|
|
|
1 |
|
00:00:07,390 |
|
So last time, we discussed how to find the |
|
|
|
2 |
|
00:00:09,890 |
|
probabilities underneath the normal curve for |
|
|
|
3 |
|
00:00:14,570 |
|
three cases. If the point lies in the lower tail, |
|
|
|
4 |
|
00:00:21,690 |
|
as this one, or upper tail, or between. So we can |
|
|
|
5 |
|
00:00:28,150 |
|
do the computations for this kind of |
|
|
|
6 |
|
00:00:32,290 |
|
probabilities. And I think we gave two examples. |
|
|
|
7 |
|
00:00:36,620 |
|
For example, if we are looking for probability of |
|
|
|
8 |
|
00:00:40,400 |
|
Z is smaller than 0.1, and Z, as we mentioned last |
|
|
|
9 |
|
00:00:45,020 |
|
time, is the standardized normal distribution that |
|
|
|
10 |
|
00:00:48,660 |
|
has mean of 0 and sigma is 1. In this case, just |
|
|
|
11 |
|
00:00:53,120 |
|
go to the table you have. Now we are looking for 0 |
|
|
|
12 |
|
00:00:57,540 |
|
.12, for example. So here we have 0.1. Under 2, we |
|
|
|
13 |
|
00:01:03,700 |
|
get this result. This value is the probability of |
|
|
|
14 |
|
00:01:07,260 |
|
Z smaller than 0.5. But you have to keep in mind |
|
|
|
15 |
|
00:01:10,480 |
|
that we have to transform first from normal |
|
|
|
16 |
|
00:01:14,460 |
|
distribution to standardized normal distribution. |
|
|
|
17 |
|
00:01:18,620 |
|
For this specific example, if we are looking for |
|
|
|
18 |
|
00:01:21,220 |
|
mean of 8 and standard deviation of 5, that's for |
|
|
|
19 |
|
00:01:26,960 |
|
the normal distribution. In this case, the z-score |
|
|
|
20 |
|
00:01:30,020 |
|
is given by this equation, which is x minus 3 |
|
|
|
21 |
|
00:01:32,920 |
|
divided by sigma. So 8.6 minus 8 divided by 5 |
|
|
|
22 |
|
00:01:37,060 |
|
gives 0.12. In this case, we can use the |
|
|
|
23 |
|
00:01:41,360 |
|
standardized normal table. Now, for the other |
|
|
|
24 |
|
00:01:46,480 |
|
case, we are looking for the probability of x |
|
|
|
25 |
|
00:01:48,700 |
|
greater than 8.6. So we are looking for the upper |
|
|
|
26 |
|
00:01:51,880 |
|
tail probability. The table we have gives the area |
|
|
|
27 |
|
00:01:56,900 |
|
to the left side. And we know that the total area |
|
|
|
28 |
|
00:02:01,640 |
|
underneath the normal curve is 1. So the |
|
|
|
29 |
|
00:02:04,760 |
|
probability of x greater than 8.6 is the same as 1 |
|
|
|
30 |
|
00:02:08,980 |
|
minus the probability of x smaller than x less |
|
|
|
31 |
|
00:02:13,720 |
|
than 8.6. So here, for 8.6, we got z squared to be |
|
|
|
32 |
|
00:02:19,100 |
|
0.12. So the probability of z greater than 0.12 is |
|
|
|
33 |
|
00:02:24,300 |
|
the same as 1 minus z of z is smaller than 0.12. |
|
|
|
34 |
|
00:02:28,850 |
|
So 1 minus the answer we just got from part A will |
|
|
|
35 |
|
00:02:33,050 |
|
get us to something. So this is the probability of |
|
|
|
36 |
|
00:02:35,930 |
|
X is greater than 8.6. So all of the time, if you |
|
|
|
37 |
|
00:02:41,170 |
|
are asking about computing the probability in the |
|
|
|
38 |
|
00:02:45,390 |
|
upper tail, you have to first find the probability |
|
|
|
39 |
|
00:02:48,290 |
|
in the lower tail, then subtract that value from |
|
|
|
40 |
|
00:02:52,630 |
|
1. So that's the probability for the upper tail. |
|
|
|
41 |
|
00:02:58,110 |
|
The last one, we are looking for probability |
|
|
|
42 |
|
00:03:02,470 |
|
between two values. For example, x. What's the |
|
|
|
43 |
|
00:03:06,970 |
|
probability of x greater than 8 and smaller than 8 |
|
|
|
44 |
|
00:03:10,190 |
|
.6? So we are looking for this area, the red area. |
|
|
|
45 |
|
00:03:16,250 |
|
So the probability of x between these two values |
|
|
|
46 |
|
00:03:18,710 |
|
actually equals the probability of x smaller than. |
|
|
|
47 |
|
00:03:23,040 |
|
8.6 minus the probability of X smaller than 8. |
|
|
|
48 |
|
00:03:27,880 |
|
And that, in this case, you have to compute two |
|
|
|
49 |
|
00:03:31,140 |
|
values of this score. One for the first value, |
|
|
|
50 |
|
00:03:34,880 |
|
which is A. This value is zero because the mean is |
