A rank test might be a better idea than using T^2, which assumes
normality of underlying variables, anyway. In the code below, you only
have one value of emp_2005 per group, so no test would be feasible
there, I would imagine.
On 1/31/06, Nicholas Potter <npotter@unm.edu> wrote:
> Hi, haven't done much with stata before, but I have a paired sample data
> set. Problem is that the sample resembles a chi-sq distribution rather
> than a normal distribution (can't assume normality). Hence I was
> thinking that using a Hotelling's T-sq. test would be more appropriate
> that a paired-sample t-test. I need the test to compare means between
> the pairs, but am uncertain of how to use the by(groupvar) option.
>
> right now I have:
>
> egen grp_a = group(emp_2005);
> hotelling emp_2003 emp_2005, by(grp_a);
>
>
> but I'm fairly sure that the group division is not what I am looking for.
>
> Any help would be much appreciated.
>
> Thanks
>
> -Nick
>
>
> *
> * 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/
>
--
Stas Kolenikov
http://stas.kolenikov.name
*
* 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/