ORIE 576 - SPRING 2004

FINAL EXAM - DUE May 12
-----------------------

1. Find a multivariate dataset suitable for a regression analysis in
   which the response variable is either binary or a count. 
   The dataset must not be taken from the course textbook.
   However, it can be from any other statistics text, from a journal,
   or from data sources on the internet. The data set must include at
   least two independent variables (predictors) not including
   polynomial and interaction terms. When you find a suitable data set,
   send me an email with a brief description of it, and a reference
   for its source. If two students choose the same data, the first one
   to email me will have priority.

2. Analyze the data using logistic regression, or a Poisson or
   negative binomial loglinear model. Consider the possibility of
   interaction terms and, if possible, discuss lack-of-fit and model
   diagnostics. 

3. Type a short report summarizing your findings. Your report should
   be self-contained, including a description of the data, the type of
   models considered, statistical methods used, and your results. Try
   to tell the story of the data. For example, what does your final
   model say about the relationships between the response and
   predictors? The target readership for your report should be your
   classmates; i.e. someone who is not familiar with your data, but
   has a similar knowledge of statistics to yourself. A typical report
   should be approximately three to four pages, including graphs.
   More credit will be given to reports that are interesting, 
   well organized, and easy to follow. 

   SUMMARIZE ONLY RELEVANT STATISTICAL RESULTS IN YOUR REPORT. 
   DO NOT INCLUDE IRRELEVANT MATERIAL, OR PAGES OF OUTPUT THAT CAN 
   BE EXPLAINED IN A FEW SENTENCES.

   IT IS ALMOST NEVER APPROPRIATE TO CUT-AND-PASTE OUTPUT INTO A
   REPORT. FOR EXAMPLE, STATISTICS ON THE OUTPUT ARE TYPICALLY
   REPORTED WITH FAR MORE ACCURACY (DECIMAL PLACES) THAN YOU NEED. 

   YOU SHOULD ATTACH YOUR CODE AND OUTPUT AS AN APPENDIX.
   HOWEVER, DO NOT INCLUDE ANY CODE IN YOUR REPORT. AN EXCEPTION 
   TO THIS WOULD BE IF YOUR GOAL WAS TO SHOW HOW TO USE THE
   SOFTWARE. HOWEVER, THAT IS NOT THE CASE HERE.