. I cannot find a job.*
____10. I want to work in the community in which
I live.
Notes : Copyright r2006 American Psychological
Association. No portion of the Work Hope Scale may be
reproduced by any means without permission in writing
from the copyright owner.
These items are similar to the items in the Work Hope
Scale.
*Item reverse scored.
Life Orientation Test (LOT/LOT-R)
(Scheier & Carver, 1985; Scheier et al., 1994 ).
Variable
Scheier and Carver (1985) define dispositional optimism as a stable personality trait characterized by general
positive expectancies.
Description
The original LOT consists of 12 statements to which respondents indicate their degree of agreement. Four
items contain positive statements assessing dispositional optimism, four items contain negative statements asses-
sing dispositional optimism, and four contain filler statements. Item response format is on a five point scale
from 0 strongly disagree to 4 strongly agree. Possible scores of overall optimism range from 0 to 32. The revised
LOT (LOT-R) consists of 10 items (three positive, three negative, three filler items). Two problematic coping items
were removed from the original LOT. Item response format mirrors that of the original measure and total scores
range from 0 to 24 ( Scheier et al., 1994 ).62 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Sample
The original LOT was validated with two separate university samples ( N5624). No differences emerged
between samples and total LOT scores were similar for males ( M521.03, SD54.56) and females ( M521.41,
SD55.22) ( Scheier & Carver, 1985 ). The LOT-R was validated with a university sample ( N52055). LOT-R scores
were similar for males ( M514.28, SD54.33) and females ( M514.42, SD54.12) ( Scheier et al., 1994 ).
Reliability
Internal Consistency
The Cronbach alpha coefficient reported for the original LOT was .76 and item-scale correlations were moder-
ate ( rs5.37 to .56) ( Scheier & Carver, 1985 ). Similarly, the alpha coefficient reported for the LOT-R was ( α5.78),
along with moderate item-scale correlations ( rs5.43 to .63) ( Scheier et al., 1994 ).
Test/C0Retest
Test/C0retest reliability for the original measure was .79 at four weeks (Scheier & Carver, 1994). Test /C0retest
reliability for the revised measure was .68 at four weeks ( r5.68), 12 weeks ( r5.60), 24 weeks ( r5.56), and
28 weeks ( r5.79) ( Scheier et al., 1994 ).
Validity
Convergent/Concurrent
LOT scores correlate positively with internal locus of control ( r5.34) and self-esteem ( r5.48) ( Scheier &
Carver, 1985 ). LOT-R scores correlate positively with self-mastery ( r5.48) and self-esteem ( r5.50). The original
and revised measures correlate strongly ( r5.95) ( Scheier et al., 1994 ).
Discriminant Validity
LOT scores do not correlate with private and public self-consciousness ( r52.04;r52.05) and correlate nega-
tively with hopelessness ( r52.47), depression ( r52.49), perceived stress ( r52.55), alienation ( rs5217 to
2.40), social desirability ( r5.26), and social anxiety ( r52.33) ( Scheier & Carver, 1985 ). LOT-R scores correlate
negatively with trait anxiety ( r52.53) and neuroticism ( rs52.36 to2.43) ( Scheier et al., 1994 ).
Construct/Factor Analytic
A principal components analysis as well as a confirmatory factor analysis supported a two-dimensional solu-
tion for positively and negatively-worded items for the LOT ( Scheier & Carver, 1985 ). Several sets of principal
components analyses yielded between 1 and 5 factors but Scheier et al. (1994) settled on a unidimensional model
of optimism for the LOT-R.
Criterion/Predictive
LOT scores correlate negatively with being bothered by physical symptoms ( rs52.22, to .31) ( Scheier & Carver,
1985). LOT-R scores correlate negatively with number of physical symptoms ( r52.21), intensity of symptom
(r52.25), mental disengagement ( r52.18), and use of drugs or alcohol ( r52.11) ( Scheier et al., 1994 ).
Location
Original 12-Item Measure
Scheier, I.H., & Carver, C.S. (1985). Optimism, coping and health: Assessment and implications of generalized
outcome expectancies on health. Health Psychology, 4 , 219/C0247.
Revised 10-Item Measure
Scheier, M.F., Carver, C.S., & Bridges, M.W. (1994). Distinguishing optimism from neuroticism (and trait anxi-
ety, self-mastery, and self-esteem): A re-evaluation of the Life Orientation Test. Journal of Personality and Social
Psychology ,67, 1063/C01078.
Results and Comments
The LOT and LOT-R are among the most commonly used measures of optimism primarily because they assess
the broadly defined construct of optimism. However, these measures primarily assess dispositional rather than
state or domain-specific optimism making its appropriate use limited to optimism as a stable personality trait.63 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
LOT-R-LIKE ITEMS
Directions: Please indicate the extent to which
you agree with each of the items using the following
response format:
05strongly disagree; 1 5disagree; 2 5neutral;
35agree; 4 5strongly agree.
____1. I know the best will happen even in the worst
of times.
____2. I know how to take time off. (filler item)
____3. Things never go right for me.*
____4. My future looks bright.
____5. Having friends is important to me. (filler item)____6. Keeping busy is important to me. (filler item)
____7. Things never go my way.*
____8. I am happy most of the time. (filler item)
____9. I can’t expect good things.*
____10. I anticipate more bad than good in my life.*
Notes : Copyright r1994 by the American Psychological
Association. No portion of the LOT-R may be repro-
duced by any means without permission in writing from
the copyright owner.
These items are similar to the items in the LOT-R.
*Item reverse scored.
Generalized Expectancy for Success Scale (GESS-R)
(Hale et al., 1992 ).
Variable
Fibel and Hale (1978) defined optimism as ‘the expectancy held by an individual that in most situations
he/she will be able to attain desired goals’ (p. 924). Hale and colleagues (1992) further noted that optimism
appears to be a stable personality trait.
Description
The original GESS consists of 30 statements to which respondents indicate their degree of agreement. Of these,
17 items contain statements about the likelihood of future successes and 13 contain statements about the likeli-
hood of future failure, reveres scored. Item response format is on a 5-point Likert-type scale from 1 ‘highly
improbable’ to 5 ‘highly probable’. Possible scores range from 30 to 150 with higher scores indicating higher
optimism ( Fibel & Hale, 1978 ). The revised GESS (GESS-R) consists of 25 items answered on the same scale
(Hale et al., 1992 ).
Sample
The original GESS was validated with three separate university samples ( N5307). No differences emerged
between samples and total GESS scores were similar for males ( M5112.15, SD513.24) and females ( M5112.32,
SD513.80) ( Fibel & Hale, 1978 ). The GESS-R was validated with a university sample (N 5199) and demonstrates
similar descriptive statistics ( M5107.00, SD55.72) ( Hale et al., 1992 ).
Reliability
Internal Consistency
Split-half reliability coefficients for the original measure were ( α5.91) for males and ( α5.90) for females
(Fibel & Hale, 1978 ). Similarly, the split-half reliability for the revised measure was ( α5.92) ( Hale et al., 1992 ).
Test/C0Retest
Test/C0retest reliability for the revised measure was .69 at six weeks ( Hale et al., 1992 ).
Validity
Convergent/Concurrent
GESS-R scores correlate positively with the LOT ( r5.40), self-esteem ( r5.46), and extraversion ( r5.16). GESS
scores correlate slightly with social desirability ( r5.26) and GESS-R scores correlate slightly but non-significantly
with Eysenck’s lie scale ( r5.18) ( Hale et al., 1992 ).64 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Divergent/Discriminant
GESS scores correlate negatively with measures of depression ( rs52.54 to2.69) and hopelessness ( rs52.31
to2.69) and GESS-R scores correlate negatively with introversion ( r52.23) and neuroticism ( r52.22)
(Hale et al., 1992 ).
Construct/Factor Analytic
The original and revised measures correlate strongly ( r5.98) ( Hale et al., 1992 ). No factor analytic evidence is
currently available for the GESS-R.
Criterion/Predictive
No criterion/predictive validity evidence is currently available for the GESS/GESS-R.
Location
Original 30-Item Measure
Fibel, B., & Hale, W.D. (1978). The Generalized Expectancy for Success Scale -A new measure. Journal of Consulting
and Clinical Psychology , 46, 924 /C0931.
Revised 25-Item Scale
Hale, W.D., Fiedler L.R., & Cochran C. D. (1992). The revised Generalized Expectancy for Success Scale:
A validity and reliability study. Journal of Clinical Psychology, 48, 517/C0521.
Results and Comments
The GESS-R demonstrates acceptable reliability and validity and as such, has been frequently used along
with the LOT/LOT-R in laboratory and field studies alike. As with the LOT/LOT-R, The GESS-R’s primary
strength lies in its broad conceptualization of optimism and measures that assess the variety of domains of
optimism.
GENERALIZED EXPECTANCY FOR SUCCESS SCALE-R
Instructions: On a scale from 1 (highly improbable)
to 5 (highly probable), how likely are the following
things to occur?
In the future I expect that I will ...
Succeed at most things I try.
Be listened to when I speak.
Carry through my responsibilities successfully.
Get the promotions I deserve.
Have successful close personal relationships.
Handle unexpected problems successfully.
Make a good impression on people I meet for the
first time.
Attain the career goals I have set for myself.
Experience many failures in my life.
Have a positive influence on most of the people with
whom I interact.
Be able to solve my own problems.
Acquire most of the things that are important to me.
Find that no matter how hard I try, things just don’t
turn out the way I would like.Be a good judge of what it takes to get ahead.
Handle myself well in whatever situation I’m in.
Reach my financial goals.
Have problems working with others.
Discover that the good in life outweighs the bad.
Be successful in my endeavors in the long run.
Be unable to accomplish my goals.
Be very successful working out my personal life.
Succeed in the projects I undertake.
Discover that my plans don’t work out too well.
Achieve recognition in my profession.
Have rewarding intimate relationships.
Find that people don’t seem to understand what I am
trying to say.
Notes : Copyright r1992 John Wiley and Sons. No por-
tion of the GESS may be reproduced by any means
without permission in writing from the copyright
owner.
Reproduced with permission.65 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Positive and Negative Expectancy Questionnaire for the Assessment of Personal
Optimism and Social Optimism /C0Extended (POSO-E)
(Schweizer & Koch, 2000).
Variable
Schweizer and Koch (2000) relied on a slightly modified version of Scheier and Carver’s (1985) definition of
optimism as a personality trait construct characterized by positive expectations. They further suggested that
optimism is both personal and socially-oriented.
Description
The POSO-E consists of 10 statements to which respondents indicate their degree of agreement. Item response-
format is on a four-point scale labeled 1 ‘incorrect,’ 2 ‘partly correct,’ 3 ‘almost correct,’ and 4 ‘completely correct,’
with total scores ranging from 10 to 40 (Schweizer & Koch, 2000).
Sample
The POSO-E was validated with an undefined sample ( N5348). No significant differences in optimism
emerged between males ( M528.47, SD54.08) and females ( M529.21, SD54.27) (overall M528.92, SD54.20)
(Schweizer & Koch, 2000).
Reliability
Internal Consistency
Cronbach alpha coefficients reported were .78 for the personal optimism subscale, .86 for the social optimism
subscale, and .87 for the self-efficacy subscale (Schweizer & Koch, 2000).
Test/C0Retest
No test /C0retest reliability evidence is currently available for the POSO-E.
Validity
Convergent/Concurrent
All POSO-E optimism subscales correlate positively with life satisfaction ( rs5.28 to .75) and negatively with
impulsiveness ( rs52.26 to 2.42), somatic complaints ( rs52.19 to 2.48), emotionality ( rs52.25 to 2.50),
depression ( rs52.26 to2.56), state anxiety ( rs52.31 to2.53), and neuroticism ( rs52.26 to2.63) (Schweizer &
Koch, 2000).
Divergent/Discriminant
None of the POSO-E optimism subscales correlate with aggressiveness ( rs52.05 to2.16), health concerns
(rs52.02 to2.09), or frankness ( rs52.07 to2.11) (Schweizer & Koch, 2000).
Construct/Factor Analytic
No factor analytic evidence is currently available for the POSO-E.
Criterion/Predictive
No criterion/predictive validity evidence is currently available for the POSO-E.
Location
Schweizer K. & Koch W. (2001). The assessment of components of optimism by POSO-E. Personality and
Individual Differences, 31 , 563/C0574.
Results and Comments
In addition to strong reliability and validity, the POSO-E is a valuable measure of three subdomains of opti-
mism. As its authors note, the POSO-E assesses feelings of personal optimism (i.e., expectations that personal
affairs will go well), feelings of social optimism (i.e., expectations that social situations will go well), and66 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
self-efficacy optimism (e.g., degree to which individuals will be able to bring out positive outcomes). Researchers
interested in these specific domains will likely find this measure of optimism especially useful.
POSO-E-SCALE
Participants respond to each statement using a
4-point scale ‘Completely correct’; ‘almost correct’;‘partly correct’; ‘incorrect’.
1.For each problem I will find a solution.
2.In difficult situations, I will find a way.
3.No task is too difficult to me.
4.I master difficult problems.
5.There is no task which is too demanding for me.
6.I even master new tasks without problems.7.I welcome every new challenge.
8.I can master difficulties.
9.I have a lot of confidence in myself.
10.I always find a solution to a problem.
Notes : Copyright r2000 Elsevier Limited. No portion
of the POSO-E Scale may be reproduced by any meanswithout permission in writing from the copyrightowner.
Reproduced with permission
Positive and Negative Expectancy Questionnaire (PANEQ)
(Oalson & Roger, 2001).
Variable
As with other conceptualizations of optimism, Oalson and Roger (2001) defined optimism as general positive
expectancies. However, they noted the importance of distinguishing optimism and pessimism as more thanopposite constructs but as separate but related constructs.
Description
The PANEQ consists of 76 statements to which respondents indicate their degree of agreement. These state-
ments reflect negative affect/pessimism, fighting spirit, and positive affect/optimism. Item response format is ona five point scale from 0 strongly disagree to 4 strongly agree. One score is obtained for each of these domainsseparately (Oalson & Roger, 2001).
Sample
The PANEQ was validated with a university sample ( N5216). No significant differences in optimism
emerged between males ( M528.47, SD54.08) and females ( M529.21, SD54.27) (overall M528.92, SD54.20)
(Oalson & Roger, 2001).
Reliability
Internal Consistency
The Cronbach alpha coefficient reported for the PANEQ optimism subscale was .75 (Oalson & Roger, 2001).
Test/C0Retest
Test/C0retest reliability for the PANEQ optimism subscale was reported after an interval of six weeks ( r5.82)
(Oalson & Roger, 2001).
Validity
Convergent/Concurrent
PANEQ optimism scores correlate positively with other measures of optimism (LOT-R r5.33; GESS-R r5.27),
and positive affect ( r5.11) (Oalson & Roger, 2001).
Divergent/Discriminant
PANEQ optimism scores do not correlate with negative affect ( r52.04) or coping ( r#2 .08) (Oalson &
Roger, 2001).67 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Construct/Factor Analytic
A principal components analysis with varimax rotation as well as a separate confirmatory factor analysis
supported three dimensions labeled: negative affect/pessimism, fighting spirit, and positive affect/optimism
(Olason & Roger, 2001 ).
Criterion/Predictive
PANEQ scores positively predict satisfaction with life ( r5.36) ( Olason & Roger, 2001 ).
Location
Olason, D.T., & Roger, D. (2001). Optimism, pessimism and ‘fighting spirit’: A new approach to assessing
expectancy and adaptation. Personality and Individual Differences, 31 , 755/C0768.
Results and Comments
The PANEQ is unique in that it conceptualizes optimism and pessimism not as two sides of the same con-
struct, but rather as two separate, albeit negatively related, constructs. Taking this into consideration, the PANEQ
is useful for assessing both optimism and pessimism as well as ‘fighting spirit’ self-confidence. This seems to
be its primary advantage above other measures which may not adequately assess these related but
distinct constructs.
PANEQ-LIKE ITEMS
Item response in on a four point Likert scale from 1
strongly disagree to 4 strongly agree.
1.My feelings often irritate me. (Negative Affect/
Pessimism)*
2.It doesn’t take much to stress me out. (Negative
Affect/Pessimism)*
3.If I had to take a test tomorrow, I would expect to
fail. (Negative Affect/Pessimism)*
4.I often imagine that the worst possible thing is about
to happen. (Negative Affect/Pessimism)*
5.I am a determined person. (Fighting Spirit)
6.I am a strong person. (Fighting Spirit)7.I am a fighter. (Fighting Spirit)
8.I am easily pleased. (Positive Affect/Optimism)
9.I often get so happy I have to be peeled off the
ceiling. (Positive Affect/Optimism)
10.I am a fortunate person. (Positive Affect/Optimism)
Notes : Copyright r2001 Elsevier Limited. No portion
of the PANEQ Scale may be reproduced by any means
without permission in writing from the copyright
owner.
These items are similar to the items in the PANEQ
Scale.
*Item reverse scored.
Cancer Patient Optimism Scale (CPOS)
(Radwin et al., 2005).
Variable
Radwin and colleagues (2005) defined optimism as the ‘patient’s belief that he or she had made appropriate
choices regarding treatment and the patient’s hopefulness about treatment outcomes’ (p. 93). Although they used
hope in their definition, Radwin et al. (2005) acknowledged the limitations of measures of hope used in medical
settings (e.g., Herth Hope Scale) to measure optimism (Radwin et al., 2005).
Description
The CPOS is part of a measure that assesses patient fortitude, trust in nurses, and authentic self-
representation. Although the scale consists of 16 items, only four are devoted to assessing patient optimism.
Participants are asked to answer each question using a six-point scale ranging from 1 ‘never’ to 6 ‘always.’
Scores are then transformed into a 0 /C0100 range, with higher numbers reflecting higher levels of optimism
(Radwin et al., 2005).68 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Sample
The CPOS was validated with a sample of cancer patients ( N566). Optimism scores ranged from 35 to 100
(M572.62, SD513.01) (Radwin et al., 2005).
Reliability
Internal Consistency
The Cronbach alpha coefficient for the CPOS was reported (.75) and item-scale correlations ranged from .44 to
.65 (Radwin et al., 2005).
Test/C0Retest
No test /C0retest reliability evidence is currently available for the CPOS.
Validity
Convergent/Concurrent
CPOS scores correlate positively with a measure of trust in nurses ( r5.33), and with a measure of authentic
self-representation ( r5.20) (Radwin et al., 2005).
Divergent/Discriminant
No discriminant validity evidence is currently available for the CPOS.
Construct/Factor Analytic
No factor analytic information is currently available for the CPOS.
Criterion/Predictive
CPOS scores are predictive of scores on a measure of fortitude ( r5.43) (Radwin et al., 2005).
Location
Radwin, L.E., Washko, M., Suchy, K.A., & Tyman, K. (2005). Development and pilot testing of four desired
health outcome scales. Oncology Nursing Forum, 32, 92/C096.
Results and Comments
The CPOS is unique in that is applies principles of optimism to a very specific (cancer patient) population.
Although the authors acknowledge that other scales have been developed to assess hope among cancer patients
(i.e., Nowotny, 1989 ), they also acknowledge the dearth of scales for assessing cancer patients.
CANCER PATIENT OPTIMISM SCALE
Respondents rank the frequency of the activity or
feeling on a Likert scale (1 5never; 2 5rarely; 3 5some
of the time; 4 5a good bit of the time; 5 5usually,
65always).
How often have you felt that your medical problem
will not work out for the best?
How often have you felt that your decisions about
how to treat your cancer were correct?How often have you felt that the cancer treatment
you chose would produce the desired outcome?
How often have you felt grim about the ways things
will work out for you?
Notes : Copyright r2005 Oncology Nursing Forum.
No portion of the Cancer Patient Optimism Scale may
be reproduced by any means without permission in
writing from the copyright owner.
Reproduced with permission.
HIV Treatment Optimism Scale (HIV-TOS)
(Van de Ven et al., 2000).69 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Variable
Van de Ven and colleagues (2008) developed a domain-specific scale to measure optimism regarding HIV
treatment programs among gay men.
Description
The HIV-TOS consists of 12 statements to which respondents indicate their degree of skepticism or optimism.
These statements concern (1) the effectiveness of HIV treatments and (2) the hope of change in the virus. Itemresponse-format is on a four-point scale ranging from 1 strongly disagree (highly skeptical) to 4 strongly agree(highly optimistic). One score is obtained for each of these three factors separately and total scores can rangefrom 12 (highly skeptical) to 48 (highly optimistic) (Van de Ven et al., 2000).
Sample
The HIV-TOS was validated with a sample of men ( N5532), most of whom were HIV positive and self-identified
as gay. The sample indicated a slightly skeptical orientation (overall M519.8, SD54.7) (Van de Ven et al., 2000).
Reliability
Internal Consistency
A Cronbach alpha coefficient reported for the HIV-TOS was .79. Item total correlations ranged from .28 to .57
(Van de Ven et al., 2000).
Test/C0Retest
No test /C0retest reliability evidence is currently available for the HIV-TOS.
Validity
Convergent/Concurrent
No convergent/concurrent validity evidence is currently available for the HIV-TOS.
Divergent/Discriminant
No discriminant validity evidence is currently available for the HIV-TOS.
Construct/Factor Analytic
Factor analysis yielded a three-factor solution (labeled: optimism, effectiveness of HIV treatments, and hope of
change in the virus) (Van de Ven et al., 2010).
Criterion/Predictive
No criterion/predictive evidence are currently available for the HIV-TOS.
Location
Van de Ven, P., Crawford, J., Kippax, S., Knox, S., & Prestage, G. (2010). A scale of optimism-scepticism in the
context of HIV treatments. AIDS Care: Psychological and Socio-Medical Aspects of HIV/AIDS, 12 , 171/C0176.
Results and Comments
The HIV-TOS is unique in that it examines treatment optimism in an important clinical population. Although
inappropriate for use in assessing more dispositional forms of optimism, the HIV Treatment Optimism Scaleadds to the increasing body of research aimed at examining state-dependent types of optimism.
HIV TREATMENT OPTIMISM SCALE
Instructions: The statements on this page are about
viral load testing and new treatments for HIV. For eachone, please tick if you strongly disagree, disagree, agree
or strongly agree. [A corresponding four-point scale
accompanied each statement.] For each question, pleasetick one box only ...if you are unsure, please give your
best guess.
____1. A person with undetectable viral load cannot
pass on the virus.70 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
____2. I’m less worried about HIV infection than I
used to be.
____3. New HIV treatments will take the worry out
of sex.
____4. If every HIV-positive person took the new
treatments, the AIDS epidemic would be over.
____5. If a cure for AIDS were announced, I would
stop practising safe sex.
____6. People with undetectable viral load do not need
to worry so much about infecting others
with HIV.
____7. Until there is a complete cure for HIV/AIDS,
prevention is still the best practice*.
____8. The availability of treatment (PEP) immediately
after unsafe sex makes safe sex less important.____9. HIV is less of a threat because the epidemic is
on the decline.
____10. HIV/AIDS is a less serious threat than it used
to be because of new treatments.
____11. It’s never safe to fuck without a condom
regardless of viral load.*
____12. Because of new treatments fewer people are
becoming infected with HIV.
Notes : Copyright r2000 Taylor and Francis. No portion
of the HIV Treatment Optimism Scale may be repro-
duced by any means without permission in writing from
the copyright owner.
*Item reverse scored.
Reproduced with permission.
FUTURE RESEARCH DIRECTIONS
As part of the conceptual bedrock of positive psychology, hope and optimism are integral to understanding
individuals’ perceptions and expectations for the future. Not surprisingly, many measures have been developed
to assess the positivity or negativity of these expectations. Although space permits only a limited discussion
of some of the more widely-used of these measures, it is clear that measures of hope generally reflect some level
of individual responsibility in bringing about a desired state. As such, measures of hope often assess feelings of
self-efficacy or motivational states that one can and will achieve a desired end goal. Conversely, measures of opti-
mism reflect an individual’s general belief that good things will happen. As such, measures of optimism tend to
assess a generalized expectancy of a desired end goal without self-efficacy or motivation that one can bring about
that state themselves. Of course, measures of hope and optimism do overlap in that both assess some positive
attitude, perception, or expectation for the future.
Although the measures discussed in this chapter purport to measure similar constructs, there is variability in
the nature of the items and response formats /C0even within the domains of hope and optimism themselves.
To date, there is little data demonstrating the inter-relations among these measures and no data demonstrating
the superiority of one measure over another. Nor is there data supporting the use of one measure among particu-
lar populations relative to others. Future research should attempt to address these limitations by explicitly exam-
ining response patterns and data patterns for the different measures together. Doing so could provide some
evidence of the preferability of one instrument over another in a given population.
Future research on hope and optimism should not only continue to develop explicit, questionnaire-based mea-
sures, but also consider developing implicit or unobtrusive measures of hope and optimism. As hope and opti-
mism seem to be desirable traits and/or states, it is possible that responses to self-report items may be subject to
evaluation apprehension. Indeed, the average hope and optimism scores in validation studies, discussed above,
seem to yield relatively high levels of these constructs. As the scales reviewed in this chapter wholly rely on
explicit self-report, participants may alter their responses consciously or unconsciously. Thus, measures using
implicit associations may provide useful in accessing both conscious and unconscious expectations relating to the
future. Similarly, unobtrusive observations or informant reports may provide an additional glimpse into indivi-
duals’ perceptions and expectations above and beyond what these individuals are willing to self-report.
In line with the current trend toward development of neurological profiles of positive psychological constructs
like savoring (e.g., Bryant, Chadwick, & Kluwe 2011 ), future research should consider the neurological basis of
hope and optimism. Theoretically, hope relies more on self-awareness, suggesting the importance of higher order
thought processes located in the prefrontal cortex, for example. Conversely, optimism might be more primal, sug-
gesting a deeper, more mammalian component to its neural profile. Future research should determine the regions
of interest in hope and optimism to further clarify and conceptually distinguish these distinct but related percep-
tions of the future. Considerable thought and effort has gone into developing the diverse measures of hope and71 FUTURE RESEARCH DIRECTIONS
II. EMOTIONAL DISPOSITIONS |
optimism discussed above. Although there is still much to learn about these distinct but related constructs, the
breadth (and depth) of measures purporting to assess hope and optimism highlight the wealth of knowledge
accumulated by research over the last several decades. Now that this information has been accumulated, it is up
to positive psychologists to continue to refine and test these measures to ensure their optimal use. Only then
might we truly understand how hope and optimism form the bedrock on which modern positive psychology
is based.
References
Affleck, G., & Tennen, H. (1996). Construing benefits from adversity: Adaptational significance and dispositional underpinnings. Journal of
Personality ,64, 899/C0922.
Ai, A. L., Peterson, C., Tice, T. N., Bolling, S. F., & Koenig, H. G. (2004). Faith-based and secular pathways to hope and optimism subconstructs
in middle-aged and older cardiac patients. Journal of Health Psychology ,9, 435/C0450.
Arnau, R. C., Martinez, P., Guzman, I. N., Herth, K., & Konishi, C. Y. (2010). A Spanish-language version of the Hearth Hope Scale:
Development and psychometric evaluation. Educational and Psychological Measurement ,70, 808/C0824.
Babyak, M. A., Snyder, C. R., & Yoshinoba, L. (1993). Psychometric properties of the Hope Scale: A confirmatory factor analysis. Journal of
Research in Personality ,27, 154/C0169.
Bailey, T. C., Eng, W., Frisch, M. B., & Snyder, C. R. (2007). Hope and optimism as related to life satisfaction. Journal of Positive Psychology ,2,
168/C0175.
Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects . New York: Harper & Row.
Beck, A. T. (1993). Beck Hopelessness Scale (BHS) . San Antonio, TX: Psychological Corporation.
Bryant, F. B., Chadwick, E. D., & Kluwe, K. (2011). Understanding the processes that regulate positive emotional experience: Unsolved pro-
blems and future directions for theory and research on savoring. International Journal of Wellbeing ,1, 107/C0126.
Bryant, F. B., & Cvengros, J. A. (2004). Distinguishing hope and optimism: Two sides of a coin, or two separate coins? Journal of Social and
Clinical Psychology ,23, 273/C0302.
Bryant, F. B., & Baxter, W. J. (1997). The structure of positive and negative automatic cognition. Cognition and Emotion ,11, 225/C0258.
Bryant, F. B., & Veroff, J. (1984). Dimensions of subjective mental health in American men and women. Journal of Health and Social Behavior ,25,
116/C0135.
Carvajal, S. C., Clair, S. D., Nash, S. G., & Evans, R. I. (1998). Relating optimism, hope, and self-esteem to social influences in deterring
substance use in adolescents. Journal of Social and Clinical Psychology ,17, 443/C0465.
Chang, E. C. (Ed.), (2001). Optimism and pessimism . Washington, DC: American Psychological Association.
Chang, E. C., D’Zurilla, T. J., & Maydeu-Olivares, A. (1994). Assessing the dimensionality of optimism and pessimism using a multimeasure
approach. Cognitive Therapy and Research ,18, 143/C0160.
Colligan, R. C., Offord, K. P., Malinchoc, M., Schulman, P., & Seligman, M. E. P. (1994). CAVEing the MMPI for an Optimism-Pessimism
Scale: Seligman’s attributional model and the assessment of explanatory style. Journal of Clinical Psychology ,50,7 1/C095.
Dember, W. N., Martin, S., Hummer, M. K., Howe, S., & Melton, R. (1989). The measurement of optimism and pessimism. Current Psychology:
Research and Reviews ,8, 102/C0119.
Dufault, K. J., & Martocchio, B. C. (1985). Hope: Its spheres and dimensions. Nursing Clinics of North America ,20, 379/C0391.
Fibel, B., & Hale, W. D. (1978). The Generalized Expectancy for Success Scale /C0A new measure. Journal of Consulting and Clinical Psychology ,
46, 924/C0931.
Fischer, M., & Leitenberg, H. (1986). Optimism and pessimism in school-aged children. Child Development ,57, 241/C0248.
Gallagher, M. W., & Lopez, S. J. (2009). Positive expectancies and mental health: Identifying the unique contributions of hope and optimism.
Journal of Positive Psychology ,4, 548/C0556.
Gillham, J. E. (Ed.), (2000). The science of optimism and hope: Research essays in honor of Martin E.P. Seligman . Philadelphia, PA: Templeton
Foundation.
Hale, W. D., Fiedler, L. R., & Cochran, C. D. (1992). The revised Generalized Expectancy for Success Scale: A validity and reliability study.
Journal of Clinical Psychology ,48, 517/C0521.
Hellman, C. M., Pittman, M. K., & Munoz, R. T. (2013). The first twenty years of the will and the ways: An examination of score reliability dis-
tribution on Snyder’s Dispositional Hope Scale. Journal of Happiness Studies ,14, 723/C0729.
Herth, K. (1991). Development and refinement of an instrument to measure hope. Scholarly Inquiry for Nursing Practice: An International Journal ,
5,3 9/C051.
Herth, K. (1992). Abbreviated instrument to measure hope: Development and psychometric evaluation. Journal of Advanced Nursing ,17,
1251/C01259.
Hinds, P. S. (1984). Inducing a definition of ‘hope’ through the use of grounded theory methodology. Journal of Advanced Nursing ,9, 357/C0362.
Kwon, P. (2002). Hope, defense mechanisms, and adjustment: Implications for false hope and defensive hopelessness. Journal of Personality ,
2002(70), 207 /C0213.
Levy, D. A. (1985). Optimism and pessimism: Relationships to circadian rhythms. Psychological Reports ,57, 1123/C01126.
Magaletta, P. R., & Oliver, J. M. (1999). The hope construct, will, and ways: Their relations with self-efficacy, optimism, and general well-
being. Journal of Clinical Psychology ,55, 539/C0551.
Malinchoc, M., Offord, K. P., & Colligan, R. C. (1995). PSM-R: Revised Optimism-Pessimism Scale for the MMPI-2 and MMPI. Journal of
Clinical Psychology ,51, 205/C0214.
Marshall, G. N., Wortman, C. B., Kusulas, J. W., Hervig, L. K., & Vickers, R. R. (1992). Distinguishing optimism and pessimism: Relations to
fundamental dimensions of mood and personality. Journal of Personality and Social Psychology ,62, 1067/C01074.72 3. MEASURES OF HOPE AND OPTIMISM
II. EMOTIONAL DISPOSITIONS |
Mroczek, D. K., Sprio, A., Aldwin, C. M., Ozer, D. J., & Bosse, R. (1993). Construct validation of optimism and pessimism in older men:
Findings from the Normative Aging Study. Health Psychology ,12, 406/C0409.
Nowotny, M. L. (1989). Assessment of hope in patients with cancer: Development of an instrument. Oncology Nursing Forum ,16,5 7/C061.
Nunn, K. P., Lewin, T. J., Walton, J. M., & Carr, V. J. (1996). The construction and characteristics of an instrument to measure personal hopeful-
ness. Psychological Medicine ,26, 531/C0545.
Olason, D. T., & Roger, D. (2001). Optimism, pessimism and ‘fighting spirit’: A new approach to assessing expectancy and adaptation.
Personality and Individual Differences ,31, 755/C0768.
Peleg, G., Barak, O., Harel, Y., Rochberg, J., & Hoofien, D. (2009). Hope, dispositional optimism and severity of depression following traumatic
brain injury. Brain Injury ,23, 800/C0808.
Pope, A. (1733/2006). An essay on man . Stockbridge, MA: Hard Press.
Post-White, J., Ceronsky, C., Kreitzer, M. J., et al. (1996). Hope, spirituality, sense of coherence, and quality of life in patients with cancer.
Oncology Nursing Forum ,23, 1571/C01579.
Rand, K. L. (2009). Hope and optimism: Latent structures and influences on grade expectancy and academic performance. Journal of
Personality ,77, 231/C0260.
Robinson-Whelen, S., Kim, C., MacCallum, R. C., & Kiecolt-Glaser, J. K. (1997). Distinguishing optimism from pessimism in older adults: Is it
more important to be optimistic or not to be pessimistic? Journal of Personality and Social Psychology ,73, 1345/C01353.
Rotter, J. B. (1954). Social learning and clinical psychology . Englewood Cliffs, NJ: Prentice-Hall.
Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health: Assessment and implications of generalized outcome expectancies. Health
Psychology ,4, 219/C0247.
Scheier, M. F., & Carver, C. S. (1992). Effects of optimism on psychological and physical well-being: Theoretical and empirical update.
Cognitive Therapy and Research ,16, 201/C0228.
Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing optimism from neuroticism (and trait anxiety, self-mastery, and self-
esteem): A reevaluation of the Life Orientation Test. Journal of Personality and Social Psychology ,67, 1063/C01078.
Schrank, B., Woppmann Mag, A., Sibitz, I., & Lauber, C. (2010). Development and validation of an integrative scale to assess hope. Health
Expectations ,14, 417/C0428.
Schweizer, K., & Koch, W. (2001). The assessment of components of optimism by POSO-E. Personality and Individual Differences ,31, 563/C0574.
Smith, T. W., Pope, M. K., Rhodewalt, F., & Poulton, J. L. (1989). Optimism, neuroticism, coping, and symptom reports: An alternative inter-
pretation of the Life Orientation Test. Journal of Personality and Social Psychology ,56, 640/C0648.
Snyder, C. R. (1989). Reality negotiation: From excuses to hope and beyond. Journal of Social and Clinical Psychology ,8, 130/C0157.
Snyder, C. R. (1994). The psychology of hope: You can get there from here . New York: Free Press.
Snyder, C. R. (1995). Conceptualizing, measuring, and nurturing hope. Journal of Counseling and Development ,73, 355/C0360.
Snyder, C. R. (Ed.), (2000). The handbook of hope: Theory, measures, and applications . San Diego, CA: Academic.
Snyder, C. R. (2002). Hope theory: Rainbows in the mind. Psychological Inquiry: An International Journal for the Advancement of Psychological
Theory ,13, 249/C0275.
Snyder, C. R., Harris, C., Anderson, J. R., Holleran, S. A., Irving, L. M., Sigmon, S. T., et al. (1991). The will and the ways: Development and
validation of an individual-differences measure of hope. Journal of Personality and Social Psychology ,60, 570/C0585.
Snyder, C. R., Hoza, B., Pelham, W. E., Rapoff, M., Ware, L., Danovsky, M., et al. (1997). The development and validation of the Children’s
Hope Scale. Journal of Pediatric Psychology ,22, 399/C0421.
Snyder, C. R., Rand, K. L., & Sigmon, D. R. (2002). Hope theory: A member of the positive psychology family. In C. R. Snyder, & S. Lopez
(Eds.), Handbook of positive psychology (pp. 257 /C0276). Oxford, UK: Oxford University Press.
Snyder, C. R., Sympson, S. C., Michael, S. T., & Cheavens, J. (2001). The optimism and hope constructs: Variants on a positive expectancy
theme. In E. Chang (Ed.), Optimism and pessimism (pp. 101 /C0125). Washington, DC: American Psychological Association.
Snyder, C. R., Sympson, S. C., Ybasco, F. C., Borders, T. F., Babyak, M. A., & Higgins, R. L. (1996). Development and validation of the State
Hope Scale. Journal of Personality and Social Psychology ,70, 321/C0335.
Staats, S. R., & Stassen, M. A. (1985). Hope: An affective cognition. Social Indicators Research ,17, 235/C0242.
Steed, L. G. (2002). A psychometric comparison of four measures of hope and optimism. Educational and Psychological Measurement ,62,
466/C0482.
Weis, R., & Speridakos, E.C. (2012). A meta-analysis of hope enhancement strategies in clinical and community settings. Psychology of Well-
Being: Theory, Research and Practice 2011, 1, www.psywb.com/content/1/1/5 .73 REFERENCES
II. EMOTIONAL DISPOSITIONS |
CHAPTER
4
Measures of Anger and Hostility in Adults
Ephrem Fernandez1, Andrew Day2and Gregory J. Boyle3
1University of Texas, San Antonio, TX, USA;2Deakin University, Geelong, Victoria, Australia;3University of
Melbourne, Parkville, Victoria, Australia
Anger lags behind anxiety and depression in terms of volume of published research. Yet, anger is widely
observed to be a problem in everyday life, its manifestation extending from the family to the workplace, and clin-
ical settings. This has spurred many scholars to develop tools for the assessment of anger. By far, most of these
tools take the form of self-report questionnaires. Some of these have been in use for more than half a century
while there are also signs of a proliferation of new instruments that coincide with the increased interest in anger
as a feature of everyday life and a target of self-regulation and intervention.
Before delving into the individual measures, a phenomenological sketch of anger and related phenomena
would be appropriate. First of all, anger has been defined in many ways with different aspects being emphasized.
However, there is general consensus that it is a feeling tied to appraised wrongdoing and coupled with action
tendencies to counter or redress the wrongdoing ( Smedslund, 1993; Wranik & Scherer, 2010 ). More broadly,
anger has been characterized in terms of patterns of psycho-physiological and facial activation. Although anger
may be deemed to have some beneficial effects, as in its role of mobilizing psychological resources, energizing
behavior, and protecting self-esteem ( Taylor & Novaco, 2005 ), it is typically regarded as a negatively valenced
emotion with potentially harmful consequences ( Fernandez, 2013 ). As Howells (2004) has suggested, ‘the argu-
ment that angry emotions, when poorly regulated, understood and expressed, make a major contribution to
human distress is a compelling one’ (p. 195).
Though sometimes used interchangeably with anger, the term hostility is more specifically reserved for fre-
quently recurring anger or anger proneness (Ramirez & Andreu, 2006); hostility is quite likely rooted in an attitu-
dinal bias or a cognitive schema of strong disapproval toward others ( Brodsky, 2011 ). By virtue of this
dispositional quality, it is akin to trait anger ( Smith, 1994 ). Aggression, which is outside the scope of this review
is defined in social psychology as behavior that is intended to harm, hurt, or damage /C0physically or psychologi-
cally (for a review of measures of aggression, please refer to Suris et al., 2004 ). Finally, violence is a subtype of
physical aggression in which the intended harm/hurt/damage actually materializes.
MEASURES REVIEWED HERE
All of the scales/measures presented in this chapter have anger as at least one of the focal constructs being
measured. In order to provide up-to-date psychometric reviews, the emphasis in this chapter is on research pub-
lished subsequent to the appearance of the first edition of this book (Robinson, Shaver, & Wrightsman, 1991).
Attention is also given to some recently constructed measures that, notwithstanding a smaller literature base, rep-
resent points of innovation in this evolving field of anger assessment (cf. Biaggio & Maiuro, 1985 ). These mea-
sures are presented in a more abbreviated format because of limited information concerning certain psychometric
criteria.
74Measures of Personality and Social Psychological Constructs.
DOI: http://dx.doi.org/10.1016/B978-0-12-386915-9.00004-8 ©2015 Elsevier Inc. and Ephrem Fernandez. All rights reserved. |
Measures Reviewed in Detail
1.Buss/C0Durkee Hostility Inventory ( Buss & Durkee, 1957 )
2.Buss/C0Perry Aggression Questionnaire ( Buss & Perry, 1992 )
3.Anger Self-Report Questionnaire ( Reynolds, Walkey, & Green, 1994; Zelin, Adler, & Meyerson, 1972 )
4.Reaction Inventory ( Cho, Kim, Kim, Wang, & Chee, 2009; Evans & Stangeland, 1971 )
5.Novaco Anger Scale and Provocation Inventory ( Novaco, 1994, 2003 )
6.Multidimensional Anger Inventory ( Siegel, 1985, 1986 )
7.State/C0Trait Anger Expression Inventory /C02nd Edn. ( Spielberger, 1988, 1991, 1999 )
Measures Reviewed Briefly
1.Anger Disorders Scale ( DiGiuseppe & Tafrate, 2004 )
2.Anger Parameters Scale ( Fernandez, Vargas, & Garza, 2010, Fernandez, Arevalo, Vargas, & Torralba, 2014 )
3.Awareness and Expression of Anger Indicator ( Catchlove & Braha, 1985 )
4.Standardized Experience of Anger Measure ( Linden et al., 1997 )
5.Anger Control Inventory ( Hoshmand & Austin, 1987 )
6.Anger Discomfort Scale ( Sharkin & Gelso, 1991 )
7.Anger Related Reactions and Goals Inventory ( Kubiak, Wiedig-Allison, Zgoriecki, & Weber, 2011 )
8.Anger Readiness to Change Questionnaire ( Williamson, Day, Howells, Bubner, & Jauncey, 2003 )
9.Short Anger Measure (Gerace & Day, 2014)
OVER VIEW OF THE MEASURES
Not surprisingly, the first psychological measures of anger or hostility emerged out of the venerable
Minnesota Multiphasic Personality Inventory which has spawned many other meas ures of affect. Basically, sub-
sets of the 550 items of the MMPI were brought togeth er to form supplementary scales. The three anger
assessment tools originating from the MMPI were the Cook-Medley Hostility Scale (Ho Scale; Cook &
Medley, 1954 ), the Overcontrolled Hostility Scale (O-H Scale; Megargee, Cook, & M endelsohn, 1967 ), and the
Hostility & Direction of Hostility Questionnaire (HDHQ; Caine, Foulds, & Hope, 1967 ). The following psycho-
metric reviews do not elaborate on these MMPI-derived measures of anger because they are really supple-
mentary scales of a much broader instrument for assessi ng psychopathology. Second, these first generation
anger measures seem to have receded into relative obscurity. Third, these measures were generally used to
obtain overall scores of hostility ra ther than different dimensions of ange r. Nevertheless, from a historical
point of view, it is helpful to point out that this was the starting point in the psychometric assessment of
anger/hostility.
