A A A

When should I use multiple regression to control the differences in groups?