Log in to SAS on Demand for Academics.
run format20.sas in the shared folder
/home/u49612056/my_shared_file_links/u49612056/BRFSS/
run 9_21_playing with brfss.data.sas
Model weight = height. (reduced model)
Model weight = height sex. (full model)
Which one is the better model? Explain.
The variables are htin4, wtkg3 and sexvar.
Write out the full model in terms of the fitted coefficients (including beta nought, beta sub 1).
If a woman is 65 inches tall, what is the predicted weight, based on this dataset?
If a man is 65 inches tall, what is the predicted weight?
Is the model homoskedastic or heteroskedastic? Discuss.
Identify 5 outliers and discuss their effect on the model.