|
|
|
51 |
|
00:03:40,020 |
|
zero. And we know that Z can be negative. If X is |
|
|
|
52 |
|
00:03:45,240 |
|
smaller than Mu, Z can be positive if X is greater |
|
|
|
53 |
|
00:03:49,460 |
|
than Mu and equals zero only if X equals Mu. In |
|
|
|
54 |
|
00:03:55,220 |
|
this case, X equals Mu, so Z score is zero. The |
|
|
|
55 |
|
00:03:59,340 |
|
other one as we got before is 0.12. So now, we |
|
|
|
56 |
|
00:04:03,220 |
|
transform actually the probability of X between |
|
|
|
57 |
|
00:04:08,870 |
|
8 and 8.6 to z-score between 0 and 0.12. In this |
|
|
|
58 |
|
00:04:13,870 |
|
case, we can use the normal theorem. Now, this |
|
|
|
59 |
|
00:04:17,170 |
|
area is, as we mentioned, just b of x smaller than |
|
|
|
60 |
|
00:04:21,130 |
|
0.12 minus b of x, b of z smaller than 0. This |
|
|
|
61 |
|
00:04:28,850 |
|
probability, we know that, 0.54878. Now, the |
|
|
|
62 |
|
00:04:34,330 |
|
probability of z smaller than 0 is one-half. |
|
|
|
63 |
|
00:04:38,860 |
|
Because the total area underneath the normal curve |
|
|
|
64 |
|
00:04:41,380 |
|
is 1, and 0 divided the curve into two equally |
|
|
|
65 |
|
00:04:45,860 |
|
parts. So the area to the right of 0 is the same |
|
|
|
66 |
|
00:04:49,100 |
|
as the area to the left of 0. So in this case, |
|
|
|
67 |
|
00:04:52,880 |
|
minus 0.5. So this is your answer. So the |
|
|
|
68 |
|
00:04:57,160 |
|
probability of Z between 8 and 8.6 is around 0.0478. |
|
|
|
69 |
|
00:05:04,320 |
|
I think we stopped last time at this point. |
|
|
|
70 |
|
00:05:09,860 |
|
This is another example to compute the probability |
|
|
|
71 |
|
00:05:17,760 |
|
of X greater than 7.4 and 8. Also, it's the same |
|
|
|
72 |
|
00:05:24,020 |
|
idea here, just find these scores for the two |
|
|
|
73 |
|
00:05:29,700 |
|
values. |
|
|
|
74 |
|
00:05:33,850 |
|
L with B of Z greater than minus 0.12 up to 0. Now |
|
|
|
75 |
|
00:05:38,670 |
|
this red area equals the area below 0. I mean B of |
|
|
|
76 |
|
00:05:47,010 |
|
Z smaller than 0 minus the probability of Z |
|
|
|
77 |
|
00:05:50,330 |
|
smaller than minus 0.12. Now by using symmetric |
|
|
|
78 |
|
00:05:57,090 |
|
probability of the normal distribution, we know |
|
|
|
79 |
|
00:06:00,470 |
|
that the probability of Z smaller than minus 0.12 |
|
|
|
80 |
|
00:06:04,620 |
|
equals the probability of Z greater than 0.12. |
|
|
|
81 |
|
00:06:11,780 |
|
Because this area, if we have Z smaller than minus |
|
|
|
82 |
|
00:06:15,520 |
|
0.12, the area to the left, that equals the area |
|
|
|
83 |
|
00:06:19,300 |
|
to the right of the same point, because of |
|
|
|
84 |
|
00:06:21,560 |
|
symmetry. And finally, you will end with this |
|
|
|
85 |
|
00:06:25,480 |
|
result. |
|
|
|
86 |
|
00:06:30,670 |
|
D of z minus 0.12, all the way up to 0, this area, |
|
|
|
87 |
|
00:06:37,630 |
|
is the same as the area from 0 up to 0.5. So this |
|
|
|
88 |
|
00:06:43,130 |
|
area actually is the same as D of z between 0 and |
|
|
|
89 |
|
00:06:46,530 |
|
0.5. So if you have a negative sign, and then take |
|
|
|
90 |
|
00:06:51,270 |
|
the opposite one, the answer will be the same |
|
|
|
91 |
|
00:06:54,390 |
|
because the normal distribution is symmetric |
|
|
|
92 |
|
00:06:57,610 |
|
around 0. The questions, I think we stopped |
|
|
|
93 |
|
00:07:04,690 |
|
here. And also we talked about empirical rules. |
|
|
|
94 |
|
00:07:10,930 |
|
The one we mentioned in chapter three, in chapter |
|
|
|
95 |
|
00:07:13,250 |
|
three. And we know that, as we mentioned before, |
|
|
|
96 |
|
00:07:16,310 |
|
that is 68.16% of the observations fall within one |
|
|
|
97 |
|
00:07:25,550 |
|
standard deviation around the mean. So this area |
|
|
|
98 |
|
00:07:30,660 |
|
from mu minus one sigma up to mu plus sigma, this |
|
|
|
99 |
|
00:07:35,380 |
|
area covers around 68%. Also |
|
|
|
100 |
|
00:07:42,080 |
|
95% or actually 95.44% of the data falls within |
|
|
|
101 |
|
00:07:50,700 |
|
two standard deviations of the mean. And finally, |
|
|
|
102 |
