That's what someone else suggested through a reference to -robvar-. But who knows whether it is what is wanted? Diego asked only for a heterogeneity index.
Nick
[email protected]
Ángel Rodríguez Laso
Wouldn't be the Levene's test of homogeneity of variances suitable?
2009/12/7 Nick Cox <[email protected]>:
> I can't see a way to answer this precisely without an idea of what index you have in mind. -egen- supports many alternatives to the standard deviation. Otherwise it may be that you seek some so-called nonparametric test.
Diego Bellavia
> I would like to measure the amount of variability I have for one specific measure in 2 different groups
> of patients. of course I could use the standard deviation for that but I would like also to
> create an index of that variability and be able to compare that index between the 2 groups with a formal stat test.
> Note that in this case I am not interested in central tendency measures (mean or median) but just in a way to quantitate
> the variability, present that as an index and compare among 2 (or more) groups.
*
* 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/
