question_id int64 1 252 | question_type stringclasses 2
values | tables_involved stringclasses 7
values | question stringlengths 187 433 | answer stringlengths 212 935 |
|---|---|---|---|---|
1 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with avg_review_score below 4.0 (sellers table), avg_return_rate_90d above 25% (refunds table), AND avg_deviation_from_median above 40 (products_order_items table), how many sellers are identified? What is their average number of refund events? | 6 sellers satisfy all three cross-table conditions. Their average number of refund events is 2.83. This cohort combines below-average customer satisfaction, elevated return rates, and above-median pricing — three independent risk dimensions drawn from three separate tables, none of which alone would trigger the same al... |
2 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have cancellation_rate > 0% (sellers table), at least one disputed refund claim (refunds table), AND sell across 3 or more distinct product categories (products_order_items table)? What is their average review score? | 3 sellers have all three conditions: active cancellations, disputed claims, and multi-category product spread. Their average review score is 4.11. Multi-category sellers with both cancellations and disputes represent a broadly exposed operational risk profile that cannot be explained by a single SKU or category problem... |
3 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have avg_review_score above 4.2 (sellers table), at least one NOT_AS_DESCRIBED refund event (refunds table), AND products with avg_deviation_from_median above 30 (products_order_items table)? What is their average deviation_from_median? | 6 sellers have high review scores yet show listing-accuracy failures combined with above-median pricing. Their average deviation_from_median is 73.27. This counterintuitive pattern — high satisfaction alongside listing-mismatch refunds and premium pricing — suggests review scores lag actual product quality or buyers di... |
4 | multi-hop reasoning | sellers + refunds + products_order_items | Among high-volume sellers with total_orders above 200 (sellers table), dispute rate above 10% on refund events (refunds table), AND 2 or more product categories (products_order_items table), how many sellers are found? What is their average disputed-claim count? | 6 high-volume multi-category sellers exceed a 10% dispute rate. Their average disputed-claim count is 2.0. Large sellers with above-median pricing accumulate refunds at scale; a 10%+ dispute rate signals that roughly 1 in 10 refund events escalate to formal conflict rather than being resolved quietly. |
5 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have at least one refund_without_return event (refunds table), average product price above 100 (products_order_items table), AND avg_review_score below 4.5 (sellers table)? What is their average refund amount? | 0 sellers combine refund-without-return events with high-price products and sub-4.5 review scores. Their average refund amount is N/A. Issuing refunds without requiring product return at price points above 100 creates direct financial exposure; paired with below-4.5 review scores the seller profile is both financially ... |
6 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have at least one SUSPECTED_COUNTERFEIT refund reason (refunds table), products priced above category median by an average deviation above 20 (products_order_items table), AND fewer than 300 total orders (sellers table)? What is their average max_deviation_from_median? | 1 sellers combine counterfeit signals with above-median pricing and a relatively small order base. Their average max deviation from category median is 34.94. Low-volume sellers with premium pricing and counterfeit flags are consistent with a new or low-velocity fraud actor: sourcing inauthentic goods, pricing them abov... |
7 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with 4 or more distinct refund reason codes (refunds table), average freight value above 20 (products_order_items table), AND avg_review_score below 4.0 (sellers table), how many sellers are identified? What is their average time_to_refund_hours? | 3 sellers show broad refund reason diversity, high freight costs, and below-average reviews. Their average time to refund is 238.36 hours. High freight spend without improving customer outcomes and combined with diverse failure modes across the refund table suggests systemic operational problems that freight investment... |
8 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have a max single-event return_rate_90d above 50% (refunds table), average Price_Dispersion above 50 (products_order_items table), AND exactly zero cancellations (sellers table)? What is their average review score? | 3 sellers show extreme peak return rates and high price dispersion yet zero cancellations. Their average review score is 3.99. Zero cancellations combined with extreme return rates indicates sellers who complete orders (hence no cancellations) but ship products that fail post-delivery — a fulfillment pattern that bypas... |
9 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have at least one DELIVERED_TOO_LATE refund event (refunds table), average freight value below 15 (products_order_items table), avg_review_score below 4.0 (sellers table), AND avg_return_rate_90d above 20%? | 3 sellers combine late-delivery refund events with low freight investment, below-average reviews, and elevated return rates. Low freight cost paired with late-delivery failures is a coherent causal chain: under-investment in shipping leads to unreliable delivery, which cascades into returns and poor satisfaction scores... |
10 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with review_volume above 100 (sellers table), dispute rate above the marketplace average of 12.0% (refunds table), 2+ product categories (products_order_items table), AND avg_return_rate above the marketplace average of 23.3%, how many are identified? What is their average review score? | 4 sellers are above-average on dispute rate and return rate, have high review volume, and span multiple categories. Their average review score is 4.02. Exceeding marketplace average on both dispute and return rate while maintaining high review volume means this segment's quality failures are visible and broadly experie... |
11 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 2 or more DEFECTIVE_ITEM refund events (refunds table), average product price above 80 (products_order_items table), AND more than 100 total orders (sellers table)? What is their average return_rate_90d? | 1 sellers combine repeated defect-based refunds with mid-to-high price points and established order volume. Their average return_rate_90d is 0.18. Recurring defect refunds at higher price points in established sellers indicate a product quality problem that is not being self-corrected — a key signal for mandatory listi... |
12 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 2 or more WRONG_ITEM_SENT refund events (refunds table), sell across at least 3 product categories (products_order_items table), AND have avg_review_score below 4.5 (sellers table)? | 1 sellers show repeated wrong-item dispatch with broad category coverage and sub-4.5 review scores. Multi-category sellers with repeated WRONG_ITEM_SENT events have elevated SKU management complexity: a diverse catalog increases pick-and-pack error rates, and the sub-4.5 review score confirms the problem is customer-vi... |