|
00:07:55,040 |
|
around most of the data, around 99.73% of the data |
|
|
|
103 |
|
00:08:00,220 |
|
falls within three subdivisions of the population |
|
|
|
104 |
|
00:08:06,860 |
|
mean. |
|
|
|
105 |
|
00:08:11,550 |
|
And now the new topic is how can we find the X |
|
|
|
106 |
|
00:08:14,970 |
|
value if the probability is given. It's vice |
|
|
|
107 |
|
00:08:18,510 |
|
versa. In the previous questions, we were asking |
|
|
|
108 |
|
00:08:21,810 |
|
about find the probability, for example, if X is |
|
|
|
109 |
|
00:08:26,130 |
|
smaller than a certain number. Now suppose this |
|
|
|
110 |
|
00:08:30,450 |
|
probability is given, and we are looking to find |
|
|
|
111 |
|
00:08:34,070 |
|
this value. I mean, for example, suppose in the |
|
|
|
112 |
|
00:08:38,710 |
|
previous examples here, |
|
|
|
113 |
|
00:08:43,380 |
|
Suppose we know this probability. So the |
|
|
|
114 |
|
00:08:46,760 |
|
probability is given. The question is, how can we |
|
|
|
115 |
|
00:08:50,220 |
|
find this value? It's the opposite, sometimes |
|
|
|
116 |
|
00:08:54,140 |
|
called backward normal calculations. |
|
|
|
117 |
|
00:09:01,660 |
|
There are actually two steps to find the x value |
|
|
|
118 |
|
00:09:05,580 |
|
for a certain probability or for a given or for a |
|
|
|
119 |
|
00:09:10,040 |
|
known probability the first step we have to find |
|
|
|
120 |
|
00:09:12,900 |
|
the z score then use this equation to find the x |
|
|
|
121 |
|
00:09:20,540 |
|
value corresponding to the z score you have and x |
|
|
|
122 |
|
00:09:25,100 |
|
is just mu plus sigma times mu so first step you |
|
|
|
123 |
|
00:09:30,120 |
|
have to find the z score |
|
|
|
124 |
|
00:09:35,350 |
|
corresponding to the probability we have. So find |
|
|
|
125 |
|
00:09:38,690 |
|
the z value for the non-probability, then use that |
|
|
|
126 |
|
00:09:43,870 |
|
z score to find the value of x by using this |
|
|
|
127 |
|
00:09:47,910 |
|
equation. So x equals mu plus z sigma. z could be |
|
|
|
128 |
|
00:09:52,010 |
|
negative, could be positive, depends on the |
|
|
|
129 |
|
00:09:54,490 |
|
probability you have. If the probability is above |
|
|
|
130 |
|
00:09:59,170 |
|
0.5, I mean 0.5 and greater than 0.5, this |
|
|
|
131 |
|
00:10:03,390 |
|
corresponds to z positive. But if z-score is negative, I'm |
|
|
|
132 |
|
00:10:08,880 |
|
sorry, if z-score is negative, then the |
|
|
|
133 |
|
00:10:10,680 |
|
probability should be smaller than 0.5. So if the |
|
|
|
134 |
|
00:10:14,880 |
|
probability is given less than 0.5, then your z |
|
|
|
135 |
|
00:10:19,360 |
|
-score should be negative, otherwise should be |
|
|
|
136 |
|
00:10:21,100 |
|
positive. So you have to be careful in this case. |
|
|
|
137 |
|
00:10:25,700 |
|
Now look at this example. Let x represent the time |
|
|
|
138 |
|
00:10:31,240 |
|
it takes in seconds to download an image file |
|
|
|
139 |
|
00:10:35,770 |
|
from the internet. The same example as we did |
|
|
|
140 |
|
00:10:39,730 |
|
before. And here we assume that x is normal |
|
|
|
141 |
|
00:10:43,590 |
|
distribution with mean of 8 and standard deviation |
|
|
|
142 |
|
00:10:46,330 |
|
of 5. Now, let's see how can we find the value of |
|
|
|
143 |
|
00:10:51,710 |
|
x such that 20% of download times are smaller than |
|
|
|
144 |
|
00:10:58,050 |
|
x. |
|
|
|
145 |
|
00:11:01,060 |
|
So, this probability is a fraction. Also, always |
|
|
|
146 |
|
00:11:04,580 |
|
the probability is between 0 and 1. So, the |
|
|
|
147 |
|
00:11:07,840 |
|
probability here is 20%. In this case, your z |
|
|
|
148 |
|
00:11:11,820 |
|
-score should be negative. Because 20% is more |
|
|
|
149 |
|
00:11:15,560 |
|
than 0.5. So, z-score should be in this side, in |
|
|
|
150 |
|
00:11:18,660 |
|
the left side. |
|
|
|
151 |
|
00:11:22,340 |
|
So, again, he asks about finding x-value such that |
|
|
|
152 |
|
00:11:26,380 |
|
20%. |
|
|
|
153 |
|
00:11:31,740 |
|
So here again we are looking for this value, for |
|
|
|
154 |
|