In the next section, we dedicate our efforts to reviewi ng seven self-report measures of anger or hostility
for which there is relatively abundant psychometric d ata. Generally, these have a longer history than other
instruments currently in use. We begin with the Buss /C0Durkee Hostility Inventory (BDHI; Buss & Durkee,
1957 ) and its successor, the Buss /C0Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992 ). This is
followed by the Anger Self-Report Questionnaire (ASR; Reynolds et al., 1994 ), the Reaction Inventory
(RI; Evans & Stangeland, 1971 ) including a Korean version ( Cho et al., 2009 ), the Novaco Anger Scale and
Provocation Inventory (NAS-PI; Novaco, 1994, 2003 ), the Multidimensional Anger Inventory (MAI; Siegel,
1986 ), and finally, the State /C0trait Anger Expression Invent ory-2nd Edition (STAXI-2; Spielberger, 1988, 1999 ).
In the subsequent section we provide brief summarie s of nine measures, most of them newer and with less
psychometric data. These include the Anger Disorders Scale (ADS; DiGiuseppe & Tafrate, 2004 ), the Anger
Parameters Scale (APS; Fernandez et al., 2010, 2014 ), the Awareness and Expression of Anger Indicator
(AEAI; Catchlove & Braha, 1985 ), the Standardized Experience of Anger Measure (SEAM; Linden et al.,
1997 ), the Anger Control Inventory (ACI; Hoshmand & Austin, 1987 ), the Anger Discomfort Scale
(ADS; Sharkin & Gelso, 1991 ), the Anger Related Reactions and Goals Inventory (ARGI; Kubiak et al., 2011 ),
the Anger Readiness to Change Questionnaire (AECQ; Williamson et al., 2003 ), and finally, the Short Anger
Measure (SAM; Gerace & Day, 2014).75 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
MEASURES REVIEWED IN DETAIL
Buss/C0Durkee Hostility Inventory (BDHI)
(Buss & Durkee, 1957 ).
Variable
Historically, anger/hostility has been viewed as a multidimensional construct as reflected, for example, in the
eight dimensional BDHI measure.
Description
The BDHI is a 75-item self-report inventory (60 direct items; 15 reverse worded items) intended to measure
aspects of anger/hostility and aggression. The BDHI comprises ‘subscales labeled Assault (10 items), Indirect
[hostility] (9 items), Irritability (11 items), Negativism (5 items), Resentment (8 items), Suspicion (10 items), Guilt
(9 items), and Verbal [hostility] (13 items). So many factor-analyses of the inventory have been published that a
meta-analysis of the factor-analyses has appeared ( Bushman, Cooper, & Lemke, 1991 ), (Leenaars & Lester, 2011 ,
p. 66). Buss and Durkee (1992) reported that, ‘The correlation between social desirability and probability of
endorsing the items was .87’ (p. 345). Even after curtailment of the range of social desirability responding, the
correlation was still .74, suggesting that the early BDHI items were unduly influenced by social desirability
responding. However, after rewriting several of the items the correlation with social desirability decreased to
.27 (males) and .30 (females), respectively. Responses are on a 6-point rating scale. The BDHI has now been
translated into several different languages.
Sample
The initial factor analytic sample comprised 173 undergraduates (85 males; 88 females), item analyses were
based on 159 (85 male; 74 female) and 120 (62 male; 58 female) undergraduates, and social desirability
responding was investigated using 120 undergraduate ‘judges’ (85 male; 35 female) and 120 undergraduate
respondents (62 male; 58 female), respectively. Buss and Durkee (1957) provided norms for 173 undergraduates
(85 males; 88 females). More recently, a Dutch adaptation of the BDHI was based on a sample of 463 undergradu-
ates ( Lange et al., 1995 ), and a Spanish adaptation of the BDHI has also been constructed ( Oquendo et al., 2001 ).
Reliability
Internal Consistency
Items selected for inclusion in the 75-item BDHI satisfied ‘internal consistency criteria’. KR-20 coefficients were
reported for the two higher-order factors as follows: Covert anger/hostility (.76) and Overt aggression (.72) (see
Pittsburgh Mind Body Center (PMBC) website /C0Retrieved January 16, 2014). http://pmbcii.psy.cmu.edu/core_c/
Buss-Durkee_Hostility_Inventory.html .
Similarly, a Dutch adaptation (BDHI-D) exhibited Cronbach alpha coefficients for the two components they
had extracted of 0.77 (Overt Aggression) and 0.79 (Covert Aggression) /C0(Lange et al., 1995 ).
Test/C0Retest
Stability coefficients over a two-week test /C0retest interval were reported by Biaggio, Supplee, and Curtis (1981)
based on a sample of 60 undergraduate students as follows: Assault (.78), Indirect Hostility (.68), Irritability (.64),
Negativism (.64), Resentment (.66), Suspicion (.68), Verbal Hostility (.77), Guilt (.72), and BDHI Total (.82).
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, in a Dutch validation study ( Lange et al.,1995 ) positive correla-
tions ( N538) were found between the BDHI-D Covert Aggression component and the SCL-90 psychopathology
subscales ( Derogatis, 1977 ) as follows: Anxiety (.58), Agoraphobia (.45), Depression (.52), Somatization (.38),
Insufficiency (.56), Sensitivity (.55), Hostility (.47), Sleeplessness (.34) and Psychoneuroticism (.64), with the VIR
(Interpersonal Relations Questionnaire; Vertommen & Rochette, 1979 ) subscales Hostility (.24), and Bitterness
(.54), and with the MMPI subscale Negativism (.29). In addition, the Overt Aggression component correlated76 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
positively with VIR Hostility (.64), VIR Bitterness (.24), and MMPI Negativism (.54). As well, the total BDHI-D
scores correlated positively with all nine of the SCL-90 psychopathology subscales ranging from .13 to .43 (for
Depression), and with VIR Hostility (.60), VIR Bitterness (.51), and MMPI Negativism (.57) (see Lange et al.,
1995). Also, the BDHI has been found to correlate positively with measures of trait anxiety ( Matthews & Saal,
1978) and with other self-report measures of anger/hostility/aggression (ranging from .40 to .70) ( Matthews,
Jamison, & Cottington,1985 ).
Divergent/Discriminant
As evidence of divergent/discriminant validity, in the Dutch validation study, negative correlations ( N538)
were found between the BDHI-D Covert Aggression component and MMPI Extraversion ( /C0.12), the NPV
(Luteijn, Starren, & Van Dijk, 1975 ) Dominance ( /C0.21), and Crowne and Marlowe’s (1960) Social Desirability
Scale (/C0.37). In addition, the BDHI-D Overt Aggression component exhibited negligible correlations with each of
the nine SCL-90 psychopathology subscales (ranging from /C0.30 to .15), and also correlated negatively with Social
Desirability ( /C0.44). As well, negative correlations were found between BHDI-D total scores and MMPI
Extraversion ( /C0.07), NPV Dominance ( /C0.19), and Social Desirability ( /C0.55) (see Lange et al., 1995 ). Also, Biaggio
(1980) reported that the BDHI Total score correlated negatively ( /C0.56) with the Marlowe /C0Crowne Social
Desirability Scale. Biaggio et al. (1981) further reported mostly negative correlations (ranging from .09 to /C0.38)
between the BDHI Total and the Personal Incidents Record (a record of all incidents which provoked anger).
Construct/Factor Analytic
As evidence of construct/factor analytic validity, Buss and Durkee (1957) reported the results of an explor-
atory factor analysis using Thurstone’s (1947) centroid method on separate relatively small samples of male
(N585) and female ( N588) undergraduates, respectively. Two factors were extracted. For both the men and
women, the first factor loaded significantly on BDHI subscales: Resentment (.59 and .57) and Suspicion (.66 and
.54), respectively. Likewise, for both men and women, the second factor loaded significantly on BDHI subscales:
Assault (.54 and .61), Indirect Hostility (.40 and .48), Irritability (.57 and .47), and Verbal Hostility (.63 and .49),
respectively. For women, there was also a significant factor loading on the BDHI Negativism subscale. The two
factors represented emotional anger/hostility, and aggressive behavior, also described as Neurotic hostility and
Expressive hostility ( Bushman et al., 1991; Felsten & Leitten, 1993; Siegman, Dembroski, & Ringel, 1987 ) (see
PMBC website). As Biaggio (1980) had previously indicated, ‘Factor analysis produced two factors: an attitudinal
component of hostility (Resentment and Suspicion subscales) and a ‘motor’ (Expressive) component (Assault,
Indirect Hostility, Irritability, and Verbal Hostility subscales)’ (pp. 289 /C0290). A meta-analysis of the many BDHI
factor-analytic studies has been carried out and it is evident that the BDHI measures both affective and behav-
ioral components of anger/hostility and aggression, respectively ( Bushman et al., 1991 ).
Criterion/Predictive
As evidence of criterion/predictive validity, Lange, Dehghoni, and De Beurs (1995) reported that BDHI scores
are predictive of aggressive behavior. Also, according to the PMBC website (see link above): (1) BDHI Expressive
hostility scores appear to relate to central serotonergic depletion ( Cleare & Bond, 1997; Coccaro et al., 1989 );
(2) During interpersonal stress, Expressive hostility scores are predictive of cardiovascular reactivity ( Felsten &
Leitten, 1993; Siegman, Anderson, Herbst, Boyle, & Wilkinson, 1992; Suarez & Williams, 1990 ); (3) Expressive
hostility scores are predictive of coronary disease among patients younger than 60 years ( Siegman, 1994;
Siegman et al., 1987 ); and (4) Both BDHI Assault and Irritability scores are reduced in individuals following SSRI
ingestion (Knuston et al., 1998). In addition, Musante, MacDougall, Dembroski, and Costa (1989) reported that,
‘The BDHI has been related (a) to behavioral measures, including shock administration and role-playing
responses to anger-provoking circumstances; (b) to greater perception of violence in binocular rivalry technique;
and (c) to expert ratings of aggressiveness’ (p. 346).
Location
Buss, A. H., & Durkee, A. (1957) . An inventory for assessing different kinds of hostility. Journal of Consulting
Psychology , 21, 343-349.
Results and Comments
The BDHI was the first major multidimensional measure of anger/hostility/aggression. It has been in use for
over five decades, and has been used in thousands of research studies. The stability of the BDH over varying77 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
time intervals might be subjected to further investigation. Also, the BDHI subscale structure has not been fully
supported in factor analytic studies that have been based on more up-to-date factor analytic methodology
(cf.Bushman et al., 1991 ;Buss & Perry, 1992 ). This deficiency motivated construction of the BPAQ which was
based on more recently devised factor analytic methods, and is reviewed next.
BDHI SAMPLE ITEMS
Scale Sample item statement
ASSAULT I have known people who pushed me so far that we came to blows.
INDIRECT I sometimes spread gossip about people I don’t like.
IRRITABILITY I can’t help being a little rude to people I don’t like.
NEGATIVISM When someone is bossy, I do the opposite of what he asks.
RESENTMENT Other people always seem to get the breaks.
SUSPICION I tend to be on my guard with people who are somewhat more friendly than I expected.
VERBAL Even when anger is aroused, I don’t use ‘strong language.’*
GUILT The few times I have cheated, I have suffered unbearable feelings of remorse.
Notes : Items currently rated on a 6-point scale:
*Reverse worded item.
The BDHI can be administered using a two-choice (agree or disagree) and six-choice response format version
(15Strongly disagree to 6 Strongly agree ). A study by Velicer, Govia, Cherico, and Corriveau (1985) concluded that
the two-choice version provided some support for the present theoretical structure, but was not stable across
administrations, whereas the six-choice version resulted in a structure that was different, but more stable across
repeated administrations.
Buss/C0Perry Aggression Questionnaire (BPAQ)
(Buss & Perry, 1992 ).
Variable
Revision of the BDHI item pool has resulted in the emergence of a new measure ( Buss & Perry, 1992 ).
The BPAQ includes four components of Physical aggression, Verbal aggression, Anger, and Hostility (cf. Lange
et al., 1995 ).
Description
Buss and Perry (1992) added additional items to the BDHI item pool and following a factor analysis of the
item inter-correlations, the resultant BPAQ consisted of 29 items structured into four subscales as follows:
Physical Aggression (9 items), Verbal Aggression (5 items), Anger (7 items), and Hostility (8 items) (see Lange
et al., 1995 ). The BPAQ utilizes a 5-point Likert-type response format. As compared with the BDHI, the BPAQ
has improved structural properties as well as capturing all three cognitive, affective, and behavioral dimensions
of the hostility construct. The BPAQ has been used in studies conducted in several different countries including
the USA, Canada, Italy, the Netherlands, Germany, Japan, Greece, and Egypt (see Abd-El-Fattah, 2007).
Sample
The BPAQ was based on an initial sample of 1253 university students (612 males; 641 females) aged from
18/C020 years ( Buss & Perry, 1992 ). The total sample consisted of three separate subsamples of undergraduates
(Ns5406, 448, and 399, respectively). Bernstein and Gesn (1997) based their analyses of the BPAQ structure on a
sample of 321 undergraduates (113 males; 208 females). In an Italian version of the BPAQ, Fossati et al. (2003) uti-
lized samples of 563 high school students, and 392 university students. Diamond, Wang, and Buffington-Vollum
(2005) utilized a sample of 383 male prisoners in testing a variety of factor models for the BPAQ, in addition to a
cross-validation sample of 403 male prisoners ( Total N 5786; M534 years, range 19 /C068 years). Abd-El-Fattah78 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
(2007) utilized a sample of 510 (265 males; 245 females) secondary school students ( M516.3 years, range 16 /C028
years).
Reliability
Internal Consistency
Buss and Perry (1992) reported Cronbach alpha coefficients for the BPAQ subscales ( N51253) as follows:
Physical Aggression (.85), Verbal Aggression (.72), Anger (.83), Hostility (.77), and BPAQ Total (.89). Fossati et al.
(2003, p. 62) reported alpha coefficients ( N5563) as follows: Physical Aggression (.81), Verbal Aggression (.53),
Anger (.72), and Hostility (.68), and ( N5392) as follows: Physical Aggression (.85), Verbal Aggression (.53),
Anger (.72), and Hostility (.78). Abd-El-Fattah (2007) reported alpha coefficients for the four BPAQ subscales as
follows: Physical Aggression (.82), Verbal Aggression (.81), Anger (.83), and Hostility (.80). Evren, Cinar, Gulec,
Celik, and Evren (2011) reported alpha coefficients ( N5166) ranging from .59 to .93 for the Turkish version of
the BPAQ. Subsequently, Demirtas-Madran (2013) reported alpha coefficients for the Turkish version ( N5220) as
follows: Physical Aggression (.78), Verbal Aggression (.48), Anger (.76), Hostility (.71), and BPAQ Total (.85).
Test/C0Retest
Buss and Perry (1992) reported stability coefficients for the BPAQ subscales ( N51253) across a nine-week
test/C0retest interval as follows: Physical Aggression (.80), Verbal Aggression (.76), Anger (.72), Hostility (.72), and
BPAQ Total (.80). Diamond et al. (2005) reported alpha coefficients for the four subscales in the Bryant and Smith
(2001) short form that ranged from .63 to .73. Likewise, Evren et al. (2011) reported two-week test /C0retest reliabil-
ity coefficients ( N5166) ranging from .54 to .84 for the Turkish version of the BPAQ. Subsequently, Demirtas-
Madran (2013) reported stability coefficients over a four-week interval as follows: Physical Aggression (.98),
Verbal Aggression (.82), Anger (.85), Hostility (.85), and BPAQ Total (.97).
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, Buss and Perry (1992) reported positive correlations between
the BPAQ subscales and various trait measures as follows: BPAQ Physical Aggression correlated .20 (men) with
Activity, .28 with Impulsiveness, .28 with Assertiveness, and .36 with Competitiveness. BPAQ Verbal Aggression
correlated .31 with Impulsiveness, .49 with Assertiveness, and .39 with Competitiveness. BPAQ Anger correlated
.43 with Emotionality .22 (men) with Activity, .42 with Impulsiveness, .40 with Assertiveness, .32 with
Competitiveness, and .20 (women) with Private self-consciousness. BPAQ Hostility correlated .52 with
Emotionality, .37 with Impulsiveness, .30 with Competitiveness, .32 (men) and .49 (women) with Public
self-consciousness, and .24 with Private self-consciousness ( Buss & Perry, 1992 ). Total BPAQ scores correlated
.35 with Emotionality, .25 (men) with Activity, .46 with Impulsiveness, .43 with Assertiveness, .46 with
Competitiveness, .20 with Public self-consciousness, and .25 (women) with Private self-consciousness. Bernstein
and Gesn (1997) reported that the BPAQ subscales exhibited positive inter-correlations (ranging from .34 to 1.00),
indicative of their convergent/concurrent validity.
Divergent/Discriminant
As evidence of divergent/discriminative validity, Buss and Perry (1992) also reported divergent/discriminant
validity evidence as follows: BPAQ Physical Aggression did not correlate significantly with Emotionality,
Activity (women), Sociability, Public self-consciousness, Private self-consciousness, or Self-esteem. BPAQ Verbal
Aggression did not correlate significantly with Emotionality, Sociability, Public self-consciousness, or Self-esteem.
BPAQ Anger did not correlate significantly with Activity (women), Sociability, Private self-consciousness (men),
and was negatively correlated with Self-esteem ( /C0.14 men; /C0.27 women). BPAQ Hostility did not correlate signif-
icantly with Activity, and was negatively correlated with Sociability ( /C0.12), and Self-esteem ( /C0.49).
Construct/Factor Analytic
As evidence of construct/factor analytic validity, Buss and Perry (1992) performed separate exploratory
principal-axis factor analyses with oblimin rotation on the BDHI item intercorrelations using samples of 406, 448,
and 399 undergraduates, respectively. Four factors were extracted and labeled: Physical Aggression, Verbal
Aggression, Anger, and Hostility. A confirmatory factor analysis carried out on the second subsample of univer-
sity students ( N5448), provided additional support for the four-factor structure. Bernstein and Gesn (1997)79 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
found support for the four-factor BPAQ structure in a sample of undergraduates ( N5321). Bryant and Smith
(2001) also found support for the four factor structure using a modified 12-item short form of the BPAQ. Fossati
et al. (2003) carried out confirmatory factor analyses in Italian samples ( Ns5392 and 563), finding strong support
for the four factor structure. Diamond et al. (2005) in confirmatory factor analyses ( Ns5383 and 403) found sup-
port for the Bryant and Smith four factor model ( χ2
(48)5110.9, χ2/df52.3, GFI 5.96, RMSEA 5.06, CFI 5.95,
NNFI5.94) Abd-El-Fattah (2007) using a sample of 510 Egyptian high school students provided support for the
four factors. In a subsequent confirmatory factor analysis, Abd-El-Fattah reported that the BPAQ comprised ‘four
first level factors that were linked by a higher order factor of general aggression’ (p. 237). CFA results for this
four-factor best-fitting model ( N5510) were: χ2
(371)5385.6 (n.s.), RMSEA 50.01, SRMR 50.02, AGFI 50.99,
PGFI50.29, TLI 50.99, PRATIO 50.85, and PNFI 50.83. The four factor structure of the BPAQ has also received
support in several other factor analytic studies including: Harris (1995) ;Meesters, Muris, Bosma, Schouten, and
Beuving (1996) ;Williams, Boyd, Cascardi, and Poythress (1996) ;Nakano (2001) [Japanese version]; von Collani
and Werner (2005) [German version]; Diamond et al. (2005) ;Vigil-Colet, Lorenzo-Seva, Codorniu-Raga, and
Morales (2005) ; and Tsorbatzoudis (2006) [Greek version]; and Demirtas-Madran (2013) [Turkish version].
Criterion/Predictive
Predictive validity evidence has been summarized by the Pittsburgh Mind Body Center (PMBC): http://pmbcii.
psy.cmu.edu/core_c/Buss_Perry_Aggression_Questionnaire.html (Retrieved January, 16, 2014). For example,
Smith and Gallo (1999) reported that (among men), the BPAQ Hostility subscale predicted increased blood pres-
sure response to interpersonal threat. PMBC also reported that, following provocation, BPAQ scores exhibit a
lowered threshold for anger, and a reduced threshold for aggression (subsequent to tryptophan depletion;
Dougherty, Bjork, Marsh, & Moeller, 1999 ) or alcohol consumption ( Giancola, 2002 ). BPAQ scores are predictive
of inflammatory processes ( Suarez, Lewis, & Kuhn, 2002 ) and C-reactive protein ( Suarez, 2004 ). Also, BPAQ
scores are predictive of the severity of coronary disease for men under 60 years ( Gidron, Davidson, & Ilia, 2001 ).
Location
Buss, A.H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality and Social Psychology ,
63, 452–459.
Results and Comments
Some 35 years after constructing the BDHI, Buss ( Buss & Perry, 1992 ) constructed a new instrument (BPAQ).
Several exploratory and confirmatory factor analytic studies using large samples in a variety of different coun-
tries have confirmed that a general anger/hostility/aggression factor can be broken down into four major sub-
factors corresponding to the BPAQ subscales labeled: Physical aggression, Verbal aggression, Anger, and
Hostility, respectively. The first two factors measure the behavioral and cognitive components while the remain-
ing two factors (anger and hostility) measure the affective component. As Fossati et al. (2003) concluded, there is
a ‘need to measure not only overall aggression but also its components’ (p. 64). There is considerable evidence
showing that the BPAQ is predictive of real-life health outcomes.
BPAQ SAMPLE ITEMS
Please rate each of the following items in terms of how characteristic they are of you. Use the following scale for
answering these items.
12 3 4 5
Extremely uncharacteristic of me Extremely characteristic of me
Scale Sample item statement
Physical aggression Given enough provocation, I may hit another person.
Verbal aggression I can’t help getting into arguments when people disagree with me.
Anger Sometimes I fly off the handle for no good reason.
Hostility I am suspicious of overly friendly strangers.
Note: Reverse scored items (items 7, 4).80 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Anger Self-Report Questionnaire (ASR)
(Reynolds et al., 1994; Zelin et al., 1972 ).
Variable
The distinction between awareness of one’s angry/hostile feelings and the expression of aggression in one’s
behavior is an important one. Accordingly, the ASR focuses on measuring anger/hostility both in terms of one’s
awareness of anger and its behavioral expression ( Zelin et al., 1972 ). As Musante et al. (1989) pointed out, the
ASR was ‘explicitly designed to remove the confounding that previous anger/hostility inventories have made
between awareness of angry feelings and expression of hostility/anger in behavior’ (p. 346).
Description
In an item analysis of the original ASR ( N5138), Zelin et al. (1972) reduced the ASR item-pool from 89 items
down to 64 items (with removal of 25 filler items). The resultant subscales were labeled as follows: Awareness
of anger, Expression of anger (comprising General, Physical, and Verbal expression subscales), Guilt,
Condemnation of Anger, and Mistrust ( Zelin et al., 1972 ; cf.Matthews et al., 1985 ;Musante et al., 1989 ;Reynolds
et al., 1994 ). A 30-item short form is also available ( Reynolds et al., 1994 ).
Sample
The initial sample ( Zelin et al., 1972 ) comprised 138 individuals used for the item analysis of the 89-item ver-
sion of the ASR, followed by samples of 82 psychiatric patients and 67 undergraduates used for validating the
psychometric properties of the abbreviated 64-item version. Reynolds et al. (1994) administered both the 89-item
and 64-item versions to a sample of 246 undergraduate students (127 males; 119 females) ranging in age from
16 to 47 years. Norms for 30-item short form were derived from the responses of 101 male and 100 female under-
graduates (Reynolds et al., p. 64).
Reliability
Internal Consistency
Split-half coefficients for the ASR subscales ranged from .64 to .82 ( Zelin et al., 1972 ). More recently, Reynolds
et al. (1994) reported split-half coefficients for the Physical expression (.64) and General expression subscales
(.66), respectively), and that 39 out of the total 89 items correlated positively with total scores, ranging from .28 to
.66 (Reynolds et al., p. 66). While the KR 20coefficient was .85 for all 89 items ( N5246), those for the ASR sub-
scales were somewhat lower, as follows: General expression (.48), Mistrust (.57), Verbal expression of anger (.70),
and Awareness of anger (.79).
Test/C0Retest
Biaggio et al. (1981) reported two-week test /C0retest stability coefficients for the ASR subscales as follows:
Awareness (.54), General (.45), Physical (.63), Verbal (.35), Guilt (.28), Condemnation (.76), Mistrust (.53), and
ASR Total (.54).
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, Biaggio (1980) reported that BDHI Total Hostility and ASR
Total Expression scores correlated positively (.64). Biaggio (p. 295) further pointed to the positive correlations
between the BDHI and ASR subscales (ranging from .28 to .78). Subsequently, Biaggio (1981, Table 1) reported
positive correlations between the ASR Total Expression scale and self-monitored incidents of Verbal expression
of anger (.38), and Physical expression of anger (.31). Likewise, Schill, Ramanaiah, and Conn (1990) reported
(N565) that ASR and BDHI scores correlated .60 (males) and .66 (females). Also, Reynolds et al. (1994, Table 1)
reported that two-thirds of the ASR subscale inter-correlations were significant ( p,.001).
Divergent/Discriminant
As evidence of divergent/discriminant validity, Zelin et al. (1972) , reported that, ‘A multi-trait, multi-method
analysis ...yielded substantial convergent and discriminant validities for the ASR scales ...Furthermore, aware-
ness lacked significant correlations with PAS scales reflecting the expression of anger, thereby demonstrating its81 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
discriminant validity’ (p. 340). Biaggio (1980, p. 296) reported that only the ASR subscales (Awareness of anger,
General Expression, Guilt, and Mistrust) correlated with BDHI Guilt and that only the BDHI subscales
(Irritability, Resentment, Suspicion, and Guilt) correlated significantly with ASR Guilt, indicative of the latter’s
discriminant validity. As well, Biaggio (1980) reported that the ASR Total Expression score correlated negatively
(/C0.40) with the Marlowe /C0Crowne Social Desirability Scale. Also, Biaggio et al. (1981) reported negative or negli-
gible correlations (ranging from .07 to /C0.31) between the ASR and three of the Personal Incidents Record
(Condemnation of Anger, Guilt, Indirect Expression) measures.
Construct/Factor Analytic
As evidence of construct/factor analytic validity, Reynolds et al. (1994) reported that, ‘Factor analyses were
conducted on the 89 item questionnaire to compare the seven factor and five factor solutions using analyses
obtained from three samples of 82 subjects’ (p. 66). Neither the 5-factor nor the 7-factor structure could be repli-
cated. Reynolds et al. subsequently undertook a factor analysis based on the item inter-correlations of the
64-item version (in a previous study, Biaggio, 1980 , had been unable to replicate the claimed ASR factors).
Again, Reynolds et al. could not find support for the purported subscale structure, leading them to conclude that
the ASR is unidimensional, comprising only a general anger factor.
Criterion/Predictive
As evidence of criterion/predictive validity, Zelin et al. (1972) reported that, ‘ASR scores were correlated with
psychiatrists’ ratings on the 16 most relevant Problem Appraisal Scales (PAS) ...the highest correlation (.41) for
the Physical expression scale is with ratings of assaultive acts on the PAS’ (p. 340). Zelin et al. also reported
several additional predictive validity correlations. For example, verbal expression of anger correlated /C0.36 with a
measure of dependency, .31 with real-life anger, belligerence and negativism, and .28 with antisocial attitudes
and acts. Awareness of anger correlated .24 with antisocial attitudes and acts, and /C0.37 with rating of obsessive-
compulsive behaviors. ASR Guilt correlated .48 with suicidal thoughts, and .33 with a measure of depression-
inferiority. ASR Mistrust correlated .33 with a measure of mistrust and suspicion.
Location
Zelin, M.T., Adler, G., & Meyerson, P.G. (1972). Anger self-report: An objective questionnaire for the measure-
ment of aggression. Journal of Consulting and Clinical Psychology, 39 , 340.
Reynolds, N.S., Walkey, F.H., & Green, D.E. (1994). The Anger Self-Report: A psychometrically sound
(30 item) version. New Zealand Journal of Psychology, 23 ,6 4/C070.
Results and Comments
The ASR could benefit by further exploration of its stability over varying time intervals. Also, Reynolds et al.
(1994) concluded that previous studies have provided ‘only modest evidence for discriminant, convergent and pre-
dictive validity of the ASR ...[and]...Factor analyses [have] provided little support for ...construct validity’ (p. 68).
In light of these inconclusive factor analytic findings, Reynolds et al. constructed a 30-item short form of the ASR
as a measure of a general anger factor. This unidimensional measure of anger is a departure from the clinical quest
to differentiate subtypes of anger and hostility already well recognized in phenomenological accounts. It is not
surprising therefore, that the ASR and its short forms have been eclipsed by alternative measures.
ASR SAMPLE ITEMS
‘I will criticize someone to their face if they deserve it.’
‘At times, I feel like smashing things.’
‘When I really lose my temper, I am capable of slapping someone.’
Notes : Items are rated on a 6-point scale ranging from: 1 5Strongly agree to 6 5Strongly disagree.
Reaction Inventory (RI)
(Cho et al., 2009; Evans & Stangeland, 1971 ).82 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Variable
Anger may arise in response to a variety of everyday stimulus situations often involving frustration of one’s
efforts. It is possible to vicariously arouse anger through invoking hypothetical imaginary scenarios. Imagining
such scenarios may ‘evoke anger, and thereby offer clues about an individual’s propensity for anger arousal.
However, hypothetical scenarios of the kind featured in the RI often run the risk of being over-specific and thus
of limited generalizability across test-takers ( Fernandez, 2008 , p. 406).
Description
The original RI was devised by Evans and Stangeland (1971) . The RI presents 76 intuitively selected items that
may incite anger, tapping into 10 hypothetical scenarios derived factor analytically (see below) including: Minor
chance annoyances, Destructive people, Unnecessary delays, Inconsiderate people, Self-opinionated people,
Frustration in business, Criticism, Major chance annoyances, People being personal, and Authority. Individuals
respond on a 5-point Likert-type rating scale indicating the degree to which they would be angered by each
scenario. In addition, a Korean adaptation of the RI has been produced ( Cho et al., 2009 ). Although the RI was con-
structed to measure ‘the number and type of incidents that arouse anger’ ( Leenaars & Lester, 2011 , p. 65), it remains
‘an instrument that has been on the periphery of anger assessment methodology’ ( Fernandez, 2008 , p. 406).
Sample
Evans and Stangeland (1971) utilized a heterogeneous sample of 275 university and non-university students
(84 males; 191 females) ranging in age from 16 to 75 years (Mdn 521 years). The total sample consisted of four
separate subsamples: Sample 1 (16 males; 29 females; Mdn 525 years), Sample 2 (10 males; 21 females; Mdn 522
years), Sample 3 (30 males; 108 females; Mdn 518 years), and Sample 4 (friends and family members who had
not attended university: 28 males; 33 females; Mdn 526 years). Subsequently, a Korean adaptation of the RI
(Cho et al., 2009 ) utilized a sample of 216 elderly depressed patients (170 women, 46 men; M571.6 years,
SD56.3) and a separate sample of 198 normal elderly individuals (128 women, 70 men; M572.6 years, SD55.9)
(Baeg, Wang, Chee, Kim, & Kim, 2011 ).
Reliability
Internal Consistency
Taking the mean item-test correlation (.46) as the starting point, and using Gaylord’s formula, the estimated
internal consistency coefficient was found to be .95 ( Evans & Stangeland, 1971 ). However, such a high level
of ‘internal consistency’ may be problematic in terms of possible item redundancy and inadequate breadth of
measurement of a scale (cf. Boyle, 1991 ).
Test/C0Retest
Biaggio et al. (1981, Table 1) using a sample of 60 undergraduate students reported a two-week test /C0retest
reliability coefficient of .70.
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, the RI total score was found to correlate positively (.52) with
the BDHI in a subsample of 45 undergraduates (16 males; 29 females), and .57 in a separate sample of 138 under-
graduates (30 males; 108 females; Mdn 518 years ( Evans & Stangeland, 1971 ). Also, Biaggio (1980) reported that
in a sample of undergraduates ( N5150), the RI correlated positively and highly (.82) with the conceptually simi-
lar NAS, showing substantial overlap in measurement variance and evidence of convergent/concurrent validity.
Divergent/Discriminant
As evidence of divergent/discriminant validity, Biaggio (1980, p.296) reported that the RI correlated negatively
(/C0.29) with the Marlowe /C0Crowne Social Desirability Scale. As further evidence of discriminative validity,
Biaggio also reported the results of a factor analysis based on the subscale inter-correlations for the BDHI, ASR,
RI, and AI measures combined ( N5150) wherein she obtained four factors. Indicative of discriminative validity,
the NAS was loaded by a separate factor (together with the RI), and did not overlap with the BDHI and ASR
measures which were loaded by other factors. Subsequently, Biaggio et al. (1981) reported negligible correlations
(ranging from .05 to /C0.22) between the RI and the Personal Incidents Record .83 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
Construct/Factor Analytic
As evidence of construct/factor analytic validity, a principal axis factor analysis ( N5275) with varimax rota-
tion of the inter-correlations of the 76 items with factor extraction number based on the eigenvalues greater than
1.0 criterion produced 10 factors, as listed above ( Evans & Stangeland, 1971 ). In the subsequent development
of a Korean version of the RI ( Cho et al., 2009 ), four higher-order factors were extracted and labeled:
Unpredictable disruptions and Disturbances, Embarrassing circumstances, Personal disrespect, and Unpleasant
encounters, respectively. Clearly, there is a multitude of situations that have the potential to provoke anger, as
recognized in the multidimensional structuring of the RI.
Criterion/Predictive
In a Korean study ( Baeg et al., 2011 ), depressive symptomatology among the elderly was positively predicted
by the RI anger factors with the sole exception of the factor labeled Unpleasant Encounters. Thus, ‘the more severe
the depressive symptoms ...the more severe the anger reaction to unpredictable disruption and disturbances,
embarrassing situations, and personal disrespected factor scores on the RI were ( p,.05)’ (Baeg et al., p. 189).
Location
Evans, D.R., & Stangeland, M. (1971). Development of the Reaction Inventory to measure anger. Psychological
Reports, 29 , 412/C0414.
Cho Y.W., Kim J.L., Kim S.Y., Wang S.K., & Chee I.S. (2009). A standardization of the Korean version of the
Reaction Inventory. Journal of the Korean Society for Biological Therapies in Psychiatry, 15 , 130/C0139.
Results and Comments
The stability of the RI over various periods of time needs to be investigated further. Also, although mere imagina-
tion of the provocative scenarios included in the RI has the potential to evoke anger, that would depend on the
degree of absorption into the imagined situations. Moreover, hypothetical provocation scenarios of the kind
included in measures such as the RI or NAS-PI, run the risk of being over-specific and may provide an anger trigger
for only a small fraction of the population. If so, this would limit the relevance of the items across respondents.
RI SAMPLE ITEMS
1.The telephone or doorbell ringing when you are busy
at something.
2.Phoney people.
3.Running out of gas.4.Having to take orders.
Note. Items are rated on a 5-point Likert-type scale rang-
ing from: 1 5Not at all to 55Very much .
Novaco Anger Scale and Provocation Inventory (NAS-PI)
(Novaco, 1994, 2003 ).
Variable
Anger disposition has been conceptualized as consisting of distinct cognitive, arousal, and behavioral compo-
nents ( Novaco, 1975 ). On the other hand, certain anger-eliciting situations can lead to the arousal of anger in
response to such provocation ( Novaco, 1994 ).
Description
The Novaco Anger Scale (NAS) was constructed by Novaco (1975) as a measure of anger reactions to various
provocations. It was revised by Novaco (1994) having 48 items relating to cognitive, behavioral, and arousal com-
ponents of anger (with the cognitive component comprising subscales of attentional focus, rumination, hostile
attitude, and suspicion). The PI comprises 25 hypothetical situations that are likely to provoke anger, distributed
across five subscales labeled: Disrespectful treatment, Unfairness/injustice, Frustration/interruptions, Annoying
traits, and Irritations. Further revisions of the NAS-PI have appeared, the most recent being the commercially
produced version ( Novaco, 2003 ).84 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Sample
An early sample comprised 353 undergraduate students ( Novaco, 1977 ). Subsequently, Novaco (1994) uti-
lized 142 hospital inpatients and another sample of 158 pat ients in undertaking initial validational studies.
Biaggio (1980) utilized a sample of 150 undergraduate students ( 72 male; 78 female) in her validation study of
the original NAS. Mills, Kroner, and Forth (1998) compared samples of 102 non-violent offenders ( M533 years,
range 19 /C069 years) with 102 violent offenders ( M528 years, range 18-55). Jones, Thomas-Peter, and Trout
(1999) investigated the NAS-PI in a sample of 58 men ( M532.47 years, SD 58.32), Monahan et al. (2001) uti-
lized a sample of 1102 mentally disturbed inpatients. Jones, Thomas-Peter, and Gangstad (2003) used a sample
of 354 anger management outpatients, Lindqvist, Waterman, and Hellstro ¨m (2003) employed a sample of 100
male undergraduates ( M533.2 years, SD512.5), Lindqvist, Waterman, and Hellstro ¨m (2005) utilized a sample
of 95 violent prisoners (ranging from 18 /C067 years), Baker, Van Hasselt, and Sellars (2008) utilized samples of
prisoners (638 males; 349 females). Hornsveld, Muris, and Kraaimaat (2011) utilized samples of 142 male foren-
sic psychiatric inpatients ( M533.16 years, range 21 /C056 years), 194 male outpatients ( M522.79 years, range
16/C056 years), and 320 secondary vocational students (160 males; M517.35 years, range 16 /C021; 160 females;
M518.36 years, range 16 /C027 years). Finally, the commercially prod uced version of the NAS-PI was validated
on an age-stratified national standardization s ample of 1546 non-clinical individuals aged from 9 /C084 years
(Novaco, 2003 ).
Reliability
Internal Consistency
Novaco (1977) reported an alpha coefficient ( N5353) for the original NAS of .96 (cf. Biaggio, 1980 ;Boyle,
1991).Novaco (1994) reported alpha coefficients ( N5126) for the NAS (.95), and PI (.95). Mills et al. (1998)
reported alpha coefficients NAS (.95) and PI (.96) in non-violent sample (N 5102), and NAS (.94) and PI (.95) in
violent sample ( N5102). Jones et al., (1999) reported alpha coefficients for NAS (.92) and PI (.92). Novaco and
Taylor (2004) reported an alpha coefficient of 0.92 for both the NAS and PI modified versions. Lindqvist et al.
(2005) reported alpha coefficients ranging from .78 to .91. Baker et al. (2008) reported alpha coefficients for men
on the NAS (.93) and PI (.92), and women on the NAS (.89) and PI (.87), respectively. Hornsveld et al. (2011)
reported alpha coefficients for the NAS subscales ranging from .77 to .95, and the PI for inpatients (.90) and out-
patients (.94), respectively.
Test/C0Retest
Novaco (1994) reported stability coefficients over an interval of two weeks for the NAS (.84) and PI (.86). Mills
et al. (1998) reported stability coefficients over a four-week interval for NAS (.89) and PI (.85). Cornell, Peterson,
and Richards (1999) reported alpha coefficients for the NAS (.94) and PI (.91). Subsequently, Hornsveld et al.
(2011) using a group of 90 forensic psychiatric outpatients, reported stability coefficients for the NAS over a four-
week interval as follows: NAS Anger (.80), NAS Cognitive (.71), NAS Arousal (.78), and NAS Behavior (.79),
respectively. They also reported a high alpha coefficient (.90) for the PI (cf. Boyle, 1991 ).
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, Biaggio (1980) had reported positive correlations ( N5150)
between the NAS and all eight BDHI subscales plus BDHI Total score (ranging from .05 to .45), with all seven
ASR subscales plus ASR Total score (ranging from .07 to .64), and with the RI (.82). Novaco (1994) reported posi-
tive correlations with both the Spielberger (1980) State/C0trait Anger Scale or STAS (.84), and the BDHI (.84),
respectively. Mills et al. (1998) reported that the NAS and PI correlated positively with the BPAQ (.79 and .68,
respectively). Cornell et al. (1999) reported that the NAS correlate .63 with the PI, and that both the NAS and PI
correlated positively with the Spielbeger STAXI scales (ranging from .26 to .90), except for the STAXI Anger
Control scale. Lindqvist et al. (2005) reported positive correlations between the NAS Total and Swedish adapta-
tions of the BPAQ (.86) and the STAXI-2 (.79). Baker et al. (2008) reported a positive correlation of .69 between
both components of the NAS-PI and STAS, and BDHI, respectively.
Divergent/Discriminant
As evidence of divergent/discriminant validity, Biaggio (1980, p. 296) reported that the NAS correlated
negatively ( /C0.26) with the Marlowe-Crowne Social Desirability Scale. Subsequently, Biaggio et al. (1981) also85 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
reported that the NAS exhibited negli gible correlations (ranging from /C0.15 to .02) with the Personal
Incidents Record. Novaco (1994) reported that NAS-PI scores were pred ictive of STAS and staff ratings of
anger. Cornell et al. (1999) reported negative corre lations between the STAXI Anger Control and NAS
(/C0.46), and PI ( /C0.21), respectively. Hornsveld et al. (2011) reported negligible or negative correlations
between the NAS-PI and the Psychopathy Checklist-Rev ised (PCL-R). As further evidence of discriminant
validity, Hornsveld et al. (p. 941) noted the lack of significant correlations between the NAS-PI and PCL-R
t o t a ls c o r e ,a sw e l la san e g a t i v ec o r r e l a t i o n( /C0.29) between the NAS total score and PCL-R Interpersonal
scale, and ( /C0.25) between the PI total score and PCL-R Affect ive scale. The NAS-PI was found to discrimi-
nate between outpatients referred for anger management versus normal healthcare employees, as well
as between forensic psychiatric patients versus normal secondary school vocational students ( Hornsveld
et al., 2011 ).
