X^2 Goodness-of-fit test
= a hypothesis test that works with independent categorical (nominal) data of sufficient size (>5)
How?
- Formulate 💭 Statistical Hypothesis pair
- → H0: P = (.25,.25,.25,.25); H1: P ≠ (.25,.25,.25,.25);
- Construct a test statistic that measures the closeness to H0
values \ k variables | a | b | c | d |
---|---|---|---|---|
expected frequency Ei | 50 | 50 | 50 | 50 |
observed frequency Oi | 35 | 51 | 64 | 50 |
difference Oi - Ei | -15 | 1 | 14 | 0 |
- Calculate chi-saquared statistics, look up critical value
- → X^2: 8.44; df: 4-1 = 3; Crit value: 7.81
- Check if the result is above the critical value
- (Calculate [[Effect Size#Cramèr’s V|💥 Effect Size#Cramèr’s V]])