00:11:35,400 |
|
the value of x, which is smaller than the area to |
|
|
|
155 |
|
00:11:40,760 |
|
the left of this x, equals 0.2. |
|
|
|
156 |
|
00:11:47,480 |
|
Now, the first step, we have to find the z-score. |
|
|
|
157 |
|
00:11:52,650 |
|
It's backward, z-score first, then x. Find a z |
|
|
|
158 |
|
00:11:56,430 |
|
-score corresponding to the probability of 0.2. |
|
|
|
159 |
|
00:12:02,510 |
|
The approximate one, the near value, I mean, to |
|
|
|
160 |
|
00:12:07,710 |
|
the 0.2 is 0.2005. Sometimes you have the exact |
|
|
|
161 |
|
00:12:12,190 |
|
value from the table you have, but most of the |
|
|
|
162 |
|
00:12:16,570 |
|
time you don't have it. So you have to look at the |
|
|
|
163 |
|
00:12:19,050 |
|
approximate value, which is very close to the one |
|
|
|
164 |
|
00:12:21,790 |
|
you have. So here, we are looking for 0.2. The |
|
|
|
165 |
|
00:12:25,840 |
|
closest value to 0.2 is 0.2005. Now, the |
|
|
|
166 |
|
00:12:30,660 |
|
corresponding value to this probability is minus 0 |
|
|
|
167 |
|
00:12:34,720 |
|
.8 all the way up to 4. So your z-score is |
|
|
|
168 |
|
00:12:40,120 |
|
negative 0.84. So this is the first step. Any |
|
|
|
169 |
|
00:12:47,840 |
|
question? Again. |
|
|
|
170 |
|
00:12:53,950 |
|
Now if we just go back to this equation, |
|
|
|
171 |
|
00:12:59,930 |
|
z equals x minus mu over sigma. A cross |
|
|
|
172 |
|
00:13:03,670 |
|
multiplication, I mean if you multiply both sides |
|
|
|
173 |
|
00:13:07,810 |
|
by sigma, you will get sigma times z equals x |
|
|
|
174 |
|
00:13:16,110 |
|
minus mu. |
|
|
|
175 |
|
00:13:32,120 |
|
Now, in this question, |
|
|
|
176 |
|
00:13:37,960 |
|
he asks about, find the value of x such that 20% |
|
|
|
177 |
|
00:13:43,160 |
|
of download times are less than x. |
|
|
|
178 |
|
00:13:50,740 |
|
Now the probability is less than 0.5, so your z |
|
|
|
179 |
|
00:13:54,080 |
|
-score should be on the left side. So here we need |
|
|
|
180 |
|
00:13:57,780 |
|
to find the value of z first. Go back to the |
|
|
|
181 |
|
00:14:03,660 |
|
normal table you have. |
|
|
|
182 |
|
00:14:07,680 |
|
This is the normal table. |
|
|
|
183 |
|
00:14:16,800 |
|
We are looking for minus 0.2. I'm sorry, we are |
|
|
|
184 |
|
00:14:21,250 |
|
looking for 0.2. So the closest value to 0.2 is |
|
|
|
185 |
|
00:14:28,910 |
|
this one, 0.2005. So this is the closest value. |
|
|
|
186 |
|
00:14:49,630 |
|
So the exact answer is sometimes not given. So the |
|
|
|
187 |
|
00:14:54,470 |
|
approximate one, minus 0.8, all the way up to 4. |
|
|
|
188 |
|
00:15:00,030 |
|
So z-score minus 0.8. Any question? |
|
|
|
189 |
|
00:15:10,330 |
|
So the value of z-score is minus 0.84. So my |
|
|
|
190 |
|
00:15: |
|
|
|
223 |
|
00:19:18,560 |
|
So z-score is 1. So we are looking for the |
|
|
|
224 |
|
00:19:22,660 |
|
probability of z greater than 1. |
|
|
|
225 |
|
00:19:28,340 |
|
1.5. 1.2. 1.5. 1.5. 1.5. So I'm looking for the |
|
|
|
226 |
|
00:19:35,220 |
|
probability of x of z greater than |
|
|
|
227 |
|
00:19:40,980 |
|
1 minus P |
|
|
|
228 |
|
00:19:48,700 |
|
of Z less than or equal to 1.5. Now go back to the |
|
|
|
229 |
|
00:19:52,700 |
|
table. |
|
|
|
230 |
|
00:20:01,540 |
|
Now 1.5 under 0. It's 0.9332. |
|
|
|
231 |
|
00:20:11,410 |
|
So, 1 minus this probability gives 0.0668. |
|
|
|
232 |
|
00:20:21,210 |
|
That's the probability of X greater than 4.4. |
|
|
|
233 |
|
00:20:28,130 |
|
So, the answer is 0.0668. |
|
|
|
234 |
|
00:20:34,870 |
|
Now, for the same question. |
|
|
|
235 |
|
00:20:41,320 |
|
What's the probability that a randomly selected |
|
|
|
236 |
|
00:20:44,380 |
|
fish will weigh between 3 and 5 pounds? |
|
|
|
237 |
|
00:20:49,660 |
|
3? |
|
|
|
238 |
|
00:20:52,940 |
|
Up to 5. So first we have to find the z-score for 3 |
|
|
|
239 |
|
00:21:01,040 |
|
out of 5. For it to be just 3 minus 3.2. Divide by |
|
|
|
240 |
|
00:21:11,260 |
|
0.8 is the first z value. Negative 0.2 divided by |
|
|
|
241 |
|
00:21:17,380 |
|
