RE: [R] Confirmatory Factor Analysis in Non-Normal case

From: John Fox <jfox_at_mcmaster.ca>
Date: Wed 23 Mar 2005 - 12:27:25 EST


Dear Liu,

I'm not aware of any way of doing what you want in R at present. You can base the analysis on polychoric correlations, which can be computed using the polycor package, and you can get consistent estimates of the factor loadings and factor correlations from the sem packages, but the likelihood won't be right.

I hope that this is of some help,
 John



John Fox
Department of Sociology
McMaster University
Hamilton, Ontario
Canada L8S 4M4
905-525-9140x23604
http://socserv.mcmaster.ca/jfox

> -----Original Message-----
> From: r-help-bounces@stat.math.ethz.ch
> [mailto:r-help-bounces@stat.math.ethz.ch] On Behalf Of yyan liu
> Sent: Tuesday, March 22, 2005 7:24 PM
> To: r-help@stat.math.ethz.ch
> Subject: [R] Confirmatory Factor Analysis in Non-Normal case
>
> Hi:
> I am doing a confirmatory factor analysis now. In the
> analysis, I have null hypothesis test which specify some
> special structure for the loading matrix.
> And the alternative is there is no such special structure.
> Then the log likelihood ratio test can be used. The problem I
> have is my data comes from a questionnaire and all the
> variables are discrete from
> 1 to 7, which makes the normality assumption a problem. I
> wonder if there is any other test can do this job or at least
> I can make some sensible transformation to make the variables
> "normal"?
> Thank you very much!
>
> liu
>
> ______________________________________________
> R-help@stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
> http://www.R-project.org/posting-guide.html



R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Wed Mar 23 12:32:24 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:30:53 EST