13 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with avg_review_score below 4.0 (sellers table), at least one DAMAGED_IN_TRANSIT refund event (refunds table), AND average freight value above 15 (products_order_items table), how many sellers are found? What is their average freight value? | 3 sellers combine transit-damage refunds with above-average freight spending and poor review scores. Their average freight value is 23.89. Higher freight cost without preventing damage-in-transit refunds suggests packaging inadequacy or carrier reliability issues that freight surcharges alone cannot address. |
14 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with more than 500 total_orders (sellers table), at least 3 refund events (refunds table), AND avg_deviation_from_median above 20 (products_order_items table), how many are identified? What is their average dispute rate on refunds? | 19 high-volume, above-median-price sellers have at least 3 refund events. Their average dispute rate is 13.0%. Large sellers with above-median pricing accumulate refunds; the dispute rate reveals whether those refunds are resolved quietly or escalate. This cohort requires monitoring of the escalation pathway, not just ... |
15 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 2 or more SIZE_FIT_issue refund events (refunds table), operate primarily in fit-sensitive categories — cama_mesa_banho, bebes, or esporte_lazer — based on top_category (products_order_items table), AND have avg_review_score below 4.2 (sellers table)? | 1 sellers in fit-sensitive categories show recurring size/fit refunds alongside below-average review scores. SIZE_FIT_issue refunds in bedding, baby, and sports categories combined with low review scores indicates persistent listing accuracy failures — size charts or product specifications are inadequate for the catego... |
16 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have a maximum time_to_refund_hours above 500 (refunds table), at least one disputed claim (refunds table), AND average product price above 50 (products_order_items table)? What is their average max_time_to_refund? | 10 sellers combine extremely slow refund processing with at least one formal dispute and mid-to-high price products. Their average max time-to-refund is 625.1 hours. Slow refunds that escalate to formal disputes at higher price points represent the highest financial-exposure scenario: both the amount at risk and the cu... |
17 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have cancellation_rate above 1% (sellers table), 2 or more ITEM_NOT_RECEIVED refund events (refunds table), AND average freight value below 20 (products_order_items table)? | 0 sellers combine active cancellations, repeated non-delivery refund events, and low freight investment. This triple signal points to a fulfillment chain failure: low freight spend → unreliable carriers → items not arriving → customers cancelling and requesting refunds. The causal chain spans from shipping economics (p... |
18 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with fewer than 50 reviews (sellers table), at least 3 refund events (refunds table), AND max product price deviation above 60 above category median (products_order_items table), how many are identified? What is their average review score? | 0 low-review-volume sellers carry high-value products and have generated multiple refund events. Their average review score is N/A. Low review volume with disproportionate refund accumulation and premium pricing creates a thin-feedback risk: the seller has insufficient public signal for buyers to assess quality, while ... |
19 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have at least one case where refund_issued=False (refunds table), avg_review_score below 4.0 (sellers table), AND at least 5 product line items recorded (products_order_items table)? | 2 sellers with an established product presence have at least one unfulfilled refund request alongside below-average reviews. Unfulfilled refunds in active sellers with poor reviews is where customer harm is most likely to persist — the seller continues operating while leaving obligations unresolved. |
20 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with 5 or more distinct refund reason codes (refunds table), more than 150 total orders (sellers table), AND average product price above 60 (products_order_items table), how many are identified? What are their average distinct reason count and average review score? | 9 mid-to-high volume sellers with above-average price points show 5+ distinct failure modes in their refund records. Average distinct reason count: 5.22, average review score: 4.08. Breadth of reason codes at this order volume and price level indicates systemic failures spanning multiple dimensions — no single root cau... |
21 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have avg_return_rate_90d above 30% (refunds table), average product price below 50 (products_order_items table), AND more than 100 total orders (sellers table)? What is their average return rate? | 5 sellers combine above-30% return rates with low-price products and meaningful order volume. Their average return rate is 36.0%. High return rates at low price points in established sellers are particularly concerning: low prices should reduce buyer expectations and thus returns, yet these sellers still generate high ... |
22 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have a dispute rate above 15% on their refunds (refunds table), operate in 2 or more product categories (products_order_items table), AND have avg_review_score between 3.8 and 4.3 (sellers table)? What is their average dispute rate? | 6 mid-range sellers with multi-category presence show above-15% dispute rates. Their average dispute rate is 30.0%. Mid-range review scores mask high dispute rates in these sellers: they pass surface-level review-score screening but have a disproportionate fraction of refund events escalating to formal disputes, repres... |
23 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have at least one SUSPECTED_COUNTERFEIT refund event (refunds table), average product price deviation above 20 (products_order_items table), AND at least one disputed claim (refunds table)? What is their average max_deviation_from_median? | 11 sellers combine counterfeit signals with above-median pricing and formal dispute escalation. Their average max product price deviation is 173.97. This triple signal — counterfeit + premium pricing + disputes — is a convergent fraud indicator: sellers charging above-market prices while generating counterfeit refunds ... |
24 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have more than 300 total_orders (sellers table), avg_return_rate below 15% (refunds table), at least one disputed refund (refunds table), AND average product price above 70 (products_order_items table)? | 1 high-volume, apparently low-risk sellers (return rate below 15%, above-average price) still have at least one disputed claim on record. This finding is important: even a single dispute in an otherwise strong performer warrants investigation — disputes from low-return-rate sellers may signal a concentrated, high-value... |
25 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with 5 or more refund events (refunds table), average time_to_refund_hours above 200 (refunds table), AND average product price deviation below 20 (products_order_items table — indicating fair pricing), how many are identified? What is their average time_to_refund? | 1 fairly-priced sellers with multiple refund events show systematically slow refund processing (304.5 hours average). Slow refunds in sellers who price near category median are not explained by pricing strategy. This pattern suggests internal refund governance failures or deliberate delay — an operational problem indep... |