0.8 minus 0.25. The other one, 5 minus 3.2 divided |
|
|
|
242 |
|
00:21:30,360 |
|
by 0.8. 1 minus 0.8 divided by 0.8 equals |
|
|
|
243 |
|
00:21:42,680 |
|
2.25. |
|
|
|
244 |
|
00:21:50,840 |
|
Just double check this result. So here, the |
|
|
|
245 |
|
00:21:57,020 |
|
probability of X between 3 and 5 equals minus 0 |
|
|
|
246 |
|
00:22:04,520 |
|
.25, smaller than Z, smaller than 2.5. |
|
|
|
247 |
|
00:22:12,650 |
|
So it's transformed from normal distribution to |
|
|
|
248 |
|
00:22:17,210 |
|
standardized normal distribution. So here instead |
|
|
|
249 |
|
00:22:20,750 |
|
of computing the probability of X between three |
|
|
|
250 |
|
00:22:23,530 |
|
and five, we are looking for the probability |
|
|
|
251 |
|
00:22:26,070 |
|
between Z between actually minus. It's minus |
|
|
|
252 |
|
00:22:31,350 |
|
because your value here is smaller than the |
|
|
|
253 |
|
00:22:34,590 |
|
average. The average was 3.2, so it should be |
|
|
|
254 |
|
00:22:37,790 |
|
negative. So z score between minus 0.25 all the |
|
|
|
255 |
|
00:22:42,630 |
|
way up to 2.25. So now, this is the probability we |
|
|
|
256 |
|
00:22:48,150 |
|
are looking for. Zero in the middle minus one |
|
|
|
257 |
|
00:22:56,590 |
|
-fourth to the left of z of zero, mu of zero. And |
|
|
|
258 |
|
00:23:03,610 |
|
this is the value of 2.25. Now we are looking |
|
|
|
259 |
|
00:23:08,560 |
|
actually for this probability. |
|
|
|
260 |
|
00:23:12,960 |
|
The area between minus 0.25 all the way up to 2.5. |
|
|
|
261 |
|
00:23:19,980 |
|
So this area equals the |
|
|
|
262 |
|
00:23:25,200 |
|
probability of Z less than 2.25 minus. |
|
|
|
263 |
|
00:23:34,280 |
|
And again, use the normal. table to give this |
|
|
|
264 |
|
00:23:38,780 |
|
value and another one. |
|
|
|
265 |
|
00:23:42,980 |
|
Any questions? So first step here, we compute the |
|
|
|
266 |
|
00:23:52,400 |
|
z-score for each value x. So the problem is |
|
|
|
267 |
|
00:23:56,880 |
|
transformed from normal distribution to |
|
|
|
268 |
|
00:24:01,380 |
|
standardized normal distribution. So it becomes z |
|
|
|
269 |
|
00:24:05,060 |
|
between minus 1.25 up to 2.25. Now, this area, |
|
|
|
270 |
|
00:24:11,960 |
|
this dashed area equals the area below 2.25 minus |
|
|
|
271 |
|
00:24:19,900 |
|
the area below minus 1.25. Now, by using the |
|
|
|
272 |
|
00:24:25,000 |
|
similar way we did before, you will compute the |
|
|
|
273 |
|
00:24:27,760 |
|
value of z. The probability of z is smaller than 2 |
|
|
|
274 |
|
00:24:30,960 |
|
.25 by using The normal table. So here, 2.2 up to |
|
|
|
275 |
|
00:24:39,580 |
|
5. So 9, 8, 7, 8. 9, 8, 7, 8. |
|
|
|
276 |
|
00:24:53,900 |
|
So the area below 2.25, 2.2, this value. All the |
|
|
|
277 |
|
00:25:00,260 |
|
way up to 5 gives 987. |
|
|
|
278 |
|
00:25:05,540 |
|
Now, what's about the probability of Z smaller |
|
|
|
279 |
|
00:25:08,860 |
|
than minus 0.25? If you go back to the Z table, |
|
|
|
280 |
|
00:25:15,380 |
|
but for the other one, the negative one. |
|
|
|
281 |
|
00:25:23,120 |
|
Minus 2 minus 0.2 up |
|
|
|
282 |
|
00:25:28,540 |
|
to 5. 0.4013 minus, |
|
|
|
283 |
|
00:25:36,620 |
|
that will give the probability between three and |
|
|
|
284 |
|
00:25:43,100 |
|
five. |
|
|
|
285 |
|
00:25:46,180 |
|
This is the second part. |
|
|
|
286 |
|
00:25:51,120 |
|
So the final answer. |
|
|
|
287 |
|
00:26:00,630 |
|
So this is the probability that the selected fish |
|
|
|
288 |
|
00:26:05,450 |
|
will weigh between three and five pounds. |
|
|
|
289 |
|
00:26:11,810 |
|
Now, other question is, for the same problem, you |
|
|
|
290 |
|
00:26:20,770 |
|
said a citation Catfish should be one of the top 2 |
|
|
|
291 |
|
00:26:28,020 |
|
% in the weight. Assuming the weights of catfish |
|
|
|
292 |
|
00:26:33,860 |
|
are normally distributed, at what weight in bounds |
|
|
|
293 |
|
00:26:38,660 |
|
should the citation, the notation be established? |
|
|
|
294 |
|
00:26:45,800 |
|
So in this board, he asked about what's the value |
|
|
|
295 |
|
00:26:50,600 |
