The discussion on EDSTAT-L of the regression model by Greg Adams has been
very interesting. I would suggest that a Poisson regression model might be
more appropriate here than a simple linear regression model, because the
dependent variable (the number of votes for Buchanan) is a count. The
Poisson model would also avoid the embarrassing possibility of predicting a
negative vote count in some of the smaller counties. The Poisson model could
incorporate an offset equal to the total number of votes in the county. You
might also want to allow for a random effect from county to county. If I get
a few spare moments, I might try that myself. I suspect that it will come to
the same general conclusion that the linear regression model and Robert
Dawson's model on the log counts.
Steve Simon, [EMAIL PROTECTED], Standard Disclaimer.
STATS: STeve's Attempt to Teach Statistics. http://www.cmh.edu/stats
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