26 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have review_volume above 50 (sellers table), avg_review_score above 4.0, at least one refund_without_return event (refunds table), AND average product price above 80 (products_order_items table)? | 0 well-reviewed, established, high-price sellers have issued at least one refund without requiring product return. At price points above 80, not requiring returns before issuing refunds creates direct financial exposure per event. The seller's strong public profile makes this a policy-compliance anomaly rather than a q... |
27 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 2 or more DELIVERED_TOO_LATE refund events (refunds table), average freight value above 18 (products_order_items table), AND zero cancellations (sellers table)? What is their average review score? | 1 sellers show repeated late-delivery refunds despite above-average freight investment, with zero cancellations. Their average review score is 3.35. High freight spend without preventing late deliveries, combined with zero cancellations, points to a carrier-side problem: the seller fulfills all orders and invests in sh... |
28 | multi-hop reasoning | sellers + refunds + products_order_items | Among low-volume sellers with fewer than 100 total_orders (sellers table), at least 2 refund events (refunds table), AND a maximum product price above 150 (products_order_items table), how many are identified? What is their average refund amount? | 1 low-volume sellers carry high-value products and have generated multiple refund events. Their average refund amount is 186.17. Low-volume, high-price sellers with multiple refunds carry disproportionate per-seller financial risk — a small number of transactions creates significant refund exposure, making this segment... |
29 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 2 or more CHANGE_OF_MIND refund events (refunds table), avg_return_rate_90d above 35% (refunds table), AND avg_review_score above 4.0 (sellers table)? What is their average return rate? | 0 well-reviewed sellers show high CHANGE_OF_MIND frequency with return rates above 35%. Their average return rate is N/A. CHANGE_OF_MIND refunds are typically coded as low-risk buyer behavior — but when combined with return rates above 35% in sellers with good review scores, it signals systematic return-for-refund abus... |
30 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers fail simultaneously on: avg_review_score below 3.8 (sellers table), avg_return_rate above 30% (refunds table), cancellation_rate > 0% (sellers table), AND 2 or more product categories (products_order_items table)? | 1 sellers fail on all four dimensions simultaneously across three data tables. A fully converged risk profile with no mitigating signals — low reviews, high returns, active cancellations, multi-category exposure — represents the highest-priority investigation cohort in the dataset. |
31 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers whose product catalog (products_order_items table) includes beleza_saude items, how many have at least one disputed refund claim (refunds table)? What is the average return_rate_90d across all refund events for these sellers (sellers table)? | 2 sellers with beleza_saude product activity have at least one disputed refund claim. The average return_rate_90d for their refund events is 22.0%. Beauty and health products carry category-specific return drivers — allergic reactions, efficacy disappointment — that escalate to disputes when not addressed proactively, ... |
32 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 2+ NOT_AS_DESCRIBED refund events, at least 1 WRONG_ITEM_SENT event (both from refunds table), AND average Price_Dispersion above 40 (products_order_items table)? What is their average number of distinct refund reasons? | 1 sellers show combined listing-accuracy failures and fulfillment errors alongside high price dispersion. Their average distinct reason count is 4.0. Co-occurring NOT_AS_DESCRIBED and WRONG_ITEM_SENT events with high price dispersion suggest chaotic SKU management: inconsistent pricing combined with inaccurate descript... |
33 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with total refund value above 500 (refunds table), avg_review_score above 4.0 (sellers table), AND 2+ product categories (products_order_items table), how many are found? What is their average total refund value? | 11 well-reviewed multi-category sellers have accumulated more than 500 in total refund payouts. Their average total refund value is 1324.73. High absolute refund values in otherwise good-performing sellers reflect scale effects: a low refund rate at high order volume still generates significant financial exposure. Abso... |
34 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have a maximum single refund amount above 400 (refunds table), zero disputed claims (refunds table), avg_review_score above 4.1 (sellers table), AND average product price above 50 (products_order_items table)? What is their average max refund amount? | 8 well-priced, well-reviewed sellers process large individual refunds without any formal disputes. Their average max refund amount is 466.38. Large refunds resolved without escalation signal operational maturity: these sellers absorb high-value refund events and resolve them internally. This is a positive pattern, not ... |
35 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have cancellation_rate > 0% (sellers table), at least one DAMAGED_IN_TRANSIT refund event AND at least one DELIVERED_TOO_LATE refund event (refunds table), AND average freight value above 15 (products_order_items table)? | 1 sellers show co-occurrence of cancellations, damage-in-transit, and late-delivery refund types despite above-average freight investment. This four-signal pattern indicates comprehensive logistics failure spanning the order lifecycle: orders are either cancelled, damaged, or late — and higher freight spend is not corr... |
36 | multi-hop reasoning | sellers + refunds + products_order_items | Among high-review-volume sellers (above 200 reviews from sellers table), avg_review_score below 4.0, dispute rate above 5% (refunds table), AND max product price deviation above 30 (products_order_items table), how many are found? What is their average review volume? | 3 high-visibility sellers (avg review volume 740.67) show simultaneously poor review scores, above-average dispute rates, and above-market pricing. These sellers have both the scale and risk signals to generate marketplace-wide reputation damage: high review volume makes their poor metrics visible to many buyers, and a... |
37 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers (across sellers, refunds, and products_order_items tables) have both at least one refund_without_return event AND at least one SUSPECTED_COUNTERFEIT refund reason code? What is their average product category count? | 1 sellers show the refund_without_return + SUSPECTED_COUNTERFEIT co-occurrence. Their average product category count is 1.0. Issuing refunds without requiring return on suspected-counterfeit items suggests the seller is aware of inauthenticity and avoids having physical evidence returned. This combination is rare but r... |
38 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have avg_return_rate_90d above 20% (refunds table), average Price_Dispersion above 60 (products_order_items table), AND avg_review_score above 3.9 (sellers table)? What is their average Price_Dispersion? | 14 sellers combine above-average return rates with very high price dispersion (avg 118.62) but maintain acceptable review scores. High price dispersion means the seller's catalog spans a very wide price range, making consistent quality control difficult. The acceptable review scores alongside elevated returns may refle... |