|
of x, for example. |
|
|
|
296 |
|
00:26:57,160 |
|
is greater than what value here. And this |
|
|
|
297 |
|
00:27:01,680 |
|
probability equals 2%. Because you said the |
|
|
|
298 |
|
00:27:07,880 |
|
citation catfish should be one of the top 2%. So |
|
|
|
299 |
|
00:27:14,720 |
|
the area in the right here, this area is 2%. |
|
|
|
300 |
|
00:27:26,000 |
|
What's the value of x in this case? So here, the |
|
|
|
301 |
|
00:27:34,080 |
|
value of x greater than a equals 0.02, and we are |
|
|
|
302 |
|
00:27:38,420 |
|
looking for this value. |
|
|
|
303 |
|
00:27:45,750 |
|
gives the area to the left side. So this |
|
|
|
304 |
|
00:27:49,170 |
|
probability actually, the area to the right is 2%, |
|
|
|
305 |
|
00:27:54,490 |
|
so the area to the left is 98%. So this is the |
|
|
|
306 |
|
00:28:00,810 |
|
same as, |
|
|
|
307 |
|
00:28:04,610 |
|
as we know, the equal sign does not matter because |
|
|
|
308 |
|
00:28:07,930 |
|
we have continuous distribution. |
|
|
|
309 |
|
00:28:12,050 |
|
continuous distribution, so equal sign does not |
|
|
|
310 |
|
00:28:14,650 |
|
matter. So now, if you ask about P of X greater |
|
|
|
311 |
|
00:28:18,510 |
|
than a certain value equals a probability of, for |
|
|
|
312 |
|
00:28:22,190 |
|
example, 0.02, you have to find the probability to |
|
|
|
313 |
|
00:28:27,330 |
|
the left, which is 0.98, because our table gives |
|
|
|
314 |
|
00:28:31,890 |
|
the area to the left. Now, we have to find the |
|
|
|
315 |
|
00:28:36,470 |
|
value of A such that Probability of X is more than |
|
|
|
316 |
|
00:28:40,820 |
|
or equal to 0.98. So again, we have to look at the |
|
|
|
317 |
|
00:28:44,820 |
|
normal table, but backwards, because this value is |
|
|
|
318 |
|
00:28:50,140 |
|
given. If the probability is given, we have to |
|
|
|
319 |
|
00:28:53,720 |
|
look inside the body of the table in order to find |
|
|
|
320 |
|
00:28:58,900 |
|
the z-score. |
|
|
|
321 |
|
00:29:03,350 |
|
x equals mu plus z sigma in order to find the |
|
|
|
322 |
|
00:29:07,850 |
|
corresponding value x. So again, go back to the |
|
|
|
323 |
|
00:29:12,290 |
|
normal table, and we are looking for 98%. The |
|
|
|
324 |
|
00:29:22,010 |
|
closest value to 98%, look here, if you stop here |
|
|
|
325 |
|
00:29:27,930 |
|
at 2, go to the right, |
|
|
|
326 |
|
00:29:33,660 |
|
Here we have 9798 or |
|
|
|
327 |
|
00:29:39,380 |
|
9803. |
|
|
|
328 |
|
00:29:42,640 |
|
So the answer might be your z-score could be 2.05 |
|
|
|
329 |
|
00:29:50,460 |
|
or 2.06. So again, in this case, the table does |
|
|
|
330 |
|
00:29:59,440 |
|
not give the exact. So the approximate one might |
|
|
|
331 |
|
00:30:04,400 |
|
be between them exactly. Or just take one of |
|
|
|
332 |
|
00:30:08,920 |
|
these. So either you can take 9798, which is |
|
|
|
333 |
|
00:30:13,140 |
|
closer to 98% than 9803, because it's three |
|
|
|
334 |
|
00:30:19,500 |
|
distant apart. So maybe we can take this value. |
|
|
|
335 |
|
00:30:24,500 |
|
Again, if you take the other one, you will be OK. |
|
|
|
336 |
|
00:30:28,540 |
|
So you take either 2.05. or 2.06. So let's take |
|
|
|
337 |
|
00:30:36,730 |
|
the first value, for example. So my x equals mu, z |
|
|
|
338 |
|
00:30:45,270 |
|
is 2.05, times sigma, 0.8. Multiply 2.05 by 8, 0 |
|
|
|
339 |
|
00:30:56,570 |
|
.8, then add 3.2, you will get What's your answer? |
|
|
|
340 |
|
00:31:08,390 |
|
3.2 plus 2 point... So around 4.8. So your answer |
|
|
|
341 |
|
00:31:17,450 |
|
is 4.84. |
|
|
|
342 |
|
00:31:23,810 |
|
Now if you go back to the problem, and suppose you |
|
|
|
343 |
|
00:31:29,470 |
|
know the value of x. |
|
|
|
344 |
|
00:31:34,250 |
|
So the probability of X. |
|
|
|
345 |
|
00:31:42,990 |
|
Double check to the answer. |
|
|
|
346 |
|
00:31:49,490 |
|
4.84. Just check. V of X greater than this value |
|
|
|
347 |
|
00:31:58,010 |
|
should |
|
|
|
348 |
|
00:32:03,290 |
|
be Two percent. Two percent. So the probability of |
|
|
|
349 |
|
00:32:09,500 |
|
