# Re: st: predict after svy logit

 From smerryman@kc.rr.com To statalist@hsphsun2.harvard.edu Subject Re: st: predict after svy logit Date Wed, 15 Jun 2005 11:49:34 -0500

```One way to get around this until it is fixed would be replace the
predictions
with :

replace temp2 = cond(temp2>0 ,1/(1+exp(-temp2)), exp(-temp2)/(1+exp(-
temp2)))

Now, running this program gives the same output:

version 8.0
qui svyset [pweight=patwt], psu(psum) strata(stratm2)
qui svylogit y x
predict double temp

version 9.0
qui svyset  psum [pweight=patwt],  strata(stratm2)
qui  svy : logit y x
predict double temp2
replace temp2 = cond(temp2>0 ,1/(1+exp(-temp2)), exp(-temp2)/(1+exp(-
temp2)))

compare temp temp2

---------- difference ---------
-
count       minimum      average
maximum
-----------------------------------------------------------------------
-
temp=temp2                    100
----------
jointly defined               100             0            0
0
----------
total                         100

Scott

----- Original Message -----
From: "Schonlau, Matthias" <matt@rand.org>
Date: Wednesday, June 15, 2005 9:42 am
Subject: st: predict after svy  logit

> Dear stata listers:
>
> In version 9, When I use predict after " svy: logit" the predicted
> probabilities range from roughly 0.8 to greater than 2.0.  I don't
> understand this. When predicting in Version 8.0 (under version
> control) I get predictions that are credible. The survey logistic
> regression give the same coefficients for both version 8.0 and
> version 9.0 , the problem is only the prediction.
>
> Below an example with partial output. Any suggestions ?
>
> Thanks,
>
> Matt
>
*
*   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/
```