Common Hypotheses
H0 : Β there is no difference in the distribution of a variable across populations or treatments
Ha : Β there is a difference in the distribution of a variable across populations or treatments
Assumptions
to test for independence: Samples are randomly selected using a simple random sample
to test for homogeneity: Samples are randomly selected using a stratified random sample or randomized experiment.
Sampling with replacement (or sample size < 10% N)
All expected counts should be greater than 5
Sampling Distribution
π2
TI Calculator:Β Stat -> Tests ->Β π2-Test...
SPSS: Analyze -> Descriptive Statistics -> Crosstabs
Enter Rowds and Colummn
Choose Statistics... Β Check Chi-Square
Choose Cells... Check Expected
For Frequency Table
Do the following steps first
Data -> Weight Cases...