About crosstabs.com

A cross-tabulation (or contingency table) shows how two categorical variables relate. crosstabs.com builds these tables from your CSV or XLSX file and computes the statistics researchers actually report: chi-square, Fisher's exact test, Cramér's V, phi, the odds ratio, and a range of ordinal and proportional-reduction-in-error measures. The full list of calculators and guides covers each one with formulas and worked examples.

Most cross-tabulation tools are paywalled, install-heavy, or built for specialists (SPSS, GraphPad, R). crosstabs.com is the opposite: instant, browser-based, and free. Students, analysts, journalists, and researchers can run a proper test without a license or a download — and because nothing is gated, it is easy to recommend and easy to cite.

Browser analysis computes statistics locally using published formulas; hosted MCP runs deterministic analysis on Crosstabs servers after explicit upload consent. The calculations are covered by an automated test suite, including exact hand-computed values and naive-reference cross-checks. Selected core cases are also compared publicly with SciPy; this is evidence for those cases, not a claim of blanket equivalence with every statistical package. Where an assumption matters — small expected counts, Fisher's exact applying only to 2×2 tables, weighting caveats — the workspace surfaces a warning rather than hiding it.

• Formulas follow standard references (Pearson chi-square, Fisher's exact via the hypergeometric distribution, Cramér's V, and the usual ordinal measures).
• Selected Pearson, Fisher, Cramér's V, and Somers' d cases are checked against a public SciPy reference manifest.
• Assumption violations are flagged in the workspace and in exports.

See the methods page for calculation behavior and review status, and the validation page for the exact parity scope and limitations.

Browser analysis and local MCP keep source files on your device. The optional hosted MCP receives files only through an explicitly consented upload workflow and processes temporary projects on Crosstabs servers. Workspace access expires 24 hours after creation; active-data cleanup and provider backup retention are separate. Keep the secret workspace token private and export needed work before expiry. Crosstabs does not run a generative AI model; your chosen AI provider has its own data policies. See privacy and data processing for details.

Start with a worked crosstab example with downloadable illustrative counts, calculations, interpretation, and limitations. For an AI workflow, follow the upload-free MCP demonstration before deciding whether to upload your own data.

Named independent statistical review remains pending. Qualified reviewers can contact support@crosstabs.com to discuss a scoped review of the published methods and reference cases. Any completed review should identify its reviewer, date, tested version, findings, and exclusions.

• Students learning the chi-square test, effect sizes, and how to read a crosstab.
• Survey and market researchers cross-tabulating responses with banners and weights.
• Analysts and journalists who need a quick, defensible categorical analysis.
• Educators who want a free tool to share with a class.

Questions, corrections, or a formula you'd like double-checked? Email support@crosstabs.com. The reporting checklist is on the support page.