1. Reproduce the hot flash analyses used in this chapter, both the dummy variable model and the MG-MACS models. Make sure your results match mine. Are there additional models you might test?   2....

1. Reproduce the hot flash analyses used in this chapter, both the dummy variable model and the MG-MACS models. Make sure your results match mine. Are there additional models you might test?

2. Figure 19.25 shows a starting model for a MG-MACS analysis of the effect of Homework on 12^th
grade GPA. The model for boys is shown. A starting model for Amos (with variable names but without cross group constraints) is available on the website (www. tzkeith.com). Also on the website are the raw data for analysis (homework means.sav). Further develop this model so that you can conduct a MG-MACS analysis (add a group for girls, make the correct cross-group constraints). In the initial model constrain latent means/intercepts for Family Background, Previous Achievement, Homework, and

Grades to zero for both groups. In a second model allow the Homework and Grades intercepts to vary for girls. Do girls have significantly higher or lower levels of true homework (once other variables are controlled)? Grades? In a third model allow the effect of Homework on Grades to vary across groups. Does homework have the same effect on grades for boys and girls, or does the effect of homework depend on sex? Interpret your findings. Make sure you answer the questions asked in the preceding

paragraph