← Return to all tools

No sign-in · no analytics · no saved or transmitted data

Chapter 11 · Counts, association, risk

Categorical Risk Lab

Test categorical patterns with chi-square, inspect which cells drive the result, and translate association into transparent business risk language.

1 Predict
2 Explore
3 Explain

Observed counts

Two customer groups are cross-classified by whether a service failure occurred. Edit the counts to test another case.

GroupFailureNo failure
New process
Current process

Expected-count check

Association readout

Chi-square
p-value
Relative risk

Cell contributions

Large contributions identify cells that depart most from independence. They do not by themselves explain why.

Canvas focus question: Report the two observed risks, relative risk, statistical conclusion, and one sentence that avoids a causal overclaim.

Biblical/Ethical Stewardship and Integrity: Use transparent denominators, avoid fear-inducing language, and consider the people represented by every count. Association should never be presented as proof of causation.