Re: Negative Adjusted R Square is a "good" thing?

Posted by Bruce Weaver on Feb 09, 2014; 11:44pm
URL: http://spssx-discussion.165.s1.nabble.com/Negative-Adjusted-R-Square-is-a-good-thing-tp5724399p5724406.html

Here are some notes on the number of explanatory variables in a linear regression model:

   http://www.angelfire.com/wv/bwhomedir/notes/linreg_rule_of_thumb.txt

Re your 3 years of data, the 3 data points for a given state would not be independent of each other, and your analysis would have to take that into account.  So it could not be an OLS regression model.  Two methods you could consider to handle those dependencies are 1) generalized estimating equations (GEE) or 2) a multilevel model, with years clustered within states.  

cynicalflyer wrote

A followup: my data is a set of scores from 3 years. I was going to average them, but if I treated each individually I would get 129 observations which under the rule of 10 would allow for 13 IVs, yes?