Date: Wed, 1 Jun 2005 20:36:38 0500
ReplyTo: schmitta@UWM.EDU
Sender: "SAS(r) Discussion" <SASL@LISTSERV.UGA.EDU>
From: "Thomas A. Schmitt" <schmitta@UWM.EDU>
Subject: Design/Analysis Quesions
ContentType: text/plain; charset=ISO88591
Hello Everyone,
I have done a simulation study using SAS. I have a design that has two factors
each with three levels that are crossed. Thus, I have a total of 9 conditions
for each simulee. Each examinee takes all 9 conditions. I’m assuming this is a
twoway withinsubjects design. Is this correct? I sampled 200 simulees from a
uniform distribution with 100 replications for each simulee. I then calculated
RMSE and Bias for each simulee across the conditions. Thus, I have a total of
200 RMSE and Bias estimates. My main questions are: (1) is this design
conceptualize correctly as a withinsubjects design, (2) what would be the
appropriate inferential analyses on the Bias and RMSE estimates I have (i.e.,
effect size (d), eta squared etc.) on the 9 conditions to determine
meaningfulness, (3) and what would be the most appropriate way to analyze this
in SAS.
Tom
