RE: st: non-parametric tests with Svy estimation

["louis boakye-yiadom" <louisby@hotmail.com>]

Before you explore the non-parametric option, it might be useful to check 
whether the variable is log normal. In other words, if the natural log of 
the "total healthcare costs" variable is normally distributed, I suppose 
with the the appropriate adjustment, you can get around the normality issue 
by using the transformed variable.

Louis


> From: anju parthan <aparthan@yahoo.com>
>
> Hi All
>
> I am using survey data. I would like to compare  total
> healthcare costs in two groups.  The variable total
> healthcare costs is not normally distributed,
> therefore one would run non-parametric statistics.
>
> Can someone tell me how I can run a non-parametric
> statistics accounting for the survey data?
>
> I have compared the two groups with lincom using
> svymean but I feel I should use non-parametric
> statistics instead of comparing the means.
>
> I am sorry if I am missing something trivial. Thanks
> for your time.
>
> Anju

*
*   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/