ABSTRACT
This chapter presents two more examples of multiple regressions, one using three independent variables and the other, four. The paths represent the regression weights and the curved arrows represent the correlations among the independent variables. If students include in the regression a variable that logically is prior to Homework out of School and Grades, but is not a common cause of Homework and Grades, the regression weight for Out-of-School Homework will not change. That is, a variable that affects only Out-of-School Homework but not Grades will not change the Homework regression coefficient. To reiterate, to interpret regression coefficients as the effects of one variable on another, the common causes of the presumed cause and presumed effect must be included in the regression. The chapter provides examples such as Grades on Parent Education, time spent on In-School Homework, and time spent on Out-of-School Homework.
