[R] Stepwise Regression + entry/exit significance level

[Peter Dalgaard]

Shubha Vishwanath Karanth wrote:
> Hi R,
> 
>  
> 
> Does the "step" function used to perform stepwise regression has the
> option to specify the entry/exit significance levels for the independent
> variables? (This is similar to the 'slentry' and 'slstay' option in
> 'Proc reg' of SAS.). Or do we have any other package which does the
> above? Thanks.
> 

That procedure is considered unreliable (someone called Frank is likely
to chime in with a reference to a book on regression modeling
strategies). Cloning obsolete SAS functionality is not high on the
priority list. It is probably not massively hard to modify the step
function to do it, though; all the building blocks would seem to be there.

-- 
Peter Dalgaard
Center for Statistics, Copenhagen Business School
Phone: (+45)38153501
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com