The Generalized Linear Model with a Continuous Outcome

a) SAS/Stata Tasks.

i) Use the SAS genmod and Stata glm commands to estimate a generalized linear model that regresses bmiR on age assuming an identity link and normally distributed errors. Call this Regression #1.

ii) Use the SAS genmod and Stata glm command to estimate an “intercept only” generalized linear model with bmiR as the outcome, assuming an identity link and normally distributed errors. Call this Regression #2.

iii) Use the SAS genmod and Stata glm command to estimate a generalized linear model that regresses bmiR on age and exfreqwR assuming an identity link and normally distributed errors. Call this Regression #3.

iv) Compare Regression #3 to Regression #2 using the Wald chi-square test and the likelihood ratio test.

b) Write-Up Tasks.

i) Compare the coeffi cient estimates in Regression #1 to the results you obtained in the chapter exercise to Chapter 8 . What do you see?

ii) Compare the coeffi cient estimates in Regression #3 to the results you obtained in the chapter exercise to Chapter 9 . What do you see?

iii) Show how to calculate by hand the z -test and the Wald chi-square test for the hypothesis that β1 = 0 in Regression #1.