chapter 15 - Chi Square Tests

Chi-square goodness-of-fit test

Used when data are frequencies. Tests whether observed frequencies match expected frequencies

Observed frequency

actual data count

Expected frequency

predicted count if the null hypothesis is true

Chi-Square

non-numerical tests and categorical variables

Multiple Regression

Predicts Y from multiple predictors (X1, X2, X3…)

Partial regression coefficient

How much a predictor affects Y after controlling for other predictors

R-squared

Proportion of variance in Y explained by all predictors together

Why you can’t add R-squared values

Because predictors often overlap in the variance they explain

Multicollinearity

Predictors are too highly correlated with each other leading to unstable regression