Construct/Factor Analytic
As evidence of construct/factor analytic validity, Jones et al. (2003) , using the NAS-PI (1994 version), carried
out an exploratory factor analysis ( N5566) that produced three factors, but these did not line up with those
Novaco (1994) had suggested previously. Monahan et al. (2001) undertook an exploratory factor analysis
(N51101) that failed to produce the expected factor structure. Novaco (2003) reported the results of an explor-
atory factor analysis based on the national standardization sample ( N51546) which also failed to find the
expected factor structure. Lindqvist et al. (2003) (NAS-PI, 1998 version; N5100) also failed to find the factor
structure claimed by Novaco (1994) .Hornsveld et al. (2011) conducted a confirmatory factor analysis on the
combined patient sample ( N5336) but could not find support for Novaco’s (1994) proposed three subscale
structure ( χ2/df52.38, GFI 5.73, CFI 5.78, RMSEA 5.06). Hornsveld et al. also carried out a CFA on the stu-
dent sample ( N5320) but again failed to confirm the three subscales ( χ2/df51.93, GFI 5.76, CFI 5.73,
RMSEA 5.05). Thus, the putative tripartite structure of the NAS (cognitive, behavioral, and arousal factors)
could not be confirmed.
Criterion/Predictive
As evidence of criterion/predictive validity, Novaco (1994) reported that the NAS significantly predicts mea-
sures of aggressive behavior from hospital records and staff ratings as well as predicting the number of convic-
tions for violent crimes ( Cornell et al., 1999 ). Likewise, Doyle and Dolan (2006) reported that among 112 violent
offenders ( M540 years, SD511.5) whom had been released into the general community, measures of risk
including the NAS significantly predicted future violent episodes.
Location
Novaco, R.W. (2003). The Novaco Anger Scale and Provocation Inventory: Manual . Los Angeles, CA: Western
Psychological Services.
www.wpspublish.com/store/p/2878/novaco-anger-scale-and-provocation-inventory-nas-pi#sthash.ym5GnyME.
dpuf (Retrieved January 17, 2014).
Results and Comments
The NAS-PI appears to provide a useful measure of affective, cognitive, and behavioral components of
anger/hostility/aggression. Nonetheless, most f actor analytic studies have failed to find support for
Novaco’s (1994) proposed three subscale structure. Even Novaco’s (2003) own factor analysis using the
large national stand ardization sample ( N51646) failed to confirm the purported three-dimensional struc-
ture. We can only conclude that, despite the great po pularity of the NAS-PI, its construct validity remains
in doubt.
NAS-PI SAMPLE ITEMS
Scale Sample items
Part A ‘Every week I meet someone I dislike’
‘When I get angry, I get really angry’
‘I can walk away from an argument’86 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Part B ‘Being criticized in front of other people for something that you have done’
‘You are overcharged by someone for a repair’
‘You are carrying a hot drink, and someone bumps into you’
Instructions : The statements in Part A are rated in terms of whether or not they apply to the subject: (1) never true;
(2) sometimes true; or (3) always true. The statements in Part B are rated in terms of how angry they would make
the subject feel, using a 4-point scale from ‘Not at all Angry’ to ‘Very angry’.
Multidimensional Anger Inventory (MAI)
(Siegel, 1985, 1986 ).
Variable
The multidimensionality of anger/hostility had been operationalized in some of the instruments discussed
above (e.g., the BDHI, BPAQ, and NAS-PI). However, a more comprehensive representation of the multidimen-
sionality of the anger/hostility construct has been proposed ( Siegel, 1986 ).
Description
The MAI is a 38-item rationally constructed multidimensional measure with several items assembled, and
adapted from previously published measures such as the BDHI. The MAI is purported to measure multiple
dimensions of anger, including Frequency, Duration, and Magnitude of anger responses, Mode of expression
(Anger-in and Anger-out), Hostile outlook, Range of anger-eliciting situations, Guilt, Brood, and Anger-discuss.
As with the NAS, RI, and ASR, a range of provocative anger-eliciting situations is incorporated into the MAI.
Responses are on a 5-point Likert-type rating scale.
Sample
Initial samples included 198 college students (74 males; 124 females) and 288 male factory workers ( M554.8
years, range 40 /C063 years). Also, a separate sample of factory workers ( N5288) was utilized in Siegel’s (1986)
study. Musante et al. (1989) utilized a sample of 82 male college students ( M/C2520 years), and 50 male faculty,
staff and senior students ( M/C2540 years). Kroner, Reddon, and Serin (1992) investigated the structure of the MAI
using a sample of 372 violent male prisoners.
Reliability
Internal Consistency
Siegel (1985) reported Cronbach alpha coefficients for the five MAI components ranging from .51 to .85
(cf.Musante et al., 1989 ). Subsequently, Siegel (1986) reported an alpha coefficient of .84 for the MAI Total score
in a college student sample ( N5198), and .89 in a sample of factory workers ( N5288). Aside from the Anger-
out scale, for the combined sample ( N5486), all scales exhibited alpha coefficients ranging from .70 to .88
(Siegel, 1986 ). For the sample of college students alpha coefficients ranged from .63 to .84, while for the sample of
factory workers, aside from the Anger-out scale, alphas ranged from .71 to .89.
Test/C0Retest
Siegel (1986) reported a stability coefficient of .75 over a 3 /C04 week test /C0retest interval based on a subsample
of 60 college students.
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, Siegel (1986) reported that the five MAI components
(see below) correlated positively with the Harburg inventory, as well as with the BDHI and NAS measures.
For example, positive correlations with the BDHI Hostility subscale were as follows: MAI Anger-arousal (.49),
MAI Range of anger-eliciting situations (.39), MAI Hostile outlook (.34), and MAI Anger-in (.62). Siegel (p. 198)87 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
also reported that the Spielberger STAI A-Trait scale correlated positively with MAI Anger-arousal (.54), with
MAI Anger-in (.54), with Anger-eliciting situations (.29), and with Hostile outlook (.26), respectively. Also, the
MAI Range of anger-eliciting situations correlated positively (.59) with the conceptually similar NAS items.
Divergent/Discriminant
As evidence of divergent/discriminant validity, Siegel (1986) reported negligible or negative correlations
between the MAI Anger-out component, and other measures of anger, as follows: Harburg Duration (.06),
Harburg Magnitude (.09), Novaco Magnitude (.06), Novaco Situations (.07), BDHI Hostility ( /C0.03), Harburg
Anger-in ( /C0.10), and Harburg Anger-out (.20), respectively. In addition, Riley and Treiber (1989) , reported that
the MAI Anger-out/Brooding scale which ‘assesses overt expression of anger combined with feelings of guilt
and brooding related to the response’ (p. 400) correlated negatively ( /C0.46) with the Framingham Anger-in mea-
sures of anger suppression.
Construct/Factor Analytic
As evidence of construct/factor analytic validity, Siegel (1986) carried out a principal components analysis
with varimax rotation on the MAI item intercorrelations in two samples ( Ns5198 and 288). For the college sam-
ple, three components were extracted and labeled: Anger-arousal (64% of the variance), Range of anger-eliciting
situations (24% of variance), and Hostile outlook (12% of variance). For the combined sample ( N5486), no fewer
than five separate components emerged and were labeled as follows: Anger-arousal (comprising: frequency, mag-
nitude, and duration of responses), Range of anger-eliciting situations, Hostile outlook, Anger-in, and Anger-out,
respectively. In a study of four well-established anger/hostility scales (BDHI, ASR, NAS, and MAI), Musante
et al. (1989) carried out a second-order principal components analysis with varimax rotation ( N5132) of the
21321 subscale intercorrelations. Although five components exhibited eigenvalues greater than 1.0, on the basis
of the Scree test ( Cattell, 1978; Cattell & Vogelmann, 1977 ), Musante et al. chose to extract only three higher-order
components that were labeled: Experience of Anger (‘representing anger-arousing and -eliciting situations and
anger awareness’); Expression of Anger (‘either physical assault or verbal expression of anger’) and Suspicion-
Guilt (‘suspicion, mistrust-suspicion, and guilt’), respectively. Riley and Treiber (1989) found that they were
unable to replicate the multi-factorial structure put forward by Siegel (1986) . Likewise, Kroner et al. (1992) also
investigated the factor structure of the MAI in a sample of 372 violent male prisoners, but claimed support for
only a two-factor structure, despite the previously demonstrated and cross-validated multidimensionality of the
MAI measure. Kneip et al. (1993) using a sample of coronary heart disease patients, reported a two-factor struc-
ture for the MAI. Using a large Finnish sample, Miller, Jenkins, Kaplan, and Salonen (1995) reported the results
of a CFA that supported the MAI rationally devised dimensions rather than the claimed multifactorial structure.
However, Dutton (1995) found support for the following MAI subscales: Anger-in, Anger-out, Magnitude of
anger, and Frequency of anger.
Criterion/Predictive
As evidence of criterion/predictive validity, the MAI Hostile Outlook scale correlated positively (.46) with the
sum of the BDHI Negativism, Resentment, and Suspicion scales ( Siegel, 1985 ).Musante et al. (1989) reported that
scores on the Structured Interview (SI) Potential for Hostility were significantly predicted by MAI Anger-Arousal
(.34), MAI Hostile outlook (.25), and MAI Range of eliciting conditions (.33), respectively (the predictive correla-
tions for MAI Anger-in and Anger-out were quite low, being .14 and 18, respectively).
Location
Siegel, J.M. (1985). The measurement of anger as a multidimensional construct. In M.A. Chesney & R.H.
Rosenman (Eds.), Anger and hostility in cardiovascular and behavioural disorders (pp. 59 /C081). New York:
Hemisphere/McGraw-Hill.
Siegel, J.M. (1986). The multidimensional anger inventory. Journal of Personality and Social Psychology, 51 ,
191/C0200.
Results and Comments
Siegel (1986) obtained virtually identical factor structures across two different samples, thereby providing
cross-validational evidence of the multidimensional structure of the MAI measure. Subsequently, Musante et al.
(1989) concluded that these anger measures reflect emotional, interpersonal, and attitudinal components, which
can be defined in terms of an experiential anger dimension, an anger expression dimension, and a hostile or88 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
suspicious-mistrustful attitudinal outlook dimension. Siegel (1986) pointed to these findings as evidence for ‘con-
ceptualizing anger as a multidimensional construct.’ (p.198). However, doubts remain as to the factor structure of
the MAI, and in addition, Eckhardt, Norlander, and Deffenbacher (2004) warned that, ‘given the weak psycho-
metric support for the anger expression items, it does not appear that the MAI offers an unambiguous assessment
of anger expression style’ (p. 28). Another caveat is that the relevance of MAI hypothetical anger provocations to
violent offenders may be questionable. This, as we shall see is a recurring theme in examining the utility of mea-
sures of anger when it comes to predicting criminal violence.
MAI SAMPLE ITEMS
Scale Example item
Frequency ‘Something makes me angry almost every day’
Duration ‘When I get angry, I stay angry for hours’
Magnitude ‘I often feel angrier than I think I should’
Hostile Outlook ‘Some of my friends have habits that annoy me very much’
Anger-in ‘I harbor grudges that I don’t tell anyone about’
Anger-Out ‘When I’m angry with someone, I let that person know’
Range of Anger-Eliciting Situations ‘I get angry when something blocks my plans’
Note. Items are rated on a 5-point Likert-type scale ranging from: ‘Completely undescriptive’ to ‘Completely
descriptive’.
State/C0Trait Anger Expression Inventory /C02nd Edition (STAXI-2)
(Spielberger, 1988, 1991, 1999 ).
Variable
The distinction between trait and state constructs has been promoted especially by Spielberger’s construction
of various state /C0trait scales/measures. The anger construct can also be viewed in terms of state and trait aspects.
However, as pointed out in Chapter 8 on Measures of Affect Dimensions by Boyle, Helmes, Matthews and Izard,
there are indeed many possible forms of affect constructs ranging all the way from transient/fleeting emotional
states, through longer lasting mood states, through motivational dynamic traits, to relatively stable enduring per-
sonality traits. While dichotomous state and trait constructs may be useful in some circumstances, there are actu-
ally many different forms of anger affect that lie somewhere on the continuum in-between these two polar
extremes that also require measurement (e.g., see Chapter 8; Davey & Day, 2007 ;Fernandez, 2008 ;Fernandez &
Kerns, 2008 ).
Description
The STAXI-2 ( Spielberger, 1999 ) which is purported to measure the experience, expression, and control of
anger, consists of 57 items, 6 scales, 5 subscales, and an Anger Expression Index (total anger expression)
(cf.Spielberger & Reheiser, 2009 ). The State Anger scale (15 items) measures anger intensity as a momentary
emotional state, while the Trait Anger scale (10 items) measures the disposition to experience angry feelings as a
personality-like trait over lengthy time periods (i.e., the individual’s disposition to become angry or angry tem-
perament). The Anger Expression scale (16 items) and the Anger Control scale (16 items) measure four anger-
related trait dimensions. As Vagg and Spielberger (2000) in their Interpretive Report (STAXI-2: IR t) indicated,
the Anger Expression-In scale measures the extent to which an individual ‘holds things in’ or suppresses anger,
whereas the Anger Expression-Out scale measures the actual expression of aggression. Concomitantly, the Anger
Control-In scale measures the extent to which an individual controls the expression of anger through attempts at
relaxation and ‘calming down’, whereas the Anger Control-Out scale measures the extent to which an individual
actively monitors and limits the overt expression of anger. As stated on the Mingarden website (see link below),
‘Scales include: State Anger, Trait Anger, Anger Expression-Out, Anger Expression-In, Anger Control-Out, Anger89 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
Control-In, and Anger Expression Index .’ Ratings of items are on a 4-point response scales that measure state anger
(intensity) as well as trait anger (frequency). Software for Windows sis available that permits an unlimited num-
ber of scoring and interpretive reports.
Sample
Spielberger and Reheiser (2004) reported that:
‘The normative samples for the STAXI-2 are based on the responses of more than 1900 individuals from two populations: a hetero-
geneous sample of 1644 normal adults (977 females, 667 males) and 274 hospitalized psychiatric patients (103 females, 171 males).
The sample of normal adults included managerial, technical, and clerical personnel: participants in stress management programs; health
care managers and professionals; insurance company employees; and undergraduate and graduate students enrolled in a large urban
university’. (p. 78)
Borteyrou, Bruchon-Schweitzer, and Spielberger (2008, 2014) utilized a large French sample of 1085 respon-
dents (539 males; 546 females; aged 18 to 70 years) in validating a French adaptation of the STAXI-2. In addition,
for the STAXI-2 C/A, the normative sample consisted of 836 individuals ( M513.77 years, SD52.54) ( Brunner &
Spielberger, 2009 ).
Reliability
Internal Consistency
Spielberger (1999) reported Cronbach alpha coefficients for the STAXI-2 scales ranging from 0.73 to 0.93
(Spielberger, 1999 ). In a sample of Hispanic undergraduates, Culhane and Morera (2010) reported alpha coeffi-
cients of 0.70 or higher. Cornell et al. (1999) reported alpha coefficients as follows: State Anger (.91), Trait Anger
(.90), Anger-In (.63), Anger-Out (.80), and Anger Control (.82). Spielberger and Reheiser (2004) reported alpha
coefficients based on the STAXI-2 normative sample ( N.1,900) ranging from .73 to .95 (Mdn 5.87), with alphas
for the AX index ranging from .75 to .82. For the STAXI-2 C/A, Brunner and Spielberger (2009) reported alpha
coefficients ( N5838) as follows: Trait Anger scale (.80), State Anger scale (.87), Anger Expression-Out (.70),
Anger Expression-In (.71), and Anger Control (.79). In a clinical sample ( N552), the corresponding alphas were:
Trait Anger scale (.88), State Anger scale (.94), Anger Expression-Out (.84), Anger Expression-In (.74), and Anger
Control (.89).
Test/C0Retest
For the French adaptation of the STAXI-2, Borteyrou et al. (2008) reported that over a two-month interval
(N5139), stability coefficients were .70 for the trait anger scale and .32 for the state anger scale, respectively.
Validity
Convergent/Concurrent
As evidence of convergent/concurrent validity, correlations between the STAXI trait anger and the Cook and
Medley (1954) Ho scale ranged from .43 to .59, and those between the STAXI with the BDHI total score ranged
from 0.66 to 0.73. Cornell et al. (1999, Table 1) reported positive inter-correlations between all of the STAXI sub-
scales (except STAXI Anger Control). Cornell et al. also reported positive correlations between the NAS and the
STAXI scales as follows: STAXI State Anger (.39), STAXI Trait Anger (.90), STAXI Anger-In (.43), STAXI Anger-
Out (.76), as well as positive correlations with the Novaco PI as follows: STAXI State Anger (.30), STAXI Trait
Anger (.63), STAXI Anger-In (.26), and STAXI Anger-Out (.49), respectively. For the STAXI-2 C/A, in the norma-
tive sample, positive correlations were found between the scales and subscales (except for Anger Expression-In
and Anger Control) ranging from .29 to .90 ( Brunner & Spielberger, 2009 ).
Divergent/Discriminant
As evidence of divergent/discriminant validity, Deffenbacher et al. (1996) [as cited in Eckhardt et al. 2004 ]
found that Trait Anger scores ‘correlated more highly with other anger-related constructs than with measures of
anxiety, depression, intoxication, phobic anxiety, paranoid thinking, and psychoticism’ (p. 29) thereby providing
support for divergent/discriminant validity. Also, Cornell et al. (1999) reported negative correlations between the
STAXI Anger Control subscale and each of the other STAXI subscales (ranging from /C0.09 to/C0.49). As further evi-
dence of discriminant validity, Cornell et al. reported negative correlations between the Novaco NAS and STAXI
Anger Control ( /C0.46), and between the Novaco PI and STAXI Anger Control ( /C0.21), respectively. For the90 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
STAXI-2 C/A, in the normative sample, negative correlations were found between both the Anger Expression-
In/Anger Control scales and the respective State and Trait scales/subscales ranging from -.03 to -.28 ( Brunner &
Spielberger, 2009 ). However, the STAXI-2 “was vulnerable to social desirability response bias…” ( McEwan,
Davis, MacKenzie, & Mullen, 2009, p. 431 ).
Construct/Factor Analytic
As evidence of construct/factor analytic validity, Fuqua et al. (1991) carried out a principal-axis factor analysis
with varimax rotation ( N5455 undergraduates) of the item intercorrelations for the 44-item STAXI, providing
support for the purported factor structure of the measure. Likewise, Forgays, Forgays, and Spielberger (1997) as
well as Forgays, Spielberger, and Forgays (1998) provided factor analytic evidence supporting the claimed STAXI
structure. Maxwell, Sukhodolsky, and Sit (2009) reported CFAs on the STAXI-2 item intercorrelations for two
Hong Kong samples ( N5489 and 775) providing support for the claimed factor structure. In a Swedish under-
graduate sample ( N5100), Lindqvist et al. (2003) found some support for the purported STAXI-2 factor structure.
De la Rubia, Gonza ´lez, and Landero (2010) reported both a principal components analysis with oblique promax
rotation and a confirmatory factor analysis of the STAXI-2-AX item intercorrelations in a sample of 226 Mexican
housewives, finding support for the purported factor structure. Borteyrou and Bruchon-Schweitzer (2014) con-
ducted both EFA and CFA analyses of the STAXI-2 item intercorrelations on a large French sample of 1085
respondents (539 males; 546 females; aged 18 to 70 years), and reported that they could only reproduce three of
the purported STAXI-2 factors.
Criterion/Predictive
As evidence of criterion/predictive validity, Spielberger and Reheiser (2010) stated that, suppressed anger, as
measured by the Anger-In subscale of the STAXI is a key factor in hypertension (e.g., Johnson, Spielberger,
Worden, & Jacobs, 1987 ;van der Ploeg, van Buuren, & van Brummelen, 1988 ). The STAXI has also been used in
studies of hardiness, well-being, and coping with stress ( Schlosser & Sheeley, 1985 ); anger in patients treated for
Hodgkin’s disease and lung cancer ( McMillan, 1984 ); Type A behavior ( Booth-Kewley, & Friedman, 1987 ); the
effects of marijuana use ( Pape, 1986; Stoner, 1988 ); and chronic pain ( Curtis, Kinder, Kalichman, & Spana, 1988 ).
Likewise, Antypa et al. (2013) reported that STAXI Anger-Out scores significantly predicted ‘anger expressed out-
wards ...in male suicidal patients compared to controls ( p,.001)’ (p. 393). Also, Desche ˆnes, Dugas, Fracalanza,
and Koerner (2012) reported that the STAXI-2 significantly predicted generalized anxiety disorder severity.
Coates and Pretty (2003) reported that the STAXI-2 Trait Anger and Anger-Out scales significantly predicted
future arthritic health status.
Location
Spielberger, C.D. (1988) Manual for the State /C0Trait Anger Expression Inventory . Odessa, FL: Psychological
Assessment Resources.
Spielberger, C.D. (1991). State /C0trait Anger Expression Inventory: Revised Research Edition: Professional
Manual. Odessa, FL: Psychological Assessment Resources.
Spielberger, C.D. (1999). STAXI-2: State /C0Trait Anger Expression Inventory Professional Manual . Odessa, FL:
Psychological Assessment Resources. www.mindgarden.com/products/staxs.htm (Retrieved January 20, 2014).
Results and Comments
As a successor to the STAXI, the STAXI-2 has had a slower rate of application in anger research, but it is the
currently preferred version. The instrument has been translated into Spanish ( Miguel-Tobal, Casado, Cano-
Vindel, & Spielberger, 2001 ) and adapted for use in Mexico (e.g., Alca´zar, Deffenbacher, & Byrne, 2011 ). With the
earlier theoretical caveats in mind, STAXI-2 results can be used to ascertain whether an individual’s anger is pri-
marily attributable to frustration or perceived maltreatment (as in state anger) or whether it is an indication of
premorbid anger-proneness (or trait anger). Even when the overt expression of anger (Anger-out) is minimal, the
STAXI-2 can capture anger that may be present where Anger-in scores are high. A person’s handling of anger
can lie between overcontrol and undercontrol and the STAXI-2 can help locate that point for purposes of treat-
ment planning.91 MEASURES REVIEWED IN DETAIL
II. EMOTIONAL DISPOSITIONS |
MEASURES REVIEWED BRIEFLY
The Anger Disorders Scale (ADS; DiGiuseppe & Tafrate, 2004 )/C0(not to be confused with the Anger Discomfort
Scale to be described later) was introduced as one of the more clinically-relevant measures for assessing anger.
The ADS consists of five domains and 18 subscales as follows: The Provocations domain includes (i) scope of
anger provocations and (ii) hurt/social rejection; the Arousal domain includes (i) physiological, (ii) duration,
and (iii) episode length; the Cognitions domain includes (i) suspicion, (ii) resentment, (iii) rumination, and
(iv) impulsivity; the Motives domain includes (i) tension reduction, (ii) coercion, and (iii) revenge; the Behaviors
domain encompasses (i) anger-in, (ii) physical aggression, (iii) verbal expression, (iv) indirect aggression, (v) pas-
sive aggression, and (vi) relational aggression. Subscale scores can be plotted as coordinates on a graph. The ADS
has been administered as a structured interview with 134 questions to be rated on a 5-point Likert-type response
scale ( Ahmed, Kingston, DiGiuseppe, Bradford, & Seto, 2012 ). Items are framed as questions with multi-choice
answers. For example, ‘My anger has been a problem for (i) ‘a week or less or not at all’, (ii) ‘a month or less’,
(iii) ‘About three months’, (iv) ‘About six months’, (v) ‘A year or more’; or ‘I get angry if someone makes me
look bad in front of others’ which is accompanied by the following options for answers: (i) ‘never’, (ii) ‘rarely’
(iii) ‘occasionally’, (iv) ‘often’, (v) ‘always’. The ADS was normed on a sample of over 1400 American adults
(aged 18 /C076 years) recruited from educational and work environments as well as the internet. Test /C0retest reli-
ability based on a college sample ( N565) was found to be .89 over a two week interval. A Short Form of the
ADS with three subscales (expression, anger-in, and vengeance) is also available.
With 18 subscales to the ADS, the number of possible profiles in which half of the subscales are elevated
approaches 50,000; this number exceeds 250,000 when all possible combinations are considered. As an alternative,
a Short Form with three subscales (expression, anger in, and vengeance) may be used.
TheAnger Parameters Scale (APS; Fernandez, 2001 ;Fernandez et al., 2010 ) was constructed in light of the frequent
distinction drawn between experience versus expression of anger and the further differentiation among subtypes of
anger experience and anger expression, Fernandez (2001) developed two separate scales: the Anger Parameters
Scale (APS) and the Anger Expressions Scale (AES). As it stands, the latter is a theoretical formulation undergoing
empirical evaluation, so the present focus is on the APS. In the APS, five anger parameters were operationalized to
delineate boundaries of anger activity: frequency, duration, intensity, latency, and threshold. The first three were
already subscales within the MAI and they have also been dependent measures in behavioral psychology.
However, as with pain and other perceptual responses, the occurrence of anger can also be measured in terms of
latency and threshold (Fernandez, 2010). Thus, we have five parameters measuring (i) how often one gets angry,
(ii) how long the anger lasts, (iii) how strong the anger is, (iv) how quick to anger, and (v) how sensitive to provoca-
tion. This is in keeping with affective chronometry for measurement of emotions ( Davidson, 1998 ) and the notion of
an emotion-generative process ( John & Gross, 2007 ). The 30 APS items, some of which are negatively-keyed are
organized into six items per parameter. Each item is rated on a 5-point Likert-type response scale reflecting the
extent to which the items are applicable to the individual test-taker. The APS has been administered to an adult
community sample ( Fernandez et al., 2010 ). Cronbach alpha coefficients were found as follows: Frequency (.85),
Duration (.90), for Intensity (.62), Latency (.88), and Threshold (.74) ( Fernandez et al., 2010 ). A principal components
analysis with oblique rotation led to the extraction of five components conforming to the five different parameters
of anger ( Fernandez et al., 2010 ). A separate PCA analysis based on the subscale inter-correlations led to a one-
component solution which the authors termed the Degree of maladaptiveness of anger. More recently the APS was
administered to a sample of incarcerated adults ( Fernandez et al., 2014 ) who showed elevations on frequency,
duration, and intensity (but not threshold or latency) as compared to adult counterparts from the community.
In summary, the APS accesses five basic parameters applicable to emotions or affective phenomena, independently
of how anger is qualitatively expressed. The parameters are internally consistent and supported by preliminary fac-
tor analytic investigation. Some examples of items include: ‘I seldom get angry’ (Frequency), ‘I stay angry only for a
short time’ (Duration), ‘It takes very little to make me angry’ (Threshold), My anger takes the form of annoyance’
(Intensity), and ‘My anger occurs immediately’ (Latency).
The Awareness and Expression of Anger Indicator (AEAI; Catchlove & Braha, 1985 ) attempts to assess Awareness
of feelings of anger as well as the Expression of anger. The AEAI uses one hypothetical vignette which is read
out to the individual who is asked two questions, one about how s/he feels, and the other about what s/he
would do. The individual is then told to imagine that the same scenario occurs on four successive mornings.
Again, s/he is asked the same two questions about feeling and expression of anger. Answers to these questions
are ranked with reference to the idealized response for each occurrence of the scenario. Based on a sample of 30
pre-surgical patients an inter-rater reliability of 0.94 was obtained. Overall, no significant correlation between92 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
awareness and expression of anger was observed. The AEAI is unique in its use of a single vignette, a very small
sample, and partly qualitative analyses of data. The last of these makes the AEAI worthy of consideration as an
adjunct to the standard quantitative rating scales used in anger assessment. The sole item consists of a hypotheti-
cal vignette as follows: ‘You have a regular arrangement with a friend to pick you up every morning at 08:15 to
go to work. He arrives at 08:35 a.m.’ (Catchlove & Braha, p. 114).
The Standardized Experience of Anger Measure (STEAM; Linden et al., 1997 ) was developed by Linden et al.
(1997) a measure of situation-specific anger who pointed out that the STAXI and most measures of anger opera-
tionalize anger as a trait construct. The authors of the STEAM generated a pool of 54 vignettes that contained at
least two of the following theorized components of anger: damage, intent, and preventability. Individuals
employed a 10 cm visual analogue scale to rate each vignette in terms of (i) degree of anger aroused by the
described situation; (ii) suitability of the vignette for different populations; and (iii) clarity of the vignette. Setting
high cutoffs on each of these criteria, a total of 12 vignettes emerged as usable for both students and the general
community. Thirty-one fellow students, friends, and family members comprised the sample used in this study.
Cronbach alpha coefficients were found to be 0.89 for college students and 0.88 in a community sample.
Test/C0retest reliability at one-month was found to be 0.83. Although significant, correlations were low between
the STEAM and the trait anger and anger-out scores on the Spielberger STAXI/STAXI-2 scales. This suggests that
the former is measuring something slightly different from the latter. The STEAM makes a good attempt to orga-
nize vignettes according to attributional theories of anger. It has acceptable reliability though factorial validity
remains to be evaluated. An example of the 12 vignettes is:
‘Your friends ask you to babysit their 10-year-old child for the evening. You have an important report for work that must be
completed by morning, so you will need quiet time to work while the child is sleeping. You try to explain this to the child, but s/he
continues to be loud and disruptive well beyond his/her normal bedtime. How anger-arousing is this situation?’
The Anger Control Inventory (ACI; Hoshmand & Austin, 1987 ) is based on a model of cognitive-behavioral and
person-situation interactions. This inventory began with 134 items, organized into 10 anger stimulus scales and six
anger response scales. Item retention was based on inter-judge agreement. The stimulus scales comprised (i) seeing
others abused, (ii) intrusion, (iii) personal devaluation, (iv) betrayal of trust, (v) minor nuisance, (vi) external control
and coercion, (vii) verbal abuse, (viii) physical abuse, (ix) unfair treatment, and (x) goal-blocking. The response scales
were (i) behavioral which was further differentiated into destructive or passive responding versus constructive or
assertive responding, (ii) cognitive which was differentiated into maladaptive cognition and cognitive skill deficit,
and (iii) arousal differentiated into duration and intensity. Two samples of 118 clinical clients and 190 undergraduate
students were used in the initial studies, respectively. Hoshmand, Austin, and Appell (1981) reported Cronbach
alpha coefficients (ranging from .54 to .81) for the stimulus scales and .76 to .89 for the response scales. One-month
retest reliability ranged from (.72 to .83) for the stimulus scales and (.73 to .83) for the response scales. When normals
were compared with batterers seeking treatment, significant differences appeared in the expected direction for four
of the stimulus scales and five of the response scales. Factor analysis of the 16 scales lent support to the presence of
an Anger Stimulus and an Anger Response Scale ( Hoshmand & Austin, 1987 ). For a sample of patients, therapist rat-
ings of anger difficulties were significant but low-to-moderate in their correlation with stimulus scales or response
scales. However, this issue of criterion validity is an important one that has been largely neglected in the anger assess-
ment literature. This inventory is well-conceived in terms of its broad division between stimulus and response scales,
and the further subdivision of the former into common types of anger elicitors.
The Anger Discomfort Scale (ADS; Sharkin & Gelso, 1991 )shares the same acronym as the Anger Disorders
Scale, but assesses something quite different: a person’s degree of intrapersonal and interpersonal discomfort
with anger. In that sense, it is associated with anger suppression. Respondents simply rate each of the 15 items
on a 4-point response scale ranging from ‘Almost never’ to ‘Almost always’. The ADS was administered to 150
undergraduates and was found to exhibit a Cronbach alpha coefficient of .81 and a test /C0retest reliability coeffi-
cient of .87. ADS scores correlated significantly and mostly with anger-in, less so with anger-out, and negatively
with anger control on the STAXI/STAXI-2. A principal axis factor analysis led to retention of four factors labeled:
intrapersonal discomfort with one’s own anger, positive views of anger, interpersonal discomfort with anger, and
perceived concomitants of anger. Koo and Park (1998, p. 60), reported that when the ADS was administered to
male and female counselor trainees, high scores were predictive of state anxiety. The ADS can be a particularly
useful tool in screening participants for readiness and motivation to enter into anger treatment programs. Here
are some examples of the 15 items making up the ADS: ‘I do not like it when I get angry’, ‘I believe that it is nat-
ural and healthy to feel angry’, ‘I create more problems for myself when I get angry’.93 MEASURES REVIEWED BRIEFLY
II. EMOTIONAL DISPOSITIONS |
TheAnger Readiness to Change Questionnaire (ARCQ; Williamson et al., 2003 ) measures readiness to change among
those identified as in need of intervention to manage problematic anger. The Transtheoretical Stages of Change
model ( Prochaska & DiClemente, 1984, 1986 ), one of the most influential models of behavior change, was originally
developed to describe the process of behavior change for addictive behavior, and postulates that individuals pass
through a series of stages involving a series of different processes when attempting to change their behavior.
The ARCQ scale is an adaptation of the Readiness to Change Questionnaire (RCQ; Heather & Rollnick, 1993)
designed to identify stages of change among problem drinkers. The 12-item measure incorporates the original RCQ
items modified to measure stages within the context of anger problems by changing the wording of each item from
alcohol to anger (e.g., ‘Sometimes I think I should try and control my anger’). The questionnaire has four items to
assess three stages of change; pre-contemplation, contemplation, and action. Responses are made on 5-point Likert-
type response scale scored from /C02 (strongly disagree) to 12 (strongly agree), with total scores for each scale calcu-
lated by summing the relevant item scores. Possible scores range from /C08t o18 on each scale. The authors recom-
mend a method of scoring (stage allocation) which classifies only participants with meaningful patterns of scores
over the three ARCQ scale scores. Patterns showing high (or low) scores on all three scales, or high scores on both
pre-contemplation and action but low on contemplation, represent non-meaningful patterns. An allocation to the
pre-contemplation stage, for example, will occur when a person scores a positive score on the pre-contemplation
scale, but a negative (or zero) score on both the contemplation and action scales. Some 418 adult male convicted
prisoners located in 16 different prisons in two states of Australia were recruited for a project evaluating anger man-
agement programs. The mean age of the sample was 28.75 years (range 518 to 62 years). Of the 411 participants
who indicated ethnicity, 266 (64.7%) classified themselves as Australian/New Zealander, 77 (18.7%) as Australian
Aboriginal, 40 (9.7%) as European, 10 (2.4%) as Asian, and 18 (4.4%) as other. The majority ( N5237) were single
(58.8%), with 123 married/defacto (30.5%), and the remainder (10.7%) divorced/widowed. Most were unskilled/
unemployed ( N5226, 58.7%), with 140 (36.4%) semi-skilled/tradesmen, and 19 professionals. Over half the sample
(N5223, 62.1%) reported current offences that included violence. Sentences being served ranged from 1 month to
26 years ( M550.0 months, SD559.0), with 137 (41.5%) first time offenders with mean number of previous offences
2.98 (ranging from 1 to 14 offences) among repeat offenders. The Cronbach alpha coefficient for the pre-
contemplation scale was (.58), for the contemplation scale (.79), and for the action scale (.78). Internal consistency
was enhanced when pre-contemplation items were reverse coded ( α5.82). The semi-structured interview of treat-
ment readiness developed by Serin (1998) and Serin and Kennedy (1997) was adapted for use as a brief 11-item
questionnaire, named the Treatment Readiness Scale (TRS). Correlations between this measure and the pre-
contemplation, contemplation, and action subscales were /C0.33, .50, and .55 respectively.
Three factors corresponding to the pre-contemplation, contemplation, and action stages were reliably identi-
fied. However, a CFA also showed that two contemplation items were loaded moderately by both the contempla-
tion and action factors. The final measurement model resulted in the addition of two extra pathways from the
action latent variable to two contemplation items. This model showed an improved model fit ( χ2
(49)5266.4,
p,.001; GFI 50.91; NNFI 50.86). There was support for the hypothesis that treatment was more beneficial in
reducing anger when ARCQ scores were initially high; either high total scores or classification to higher stages
(i.e., contemplation, preparation, action). The interaction between group and readiness to change was tested using
moderated multiple regression. The variables combined explained a significant 4.1% of the variance in standard-
ized anger change scores (F (2,271)55.78, p5.004). Only readiness to change was a significant predictor of anger
change (readiness to change, β5/C0.20; group, β5/C0.02). The ARCQ may be a useful tool for selecting participants
for anger management programs. It may also help to identify when alternative interventions are required for
those who are less ready for anger management. Sample items include: ‘Controlling anger better would be point-
less for me’ (Pre-contemplation); ‘I’m entitled to get angry, but sometimes I go too far’ (Contemplation); ‘Anyone
can talk about wanting to do something about anger, but I am actually doing something about it’ (Action).
The Short Anger Measure (SAM; Gerace & Day, 2014) measures angry feelings and aggressive impulses. The
SAM is a 12-item self-report measure of angry feelings and aggressive impulses. Originally developed for use as
a brief measure of anger for use with adolescents identified as ‘at risk’, the SAM differs from trait measures of
anger in that it asks respondents to rate their anger over the last week. It is intended for use as brief measure
of anger for use in forensic populations which can be used for screening purposes before treatment is offered
and to measure change in anger over time. SAM items were selected to address the frequency of the experience
and of the expression of anger. Respondents are asked to answer on a 5-point Likert-type response scale from
‘Never’ to ‘Very often’ over the previous week. Sample items are ‘I felt angry’, ‘Something annoyed me and I
couldn’t get it out of my mind’, ‘I felt like I was ready to explode’, ‘Other people or things got on my nerves’.
Total scores on the SAM in the development study ranged from 12 to 58, with a mean score of 23.56 ( SD59.31).94 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Some 73 adult male offenders in both community and prison settings in South Australia participated in the devel-
opment of the SAM. Of these 49 were in prison and 24 were in the community (reporting to a probation and
parole or community corrections center). The average age of participants was 30.38 years ( SD58.55;
Range519/C060 years) and the average level of education completed was Year 10. The full scale scores exhibited a
high Cronbach alpha coefficient of .91 (cf. Boyle, 1991 ). A subgroup of participants completed the SAM twice
over a two-week interval. The test /C0retest reliability coefficient (.74) suggested that the scores were relatively
stable, but still amenable to change. Correlations between the total SAM score and the subscales of the State /C0trait
Anger Expression Inventory-2 (STAXI-2, Spielberger, 1999 ) were used to assess concurrent validity. Scores on the
total SAM score were moderately associated with those obtained from the Trait Anger subscale of the STAXI
(.33), indicating that the SAM measures a construct that is related to trait anger. Scores were also moderately
correlated with Anger Expression-Out (.58), Anger Control-Out ( /C0.52), and Anger Control-In ( /C0.39) scales of
the STAXI-2, with weaker correlations between the measure and State Anger (.39) and Anger Expression-In (.26).
Anexploratory Principal Axis Factoring with an oblimin rotation was used to assess the factor structure of the
measure. Factor analysis identified two dimensions. The correlation between factors was high (.67) and the
authors suggest that persons administering the measure should use a total scale score. This study suggests that
the SAM may be a useful screening tool for problematic anger when only brief assessment is possible, and as an
outcome measure in treatment evaluation or in assessing changes in anger over time.
FUTURE RESEARCH DIRECTIONS
The aim of this chapter was to provide an overview of some of the most widely used and emerging self-report
measures of anger/hostility in adults. Even a cursory inspection of the 16 self-report measures reviewed in this
chapter shows that in general they do live up to psychometric standards of reliability and validity and are conve-
nient to administer and score. Where the substantive differences exist are in the particular construct of interest
(e.g., anger or hostility, arousal versus expression of anger, and different subtypes of anger). The choice of which
instrument to use is also dictated by the population of interest (e.g., community, psychiatric, forensic). As well,
the selection of a scale/measure hinges on the specific question being posed, whether this be related to classifica-
tion and diagnosis, case conceptualization, making decisions about the course and efficacy of treatment, or testing
hypothesized relationships between anger and other aspects of health and human functioning.
Nevertheless, some measures are lacking in data about test /C0retest reliability. Especially with state /C0trait mea-
sures such as the STAXI-2, it would be expected that the trait scale would exhibit temporal stability whereas the
state scale would exhibit much lower test /C0retest correlations, if it is truly sensitive to situational variability. Also,
there are many different forms of anger and hostility ranging all the way from transitory emotional states,
through longer lasting mood states, through motivational dynamic traits, and finally relatively stable enduring
personality traits.
One ongoing problem is the fact that self-report scales/measures (including those reviewed in this chapter)
are prone to motivational and response distortion. Clearly, a range of factors may be responsible for biased
responding, ranging all the way from deliberate conscious dissimulation to inadequate self-insight and/or lack of
conscious awareness of one’s anger and hostility. Responses to items included in self-report measures of anger
and hostility are easily distorted, and therefore, treatments and interventions based on scores from such scales
need to be considered carefully. One avenue of possible advancement would be to develop objective (T-data) test
measures of anger and hostility, in order to minimize the impact of such conscious or unconscious motivational
biases in responding (cf. Boyle, Stankov, & Cattell, 1995, p. 435; Schuerger, 2008).
A further ongoing problem for many of the self-report anger/hostility measures presented in this chapter, is
that the application of factor analytic methods often leaves much to be desired. A recurring theme is the use of
less than optimal factor analytic methods, particularly the now essentially outmoded principal components anal-
ysis plus varimax (‘Little Jiffy’) procedure which, at best, produces somewhat crude approximations to the actual
structure of the various measures (see Cattell, 1978 ). Likewise, use of varimax rotation can be problematic in arti-
factually imposing orthogonality onto the components, when in fact, the components may be correlated in real
life. Extracting and rotating components however, rather than factors, necessarily means that both measurement
and unique error variance contaminates the resultant components. Nonetheless, advances in factor analytic meth-
ods, including use of confirmatory factor analysis have already resulted in construction of new measures of anger
and hostility some of which are described in this chapter.95 FUTURE RESEARCH DIRECTIONS
II. EMOTIONAL DISPOSITIONS |
References
Ahmed, A. G., Kingston, D. A., DiGiuseppe, R., Bradford, J. M., & Seto, M. C. (2012). Developing a clinical typology of dysfunctional anger.
