<p> </p><p></p><p>ABSTRACT<br/>The main problem in any model-building situation is to choose from a large set of covariates those that<br/>should be included in the “best” model. A decision to keep a variable in the model might be based on the<br/>clinical or statistical significance. There are several variable selection algorithms embedded in SAS PROC<br/>LOGISTIC. Those methods are mechanical and as such carry some limitations. Hosmer and Lemeshow<br/>describe a purposeful selection of covariates algorithm within which an analyst makes a variable selection<br/>decision at each step of the modeling process. In this paper we introduce a macro, %PurposefulSelection,<br/>which automates that process. The macro is based on the following algorithm: (1) fit a univariate model with<br/>each covariate, (2) select as candidates for a multivariate model those significant at some chosen alpha<br/>level, (3) identify those variables that are not significant in the multivariate model at some arbitrary alpha<br/>level, (4) fit a reduced model and evaluate confounding by change in parameter estimates, (5) repeat steps<br/>3 and 4 until the model contains significant covariates and/or confounders and (6) add back in the model,<br/>one at a time, any variable not originally selected, keep any that are significant, and reduce the model<br/>following steps 3 and 4. At the end of step 6, the analyst will have a “main effects model.” Performance of<br/>the macro is illustrated with the application to the Hosmer and Lemeshow Worchester Heart Attack Study<br/>(WHAS) data.<br/>Keywords: logistic regression, SAS PROC LOGISTIC, variable selection algorithm, purposeful selection,<br/>confounding</p>