X greater than this value should be equal to one |
|
|
|
350 |
|
00:32:13,440 |
|
zero. So this problem is called backward normal |
|
|
|
351 |
|
00:32:18,980 |
|
calculation because here first step we find the |
|
|
|
352 |
|
00:32:24,960 |
|
value of this score corresponding to this |
|
|
|
353 |
|
00:32:28,040 |
|
probability. Be careful. The probability of X |
|
|
|
354 |
|
00:32:33,420 |
|
greater than 2 is 0.02. So my value here should be |
|
|
|
355 |
|
00:32:38,740 |
|
to the right. Because it says greater than A is |
|
|
|
356 |
|
00:32:44,980 |
|
just 2%. If you switch the position of A, for |
|
|
|
357 |
|
00:32:51,240 |
|
example, if A is on this side, And he asked about |
|
|
|
358 |
|
00:32:57,130 |
|
E of X greater than E is 2%. This area is not 2%. |
|
|
|
359 |
|
00:33:02,850 |
|
From A up to infinity here, this area is not 2% |
|
|
|
360 |
|
00:33:10,050 |
|
because at least it's greater than 0.5. Make |
|
|
|
361 |
|
00:33:14,070 |
|
sense? So your A should be to the right side. |
|
|
|
362 |
|
00:33:17,590 |
|
Because the value of X greater than E, 2% is on |
|
|
|
363 |
|
00:33:21,810 |
|
the other side. Let's move to the next one. |
|
|
|
364 |
|
00:33:36,180 |
|
For the same question. |
|
|
|
365 |
|
00:33:52,790 |
|
Again, the owner of Catfish Market determined the |
|
|
|
366 |
|
00:33:57,150 |
|
average weight of a catfish 3.2 with |
|
|
|
367 |
|
00:34:00,930 |
|
standardization 0.8 and we are assuming the |
|
|
|
368 |
|
00:34:04,670 |
|
weights of catfish are normally distributed, kiosk |
|
|
|
369 |
|
00:34:08,270 |
|
above. Above what weight? Do 89.8% of the weights |
|
|
|
370 |
|
00:34:17,390 |
|
care? |
|
|
|
371 |
|
00:34:20,630 |
|
Above? And above, so x greater than. X minus. And |
|
|
|
372 |
|
00:34:27,980 |
|
98, 89, sorry, 89. So this is a percentage he's |
|
|
|
373 |
|
00:34:34,900 |
|
looking for. 89.8%. Now what's the value of A? Or |
|
|
|
374 |
|
00:34:45,860 |
|
above what weight? Do 89.8% of the weights occur? |
|
|
|
375 |
|
00:34:57,730 |
|
You just make the normal curve in order to |
|
|
|
376 |
|
00:35:02,670 |
|
understand the probability. Now, A should be to |
|
|
|
377 |
|
00:35:07,010 |
|
the right or to the left side? Imagine A in the |
|
|
|
378 |
|
00:35:11,550 |
|
right side here. Do you think the area above A is |
|
|
|
379 |
|
00:35:15,990 |
|
89%? It's smaller than 0.5 for sure. So it should |
|
|
|
380 |
|
00:35:21,670 |
|
be the other side. So this is your 8. Now, this |
|
|
|
381 |
|
00:35:29,950 |
|
area makes sense that it's above 0.5. It's 0.8980. |
|
|
|
382 |
|
00:35:38,750 |
|
Now, B of X greater than equals this value. And |
|
|
|
383 |
|
00:35:42,850 |
|
again, the table gives the area to the left. So |
|
|
|
384 |
|
00:35:46,990 |
|
this is actually X less than A, 1 minus this |
|
|
|
385 |
|
00:35:53,740 |
|
value, equals 0.1020. |
|
|
|
386 |
|
00:36:01,480 |
|
Now go back to |
|
|
|
387 |
|
00:36:08,760 |
|
the normal table. Here it's 0.1020. So it should |
|
|
|
388 |
|
00:36:14,400 |
|
be negative. I mean, your z-scope should be |
|
|
|
389 |
|
00:36:17,500 |
|
negative. |
|
|
|
390 |
|
00:36:22,000 |
|
Now look at 0.102. |
|
|
|
391 |
|
00:36:28,560 |
|
Exactly this value. 0.102 is minus 1.2 up to 7. So |
|
|
|
392 |
|
00:36:38,680 |
|
minus 1.27. |
|
|
|
393 |
|
00:36:49,120 |
|
Minus 1.2. All the way up to 7 is 0.102. So the |
|
|
|
394 |
|
00:36:57,900 |
|
corresponding z-score is minus 1.17. Now x again |
|
|
|
395 |
|
00:37:04,160 |
|
equals mu plus z sigma. |
|
|
|
396 |
|
00:37:10,280 |
|
Mu is 3.2 plus z is negative 1.17 times sigma. |
|
|
|
397 |
|
00:37:24,250 |
|
So it's equal to 3.2 minus 127 times 0.3. By |
|
|
|
398 |
|
00:37:34,390 |
|
calculator, you'll get the final result. |
|
|
|
399 |
|
00:37:51,120 |
|
If the probability is smaller than 0.5, then this |
|
|
|
400 |
|
00:37:56,180 |
|
one is negative. Go to the other one. If the |
|
|
|
401 |
|
00:38:00,480 |
|
probability is above 0.5, then use the positive z |
|
|
|
402 |
|
00:38:04,040 |
