Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Differences between stata 9 and 8.2 - factor analysis


From   khigbee@stata.com
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Differences between stata 9 and 8.2 - factor analysis
Date   Mon, 20 Jun 2005 14:32:05 -0500

Daniel Schneider <daniel.schneider@stanford.edu> asks:

> When I use a factor analysis, pcf:
> 
> -- factor grata-gratn gratr if s_func!=4, pcf
> 
> And then do a rotation:
> 
> -- rotate, varimax
> 
> 
> My results differ between Stata 9 and 8.2. I am aware that factor
> analysis might be differing between different software, but this seems
> to be a bit strange. My results in Stata 8.2 are:
> 
> <stata 8.2 and 9 output cut>

I didn't compare every number in the two tables, but it appears
that all that is different is a sign flip for the third column.

The sign of each column is arbitrary.  Think about it
geometrically.  Each column (vector) provides an axis for the
described subspace.  A vector and its negative describe the exact
same axis.

This is true for many of the multivariate methods -- the sign is
arbitrary.

As to why the difference between Versions 8.2 and 9, it is likely
due to most of the code being moved from internal source code
(Version 8.2) to user-readable ado code (Version 9).

Ken Higbee    khigbee@stata.com
StataCorp     1-800-STATAPC

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index