Statalist


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

st: Generating random variable with specified effect size between groups


From   "Barth Riley" <BarthRiley@comcast.net>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Generating random variable with specified effect size between groups
Date   Sat, 19 Jul 2008 04:56:21 -0500

Dear Statlist

I want to generate a normal random variable (like a z score) with a
specified effect size when comparing two groups of equal size. The effect
size is based on an independent t test:

			 (mean 2 -mean1/pooled sd)

The effect sizes I get are usually smaller than what I wanted (e.g., .175
instead of .2). I am having trouble simulating data in which both the mean
difference equals the desired effect size and the pooled SD = 1. I know that
the group SDs have to be smaller than the SD for all simulated observations,
but I don't know by how much or how to simulate this. If anyone has
suggestions on how to do this, I would very much appreciate it.

Barth



*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index