Here is what i got when I run the suggested code, with a slight
modification taking the survey design into account:
------------
. logit mis V781_R [pweight=weight], cluster(psu), if
SexActiveHIVNegMale==1
(sum of wgt is 3.2763e+03)
Iteration 0: log pseudolikelihood = -1188.0902
Iteration 1: log pseudolikelihood = -1183.4275
Iteration 2: log pseudolikelihood = -1183.289
Iteration 3: log pseudolikelihood = -1183.2888
Logistic regression Number of obs = 3130
Wald chi2(1) = 5.37
Prob > chi2 = 0.0205
Log pseudolikelihood = -1183.2888 Pseudo R2 = 0.0040
(Std. Err. adjusted for 357 clusters in psu)
------------------------------------------------------------------------------
| Robust
mis | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
V781_R | -.6133271 .2646202 -2.32 0.020 -1.131973 -.0946809
_cons | -1.879409 .1766616 -10.64 0.000 -2.22566 -1.533159
------------
the significance mean then that the DV (V781_R) negatively predicts
the missing value (mis). What does that mean? ...
thanks - cy
On Tue, Jul 14, 2009 at 4:42 PM, Maarten buis<[email protected]> wrote:
>
> --- On Tue, 14/7/09, Chao Yawo wrote:
>> So the missing values result from interview errors, and the
>> errors are not related to my DV. In fact, the DV had only
>> 161 missing variables.
>
> This is something you can check (assume that rep78 is your
> unsafe sex variable and mpg is your dependent variable ):
>
> *------ begin example ------
> sysuse auto, clear
> gen mis = missing(rep78)
> logit mis mpg
> *------- end example -------
>
>> If I ignore the errors on that single IV then it implies I
>> will have to accept the lower N (sample size) my analysis,
>> and explain that in my write-up (that changes in sample
>> size for the regression result from missing values on some
>> of the covariates??
>
> This is very common, look up some leading empirical
> publications in your discipline to see what the most common
> formulation in your discipline is. You can do more: you have
> reason why your missing data is not related to your dependent
> variable, and you checked that, and the Allison reference I
> gave in my previous post explains why your results are not
> biased by these missing values.
>
> -- Maarten
>
> -----------------------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
> http://home.fsw.vu.nl/m.buis/
> -----------------------------------------
>
>
>
>
> *
> * 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/
>
*
* 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/
