Chi-square : GF = , GF compare, Independence


1. All use frequency counts as the data

2. All use the same chi-square formula

3. Once you have your O and E tables, the rest of the procedure is the same


1. Each addresses a different question

2. Thus hypotheses differ in form

3. Df formula different for GF (C-1) and IND (C-1)(R-1). Note: for GF, C = cells or categories. For IND, C= number of columns, R = number of rows

4. **How you generate expected frequencies differs**

GF = : divide 100% evenly, then use pn

GF compare : use % from comparion pop., then use pn

IND : calculate column and row totals, then use fe = fc fr / n formula