Journal of Affective Disorders ,136, 139/C0148.
Alca´zar, R. J., Deffenbacher, J. L., & Byrne, Z. S. (2011). Assessing the factor structure of the anger expression inventory (ML-STAXI) in a
Mexican sample. International Journal of Psychology and Psychological Therapy ,11, 307/C0318.
Antypa, N., Giegling, I., Calati, R., Schneider, B., Hartmann, A., Friedl, M., et al. (2013). MAOA and MAOB polymorphisms and anger-related
traits in suicidal participants and controls. Archives of Psychiatry and Clinical Neuroscience ,263, 393/C0403.
Baeg, S., Wang, S. K., Chee, I. S., Kim, S. Y., & Kim, J. L. (2011). Anger in elderly patients with depressive disorders. Psychiatry Investigation ,
8, 186/C0193.
Baker, M. T., Van Hasselt, V. B., & Sellars, A. H. (2008). Validation of the Novaco Anger Scale in an incarcerated offender population. Criminal
Justice and Behavior ,35, 741/C0754.
Bernstein, I. H., & Gesn, P. R. (1997). On the dimensionality of the Buss/Perry Aggression Questionnaire. Behaviour Research and Therapy ,
35, 563/C0568.
Biaggio, M. K. (1980). Assessment of anger arousal. Journal of Personality Assessment ,44, 289/C0298.
Biaggio, M. K., & Maiuro, R. D. (1985). Recent advances in anger assessment. In C. D. Spielberger, & J. N. Butcher (Eds.), Advances in personal-
ity assessment (Vol. 5, pp. 71 /C0101). Hillsdale, NJ: Erlbaum.
Biaggio, M. K., Supplee, K., & Curtis, N. (1981). Reliability and validity of four anger scales. Journal of Personality Assessment ,45, 639/C0648.
Booth-Kewley, S., & Friedman, H. S. (1987). Psychological predictors of heart disease: A quantitative review. Psychological Bulletin ,101,
343/C0362.
Borteyrou, X., & Bruchon-Schweitzer, M. (2014). French adaptation and validation of the Questionnaire of Anger STAXI-2 of C.D. Spielberger.
Retrieved from www.ugr.es/ Bcmetodo/pdf/comunicaciones/borteyrou.pdf Accessed on January 20, 2014.
Borteyrou, X., Bruchon-Schweitzer, M., & Spielberger, C. D. (2008). The French adaptation of the STAXI-2, C.D. Spielberger’s State-trait anger
expression inventory. L’Encephale ,34, 249/C0255.
Boyle, G. J. (1991). Does item homogeneity indicate internal consistency or item redundancy in psychometric scales? Personality and Individual
Differences ,12, 291/C0294.
Boyle, G. J., Stankov, L., & Cattell, R. B. (1995). Measurement and statistical models in the study of personality and intelligence. In D. H.
Saklofske, & M. Zeidner (Eds.), International handbook of personality and intelligence (pp. 417 /C0446). New York: Plenum.
Brodsky, S. L. (2011). Hostility and scorn. In S. L. Brodsky (Ed.), Therapy with coerced and reluctant clients (pp. 141 /C0155). Washington, DC:
American Psychological Association.
Brunner, T. M., & Spielberger, C. D. (2009). State-trait anger expression inventory-2 tchild and adolescent (STAXI-2 tC/A). Odessa, FL:
Psychological Assessment Resources.
Bryant, F. B., & Smith, B. D. (2001). Refining the architecture of aggression: A measurement model for the Buss /C0Perry Aggression
Questionnaire. Journal of Research in Personality ,35, 138/C0167.
Bushman, B. J., Cooper, H. M., & Lemke, K. M. (1991). Meta-analysis of factor analyses: An illustration using the Buss /C0Durkee Hostility
Inventory. Personality and Social Psychology Bulletin ,17, 344/C0349.
Buss, A. H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility. Journal of Consulting Psychology ,21, 343/C0349.
Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality and Social Psychology ,63, 452/C0459.
Caine, T. M., Foulds, G. A., & Hope, K. (1967). Manual of the Hostility and Direction of Hostility Questionnaire (HDHQ) . London: University of
London Press.
Catchlove, R. F., & Braha, R. E. D. (1985). A test to measure the awareness and expression of anger. Psychotherapy and Psychosomatics ,43,
113/C0119.
Cattell, R. B. (1978). The scientific use of factor analysis in behavioral and life sciences . New York: Plenum.
Cattell, R. B., & Vogelmann, S. (1977). A comprehensive trial of the scree and K.G. criteria for determining the number of factors. Multivariate
Behavioral Research ,12, 289/C0325.
Cho, Y. W., Kim, J. L., Kim, S. Y., Wang, S. K., & Chee, I. S. (2009). A standardization of the Korean version of the Reaction Inventory. Journal
of the Korean Society of Biological Therapies in Psychiatry ,15, 130/C0139.
Cleare, A. J., & Bond, A. J. (1997). Does central serotonergic function correlate inversely with aggression? A study using D-fenfluramine in
healthy subjects. Psychiatry Research ,69,8 9/C095.
Coates, P. D., & Pretty, G. H. (2003). Trait anger predicts arthritic health status. Australian Journal of Psychology ,55, 172.
Coccaro, E. F., Siever, L. J., Klar, H. M., Maurer, G., Cochrane, K., Cooper, T. B., et al. (1989). Serotonergic studies in patients with affective
and personality disorders. Archives of General Psychiatry ,46, 587/C0599.
Cook, W. W., & Medley, D. M. (1954). Proposed hostility and pharisaic-virtue scales for the MMPI. Journal of Applied Psychology ,38, 414/C0418.
Cornell, D. G., Peterson, C. S., & Richards, H. (1999). Anger as a predictor of aggression among incarcerated adolescents. Journal of Consulting
and Clinical Psychology ,67, 108/C0115.
Crowne, D. P., & Marlowe, D. (1960). A scale of social desirability independent of pathology. Journal of Consulting Psychology ,24, 349/C0354.
Culhane, S. E., & Morera, O. F. (2010). Reliability and validity of the Novaco anger scale and provocation inventory (NAS-PI) and
state/C0trait anger expression inventory-2 (STAXI-2) in Hispanic and non-Hispanic white student samples. Hispanic Journal of Behavioral
Sciences ,32, 586.
Curtis, G., Kinder, B., Kalichman, S., & Spana, R. (1988). Affective differences among of chronic pain patients. Anxiety Research: An International
Journal ,1,6 5/C073.
Davey, L., & Day, A. (2007). The poetics of anger control: Metaphorical conceptualizations of anger expression in violent offenders. In
E. I. Clausen (Ed.), Psychology of anger . Hauppauge, NY: Nova Science Publishers.
Davidson, R. J. (1998). Affective style and affective disorders: Perspectives from affective neuroscience. Cognition and Emotion ,12, 307/C0330.96 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Deffenbacher, J. L., Oetting, E. R., Thwaites, G. A., Lynch, R. S., Baker, R. S., Thacker, S., et al. (1996). State /C0Trait Anger Theory and the utility
of the Trait Anger Scale. Journal of Counseling Psychology ,43, 131/C0148.
De la Rubia, J. M., Gonza ´lez, M. T., & Landero, R. (2010). Factor structure of the STAXI-2-AX and its relationship to burnout in housewives.
Spanish Journal of Psychology ,13, 418/C0430.
Demirtas-Madran, H. A. (2013). The reliability and validity of the Buss-Perry Aggression Questionnaire (BAQ) /C0Turkish version. Turkish
Journal of Psychiatry ,24, 124/C0129.
Derogatis, L. R. (1977). SCL-90. Administration, scoring and procedures manual-1 for the R(evised) version . Baltimore, MD: Johns Hopkins
University School of Medicine.
Desche ˆnes, S. S., Dugas, M. J., Fracalanza, K., & Koerner, N. (2012). The role of anger in generalized anxiety disorder. Behavior Therapy ,33,
215/C0233.
Diamond, P. M., Wang, E. W., & Buffington-Vollum, J. (2005). Factor structure of the Buss-Perry Aggression Questionnaire (BPAQ) with men-
tally ill male prisoners. Criminal Justice and Behavior ,32, 546/C0564.
DiGiuseppe, R., & Tafrate, R. C. (2004). Anger disorders scale: Manual . Toronto, Ontario: Canada: Multi Health Systems, Inc.
Dougherty, D. M., Bjork, J. M., Marsh, D. M., & Moeller, F. G. (1999). Influence of trait hostility on tryptophan depletion-induced laboratory
aggression. Psychiatry Research ,88, 227/C0232.
Doyle, M., & Dolan, M. (2006). Predicting community violence from patients discharged from mental health services. British Journal of
Psychiatry ,189, 520/C0526.
Dutton, D. G. (1995). Intimate abusiveness. Clinical Psychology: Science and Practice ,2, 207/C0224.
Eckhardt, C., Norlander, B., & Deffenbacher, J. (2004). The assessment of anger and hostility: A critical review. Aggression and Violent Behavior ,
9,1 7/C043.
Evans, D. R., & Stangeland, M. (1971). Development of the Reaction Inventory to measure anger. Psychological Reports ,29, 412/C0414.
Evren, C., Cinar, O., Gulec, H., Celik, S., & Evren, B. (2011). The validity and reliability of the Turkish version of the Buss-Perry Aggression
Questionnaire in male substance dependent inpatients. Journal of Psychiatry and Neurological Sciences ,24, 283/C0295.
Felsten, G., & Leitten, C. L. (1993). Expressive, but not neurotic hostility is related to cardiovascular reactivity during a hostile competitive
task. Personality and Individual Differences ,14, 805/C0813.
Fernandez, E. (2001). The anger parameters and expressions scales . Washington, DC: US Library of Congress.
Fernandez, E. (2008). The angry personality: A representation on six dimensions of anger expression. In G. J. Boyle, G. Matthews, & D. H.
Saklofske (Eds.), The SAGE handbook of personality theory and assessment, Vol. 2: Personality measurement and testing (pp. 402 /C0419). Los
Angeles, CA: Sage.
Fernandez, E. (2010). Toward an integrative psychotherapy for maladaptive anger. In M. Potegal, G. Stemmler, & C. Spielberger (Eds.),
The international handbook of anger: Constituent and concomitant biological, psychological, and social processes (pp. 499 /C0514). New York: Springer.
Fernandez, E. (2013). Anger dysfunction and its treatment. In E. Fernandez (Ed.), Treatments for anger in specific populations: Theory, application,
and outcome (pp. 1/C014). New York: Oxford University Press.
Fernandez, E., Arevalo, I., Vargas, R., & Torralba, A. (2014). Norms for five parameters of anger: How do incarcerated adults differ from the
community? International Journal of Forensic Mental Health ,13,1 8/C024.
Fernandez, E., & Kerns, R. D. (2008). Anxiety, depression, and anger: The core of negative affect in medical populations. In G. J. Boyle, G.
Matthews, & D. H. Saklofske (Eds.), The SAGE handbook of personality theory and assessment, Vol. 1: Personality theories and models
(pp. 659 /C0676). Los Angeles, CA: Sage.
Fernandez, E., Vargas, R., & Garza, C. (2010, May). Five parameters for mapping the angry person: Results from a community sample . Paper pre-
sented at the Association for Psychological Science 22nd Annual Convention, Boston.
Forgays, D. G., Forgays, D. K., & Spielberger, C. D. (1997). Factor structure of the State /C0trait Anger Expression Inventory. Journal of Personality
Assessment ,69, 497/C0507.
Forgays, D. K., Spielberger, C. D., & Forgays, D. G. (1998). Factor structure of the State /C0trait Anger Expression Inventory for middle-aged
men and women. Assessment ,1998(5), 141 /C0155.
Fossati, A., Di Ceglie, A., & Acquarini, E. (1999). Proprieta /C18psicometriche della versione italiana del questionario di Buss e Durkee per la valu-
tazione dell’ostilita’ e della colpa in soggetti non clinici. Ricerche di Psicologia ,23,5 7/C078.
Fossati, A., Maffei, C., Acuarini, E., & Di Ceglie, A. (2003). Multigroup confirmatory component and factor analyses of the Italian version of
the Aggression Questionnaire. European Journal of Psychological Assessment ,19,5 4/C065.
Fuqua, D. R., Leonard, E., Masters, M. A., Smith, R. J., Campbell, J. L., & Fischer, P. C. (1991). A structural analysis of the State /C0trait Anger
Expression Inventory. Educational and Psychological Measurement ,51, 439/C0446.
Gerace, A., & Day, A. (2014). The Short Anger Measure (SAM): Development of a measure to assess anger in forensic populations. Journal of
Forensic Nursing ,10,4 4/C049.
Giancola, P. R. (2002). Alcohol-related aggression in men and women: The influence of dispositional aggressivity. Journal of Studies on Alcohol ,
63, 696/C0708.
Gidron, Y., Davidson, K., & Ilia, R. (2001). Development and cross-cultural and clinical validation of a brief comprehensive scale for assessing
hostility in medical settings. Journal of Behavioral Medicine ,24,1/C015.
Harris, J. A. (1995). Confirmatory factor analysis of the Aggression Questionnaire. Behavior Research and Therapy ,33, 991/C0993.
Hornsveld, R. H. J., Muris, P., & Kraaimaat, F. W. (2011). The Novaco Anger Scale /C0Provocation Inventory (1994 version) in Dutch forensic
psychiatric patients. Psychological Assessment ,23, 937/C0944.
Hoshmand, L. T., & Austin, G. W. (1987). Validation studies of a multifactor cognitive-behavioral anger control inventory. Journal of Personality
Assessment ,51, 417/C0432.
Hoshmand, L. T., Austin, G. W., & Appell (1981, August). The diagnosis and assessment of anger control problems . Paper presented at the
American Psychological Association Annual Convention, Los Angeles.
Howells, K. (2004). Anger and its links to violent offending. Psychiatry, Psychology and Law ,11, 189/C0196.97 REFERENCES
II. EMOTIONAL DISPOSITIONS |
John, O. P., & Gross, J. J. (2007). Individual differences in emotion regulation . New York: Guilford.
Johnson, E., Spielberger, C. D., Worden, T., & Jacobs, G. (1987). Emotional and familial determinants of elevated blood pressure in black and
white adolescent males. Journal of Psychosomatic Research ,31, 287/C0300.
Jones, J., Thomas-Peter, B., & Gangstad, B. (2003). An investigation of the factor structure of the Novaco Anger Scale. Behavioural and Cognitive
Psychotherapy ,31, 429/C0437.
Jones, J. P., Thomas-Peter, B., & Trout, A. (1999). Normative data for the Novaco Anger Scale from a non-clinical sample and implications for
clinical use. British Journal of Clinical Psychology ,38, 417/C0424.
Kneip, R. C., Delamater, A. M., Ismond, T., Milford, C., Salvia, L., & Schwartz, D. (1993). Self- and spouse ratings of anger and hostility as pre-
dictors of coronary heart disease. Health Psychology ,12, 301/C0307.
Koo, J. G., & Park, S. S. (1998). Counselor’s reactions to an angry client. Korean Journal of Counseling & Psychotherapy ,10,4 3/C061.
Kroner, D. G., Reddon, J. R., & Serin, R. C. (1992). The multidimensional anger inventory: Reliability and factor structure in an inmate sample.
Educational and Psychological Measurement ,52, 687/C0693.
Kubiak, T., Wiedig-Allison, M., Zgoriecki, S., & Weber, H. (2011). Habitual goals and strategies in anger regulation: Psychometric evaluation
of the Anger-Related Reactions and Goals Inventory (ARGI). Journal of Individual Differences ,32,1/C013.
Lange, A., Dehghoni, B., & De Beurs, E. (1995). Validation of the Dutch adaptation of the Buss-Durkee Hostility Inventory. Behaviour Research
and Therapy ,33, 229/C0233.
Lange, A., Pahlich, A., Sarucco, M., Smits, G., Dehghani, B., & Hanewald, G. (1995). Psychometric characteristics and validity of the Dutch
adaptation of the Buss-Durkee Hostility Inventory (the BDHI-D). Behaviour Research and Therapy ,33, 223/C0227.
Leenaars, L., & Lester, D. (2011). Indirect aggression and victimization are positively associated in emerging adulthood: The psychological
functioning of indirect aggressors and victims. Journal of College Student Development ,52,6 2/C076.
Linden, W., Leung, D., Chawla, A., Stossel, C., Rutledge, T., & Tanco, S. A. (1997). Social determinants of experienced anger. Journal of
Behavioral Medicine ,20, 415/C0432.
Lindqvist, J. K., Waterman, A. M., & Hellstro ¨m, A. (2003). Swedish adaptations of the Novaco Amger Scale-1998, the Provocation Inventory,
and the State /C0trait Anger Expression Inventory-2. Social Behavior and Personality ,31, 773/C0788.
Lindqvist, J. K., Waterman, A. M., & Hellstro ¨m, A. (2005). Internal reliability and construct validity of the Novaco Anger Scale-1998-S in a
sample of violent prison inmates in Sweden. Psychology: Crime & Law ,11, 223/C0237.
Luteijn, F., Starren, J., & Van Dijk, H. (1975). Nederlandse Persoonlijkheidsvragenlijst, Handleiding (Dutch Personality Inventory) . Lisse, The
Netherlands: Swets & Zeitlinger B.V.
Matthews, K. A., Jamison, J. W., & Cottington, E. M. (1985). Assessment of Type A, anger and hostility: A review of scales through 1982.
In A. M. Ostfeld, & E. D. Eaker (Eds.), Measuring psychosocial variables in epidemiologic studies of cardiovascular disease (NIH
Publication No. 85-2270) . Bethesda, MD: National Institutes of Health.
Matthews, K. A., & Saal, F. E. (1978). Relationship of the Type A coronary-prone behavior pattern to achievement, power, and affiliation
motives. Psychosomatic Medicine ,40, 630/C0636.
Maxwell, J. P., Sukhodolsky, D. G., & Sit, C. H. P. (2009). Preliminary validation of a Chinese version of the State /C0trait Anger Expression
Inventory-2. Asian Journal of Social Psychology ,12,1/C011.
McEwan, T. E., Davis, M. R., MacKenzie, R., & Mullen, P. (2009). The effects of social desirability response bias on STAXI-2 profiles in a clini-
cal forensic sample. British Journal of Clinical Psychology ,48, 431/C0436.
McMillan, S. C. (1984). A comparison of levels of anxiety and anger experienced by two groups of cancer patients during therapy for Hodgkin’s disease
and small cell lung cancer . Master’s thesis, Tampa, FL: University of South Florida.
Meesters, C., Muris, P., Bosma, H., Schouten, E., & Beuving, S. (1996). Psychometric evaluation of the Dutch version of the Aggression
Questionnaire. Behavior Research and Therapy ,34, 839/C0843.
Megargee, E. I., Cook, P. E., & Mendelsohn, G. A. (1967). Development and validation of an MMPI Scale of Assaultiveness in overcontrolled
individuals. Journal of Abnormal Psychology ,72, 519/C0528.
Miguel-Tobal, J. J., Casado, M. I., Cano-Vindel, A., & Spielberger, C. D. (2001). Inventario de Expression de Ira Estado-Rasgo: STAXI-2 . Madrid:
TEA Ediciones.
Miller, T. Q., Jenkins, C. D., Kaplan, G. A., & Salonen, J. T. (1995). Are all hostility scales alike: Factor structure and covariation among mea-
sures of hostility. Journal of Applied Social Psychology ,25, 1142/C01168.
Mills, J. F., Kroner, D. G., & Forth, A. E. (1998). Novaco Anger Scale: Reliability and validity within an adult criminal sample. Assessment ,5,
237/C0248.
Monahan, J., Steadman, H. J., Silver, E., Appelbaum, P. S., Robbins, P. C., Mulvey, E. P., et al. Rethinking risk assessment: The MacArthur study of
mental disorder and violence . Oxford, England: Oxford University Press.
Musante, L., MacDougall, J. M., Dembroski, T. M., & Costa, P. T. (1989). Potential hostility and dimensions of anger. Health Psychology ,8,
343/C0354.
Nakano, K. (2001). Psychometric evaluation on the Japanese adaptation of the Aggression Questionnaire. Behaviour Research and Therapy ,39,
853/C0858.
Novaco, R. W. (1975). Anger control: The development and evaluation of an experimental treatment . Lexington, MA: DC Health.
Novaco, R. W. (1977). Stress inoculation: A cognitive therapy for anger and its application to a case of depression. Journal of Consulting and
Clinical Psychology ,45, 600/C0608.
Novaco, R. W. (1994). Anger as a risk factor for violence among the mentally disordered. In J. Monahan, & H. J. Steadman (Eds.), Violence and
mental disorder (pp. 21 /C059). Chicago, IL: University of Chicago Press.
Novaco, R. W. (2003). The Novaco Anger scale and provocation inventory: Manual . Los Angeles, CA: Western Psychological Services.
Novaco, R. W., & Taylor, J. L. (2004). Assessment of anger and aggression in male offenders with developmental disabilities. Psychological
Assessment ,16,4 2/C050.
Oquendo, M. A., Graver, R., Baca-Garcia, E., Morales, M., Montalvan, V., & Mann, J. J. (2001). Spanish adaptation of the Buss-Durkee Hostility
Inventory (BDHI). European Journal of Psychiatry ,15, 101/C0112.98 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
Pape, N. E. (1986). Emotional reactions and anger coping strategies of anger suppressors and expression . Doctoral dissertation, Tampa, FL: University
of South Florida.
Prochaska, J. O., & Di Clemente, C. C. (1984). The transtheoretical approach: Crossing traditional boundaries of therapy . Homewood, IL: Dow Jones
Irwin.
Prochaska, J. O., & DiClemente, C. C. (1986). Towards a comprehensive model of change. In W. R. Miller, & N. Heather (Eds.), Treating addic-
tive behaviours: Processes of change (pp. 3/C027). New York: Plenum.
Reynolds, N. S., Walkey, F. H., & Green, D. E. (1994). The anger self report: A psychometrically sound (30 item) version. New Zealand Journal
of Psychology ,23,6 4/C070.
Riley, W. T., & Treiber, F. A. (1989). The validity of multidimensional self-report anger and hostility measures. Journal of Clinical Psychology ,45,
397/C0404.
Schill, T., Ramanaiah, N. V., & Conn, S. R. (1990). Development of covert and overt hostility scales from the Buss-Durkee Inventory.
Psychological Reports ,67, 671/C0674.
Schlosser, M. B., & Sheeley, L. A. (1985, August). Subjective well-being and the stress process . Paper presented at the 93rdannual convention of
the American Psychological Association, Los Angeles, CA.
Schuerger, J. M. (2008). The Objective-Analytic Test Battery. In G. J. Boyle, G. Matthews, & D. H. Saklofske (Eds.), The SAGE handbook of person-
ality theory and assessment, vol. 2: Personality measurement and testing (pp. 402 /C0419). Los Angeles, CA: Sage.
Serin, R. (1998). Treatment responsivity, intervention, and reintegration: A conceptual model. Forum on Corrections Research ,10,2 9/C032.
Serin, R., & Kennedy, S. (1997). Treatment readiness and responsivity: Contributing to effective correctional programming . Research Report, R-54.
Ottawa, ON: Correctional Services of Canada.
Sharkin, B. S., & Gelso, C. J. (1991). The anger discomfort scale: Beginning reliability and validity data. Measurement and Evaluation in
Counseling and Development ,24,6 1/C068.
Siegel, J. M. (1985). The measurement of anger as a multidimensional construct. In M. A. Chesney, & R. H. Rosenman (Eds.), Anger and hostility
in cardiovascular and behavioural disorders (pp. 59 /C081). New York: Hemisphere/McGraw-Hill.
Siegel, J. M. (1986). The multidimensional anger inventory. Journal of Personality and Social Psychology ,51, 191/C0200.
Siegman, A. W. (1994). Cardiovascular consequences of expressing and repressing anger. In A. W. Siegman, & W. Timothy (Eds.), Anger, hostil-
ity, and the heart (pp. 173 /C0197). Hillsdale, NJ: Erlbaum.
Siegman, A. W., Anderson, R., Herbst, J., Boyle, S., & Wilkinson, J. (1992). Dimensions of anger-hostility and cardiovascular reactivity in pro-
voked and angered men. Journal of Behavioral Medicine ,15, 257/C0272.
Siegman, A. W., Dembroski, T. M., & Ringel, N. (1987). Components of hostility and the severity of coronary artery disease. Psychosomatic
Medicine ,49, 127/C0135.
Smedslund, J. (1993). How shall the concept of anger be defined? Theory and Psychology ,3,5/C033.
Smith, T. W. (1994). Concepts and methods in the study of anger, hostility, and health. In A. W. Siegman, & T. W. Smith (Eds.), Anger, hostility,
and the heart (pp. 23 /C043). Hillsdale, NJ: Erlbaum.
Smith, T. W., & Gallo, L. C. (1999). Hostility and cardiovascular reactivity during marital interaction. Psychosomatic Medicine ,61, 436/C0445.
Spielberger, C. D. (1980). Preliminary manual for the State /C0trait Anger Scale (STAS) . Tampa, FL: University of South Florida.
Spielberger, C. D. (1988). Manual for the State /C0Trait Anger Expression Inventory . Odessa, FL: Psychological Assessment Resources.
Spielberger, C. D. (1991). State/C0trait anger expression inventory: Revised research edition: Professional manual . Odessa, FL: Psychological
Assessment Resources.
Spielberger, C. D. (1999). STAXI-2: State /C0Trait anger expression inventory professional manual . Odessa, FL: Psychological Assessment Resources.
Spielberger, C. D., & Reheiser, E. C. (2004). Measuring anxiety, anger, depression, and curiosity as emotional states and personality traits with
the STAI, STAXI, and STPI. In M. Hersen, D. L. Segal, & M. Hilsenroth (Eds.), Comprehensive handbook of psychological assessment (Vol. 2):
Personality assessment (pp. 74 /C080). New York: Wiley.
Spielberger, C. D., & Reheiser, E. C. (2009). Assessment of emotions: Anxiety, anger, depression, and curiosity. Applied Psychology: Health and
Well-Being ,1, 271/C0302.
Spielberger, C. D., & Reheiser, E. C. (2010). The nature and measurement of anger. In M. Potegal, G. Stemmler, & C. D. Spielberger (Eds.),
International handbook of anger: Constituent and concomitant biological, psychological, and social processes . New York: Springer.
Stoner, S. B. (1988). Undergraduate marijuana use and anger. Journal of Psychology ,122, 343/C0347.
Suarez, E. C. (2004). C-reactive protein is associated with psychological risk factors of cardiovascular disease in apparently healthy adults.
Psychosomatic Medicine ,66, 684/C0691.
Suarez, E. C., Lewis, J. G., & Kuhn, C. (2002). The relation of aggression, hostility, and anger to lipopolysaccharide-stimulated tumor necrosis
factor (TNF)-alpha by blood monocytes from normal men. Brain, Behavior and Immunity ,16, 675/C0684.
Suarez, E. C., & Williams, R. B. (1990). The relationships between dimensions of hostility and cardiovascular reactivity as a function of task
characteristics. Psychosomatic Medicine ,52, 558/C0570.
Suris, A., Lind, L., Emmett, G., Borman, P. D., Kashner, M., & Barratt, E. S. (2004). Measures of aggressive behavior: Overview of clinical and
research instruments. Aggression and Violent Behavior ,9, 165/C0227.
Taylor, J. L., & Novaco, R. W. (2005). Anger treatment for people with developmental Disabilities . Chichester, UK: Wiley.
Thurstone, L. L. (1947). Multiple-factor analysis . Chicago, IL: University of Chicago Press.
Tsorbatzoudis, H. (2006). Psychometric evaluation of the Greek version of the Aggression Questionnaire. Perceptual and Motor Skills ,102,
703/C0718.
Vagg, P. R., & Spielberger, C. D. (2000). State /C0trait Anger Expression Inventory tInterpretive Report (STAXI-2: IR t). PAR Psychological
Assessment Resources, Inc. 16204 North Florida Ave., Lutz, FL 33549 1.800.331.8378. www.parinc.com Copyright r1979, 1986, 1988, 1989,
1991, 1996, 1999 by Psychological Assessment Resources, Inc. www4.parinc.com/WebUploads/samplerpts/STAXI2_IR.pdf
van der Ploeg, H. M., van Buuren, E. T., & van Brummelen, P. (1988). The role of anger in hypertension. Psychotherapy and Psychosomatics ,43,
186/C0193.99 REFERENCES
II. EMOTIONAL DISPOSITIONS |
Velicer, W. F., Govia, J. M., Cherico, N. P., & Corriveau, D. P. (1985). Item format and the structure of the Buss-Durkee Hostility Inventory.
Aggressive Behavior ,11,6 5/C082.
Vertommen, H., & Rochette, F. (1979). Handleiding vragenlijst interpersoonlijke relaties . Lisse, The Netherlands: Swets & Zeitlinger B.V.
Vigil-Colet, A., Lorenzo-Seva, U., Codorniu-Raga, M. J., & Morales, F. (2005). Factor structure of the Aggression Questionnaire among different
samples and languages. Aggression Behavior ,31, 601/C0608.
von Collani, G., & Werner, R. (2005). Self-related and motivational constructs as determinants of aggression. An analysis and validation of a
German version of the Buss-Perry Aggression Questionnaire. Personality and Individual Differences ,38, 1631/C01643.
Williams, T. W., Boyd, J. C., Cascardi, M. A., & Poythress, N. (1996). Factor structure and convergent validity of the Aggression Questionnaire
in an offender population. Psychological Assessment ,8, 398/C0403.
Williamson, P., Day, A., Howells, K., Bubner, S., & Jauncey, S. (2003). Assessing offender motivation to address problems with anger.
Psychology, Crime and Law ,9, 295/C0307.
Wranik, T., & Scherer, K. R. (2010). Why do I get angry? A componential appraisal approach. In M. Potegal, G. Stemmler, & C. Speilberger
(Eds.), International handbook of anger (pp. 243 /C0266). New York: Springer.
Zelin, M. T., Adler, G., & Meyerson, P. G. (1972). Anger self-report: An objective questionnaire for the measurement of aggression. Journal of
Consulting and Clinical Psychology ,39, 340.100 4. ANGER/HOSTILITY MEASURES
II. EMOTIONAL DISPOSITIONS |
CHAPTER
5
Measures of Life Satisfaction
Across the Lifespan
Marco Weber, Claudia Harzer, E. Scott Huebner and Kimberly J. Hills
University of South Carolina, Columbia, SC, USA
The notion of life satisfaction or perceived quality of life has been of central importance to several areas of
inquiry including gerontology, mental health, positive psychology, quality of life, and social indicators research.
It is frequently used as a criterion measure in studies of the determinants of optimal well-being among adults
and children ( Lyubomirsky, King, & Diener 2005 ). Studies have also suggested that it is not only an important
psychological ‘outcome’, but it is also an important determinant of a variety of adaptive life outcomes, such as
health, longevity, quality relationships, and vocational success ( Lyubomirsky et al., 2005 ). In addition to research
applications, practical applications have been numerous as well, including clinical ( Frisch, 2006 ), school
(Gilman & Huebner, 2003 ), and business ( Judge, Thoresen, Bono, & Patton 2001 ) settings. Diener (2000) has issued
a call for nations to undertake the development of ongoing national indexes of well-being, including measures of
life satisfaction as core indicators.
Since Andrews and Robinson’s (1991) contribution to the last edition of this text, research on the measurement
and correlates of life satisfaction and related variables has increased dramatically. This increase has included sub-
stantial attention to a previously neglected population, that is, children and adolescents. Thus, research efforts
are now being undertaken across almost the entire lifespan. Consistent with the review of Andrews and
Robinson (1991) , the research continues to be ‘extensive, broad-ranging, and conceptually diffuse’ (p. 61).
Despite this widespread use of the construct in research and practice contexts, challenges remain with its defi-
nition, measurement, and interpretation. With respect to its definition, life satisfaction is often confused with pos-
itive affect (aka ‘happiness’) as well as the absence of negative affect. These constructs are often used
interchangeably in the literature, yet they are already differentiable (using factor analytic procedures) by children
as young as age 8 years ( Huebner, 1991a ). On the one hand, positive affect refers to individuals’ experience of fre-
quent positive emotions, such as joy, interest, and excitement, while negative affect refers to individuals’ experi-
ence of frequent negative emotions, such as guilt, anxiety, and anger. On the other hand, life satisfaction is
typically defined as a cognitive (vs. emotional) judgment of the positivity of one’s life as a whole and/or with
specific life domains (e.g., family, work, school) ( Diener, 1994 ). In addition to being conceptually and statistically
distinguishable, these constructs show different correlates and can diverge across time ( Diener, 1994 ).
Furthermore, life satisfaction reports tend to be more reliable across time ( Diener & Larsen, 1984 ). Thus, the over-
all conclusion from the extant body of research suggests that the constructs of life satisfaction and affect are
related somewhat, but show considerable discriminant validity, indicating the need to measure them separately
in studies of adult and child well-being and quality of life. Within the present chapter we will focus on this more
cognitive element of well-being, that is, on the assessment of life satisfaction.
Some measurement and interpretation issues also remain unresolved. Most measures of life satisfaction have
been derived from one of three separate theoretical models. Some measures are based on a ‘bottom up’ model, in
which a ‘general’ life satisfaction score is calculated by summing responses to a variety of domain-specific items
(e.g., my family life is good, I enjoy going to work, school is fun). In some cases, weighted responses are deter-
mined based on an individual’s judgment of the importance of the particular domain. An example of a ‘bottom
101Measures of Personality and Social Psychological Constructs.
DOI: http://dx.doi.org/10.1016/B978-0-12-386915-9.00005-X ©2015 Elsevier Inc. All rights reserved. |
up’ scale for adults is Frisch’s (1992) Quality of Life Inventory , and an example of such a scale for children is the
Perceived Life Satisfaction Scale (Adelman, Taylor, & Nelson 1989 ).
Other measures have been based on a ‘top down’ model, in which a ‘global’ life satisfaction score is deter-
mined from one or more items that are domain-free in nature (e.g., my life is good vs. my family life is good).
Such measures allow the respondents to formulate their responses based upon their own unique circumstances
and standards ( Pavot & Diener, 1993 ). This approach contrasts with the aforementioned ‘general’ approach as
well as the ‘multidimensional’ approach (see below) in which the items reflect content (e.g., nature and number
of domain-based items) determined by the test developer. An example of the ‘global’ type of scales for adults is
theSatisfaction with Life Scale (Diener, Emmons, Larsen, & Griffin 1985 ), and an example of a scale for children is
theStudents’ Life Satisfaction Scale (Huebner, 1991b ).
Finally, some measures of life satisfaction emphasize multidimensional assessments of specific life domains,
which are considered to be important to most, if not all, individuals of interest. Such instruments thus yield pro-
files of individuals’ life satisfaction judgments, providing greater emphasis on the domain scores rather than an
overall ‘general’ or ‘global’ rating. Thus, individuals may report varying levels of ratings in different life
domains, yielding more differentiated responses. Some research suggests the importance of such contextualized
profiles given that overall ratings of life satisfaction may mask important domain-based differences. For example,
Antaramian, Huebner, and Valois (2008) found that adolescents’ ratings of their family (but not general) satisfac-
tion were related to their current family structure (i.e., intact vs. non-intact families). An example of this type of
scale for adults is the Extended Satisfaction with Life Scale (Alfonso, Allison, Rader, & Gorman 1996 ), and an exam-
ple of a scale for children is the Multidimensional Students’ Life Satisfaction Scale (Huebner, 1994 ).
Empirical investigations have yet to resolve the ‘bottom up’ vs. ‘top down’ controversy. In fact, some research-
ers have concluded that the effects are bidirectional ( Schimmack, 2008 ). Whatever the case, given the conceptual
distinctions among the various measures, it appears that the decision as to which measure to select is best deter-
mined by the particular purpose of the researcher or practitioner. It remains for the user to carefully consider the
strengths and limitations of each approach and the associated measures.
Although multi-method assessments are recommended ( Diener, Inglehard, & Tay 2013 ), self-report has been
the main method of life satisfaction assessment. Other efforts have been undertaken (e.g., reports of others, obser-
vations of smiling behavior, heart rate reactivity, memory for positive and negative events), but have been rela-
tively understudied. Given the subjective nature of life satisfaction, self-reports are likely to remain the favored
approach to assessment. A number of cautions have been raised regarding the validity of self-reports of life satis-
faction. These cautions include response styles (e.g., social desirability, number use) and situational influences
(e.g., priming, mood, item order, and cultural effects), which can increase errors of measurement.
Nevertheless, Diener et al. (2013) have recently summarized the existing literature and concluded that such
artifacts can usually be controlled. They further concluded that the overall literature supports the construct valid-
ity of self-report life satisfaction scales, showing that they reveal important information about individuals’ per-
ceptions of their lives. Nevertheless, some measures have stronger empirical support than others for some
purposes. Thus, specific strengths and weaknesses of several widely used measures are discussed below.
The purpose of this chapter is to present information on a number of instruments used to measure life satisfac-
tion across the lifespan. It is hoped that this information will be of use to researchers interested in measuring life
satisfaction variables as well as practitioners (e.g., psychologists, social workers, medical personnel) who wish to
assess the perceived quality of life of individuals and/or groups.
We used six criteria to guide our selection of measures. First, we tried to include one-dimensional and multidi-
mensional self-report measures across all major age groups (youth, adults, seniors/third age) where possible.
However, we excluded measures of life satisfaction that focused on single specific domains (e.g., health-related
quality of life, job satisfaction, quality of school life). Second , we selected measures that primarily reflect cognitive
judgments, omitting any that confounded affective (e.g., positive emotions) and cognitive judgments (e.g., Oxford
Happiness Questionnaire; Hills & Argyle, 2002 ).Third , the final version of the instrument had to be developed
within the past two decades or used in numerous research studies during the past two decades, yielding approxi-
mately 30 citations or more (Web of Knowledge). Fourth , there had to be some information available related to
normative samples, reliability, and validity. Fifth, the instrument had to be readily available to researchers, that
is, available at no cost to the user. Sixth , we included only measures published after the prior addition of this
book (i.e., 1991) with one exception. We included the Satisfaction with Life Scale ( Diener et al., 1985 ), which is
the most widely cited measure of global life satisfaction for adults. In meeting the above criteria, we omitted a
few noteworthy published instruments, but provide readers with the references (e.g., Frisch, 1994 ;Schalock,
Keith, Hoffman, & Karan 1989 ).102 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
For organizational purposes, the measures were divided into two groups. The first group includes measures
constructed for adults while the second group includes measures constructed for children (primarily for ages
8/C018 years).
MEASURES REVIEWED HERE
Measures of Life Satisfaction for Adults
1.Satisfaction with Life Scale ( Diener et al., 1985 )
2.Temporal Satisfaction with Life Scale ( Pavot, Diener, & Suh. 1998 )
3.Quality of Life Index ( Ferrans & Powers, 1985, 1992 )
4.Personal Wellbeing Index ( Cummins, Eckersley, Pallant, Van Vugt, & Misajon 2003; International Wellbeing
Group, 2013 )
5.Extended Satisfaction with Life Scale ( Alfonso et al., 1996 )
6.Quality of Life Enjoyment and Satisfaction Questionnaire ( Endicott, Nee, Harrison, & Blumenthal 1993;
Ritsner, Kurs, Gibel, Ratner & Endicott 2005 )
7.Life Satisfaction Index for the Third Age ( Barrett & Murk, 2006 )
Measures of Life Satisfaction for Youth
1.Students’ Life Satisfaction Scale ( Huebner, 1991b )
2.Perceived Life Satisfaction Scale ( Adelman et al., 1989 )
3.Multidimensional Students’ Life Satisfaction Scale ( Huebner, 1994 )
OVER VIEW OF THE MEASURES
The measures selected here for review represent different models (i.e., bottom up vs. top down, unidimen-
sional vs. multidimensional) of assessment of life satisfaction and perceived quality of life. The first seven mea-
sures were predominantly developed for adults and seniors, whereas the last three are assessment tools of life
satisfaction of young people (aged 8 years and above).
We start with the Satisfaction with Life Scale (Diener et al., 1985 ) as a measure of global life satisfaction. This
brief scale consists of five items yielding a total score representing satisfaction with life as a whole. Next, the
Temporal Satisfaction with Life Scale (Pavot et al., 1998 ), which also follows the top down strategy, added a new
perspective to the assessment of global life satisfaction by measuring three temporal dimensions of life satisfac-
tion (i.e., past, present, and future life satisfaction). This measure consists of 15 items yielding a total life satisfac-
tion score, and three subscale scores (one for each temporal dimension). The Quality of Life Index (Ferrans &
Powers, 1985, 1992 ) assesses quality of life focusing on an individual’s judgments of satisfaction with and impor-
tance of different life domains (e.g., physical health and functioning). The measure consists of 32 items that are
rated with respect to the extent of satisfaction and importance. Subscale scores (multidimensional model) as well
as a total scale score (bottom up strategy) can be obtained. The Personal Wellbeing Index (e.g., International
Wellbeing Group, 2013 ) assesses adults’ quality of life utilizing one item that asks for satisfaction with life as a
whole, and additional eight domain-specific items (e.g., for satisfaction with standard of living). The domain-
specific items can be used individually (multidimensional model), or can be summarized to a total score (bottom
up) of quality of life. With the Extended Satisfaction with Life Scale (Alfonso et al., 1996 ), we review a measure
focusing on an individual’s satisfaction in life with respect to eight life domains (e.g., social life). The 50 items
yield one total life satisfaction score, and eight domain-specific subscale scores. As another example following a
multidimensional model, the Quality of Life Enjoyment and Satisfaction Questionnaire (e.g., Endicott et al., 1993 )
assesses health-related quality of life. Sixty items form the basic measure, which includes five subscales (e.g.,
physical health). Furthermore, a General QOL Index can be obtained. As the last measure in the group of assess-
ments in adulthood, we review the Life Satisfaction Index for the Third Age (Barrett & Murk, 2006 ) assessing life sat-
isfaction of people in their ‘third age’ (i.e., seniors). This measure comprises 35 items assessing five different
subscales (e.g., zest vs. apathy). A total life satisfaction score can also be computed. This measure is a multidi-
mensional one that utilizes the bottom up strategy to assess satisfaction with life.103 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
The next section focuses on a group of measures of life satisfaction in children and adolescents. We start with
theStudents’ Life Satisfaction Scale (Huebner, 1991b ). This brief assessment tool consists of seven items that lead
to a total score of young peoples’ satisfaction with life as a whole. Next, the Perceived Life Satisfaction Scale
(Adelman et al., 1989 ) focuses on young peoples’ satisfaction in different life domains (e.g., material and physical
well-being). Nineteen items yield a total score of perceived satisfaction with life. Finally, the Multidimensional
Students’ Life Satisfaction Scale (Huebner, 1994 ) is a multidimensional assessment of satisfaction with specific
domains (e.g., satisfaction with friendships). Its 40 items yield a total life satisfaction score, and five domain-
specific subscale scores.
