<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=windows-1252">
<META content="MSHTML 5.50.4207.2601" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>I have a question regarding doing forward stepwise
regression using the step function and an lm model.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>I am trying to run a stepwise procedure that
will add a predictor variable, one at a time, conduct an F test to determine
whether or not to drop any previous terms, and terminate when no
other predictors qualify to be added. </FONT><FONT face=Arial
size=2>Using the step function and looking at the example in the book, I
have the following questions.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Is the use of step (lm, direction="forward")
appropriate for forward stepwise regression?</FONT></DIV>
<DIV><FONT face=Arial size=2>Should all terms be included in the original lm
model?</FONT></DIV>
<DIV><FONT face=Arial size=2>Should all possible models be included in the scope
parameter? Can I generate all possible models with a single
command?</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV><FONT face=Arial size=2>Thank you for your
comments.</FONT></DIV></BODY></HTML>