Chi-Square and Categorical Data Walkthrough
Test relationships between categories.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Walk me through analysing categorical data. Table of counts: [COUNTS_TABLE]. Research question: [QUESTION]. Explain the idea of expected counts and compute them for my table, then compute the chi-square statistic step by step, the degrees of freedom, and the p-value using a tool I can verify. Check the assumptions, such as minimum expected counts, and suggest an alternative when they fail. Add an effect size measure and a plot of standardised residuals to show where the differences are. Write a sentence for a report that avoids overclaiming and explain what the test does not show. Add a worked example with an invented two-by-three table, showing every calculation in order, and the code that confirms it. Include a note on paired categorical data and when a different test is needed, and a short description of how to report an odds ratio or a risk difference for a two-by-two table. Also give a short guide to reading the residual plot and a warning about interpreting cells with small counts.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.