More detailed descriptions of the life satisfaction measures reviewed in this chapter follow below. For each
measure, we present information on (1) theoretical conceptualization and key references; (2) description of the
items and response options; (3) sample(s); (4) reliability; (5) validity; (6) location; (7) psychometric strengths and
weaknesses; and (8) a sample of the actual measure.
Satisfaction with Life Scale (SWLS)
(Diener et al., 1985 )
Variable
Diener et al. (1985) defined global life satisfaction as a cognitive judgment of the quality of one’s own life.
Cognitive ‘judgments of satisfaction are dependent upon a comparison of one’s circumstances with what is
thought to be an appropriate standard. It is important to point out that the judgment of how satisfied people are
with their present state of affairs is based on a comparison with a standard which each individual sets for him or
herself; it is not externally imposed’ ( Diener et al., 1985 , p. 71).
Description
The SWLS consists of five items assessing individuals’ global life satisfaction. Respondents rate their agree-
ment with each item on a 7-point Likert-type response scale (1 5strongly disagree to 75strongly agree ). A total
score can be obtained by summing the responses to all five items. Scores can be interpreted in terms of relative
life satisfaction by comparing individual scores with scores from normative samples, but also absolutely
(i.e., 5 /C095extremely dissatisfied; 10 /C0145dissatisfied; 15 /C0195slightly dissatisfied; 20 5neutral; 21 /C0255
slightly satisfied; 26 /C0305satisfied; 31 /C0355extremely satisfied [ Pavot & Diener, 1993 ]).
Sample
Three samples were studied to develop the SWLS ( Diener et al., 1985 ). Sample 1 ( N5176 undergraduates; no
further information on age) was used to examine the psychometrics (e.g., M, SD, internal consistency, test /C0retest
reliability) and the validity of the measure. Sample 2 ( N5163 undergraduates; no further information on age)
was used to examine convergent and discriminant validity. With sample 3 ( N553; individuals with a mean age
of 75 years), the properties of the measure in elderly persons were examined. Further normative data for different
groups (e.g., students, adults, health-related samples) were reported in Pavot and Diener (1993, 2008) .Diener
et al. (1985) reported a mean life satisfaction score of 23.5 ( SD56.43) for sample 1.
Reliability
Internal Consistency
The 5-item SLWS shows a high Cronbach alpha coefficient of α5.87 (Diener et al., 1985 ). Likewise, Pavot and
Diener (1993) reported alpha coefficients ranging between α5.79 and α5.89.
Test/C0Retest
Two-month test /C0retest reliability of the SWLS is rtt5.82 (Diener et al., 1985 ). Further research on the test /C0ret-
est stability yielded coefficients of rtt5.83 (2 weeks), rtt5.84 (1 month), rtt5.64 and .82 (2 months), rtt5.50
(10 weeks), and rtt5.54 (4 years) (cf. Pavot & Diener, 1993 ).104 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Validity
Convergent/Concurrent
The SWLS exhibited positive correlations with several measures of subjective well-being ( Diener et al., 1985 ).
For example, the SWLS correlated with the Fordyce (1978) single item measure of happiness ( r5.57 to r5.58),
the Fordyce percent of time happy question ( r5.58 to r5.62), Bradburn’s (1969) measures of positive affect
(r5.50 to r5.51), and Andrews and Withey’s (1976) Delighted /C0Terrible scale ( r5.62 to r5.68) (sample 1
and 2). In the elderly sample, the SWLS correlated with the Life Satisfaction Index ( Adams, 1969 ) with a moder-
ate coefficient of .46 (sample 3; Diener et al., 1985 ).
Divergent/Discriminant
The SWLS exhibited negative correlations with Bradburn’s (1969) measure of negative affect ( r52.32 to
r52.37). The SWLS also exhibited a non-significant relationship ( r5.02) with the Marlowe /C0Crown scale
(Crowne & Marlowe, 1964 ), suggesting that SWLS scores are unrelated to a socially desirability response style
(sample 2; Diener et al., 1985 ).
Construct/Factor Analytic
Factorial validity was estimated via a principal axis factor analysis of the item intercorrelations in sample 1
(N5176). A Cattell scree plot of eigenvalues ( Cattell, 1978 ) suggested that the five items of the SWLS formed a
single factor ( Diener et al., 1985 ).
Criterion/Predictive
Evidence of predictive validity is presented in Pavot, Diener, Colvin, and Sandvik (1991) .
Location
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life Scale. Journal of
Personality Assessment ,49,7 1/C075.
Results and Comments
Substantial research evidence supports the reliability and validity of the SWLS as a measure of adults’ global
life satisfaction. This widely used and brief assessment tool can be integrated in large-scale assessments, but also
in all other kinds of studies where a measure of domain-free life satisfaction is of interest. The SWLS has been
adapted into several languages (e.g., Arabic, Chinese, Hebrew, Russian, Setswana). Furthermore, the SWLS
inspired the development of further multi-item measures of life satisfaction, some of which are presented in
this chapter.
The SWLS appears to be a useful measure of global life satisfaction within a broad age range (e.g., undergradu-
ates, middle-aged adults, and elderly individuals). Recently, a version for children has been developed (SWLS-C;
Gadermann, Schonert-Reichl, & Zumbo 2010 ) based on the SWLS. The five items and the response format have
been adapted for children in grades 4 to 7. Research on the SWLS-C suggests that it is a promising measure of
children’s global life satisfaction, but additional research would be beneficial ( Gadermann et al., 2010 ).
SATISFACTION WITH LIFE SCALE
Below are five statements that you may agree or dis-
agree with. Using the 1 /C07 scale below, indicate your
agreement with each item by placing the appropriate
number on the line preceding that item. Please be open
and honest in your responding.
1/C0Strongly disagree; 2 /C0Disagree; 3 /C0Slightly dis-
agree; 4 /C0Neither agree nor disagree; 5 /C0Slightly agree;
6/C0Agree; 7 /C0Strongly agree
1.In most ways my life is close to my ideal.2.The conditions of my life are excellent.
3.I am satisfied with my life.
4.So far I have gotten the important things I want
in life.
5.If I could live my life over, I would change almost
nothing.
Notes :S o u r c e : http://internal.psychology.illinois.edu/
Bediener/SWLS.html (Retrieved May 9, 2014).
Reproduced with permission.105 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Temporal Satisfaction with Life Scale (TSWLS)
(Pavot et al., 1998 )
Variable
The TSWLS measures three temporal dimensions of life satisfaction /C0past, present, and future life satisfaction.
The TSWLS is based on the Satisfaction with Life Scale (SWLS; Diener et al., 1985 ).Pavot et al. (1998) rephrased
the five original SWLS items with respect to all three time frames of interest.
Description
The TSWLS consists of 15 items assessing individuals’ global life satisfaction, separated on the basis of three
temporal dimensions (i.e., the past, present, and future satisfaction). Respondents rate their agreement with each
item on a 7-point Likert-type response scale (1 5strongly disagree to 75strongly agree ). All 15 items are positively
keyed. Four scores can be obtained for analyses. First, summing all 15 items leads to a total score ranging
between 15 and 105. Second, summing the five items for each temporal dimension (past satisfaction, present sat-
isfaction, and future satisfaction, respectively) provides three subscale scores ranging from 5 to 35.
Sample
Pavot et al. (1998) presented three samples. Sample 1 consisted of 157 students (mainly aged from 18 to 25
years) who filled in the TSWLS on three occasions with a 4-week interval between administrations 1 and 2, and a
5-week interval between administrations 2 and 3. Total score means were 63.61 ( SD516.03), 69.39 ( SD516.20),
and 69.96 ( SD516.42) for administration 1, 2, and 3, respectively. Means for the three subscales past, present,
and future satisfaction were 20.98 ( SD56.58), 21.79 ( SD55.96), and 26.32 ( SD54.81), respectively. The differ-
ences between means of past vs. future, and present vs. future were statistically significant. There were no signifi-
cant differences between means of past vs. present satisfaction subscales.
Sample 2 consisted of 294 adults ( M559 years; ranging from 25 to 88 years) and yielded a mean for the total
score of 70.80 ( SD514.83). Means for the three subscales were 22.52 ( SD56.81), 24.49 ( SD56.37), and 23.79
(SD55.93) for past, present, and future satisfaction, respectively. The differences between means of past vs. pres-
ent, and past vs. future were statistically significant. There were no significant differences between means of pres-
ent vs. future satisfaction subscales.
Sample 3 consisted of 66 older adults ( M579 years; ranging from 61 to 99 years) who filled in the TSWLS on
two occasions over a 4 /C06 week interval with total score means of 72.89 ( SD519.30), and 74.28 ( SD515.90).
Means for the three subscales were 23.34 ( SD57.24), 25.93 ( SD55.54), and 24.83 ( SD55.94) for past, present,
and future satisfaction, respectively. The differences between means of past vs. present, and present vs. future
were statistically significant. No significant differences were observed between means of past vs. future satisfac-
tion subscales.
Reliability
Internal Consistency
The total 15-item TSLWS exhibited Cronbach alpha coefficients ranging from α5.91 to α5.93 in all three
samples.
Test/C0Retest
In the sample of students (sample 1), test /C0retest reliability coefficients were rtt5.83 (4 weeks), rtt5.88 (5 weeks),
and rtt5.82 (9 weeks), respectively. No test /C0retest correlation coefficients were reported for sample 2 and 3.
Validity
Convergent/Concurrent
According to Pavot et al. (1998) , the total TSWLS correlated .89 and .74 with the original 5-item SWLS in sam-
ple 1 and 2, respectively. In sample 1, the total TSWLS correlated substantially with other measures of subjective
well-being, for example, with the Andrews and Withey (1976) Delighted /C0Terrible scale ( r5.75). In sample 1,
self-rated optimism ( Scheier & Carver, 1985 ) correlated positively with past ( r5.48), present ( r5.53), and future
satisfaction ( r5.64). In sample 2, satisfaction was positively correlated with SWLS exhibiting coefficients of106 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
r5.72 (past satisfaction), r5.92 (present satisfaction), and r5.59 (future satisfaction). In sample 3, self-rated
optimism ( Scheier & Carver, 1985 ) exhibited positive correlations of r5.45,r5.22, and r5.38 with past, present,
and future satisfaction, respectively. In sample 3, correlations between extraversion ( Costa & McCrae, 1992 ) and
satisfaction were r5.34 (past satisfaction), r5.19 (present satisfaction), and r5.40 (future satisfaction).
Divergent/Discriminant
In sample 1, the TSWLS yielded significant but lower coefficients with measures of the affective components
of subjective well-being (SWB) than with measures of the cognitive component (e.g., SWLS; Diener et al., 1985 ).
In sample 3, correlations between the TSWLS and peer-rated optimism ( Scheier & Carver, 1985 ) and self-rated
openness ( Costa & McCrae, 1992 ) were not significant. Negative correlations were obtained between neuroticism
(Costa & McCrae, 1992 ) and satisfaction, with r52.58 (past satisfaction), r52.46 (present satisfaction), and
r52.52 (future satisfaction) in sample 3.
Construct/Factor Analytic
An orthogonally rotated principal components analysis of the data from sample 2 yielded a three-dimensional
solution. Three components that represented the items of past satisfaction, present satisfaction, and future satis-
faction, respectively, with eigenvalues greater than 1 explained 73.80% of the total variance.
Criterion/Predictive
Correlations between self-reported health and the TSWLS yielded coefficients of r5.41, r5.32, r5.39, and
r5.45 for past, present, and future satisfaction as well as the total score of the TSWLS, respectively (sample 3).
A hierarchical regression analysis with a composite score for SWB (i.e., peer-reported SWLS and peer-reported
happiness scores [ Fordyce, 1977 ]) as the criterion variable, and the three time dimensions of the TSWLS as pre-
dictors, yielded significant R2-changes for the future satisfaction score in step 3, over and above the significant
increment of present satisfaction in step 1. Past satisfaction, entered in step 2, failed to explain significant addi-
tional variance. Nevertheless, the results show the incremental validity due to adding in the temporal aspect
when assessing life satisfaction among adults.
Location
Pavot, W., Diener, E., & Suh., E. (1998). The Temporal Satisfaction with Life Scale. Journal of Personality
Assessment ,70, 340/C0354.
Results and Comments
Research on the TSWLS has provided promising, preliminary evidence of its reliability and validity as a mea-
sure of global life satisfaction suitable for young, middle, and older-aged individuals (from 18 to 99 years).
Divergence across a number of validation variables (e.g., optimism, extraversion) among the three time frames
was reported, which supports the meaningfulness of the TSWLS scores. The TSWLS is based on the idea that it
can be useful when judging life as a whole to have the opportunity to differentiate past, present, and future-
related perspectives. The inclusion of the temporal dimensions opens the possibility of research investigating the
roles and antecedents of life satisfaction from a developmental perspective.
TEMPORAL SATISFACTION WITH LIFE SCALE
Below are 15 statements with which you may
agree or disagree. These statements concern either
your past, present, or future. Using the 1 /C07s c a l e
below, indicate your agreement with each item by
placing the appropriate number on the line preced-
ing that item. Please be open and honest in your
responding.
1/C0Strongly disagree; 2 /C0Disagree, 3 /C0Slightly dis-
agree; 4 /C0Neither agree nor disagree; 5 /C0Slightly agree,
6/C0Agree; 7 /C0Strongly agree1.If I had my past to live over, I would change
nothing.
2.I am satisfied with my life in the past.
3.My life in the past was ideal for me.
4.The conditions of my life in the past were excellent.
5.I had the important things I wanted in my past.
6.I would change nothing about my current life.
7.I am satisfied with my current life.
8.My current life is ideal for me.
9.The current conditions of my life are excellent.107 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
10.I have the important things I want right now.
11.There will be nothing that I will want to change
about my future.
12.I will be satisfied with my life in the future.
13.I expect my future life will be ideal for me.14.The conditions of my future life will be excellent.
15.I will have the important things I want in the future.
Notes :r1997 by William Pavot.
Reproduced with permission.
Quality of Life Index (QLI)
(Ferrans & Powers, 1985, 1992 )
Variable
Ferrans and Powers (1992) defined quality of life ‘as a person’s sense of well-being that stems from satisfaction
or dissatisfaction with the areas of life that are important’ for the person (p. 29). In choosing the label ‘satisfac-
tion’ they highlighted the idea that quality of life is determined by a judgment and evaluation of life conditions.
Following Campbell (1981) ,Ferrans and Powers (1985) considered satisfaction of needs in different life domains
as being relevant for an individual’s quality of life with the perceived discrepancy and achievement of needs
ranging from fulfillment to deprivation. Need in this approach was defined as ‘the amount of a particular reward
that a person may require’ ( Ferrans & Powers, 1985 , p. 17).
Description
The QLI measures quality of life considering an individual’s satisfaction with and importance of different life
domains (e.g., health care, physical health and functioning, marriage, family, friends, stress, standard of living,
occupation). Item content was derived from a literature review on quality of life and reports of patients about the
effects of hemodialysis on their quality of life. The QLI consists of 32 items that are rated twice, once for the
extent of satisfaction (1 5very dissatisfied to 65very satisfied ) and once for their importance (1 5very unimportant
to 65very important ). For use with dialysis patients, three more items measure aspects related to dialysis treat-
ment. According to Ferrans and Powers (1985) , a total score and four subscales can be computed, namely health
and functioning (12 items), socioeconomic (9 items), psychological/spiritual (7 items), and family (4 items).
Sample
The first sample ( Ferrans & Powers, 1985 ) consisted of 88 graduate students (97% females) with a mean age of
33.1 years (ranging from 23 to 52 years). The students completed the QLI in a paper-pencil questionnaire format.
The second sample (clinical sample; Ferrans & Powers, 1985 ) consisted of 37 dialysis patients (72% males) with a
mean age of 50.0 years (ranging from 24 to 75 years). The patients answered the QLI items in an interview for-
mat. The third sample (dialysis patients; Ferrans & Powers, 1992 ) consisted of 349 in-unit hemodialysis patients
(56% males) with a mean age of 55.2 years (ranging from 25 to 84 years). This sample was representative of the
population of adult in-unit hemodialysis patients in terms of gender ratio, months in dialysis, presence of diabe-
tes mellitus, or primary cause of renal failure, but it was slightly older, and had a higher ratio of white patients.
This sample responded to the QLI items in a paper-pencil questionnaire format.
The first and the second sample were used for initial examination of reliability and validity of the QLI total
scale. The third sample was utilized to study the factor structure of the QLI, and to further examine validity and
reliability of the total scale and the subscales.
Reliability
Internal Consistency
Cronbach alpha coefficients for the total scale ranged from α5.90 to α5.93 for the three samples. Alpha coef-
ficients for the subscales were α5.87 (health and functioning), α5.82 (socioeconomic), α5.90 (psychological/
spiritual), and α5.77 (family).
Test/C0Retest
Test/C0retest correlations were rtt5.87 for students ( n569; two-weeks interval) and rtt5.81 for patients ( n520;
one-month interval) ( Ferrans & Powers, 1985 ).108 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Validity
Convergent/Concurrent
The correlations of the QLI and a one-item rating of the overall satisfaction with life (How satisfied are you
with your life? 1 5very dissatisfied to 65very satisfied ) were r5.75 for the students, r5.65 for the patients of the
second sample, and r5.77 for the patients in the third sample ( Ferrans & Powers, 1992 ).Ferrans and Powers
(1992) showed that the means for the socioeconomic subscale differed between lower (less than $10,000 annual
family income; M519.58) and higher income participants ($10,000 and more annual family income; M521.70).
As expected, all other subscales and the total scale did not show any differences between the lower and higher
income participants.
Divergent/Discriminant
Atkinson, Zibin, and Chuang (1997) reported that the overall score of the QLI was not related to educational
level, employment status, years of unemployment, housing arrangement, and social involvement in patients with
schizophrenia ( N569;rs ranging from .01 to .16).
Construct/Factor Analytic
Ferrans and Powers (1992) reported results from a factor analysis with oblique promax rotation utilizing the
satisfaction responses weighted by importance. Examinations of the Scree plot, relative chi-square, simple struc-
ture, and presence of trivial factors suggested extraction of four factors explaining 91% of the variance. The inter-
correlations among the factors ranged from .31 to .60. A higher-order factor analysis yielded a single broad QLI
factor. The first-order factors health and functioning, socioeconomic, psychological/spiritual, and family exhib-
ited loadings of .70, .69, .85, and .48, respectively, on the higher-order QLI factor ( Ferrans & Powers, 1992 ).
Criterion/Predictive
According to Haywood, Garratt, Schmidt, Mackintosh, and Fitzpatrick (2004, p. 72) , following recovery from a
period of intensive care, ‘greater perceived health (and future health), greater social support, and hospital read-
mission explained 51% of the variance in higher QLI scores.’
Location
Ferrans, C. E., & Powers, M. J. (1985). Quality of Life Index: Development and psychometric properties.
Advances in Nursing Science ,8,1 5/C024.
Ferrans, C. E., & Powers, M. J. (1992). Psychometric assessment of the Quality of Life Index. Research in
Nursing and Health, 15, 29/C038.
Results and Comments
The QLI is one of the few measures of life satisfaction for adults that contain multiple items for each domain.
It provides a total score for quality of life and scores for the four subscales: health and functioning, socioeco-
nomic, psychological/spiritual, and family. Results of research with the QLI to date suggest satisfactory reliabil-
ity and validity, although further evidence for predictive validity and construct validity, especially discriminant
validity, is needed. The measure appears to demonstrate usefulness in clinical settings, and it also appears
suitable for healthy people. Different versions have been developed for very different groups of patients (e.g.,
cardiac, respiratory, cancer, and burn patients). Each of these versions includes the same set of 32 core items that
are judged with respect to the individual’s satisfaction and importance. Additionally, items are added to those
core items to address group-specific issues.
Results presented by Ferrans and Powers (1992) showed that the socioeconomic subscale differentiates
between lower and higher income participants. Future research is needed to further investigate the discrimina-
tion of different groups like healthy vs. unhealthy participants, who would be expected to show different means
on the subscale of health and functioning. Additionally, Ferrans and Powers (1992) provided several ideas
regarding how to improve the QLI (e.g., items relating to leisure activities or social support could be split up into
different components to reduce conceptual ambiguity). Finally, the replicability of the factor structure should be
addressed in future studies as well, because Ferrans and Powers (1992) highlighted some sample-specific aspects
of the reported factorial structure of the QLI in in-unit hemodialysis patients.109 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
QUALITY OF LIFE INDEX
PART 1: For each of the following, please choose
the answer that best describes how satisfied you are
with that area of your life. Please mark your answer
by circling the number. There are no right or wrong
answers.
1/C0Very dissatisfied; 2 /C0Moderately dissatisfied;
3/C0Slightly dissatisfied; 4 /C0Slightly satisfied; 5 /C0
Moderately satisfied; 6 /C0Very satisfied
HOW SATISFIED ARE YOU WITH:
1.Your health?
2.Your health care?
3.The amount of pain that you have?
4.The amount of energy you have for everyday
activities?
5.Your ability to take care of yourself without help?
6.The amount of control you have over your life?
7.Your chances of living as long as you would like?
8.Your family’s health?
9.Your children?
10.Your family’s happiness?
11.Your sex life?
12.Your spouse, lover, or partner?
13.Your friends?
14.The emotional support you get from your family?
15.The emotional support you get from people other
than your family?
16.Your ability to take care of family responsibilities?
17.How useful you are to others?
18.The amount of worries in your life?
19.Your neighborhood?
20.Your home, apartment, or place where you live?
21.Your job (if employed)?
22.Not having a job (if unemployed, retired, or
disabled)?
23.Your education?
24.How well you can take care of your financial needs?
25.The things you do for fun?
26.Your chances for a happy future?
27.Your peace of mind?
28.Your faith in God?
29.Your achievement of personal goals?
30.Your happiness in general?
31.Your life in general?
32.Your personal appearance?
33.Yourself in general?
PART 2: For each of the following, please choose the
answer that best describes how important that area ofyour life is to you. Please mark your answer by circling
the number. There are no right or wrong answers.
1/C0Very unimportant; 2 /C0Moderately unimportant;
3/C0Slightly unimportant; 4 /C0Slightly important; 5 /C0
Moderately important; 6 /C0Very important
HOW IMPORTANT TO YOU IS:
1.Your health?
2.Your health care?
3.Having no pain?
4.Having enough energy for everyday activities?
5.Taking care of yourself without help?
6.Having control over your life?
7.Living as long as you would like?
8.Your family’s health?
9.Your children?
10.Your family’s happiness?
11.Your sex life?
12.Your spouse, lover, or partner?
13.Your friends?
14.The emotional support you get from your family?
15.The emotional support you get from people other
than your family?
16.Taking care of family responsibilities?
17.Being useful to others?
18.Having no worries?
19.Your neighborhood?
20.Your home, apartment, or place where you live?
21.Your job (if employed)?
22.Having a job (if unemployed, retired, or disabled)?
23.Your education?
24.Being able to take care of your financial needs?
25.Doing things for fun?
26.Having a happy future?
27.Peace of mind?
28.Your faith in God?
29.Achieving your personal goals?
30.Your happiness in general?
31.Being satisfied with life?
32.Your personal appearance?
33.Are you to yourself?
Notes: Presented are the items of the QLI Generic
Version-III.
Source: www.uic.edu/orgs/qli/questionaires/question-
nairehome.htm (Retrieved May 9, 2014).
rCopyright 1984 & 1998 Carol Estwing Ferrans and
Marjorie J. Powers.
Reproduced with permission.110 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Personal Wellbeing Index (PWI)
(Cummins et al., 2003; International Wellbeing Group, 2013 ).
Variable
There are hundreds of well-being measures available, but many of them are designed for highly specific sub-
groups of individuals (e.g., cancer patients, individuals with disabilities), and therefore, are not useful in examin-
ing the general population. To the contrary, measures for the general population often cannot be used within
specific groups (cf. International Wellbeing Group, 2013 ). Furthermore, most of those measures do not clearly
discriminate between objective and subjective aspects of quality of life ( International Wellbeing Group, 2013 ).
The Personal Wellbeing Index (PWI) was thus constructed for the assessment of the subjective dimension of qual-
ity of life in both the general population and specific subgroups. The PWI is based on the Comprehensive
Quality of Life Scale ( Cummins, 1997 ), and is part of the Australian Unity Wellbeing Index ( Cummins et al.,
2003) comprising both the PWI and the National Wellbeing Index (NWI). The PWI assesses adults’ satisfaction
with life as a whole, and satisfaction with life in different life domains.
Description
The PWI consists of eight items assessing individuals’ life satisfaction. One item assesses satisfaction with life
as a whole, whereas seven additional items assess domain-specific satisfaction (i.e., satisfaction with standard of
living, health, achievements, personal relationships, personal safety, community connectedness, and future secu-
rity). In a revised version ( International Wellbeing Group, 2013 ), the domain of religiousness has been added.
Respondents rate their agreement with each item on an 11-point Likert-type response scale (0 5completely dissatis-
fiedto 105completely satisfied ). Scoring of the PWI involves analyzing each of the 7 (excluding religiousness) or
8 (including religiousness) domain items as a separate variable, or computing a total score that can be obtained
by averaging all 7 or 8 domain-specific items ( International Wellbeing Group, 2013 ). The scores are then
converted using the formula ‘(score/10) 3100’ to produce percentage of scale maximum units on a 0 to 100
distribution (cf. Cummins et al., 2003 ).
Sample
Cummins et al. (2003) reported findings from a nationally representative sample of 2,000 Australians (ranging
in age from 18 to 76 1years). This sample yielded a mean score of 75.48 ( SD519.67) for the single item on satis-
faction with life as a whole. Means were reported for satisfaction with standard of living ( M575.78, SD519.50),
health ( M573.97, SD521.38), achievements ( M573.48, SD518.51), personal relationships ( M578.44,
SD521.22), personal safety ( M575.40, SD520.25), community connectedness ( M568.98, SD520.84), and
future security ( M569.29, SD521.24). For the composite score of the PWI consisting of all seven domains, a
mean of 73.48 ( SD513.57) was found ( Cummins et al., 2003 ).
Reliability
Internal Consistency
The PWI exhibited Cronbach alpha coefficients ranging between α5.70 and α5.85 across several studies
(International Wellbeing Group, 2013 ).
Test/C0Retest
Across a 1 /C02 week interval, the PWI yielded an intraclass correlation of .84 ( Lau & Cummins, 2005 ).
Validity
Convergent/Concurrent
Cummins et al. (2003) reported that the PWI total score correlated substantially with satisfaction with life as a
whole ( r5.67). Participants within the two highest impact groups of a happy event scored higher on the PWI as
compared with lower impact groups ( Cummins et al., 2003 ). The International Wellbeing Group (2013) reported a
correlation of .78 between the PWI and the SWLS ( Diener et al., 1985 ).Renn et al. (2009) reported positive correlations
between PWI, and both the environmental mastery subscale ( r5.59), and the self-acceptance subscale ( r5.68) of the
Scales of Psychological Well-Being (SPWB; Ryff & Keyes, 1995 ) in an Austrian sample ( N5573 college students).111 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Divergent/Discriminant
Renn et al. (2009) reported that the PWI was weakly correlated with the SPWB scales of autonomy (r5.28),
personal growth (r5.22), and purpose in life (r5.17) in an Austrian sample ( N5573 college students) .
Construct/Factor Analytic
Cummins et al. (2003) reported results of a principal components analysis (with oblimin rotation) based on the
intercorrelations of the seven domain-based items of the PWI, as well as three additional items from the National
Wellbeing Index (NWI; i.e., economic situation, state of the environment, social conditions). This analysis yielded
two components with one representing all seven PWI items and another representing the three items of the NWI.
The loadings of the PWI items ranged from .51 to .72. The two indices (i.e., PWI and NWI) correlated positively
(r5.44). The items of the 8-domain version also consistently form a single factor accounting for about 50% of the
variance ( International Wellbeing Group, 2013 ).
Criterion/Predictive
According to Cummins et al. (2003) , on the domain level, satisfaction with standard of living emerged as the
most substantial predictor ( β5.32) of overall satisfaction, followed by satisfaction with achievement and relation-
ships with β5.20 and β5.19, respectively. Satisfaction with health, future security, and community also emerged
as significant predictors of overall satisfaction with β5.14,β5.10, and β5.08, respectively. Satisfaction with
safety did not contribute to explaining variance in satisfaction with life as a whole, but it contributed to predict-
ing other variables (e.g., social capital, National Wellbeing Index, life in Australia).
Location
Cummins, R. A., Eckersley, R., Pallant, J., Van Vugt, J., & Misajon, R. (2003). Developing a national index of
subjective wellbeing: The Australian Unity Wellbeing Index. Social Indicators Research ,64, 159/C0190.
International Wellbeing Group (2013) .Personal Wellbeing Index: 5th Edition . Melbourne, Australia: Australian
Centre on Quality of Life, Deakin University. Source: www.deakin.edu.au/research/acqol/iwbg/wellbeing-
index/index.php (Retrieved May 9, 2014).
Results and Comments
Research into the psychometric properties of the PWI has been somewhat sparse, but the findings to date
are promising. The PWI seems to be useful for a broad age-range from younger adults to elderly people (18 to
761years). For young people, a pre-school version (PWI-PS; Cummins & Lau, 2005a ), and a school-children and
adolescents’ version (PWI-SC; Cummins & Lau, 2005b ) have been developed. Furthermore, a version has been
developed for individuals with intellectual disabilities, the PWI-ID ( Cummins & Lau, 2005c ).
As the PWI is a brief measure of subjective quality of life, it is useful for several research purposes, but it also
may be useful for applied professionals in different areas (e.g., counseling, program evaluation). Like the interna-
tionally used SWLS ( Diener et al., 1985 ), the PWI has been used in several languages as a brief scale assessing sat-
isfaction with life as a whole and domain-specific satisfaction in adults.
PERSONAL WELLBEING INDEX
The following questions ask how satisfied you feel,
on a scale from zero to 10. Zero means you feel no satis-
faction at all and 10 means you feel completely satisfied.
Part I (Optional item): Satisfaction
with Life as a Whole
Thinking about your own life and personal circum-
stances, how satisfied are you with your life as a whole?
Part II: Personal Wellbeing Index
How satisfied are you with ...1.your standard of living?
2.your health?
3.what you are achieving in life?
4.your personal relationships?
5.how safe you feel?
6.feeling part of your community?
7.your future security?
8.your spirituality or religion?
Note: Source: www.deakin.edu.au/research/acqol/
iwbg/wellbeing-index/index.php (Retrieved May 9,
2014). Reproduced with permission.112 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Extended Satisfaction with Life Scale (ESWLS)
(Alfonso et al., 1996 )
Variable
Alfonso et al. (1996) defined the cognitive component of subjective well-being as the ‘intellectual evaluation of
one’s life satisfaction either globally or with respect to specific life domains’ (p. 276; cf. Andrews & Robinson,
1991;George, 1981 ;Myers & Diener, 1995 ;Pavot & Diener, 1993 ). The selected domains of interest were those
highlighted within well-known, sophisticated approaches to subjective well-being and life satisfaction as well
as those being most strongly related to overall subjective well-being and those applicable to most people
(e.g., Alfonso, 1995 ;Alfonso, Allison, & Dunn 1992 ;Diener, 1984 ;Myers & Diener, 1995 ). These domains are
social life, sexual life, relationship, self, physical appearance, family life, school life, and job.
Description
The ESWLS measures an individual’s satisfaction with life in general, and with respect to eight life domains
(i.e., social life, sexual life, relationships, self, physical appearance, family life, school life, and job) with a total of
50 items, using a 7-point Likert-type response format (ranging from 1 5strongly disagree to 75strongly agree ).
All subscales are measured with five items except for job satisfaction, which comprises 10 items.
Items of the ESWLS (except school life and job) were based on the phrasing of the SWLS items ( Diener et al.,
1985). The SWLS items were replicated with minor modifications. For example, Alfonso et al. (1996) changed
‘In most ways my life is close to my ideal’ to ‘In most ways my social life is close to my ideal’. The SWLS item
‘If I could live my life over, I would change almost nothing’ was modified to ‘I am generally pleased with my life
I lead’ to exclude the regret-related connotation of this item (cf. Alfonso et al., 1996 ). The items measuring school
life derived from the Perceived Quality of Academic Life scale (PQAL; Okun, Kardash, Stock, Sandler, & Baumann
1986), but were converted to the wording of the rest of the items of the ESWLS. Job satisfaction items were based
on the Minnesota Satisfaction Questionnaire /C0Short Form (MSQ; Weiss, Dawis, England, & Lofquist 1967 ), but are
not identical with respect to wording.
Sample
The total sample was comprised of 302 undergraduate students, which were mainly white, single, and had
middle class backgrounds. The exact number of respondents varied across subscales from 182 for job satisfaction
to 302 for satisfaction with self. A total of 109 participants completed the ESWLS twice allowing for analyzing the
test/C0retest reliability. Means of the five-item scales ranged from 19.6 (physical appearance) to 25.1 (social life,
self) with standard deviations ranging from 4.7 (school life) to 8.3 (family life). The 10-item subscale of job satis-
faction had a mean of 48.2 with a standard deviation of 11.0 (cf. Alfonso et al., 1996 ).
Reliability
Internal Consistency
For the five-item scales, Cronbach alpha coefficients were high ranging from α5.81 (school life) to α5.96
(social life, sex life, relationship). Job satisfaction exhibited a high alpha coefficient as well ( α5.88) ( Alfonso
et al., 1996 ).
Test/C0Retest
Two-week test /C0retest reliability coefficients ranged from rtt5.74 (school life) to rtt5.87 (sex life) ( Alfonso
et al., 1996 ).
Validity
Convergent/Concurrent
Intercorrelations among the subscales of the ESWLS, and correlations between the ESWLS and the Rosenberg
Self Esteem Scale (RSE; Rosenberg, 1965 ), and Self-Deception Subscale of the Balanced Inventory of Desirable
Responding (BIDR; Paulhus, 1991 ) were reported as indicators of convergent validity in Alfonso et al. (1996) .
The domain-related subscales of the ESWLS showed positive correlations with the general satisfaction score with
coefficients ranging from r5.28 (school) to r5.63 (self). The intercorrelations among the domain specific sub-
scales ranged from r5.11 (relationship /C0physical appearance) to r5.57 (relationship /C0sex life). The subscale of113 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
self exhibited the strongest correlations with the RSE and the Self-Deception Subscale of the BIDR, with coeffi-
cients of r5.59 and r5.49, respectively. The other subscales showed correlation coefficients between r5.12 (job)
and r5.48 (general life) with the RSE, and between r5.11 (job) and r5.35 (social life) with the Self-Deception
Subscale of the BIDR.
Divergent/Discriminant
The Impression Management (IM) subscale of the BIDR and the Agreement Response Scale (ARS; Couch &
Keniston, 1960 ) were utilized as measures for social desirability and acquiescence, respectively. The ESWLS sub-
scales were only modestly related to IM with correlation coefficients ranging between r5.09 (physical appear-
ance) and r5.31 (school). The ARS was not related to the ESWLS subscales (coefficients ranged between r52.16
and r5.14).
Construct/Factor Analytic
In a principal components analysis (with direct oblimin rotation) of the intercorrelations of 40 items of the
ESWLS (i.e., without the job subscale), Alfonso et al. (1996) extracted eight components that explained 77% of
the variance. In a confirmatory factor analysis (CFA), indices indicated a modest fit, with NFI 5.79, AGFI 5.69,
and RMSR 5.06. Correlations among the components ranged from r5.00 (physical appearance /C0school) to
r5.46 (general life /C0social life), with a median of r5.26. Correlations among the CFA latent traits ranged from
r5.07 (physical appearance /C0school life) to r5.75 (general life /C0self) with a median of r5.34.
Criterion/Predictive
Evidence for criterion and predictive validity has been provided by Pavot and Diener (1993) , as well as
Morden and Ostiguy (2005) .
Location
Alfonso, V. C., Allison, D. B., Rader, D. E., & Gorman, B. S. (1996). The Extended Satisfaction with Life Scale:
Development and psychometric properties. Social Indicators Research ,38,2 7 5/C0301.
Results and Comments
The ESWLS is one of the few multidimensional measures of life satisfaction for adults that contain multiple
items for each domain. While the research base for the ESWLS is relatively sparse, the findings are promising.
The preliminary results for the ESWLS suggested satisfactory internal consistency and test /C0retest reliability as
well as evidence of validity in a sample of college students. The ESWLS is a measure that can potentially be used
in various age groups, ranging from adolescents to elderly. The authors of the ESWLS indicated that an individ-
ual should have a reading level of at least the seventh grade to be able to understand the ESWLS. However, stud-
ies of the psychometric properties of the measure with different populations (e.g., older adults, psychiatric
patients) are needed to assess its range of applicability.
EXTENDED SATISFACTION WITH LIFE SCALE
Below are some statements with which you may agree
or disagree. Use the scale below to show your agreement
with each item. Place the number on the line for that item.
Please be open and honest in your answers.
1/C0Strongly disagree; 2 /C0Disagree; 3 /C0Slightly dis-
agree; 4 /C0Neither agree nor disagree; 5 /C0Slightly agree;
6/C0Agree; 7 /C0Strongly agree
1.In most ways my life is close to my ideal.
2.The conditions of my life are excellent.
3.I am satisfied with my life.
4.So far I have gotten the important things I want
from life.
5.I am generally pleased with the life I lead.
6.In most ways my social life is close to my ideal.7.The conditions of my social life are excellent.
8.I am satisfied with my social life.
9.So far I have gotten the important things I want
from my social life.
10.I am generally pleased with the social life I lead.
11.In most ways my sex life is close to my ideal.
12.The conditions of my sex life are excellent.
13.I am satisfied with my sex life.
14.So far I have gotten the important things I want
from my sex life.
15.I am generally pleased with the quality of my
sex life.
16.In most ways my actual self is close to my ideal self.
17.As an individual I consider myself excellent.114 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
18.I am satisfied with my person or self as an
individual.
19.So far I have gotten the important things I want
from myself.
20.I am generally pleased with myself as an individual.
21.In most ways my actual physical appearance is close
to my ideal physical appearance.
22.I consider my physical appearance excellent.
23.I am satisfied with my physical appearance.
24.There is nothing about my physical appearance that
I would like to change.
25.I am generally pleased with my physical appearance.
The questions below pertain to your current
‘immediate’ family not your ‘extended’ family.
26.In most ways my family life is close to my ideal.
27.The conditions of my family life are excellent.
28.I am satisfied with my family life.
29.So far I have gotten the important things I want
from my family life.
30.I am generally pleased with the quality of my
family life.
Do you go to school? ○yes○no (if not, skip
the next 5 questions)
31.The education I get at school is great.
32.I like or respect the other students at school.
33.I am satisfied with my classes.
34.So far I have learned the important things I wanted
at school.
35.I am generally pleased with the quality of my
teachers.
Do you have a job? ○yes○no (if not, skip the
next 10 questions)36.The chance for advancement on my job is good.
37.I like the company policies and practices.
38.I like or respect my co-workers.
39.I am pleased with the praise I get for doing a
good job.
40.I am given enough freedom to use my own
judgment.
41.I like the way my job provides for steady
employment.
42.My boss handles his or her employees well.
43.I am happy with the competence of my supervisor.
44.The working conditions of my job are excellent.
45.Overall, I am satisfied with my job.
Are you now in an ‘exclusive’ relationship?
○yes (please answer the questions below based
on your current relationship)
○no, but I have been in the past (please answer
the questions below based on your past
relationship)
○no, and I have not been in the past (you may
stop here)
46.In most ways my relationship/marriage is close to
my ideal.
47.The conditions of my relationship/marriage are
excellent.
48.I am satisfied with my relationship/marriage.
49.So far I have gotten the important things I want
from my relationship/marriage.
50.I am generally pleased with the quality of my
relationship/marriage.
Note: Reproduced with permission.
Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q) and Abbreviated
Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q-18)
(Endicott et al., 1993; Ritsner et al., 2005 )
Variable
These measures are designed to assess the degree of enjoyment and satisfaction experienced by individuals in
different areas of daily functioning ( Endicott et al., 1993 ) in order to measure the health-related quality of life
(QOL; Ritsner et al., 2005 ). The Q-LES-Q and its abbreviated form (Q-LES-Q-18), primarily measure individuals’
judgments of ‘how they feel about what they have, how they are functioning, and their ability to derive pleasure
from their life activities’ ( Endicott et al., 1993 , p. 321).
Description
Q-LES-Q
The Q-LES-Q consists of 93 rationally derived items and comprises a total of eight subscales and two individ-
ual items (satisfaction with medication; overall satisfaction and contentment) with a 5-point Likert-type response
format (ranging from 1 5not at all or never to 55frequently or all of the time ). A subset of 60 items forms the core
of the Q-LES-Q (also labeled as basic Q-LES-Q) and is answered by all participants. This core comprises five115 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
subscales, namely physical health (13 items), subjective feelings (14 items), leisure time activities (6 items), social
relationships (11 items), and general activities (14 items) as well as the two individual items. The three subscales
of work (13 items), household duties (10 items), and school/course work (10 items) are given to participants for
whom those activities are applicable. The General QOL Index is comprised of the average score of all items of the
basic Q-LES-Q. The questionnaire takes between 40 and 45 minutes to administer.
Q-LES-Q-18
The Q-LES-Q-18 consists of 18 items with a 5-point Likert-type response format (ranging from 1 5not at all or
never to 55frequently or all of the time ) and measures four of the five subscales of the basic Q-LES-Q. These sub-
scales are physical health (4 items), subjective feelings (5 items), leisure time activities (3 items), and social rela-
tionships (5 items). The General QOL Index is comprised of the average score of all items of the Q-LES-Q-18.
