Conduct each of the parent involvement analyses reported in this chapter, using the NELS data. The variables, as listed in NELS, are: Underrepresented Minority = URM;

Family Background = BySES; Previous Achievement = ByTests; Parent Involvement =

Par_Inv; and GPA = FfuGrad. Compare your results to mine.

a. Make sure you understand what happens when a common cause is omitted versus a simple cause of only one of the variables of interest (Figures 13.3 through 13.4). Is Family Background a common cause or a simple cause of Parent Involvement and GPA? Try deleting it from the model; what happens to the path from Involvement to GPA?

b. Analyze a model without Parent Involvement. Calculate direct, total, and indirect effects for each variable on GPA. Do the same for the model shown in Figure 13.3.

Compare the tables of direct, indirect, and total effects.

c. Analyze a model like Figure 13.3, but in which a path is drawn from URM to Family Background. Now analyze a model in which the path is drawn from Family Background to URM. Which model is correct? How did you make this decision? What effect, if any, did this change in direction have on the estimate of the effect of Parent Involvement on GPA?