[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Appropriate parametric and nonparametric test

From   Maarten buis <>
Subject   Re: st: Appropriate parametric and nonparametric test
Date   Sat, 20 Oct 2007 10:55:09 +0100 (BST)

--- Anita Sayal <> wrote:
> what would you recommend the ideal parametric and
> non-parametric test for 2 small samples one with 25
> obs. the other with 31.(both unrelated models) 
> OLS did not work in both occassions I found the
> classical assumptions did  not hold so I was looking
> for a non-parametric counterpart such as..Chi-squared
> test that did not work either..

A chi-square test tests something very different than
a OLS regression, so before you start worying about
which test is ideal, you should take some time to 
think about what it is you want to test exactly.

Given that you started out with OLS (and because it is
a very common thing to want to test) I assume that you
want to test whether the central tendency of sample 1
is the same as the central tendency in sample 2. If 
you think it is ok to look at the median instead of the
mean (when doing OLS you are comparing means), than you
could have a look at Roger Newson's -somersd- package 
on -ssc-. Don't forget to get the ancillary files, they
are a valuable description of what that package does.

Hope this helps,

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Want ideas for reducing your carbon footprint? Visit Yahoo! For Good
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index