Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: RE: st: Repeated Measures ANOVA and Missing Values

From   Philip Ender <>
Subject   Re: Re: RE: st: Repeated Measures ANOVA and Missing Values
Date   Mon, 17 Oct 2011 13:56:47 -0700

<> wrote:
what is considered to be small sample size for xtmixed? from what I
understood from you xtmixed Chicago lecture with small sample size
xtmixed may cause type I error, if I am correct please advise about a
good sample size (total sample size) for xtmixed.
Dear Amir,

The issue is not the sample size for -xtmixed- but what sample size do
you need to have confidence in the chi-square p-values.  I'm afraid
there is no simple easy answer to this question.  Most experimental
designs tend to have relatively small sample sizes and report F-ratios
and associated p-values.  If I don't have repeated measures, I use
-anova- or -regress-.  However, for repeated measures with missing
observations I use -xtmixed-.  Back when I was a S*S user I used -proc
mixed- which gives results as F-ratios with denominator degrees of
freedom.  A lot of my psychology clients use S*SS -mixed- which also
yields F-ratios and p-values.  I try to obtain the same results in
Stata, my preferred package, by estimating reasonable denominator
degrees of freedom.

Phil Ender
UCLA Statistical Consulting Group
"Money can't buy you happiness, but it can buy you a good statistical package."
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index