39 | multi-hop reasoning | sellers + refunds + products_order_items | How many established sellers (total_orders > 100 from sellers table) have at least one unfulfilled refund request (refunds table), avg_review_score below 4.3, AND average product price above 40 (products_order_items table)? | 13 established sellers with mid-range products have at least one unfulfilled refund request combined with below-average review scores. Unfulfilled refunds in established sellers are particularly actionable: sufficient order history exists to understand refund processes, yet at least one obligation remains unresolved. A... |
40 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have 3 or more distinct refund reason codes (refunds table), cancellation_rate > 0% (sellers table), avg_deviation_from_median above 25 (products_order_items table), AND avg_review_score below 4.2? What is their average review score? | 10 sellers fail across all three tables simultaneously: diverse refund failures, active cancellations, above-median pricing, and below-average reviews. Their average review score is 4.02. Broad refund reason diversity combined with cancellations in above-median-price sellers with poor reviews represents the complete co... |
41 | multi-hop reasoning | sellers + refunds + products_order_items | Across all product categories with at least 5 sellers (products_order_items table joined with refunds table via seller_id), which category has the highest average return_rate_90d? What is that rate, and what is the average dispute rate for sellers in that category? Also report the seller-level review score average from... | The category with the highest average return rate (among those with 5+ sellers) is 'eletronicos' with an average return_rate_90d of 34.0%. Sellers in this category have an average dispute rate of 0.0%. This requires pulling product category from products_order_items, return rates from refunds, and seller review scores ... |
42 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category (with at least 5 sellers) has the lowest average return_rate_90d when joining products_order_items with refunds via seller_id? What is its average refund amount and how many distinct sellers operate in it? | The category with the lowest average return rate among those with 5+ sellers is 'consoles_games' with avg return_rate_90d of 13.0%. Average refund amount for sellers in this category: 275.2. Number of sellers: 5. |
43 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category (with at least 5 sellers) has the highest average deviation_from_median across its product prices (products_order_items table)? What is the average return rate for sellers operating in that category (refunds table), and what is their average review score (sellers table)? | The category with the highest average price deviation from median is 'perfumaria' (avg deviation 90.89). Sellers in this category have an average return rate of 22.0% and average refund-event dispute rate of 24.0%. Three-table join required: product category data, refund rates, and seller reviews. |
44 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category (with at least 5 sellers, joining products_order_items and refunds via seller_id) has the highest average dispute rate on refund events? How many sellers operate in this category, and what is their average time_to_refund_hours? | The category with the highest average dispute rate is 'perfumaria' at 24.0%. It has 7 sellers with an average time_to_refund of 179.95 hours. High dispute rates in a category signal either product complexity (buyers unsure of expectations) or seller behavior patterns concentrated in that segment. |
45 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category has the slowest average refund processing time (avg_time_to_refund_hours) when joining products_order_items with refunds via seller_id, restricting to categories with 5+ sellers? What is the average return rate and average price for that category? | The category with the slowest average refund processing is 'automotivo' at 252.25 hours. Its average return rate is 19.0% and average product price is 95.58. Slow refund processing combined with the category's return and pricing profile requires examining whether operational or policy factors drive the delay. |
46 | multi-hop reasoning | sellers + refunds + products_order_items | How many product categories (with 5+ sellers, joining products_order_items and refunds via seller_id) have an average dispute rate above the overall marketplace dispute rate of 12.0%? Name up to five such categories. | 6 product categories have average dispute rates above the marketplace average of 12.0%. Categories include: cama_mesa_banho, informatica_acessorios, moveis_decoracao, perfumaria, relogios_presentes.... Categories with above-average dispute rates signal segments where refund events escalate disproportionately — either d... |
47 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category has the most sellers with avg_review_score below 4.0, when joining products_order_items with sellers via seller_id (requiring at least 5 sellers per category)? How many such sellers are in that category, and what is the category's overall average review score? | The category with the most below-4.0-review sellers is 'cama_mesa_banho' with 23 sellers scoring below 4.0. The category's overall average review score is 3.92. This requires joining products_order_items to sellers on seller_id to compute per-category review distributions — a two-table join that, combined with refund c... |
48 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category has the highest average cancellation_rate when joining products_order_items and sellers via seller_id (5+ sellers per category)? How many sellers in that category have any cancellations, and what is the category average cancellation rate? | The category with the highest average cancellation rate is 'utilidades_domesticas' at 1.0%. 16 out of 12 sellers in this category have at least some cancellations. A category-level view of cancellation rates surfaces systemic fulfillment issues invisible when looking at individual seller metrics. |
49 | multi-hop reasoning | sellers + refunds + products_order_items | Which product category (with at least 5 sellers) has the highest fraction of its sellers appearing in the refunds table, when linking products_order_items to refunds via seller_id? What fraction is that, and how many total sellers operate in that category? | The category where the highest fraction of sellers also have refund records is 'bebes' with 83.0% of its 8 sellers appearing in the refunds table. High refund-presence rates in a category indicate either genuine return activity or a segment where refund requests are common, warranting category-specific return policy re... |
50 | multi-hop reasoning | sellers + refunds + products_order_items | For the informatica_acessorios (computer accessories) category: how many unique sellers operate in it (products_order_items table), what fraction have refund records (refunds table), and what are the average return_rate_90d and average review score (sellers table) for those with refund records? | The informatica_acessorios category has 33 unique sellers. 17 of them (51.52%) have refund records. Among those with refunds: average return_rate_90d = 24.0%, average review score = 3.98, average dispute rate = 15.0%. This three-table profile is required for accurate category health assessment — product table alone onl... |
51 | multi-hop reasoning | sellers + refunds + products_order_items | For the cama_mesa_banho (bed/bath) category: how many sellers operate in it (products_order_items), what is their average review score (sellers table), and what is the average return_rate_90d for those with refund records (refunds table)? | 29 sellers operate in cama_mesa_banho. 17 have refund records. Average review score: 3.85. Average return_rate_90d: 0.25. Average refund amount: 257.18. Average cancellation rate: 1.0%. Average dispute rate: 0.14. Most common refund reason: ITEM_NOT_RECEIVED. Average product price: 120.48, avg deviation from median: 67... |
