Chi-Square Test of Association
Tests an r×c cross table for association: expected counts, contributions, Pearson χ² and likelihood ratio G².
Description
Expected counts E = row total · column total / n. Pearson χ² = Σ (O − E)²/E, G² = 2 Σ O·ln(O/E), both with (r − 1)(c − 1) degrees of freedom. Rows and columns without observations are dropped. Warning when an expected count is below 1 or more than 20 % are below 5. For 2×2 also Yates' correction and Fisher's exact test.
Formulas
Assumptions
- Independent observations, each in exactly one cell
- Expected counts ≥ 1 and at most 20 % below 5
Limitations
- Asymptotic test; unreliable for small expected counts
References
- Minitab Methods and Formulas: Chi-Square Test for Association
- Melzer, A. (2019), Six Sigma – Kompakt und praxisnah, §9.11