Homework #4 STAT 505 Fall 2011

Due date: Sept 30, 2011
Write up a 2-3 page report on the FEV data. The variables (listed in line 1) are:
age
in years. Note that these are kids.
fev
continuous measure (liters) of forced expiration volume. It measures the size and health of a person's lungs. Larger numbers are preferred.
height
continuous measure (inches)
sex
discrete/nominal (Female coded 0, Male coded 1)
smoke
discrete/nominal (Nonsmoker coded 0, Smoker coded 1)
The big question: "Is there an effect of smoking on forced expiration volume after adjusting for other variables?" Secondly, "Does the effect (or lack thereof) depend on gender?"

I expect to see plots in which you explore the data and discussion of the plots. I want you to think about the context and explain it to me. Why is height positively correlated with fev? Why are smokers generally taller than non-smokers?

You will need to build a reasonable model for fev based on the other variables including possible interactions and possible curvature and justify your choice. Check diagnostic plots for possible problems. If you see a problem with heteroscedasticity, address it with a transformation.

Write up your findings as a report in a word processor or using LaTeX. Include as an appendix the computer code for fitting the final model in R. Your exploratory work can be done in either R or SAS. Include that in the appendix as well. Plots should be incorporated into the report text, not all at the end, but close to where you discuss them. Refer to the computer output as needed to answer the above questions and justify any conclusions, but do not include any computer code or output in the main report body. Use courier (or another fixed-width) font for all computer input/output.


Author: Jim Robison-Cox
Last Updated: