Multiple Regression Model Selection techniques

Shivi82 — Wed, 05 Nov 2014 04:56:56 GMT

Hi,

I am building a multiple logistic regression model in sas. The model is significant after consulting the concordant and C statisitc value. Also the other statisitc such as discordant, Somer's D, multicollinearity, AIC are under the allowed limits.

The residuals also meet the assumptions of the model. However I have a question - do i still need to use the model selection techniques - forward, backward or stepwise regression. What i have learnt so far from reading literature is that these techniques could slow down the modeling process.

Could you please advice under what circumstances it is best to use these selection techniques and should there be a minimum number of independent variables while doing so.

Thanks you. Shivi

Re: Multiple Regression Model Selection techniques

PaigeMiller — Wed, 05 Nov 2014 14:06:36 GMT

Residuals and all the tests you mentioned really don't address the issue of whether you need to add other variables into the model. That is an entirely separate question. These residuals and tests tell you things about the model you have fit, and you may or may not be satisfied with the model. They do not tell you if another variable would make a better model.

If a particular modeling technique is needed, then you ought to use it regardless of the fact that it might slow down the modeling process.

Could you please advice under what circumstances it is best to use these selection techniques and should there be a minimum number of independent variables while doing so.

I'm not a fan of forward backward and stepwise, I prefer something like Partial Least Squares, even in the logistic case (although I don't think you can actually do that in SAS without writing your own code). The minimum number of independent variables changes from problem to problem.

So to summarize, the only way to know if you should add additional variables into your model is to actually try it and see what the results are.

topic Multiple Regression Model Selection techniques in SAS Procedures

Multiple Regression Model Selection techniques

Re: Multiple Regression Model Selection techniques