ABSTRACT

This chapter discusses such interactions, and regression in which there is a curve in the regression line. Such curves can be considered cases in which a variable interacts with itself in its effect on some outcome variable. There is little difference between testing an interaction between two continuous variables and testing an interaction between a categorical and continuous variable. Although the probing of a statistically significant interaction is slightly more complex when both variables are continuous, the basic steps are the same. With interactions between categorical and continuous variables, it is relatively easy to probe the interaction through graphing, because one variable already represents a limited number of categories. To reiterate, there are several possibilities for exploring, through graphing, a statistically significant interaction with two continuous variables. The chapter examines in several ways the effect of homework on achievement and grades.