52 | multi-hop reasoning | sellers + refunds + products_order_items | For the beleza_saude (beauty/health) category: how many sellers operate in it (products_order_items), what is their average review score (sellers table), and what fraction of sellers in this category have at least one disputed refund (refunds table)? | 36 sellers operate in beleza_saude. 21 have refund records. Average review score: 4.02. Average return_rate_90d: 0.21. Average refund amount: 227.62. Average cancellation rate: 0.0%. Average dispute rate: 0.03. Most common refund reason: NOT_AS_DESCRIBED. Average product price: 98.75, avg deviation from median: 60.67. ... |
53 | multi-hop reasoning | sellers + refunds + products_order_items | For the moveis_decoracao (furniture/decoration) category: how many unique sellers operate in it (products_order_items), what is their average product price and average deviation from category median, and what is the average return_rate_90d for those with refund records? | 25 sellers operate in moveis_decoracao. 10 have refund records. Average review score: 4.04. Average return_rate_90d: 0.28. Average refund amount: 257.5. Average cancellation rate: 0.0%. Average dispute rate: 0.09. Most common refund reason: OTHER. Average product price: 99.44, avg deviation from median: 56.62. Furnitur... |
54 | multi-hop reasoning | sellers + refunds + products_order_items | For the esporte_lazer (sports/leisure) category: how many sellers operate in it (products_order_items), and for those sellers who also appear in the refunds table, what is the most common refund reason code and the average return_rate_90d? | 25 sellers operate in esporte_lazer. 13 have refund records. Average review score: 4.15. Average return_rate_90d: 0.18. Average refund amount: 275.39. Average cancellation rate: 0.0%. Average dispute rate: 0.06. Most common refund reason: SIZE_FIT_issue. Average product price: 93.84, avg deviation from median: 55.19. S... |
55 | multi-hop reasoning | sellers + refunds + products_order_items | For the telefonia (mobile/phone) category: how many sellers operate in it (products_order_items), what is their average cancellation_rate (sellers table), and what is the average refund amount for those with refund records (refunds table)? | 20 sellers operate in telefonia. 15 have refund records. Average review score: 4.03. Average return_rate_90d: 0.22. Average refund amount: 228.45. Average cancellation rate: 0.0%. Average dispute rate: 0.12. Most common refund reason: SUSPECTED_COUNTERFEIT. Average product price: 66.03, avg deviation from median: 44.56... |
56 | multi-hop reasoning | sellers + refunds + products_order_items | Compare the cama_mesa_banho and informatica_acessorios categories across three dimensions using three tables: (1) average review score from sellers table, (2) average return_rate_90d from refunds table, and (3) average deviation_from_median from products_order_items table. Which category performs worse on each dimensio... | cama_mesa_banho: avg review score 3.85, avg return rate 25.0%, avg price deviation 67.28. informatica_acessorios: avg review score 3.98, avg return rate 24.0%, avg price deviation 63.53. The category scoring worse on more dimensions represents a higher-risk segment requiring more intensive monitoring. |
57 | multi-hop reasoning | sellers + refunds + products_order_items | Compare the beleza_saude and telefonia categories across three dimensions using three tables: (1) average review score from sellers table, (2) average return_rate_90d from refunds table, and (3) average deviation_from_median from products_order_items table. Which category performs worse on each dimension? | beleza_saude: avg review score 4.02, avg return rate 21.0%, avg price deviation 60.67. telefonia: avg review score 4.03, avg return rate 22.0%, avg price deviation 44.56. The category scoring worse on more dimensions represents a higher-risk segment requiring more intensive monitoring. |
58 | multi-hop reasoning | sellers + refunds + products_order_items | Compare the moveis_decoracao and esporte_lazer categories across three dimensions using three tables: (1) average review score from sellers table, (2) average return_rate_90d from refunds table, and (3) average deviation_from_median from products_order_items table. Which category performs worse on each dimension? | moveis_decoracao: avg review score 4.04, avg return rate 28.0%, avg price deviation 56.62. esporte_lazer: avg review score 4.15, avg return rate 18.0%, avg price deviation 55.19. The category scoring worse on more dimensions represents a higher-risk segment requiring more intensive monitoring. |
59 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have avg_return_rate_90d above 30% (refunds table)? Among those, what are their average product price and average number of product categories (both from products_order_items table), and what is their average review score (sellers table)? | 29 sellers have an average return rate above 30%. Their average product price is 99.22, average category count is 1.34, and average review score is 3.99. Sellers with return rates above 30% that also span multiple categories and price ranges represent a quality problem that cuts across their entire catalog, not a singl... |
60 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have avg_return_rate_90d above 25% (refunds table)? Among those, what are their average product price and average number of product categories (both from products_order_items table), and what is their average review score (sellers table)? | 45 sellers have an average return rate above 25%. Their average product price is 106.62, average category count is 1.47, and average review score is 4.01. Sellers with return rates above 25% that also span multiple categories and price ranges represent a quality problem that cuts across their entire catalog, not a sing... |
61 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have avg_return_rate_90d above 40% (refunds table)? Among those, what are their average product price and average number of product categories (both from products_order_items table), and what is their average review score (sellers table)? | 5 sellers have an average return rate above 40%. Their average product price is 100.62, average category count is 1.6, and average review score is 4.11. Sellers with return rates above 40% that also span multiple categories and price ranges represent a quality problem that cuts across their entire catalog, not a single... |
62 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have a refund dispute rate above 20% (refunds table)? For those sellers, what is their average product price (products_order_items table), average refund count, and average review score (sellers table)? | 23 sellers have a dispute rate above 20% on their refunds. Their average product price is 115.53, average refund count is 3.74, and average review score is 4.03. Dispute rates above 20% indicate a pattern of unresolved conflicts: 1 in 5 refund events escalates to formal dispute, a rate that demands structural intervent... |
63 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers have a refund dispute rate above 15% (refunds table)? For those sellers, what is their average product price (products_order_items table), average refund count, and average review score (sellers table)? | 27 sellers have a dispute rate above 15% on their refunds. Their average product price is 112.48, average refund count is 3.96, and average review score is 4.03. Dispute rates above 15% indicate a pattern of unresolved conflicts: 1 in 6 refund events escalates to formal dispute, a rate that demands structural intervent... |
