The dataset StudentSurvey, introduced on page 4, contains information on hours of Exercise per week and GPA. Here we use a slightly modified version called GPAGender which eliminates missing values (leaving n = 343 students) and codes the gender with 1 for males and 0 for females in a new GenderCode variable. This allows us to use information on gender in a regression model.
(a) Test for an association between Exercise and GPA using the data in GPAGender. Give the p-value and make a conclusion in context.
(b) Tests for difference in means reveal that gender is significantly associated with both GPA and Exercise (males have lower GPAs and exercise more on average), so gender may be a confounding variable in the association between Exercise and GPA. Use multiple regression to determine whether Exercise is a significant predictor of GPA, even after accounting for gender as coded in GenderCode.

Dataset StudentSurvey, introduced on page 4

For several years, a first-day survey has been administered to students in an introductory statistics class at one university. Some of the data for a few of the students are displayed in Table 1.1. A more complete table with data for 362 students and 17 variables can be found in the file StudentSurvey.