← Spearman All techniques Next: Multiple Regression →

18 — Chi-square Test of Independence

Two categorical variables. Are they linked, or do they vary independently? Compare observed counts in each cell to the count you'd expect if rows and columns were unrelated: E = (row total × col total) / grand total.

χ² statistic
df = (r−1)(c−1)
p-value
Decision (α=0.05)
OBSERVED & EXPECTED COUNTS
(O−E)²/E CONTRIBUTIONS
50%
Reading the cells: when association is 0, observed ≈ expected, all contributions near zero, χ² is small. As association grows, certain cells deviate more (big orange numbers) and χ² climbs.