Statalist The Stata Listserver

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

Re: st: Repeated measures with no between-participants vars?

From   Joseph Coveney <>
To   Statalist <>
Subject   Re: st: Repeated measures with no between-participants vars?
Date   Tue, 06 Jun 2006 18:18:59 +0900

Eran wrote:

I am trying to assess the way in which participants' evaluations of 3
different stimuli change over time, as indicated in a series of
measurements. I am interested in checking whether the pattern of changes
in those evaluations is different for the different stimuli.

In other words, the two within-participants variables are (1) type of
stimulus, and (2) time of measurement. There are no between-participant

This seems like a simple task, but I haven't been able to get Stata to do
this for me. I tried a number of variations on the repeated ANOVA command,
but no luck. I would greatly appreciate any advice.


If your study's experimental design is suitable, then you can analyze the
results with ANOVA in a manner analogous to that of a cross-over
"bioequivalence" study.  The model has factors of, for example, period (time
in your case), treatment (stimulus type in your case) and assigned sequence
of treatments (sequence or pattern in which you presented the various
stimulus types to the participants--which is a between-participants
variable, by the way), and one or more interaction terms.

Stata has a whole suite of commands for analysis of these types of studies.
Type -help pk- in the command window to see the intial help screen and go
from there.  Look especially at -pkcross-.  For your study, substitute
"participant's evaluation score" for whatever pharmacokinetic response
variable that the help files might mention.  (The syntax diagram in the help
file for -pkcross- is fairly general, referring to "outcome" for the
response variable.)  If you're interested in checking whether the pattern of
changes differs, then, as you read the help file, look for one
of -pkcross-'s parameterizations that estimates period-by-treatment
interaction (time-by-stimulus-type interaction).

Joseph Coveney

*   For searches and help try:

© Copyright 1996–2022 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index