206 women and 88 men enrolled in classes requiring different amounts o
f physical activity were administered the Personal Incentives for Exer
cise Questionnaire. A two-way multivariate analysis of variance, with
course type and gender as the categorical independent variables and 10
subscale scores representing incentives to exercise as the multivaria
te dependent variables, was completed. Canonical discriminant analysis
was used to identify which of the incentives is most useful in discri
minating among participants in active and less active classes or betwe
en men and women. For main effects, analysis indicated an over-all sig
nificant difference for both course type and gender. Men scored higher
than women on activity and on competition, for example, while women s
cored higher on appearance and weight management.