Re: st: RE: Re: Capture syntax

[Janet Hill <janethill73@yahoo.co.uk>]

Thank you for that I agree with the 0.05 but
unfortunately I did not write the protocol.

The object was to list all the variables that had
p<0.05 which worked for ache (0.047)  but not for diet
(0.114). I realise that I am doing something stupid in
the syntax but I can't see it. I thought that I could
"automate" the process rather than typing in 50
ologits and examining the results.
Janet
--- Nick Cox <n.j.cox@durham.ac.uk> wrote:

> I don't follow this. Your results for diet show a
> P-value of 0.0407
> which Stata is evidently happy to consider as <
> 0.05. 
> 
> A quite separate point is that to many conservative
> statisticians your
> P-values no longer have the orthodox interpretation
> after this shotgun
> screening process. 
> 
> Nick
> n.j.cox@durham.ac.uk 
> 
> Janet Hill
> 
> I have to run typically 50 variables separately in
> ologit as part of a screening process and some of
> the
> models fail with no convergence. I have tried the
> following to trap the error and continue the
> analysis:
> 
> set more off
> foreach x of var chew speech appear1 ......{
> capture quietly ologit `x' wmean if surface==1
> 	if _rc==0 & chi2tail(e(df_m),e(chi2)) < 0.05 di
> "`x'"
> }
> set more on
> 
> Hoping just to print out those  variables with a
> model
> probability < 0.05.
> 
> This typically produces several variables:
> ache
> diet
> glass
> etc.
> When I subsequently analyse these variables the
> model
> associated with ache has p<0.05, but diet does not
> 
> .. ologit ache wm if surface==1,nolog
> 
> Ordered logistic regression                      
> Number of obs   =         15
>                                                   LR
> chi2(1)      =       4.83
>                                                  
> Prob
> > chi2     =     0.0280
> Log likelihood = -16.888841                      
> Pseudo R2       =     0.1251
> 
>
------------------------------------------------------------------------
> ------
>         ache |      Coef.   Std. Err.      z   
> P>|z| 
>    [95% Conf. Interval]
>
-------------+----------------------------------------------------------
> ------
>        wmean |    1.23758   .6219381     1.99  
> 0.047 
>    .0186035    2.456556
> .. ologit ache wm if surface==1,nolog
> 
> 
> .. ologit diet wm if surface==1,nolog
> 
> Ordered logistic regression                      
> Number of obs   =         15
>                                                   LR
> chi2(1)      =       4.19
>                                                  
> Prob
> > chi2     =     0.0407
> Log likelihood = -3.7955935                      
> Pseudo R2       =     0.3556
> 
>
------------------------------------------------------------------------
> ------
>         diet |      Coef.   Std. Err.      z   
> P>|z| 
>    [95% Conf. Interval]
>
-------------+----------------------------------------------------------
> ------
>        wmean |  -1.967477    1.24521    -1.58  
> 0.114 
>   -4.408044    .4730897
> 
> I would be grateful if somebody could point out my
> programming error.
> 
> *
> *   For searches and help try:
> *  
> http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
> 



      __________________________________________________________
Not happy with your email address?.
Get the one you really want - millions of new email addresses available now at Yahoo! http://uk.docs.yahoo.com/ymail/new.html

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/