s-news
[Top] [All Lists]

Factor analysis, method=MLE, in Splus and SAS:

To: s-news <s-news@wubios.wustl.edu>
Subject: Factor analysis, method=MLE, in Splus and SAS:
From: Aldi Kraja <aldi@wubios.wustl.edu>
Date: Mon, 15 Sep 2003 13:31:50 -0500
Organization: Washington University
Reply-to: aldi@wubios.wustl.edu
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax)
Hi,
Does anybody know what could be a good reason(s) that Factor analysis results from Splus6 and version 9 of SAS are a little different? Scores of Factors 1 and 2 are similar in both analysis, but they change starting at factor3 and 4.

Here is a sample from the data run in Splus and SAS:
In Splus6 in Linux OS server I applied factanal function with method=mle.
The output of the scores follow with symbols Factor1mle etc.
In SAS version 9 for windows I applied the procedure:
proc factor method=ml score
The output of the scores follow with symbols Factor1 etc.
In both cases I kept 4 factors.

Obs Factor1mle Factor2mle Factor3mle Factor4mle Factor1 Factor2 Factor3 Factor4

949 -0.248375 1.048772 0.631290 -1.173078 -0.29346 1.08845 -1.07569 -0.71018 950 -0.058311 0.361250 1.282595 -0.755245 0.03378 0.40903 -0.66505 -1.31753 951 -0.219713 -1.351065 0.576419 -0.684653 -0.34923 -1.30681 -0.57086 -0.71870 952 2.204151 -0.014672 1.059058 -1.248900 2.04540 -0.15122 -1.63148 -0.81412 953 0.328395 -0.720541 0.187006 -0.666158 0.16131 -0.73544 -0.69851 -0.22287

TIA,
Aldi

--




<Prev in Thread] Current Thread [Next in Thread>
  • Factor analysis, method=MLE, in Splus and SAS:, Aldi Kraja <=