st: Stepwise Regression using Multiple Imputed Data - stata 11

[Russell Simco <simco_r@hotmail.com>]

Hi,

 

The "stepwise" model selection program is not available when using the "mi estimate:" prefix in stata-11. I would like to do this manually using "mi estimate: ologit". 

 

Having read the "mi" manual, I still am not sure if I am looking at the correct "P value" to use. 

 

Can someone explain what the Prob > F output, which signifies - is this a Wald test? - and would using this P value be equivalent to running the stepwise selection procedure using the wald option? I know that likelihood scores are not averaged over the imputations in the output of "mi estimate:". Below is the output in text - not sure if it will show up.

 

 

 

Model F test:       Equal FMI                     F(   x, xxx  .) =       xxx
Within VCE type:          OIM                     Prob > F        =       xxx

------------------------------------------------------------------------------
   depvar    |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
  indepvar   |  xxxxxxxx.     xxxxx       xxx   xxxx         xxxx       xxxxx
     indepvar|  xxxxxxxx.     xxxxx       xxx   xxxx         xxxx       xxxxx
-------------+----------------------------------------------------------------
       /cut1 |  xxxxxxxx.     xxxxx       xxx   xxxx         xxxx       xxxxx
       /cut2 |  xxxxxxxx.     xxxxx       xxx   xxxx         xxxx       xxxxx
------------------------------------------------------------------------------


 

Thanks for your help,

RS 		 	   		  
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/