|
-score. So what's the answer? 2.18. |
|
|
|
403 |
|
00:38:12,680 |
|
Be careful. In the previous one, We had a |
|
|
|
404 |
|
00:38:19,850 |
|
probability of X greater than A equal 2%. In |
|
|
|
405 |
|
00:38:28,870 |
|
this case, the value of A, for example, is located |
|
|
|
406 |
|
00:38:33,070 |
|
in the upper tail here. |
|
|
|
407 |
|
00:38:40,210 |
|
For this part, you ask about B of X greater than A |
|
|
|
408 |
|
00:38:45,530 |
|
equal 0.89. It's here more than 0.5 |
|
|
|
445 |
|
00:42:54,950 |
|
Below minus 3, I mean smaller than minus 3, or |
|
|
|
446 |
|
00:42:59,010 |
|
above 3, these points are suspected to be |
|
|
|
447 |
|
00:43:04,010 |
|
outliers. |
|
|
|
448 |
|
00:43:09,230 |
|
So any point, any data value smaller than minus 3 |
|
|
|
449 |
|
00:43:16,230 |
|
in this form, or above plus 3 is considered to be |
|
|
|
450 |
|
00:43:22,330 |
|
an outlier. If we compute the lower limit, which |
|
|
|
451 |
|
00:43:30,250 |
|
is Q1 minus 1.5 IQR over the upper limit. |
|
|
|
452 |
|
00:43:40,170 |
|
So we |
|
|
|
453 |
|
00:43:47,490 |
|
have lower limit, upper limit. So lower limit. |
|
|
|
454 |
|
00:43:55,400 |
|
And upper limit. |
|
|
|
455 |
|
00:43:59,980 |
|
Any data point below lower limit or above upper |
|
|
|
456 |
|
00:44:08,420 |
|
limit is considered to be a type. So therefore, we |
|
|
|
457 |
|
00:44:13,080 |
|
have two methods to determine or to examine if the |
|
|
|
458 |
|
00:44:18,320 |
|
observation is enough there or not. One by using |
|
|
|
459 |
|
00:44:20,960 |
|
this score, straightforward. And the other one, we |
|
|
|
460 |
|
00:44:24,060 |
|
have to look for the lower limit and upper limit. |
|
|
|
461 |
|
00:44:28,960 |
|
The other method by using software and later on |
|
|
|
462 |
|
00:44:34,200 |
|
you will take SPSS in order to determine if the |
|
|
|
463 |
|
00:44:39,380 |
|
data is normally distributed by using something |
|
|
|
464 |
|
00:44:43,100 |
|
called QQ plot or normal probability plot. So I'm |
|
|
|
465 |
|
00:44:52,540 |
|
going to skip this part. Because data is taken by |
|
|
|
466 |
|
00:44:56,710 |
|
using software. But in general, |
|
|
|
467 |
|
00:45:04,330 |
|
you may look at this graph. Generally speaking, if |
|
|
|
468 |
|
00:45:11,750 |
|
you have a probability plot of a data, and the |
|
|
|
469 |
|
00:45:17,730 |
|
points lie on a straight line, or close to it, in |
|
|
|
470 |
|
00:45:25,530 |
|
this case, the distribution is normal. It's hard |
|
|
|
471 |
|
00:45:29,650 |
|
to make this graph manual. It's better to use |
|
|
|
472 |
|
00:45:33,390 |
|
software. But at least if we have this graph, and |
|
|
|
473 |
|
00:45:38,510 |
|
the points are close to the straight line. I mean, |
|
|
|
474 |
|
00:45:42,250 |
|
the points are either on the straight line, lies |
|
|
|
475 |
|
00:45:45,950 |
|
on the straight line, or close to it. In this case, |
|
|
|
476 |
|
00:45:50,870 |
|
the data is normally distributed. If the data |
|
|
|
477 |
|
00:45:55,030 |
|
points scattered away of the straight line, then |
|
|
|
478 |
|
00:46:00,870 |
|
the distribution is not normal, either skewed to |
|
|
|
479 |
|
00:46:04,390 |
|
the right or skewed to the left. So for this |
|
|
|
480 |
|
00:46:08,690 |
|
specific graph, the plot is normally distributed, |
|
|
|
481 |
|
00:46:14,290 |
|
approximately normally distributed. Because most |
|
|
|
482 |
|
00:46:17,550 |
|
of the points here lie close to the line and few |
|
|
|
483 |
|
00:46:24,260 |
|
are scattered away. Or it means that there are few |
|
|
|
484 |
|
00:46:29,660 |
|
outliers in this case, we can ignore these values. |
|
|
|
485 |
|
00:46:33,100 |
|
So here the plot is approximately a straight line |
|
|
|
486 |
|
00:46:35,620 |
|
except for a few outliers at the low and the |
|
|
|
487 |
|
00:46:40,360 |
|
right, those points. So generally speaking, the |
|
|
|
488 |
|
00:46:44,780 |
|
distribution is normal distribution. That's all |
|
|
|
489 |
|
00:46:51,200 |
|
for this chapter. |
|
|
|
|