The completion of the questionnaire takes between 10 and 12 minutes. The items of the basic Q-LES-Q were
selected for the Q-LES-Q-18 in a two-stage process. In the first stage, items were selected that best predicted each
of the subscales of the basic Q-LES-Q. In the second stage, a factor analysis revealed that the subscale of general
activity could not be extracted. Hence, it was excluded from the abbreviated form of the Q-LES-Q.
Sample
Endicott et al. (1993) reported results from 85 outpatients who met the DSM-III-R criteria for major depression.
Their mean age was 39.1 years (ranging from 18 to 63 years) and 59% were female. All participants responded to
the 93 items of the Q-LES-Q.
Ritsner et al. (2005) reported results from 339 inpatients (70.8% male) who met the DSM-IV criteria for schizo-
phrenia ( n5237), schizoaffective, and mood (major depression or bipolar) disorders (total nfor the last two
groups 5102). Their mean age was 38.5 years. All participants responded to the 60 items of the basic Q-LES-Q.
A subset of 199 patients (74.9% male; age: M538.9; n5148 with schizophrenia, n551 with schizoaffective/
mood disorder) completed the basic Q-LES-Q again 16 months after the first assessment.
Furthermore, Ritsner et al. (2005) reported results based on the basic Q-LES-Q of 133 outpatients (76.7% males)
with a mean age of 39.6 years. They all fulfilled the DSM-IV criteria for schizophrenia.
Finally, 175 healthy participants (27.1% male; age: M538.4) completed the basic Q-LES-Q. Ritsner et al. (2005)
also reported results for the Q-LES-Q-18 whose scales were extrapolated from the items of the Q-LES-Q.
Reliability
Internal Consistency
Cronbach alpha coefficients of the Q-LES-Q subscales ranged from α5.90 (general activities) to α5.96 (work)
among outpatients with major depression ( Endicott et al., 1993 ). Among inpatients, alpha coefficients for the basic
Q-LES-Q measure ranged from α5.87 (social relationships) to α5.91 (subjective feelings), while the alpha coeffi-
cient for the General QOL Index wasα5.96 (Ritsner et al., 2005 ).
Ritsner et al. (2005) reported alpha coefficients of the subscales of the Q-LES-Q-18 ranging from α5.76 (social
relationships) to α5.80 (leisure time activities), from α5.82 (subjective feelings) to α5.89 (leisure time activi-
ties), and from α5.74 (leisure time activities) to α5.78 (social relationships), for the inpatients, outpatients, and
healthy sample, respectively. The extrapolated general activity scale yielded high alpha coefficients as well
(i.e., α5.78,α5.85, and α5.71 for the inpatients, outpatients, and healthy sample, respectively). The total score
of the Q-LES-Q-18 yielded alpha coefficients of α5.93,α5.96, and α5.88 in the inpatients, outpatients, and
healthy sample, respectively (cf. Ritsner et al., 2005 ).
Test/C0Retest
Test/C0retest correlations (intraclass correlations /C0ICC; two-week interval) of the Q-LES-Q scores of 54 outpati-
ents with major depression disorder ranged from .63 for leisure activities to .89 for school/course work.
However, actual sample size differed for the subscales (i.e., n56 for school/course work; n539 for work; n552
for household duties; remaining scales N554). Furthermore, test /C0retest reliability for the basic Q-LES-Q (and
the Q-LES-Q-18) was examined in a sample of 33 randomly selected outpatients and 35 healthy individuals who
completed the Q-LES-Q twice within a two-week interval. The ICCs of the General QOL Index were .91 and .90
for the Q-LES-Q and the Q-LES-Q-18 in outpatients, respectively as well as .87 and .86 in healthy subjects, respec-
tively. The subscales showed ICCs ranging from .79 to .87 in the basic Q-LES-Q and from .71 to .83 in the Q-LES-
Q-18 for both the outpatients and the healthy sample.116 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Validity
Convergent/Concurrent
In the inpatient sample, positive correlations between the corresponding scales of the basic Q-LES-Q and its
abbreviated version (Q-LES-Q-18) were r5.97 for the General QOL Index, r5.96 for leisure time activities,
r5.95 for subjective feelings, r5.95 for social relationships, and r5.93 for physical health ( Ritsner et al., 2005 ).
The Quality of Life Scale (QLS; Heinrichs, Hanlon, & Carpenter 1984 ) scores of outpatients significantly corre-
lated with the General QOL Index and the subscales of the basic Q-LES-Q ( r5.55 to r5.64) and the abbreviated
Q-LES-Q-18 ( r5.51 to r5.64;Ritsner et al., 2005 ).
Divergent/Discriminant
Correlational analyses between the Q-LES-Q and various depression measures revealed small to moderate
negative correlations ( Endicott et al., 1993 ). For example, the Q-LES-Q total score correlated negatively with the
Beck Depression Inventory (r52.36;Beck & Beamesderfer, 1974 ). Also, the general index of the Q-LES-Q-18 exhib-
ited negative associations with measures of depression ( r52.29 with the Montgomery and Asberg Depression
Rating Scale ;Montgomery & Asberg, 1979 ), negative symptoms ( r52.19 with the Positive and Negative Syndrome
Scale ;Kay, Fiszbein, & Opler 1987 ), and general psychopathology ( r52.28 with the Positive and Negative
Syndrome Scale )(Ritsner et al., 2005 ).Ritsner et al. (2005) showed that the means of the basic Q-LES-Q (general
activities subscale excluded) and the Q-LES-Q-18 differed between inpatients and healthy participants as well as
between subgroups of the inpatients. Healthy participants always obtained higher scores on the General QOL
Index and subscales than did inpatients. Schizophrenic patients were less satisfied with their social relationships
than patients with a mood/schizoaffective disorder, but did not differ in their levels of satisfaction on any of the
other subscales or the General QOL Index (Ritsner et al., 2005 ).Endicott et al. (1993) reported negative correlations
between the Q-LES-Q and the Clinical Global Impressions Severity of Illness Ratings (NIMH, 1985 ), ranging from
r52.51 (social relationships) to r52.68 (subjective feelings).
Construct/Factor Analytic
Ritsner et al. (2005) presented results of an exploratory principal components analysis performed on the inter-
correlations of the Q-LES-Q-18 items. Four components were extracted (with eigenvalues $1) which were sub-
jected to varimax rotation and subsequently labeled social relationships, physical health, subjective feelings, and
leisure time activities. The hypothesized fifth general activity dimension could not be replicated and was there-
fore not included in the Q-LES-Q-18 measure.
Criterion/Predictive
Predictive validity of the Q-LES-Q and Q-LES-Q-18 measures has been summarized in Ritsner et al. (2005) .
Location
Endicott, J., Nee, J., Harrison, W., & Blumenthal, R. (1993). Quality of Life Enjoyment and Satisfaction
Questionnaire: A new measure. Psychopharmacology Bulletin ,29, 321/C0326.
Ritsner, M., Kurs, R., Gibel, A., Ratner, Y., & Endicott, J. (2005). Validity of an abbreviated Quality of Life
Enjoyment and Satisfaction Questionnaire (Q-LES-Q-18) for schizophrenia, schizoaffective, and mood disorder
patients. Quality of Life Research, 14, 1693/C01703.
Results and Comments
The Q-LES-Q and its abbreviated form (i.e., the Q-LES-Q-18) are multidimensional measures of life satisfaction
for adults that contain multiple items for each domain. Studies of the Q-LES-Q and the Q-LES-Q-18 suggested
satisfactory item homogeneity, test /C0retest reliability, and evidence of validity. These measures appear to have
good usefulness in clinical (i.e., psychiatric) settings, and they are also suitable for use with healthy people.
However, further studies of the psychometric properties of the measures in different populations (e.g., older
adults, psychiatric patients other than schizophrenics, and patients with schizoaffective and mood disorders) are
needed to assess the range of applicability of the Q-LES-Q and the Q-LES-Q-18.117 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
ABBREVIATED QUALITY OF LIFE ENJOYMENT
AND SATISFACTION QUESTIONNAIRE
This questionnaire is designed to help assess the
degree of enjoyment and satisfaction experienced during
the past week.
1/C0Not at all or never; 2 /C0Rarely; 3 /C0Sometimes;
4/C0Often or most of the time; 5 /C0Frequently or all
the time
During the past week how much of the time have you:
1.Felt in at least very good physical health?
2.Been free of worry about your physical health?
3.Felt good physically?
4.Felt full of pep and vitality?
5.Felt satisfied with your life?
6.Felt happy or cheerful?
7.Felt able to communicate with others?
8.Felt able to travel about to get things done when
needed (walk, use car, bus, train, or whatever is
available as needed)?
9.Felt able to take care of yourself?
The following questions refer to leisure time
activities such as watching T.V., reading the paper
or magazines, tending house plants or gardening,
hobbies, going to museums or the movies, or to
sports events, sports, etc.:
10.How often did you enjoy leisure time activities?
11.How often did you concentrate on the leisure
activities and pay attention to them?
12.If a problem arose in your leisure activities, how often
did you solve it or deal with it without undue stress?During the past week how often have you:
13.Looked forward to getting together with friends or
relatives?
14.Enjoyed talking with co-workers or neighbors?
15.Felt affection toward one or more people?
16.Joked or laughed with other people?
17.Felt you met the needs of friends or relatives?
Taking everything into consideration, during the
past week how satisfied have you been with your ...
18.Medication? (if not taking any check here __ and
leave item blank)
19.*Social relationships?
20.*Ability to function in daily life?
21.*Economic status?
22.*Overall sense of well being?
23.*How would you rate your overall life satisfaction
and contentment during the past week?
Notes : Domains: Physical health (items 1 /C04), subjective
feelings (items 5 /C09), leisure time activity (items 10 /C012),
social relationships (items 13 /C017), satisfaction with med-
ication (item 18), general activities (items 19 /C022), life sat-
isfaction (item 23). Domain score is average of scores of
all items in that domain.
*Final Q-LES-Q-18 version does not include last 5 items
(19/C023). General Quality of Life Index 5average of
scores of all 18 items.
Reproduced with permission.
Life Satisfaction Index for the Third Age (LSITA)
(Barrett & Murk, 2006 )
Variable
Barrett and Murk (2006) aimed at measuring the life satisfaction of older people referring to individuals’ eva-
luations of their general feelings of well-being to identify ‘successful’ aging. They defined life satisfaction as a
latent construct consisting of five major components: (a) z est vs. empathy (i.e., enthusiasm of response to life in
general, independent of type of activity); (b) resolution and fortitude (i.e., active acceptance of personal responsibil-
ity for one’s own life); (c) congruence between desired and achieved goals (i.e., believing that major goals have been
achieved); (d) self-concept (i.e., being concerned with one’s appearance, judging oneself as wise and competent,
and not feeling old); and (e) mood tone (i.e., positive affective responses, like happiness and optimism).
Description
The LSITA is a 35-item self-rating questionnaire (zest vs. apathy: 7 items; resolution and fortitude: 9 items;
congruence of goals: 5 items; self-concept: 5 items; mood tone: 9 items) utilizing a 6-point Likert-type response
format (ranging from 1 5strongly disagree to 65strongly agree ). Twenty of the LSITA items were derived from an
earlier measure of life satisfaction in older people ( Life Satisfaction Index , LSI; Neugarten, Havighurst, & Tobin
1961). Additionally, Barrett and Murk expanded the former LSI to include 15 new items that reflected each of the118 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
five postulated components of life satisfaction. A panel of experts provided support for the content validity of
the items. Five domain scores and a total score are created by summing the related item responses (15 items are
reverse keyed).
Sample
Barrett and Murk (2006) reported data from 654 participants, all of whom were 50 years of age or older. These
adults, all within their ‘third age’ (i.e., seniors), were Midwestern US adults acquired at third age learning events,
retirement centers, church events, community centers, and from the general public. In this sample, the five
domains exhibited mean scores for zest vs. apathy ( M530.5, SD55.2), resolution and fortitude ( M539.4,
SD54.8), congruence of goals ( M522.2, SD53.0), self-concept ( M525.2, SD53.3), and mood tone ( M532.9,
SD56.3) as well as the total score ( M5151.0, SD519.5) ( Barrett, 2006 ).
Reliability
Internal Consistency
Barrett and Murk (2006) reported Cronbach alpha coefficients ranging from α5.56 to α5.84 for the five
domains. The LSITA total score yielded an alpha coefficient of α5.93.
Test/C0Retest
No test /C0retest reliability coefficients have been reported to-date.
Validity
Convergent/Concurrent
The LSITA scores correlate positively with the Salamon-Conte Life Satisfaction in the Elderly Scale (SCLSES;
Salamon & Conte, 1984 ). For example, Barrett and Murk (2006) reported a correlation of r5.78 between the total
scores of the LSITA and SCLSES. The corresponding domain scores showed correlations between r5.56 (LSITA
congruence of goals /C0SCLSES goals) and r5.75 (LSITA zest vs. apathy /C0SCLSES daily activities). The correla-
tion between the LSITA and the SWLS ( Diener et al., 1985 ) also exceeded r5.50 (Barrett & Murk, 2006 ).
Divergent/Discriminant
No discriminant validity coefficients have been reported to-date.
Construct/Factor Analytic
A confirmatory factor analysis of the LSITA was carried out at the domain level and showed that the five
domains were well represented in a one-factor solution (e.g., CFI 5.94, NFI 5.94) ( Barrett & Murk, 2006 ).
Loadings ranged from .68 (congruence of goals) to .89 (mood tone).
Criterion/Predictive
Criterion validity evidence for the LSITA has been reported in Barrett and Murk (2006) .
Location
Barrett, A. J., & Murk, P. J. (2006). Life Satisfaction Index for the Third Age (LSITA): A measurement of suc-
cessful aging. In E. P. Isaac (Ed.), Proceedings of the 2006 Midwest Research-to-Practice Conference in Adult,
Continuing, Extension, and Community Education: Impacting adult learners near and far (pp. 7/C012). St. Louis, MO:
University of Missouri, St. Louis.
Results and Comments
Although Barrett and Murk (2006) devised 15 new items that reflected each of the five postulated components
of life satisfaction, the LSITA appears to be a unidimensional measure of life satisfaction for elderly people.
Including the aspects of zest vs. apathy suggests that the LSITA is not strictly a measure of life satisfaction, but
also incorporates affective components. The preliminary results of the LSITA reported by Barrett and Murk
(2006) indicated good item homogeneity for the total LSITA, and the domains (except for the self-concept
domain). Although there is evidence for the convergent validity of the LSITA total and domain subscales in
elderly samples, the research base for the LSITA is limited. Future studies are needed to assess its test /C0retest
reliability, factor structure at the item level, and validity.119 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
LIFE SATISFACTION INDEX FOR THE THIRD AGE
Directions: There are some statements about life in
general that people feel differently about. Please read
each statement on the list and circle the answer that
most closely reflects your attitude toward the statement.
There are no right or wrong answers and your opinion
on each of the statements is important. Thank you for
your confidential participation in this survey.
1.The things I do are as interesting to me as they
ever were.
2.As I grow older, things seem better than I thought
they would be.
3.Everything I have attempted in life has failed. ( /C0)
4.I get respect for the wisdom of my age and
experience.
5.This is the dreariest time of my life. ( /C0)
6.I would enjoy my life more if it were not so dull. ( /C0)
7.Life has not been good to me. ( /C0)
8.I have gotten more of the breaks in life than most of
the people I know.
9.The best of life is behind me. ( /C0)
10.I am just as happy as when I was younger.
11.I enjoy everything that I do.
12.I have been unable to do things right. The deck has
been stacked against me. ( /C0)
13.I achieved in my life what I set out to do.
14.I feel my age, but it does not bother me.
15.I am frequently down in the dumps. ( /C0)
16.I expect interesting and pleasant things to happen to
me in the future.
17.I have made both good and bad choices in my life
and I can live with the results.
18.As I look back on my life I am well satisfied.19.Compared to other people my age, I make a good
appearance.
20.I am appreciated by people who know me.
21.My life is great.
22.I would not change my past life even if I could.
23.When I think back over my life, I didn’t get the
important things I wanted. ( /C0)
24.I feel old and tired. ( /C0)
25.These are the best years of my life.
26.The things that I do are boring or monotonous. ( /C0)
27.Compared to other people my age, I’ve made a lot
of foolish decisions in my life. ( /C0)
28.I have gotten pretty much what I expected out
of life.
29.Everything is just great.
30.My life could be happier than it is now. ( /C0)
31.I have made plans for things I’ll be doing a month
from now.
32.I did it my way.
33.In spite of what people say, the fate of the average
person is getting worse, not better. ( /C0)
34.Compared to other people I often get depressed or
down in the dumps. ( /C0)
35.As I age I get more irritable. ( /C0)
Notes :
(/C0) Reverse keyed item.
Domains: Zest vs. apathy (items 1, 6, 11, 16, 21, 26, and
31), resolution & fortitude (items 2, 7, 12, 17, 22, 27, 28,
32, and 33), congruence of goals (items 3, 8, 13, 18, and
23), self-concept (items 4, 9, 14, 19, and 24), mood tone
(items 5, 10, 15, 20, 25, 29, 30, 34, and 35).
Reproduced with permission.
Students’ Life Satisfaction Scale (SLSS)
(Huebner, 1991b )
Variable
Huebner (1991b) constructed the SLSS for the assessment of satisfaction with life as a whole, based on Diener’s
(1984) conceptualization of global life satisfaction.
Description
The SLSS, a self-report measure for 8 to 18-year-old children and adolescents, is composed of seven items
(two of them are reverse keyed). To facilitate judgments of life overall, items were written to be domain-free in
nature (e.g., ‘I have a good life’ instead of ‘I have a good family life’). The items were designed to be simple and
easy to read, and thus, to be appropriate for students representing a wide range of ages and intellectual abilities.
Whereas in early studies (e.g., Huebner, 1991b ), the SLSS used a 4-point answer format ranging from 1 5never
to 45always , the current form uses a 6-point Likert-type response format ranging from 1 5strongly disagree to
65strongly agree (cf.Huebner, Suldo, & Valois 2005 ).120 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Sample
Huebner (1991b) reported findings for a sample consisting of a total of 329 children aged 8 to 14 years.
Utilizing the 4-point response format, the mean of the SLSS (sum of 7 items) was found to be 20.84 ( SD54.51).
Using a larger sample of 1,188 adolescents, Huebner et al. (2005) reported a mean of the SLSS (averaged over
7 items) of 4.21 ( SD51.14; utilizing the 6-point response format).
Reliability
Internal Consistency
The 7-item SLSS exhibited a Cronbach alpha coefficient of α5.82 ( Huebner, 1991b ).Huebner et al. (2005)
reported alpha coefficients ranging from α5.73 to α5.86 in all age groups (8 /C018 years).
Test/C0Retest
One- to two-weeks test /C0retest reliability of the SLSS was found to be rtt5.74 ( Huebner, 1991b ). Further
research on test /C0retest stabilities yielded coefficients of rtt5.76 (1/C02 weeks), rtt5.64 (one month), rtt5.55
(4 months), rtt5.53 (one year), and rtt5.51 (2 years) (e.g., Huebner et al., 2005 ;Marques, Pais-Ribeiro, & Lopez
2011;Weber, Ruch, & Huebner 2013 ).
Validity
Convergent/Concurrent
The SLSS total score revealed a positive correlation of r5.62 with the Andrews-Withey (1976) life satisfaction
item, as well as correlations with measures of positive affect (e.g., Bradburn’s, 1969 , happiness item, r5.36;
Happiness subscale of the Piers-Harris Children’s Self-Concept Scale ,Piers, 1984 ,r5.53; cf. Huebner, 1991b ).
The SLSS also showed a positive relationship with the Piers-Harris total self-concept score (r5.53;Huebner, 1991b ).
Dew and Huebner (1994) reported a positive correlation ( r5.58) between the SLSS and the Perceived Life
Satisfaction Scale (PLSS; Adelman et al., 1989 ).
Divergent/Discriminant
Huebner (1991b) showed that the SLSS was not significantly associated with a social desirability response set
(r5.05). Huebner et al. (2005) listed several more results on its discriminant validity, for example, the SLSS
showed non-significant relationships with IQ and school grades ( Huebner & Alderman, 1993 ). Additionally, the
SLSS as a measure of the cognitive component of subjective well-being was shown to be distinguishable from the
affective components of subjective well-being utilizing conjoint factor analysis (cf. Huebner et al., 2005 ).Huebner
(1991c) reported a negative correlation between SLSS and a measure of anxiety ( r5/C0.51; Revised Children’s
Manifest Anxiety Scale ,Reynolds & Richmond, 1985 ).
Construct/Factor Analytic
Huebner (1991b) computed a principal components analysis (extraction criterion was eigenvalues .1), and
found that the first unrotated principal component accounted for most of the variance. Unidimensionality of the
SLSS has been reported in several studies (for a review see Huebner et al., 2005 ).
Criterion/Predictive
Predictive validity evidence for the SLSS has been reported by Huebner (1991a) . Also, Huebner et al. (2005,
p. 50) reported that, ‘Predictive validity studies suggest that the SLSS predicts important mental health behaviors
independently and interactively with measures of stressful life events.’
Location
Huebner, E. S. (1991b). Initial development of the Students’ Life Satisfaction Scale. School Psychology
International ,12, 231/C0240.
Results and Comments
Research on the SLSS has provided support for its reliability and validity as a measure of global life satisfac-
tion in children and adolescents aged between 8 and 18 years, covering elementary, middle, and high school stu-
dents. The SLSS has been evaluated in studies in several nations (e.g., Germany, Portugal); however, additional
cross-national studies are needed. Because professionals and researchers often require such a brief measure of121 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
global life satisfaction, the SLSS can be a useful assessment tool for applied professionals (e.g., screening pur-
poses) and researchers in this field (e.g., large-scale assessments). Additional research on its sensitivity to chang-
ing life conditions and interventions would be particularly beneficial.
STUDENTS’ LIFE SATISFACTION SCALE
We would like to know what thoughts about life you
have had during the past several weeks. Think about how
you spend each day and night and then think about how
your life has been during most of this time. Here are
some questions that ask you to indicate your satisfaction
with your overall life. Circle the words next to each state-
ment that indicate the extent to which you agree or dis-
agree with each statement. It is important to know what
you REALLY think, so please answer the questions the
way you really think, not how you should think. This is
NOT a test. There are NO right or wrong answers.
1/C0Strongly disagree; 2 /C0Moderately disagree;
3/C0Mildly disagree; 4 /C0Mildly agree; 5 /C0Moderately
agree; 6 /C0Strongly agree1.My life is going well.
2.My life is just right.
3.I would like to change many things in my life.*
4.I wish I had a different kind of life.*
5.I have a good life.
6.I have what I want in life.
7.My life is better than most kids.
Note: *Reverse keyed item.
Copyright r1994 by the American Psychological
Association. Source: http://artsandsciences.sc.edu/psyc/
faculty/Scott_Huebner (Retrieved May 9, 2014).
Reproduced with permission.
Perceived Life Satisfaction Scale (PLSS)
(Adelman et al., 1989 )
Variable
The Perceived Life Satisfaction Scale (PLSS) was developed within the context of studying psychosocial pro-
blems in young people of ages 8 to 18 years. As most of the prior research in this age group focused on dissatis-
faction with school experiences in particular, the PLSS was developed to expand the range of children’s reports
of their satisfaction or dissatisfaction with respect to additional major aspects of their daily lives (i.e., material
and physical well-being, relationships, environment, personal development and fulfillment, and recreation and
entertainment).
Description
The PLSS consists of 19 items that measure the degree of satisfaction/dissatisfaction with different facets of
minors’ lives. The PLSS utilizes a 6-point response scale that needs to be converted when scoring for dissatisfac-
tion of the respondents’ answers (i.e., low ratings, 1 and 2, are scored as 2; moderate ratings, 3 and 4, are rated as
1; high ratings, 5 and 6, are rated as 0). The PLSS scoring ranges between 0 (low dissatisfaction) and 38 (high dis-
satisfaction) by summing all 19 items. Researchers have also calculated simple sums of items to create a total
score (e.g., ranging from 1 5very dissatisfied to 65very satisfied ), yielding a total score of between 19 and 114,
where higher scores represent higher life satisfaction (e.g., Huebner & Dew, 1993a,b ).
Sample
Four samples have been reported in Adelman et al. (1989) . Samples 1, 2, and 3 consisted of 221, 179, and 68
school students, respectively, with mean ages of 14.5 years, 13.2 years, and 12.9 years, respectively. Sample 1 was
mostly represented by ethnic minorities (70%), while samples 2 and 3 were not (i.e., 16% and 13%, respectively).
The fourth sample was a mental health sample consisting of students who were referred by their schools to a
mental health center for treatment. It consisted of 47 children with a mean age of 11.1 years. About one fourth
(24%) were represented by ethnic minorities. Twenty-five students showed emotionally based pervasive school
behavior problems, 15 students showed underachievement, and seven students showed school avoidance.
The means of the total PLSS were 8.6 ( SD55.9; sample 1), 7.0 ( SD54.1; sample 2), and 7.8 ( SD54.7; sample 3).122 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
The mean of the total PLSS in the mental health sample was 9.7 ( SD56.8), and was higher than all means of
sample 1, 2, and 3 indicating that the mental health sample was more dissatisfied. In sample 1, and in the mental
health sample, the older students were more likely to report higher dissatisfaction than the younger students
(both rs5.33). In the mental health sample, girls scored more highly ( M512.8, SD58.2) than boys ( M58.4,
SD55.7).
Reliability
Internal Consistency
Huebner and Dew (1993a) reported a Cronbach alpha coefficient of α5.89 for the total scale in a sample of
222 students in grades 8 to 12 (mean age of 15.5 years, SD51.5).
Test/C0Retest
Adelman et al. (1989) reported a test /C0retest correlation of rtt5.85, without specifying the time interval.
Validity
Convergent/Concurrent
The PLSS correlated positively ( r5.58) with the Students’ Life Satisfaction Scale (SLSS) ( Dew & Huebner, 1994 ).
Adelman et al. (1989) reported that the PLSS correlated positively with depression levels ( Children’s Depression
Inventory ;Adelman et al., 1989 ) in the mental health sample ( r5.55). Hence, students who reported higher dissat-
isfaction reported more symptoms of depression. Huebner and Dew (1993a) reported that more satisfied students
reported a higher general self-concept ( r5.48; assessed with SDQ-II; Marsh, 1988 ). PLSS scores correlated with
parent-rated life satisfaction, showing a coefficient of r5.42 (Huebner & Dew, 1993a ).Smith, Adelman, Nelson,
Taylor, and Phares (1987) showed that the PLSS is positively correlated ( r5.60) with control satisfaction
(assessed with a single item; cf. Smith et al., 1987 ), but less so ( r5.35) with happiness (assessed with a single
item; cf. Smith et al., 1987 ).
Divergent/Discriminant
The PLSS demonstrated a small inverse correlation of r52.29 with perceived control at school ( Perceived
Control at School Scale ;Adelman et al., 1989 ).
Construct/Factor Analytic
Huebner and Dew (1993b) conducted a principal components analysis with a promax oblique rotation on the
sample of 222 students resulting in a four-component solution showing that the PLSS provides multidimensional
assessment of life satisfaction.
Criterion/Predictive
For psychologically disturbed children, PLSS scores correlated positively ( r5.55) with a child depression measure,
and negatively ( r52.54) with expectations of improvement both at school and at home ( Bender, 1997 ).
Location
Adelman, H. S., Taylor, L., & Nelson, P. (1989). Minor’s dissatisfaction with their life circumstances. Child
Psychiatry and Human Development ,20, 135/C0147.
Results and Comments
Research on the PLSS has provided preliminary information on its reliability and validity as a measure of dis-
satisfaction/satisfaction in children and adolescents (aged 8 to 18 years) studying psychosocial problems in the
general population and individuals with mental health problems. The PLSS is a promising measure. It was
designed to assess general life satisfaction by evaluating satisfaction across a variety of life experiences.
Nevertheless, although there is evidence for the usefulness of the total score, some research suggests that the
PLSS is multidimensional in nature ( Huebner & Dew, 1993b ). Additional research is needed to clarify its psycho-
metric properties, especially its dimensionality and test /C0retest reliability.123 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
PERCEIVED LIFE SATISFACTION SCALE
We are concerned about knowing what you like and
what you dislike. We know that not all students see
things in the same way. We’re going to read some things
to you so you can tell us how satisfied you are with
each of them. How satisfied do you usually feel when
you think about ...
Answer scale from: 1 5Low dissatisfaction to
65High dissatisfaction. Alternatively, answers can be
scored from 1 5Very dissatisfied to 6 5Very satisfied.
1.the amount of spending money you usually have?
2.the amount of time you can spend doing anything
you want?
3.the amount of control you have over your life?
4.going to school?
5.the opportunities you have to learn new things and
improve your skills?
6.your physical appearance, such as your height,
weight, hairstyle?7.your progress at school compared to others in
your classroom?
8.the way you get along with your mother?
9.the way you get along with your father?
10.how physically fit and energetic you are?
11.the amount of time you can spend watching TV?
12.the type of clothes you wear?
13.nonschool activities such as hobbies, sports?
14.the type of neighborhood where you live?
15.the type of place (home, apartment, etc.) where
you live?
16.the way you get along with your friends?
17.the goals you have set for your future?
18.the numbers of friends you have?
19.the type of job you’ll get when you stop going
to school?
Note: Reproduced with permission.
Multidimensional Students’ Life Satisfaction Scale (MSLSS)
(Huebner, 1994 )
Variable
The MSLSS was designed to provide a multidimensional assessment of children’s life satisfaction, for more
focused diagnostic, prevention, and intervention efforts. The MSLSS provides a profile of young people’s satisfac-
tion with important, specific domains (e.g., school, family, and friends) in their lives. Furthermore, a total score
can be calculated to assess general, overall life satisfaction. The MSLSS has been developed for children and ado-
lescents across a wide range of age (grades 3 /C012) and ability levels (e.g., children with mild developmental dis-
abilities through gifted children).
Description
The MSLSS consists of 40 items assessing domain-specific satisfaction with family (7 items), friends (9 items),
school (8 items), living environment (9 items), and self (7 items). Ten of the 40 items are reverse keyed. The items
were designed to be simple and easy to read, and thus, to be adequate for a wide range of age and intellectual
ability. Whereas in early studies (e.g., Huebner, 1994 ), the MSLSS used a 4-point response format ranging from
15never to 45almost always , the current form uses a 6-point Likert-type response format ranging from
15strongly disagree to 65strongly agree (cf.Huebner, 2001 ). As the domains consist of unequal number of items,
the five domain scores and the total score are created by summing the related item responses and dividing by
the number of domain (or total) items.
Sample
Sample 1 ( N5312) consisted of grades 3 /C08 children having a mean age of 10.90 years. Sample 2 ( N5413) con-
sisted of grades 3 /C05 children with a mean age of 8.97 years. Both samples used the 4-point response
format. Sample 1 yielded domain-specific means for the subscales labeled: friends ( M53.31, SD50.57), family
(M53.10, SD50.64), school ( M52.65, SD50.64), self ( M53.13, SD50.63), and living environment ( M53.11,
SD50.62). Sample 2 yielded domain-specific means for the subscales labeled: friends ( M53.30, SD50.56), fam-
ily ( M53.34, SD50.60), school ( M53.02, SD50.68), self ( M53.21, SD50.61), and living environment
(M53.27, SD50.62) (cf. Huebner, 1994 ).124 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Reliability
Internal Consistency
The MSLSS exhibited a Cronbach alpha coefficient of α5.92 for the total scale (in both samples), and alpha
coefficients between α5.82 and α5.85 (sample 1), and between α5.78 and α5.83 (sample 2) for the domain-
specific subscales ( Huebner, 1994 ).
Test/C0Retest
Test/C0retest coefficients for two- and four-week time periods have been reported falling mostly in the rtt5.70
tortt5.90 range (cf. Huebner, 2001 ).
Validity
Convergent/Concurrent
Convergent validity was evaluated through the pattern of correlations between the MSLSS and a self-concept
measure ( Self Description Questionnaire-I ;Marsh, 1990 ). For example, school-related and self-related self-concept
measures were expected to show the highest correlations with the corresponding subscales of the MSLSS
(i.e., school-related satisfaction, and self-related satisfaction), confirmed by correlations of r5.45 and r5.62,
respectively ( Huebner, 1994 ). Also, the friends-related subscale of the MSLSS showed the highest correlation with
another measure of satisfaction with friendships ( r5.56; Children’s Loneliness and Social Dissatisfaction Scale ;
Asher, Hymel, & Renshaw 1984 ), and the school-related subscale of the MSLSS showed the highest correlations
with a measure on quality of school life ( r5.68;Quality of School Life Scale ;Epstein & McPartland, 1977 ).
Divergent/Discriminant
The analyses also provided evidence for the discriminant validity of the MSLSS as lower coefficients with
respect to non-targeted measures were also found (e.g., a correlation between satisfaction with friendships and
school-related self-concept of r5.22). Likewise, the MSLSS friends-related subscale and the peer relations
self-concept measure exhibited a modest correlation of r5.27 (cf. Huebner, 1994 ). Other validity evidence
(e.g., correlations with parents and teachers reports, and measures of social desirability) was summarized by
Huebner (2001) .
Construct/Factor Analytic
In sample 1, a principal components analysis with oblique rotation yielded a five-dimensional solution. This
five-factorial solution explained 49.5% of the total variance. Huebner (1994) reported congruence coefficients of
the solution in sample 1 with a solution computed for sample 2 showing coefficients of between .98 and .99,
indicative of a very high degree of congruence between both solutions.
Criterion/Predictive
Predictive validity of a brief version of the MSLSS was investigated by Huebner, Antaramian, Hills, Lewis,
and Saha (2011) who reported, for example, that the scale is longitudinally predictive of ‘multiple indices of stu-
dents’ engagement in their schooling’ (p. 161).
Location
Huebner, E. S. (1994). Preliminary development and validation of a multidimensional life satisfaction scale for
children. Psychological Assessment ,6, 149/C0158.
Results and Comments
Research has provided considerable evidence for the reliability and validity of the MSLSS with elementary,
middle, and high school students. Children as young as 8 years of age appear to be able to differentiate satisfac-
tion across important life domains. Measures of domain-based life satisfaction are thus expected to complement
global measures, providing more nuanced measures of subjective quality of life for research and practice
purposes (e.g., counseling). Some cross-cultural validation research has been done (e.g., Gilman et al., 2008 ).
Nevertheless, additional validation studies are needed to determine its meaningfulness in other nations.
Recently, an abbreviated version of the MSLSS has been presented ( Huebner, Zullig, & Saha 2012 ), which con-
sists of 30 items, excluding the 10 reverse keyed items from the original MSLSS ( Huebner, 1994 ). Although the
research is limited, this abbreviated version showed a good fit with the expected factor structure and promising125 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
reliability ( Huebner et al., 2012 ). Furthermore, a promising 5-item measure exists, the Brief Multidimensional
Students’ Life Satisfaction Scale (BMSLSS; Seligson, Huebner & Valois 2003 ), which assesses domain-specific satis-
faction with family, friendships, school experiences, self, and living environment with one item for each domain.
MULTIDIMENSIONAL STUDENTS’ LIFE SATISFACTION SCALE
We would like to know what thoughts about life
you’ve had during the past several weeks. Think about
how you spend each day and night and then think about
how your life has been during most of this time. Here
are some questions that ask you to indicate your satis-
faction with life. Circle the number (from 1 to 6) next to
each statement that indicates the extent to which you
agree or disagree with each statement. It is important to
know what you REALLY think, so please answer the
question the way you really feel, not how you think you
should. This is NOT a test. There are NO right or wrong
answers. Your answers will NOT affect your grades, and
no one will be told your answers.
1/C0Strongly disagree; 2 /C0Moderately disagree;
3/C0Mildly disagree; 4 /C0Mildly agree; 5 /C0Moderately
agree; 6 /C0Strongly agree
1.My friends are nice to me.
2.I am fun to be around.
3.I feel bad at school.*
4.I have a bad time with my friends.*
5.There are lots of things I can do well.
6.I learn a lot at school.
7.I like spending time with my parents.
8.My family is better than most.
9.There are many things about school I don’t like.*
10.I think I am good looking.
11.My friends are great.
12.My friends will help me if I need it.
13.I wish I didn’t have to go to school.*
14.I like myself.
15.There are lots of fun things to do where I live.16.My friends treat me well.
17.Most people like me.
18.I enjoy being at home with my family.
19.My family gets along well together.
20.I look forward to going to school.
21.My parents treat me fairly.
22.I like being in school.
23.My friends are mean to me.*
24.I wish I had different friends.*
25.School is interesting.
26.I enjoy school activities.
27.I wish I lived in a different house.*
28.Members of my family talk nicely to one another.
29.I have a lot of fun with my friends.
30.My parents and I do fun things together.
31.I like my neighborhood.
32.I wish I lived somewhere else.*
33.I am a nice person.
34.This town is filled with mean people.*
35.I like to try new things.
36.My family’s house is nice.
37.I like my neighbors.
38.I have enough friends.
39.I wish there were different people in my
neighborhood.*
40.I like where I live.
Notes :
*Reverse keyed item.
Source: http://artsandsciences.sc.edu/psyc/faculty/
Scott_Huebner (Retrieved May 9, 2014).
Reproduced with permission.
FUTURE RESEARCH DIRECTIONS
The breadth and depth of life satisfaction measures available has increased considerably over the last several
decades. A variety of measures now exist that offer reliable methods for assessing a person’s cognitive judgments
of their quality of life for different age groups, different clinical populations, using a variety of formats
(top down vs. bottom up; unidimensional vs. multidimensional; brief vs. comprehensive). Most of these scales
have been published since Andrews and Robinson’s (1991) contribution to the previous edition of this text,
reflecting a significant expansion of measures and corresponding research to the literature base.
The measures reviewed within this chapter vary in the number and quality of reliability and validity studies;
however, all have provided some level of research evidence supporting their use for various purposes, especially
research purposes. Research to date suggests that the measure with the most reliability and validity evidence
across college-age, middle-age, and senior adults of adult global life satisfaction, for use with the general popula-
tion is the SWLS. Multidimensional measures of life satisfaction show promising reliability and validity evidence126 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
(e.g., ESWLS, Q-LES-Q[-18], QLI, PWI, TSWLS), with many offering global and domain-specific information.
Some of these measures were designed to assess specific age groups (e.g., LSITA), specific clinical populations
(e.g., Q-LES-Q[-18]), and/or have been expanded into different versions to provide a more nuanced assessment
of specific clinical populations (e.g., QLI). Many of these measures have been translated into various languages
and used internationally, especially the SWLS. A recent summary of the overall status of life satisfaction measure-
ment research, focusing on adults, has been completed by Diener et al. (2013) .
The development of life satisfaction measures for youth has been a notable area of progress. A number of mea-
sures have been developed to assess life satisfaction for youth aged 8 /C018 years and many of the adult life satis-
faction measures have been adapted for youth (e.g., SWLS-C, PWI-SC). Some initial attention has been given to
the development of measures targeting younger populations (e.g., PWI-PS); however, given the challenge
involved in obtaining subjective perceptions of young children, more research is needed investigating the validity
and utility of such measures. The measurement of subjective well-being in young children (age 8 years and
under) remains a key area of opportunity for research. Youth measures include global and multidimensional
assessments of life satisfaction, and extensive work has demonstrated the reliability and validity of many of these
measures, especially the SLSS and MSLSS ( Proctor, Linley, & Maltby 2009a ). As with adult measures, much work
remains in further establishing the psychometric evidence for many of these measures, especially across cultures
(cf.Casas et al., 2012 ). Recent reviews of youth life satisfaction measures are available ( Huebner & Hills, 2013;
Proctor, Linley, & Maltby 2009b ). Consistent with Andrews and Robinson’s (1991) recommendation, most life sat-
isfaction measures developed in the last two decades incorporate a scale with five or more response categories.
Competing criteria should also be considered that include age and functional capacity of the participants. For a
current review see Diener et al. (2013) .
It is important to understand the different factors found to influence life satisfaction depending on context
(e.g., culture, age) in order to make an informed decision in selecting an appropriate measure. For example,
although the SWLS is an internationally used, empirically supported measure of well-being, the PWI has been
used extensively in Australian populations.
Substantial progress has been made in the development of life satisfaction measures for adults, youths, and
specific clinical populations. Research over the past two decades has increased our understanding of the relation-
ship between life satisfaction and the social and psychological factors that influence satisfaction across the life-
span ( Lyubomirsky et al., 2005; Proctor et al., 2009a ). Further, such research has laid the groundwork for
investigating appropriate causal models of well-being.
The usefulness of life satisfaction measures has gained attention at the macro-applied level of social policy-
making and many countries have expanded their use of these measures in understanding social impact
(International Wellbeing Group, 2013 ). National accounts of subjective well-being (including life satisfaction mea-
sures) for children and adults are being considered and adopted by various nations. This will also provide the
opportunity to investigate the psychometric properties and utility of life satisfaction measures in different con-
texts. Given the widespread international interest in national accounts of well-being, perhaps the greatest need
for future research involves studies of the comparability of reliability, construct validity, and predictive validity
of different life satisfaction measures across different groups, such as nations, cultures, ethnic groups, genders,
and age groups (see Gilman et al., 2008 ). Research into the usefulness of life satisfaction measures in the public
policymaking arena is also essential.
The progress made to-date in the development and validation of measures of life satisfaction is commendable,
and future research expanding the reliability/validity evidence for use of these measures within general and spe-
cific populations will aid in continuing to resolve many of the questions that remain in understanding the nature
and impact of satisfaction across the lifespan.
References
Adams, D. L. (1969). Analysis of a life satisfaction index. Journal of Gerontology ,24, 470/C0474. Available from http://dx.doi.org/doi:10.1093/
geronj/24.4.470 .
Adelman, H. S., Taylor, L., & Nelson, P. (1989). Minor’s dissatisfaction with their life circumstances. Child Psychiatry and Human Development ,
20, 135/C0147. Available from http://dx.doi.org/doi:10.1007/BF00711660 .
Alfonso, V. C. (1995). Measures of quality of life, subjective well-being, and satisfaction with life. In D. A. Allison (Ed.), Handbook of assessment
methods for eating behaviors and weight-related problems: Measures, theory and research (pp. 23 /C080). Thousand Oaks, CA: Sage.
