Re: Re: st: factor analysis on tetrachoric correlation

[Richard Williams <Richard.A.Williams.5@nd.edu>]

At 03:00 PM 1/23/2004 +0100, jean-benoit.hardouin@club-internet.fr wrote:
> OK, thank you for your responses, but my question was not enough precise :
> I want to realize a factor analysis based on the tetrachoric coefficients 
> matrix WITH STATA. The Stata program tetrac (author :  Philip B. Ender) 
> permit to create this matrix, but I don't know how realize WITH STATA a 
> factor analysis on a such matrix (the factor Stata program don't permit to 
> use a correlation matrix different of the classical (pearson rho 
> coefficients) correlation matrix). The procedure "factor" of SAS, for 
> example, permit to choose a dataset as the working correlation matrix in 
> defining
> the dataset as a correlation matrix.
Jean, perhaps I wasn't clear, but -corr2data- is a Stata command.  It is 
quite useful when you have the means and covariances but not the original 
data.  In your case, it might be a way of faking Stata into analyzing 
something other than the usual Pearson correlation matrix.

My one concern, as I said before, is that I am not sure whether factor 
analysis using tetrachoric correlations only requires the input of the 
tetrachoric correlations, or whether the computations are supposed to 
proceed differently when you have tetrachoric correlations.

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/