How to use DeepStats

Arrange your dataset so each row is one observation and each column is one variable. Save it as CSV or XLSX. Missing cells can stay empty — DeepStats will flag them during validation. If you don't have a file yet, you can paste rows directly in the next step.

Open the analyzer and either drag your file into the upload area or paste cells into the built-in spreadsheet. Your data stays in the grid throughout the session so you can edit values or add columns before running an analysis.

Mark every column as metric (continuous numeric), nominal (unordered categories) or ordinal (ranked categories). This controls which tests are available — a t-test, for example, only shows up when you have a metric and a nominal variable to compare.

Browse the catalog by category or search by name. Assign variables to the roles the test requires (group, outcome, covariates, etc.). If assumptions are likely to be violated — non-normality, unequal variances, small samples — DeepStats will flag it before you run.

Each analysis returns a results table with the statistics you need to report, a chart appropriate to the method, and a plain-English interpretation explaining what the numbers mean. Reference values, degrees of freedom and effect sizes are all included where applicable.

Download the results table as CSV, save the chart as an image, or copy the interpretation into your report. With a free account you can save the full analysis to your history and revisit it later.