Alfonso, V. C., Allison, D. B., & Dunn, G. M. (1992). Sexual fantasy and satisfaction: A multidimensional analysis of gender differences. Journal
of Psychology and Human Sexuality ,5,1 9/C037. Available from http://dx.doi.org/doi:10.1300/J056v05n03_03 .127 REFERENCES
II. EMOTIONAL DISPOSITIONS |
Alfonso, V. C., Allison, D. B., Rader, D. E., & Gorman, B. S. (1996). The Extended Satisfaction with Life Scale: Development and psychometric
properties. Social Indicators Research ,38, 275/C0301. Available from http://dx.doi.org/doi:10.1007/BF00292049 .
Andrews, F. M., & Robinson, J. P. (1991). Measures of subjective well-being. In J. P. Robinson, P. R. Shaver, & L. W. Wrightsman (Eds.),
Measures of personality and social psychological attitudes (pp. 61 /C0114). New York, NY: Academic Press, Inc.
Andrews, F. M., & Withey, S. B. (1976). Social indicators of well-being . New York, NY: Plenum.
Antaramian, S. P., Huebner, E. S., & Valois, R. F. (2008). Adolescent life satisfaction. Applied Psychology: An International Review ,57, 112/C0126.
Available from http://dx.doi.org/doi:10.1111/j.1464-0597.2008.00357.x .
Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development ,55, 1456/C01464.
Atkinson, M., Zibin, S., & Chuang, H. (1997). Characterizing quality of life among patients with chronic mental illness: A critical examination
of the self-report methodology. The American Journal of Psychiatry ,154,9 9/C0105.
Barrett, A. J. (2006). LSITA scoring matrix. Unpublished document, Ball State University.
Barrett, A. J., & Murk, P. J. (2006). Life Satisfaction Index for the Third Age (LSITA): A measurement of successful aging. In E. P. Isaac (Ed.),
Proceedings of the 2006 Midwest Research-to-Practice Conference in Adult, Continuing, Extension, and Community Education: Impacting adult lear-
ners near and far (pp. 7/C012). St. Louis, MO: University of Missouri-St. Louis.
Beck, A. T., & Beamesderfer, A. (1974). Assessment of depression: The depression inventory. In P. Pichot (Ed.), Modern problems in pharmacopsy-
chiatry (pp. 151 /C0169). Basel, Switzerland: S. Karger.
Bender, T. A. (1997). Assessment of subjective well-being during childhood and adolescence. In G. D. Phye (Ed.), Handbook of classroom assess-
ment: Learning, achievement and adjustment (pp. 199 /C0225). San Diego, CA: Academic.
Bradburn, N. M. (1969). The structure of psychological well-being . Chicago, IL: Aldine.
Campbell, A. (1981). The sense of well-being in America . New York, NY: McGraw-Hill.
Casas, F., Sarriera, J., Abs, P., Coenders, M., Alforo, J., Saforcado, E., et al. (2012). Subjective indicators of personal well-being among adoles-
cents. Performance and results for different scales in Latin-language speaking countries: A contribution to the international debate. Child
Indicators Research ,5,1/C028. Available from http://dx.doi.org/doi:10.1007/s12187-011-9119-1 .
Cattell, R. B. (1978). The scientific use of factor analysis in the behavioral and life sciences . New York, NY: Plenum.
Costa, P. T., Jr., & McCrae, R. R. (1992). Revised NEO Personality lnventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional man-
ual. Odessa, FL: Psychological Assessment Resources, Inc.
Couch, A., & Keniston, K. (1960). Yeasayers and naysayers: Agreeing response set as a personality variable. Journal of Abnormal and Social
Psychology ,60, 151/C0174. Available from http://dx.doi.org/doi:10.1037/h0040372 .
Crowne, D. P., & Marlowe, D. (1964). The approval motive . New York, NY: Wiley.
Cummins, R. A. (1997). Comprehensive Quality of Life Scale /C0Adult Manual (5th ed.). Melbourne, Australia: School of Psychology, Deakin
University.
Cummins, R. A., Eckersley, R., Pallant, J., Van Vugt, J., & Misajon, R. (2003). Developing a national index of subjective wellbeing: The
Australian Unity Wellbeing Index. Social Indicators Research ,64, 159/C0190. Available from http://dx.doi.org/doi:10.1023/A:1024704320683 .
Cummins, R. A., & Lau, A. L. D. (2005a). Personal Wellbeing Index /C0Pre School (PWI-PS) /C0Manual . Melbourne/Hong Kong, Australia/China:
School of Psychology, Deakin University and Department of Rehabilitation Sciences, Hong Kong Polytechnic University.
Cummins, R. A., & Lau, A. L. D. (2005b). Personal Wellbeing Index /C0School Children (PWI-CS) /C0Manual . Melbourne, Australia: Australian
Centre on Quality of Life, School of Psychology, Deakin University.
Cummins, R. A., & Lau, A. L. D. (2005c). Personal Wellbeing Index /C0Intellectual Disability (PWI-ID) /C0Manual . Melbourne, Australia: Australian
Centre on Quality of Life, School of Psychology, Deakin University.
Dew, T., & Huebner, E. S. (1994). Adolescents’ perceived quality of life: An exploratory investigation. Journal of School Psychology ,32, 185/C0199.
Available from http://dx.doi.org/doi:10.1016/0022-4405(94)90010-8 .
Diener, E. (1984). Subjective well-being. Psychological Bulletin ,95, 542/C0575. Available from http://dx.doi.org/doi:10.1037/0033-2909.95.3.542 .
Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social Indicators Research ,31, 103/C0157. Available from http://
dx.doi.org/doi:10.1007/BF01207052 .
Diener, E. (2000). Subjective well-being: The science of happiness and a proposal for a national index. American Psychologist ,55,3 4/C043.
Available from http://dx.doi.org/doi:10.1037/0003-066X.55.1.34 .
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment ,49,7 1/C075.
Available from http://dx.doi.org/doi:10.1207/s15327752jpa4901_13 .
Diener, E., Inglehard, R., & Tay, L. (2013). Theory and validity of life satisfaction scales. Social Indicators Research ,112, 497/C0527. Available from
http://dx.doi.org/doi:10.1007/s11205-012-0076-y .
Diener, E., & Larsen, R. J. (1984). Temporal stability and cross-situational consistency of affective, behavioral, and cognitive responses. Journal
of Personality and Social Psychology ,47, 871/C0883. Available from http://dx.doi.org/doi:10.1037/0022-3514.47.4.871 .
Endicott, J., Nee, J., Harrison, W., & Blumenthal, R. (1993). Quality of Life Enjoyment and Satisfaction Questionnaire: A new measure.
Psychopharmacology Bulletin ,29, 321/C0326.
Epstein, J. C., & McPartland, J. M. (1977). The Quality of School Life Scale . Riverside, CA: Riverside.
Ferrans, C. E., & Powers, M. J. (1985). Quality of Life Index: Development and psychometric properties. Advances in Nursing Science ,8,1 5/C024.
Ferrans, C. E., & Powers, M. J. (1992). Psychometric assessment of the Quality of Life Index. Research in Nursing and Health ,15,2 9/C038.
Fordyce, M. W. (1977). The happiness measure: A sixty-second index of emotional well-being and mental health . Unpublished manuscript, Edison
Community College, Ft. Myers, FL.
Fordyce, M. W. (1978). Prospectus: The self-description inventory . Unpublished paper, Edison Community College, Ft. Myers, FL.
Frisch, M. B. (1992). Use of the Quality of Life Inventory in problem assessment and treatment planning for cognitive therapy of depression.
In A. Freeman, & F. M. Dattlio (Eds.), Comprehensive casebook of cognitive therapy (pp. 27 /C052). New York, NY: Plenum.
Frisch, M. B. (1994). Manual and treatment guide for the Quality of Life Inventory or QOLI . Minneapolis, MN: Pearson.
Frisch, M. B. (2006). Quality of life therapy: Applying a life satisfaction approach to positive psychology and cognitive therapy . New York, NY: Wiley.128 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
Gadermann, A. M., Schonert-Reichl, K. A., & Zumbo, B. D. (2010). Investigating validity evidence of the Satisfaction with Life Scale adapted
for children. Social Indicators Research ,96, 229/C0247. Available from http://dx.doi.org/doi:10.1007/s11205-009-9474-1 .
George, L. K. (1981). Subjective well-being: Conceptual and methodological issues. Annual Review of Gerontology and Geriatrics ,2, 345/C0382.
Gilman, R., & Huebner, E. S. (2003). A review of life satisfaction research with children and adolescents. School Psychology Quarterly ,18,
192/C0205. Available from http://dx.doi.org/doi:10.1521/scpq.18.2.192.21858 .
Gilman, R., Huebner, E. S., Tian, L., Park, N., O’Byrne, J., Schiff, M., et al. (2008). Cross-national adolescent multidimensional life satisfactio n
reports: Analyses of mean scores and response style differences. Journal of Youth and Adolescence ,37, 142/C0154. Available from http://dx.
doi.org/doi:10.1007/s10964-007-9172-8 .
Haywood, K. L., Garratt, A. M., Schmidt, L. J., Mackintosh, A. E., & Fitzpatrick, R. (2004). Health status and quality of life in older people: A struc-
tured review of patient-reported health instruments . National Centres for Health Outcome Developments, Oxford University.
Heinrichs, D. W., Hanlon, T. E., & Carpenter, W. T. (1984). The Quality of Life Scale: An instrument for rating the schizophrenic deficit syn-
drome. Schizophrenia Bulletin ,10, 388/C0398. Available from http://dx.doi.org/doi:10.1093/schbul/10.3.388 .
Hills, P., & Argyle, M. (2002). The Oxford Happiness Questionnaire: A compact scale for the measurement of psychological well-being.
Personality and Individual Differences ,33, 1073/C01082. Available from http://dx.doi.org/doi:10.1016/S0191-8869(01)00213-6 .
Huebner, E. S. (1991a). Further validation of the Student’s Life Satisfaction Scale: Independence of satisfaction and affect ratings. Journal of
Psychoeducational Assessment ,9, 363/C0368. Available from http://dx.doi.org/doi:10.1177/073428299100900408 .
Huebner, E. S. (1991b). Initial development of the Students’ Life Satisfaction Scale. School Psychology International ,12, 231/C0240. Available from
http://dx.doi.org/doi:10.1177/0143034391123010 .
Huebner, E. S. (1991c). Correlates of life satisfaction in children. School Psychology Quarterly ,6, 103/C0111. Available from http://dx.doi.org/
doi:10.1037/h0088805 .
Huebner, E. S. (1994). Preliminary development and validation of a multidimensional life satisfaction scale for children. Psychological
Assessment ,6, 149/C0158. Available from http:/ /dx.doi.org/doi:10.1037/1040-3590.6.2.149 .
Huebner, E. S. (2001). Manual for the Multidimensional Students’ Life Satisfaction Scale . Department of Psychology, University of South Carolina,
Columbia, SC, USA. Available from http:/ /www.psych.sc.edu/faculty/Scott_Huebne .
Huebner, E. S., & Alderman, G. L. (1993). Convergent and discriminant validation of a children’s life satisfaction scale: Its relationship to self-
and teacher-reported psychological problems and school functioning. Social Indicators Research ,30,7 1/C082. Available from http://dx.doi.
org/doi:10.1007/BF01080333 .
Huebner, E. S., Antaramian, S. P., Hills, K. J., Lewis, A. D., & Saha, R. (2011). Stability and predictive validity of the Brief Multidimensional
Students’ Life Satisfaction Scale. Child Indicators Research ,4, 161/C0168. Available from http://dx.doi.org/doi:10.1007/s12187-010-9082-2 .
Huebner, E. S., & Dew, T. (1993a). Validity of the Perceived Life Satisfaction Scale. School Psychology International ,14, 355/C0360. Available from
http://dx.doi.org/doi:10.1177/0143034393144005 .
Huebner, E. S., & Dew, T. (1993b). Is life satisfaction multidimensional? The factor structure of the perceived life satisfaction scale. Journal of
Psychoeducational Assessment ,11, 345/C0350 Available from ,http://dx.doi.org/doi:10.1177/073428299301100405 .
Huebner, E. S., & Hills, K. J. (2013). Assessment of children’s subjective well-being. In D. H. Saklofske, V. Schwean, & C. R. Reynolds (Eds.),
Oxford handbook of psychological assessment (pp. 773 /C0787). Oxford, UK: Oxford University Press.
Huebner, E. S., Suldo, S. M., & Valois, R. F. (2005). Children‘s life satisfaction. In K. A. Moore, & L. H. Lippman (Eds.), What do children need to
flourish: Conceptualizing and measuring indicators of positive development (pp. 41 /C059). New York, NY: Springer.
Huebner, E. S., Zullig, K. J., & Saha, R. (2012). Factor structure and reliability of an abbreviated version of the Multidimensional Students’ Life
Satisfaction Scale. Child Indicators Research ,5, 651/C0657. Available from http://dx.doi.org/doi:10.1007/s12187-012-9140-z .
International Wellbeing Group (2013). Personal wellbeing index (5th Ed., Available from ,http://www.deakin.edu.au/research/acqol/iwbg/
wellbeing-index/index.php .. Melbourne, Australia: Australian Centre on Quality of Life, Deakin University.
Judge, T. A., Thoresen, C. J., Bono, J. E., & Patton, G. K. (2001). The job satisfaction-job performance relationship: A qualitative and quantita-
tive review. Psychological Bulletin ,127, 376/C0407. Available from http://dx.doi.org/doi:10.1037/0033-2909.127.3.376 .
Kay, S. R., Fiszbein, A., & Opler, L. A. (1987). The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophrenia Bulletin ,13,
261/C0276. Available from http://dx.doi.org/doi:10.1093/schbul/13.2.261 .
Lau, A. L. D., & Cummins, R. A. (2005). Test/C0retest reliability of the Personal Wellbeing Index . Unpublished research report, Hong Kong
Polytechnic University.
Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent positive affect: Does happiness lead to success? Psychological Bulletin ,
131, 803/C0855. Available from http://dx.doi.org/doi:10.1037/0033-2909.131.6.803 .
Marques, S. C., Pais-Ribeiro, J. L., & Lopez, S. J. (2011). The role of positive psychology constructs in predicting mental health and academic
achievement in children and adolescents: A two-year longitudinal study. Journal of Happiness Studies ,12, 1049/C01062. Available from
http://dx.doi.org/doi:10.1007/s10902-010-9244-4 .
Marsh, H. W. (1988). Self-Description Questionnaire-II . San Antonio, TX: Psychological Corporation.
Marsh, H. W. (1990). Self-Description Questionnaire-I manual . Campbelltown NSW, Australia: University of Western Sydney.
Montgomery, S. A., & Asberg, M. (1979). A new depression scale designed to be sensitive to change. British Journal of Psychiatry ,134, 382/C0389.
Available from http://dx.doi.org/doi:10.1192/bjp.134.4.382 .
Morden, P., & Ostiguy, L. (2005). Gender differences in the centrality of work, leisure, family and community and models predicting life satisfaction .
Paper presented at the Eleventh Canadian Congress on Leisure Research, Malaspina University-College, Nanaimo, B.C., May 17-20.
Myers, D. G., & Diener, E. (1995). Who is happy? Psychological Science ,6,1 0/C019. Available from http://dx.doi.org/doi:10.1111/j.1467-
9280.1995.tb00298.x .
National Institute of Mental Health (1985). Special feature: Rating scales and assessment instruments for use in pediatric psychopharmacology
research. Psychopharmacology Bulletin ,21, 714/C01124.
Neugarten, B. J., Havighurst, R. J., & Tobin, S. S. (1961). The measurement of life satisfaction. Journal of Gerontology ,16, 134/C0143. Available
from http://dx.doi.org/doi:10.1093/geronj/16.2.134 .129 REFERENCES
II. EMOTIONAL DISPOSITIONS |
Okun, M. A., Kardash, C. A., Stock, W. A., Sandler, I. N., & Baumann, D. J. (1986). Measuring perceptions of the quality of academic life
among college students. Journal of College Student Personnel ,27, 447/C0451.
Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R. Shaver, & L. W. Wrightsman (Eds.), Measures of per-
sonality and social psychological attitudes (pp. 17 /C050). New York, NY: Academic.
Pavot, W., & Diener, E. (1993). Review of the Satisfaction with Life Scale. Psychological Assessment ,5, 164/C0172. Available from http://dx.doi.
org/doi:10.1037/1040-3590.5.2.164 .
Pavot, W., & Diener, E. (2008). The Satisfaction with Life Scale and the emerging construct of life satisfaction. Journal of Positive Psychology ,3,
137/C0152. Available from http://dx.doi.org/doi:10.1080/17439760701756946 .
Pavot, W., Diener, E., & Suh, E. (1998). The Temporal Satisfaction with Life Scale. Journal of Personality Assessment ,70, 340/C0354. Available from
http://dx.doi.org/doi:10.1207/s15327752jpa7002_11 .
Pavot, W., Diener, E., Colvin, C. R., & Sandvik, E. (1991). Further validation of the Satisfaction with Life Scale: Evidence for the cross-method
convergence of well-being measures. Journal of Personality Assessment ,57, 149/C0161. Available from http://dx.doi.org/doi:10.1207/
s15327752jpa5701_17 .
Piers, E. (1984). Piers-Harris Children’s Self-Concept Scale . Los Angeles, CA: Western Psychological Services.
Proctor, C., Linley, P. A., & Maltby, J. (2009a). Youth life satisfaction: A review of the literature. Journal of Happiness Studies ,10, 583/C0630.
Available from http://dx.doi.org/doi:10.1007/s10902-008-9110-9 .
Proctor, C., Linley, P. A., & Maltby, J. (2009b). Youth life satisfaction measures: A review. The Journal of Positive Psychology ,4, 128/C0144.
Available from http://dx.doi.org/doi:10.1080/17439760802650816 .
Renn, D., Pfaffenberger, N., Platter, M., Mitmansgruber, H., Cummins, R. A., & Ho ¨fer, S. (2009). International Well-being Index: The Austrian
Version. Social Indicators Research ,90, 243/C0256. Available from http://dx.doi.org/doi:10.1007/s11205-008-9255-2 .
Reynolds, C. R., & Richmond, B. (1985). Revised Children’s Manifest Anxiety Scale . Los Angeles, CA: Western Psychological Services.
Ritsner, M., Kurs, R., Gibel, A., Ratner, Y., & Endicott, J. (2005). Validity of an abbreviated Quality of Life Enjoyment and Satisfaction
Questionnaire (Q-LES-Q-18) for schizophrenia, schizoaffective, and mood disorder patients. Quality of Life Research ,14, 1693/C01703.
Available from http://dx.doi.org/doi:10.1007/s11136-005-2816-9 .
Rosenberg, M. (1965). Society and the adolescent self-image . Princeton, NJ: Princeton University Press.
Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological well-being revisited. Journal of Personality and Social Psychology ,69,
719/C0727. Available from http://dx.doi.org/doi:10.1037/0022-3514.69.4.719 .
Salamon, M. I., & Conte, V. A. (1984). Manual for the Salamon-Conte Life Satisfaction in the Elderly Scale . Odessa, FL: Psychological Assessment
Resources.
Schalock, R. L., Keith, K. D., Hoffman, K., & Karan, O. C. (1989). Quality of life: Its measurement and use. Mental Retardation ,27,2 5/C031.
Scheier, M. F., & Carver, C. S. (1985). Optimism, coping and health: Assessment and implications of generalized outcome expectancies. Health
Psychology ,4, 219/C0247. Available from http://dx.doi.org/doi:10.1037/0278-6133.4.3.219 .
Schimmack, U. (2008). The structure of subjective well-being. In M. Eid, & R. J. Larsen (Eds.), The science of subjective well-being (pp. 97 /C0123).
New York, NY: Guilford.
Seligson, J. L., Huebner, E. S., & Valois, R. F. (2003). Preliminary validation of the Brief Multidimensional Students’ Life Satisfaction Scale
(BMSLSS). Social Indicators Research ,61, 121/C0145. Available from http://dx.doi.org/doi:10.1023/A:1021326822957 .
Smith, D. C., Adelman, H. S., Nelson, P., Taylor, L., & Phares, V. (1987). Students’ perception of control at school and problem behavior and
attitudes. Journal of School Psychology ,25, 167/C0176. Available from http://dx.doi.org/doi:10.1016/0022-4405(87)90026-4 .
Weber, M., Ruch, W., & Huebner, E. S. (2013). Adaptation and initial validation of the German version of the Students’ Life Satisfaction Scale
(German SLSS). European Journal of Psychological Assessment ,29, 105/C0112. Available from http://dx.doi.org/doi:10.1027/1015-5759/
a000133 .
Weiss, D. J., Dawis, R. V., England, G. W., & Lofquist, L. H. (1967). Manual of the Minnesota Satisfaction Questionnaire . University of Minnesota,
MN: Minnesota Studies in Vocational Rehabilitation.130 5. MEASURES OF LIFE SATISFACTION ACROSS THE LIFESPAN
II. EMOTIONAL DISPOSITIONS |
CHAPTER
6
Measures of Self-Esteem
M. Brent Donnellan1, Kali H. T rzesniewski2and Richard W . Robins3
1Michigan State University, East Lansing, MI, USA;2University of California, Davis, CA, USA;
3University of California, Davis, CA, USA
Self-esteem is one of the most widely studied constructs in the social sciences as it unites basic and applied
researchers from psychology to education to sociology (see e.g., Donnellan, Trzesniewski, & Robins, 2011 ;
Robins, Tracy, & Trzesniewski, 2008a ;Robins, Trzesniewski, & Schriber, 2008b ;Zeigler-Hill, 2013 ). It is also one
of the oldest constructs in scientific psychology as it was first described by William James (see e.g., 1985/1892).
In light of these two facts, it is perhaps unsurprising tha t self-esteem is one of the more controversial constructs
in literature. There are ongoing debates about the caus al role of self-esteem for life outcomes, the degree to
which self-esteem is a cultural universal as opposed to a construct limited to Western/individualistic cultures,
as well as disagreement over whether self-esteem is more trait-like or state-like (see Donnellan et al., 2011 ). In
light of these debates, the assessment of self-esteem is an important and even critic al issue because measure-
ment lies at the heart of empirical research. According ly, the focus of this chapter is to update and extend pre-
vious work detailing the assessment of self-esteem by Blascovich and Tomaka (1991) and Heatherton and
Wyland (2003) .
James (1985/1892) specified that self-esteem is ‘determined by the ratio of our actualities toour supposed
potentialities ’ (italics added, p. 54). Subsequent treatments emphasize that self-esteem involves feelings of self-
acceptance and self-respect (e.g., Rosenberg, 1989 ). In a nutshell, self-esteem is the subjective assessment of
one’s worth as a person. If someone feels good about herself then she is said to have high self-esteem, whereas
someone who feels badly about herself is said to have lo w self-esteem. These self-eva luations need not be tied
to objective standards because the critical element o f the construct is the subjective evaluation. Self-
evaluations can occur with respect to specific domain s, such as athletics and physical appearance, or at the
general level in terms of an overarching evaluation of t h es e l fa saw h o l e .T h ef o r m e rc a s er e f e r st od o m a i n
specific self-evaluations (or domain specific self-est eem) whereas the latter refers to global or general self-
esteem.
Given the subjective nature of the construct, self-este em is typically assessed with self-report scales. There
are apparently many different self-esteem measures and one review article even suggested a figure of 200
(Sheff & Fearon, 2004 ). However, most of these measures are used in frequently and the current research litera-
ture in mainstream journals largely consists of studies based on a relatively small set of commonly used instru-
ments. Indeed, the previous version of this chapter by Blascovich and Tomaka (1991) reviewed just 11
measures of self-esteem and closely related constructs . They concluded that the Ros enberg Self-Esteem scale
(RSE; e.g., Rosenberg, 1989 ) was the most popular measure of the construct as it accounted for at least 25% of
all citations to self-esteem measures in major journa ls. The next closest measure accounted for around 18% of
citations, whereas none of the other measures accounted for more than 10% of the total citations to self-esteem
measures. In other words, a small set of measures accou nted for the majority of citations related to the assess-
ment of self-esteem.
We conducted an analysis similar to the one performed by Blascovich and Tomaka (1991) to estimate the
prevalence of citations to different self-esteem measures. Results are displayed in Table 6.1 . The continued
131Measures of Personality and Social Psychological Constructs.
DOI: http://dx.doi.org/10.1016/B978-0-12-386915-9.00006-1 ©2015 Elsevier Inc. All rights reserved. |
prominence of the Rosenberg scale continues through 2 013. We suspect that for many researchers, the RSE is
the standard measure of global self-esteem. In light of the importance of this scale, we will summarize recent
work on this measure, in addition to covering several other measures that were either not yet developed
(e.g., the single item self-esteem scale; Robins, Hendin, & Trzesniewski, 2001a ;R o b i n s ,T r a c y ,T r z e s n i e w s k i ,
Potter, & Gosling, 2001b ) or not well-covered by the Blascovich and Tomaka (1991) chapter, such as the Harter
scales (e.g., Harter & Pike, 1984 ) and the Self-Description Questionnaires (e.g., Marsh, 1992a,b ;s e ea l s o Boyle,
1994 and Byrne, 1996 ). Readers interested in reviews of measures not reported here should consult Blascovich
and Tomaka (1991) orHeatherton and Wyland (2003) . In total, we review the five measures listed below.
The first three are global self-esteem measures and th e last two are domain-specific inventories that also
include a general or global scale.
TABLE 6.1 Commonly Used Measures of Self-Esteem and Self-Competence
Measure Citation Approximate age Cites (%) Past 10-yr
cites (%)% Last edition
(1991)
Rosenberg Self-Esteem
scaleRosenberg (1989) adolescence and older 18,216(49%) 11,600 (53%) 25%
Self-Perception Profile Harter /C0 8,672 (23%) 5,065 (23%) /C0
Early Child Harter and Pike (1984) 5 to 7 1068 612 /C0
Child Harter (1982/1985) 8 to 12 5939 3336 /C0
Adolescence (SDQ II) Harter (1988) 13 to 18 1312 895 /C0
College Students Neemann and
Harter (1986)college aged 210 133 /C0
Adults Messer and Harter (1986) 20 to 60 143 89 /C0
Self-Esteem Inventory Coopersmith (1967) middle childhood
and older5,448 (15%) 1,970 (9%) 18%
Single-Item Self-Esteem Robins et al. (2001a) middle childhood
and older697 (2%) 672 (3%) /C0
Janis-Field Feelings
of InadequacyEagly (1967) ;Fleming and
Courtney (1984) ;Janis and
Field (1959)adolescence and older 1,320 (4%) 576 (3%) 5%
Self-Description
QuestionnaireMarsh /C0 1,775 (5%) 1,199 (6%)
Young Child Marsh, Craven, and Debus
(1991) ;Marsh and
Holmes (1990)5 to 8 399 289 /C0
Preadolescence (SDQ I) Marsh, Bames, Cairns, &
Tidman (1984) ;Marsh (1988)8 to 12 657 408 /C0
Adolescence (SDQ II) Marsh (1988) ;Marsh, Ellis,
Parada, Richards, and
Heubeck (2005) ;
Marsh (1992b)12 to 18 107 107 1%
Late Adolescence
(SDQ III)Marsh (1992c) ; Marsh and
O’Neill (1984)18 to 24 612 395
Self-Liking Self-
CompetenceTafarodi and Swann
(1995/2001)adolescence and older 596 (2%) 522 (2%) /C0
Texas Social Behavior
InventoryHelmreich and Stapp (1974) adolescence and older 362 (1%) 106 (0%) 3%
Total Citations /C0 37,086 21,710
Note: Citations based on scholar.google.com on 06-22-2013 (past 10-year search restricted to 2003 /C02013).132 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
MEASURES REVIEWED HERE
1.The Rosenberg Self-Esteem Scale ( Rosenberg, 1989 )
2.The Single-Item Self-Esteem Scale ( Robins et al., 2001a )
3.Self-Liking and Self-Competence Questionnaires (e.g., Tafarodi & Swann, 1995 )
4.The series of Harter Self-Perception Profiles (e.g., Harter & Pike, 1988)
5.The series of Self-Description Questionnaires (e.g., Marsh, 1992a )
OVER VIEW OF THE MEASURES
Our goal in this chapter is to help researchers make informed decisions when selecting measures of self-
esteem for their own research and when evaluating the existing literature. Thus, the current chapter reviews
some of the most widely used self-report measures in the literature. We first provide a review of the Rosenberg
Self-Esteem Scale (RSE) /C0(Rosenberg, 1989 ) with material updated since the Blascovich and Tomaka (1991) chap-
ter, given the ongoing popularity of this instrument. In particular, we summarize new data on the correlates of
the RSE and describe debates about the factor structure of this instrument. We then describe the psychometric
properties of a popular single-item self-esteem measure, the Single-Item Self-Esteem Scale (SISE; Robins et al.,
2001a ). This measure was designed to approximate the RSE and is useful when researchers have very limited
time to assess global self-esteem. We then describe the Self-Liking and Self-Competence Scale /C0Revised (SLSC-R;
Tafarodi & Swann, 1995, 2001 ). This instrument provides a way to separately assess perceptions of self-worth
(i.e., self-liking) from perceptions of personal efficacy and self-regard of one’s capabilities (i.e., self-competence).
We conclude the chapter by reviewing two of the mostly widely used multiple domain inventories /C0the series
of Self-Perception Profiles developed by Harter (e.g., Harter & Pike, 1988) and the series of Self-Description
Questionnaires developed by Marsh (e.g., Marsh, 1992a,b ; see also the comprehensive reviews by Boyle, 1994 ;
andByrne, 1996 ). Both of these measures include an overall global scale as well as scales designed to assess per-
ceptions of self-worth relevant for individuals at different developmental periods (e.g., childhood, adolescence,
and young adulthood).
We will not cover measures of implicit self-esteem in this chapter. Implicit self-esteem refers either to: (1) an
automatic and preconscious self-evaluation that is distinct from explicit self-esteem (i.e., the construct assessed
by self-report measures); or (2) feelings of self-worth that individuals are unwilling or unable to disclose on self-
report measures (see Buhrmester, Blanton, & Swann, 2011 ). The first account suggests two distinct constructs /C0
implicit versus explicit self-esteem, whereas the second account of implicit self-esteem suggests the need for
alternative measurement techniques to self-report scales. The two prominent measures of implicit self-esteem
are the Implicit Association Test (IAT; e.g., Greenwald & Farnham, 2000 ) and the Name-Letter Test (NLT;
Nuttin, 1985 ; see also Koole, Dijksterhuis, & van Knippenberg, 2001 ). These two scales are only weakly corre-
lated with each other ( r5.08 from a meta-analysis of nine studies in Buhrmester et al., 2011 ) and the fact that
the two implicit measures do not strongly converge seems problematic because it suggests that the two implicit
measures do not capture the same underlying attribute. Likewise, questions have been raised about the
criterion-related validity of implicit self-esteem measures ( Buhrmester et al., 2011 ). Moreover, neither of these
two measures are strongly correlated with explicit measures of self-esteem (e.g., the meta-analytic correlation
between the IAT and the RSE was .13 based on 11 studies summarized in Hofmann, Gawronski, Gschwendner,
Le, & Schmitt, 2005 ) and the meta-analytic correlation between the Name-Letter Test and the RSE was only .12
[based on 19 studies summarized in Krizan & Suls, 2008 ] (see Buhrmester et al., 2011 ). Thus, we believe it
would be premature to review implicit measures in light of the ongoing issues in the literature. Additional
details about the measurement of implicit self-esteem can be found in Buhrmester et al. (2011) and Bosson,
Swann, and Pennebaker (2000) . A quick and accessible primer on the IAT is found in Robins et al. (2008a,b) .
We return to the importance of developing implicit measures of self-esteem in our final discussion of future
research directions.
Rosenberg Self-Esteem Scale (RSE)
(Rosenberg, 1989 ).133 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Variable
Rosenberg (1989) defined global self-esteem as the overall attitude one holds about oneself, ranging from nega-
tive to positive. High self-esteem entails the belief that one is ‘good enough’ whereas low self-esteem is associated
with self-rejection and a general lack of self-respect.
Description
According to Blascovich and Tomaka (1991, p. 121) :
The 10 items that make up the Rosenberg Self-Esteem Scale [RSE] were designed to optimize ease of administration, economy of time,
unidimensionality, and face validity. Self-Esteem Scale items require the respondent to report feelings about the self directly. Although
originally designed as a Guttman-type scale, the [RSE] is typically scored using a four-point response format (strongly agree, agree, dis-
agree, strongly disagree) resulting in a scale range of 10 /C040 with higher scores representing higher self-esteem. Some authors, however,
have adopted more familiar Likert-type response formats employing 5- or 7-point scales resulting in broader ranges of [RSE] scores.
While contemporary investigators typically use the 10 Rosenberg items with a 4-point rating scale ( Heatherton &
Wyland, 2003 ), it is also common for researchers to add a neutral response option and thus use a 5-point Likert-type
response scale (e.g., Donnellan, Kenny, Trzesniewski, Lucas, & Conger, 2012 ;Gray-Little, Williams, & Hancock,
1997). Other response options have also been reported in the literature (see discussion pertaining to number of
response scale points in Chapter 1, this volume).
Sample
Details of the original sample ( N55,024) used in the construction of the RSE were provided in Blascovich
and Tomaka (1991, p. 121) . Subsequently, Schmitt and Allik (2005) investigated the psychometric properties of
the RSE in several countries using samples ranging from N559 (Cyprus) to N52,782 (United States) with a
combined sample of N516,998 (see below). Sinclair et al. (2010) also administered the RSE online to a sample
ofN5503 respondents (see below). Even larger combined samples have been reported in the literature
(e.g., N520,332; Rosenthal, Matthew Montoya, Ridings, Rieck, & Hooley, 2011 , see below).
Reliability
Internal Consistency
Cronbach alpha coefficients for the RSE are usually above .80 and values above .90 have been reported in the
literature ( Heatherton & Wyland, 2003 ). For example, Gray-Little et al. (1997) reported an alpha coefficient of .88
for a sample of 1,234 college students. Likewise, Zeigler-Hill (2010) reported an alpha coefficient of .88 for a sam-
ple of 1,422 university students. Schmitt and Allik (2005) administered the RSE to participants from 53 countries
and reported alpha coefficients ranging from .45 (Democratic Republic of the Congo; N5183) to .90 (Israel and
the United Kingdom; N5389 and N5480, respectively). The average alpha coefficient was .81 (total
N516,998). Sinclair et al. (2010) collected data from an internet sample designed to match the population of the
United States and reported an alpha coefficient of .91 ( N5503).
Test/C0Retest
Ackerman and Donnellan (2013) reported a two-week test /C0retest reliability coefficient of .80 ( N5143 univer-
sity students). Donnellan and McAdams (2013) reported a stability coefficient of .69 for a sample of college stu-
dents assessed initially in the middle of their first semester and then again in the middle of their second semester
(N5347). Donnellan et al. (2012) reported even higher stability coefficients for individuals aged in their early
30s assessed over a two-year span ( r5.77,N5399). Recently, Kuster and Orth (2013) suggested that the long-
term stability coefficient for items from the RSE is around .40 for intervals approaching up to 30 years, based on
a sample of 3,180 participants. All of these findings suggest that the RSE captures variance related to a reasonably
stable individual difference construct (cf. Donnellan et al., 2012 ;Trzesniewski, Donnellan, & Robins, 2003 ).
Validity
Convergent/Concurrent
Zeigler-Hill (2010) reported a positive correlation of .71 with the State Self-Esteem Scale ( Heatherton & Polivy,
1991). He also reported that the RSE was strongly associated with both the Self-Liking ( r5.90) and Self-
Competence ( r5.71) scales developed by Tafarodi and Swann (2001) . The RSE correlated with the Harter global
subscale .74 and .75 ( Donnellan, Trzesniewski, Conger, & Conger, 2007 ; Ferrier et al., 2008). Buhrmester et al. (2011)134 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
suggested that the RSE converges so highly with other global measures of self-esteem like the Self-Liking and Self-
Competence scales from Tafarodi and Swann (2001) that they can be combined into a single composite variable cap-
turing explicit self-esteem ( Buhrmester et al., 2011 , p. 369). Thus, there is good evidence that several of the well-
used global measures of self-esteem in the literature are highly interrelated. The RSE also correlates positively with
informant reports of life-satisfaction and pleasant affect ( r5.42 and r5.39,N5141; Schimmack & Diener,
2003). Likewise, Buhrmester et al. (2011) reported that rater evaluations of global self-esteem are associated with
explicit self-esteem ( r5.45). Self-esteem is positively correlated with psychological well-being and life satisfaction.
These coefficients range from .30 to .60 or higher (e.g., Buhrmester et al., 2011 ;Robins et al., 2001a ;Rosenthal et al.,
2011). For example, Sinclair et al. (2010) reported that the RSE correlates .51 with the mental health component of
the SF-8 Health Survey ( Ware, Kosinski, Dewey, & Gandek, 2001 ). Scores on the RSE correlate positively with a
measure of extraversion (e.g., r5.38;Robins et al., 2001a ;Schmitt & Allik, 2005 ;Watson, Suls, & Haig, 2002 )a n d
this association remains relatively unchanged when controlling for social desirability ( Robins et al., 2001b ).
Divergent/Discriminant
RSE scores correlate negatively with measures of depression. For example, Sinclair et al. (2010) reported that the
RSE correlated 2.62 with the 7-item depression scale from the Depression Anxiety Stress scales ( Lovibond &
Lovibond, 1995 ). Likewise, scores on the Rosenberg are negatively correlated with measures of neuroticism
(e.g., r52 .50; Robins et al., 2001a ; see also Judge, Erez, Thoresen, & Bono, 2002 ;Rosenthal et al., 2011 ;
Watson et al., 2002 ). Although the causal relationship between self-esteem and aggression is contentious
(e.g., Baumeister, Campbell, Krueger, & Vohs, 2003 ;Donnellan, Trzesniewski, Robins, Moffitt, & Caspi, 2005 ), scores
on the RSE are negatively correlated with self-reports of aggression and anger (meta-analytic r52 .29 based on 8
studies; N520,332; Rosenthal et al., 2011 ). While RSE scores correlate positively with some global measures of nar-
cissism ( #.30) ( Trzesniewski, Donnellan, & Robins, 2008 ); importantly, the RSE exhibits a small negative correlation
with elements of narcissism related to entitlement and exploitativeness (meta-analytic r52 11,Rosenthal et al.,
2011;s e ea l s o Ackerman & Donnellan, 2013 ). One interpretation is that self-esteem is more strongly related to adap-
tive rather than maladaptive components of narcissism. This is consistent with Rosenberg’s contention that global
self-esteem is distinct from the sense that one feels better than others (see Donnellan et al., 2005 ;2011). Also, scores
on the RSE are largely distinct from academic outcomes and socioeconomic status ( Blascovich & Tomaka, 1991 ).
Construct Factor Analytic
There is ongoing discussion about the underlying factor structure of the RSE with proposed solutions ranging
from one to five factors (see e.g., Corwyn, 2000 ;DiStefano & Motl, 2006 ;Gana et al., 2013 ;Greenberger, Chen,
Dmitrieva, & Farruggia, 2003 ;Marsh, 1996 ;Marsh, Scalas, & Nagengast, 2010 ;Quilty, Oakman, & Risko, 2006 ;
Supple, Su, Plunkett, Peterson, & Bush, 2012 ;Tafarodi & Milne, 2002 ). For example, Tafarodi and Milne (2002)
reported a 5-factor model using data from 836 Canadian university students, whereas Quilty et al. (2006) found
support for a 3-factor model using data from 503 Canadian university students and another sample of 501 adults
from Oregon. DiStefano and Motl (2006) also found support for a 3-factor model using data from 757 university
students from the United States. The basic concern is whether the RSE assesses a unidimensional construct or a
number of conceptually distinct latent factors. This issue is important given that the vast majority of research
with the RSE is based on composite scores which are basically the sums (or averages) of the responses to the
10 items. This strategy may be inappropriate if the items assess multiple substantive factors. The Rosenberg
Self-Esteem Scale table below provides a key to the major solutions for the RSE.
While this is currently a contentious area of psychometric research, at least two consistent results are emerging
from this literature. First, the 5-factor model proposed by Tafarodi and Milne (2002) is difficult to fit to real data
(Marsh et al., 2010 ;N52,213 10th graders from the United States). Tafarodi and Milne fit a model that included
five latent factors labeled: global self-esteem, a positive method factor, a negative method factor, an assessment
factor, and an acceptance factor. Assessment refers to items that involve content related to self-evaluation whereas
acceptance refers to items with content that appear to capture feelings of self-acceptance ( Tafarodi & Milne, 2002 ,
p. 448). These two latent factors correspond to the conceptual attributes of self-competence and self-liking, respec-
tively (see the discussion of the Self-Liking/Self-Competence Scale). Despite any conceptual appeal of these two
dimensions, the psychometric issue is that the 5-factor model for the RSE proposed by Tafarodi and Milne does
not appear to replicate well according to the published literature.
Second, models that specify one general factor and two method factors capturing the positively keyed and nega-
tively keyed items tend to fit the observed data better than alternatives such as a single-factor model or a 2-factor
model ( DiStefano & Motl, 2006; Gana et al., 2013; Marsh et al., 2010; Quilty et al., 2006 ). Specifically, there is emerging135 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
evidence from factor analytic studies showing a single substantive general global factor and two seemingly
stable response style tendencies, one capturing variance associated with the negatively keyed items and a second cap-
turing variance associated with positively keyed items. However, research also suggests that a single global self-
esteem factor explains a considerable amount of variance in the RSE items (e.g., Donnellan et al., 2012 ;Schmitt &
Allik, 2005 ). Indeed, Tafarodi and Milne (2002) noted that a ‘common factor accounted for the lion’s share of reliable
variance across the [RSE] items’ (p. 456) even when they could fit a complicated 5-factor model. This may suggest that
the RSE items assess an essentially unidimensional construct (see Slocum-Gori, Zumbo, Michalos, & Diener, 2009 ).
[To facilitate additional research, we have prepared an Mplus script (that specifies various factor models for the RSE)
that can be obtained upon request. This script assumes that the ordering of the RSE items follows the order presented
in the Blascovich and Tomaka (1991) chapter (see their p. 123) and the item order listed in this chapter.]
