Re: Stepwise versus Enter method in regression

Posted by Yang, Hongwei on Feb 11, 2014; 8:39pm
URL: http://spssx-discussion.165.s1.nabble.com/Stepwise-versus-Enter-method-in-regression-tp5724452p5724500.html

My understanding is that these days people do not like the stepwise method any more in applied research.

 

Instead, the preferred approach seems to be a hierarchical approach to building regression models. This approach is certainly based on the Enter method, instead of the Stepwise method. The Enter method is used each time a candidate in a hierarchy of models is fitted.

 

In addition, there is also a recommendation in the literature on the use of the all-possible-subsets method (rather than stepwise method) to evaluate all possible models from the model space. Then, the top models from the all-possible-subsets method is further evaluated after taking into account theoretical considerations. In SPSS, the all-possible-subsets method is available under the Automatic Linear Modeling platform, not under the traditional Linear Regression platform.

 

Hongwei Yang, University of Kentucky

 

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Norberto Hernandez
Sent: Tuesday, February 11, 2014 2:45 PM
To: [hidden email]
Subject: Stepwise versus Enter method in regression

 

Hi everybody, I am looking for a recommendation, I heard that is better to use Enter besides Stepwise method in regression, but I couldn't find any useful reference about it. 

 

Could you please help me!

Norberto