Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Sample size for ttest - (am I wrong?)

From   David Airey <david.airey@Vanderbilt.Edu>
Subject   st: Sample size for ttest - (am I wrong?)
Date   Wed, 19 Feb 2003 09:59:06 -0600

A better book on power analysis for most will be:

Keven R. Murphy and Brett Myors. 1998. Statistical Power Analysis: A Simple and General Model for Traditional and Modern Hypothesis Tests. LEA Publishers. ISBN 0-8058-2947-4.

It's cheap, about 100 pages, well written, and doesn't skimp on topics. It also pushes minimum effect null hypothesis which are arguably more meaningful in the context of power analyses than zero point null hypotheses. With traditional null hypotheses, regardless of the size of effect greater than zero, you will be guaranteed enough power with enough N, and when you can't reject, you can't really accept the null as true either. If you don't know the difference between significance and effect size in statistical testing, this is dangerous. With a minimum effect size null hypothesis, in and experiment with enough power, you can reject an effect you judge to be negligible and a nonsignificant result means something (see Chapter 4). Note, unlike zero point null hypotheses, it is harder to reject minimum effect size null hypotheses with larger N.

My suggestion would be to get your friend to read this book.


* For searches and help try:

© Copyright 1996–2019 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index