64 | multi-hop reasoning | sellers + refunds + products_order_items | For sellers operating in technical/tools categories (ferramentas_jardim, automotivo in products_order_items table): how many unique sellers are there, what is their average review score (sellers table), and what is the average return_rate_90d and dispute rate for those with refund records (refunds table)? | 30 sellers operate in technical/tools categories. Average review score: 4.19. Among those with refund records (11 sellers): average return_rate_90d = 18.0%, average dispute rate = 3.0%. Segment-specific profiling across three tables enables targeted category policy rather than one-size-fits-all thresholds. |
65 | multi-hop reasoning | sellers + refunds + products_order_items | For sellers operating in children/family categories (brinquedos, bebes in products_order_items table): how many unique sellers are there, what is their average review score (sellers table), and what is the average return_rate_90d and dispute rate for those with refund records (refunds table)? | 23 sellers operate in children/family categories. Average review score: 4.13. Among those with refund records (13 sellers): average return_rate_90d = 25.0%, average dispute rate = 0.0%. Segment-specific profiling across three tables enables targeted category policy rather than one-size-fits-all thresholds. |
66 | multi-hop reasoning | sellers + refunds + products_order_items | For the relogios_presentes category: what is the total order volume (sellers + products_order_items joined), average review score (sellers table), and average return_rate_90d for sellers with refund records (refunds table)? Is the category's return rate above or below the marketplace average of 23.3%? | Category relogios_presentes: total order volume = 32178, average review score = 4.06, average price = 132.84. Among sellers with refund records: average return_rate_90d = 0.22. Below marketplace average of 23.3%. Three-table join required for this volume-weighted category health view. |
67 | multi-hop reasoning | sellers + refunds + products_order_items | For the cama_mesa_banho category: what is the total order volume (sellers + products_order_items joined), average review score (sellers table), and average return_rate_90d for sellers with refund records (refunds table)? Is the category's return rate above or below the marketplace average of 23.3%? | Category cama_mesa_banho: total order volume = 30977, average review score = 3.92, average price = 90.25. Among sellers with refund records: average return_rate_90d = 0.25. Above marketplace average of 23.3%. Three-table join required for this volume-weighted category health view. |
68 | multi-hop reasoning | sellers + refunds + products_order_items | For the ferramentas_jardim category: what is the total order volume (sellers + products_order_items joined), average review score (sellers table), and average return_rate_90d for sellers with refund records (refunds table)? Is the category's return rate above or below the marketplace average of 23.3%? | Category ferramentas_jardim: total order volume = 26071, average review score = 4.1, average price = 64.22. Among sellers with refund records: average return_rate_90d = 0.19. Below marketplace average of 23.3%. Three-table join required for this volume-weighted category health view. |
69 | multi-hop reasoning | sellers + refunds + products_order_items | Compare high-volume sellers (total_orders > 300, sellers table) vs low-volume sellers (total_orders ≤ 100) on: average return_rate_90d, average dispute rate (refunds table), and average product price deviation (products_order_items table). Which group poses greater normalized refund risk? | High-volume sellers (n=51): avg return rate 24.0%, avg dispute rate 11.0%, avg price deviation 57.08. Low-volume sellers (n=17): avg return rate 20.0%, avg dispute rate 6.0%, avg price deviation 66.69. Normalized risk comparison across three tables reveals whether scale reduces or amplifies per-event risk signals. |
70 | multi-hop reasoning | refunds + products_order_items + orders | How many orders in the orders table have a positive order_delivery_delay (delivered later than estimated)? What is their average delay in days? Cross-reference: among the refund events in the refunds table with refund_reason_code = DELIVERED_TOO_LATE, what is the average return_rate_90d, and what does the mismatch betw... | 35 orders have positive delivery delay. Average delay: 10.77 days. In the refunds table, 59 refund events carry the DELIVERED_TOO_LATE reason, with average return_rate_90d of 27.0%. The mismatch between delayed order count and late-delivery refund count reveals that most delayed orders do NOT generate refund events — t... |
71 | multi-hop reasoning | sellers + refunds + orders | How many orders have order_status = 'canceled' in the orders table? Among sellers with cancellation_rate > 1% (sellers table), how many have at least one disputed refund (refunds table)? What does the co-occurrence of high cancellation rate and disputed refunds suggest about operational failure modes? | 1 orders carry canceled status in the orders table. 13 sellers have cancellation_rate > 1%, and of these, 0 also have at least one disputed refund. Sellers with both active cancellations and disputed refunds have dual failure modes: they cancel orders they cannot fulfill AND face formal disputes on the orders they do a... |
72 | multi-hop reasoning | sellers + refunds + orders | How many orders have order_delivery_delay above 7 days (orders table)? What is the count of DELIVERED_TOO_LATE refund events (refunds table) from sellers with avg_review_score below 4.0 (sellers table)? How does the late-delivery signal differ between the order-level and refund-level tables? | 14 orders were delivered more than 7 days late. 59 total DELIVERED_TOO_LATE refund events exist; among those from sellers with avg_review_score below 4.0, 18 events. Order-level delays measure logistics timing; refund-level late-delivery events measure customer response to delays — the gap between the two quantifies cu... |
73 | multi-hop reasoning | sellers + refunds + orders | For delivered orders (orders table): what is the average order_delivery_delay in days? For DELIVERED_TOO_LATE refund events (refunds table): what is the average time_to_refund_hours? Combining these two metrics with seller review scores (sellers table), what does a high time-to-refund on late-delivery refunds indicate? | Among delivered orders, average delivery delay: -12.26 days. For DELIVERED_TOO_LATE refund events, average time_to_refund: 125.46 hours. A high time-to-refund on late-delivery cases (where the seller is already responsible for the delay) indicates a secondary failure: not only was the product late, but the refund resol... |
74 | multi-hop reasoning | sellers + refunds + orders | How many orders currently have order_status = 'shipped' (orders table)? Among sellers with any canceled_orders > 0 (sellers table), what fraction also have ITEM_NOT_RECEIVED refund events (refunds table)? What does the presence of non-delivered order statuses combined with non-delivery refund events reveal about fulfil... | 12 orders have status 'shipped'. 51 sellers have canceled orders; of these, 11 also have ITEM_NOT_RECEIVED refund events. Non-delivered order statuses and non-delivery refund events measure the same fulfillment failure from different angles: orders table captures the seller's fulfillment state, while refunds table capt... |
75 | multi-hop reasoning | sellers + refunds + orders | How many orders currently have order_status = 'invoiced' (orders table)? Among sellers with any canceled_orders > 0 (sellers table), what fraction also have ITEM_NOT_RECEIVED refund events (refunds table)? What does the presence of non-delivered order statuses combined with non-delivery refund events reveal about fulfi... | 3 orders have status 'invoiced'. 51 sellers have canceled orders; of these, 11 also have ITEM_NOT_RECEIVED refund events. Non-delivered order statuses and non-delivery refund events measure the same fulfillment failure from different angles: orders table captures the seller's fulfillment state, while refunds table capt... |