Criterion/Predictive
Several longitudinal studies have evaluated the predictive validity of scores on the RSE or scores on a subset
of items from the RSE. For example, Trzesniewski et al. (2006) found that z-scores on the RSE (averaged across
ages 11, 13, and 15 years) predicted adult outcomes including mental health problems, such as major depressive
disorder (Odds Ratio 51.21) and adult anxiety disorder (Odds Ratio 51.45), as well as adult criminal convic-
tions for violent offenses (Odds Ratio 51.25). Trzesniewski et al. also found that low adolescent self-esteem pre-
dicted informant-reported work problems ( β5.13). Many of the findings of Trzesniewski et al. have held up in
more recent analyses using cross-lagged modeling strategies that also control for prior levels of an outcome when
considering predictive statistical effects for the RSE. For example, Kuster, Orth, & Meier (2013) found that self-
esteem predicted future job satisfaction, controlling for previous levels (e.g. β5.13). Self-esteem also predicted
future employment in a sample of 600 individuals (e.g. β5.17). Moreover, the results of Kuster et al. suggest
that self-esteem was not consistently predicted by job conditions while controlling for previous levels of self-
esteem (see also Orth, Robins, & Widaman, 2012 ). This finding places constraints on inferences about the strength
of reciprocal relations between self-esteem and job conditions. In addition, scores on RSE positively predicted
future levels of self-reported mental health such as depression ( β52 .20) and negative affect ( β52 .13) as
well as self-reported physical health ( β52 .11) while controlling for prior levels using a data from over 1,800
participants enrolled in a long-term longitudinal study of multiple generations ( Orth et al., 2012 ). These authors
also found that self-esteem predicted future job satisfaction ( β5.14). Thus, Orth et al. provided compelling evi-
dence for the predictive validity of scores on the RSE for important life outcomes.
The prospective effect of self-esteem on future depression has been replicated several times in reports from
Orth, Robins, and Roberts (2008) ;Orth, Robins, and Meier, (2009a); Orth, Robins, Trzesniewski, Maes, and
Schmitt (2009b) , and Orth et al. (2009a,b) . Indeed, a meta-analysis of 77 studies linking self-esteem and depres-
sion suggested the same prospective effect of self-esteem on future depression ( Sowislo & Orth, 2013 ; meta-
analytic estimate was β52 .16). Self-esteem is also prospectively linked with anxiety and there are some
indications that this association is reciprocal ( Sowislo & Orth, 2013 ). The prospective meta-analytic effect captur-
ing how strongly self-esteem predicted anxiety controlling for prior levels (i.e., a standardized regression coeffi-
cient) was 2.10 whereas the prospective effect for anxiety predicting self-esteem was 2.08. The major caveat is
that effect sizes tend to be modest (especially when controlling for prior levels of criterion-variables), a result that
is perfectly consistent with the idea that single individual differences cannot have large effects on multiply deter-
mined outcomes (see Ahadi & Diener, 1989 ).
Finally, the RSE correlates positively with the Marlowe /C0Crowne Social Desirability Scale ( Crowne & Marlowe,
1960)(r5.22 based on data from over 7,000 college students; Trzesniewski et al., 2008 ). However, there is little
evidence that controlling for social desirability or impression management substantially alters the criterion-
related validity of measures of personality constructs (see Barrick & Mount, 1996 ;Li & Bagger, 2006 ) or the RSE
in particular ( Moorman & Podsakoff, 1992 ).
Location
Rosenberg, M. (1989). Society and the adolescent self-image (Revised edition). Middletown, CT: Wesleyan
University Press.
Results and Comments
The RSE is still the most widely used measure of self-esteem. It appears to generate reliable scores that have
predictive validity. Nonetheless there are concerns about the RSE. For example, Byrne (1996) raised concerns
about the dimensionality of the RSE and we noted this is an ongoing area of research. Moreover, relatively little
research has been conducted to establish the nomological network of the positive and negative method factors136 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
(cf.DiStefano & Motl, 2006 ;Quilty et al., 2006 ). These method factors are stable over time (e.g., see Gana et al.,
2013; Marsh, 2010), but much more work on the importance of the RSE method factors is needed. Byrne (1996)
raised a second concern about focusing on global self-esteem as opposed to multiple dimensions of self-esteem
(cf.Marsh & Craven, 2006 ). This is a concern we address in more detail on the sections describing multidimen-
sional self-esteem inventories. A final issue with the RSE concerns the limitations of subjective self-reports.
Individuals may respond carelessly to survey items or they may distort their responses due to inadequate self-
insight, or for reasons of self-deception or self-enhancement (see Boyle, 1994 , p. 650). This possibility motivates
the search for implicit self-esteem measures designed to gain ‘a reliable window into what people think about
themselves but cannot or will not report’ ( Buhrmester et al., 2011 , p. 377).
ROSENBERG SELF-ESTEEM SCALE
Instructions: Below is a list of statements dealing with your general feelings about yourself. Please indicate the
extent to which you agree or disagree with each statement.
Item Method
factorTafarodi & Milne
(2002) dimension
1 I feel that I’m a person of worth, at least on an equal plane with
others. ( Rosenberg, 1989 )/C0OR/C0Positive AS*
I feel that I am a person of worth, at least on an equal basis with
others. ( Blascovich & Tomaka, 1991 )
2 I feel that I have a number of good qualities. Positive AS
3 All in all, I am inclined to feel that I am a failure. Negative AS*
4 I am able to do things as well as most other people. Positive AS
5 I feel I do not have much to be proud of. Negative AS*
6 I take a positive attitude toward myself. Positive AC
7 On the whole, I am satisfied with myself. Positive AC
8 I wish I could have more respect for myself. Negative AC
9 I certainly feel useless at times. Negative AC
10 At times I think I am no good at all. Negative AC
Notes :
AS5Assessment (Self-Competence); AC 5Acceptance (Self-Liking). *Indicates a non-significant (and thereby
weak) factor loading in the Tafarodi and Milne (2002) analysis.
For the original instructions see www.socy.umd.edu/quick-links/rosenberg-self-esteem-scale (Retrieved January
6, 2014).
The RSE was originally designed as a 4-point Guttman scale, but is often measured on a 5-point Likert-type scale,
ranging from (1) Strongly disagree to (5) Strongly agree.
Items and a note about permission to use the items is available at: www.socy.umd.edu/quick-links/rosenberg-
self-esteem-scale (Retrieved January 1, 2014).
According to the website, Rosenberg’s family has given permission to use the scale for educational and profes-
sional research without charge. Their only request is that copies of published works are sent to this address: The
Morris Rosenberg Foundation, c/o Department of Sociology, 2112 Art-Sociology Building, University of
Maryland, College Park, MD 20742-1315, USA.
Source : Reproduced with permission.
Single-Item Self-Esteem Scale (SISE)
(Robins et al., 2001a ).
Variable
Global self-worth or the overall attitude that one holds about oneself can be measured on a single-item scale
ranging from negative to positive self-esteem ( Robins et al., 2001a ).137 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Description
The SISE was developed to serve as a proxy for the RSE in research contexts in which time constraints severely
limit the number of items that can be administered (e.g., large-scale surveys, pre-screening packets, longitudinal
studies, experience sampling studies). The SISE is a highly face valid measure that assesses an individual’s
explicit knowledge about one’s global self-evaluation. The SISE provides a very brief, standardized measure that
bypasses the need for individual researchers to abbreviate measures of global self-esteem. The SISE asks the par-
ticipant to rate the statement ‘I have high self-esteem’ on a 5-point Likert-type scale, ranging from ‘not very true
of me’ to ‘very true of me.’ Some researchers have used alternative item wording (‘My self-esteem is high’, ‘I see
myself as someone who has high self-esteem’), scale anchors (‘strongly disagree’ and ‘strongly agree’), and
response formats (7- and 9-point Likert-type rating scales), apparently with little/no appreciable effects on
the psychometric properties of the measure (see discussion regarding number of response options in Chapter 1,
this volume).
Sample
The SISE was initially developed and validated using data from four studies, a 4-year longitudinal study of
college students, a cross-sectional study of college students, a community sample, and a sample of 4th to 8th
grade children. In a separate study, Ackerman, Brecheen, Corker, Donnellan, and Witt (2013) investigated the
test/C0retest reliability of the SISE in a sample of ( N5300) undergraduates (see below).
Reliability
Internal Consistency
This is not relevant given the single item nature of the measure.
Test/C0Retest
Robins et al. (2001a) reported a test /C0retest coefficient of .75, using the Heise (1969) procedure for estimating
test/C0retest reliability from the pattern of auto-correlations over six waves of data spanning four years of college.
Vazire and Mehl (2007) reported a test /C0retest coefficient of .79, as cited in Vazire, Naumann, Rentfrow, &
Gosling (2008, p. 1442) .Dollinger and Malmquist (2009) reported an 11-week test /C0retest coefficient of .55. Using
unpublished data from approximately 300 college students at a Southwestern University followed over a
1-semester interval, Ackerman et al. (2013) reported a test /C0retest coefficient of .53 for the SISE (compared with
.67 for the RSE). The test /C0retest reliability of the SISE compares favorably with that of other single-item mea-
sures, according to a meta-analysis of single-item scales conducted by Postmes, Haslam, and Jans (2012) .
Validity
Convergent/Concurrent
Robins et al. (2001a) reported positive correlations between the SISE and RSE of .75 in a sample of college stu-
dents, .74 in a second sample of college students, and .80 in a community sample. The strong convergence
between the SISE and the RSE measures held: (1) for males and females; (2) for different ethnic groups; (3) for
both college students and community members; (4) for different occupational statuses; (5) across four years of
college; and (6) for both a 5-point and 7-point rating scale. The SISE also exhibits moderate convergent validity
during childhood, correlating .52 with the Harter (1985) Global Self-Worth (GSE) scale in a sample of 9- to
13-year olds; in the same sample, the correlation between parent reports of their children’s self-esteem using
the SISE and the GSE was .74 ( Robins et al., 2001a ). Using data from over 1,200 first-year students at a
Southwestern University, Ackerman et al. (2013) found a correlation of .68 between the SISE and RSE. Finally,
the SISE was correlated .79 with a composite of the Tafarodi and Swann (2001) self-liking and self-competence
scales/C0revised ( Gebauer, Riketta, Broemer, & Maio, 2008 ). Overall, these correlations are as high, if not higher,
than those typically found between the various multi-item global self-esteem scales commonly reported in
the literature.
Robins et al. (2001a) found that higher scores on the SISE were as sociated with being male; more positive
domain-specific self-evaluations; higher levels of Big Five Extraversion and Conscientiousness; lower
levels of Big Five Neuroticism and shyness; higher levels of optimism; higher levels of psychological and
physical health; and higher levels of peer-rated perfor mance in a group interaction task. The overall pattern
of convergent correlations was nearly identical for the SISE and the RSE (correlation between the two138 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
columns of correlations 5.98), supporting the validity of both scales. In terms of cross-method conver-
gence correlations, child and parent reports of a ch ild’s self-esteem using the SISE were correlated
.23 ( Robins et al., 2001a ); this level of agreement, although low, is not usual in studies of children
(Meyer et al., 2001 ). Also, the SISE correlates positively (.30) with self-report measures of self-enhancement
(Robins et al., 2001a ).
Divergent/Discriminant
Robins et al. (2001a) found that the SISE was not substantially co rrelated with Big Five Agreeableness and
Openness to Experience ( r5.04 and r5.11, respectively); peer ratings of c ooperativeness, competitiveness,
and creativity in a group interaction task (r5.08, .04, and .10, respectively) ; several measures of academic
success (e.g., grades, test scores; rsr a n g ef r o m 2.10 to .05); or socioeconomic status ( r52 .02). In terms of
socially desirable responding, the SISE is not correl ated with measures of impression management but is
moderately correlated with self-report and cr iterion-based measures of self-enhancement ( Robins et al.,
2001a ;r5.05 and r5.30, respectively). Jonason and Webster (2010) found that the SISE was not signifi-
cantly or positively related to measures of psy chopathy, Machiavellianism, or narcissism ( r52 .09,2.09,
and2.13, respectively), although Ames, Rose, and Anderson (2006) reported a correlation (.24) between the
SISE and narcissism as assessed with a variant of t he Narcissistic Personality Inventory (NPI; Raskin &
Terry, 1988 ). Such correlations are likely to be observed w hen measures of narcissism include attributes
linked with social dominance and extraversion (see Brown & Zeigler-Hill, 2004 ). Also, in the Ackerman et al.
(2013) study, the SISE did not correlate with the explo itative/entitlement dimension of the NPI ( r52 .02,
N51,256).
Construct Factor Analytic
Factor analytic evidence is not relevant given the single item nature of the scale.
Criterion/Predictive
The SISE scores measured at the beginning of college were often predictive of a number of variables aggre-
gated across the subsequent four years ( Robins et al., 2001a ). In addition, scores on the SISE were found to be
predictive of cardiac vagal tone, suggesting better emotion regulation ( Schwerdtfeger & Scheel, 2012 ), and faster
physiological (e.g., blood pressure) habituation to a stressful socially evaluative context such as public speaking
(Elfering & Grebner, 2012 ).
Location
Robins, R.W., Hendin, H.M., & Trzesniewski, K.H. (2001). Measuring global self-esteem: Construct validation
of a single-item measure and the Rosenberg Self-Esteem Scale. Personality and Social Psychology Bulletin, 27 ,
151/C0161.
Results and Comments
The SISE is the briefest measure of global self-esteem available. Despite its brevity, the pattern of convergent
and discriminant correlates of the SISE closely mimic those of other more established measures of global self-
esteem, such as the RSE. However, there are also several notes of caution to consider before using the SISE. First,
it has lower test /C0retest reliability than multi-item measures of global self-esteem. This is likely due to the fact
that random measurement errors are typically larger for items than scales (see discussion in Chapter 1, this vol-
ume). Second, because it relies on the individual’s ability to know and acknowledge explicit feelings of global
self-worth, it may not be suitable for children. Third, the SISE may be more susceptible to extremity and acquies-
cence response styles because it has only one positively keyed item. Fourth, the SISE is a blatantly face valid mea-
sure and shows some susceptibility to socially desirable responding, in particular to self-deceptive enhancement.
However, this concern may be equally applicable to the RSE and other self-report measures of global self-esteem
(see Robins et al., 2001a , Study 2). Fifth, although the SISE has been used in numerous studies conducted with
North American samples, its reliability and validity in other cultures has yet to be systematically explored.
In summary, the accumulated evidence suggests that the SISE provides a practical alternative to longer measures
of global self-esteem.139 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
SINGLE-ITEM SELF-ESTEEM SCALE
Instructions for participants:
Please indicate the extent to which you agree with this statement.
1.I have high self-esteem
Notes: Generally used with a 5-point scale ranging from 1 5not very true of me to 5 5very true of me.
Source : Reproduced with permission.
Self-Liking/Self-Competence Scale /C0Revised (SLSC-R)
(Tafarodi & Swann, 2001 ).
Variable
Tafarodi and Swann (1995) argued that global self-esteem has two distinct components, one corresponding to
a personal sense of worth (termed self-liking) and one corresponding to a sense of personal efficacy (termed self-
competence). Self-liking is a general feeling towards the self, such as feeling positive affect when thinking of the
self. Self-competence is the feeling of being capable and in control, and the belief that one will be successful in
the future.
Description
The SLSC is a 16 -item self-report scale that includes items that measure self-liking and self-competence that
are rated on a 5-point Likert-type response scale. The two 8-item subscales correlate from .47 to .59 ( Mar,
DeYoung, Higgins, & Peterson, 2006; Meagher & Aidman, 2004; Tafarodi & Swann, 2001; Tafarodi, Wild, & Ho,
2010; Vandromme, Hermans, Spruyt, & Eelen, 2007; Wilkinson, 2010 ). Fifty percent of the items are reverse-
worded (see discussion in Chapter 1). Tafarodi and Swann (1985) argue that these dimensions are related, but
substantially distinct; thus, only studies that evaluated distinct self-liking and self-competence scales are
reviewed here.
Sample
Three large samples, all university students ( Ns51,053, 835, and 844, respectively) were used in the
construction/validation of the original SLSC measure ( Tafarodi & Swann, 1995 ). The revised SLSC scale was based
on two further samples of university students ( Ns51,325 and N5298, respectively) ( Tafarodi & Swann, 2001 ).
Reliability
Internal Consistency
Cronbach alpha coefficients were reported for self-liking (ranging from .70 to .98) and self-competence
(ranging from .56 to .92) across the original and revised subscales ( Aidman, 1998 ; Bosson & Swann, 1991;
Cicero & Kerns, 2011 ;Mar et al., 2006 ;Oakes et al., 2008 ;Riketta & Zieglet, 2006 ;Sasaki et al., 2010 ;Silvera et al.,
2001;Song, Thompson, & Ferrer, 2009 ;Tafarodi & Swann, 1995,2001 ;Vandromme et al., 2007 ;Wilkinson,
2010). In addition, split-half reliability coefficients were high for both self-liking (.92) and self-competence
(.75; Vandromme et al., 2007 ).
Test/C0Retest
Self-liking exhibited a test /C0retest reliability coefficient of .80 over a 3-week interval and self-competence exhib-
ited a stability coefficient of .78 ( Tafarodi & Swann, 1995 ). Similar figures have been reported for the revised
scales (.75 and .78, respectively, N5138; Tafarodi & Swann, 2001 ).140 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
Validity
Convergent/Concurrent
Self-liking and self-competence both have strong positive correlations with the RSE. The correlations for the
RSE and self-liking measured by the SLSC scale ranged from .74 to .88 and for the RSE and self-competence
(from .53 to .88; Cicero & Kerns, 2011 ;Mar et al., 2006 ;Vandromme et al., 2007 ). Both self-liking ( β5.71 study 1,
β5.62 study 2) and self-competence ( β5.28 study 1, β5.37 study 2) independently predicted scores on the
RSE in regression analyses with both scales in the model ( Mar et al., 2006 ), and combined the two scales explain
most of the variance in the RSE (R2ranging from .79 to .96; Mar et al., 2006 ;Tafarodi et al., 2010 ). Moderate corre-
lations (ranging from .34 to .57) were found between reporters (mothers, fathers, and children) ( Tafarodi &
Swann, 2001 ). Correlations with self-serving attributions were .22 for self-liking and .24 for self-competence
(Cicero & Kerns, 2011 ). Self-liking scores correlated positively with reports of parental warmth ( r5.24 for men
and r5.32 for women), and also correlated .16 with a measure of social desirability, while self-competence cor-
related .23 with a measure of social desirability ( Tafarodi & Swann, 1995 ). Self-competence scores also correlated
positively with self-ratings of abilities ( r5.30 for men and r5.48 for women) ( Tafarodi & Swann, 1995 ).
Relations with self-deception were also independent and of similar magnitude (around .60 for self-liking and .50
for self-competence; Mar et al., 2006 ).
Divergent/Discriminant
Self-liking scores correlated around 2.30 with self-doubt while self-competence scores correlated 2.63 with
self-doubt ( Greenway et al., 2003 ). Self-liking and self-competence correlated 2.63 and 2.57, respectively with
neuroticism in one study ( Ramsdel, 2008 ) and2.68 and 2.41 in another ( Vandromme et al., 2007 ). Self-liking cor-
related 2.45 with depression and self-competence 2.35 with depression ( Greenway et al., 2003 ).Tafarodi and
Swann (1995) found that the self-liking and self-competence relations with depression were independent and
approximately equal ( 2.30 for self-liking and 2.20 for self-competence), but others have found that the magni-
tude of the relations varies by the type of depression measure used ( Silvera et al., 2001 ). Self-competence scores
were not significantly correlated with reports of parental warmth ( r52 .11 for men and r5.01 for women)
(Tafarodi & Swann, 1995 ). Likewise, self-liking scores were not correlated with self-ratings of ability ( r5.06 for
men and r5.04 for women) ( Tafarodi & Swann, 1995 ).
Construct Factor Analytic
Competing factor models were tested using confirmatory factor analytic methods based on relatively large
samples (e.g., N51,325 in Tafarodi & Swann, 2001 ). The model representing separate self-liking and self-
competence factors fit the data best (compared with a single factor, positive and negative method factors, and
hybrids of the two; Tafarodi & Swann, 1995,2001 ;Vandromme et al., 2007 ). However, Aidman (1998) conducted a
principal component analysis (PCA) on a sample of Australian university students ( N5480) and found three
components: the first comprised the positively worded items, the second captured the negatively worded self-
liking items, and the third captured the positively worded self-competence items. Silvera et al. (2001) conducted
a confirmatory factor analysis with 372 Norwegian college students providing support for the self-liking and
self-competence factors in addition to negative and positive method factors.
Criterion/Predictive
Evidence of criterion/predictive validity is not currently available.
Location
Tafarodi, R.W., & Swann, W.B. Jr., (2001). Two-dimensional self-esteem: Theory and measurement. Personality
and Individual Differences, 31 , 653/C0673.
Results and Comments
There is a potentially interesting conceptual distinction between self-liking and self-competence that could be
a fruitful line for future theoretical and empirical work. However, the two subscales of the SLSC tend to be mod-
erately correlated (e.g., r5.62;Zeigler-Hill, 2010 ). This level of association has perhaps motivated some to aggre-
gate both scores into a single composite (see Buhrmester et al., 2011 ). It might be profitable in future work to
develop scales that are more independent of each other. Likewise, additional work is needed to further identify
the nomological networks of these two constructs as assessed by the SLSC measure.141 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
SELF-LIKING AND SELF-COMPETENCE SCALE /C0REVISED (SLSC-R)
Instructions: These items concern your general thoughts and feelings about yourself.
Please indicate the extent to which you agree or disagree with each item using the 5-point scale below:
Scale Item Scoring
1. I tend to devalue myself. L2
2. I am highly effective at the things I do. C 1
3. I am very comfortable with myself. L 1
4. I am almost always able to accomplish what I try for. C 1
5. I am secure in my sense of self-worth. L 1
6. It is sometimes unpleasant for me to think about myself. L 2
7. I have a negative attitude toward myself. L 2
8. At times, I find it difficult to achieve the things that are important to me. C 2
9. I feel great about who I am. L1
10. I sometimes deal poorly with challenges. C 2
11. I never doubt my personal worth. L 1
12. I perform very well at many things. C 1
13. I sometimes fail to fulfill my goals. C 2
14. I am very talented. C1
15. I do not have enough respect for myself. L 2
16. I wish I were more skillful in my activities. C 2
Notes:
Used with a response scale ranging from 1 5Strongly disagree to 5 5Strongly agree.
Scoring: C 5self-competence; L 5self-liking; 25 negatively-keyed item; 15 positively-keyed item.
www.psych.utoronto.ca/users/tafarodi/ . (Retrieved January 1, 2014).
http://homepage.psy.utexas.edu/homepage/faculty/swann/research_materials.htm . (Retrieved January 1, 2014).
Permission not needed for non-commercial use.
Source : Reproduced with permission.
Harter Self-Perception Profile (SPP)
(e.g., Harter, 2012a,b ;Neemann & Harter, 2012 ).
Variable
The family of SPP profiles each assesses global self-worth as well as multiple specific domains of self, relevant
for individuals at different developmental periods ( Neemann & Harter, 2012 ).
Description
The original self-perception profile was developed for children aged 8 to 13 years and was briefly detailed in
theBlascovich and Tomaka (1991) chapter. We expand on their treatment in this section to cover the entire family
of Harter measures. There are different profiles for use with children, adolescents, college students, adults, and
most recently, older adults. Each profile assesses multiple domains that are developmentally appropriate and the
number of domains increases with age. The SPP presents two statements and asks the individual to choose the
one that is most self-descriptive, and then rate it as ‘A Lot Like Me’ or ‘A Little Like Me.’ These responses
are rated on a 4-point scale. The two-step, forced-choice format was originally designed to reduce socially-
desirable responses but Marsh (1992a) has queried this claim. The scale for young children (preschool to third
grade; boy and girl versions) is administered individually in pictorial format, eliminating the need for reading
skills. However, no global subscale is available for young children because it was argued that children under
eight years of age cannot communicate feelings of global self-worth ( Harter & Pike, 1984 ). The scales for other
ages use the same format as the scale for young children, except words are used instead of pictures.142 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
The pictorial scale for young children (aged 4 to 7 years) contains four subscales labeled: cognitive compe-
tence, physical competence, peer acceptance, and maternal acceptance; however, analyses suggested only two
dimensions based on samples of N5145 and N5104 children /C0competence and acceptance ( Harter &
Pike, 1984 ). The child scale (aged 8 to 13 years) has six subscales labeled: scholastic, social, athletic, physical,
behavioral, and global self-worth. The adolescent scale (aged 14 to 19 years) has nine subscales labeled: scholastic
competence, social competence, athletic competence, physical appearance, job competence, romantic appeal,
behavioral conduct, close friendship, and global self-worth. The college student scale (aged 17 to 23 years) has
13 subscales labeled: creativity, intellectual ability, scholastic competence, job competence, athletic competence,
appearance, romantic relationships, social acceptance, close friends, parent relationships, sense of humor, moral-
ity, and global self-worth. The adult scale (aged 20 to 60 years) has 12 subscales: intelligence, job competence, ath-
letic competence, physical appearance, sociability, close friendship, intimate relationships, morality, sense of
humor, nurturance, household management, adequacy as a provider, and global self-worth. The older adult scale
(aged 60 years and older) has 14 subscales labeled: relationship with friends, family relationships, nurturance,
adequacy as a provider, job competence, cognitive abilities, personal/household management, leisure activities,
health status, physical appearance, morality, life satisfaction, reminiscence, and global self-worth.
Sample
The original samples were primarily drawn from Colorado in the USA and were mostly Caucasian. Sample
sizes for the development of the Self-Perception Profile for Children (e.g., Harter, 2012a ) ranged from N5220
(Grade 5) to N5928 (Grade 6) with a total sample of 2,744 students in Grades 3 to 8. Sample sizes for the devel-
opment of the Self-Perception Profile for Adolescents (e.g., Harter, 2012b ) ranged from N5165 (Grade 8) to
N5361 (Grade 11) with a total sample of 1,099 students in Grades 8 to 11. The Self-Perception Profile for
College Students ( Neemann & Harter, 2012 ) was developed using data from 300 college students. The Self-
Perception Profile for Adults ( Messer & Harter, 2012 ) was developed using data from 356 parents (88% women).
Reliability
Internal Consistency
Harter reported Cronbach alpha coefficients ranging from .71 to .92 across all of the subscales and ages
(Harter, 1985 ; 2012; Neemann & Harter, 2012 ).Rose, Hands, and Larkin (2012) reported alpha coefficients ranging
from .68 to .87 for the adolescent scale, consistent with several other studies conducted in different counties
(seeRose et al., 2012 ). Alpha coefficients ranged from .70 to .89 for the adult scales ( Donnellan et al., 2007 ).
Test/C0Retest
Harter (1984) reported stability coefficients for eight samples for the child scale. These ranged from .70 to .87
for three months and .69 to .80 for nine months. A meta-analysis found a long-term stability coefficient of .43
(average time interval three years) for the global self-worth scale ( Trzesniewski et al., 2003 ).Donnellan et al.
(2007) reported four-year stability correlations ranging from .38 to .77 for the adult scales.
Validity
Convergent/Concurrent
The adult global subscale correlated .74 with the RSE Scale ( Donnellan et al., 2007 ) and the college student
global scale correlated .75 with the RSE (Ferrier et al., 2008). The child subscales from the SPP correlated as
expected with Marsh’s Self-Description Questionnaire (SDQ-I) using data from 508 students ( Marsh &
Gouvernet, 1989 ). The physical scales from both inventories were correlated at .67, the social scale from the
Harter inventory was correlated .74 with the peer scale for the SDQ-I, the cognitive scale from the Harter inven-
tory was correlated .60 with the total academic scales, and the general scales from both were correlated at .57
(Marsh & Gouvernet, 1989 ).
The child scales correlated positively with teacher reports of scholastic competence ( r5.35, .43, and .60 for
students in Grades 2, 3, and 4, respectively; Boivin, Vitaro, & Gagnon, 1992 ), teacher reports of peer acceptance
(r5.19, .28, and .33 for students in Grades 2, 3, and 4, respectively; Boivin et al., 1992 ) and teacher reports of
peer status ( r5.27, .32, and .38 for students in Grades 2, 3, and 4, respectively; Boivin et al., 1992 ). In a sample
of 107 Mexican-American school children, the scholastic scale was correlated with a positive school attitude
(r5.58) and the physical ( r5.41), behavioral ( r5.64), and global scales ( r5.51) ( Hess & Petersen, 1996 ).143 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Divergent/Discriminant
Self-perception scores for the physical appearance scale did not correlate with teacher reports ( rs range from
.08 to .12; Boivin et al., 1992 ). The adolescent version was unrelated or weakly related to social desirability with
correlations ranging from .02 (for close friends) to .29 (for physical appearance), similar for the original forced-
choice and alternative Likert-type format ( Wichstrom, 1995 ). The global Harter scales for children correlated
negatively both with internalizing problems ( r52 .22) and externalizing problems ( r52 .30) as well as
with measures of anxiety and depression ( r52 .56 and 2.67, respectively) in samples of Dutch children
(Muris, Meesters, & Fijen, 2003 ).
Construct Factor Analytic
Several studies have evaluated the factor structure of the Harter measures, although these studies have tended to
omit the global items and it is not clear whether the results would be different if the global items were included in
the analyses. Some studies have replicated the intended factor structure ( Harter, 1985 ;2 0 1 2 ; Neemann & Harter,
2012) and using diverse samples ( Boivin et al., 1992; Ferrier & Martens, 2008; Granleese & Joseph, 1993; Miller, 2000;
Muris et al., 2003; Rose et al., 2012; Van Dongen-Melman, Koot, & Verhulst, 1993; Worth et al., 1996 ), but others
have been unable to replicate this factor structure or have only partially replicated it ( Egberink & Meijer, 2011; Eiser,
Eiser, & Havermans, 1995; Rose et al., 2012; Thomson & Zand, 2002; Wichstrom, 1995 ).
Criterion/Predictive
Several studies have shown that the self-perception profile subscales predict hypothesized outcomes over
time. For example, using the child scale, lower scores on the physical appearance scale predicted later depression
(β52 .12;Kim-Spoom, Ollendick, & Seligman, 2012 ) and lower scores on the social competence scale predicted
later internalizing and externalizing behaviors ( β52 .19 and β52 .18, respectively; Bornstein, Hahn, &
Haynes, 2010 ). Lower scores on the physical appearance scale but not on the social competence scale, predicted
later bulimic symptoms, controlling for prior symptom levels ( β52 .16 and β52 .02, respectively; Holm-
Denoma & Hankin, 2010 ). Scores on the scholastic competence scale predicted later school engagement ( β5.18
Chen, Hughes, Liew, & Kwok, 2010 ). Similar predictive validity findings have been found using the adolescent
scale: lower scores on the global self-worth predicted later disordered eating ( β52 .11 to 2.21; Kansi,
Wichstrom, & Bergman, 2005 ;β52 .20 to 2.31;Salafia, Gondoli, Corning, Bucchianeri, & Godinez, 2009 ),
declines in depressive symptoms ( β52 .30; Burwell & Shirk, 2006 ), and higher social acceptance scores
predicted later peer-reported companionship, lower hostility, and lower withdrawal ( β5.19, β52 .31, and
β52 .23, respectively; McElhaney, Antonishak, & Allen, 2008 ).
Location
Harter, S. (2012a). The self-perception profile for children: Manual and questionnaires. Unpublished manuscript,
University of Denver, Colorado, USA.
Harter, S. (2012b). The self-perception profile for adolescents: Manual and questionnaires. Unpublished manuscript,
University of Denver, Colorado, USA.
Scales and manuals can be downloaded from: http://portfolio.du.edu/SusanHarter (Retrieved January 1,
2014).
Results and Comments
The SPP measures have several benefits. Foremost, there are scales available for different age groups so these
instruments permit a developmentally sensitive approach to assessing self-evaluations across the life span.
Moreover, the scales appear to have predictive validity for important outcomes. However, there are also several
notes of caution to consider before deciding to use the SPP. First, some participants do not understand the
forced-choice format. In some of our own data, we found roughly 20% of children completed the scale incor-
rectly, when administered in a group setting, making their data unusable (cf. Eiser et al., 1995 ). Harter has since
updated her manuals providing detailed instructions for administering the scales to reduce this potential loss of
data. Also, the forced-choice format may not be necessary. Wichstrom (1995) tested the forced-choice format and
a standard Likert-type format and found the latter resulted in the same factorial structure, higher reliabilities,144 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
and better convergent validity. Moreover, the Likert-type version showed just as low correlations with social
desirability as did the original scale.
Second, although the SPP scales form a life-span measure, the domains differ by age and there is some vari-
ability in the item wording for the domains that do cut across age groups. Thus, for many analyses, age differ-
ences will be somewhat confounded by differences in item content. Third, global self-esteem is not included until
age 8 years because Harter (e.g., Harter & Pike, 1984 ) posited that children younger than 8 years are not able to
make global evaluations about the self. However, Marsh et al. (1991) questioned Harter’s claim, and presented
evidence of a well-defined global self-worth factor that was distinct from domain-specific factors in a sample of 6
to 8-year-olds (cf. Marsh, Ellis, & Craven, 2002 ). In addition, Verschueren, Buyck, and Marcoen (2001) assessed
global self-esteem in 5-year-olds through a puppet interview and found good correspondence with self-reported
global self-esteem at age 8 years. Thus, the omission of a global scale on the Harter measure for the youngest chil-
dren may be a limitation.
Fourth, more evidence is needed to support the number and type of domains for each age group.
For example, Marsh (1992a) has made a convincing case for the value of assessing self-competence in the
separate domains of mathematics and reading/verbal a bilities. The Harter scales do not separately assess
these domains and this might be a concern, especially give n purported gender differences of self-perceptions
in these domains. Conversely, there might be too many domains on some SPP versions. These domains
add length without necessarily providing psychologic ally important information. In short, more work is
needed to establish the validity and necessity of all t he domains assessed by the various profiles developed
by Harter.
In sum, there are a number of positive features of the Harter SPP measures ( Byrne, 1996 ). The global self-
esteem scale converges strongly with the RSE and the re is little evidence (as yet) that one global scale
is superior to the other. Our reservations are mostl y with respect to the response format of the items and
t h en e e df o rm o r ee v i d e n c ef o rt h ev a l i d i t yf o re a c ho f the subscales. The bottom line is that if researchers
wish to use a multidimensional measure of the self-c oncept, the decision will likely boil down to a choice
between the Harter family of measures or the Self-Des cription Questionnaires (SDQ), discussed next.
ADMINISTRATION INSTRUCTIONS FOR THE HARTER
SELF-PERCEPTION PROFILE FOR CHILDREN
Instructions to the child
We have some sentences here and, as you can see
from the top of your sheet where it says “What I am
like”, we are interested in what each of you is like, what
kind of a person you are like. This is a survey, nota test.
There are no right or wrong answers. Since kids are very
different from one another, each of you will be putting
down something different.
First, let me explain how these questions work. There
is a sample question at the top, marked (a). I’ll read it
out loud and you follow along with me. ( Examiner reads
the sample question .) This question talks about two kinds
of kids, and we want to know which kinds are most
likeyou.
(1)So, what I want you to decide first is whether you
are more like the kids on the left side who
would rather play outdoors, or whether you are
more like the kids on the right side who wouldrather watch T.V. Don’t mark anything yet, but first
decide which kinds of kids are most like you , and go
to that side of the sentence.
(2)Now the second thing I want you to think about,
now that you have decided which kinds of kids are
most like you, is to decide whether that is only sort
of true for you,o rreally true for you . If it’s only sort of
true, the put an X in the box under Sort of True for
me; if it’s really true for you, then put an X in that
box, under Really True for me.
(3)For each sentence, you only check onebox.
Sometimes it will be on one side of the page, another
time it will be on the other side of the page, but you
can only check one box for each sentence. Y OU DON’T
CHECK BOTH SIDES ,JUST THE ONE SIDE MOST LIKE YOU .
(4)OK, that one was just for practice. Now we have
some more sentences that I will read out loud. For
each one, just check one box—the one that goes with
what is true for You, what you are most like.145 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
HARTER SPP SAMPLE ITEMS
Really
True
for meSort of
True
for meSort of
True
for meReally
True
for me
Sample Sentence
a.&& Some kids would
rather play outdoors in
their spare timeBUT Other kids would rather
watch T.V.&&
1.&& Some kids feel that
they are very good at
their school workBUT Other kids worry about
whether they can do the
school work assigned to them&&
2.&& Some kids find it hard
to make friendsBUT Other kids find it pretty easy
to make friends&&
3.&& Some kids do very well
at all kinds of sportsBUT Other kids don’t feel that
they are very good when it
comes to sports&&
4.&& Some kids are happy
with the way they lookBUT Other kids are nothappy
with the way they look&&
5.&& Some kids often do not
like the way they
behaveBUT Other kids usually like the
way they behave&&
6.&& Some kids are often
unhappy with
themselvesBUT Other kids are pretty pleased
with themselves&&
Notes:
Sample items are from the child (age 8 to 13 years) profile.
A scoring key is provided in the user manuals: http://portfolio.du.edu/SusanHarter Reproduced with
permission.
Sample items from the adolescent profile:
Some teenagers are very happy being the way they are BUT Other teenagers often wish they were different.
(Global Self-Worth Scale)
Some teenagers feel that they are pretty intelligent BUT Other teenagers question whether they are intelligent.
(Scholastic Competence Scale)
Sample items from the college students profile:
Some students usually like themselves as a person BUT Other students often don’t like themselves as a person.
(Global Self-Worth Scale)
Some students do not feel they are very mentally able BUT Other students feel they are very mentally able.
(Intellectual Ability Scale)
Sample items from the adults profile:
Some adults like the way they are leading their lives BUT Other adults don’t like the way they are leading their
lives. (Global Self-Worth Scale)
When some adults don’t understand something, it makes them feel stupid BUT Other adults don’t necessarily feel
stupid when they don’t understand. (Intelligence Scale)
Please note that these are sample items and complete inventories are available at Susan Harter’s website.
Furthermore, Susan Harter requests that all items from each domain scale are administered. Do not shorten the
domain scales without explicit permission. However, researchers do not need to administer all scales depending
on their research needs.146 6. MEASURES OF SELF-ESTEEM
II. EMOTIONAL DISPOSITIONS |
Self-Description Questionnaire (SDQ-1; SDQ-II; SDQ-III)
(Marsh, 1992a,b,c ).
Variable
The basic underlying measurement philosophy of the SDQ scales is to assess aspects of the self-concept in
domains that are developmentally relevant (such as academic domains and peer relationships). Individuals might
feel positively about their level of fitness but not like their academic abilities so obtaining scores in different
domains can provide a more detailed profile of an individual’s self-concept.
Description
Marsh and his colleagues have created a family of measures of the domains of the self-concept (see Byrne,
1996) based on theoretical considerations discussed extensively in Marsh (1990a) and Marsh and Craven (2006) .
The various SDQ measures have developmentally appropriate scales for assessing different domains of the self-
concept relevant for individuals of different ages. Each SDQ inventory has a global scale modeled on the RSE so
that researchers can evaluate how strongly each domain is related to global self-esteem, the construct that is pos-
ited to sit at the highest level of the self-concept hierarchy (see Marsh & Craven, 2006 ).
SDQ-I ( Marsh, 1992a ) is a 76-item measure designed to assess aspects of the self-concept relevant for pre-
adolescent children aged 8 to 12 years. A preliminary evaluation of the SDQ-I was presented in Blascovich and
Tomaka (1991, pp. 144 /C0147). Items are answered on a 5-point scale ranging from ‘False’ to ‘True’ with a midpoint
labeled ‘Sometimes false/Sometimes true’. Although originally designed for those in middle to late childhood,
the items may be used with children as young as second grade. There is a version for even younger children
(Kindergarten: aged 5 to 7 years) using a one-on-one interview format ( Marsh et al., 1991 ). The SDQ-I includes
seven scales designed to measure broad domains of the self-concept relevant for preadolescents, four from the
non-academic realm (labeled: Physical Abilities, Physical Appearance, Peer Relations, and Parent Relations) and
three from the academic realm (labeled: Reading, Mathematics, and General School Subjects). Many of the items
on the SDQ-I are similar to items on the SDQ-II and SDQ-III.
The SDQ-II ( Marsh, 1992b ) is a 102-item measure designed to assess aspects of the self-concept relevant for
adolescents roughly between the ages of 12 and 18 years (Middle School and High School). Items are rated on a
6-point scale ranging from ‘False’ to ‘True’. There is a general self-esteem scale and 10 domain scales - seven
non-academic (labeled: Physical Abilities, Physical Appearance, Same-Sex Peer Relations, Opposite-Sex Peer
Relations, Parent Relations, Emotional Stability, and Honesty) and three academic (labeled: Verbal, Mathematics,
and General Academics).
The SDQ-III ( Marsh, 1992c ) is a 136-item measure designed to assess aspects of the self-concept relevant for
late adolescents and college students. The heavy focus on aspects of academic self-concept means that many
scales will be less relevant for adults who are not at school. Byrne (1996) noted that certain domains relevant to
older adults (e.g., domains of work and parenthood) are not well-represented in the SDQ-III. Items are rated on
an 8-point scale ranging from ‘Definitely False’ to ‘Definitely True’. There is a general self-esteem scale and 12
domain scales /C0eight non-academic (labeled: Physical Abilities, Physical Appearance, Same-Sex Peer Relations,
Opposite-Sex Peer Relations, Parent Relations, Emotional Stability, Honesty, and Spiritual Values/Religion) and
four academic (labeled: Verbal, Mathematics, Problem Solving, and General Academics).
Sample
Data on the psychometric properties of the SDQ-I is based on 3,562 students from New South Wales,
Australia. The global self-esteem scale was later developed for the SDQ-I using an 8-item general scale based on
the RSE (see Marsh, Smith, & Barnes, 1985b, pp. 585 /C0586) and evaluated using data from 559 5th grade
Australian school students. Initial information concerning the psychometric properties of SDQ-II were based on
901 Australian high school students as reported in Marsh et al. (1985a,b) , later refined using a large normative
sample of 5,494 5th grade Australian school students ( Marsh, 1992b ). Normative data on the SDQ is based on
9,187 Australian high school students ( Marsh et al., 2005a,b ) and 17,544 students from the United States who
completed 21 items from the SDQ as part of the National Educational Longitudinal Survey of 1988 ( Marsh, 1994 ).
Normative data on the SDQ-III is available on an Australian sample ( N52,436; Marsh, 1992c ) and a Canadian
sample ( N5991; Byrne, 1988 ).147 OVERVIEW OF THE MEASURES
II. EMOTIONAL DISPOSITIONS |
Subsets and Splits