Download the Data — Aggregated Cookware Datasets
Free, reproducible aggregates behind this site: brand scores, type x problem rates, material effects, durability trajectories and every mined finding.
Every number on this site is computed offline in Python from aggregated Amazon review data (2023 vintage). So that the work is reproducible and citable, the aggregates themselves are published here as plain CSV/JSON.
License: CC BY 4.0 (aggregates only; cite AMZ Reviews Data / www.amzreviewsdata.com )
Files
| File | Rows | Size |
|---|---|---|
brand-scores.csv
|
3,166 | 200 KB |
type-problem-matrix.csv
|
93 | 9 KB |
material-effects.csv
|
38 | 4 KB |
durability-by-material.csv
|
5 | 0 KB |
durability.json
|
20 | 45 KB |
findings.csv
|
137 | 110 KB |
site-facts.json
|
1 | 4 KB |
What is in each file
- brand-scores.csv — one row per brand: the six review-derived dimensions, product count, sampled reviews and whether the brand passed the minimum-sample threshold.
- type-problem-matrix.csv — for every cookware type and problem we measure: the type’s mention rate vs the rest of the catalog, the ratio and the z-score.
- material-effects.csv — material-vs-material differences on a specific problem, with rates, sample sizes and adjusted p-values.
- durability-by-material.csv — how far ratings drift from early to late reviews, per material.
- findings.csv — every mined finding with its claim, metric, sample and limitations.
- site-facts.json — the single source of the corpus numbers quoted on the site.
Method and limits
The method is published on the scoring standards page, and the limits — 2023 vintage, self-selection, keyword labelling, association rather than causation — are on Data Sources & Limitations . Read them before reusing the data.
- The corpus is {< fact “total_reviews” >} review records across {< fact “catalog_products” >} products.
- Aggregates only. We never distribute raw reviews, and no reviewer identity is present.
- Rates are the share of sampled reviews that mention a problem, not a measured failure rate.
Citation
AMZ Reviews Data, Aggregated cookware review datasets, 2023 vintage, https://www.amzreviewsdata.com/data/
Corrections
If a file looks wrong, tell us via Contact ; corrections are logged on the changelog .