76 | multi-hop reasoning | sellers + refunds + orders | How many orders currently have order_status = 'processing' (orders table)? Among sellers with any canceled_orders > 0 (sellers table), what fraction also have ITEM_NOT_RECEIVED refund events (refunds table)? What does the presence of non-delivered order statuses combined with non-delivery refund events reveal about ful... | 2 orders have status 'processing'. 51 sellers have canceled orders; of these, 11 also have ITEM_NOT_RECEIVED refund events. Non-delivered order statuses and non-delivery refund events measure the same fulfillment failure from different angles: orders table captures the seller's fulfillment state, while refunds table ca... |
77 | multi-hop reasoning | refunds + products_order_items + orders | How many total delayed orders (order_delivery_delay > 0) are in the orders table? Among sellers in the products_order_items table with average freight_value below 15 who also have DELIVERED_TOO_LATE refund events (refunds table), what is their average delay-related refund return_rate_90d? What does low freight investme... | 35 orders have positive delivery delay. Sellers with avg_freight below 15 AND DELIVERED_TOO_LATE refunds (n=9): their average return_rate_90d across DELIVERED_TOO_LATE events is 0.22. Low freight investment is predictive of late-delivery severity: sellers who under-invest in shipping not only deliver late more often bu... |
78 | multi-hop reasoning | sellers + refunds + orders | How many orders experienced a delivery delay of any (orders table)? Among sellers with DELIVERED_TOO_LATE refund events AND cancellation_rate > 0 (both from sellers + refunds tables), how many sellers have this dual signal? What does the combination of order-level delays, refund-level complaints, and seller-level cance... | 35 orders have delivery delay any. 18 sellers have both DELIVERED_TOO_LATE refund events AND positive cancellation rates. The triple-table combination reveals a seller cohort with comprehensive order fulfillment failure: late deliveries confirmed at the order level, formalized as refund complaints, combined with a patt... |
79 | multi-hop reasoning | sellers + refunds + orders | How many orders experienced a delivery delay of > 7 days (orders table)? Among sellers with DELIVERED_TOO_LATE refund events AND cancellation_rate > 0 (both from sellers + refunds tables), how many sellers have this dual signal? What does the combination of order-level delays, refund-level complaints, and seller-level ... | 14 orders have delivery delay > 7 days. 18 sellers have both DELIVERED_TOO_LATE refund events AND positive cancellation rates. The triple-table combination reveals a seller cohort with comprehensive order fulfillment failure: late deliveries confirmed at the order level, formalized as refund complaints, combined with a... |
80 | multi-hop reasoning | sellers + refunds + orders | How many orders experienced a delivery delay of > 14 days (orders table)? Among sellers with DELIVERED_TOO_LATE refund events AND cancellation_rate > 0 (both from sellers + refunds tables), how many sellers have this dual signal? What does the combination of order-level delays, refund-level complaints, and seller-level... | 9 orders have delivery delay > 14 days. 18 sellers have both DELIVERED_TOO_LATE refund events AND positive cancellation rates. The triple-table combination reveals a seller cohort with comprehensive order fulfillment failure: late deliveries confirmed at the order level, formalized as refund complaints, combined with a... |
81 | multi-hop reasoning | sellers + refunds + orders | What is the ratio of DELIVERED_TOO_LATE refund events (refunds table) to total delayed orders (order_delivery_delay > 0, orders table)? Among sellers with avg_review_score < 4.0 (sellers table), what fraction have DELIVERED_TOO_LATE refund events? Does review score predict late-delivery complaint rate? | Total delayed orders: 35. Total DELIVERED_TOO_LATE refund events: 59. Ratio: 1.69 refund events per delayed order. Among below-4.0-score sellers in the sellers+refunds+products master dataset, 26.0% have at least one DELIVERED_TOO_LATE event vs 21.0% for above-4.0 sellers — confirming review score as a predictor of lat... |
82 | multi-hop reasoning | sellers + refunds + orders | Across all 600 orders in the orders table, what is the distribution of order_delivery_delay (negative = early, positive = late, null = not delivered)? For orders with delay > 0, what is the 90th percentile delay? Cross-referencing with refunds, what fraction of DELIVERED_TOO_LATE refund events come from sellers whose t... | Orders delivered early (delay < 0): 538. On time (delay = 0): 7. Late (delay > 0): 35. Not delivered: 20. 90th percentile delay: 27.2 days. DELIVERED_TOO_LATE refunds from sellers with total_orders > 200: 36.0% — requiring all three tables to compute. |
83 | multi-hop reasoning | sellers + refunds + products_order_items | Compare the bottom 20% of sellers by avg_review_score (below 3.81) vs top 20% (above 4.38) from the sellers table on: (1) average return_rate_90d, (2) average dispute rate, (3) average product price deviation — all requiring refunds and products_order_items tables. Which group shows stronger correlation between review ... | Bottom 20% sellers (n=16): avg return rate 23.0%, avg dispute rate 6.0%, avg price deviation 34.11. Top 20% (n=8): avg return rate 445.0%, avg dispute rate 4.0%, avg price deviation 45.77. The bottom quintile's elevated return and dispute rates confirm review score as a valid cross-table risk proxy — but the magnitude ... |
84 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: avg_review_score < 3.5 (sellers table), avg_return_rate > 0.35 (refunds table), n_reasons >= 3 (products/refunds table), avg_price > 70 (products/refunds table)? Report their average review score and average refund amount. | 1 sellers satisfy all four cross-table conditions. Average review score: 3.4. Average refund amount: 248.68. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
85 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: total_orders > 200 (sellers table), avg_return_rate > 0.2 (refunds table), n_disputed >= 1 (products/refunds table), avg_dispersion > 50 (products/refunds table)? Report their average review score and average refund amount. | 6 sellers satisfy all four cross-table conditions. Average review score: 4.05. Average refund amount: 277.86. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
86 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: cancellation_rate > 0.005 (sellers table), pct_disputed > 0.1 (refunds table), avg_deviation > 30 (products/refunds table), avg_freight > 20 (products/refunds table)? Report their average review score and average refund amount. | 0 sellers satisfy all four cross-table conditions. Average review score: N/A. Average refund amount: N/A. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
87 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: review_volume > 100 (sellers table), avg_return_rate > 0.25 (refunds table), n_categories >= 2 (products/refunds table), max_return_rate > 0.5 (products/refunds table)? Report their average review score and average refund amount. | 5 sellers satisfy all four cross-table conditions. Average review score: 4.05. Average refund amount: 264.29. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
88 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: total_orders > 150 (sellers table), n_counterfeit >= 1 (refunds table), avg_deviation > 25 (products/refunds table), n_disputed >= 1 (products/refunds table)? Report their average review score and average refund amount. | 10 sellers satisfy all four cross-table conditions. Average review score: 4.02. Average refund amount: 266.62. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
89 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: avg_review_score > 4.0 (sellers table), n_refunds >= 3 (refunds table), max_deviation > 60 (products/refunds table), pct_disputed > 0.05 (products/refunds table)? Report their average review score and average refund amount. | 9 sellers satisfy all four cross-table conditions. Average review score: 4.16. Average refund amount: 264.96. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
90 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: cancellation_rate == 0.0 (sellers table), max_return_rate > 0.4 (refunds table), avg_price > 80 (products/refunds table), n_refund_without_ret >= 1 (products/refunds table)? Report their average review score and average refund amount. | 0 sellers satisfy all four cross-table conditions. Average review score: N/A. Average refund amount: N/A. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
91 | multi-hop reasoning | sellers + refunds + products_order_items | How many sellers simultaneously satisfy all four conditions: review_volume > 200 (sellers table), avg_review_score > 3.9 (refunds table), avg_return_rate > 0.2 (products/refunds table), n_late >= 2 (products/refunds table)? Report their average review score and average refund amount. | 0 sellers satisfy all four cross-table conditions. Average review score: N/A. Average refund amount: N/A. Four-signal convergence across three tables is rare by definition; the precise count and profile of this cohort is only computable by retrieving and joining sellers, refunds, and products_order_items data. |
92 | multi-hop reasoning | sellers + refunds + products_order_items | What are the 75th and 90th percentile values of avg_return_rate across sellers in the three-table joined dataset (sellers + refunds + products_order_items)? How many sellers are above each threshold, and what is the average review score for sellers above the 90th percentile? | 75th percentile of avg_return_rate: 0.3 (29 sellers above). 90th percentile: 0.37 (13 sellers above). Average review score of sellers above 90th percentile: 4.1. Percentile-based thresholds derived from the joined dataset are more accurate risk cutoffs than fixed universal thresholds, since they account for the actual ... |
93 | multi-hop reasoning | sellers + refunds + products_order_items | What are the 75th and 90th percentile values of pct_disputed across sellers in the three-table joined dataset (sellers + refunds + products_order_items)? How many sellers are above each threshold, and what is the average review score for sellers above the 90th percentile? | 75th percentile of pct_disputed: 0.0 (124 sellers above). 90th percentile: 0.33 (19 sellers above). Average review score of sellers above 90th percentile: 4.06. Percentile-based thresholds derived from the joined dataset are more accurate risk cutoffs than fixed universal thresholds, since they account for the actual d... |
94 | multi-hop reasoning | sellers + refunds + products_order_items | What are the 75th and 90th percentile values of avg_price across sellers in the three-table joined dataset (sellers + refunds + products_order_items)? How many sellers are above each threshold, and what is the average review score for sellers above the 90th percentile? | 75th percentile of avg_price: 120.67 (31 sellers above). 90th percentile: 195.74 (13 sellers above). Average review score of sellers above 90th percentile: 4.1. Percentile-based thresholds derived from the joined dataset are more accurate risk cutoffs than fixed universal thresholds, since they account for the actual d... |
95 | multi-hop reasoning | sellers + refunds + products_order_items | What are the 75th and 90th percentile values of n_reasons across sellers in the three-table joined dataset (sellers + refunds + products_order_items)? How many sellers are above each threshold, and what is the average review score for sellers above the 90th percentile? | 75th percentile of n_reasons: 3.0 (35 sellers above). 90th percentile: 4.0 (23 sellers above). Average review score of sellers above 90th percentile: 4.01. Percentile-based thresholds derived from the joined dataset are more accurate risk cutoffs than fixed universal thresholds, since they account for the actual distri... |
96 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with at least one SUSPECTED_COUNTERFEIT refund event (refunds table), what is their average review score (sellers table) and average product price and deviation_from_median (products_order_items table)? How does this cohort compare to the overall dataset average on these three metrics? | 23 sellers have at least one SUSPECTED_COUNTERFEIT refund event. Their profile: avg review score 4.0 (vs dataset avg 4.0541), avg product price 109.19, avg deviation from category median 53.69. Below-average review scores for this reason code, combined with above price deviations, supports the hypothesis that SUSPECTED... |
97 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with at least one ITEM_NOT_RECEIVED refund event (refunds table), what is their average review score (sellers table) and average product price and deviation_from_median (products_order_items table)? How does this cohort compare to the overall dataset average on these three metrics? | 25 sellers have at least one ITEM_NOT_RECEIVED refund event. Their profile: avg review score 4.1 (vs dataset avg 4.0541), avg product price 88.21, avg deviation from category median 47.21. Above-average review scores for this reason code, combined with above price deviations, does not confirm the hypothesis that ITEM_N... |
98 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with at least one NOT_AS_DESCRIBED refund event (refunds table), what is their average review score (sellers table) and average product price and deviation_from_median (products_order_items table)? How does this cohort compare to the overall dataset average on these three metrics? | 32 sellers have at least one NOT_AS_DESCRIBED refund event. Their profile: avg review score 4.03 (vs dataset avg 4.0541), avg product price 108.36, avg deviation from category median 58.97. Below-average review scores for this reason code, combined with above price deviations, supports the hypothesis that NOT_AS_DESCRI... |
99 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with at least one DEFECTIVE_ITEM refund event (refunds table), what is their average review score (sellers table) and average product price and deviation_from_median (products_order_items table)? How does this cohort compare to the overall dataset average on these three metrics? | 29 sellers have at least one DEFECTIVE_ITEM refund event. Their profile: avg review score 4.02 (vs dataset avg 4.0541), avg product price 89.06, avg deviation from category median 62.33. Below-average review scores for this reason code, combined with above price deviations, supports the hypothesis that DEFECTIVE_ITEM i... |
100 | multi-hop reasoning | sellers + refunds + products_order_items | Among sellers with at least one WRONG_ITEM_SENT refund event (refunds table), what is their average review score (sellers table) and average product price and deviation_from_median (products_order_items table)? How does this cohort compare to the overall dataset average on these three metrics? | 26 sellers have at least one WRONG_ITEM_SENT refund event. Their profile: avg review score 4.03 (vs dataset avg 4.0541), avg product price 96.06, avg deviation from category median 59.03. Below-average review scores for this reason code, combined with above price deviations, supports the hypothesis that WRONG_ITEM_SENT... |
End of preview. Expand in Data Studio
README.md exists but content is empty.
- Downloads last month
- 7