Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Re: Re: Why do Stata Cronbach's Alpha values not match SAS?

From   "Joseph Coveney" <>
To   <>
Subject   st: Re: Re: Why do Stata Cronbach's Alpha values not match SAS?
Date   Fri, 29 Apr 2011 13:12:39 +0900

Benjamin Gramig wrote:

. . .

I feel I still need to use SAS for the accompanying PCA of the items in the
scale because I cannot figure out how to output factor loadings or apply
rotations following -polychoric pca-. It is my understanding that for
ordinal/Likert data I need to use polychoric correlations (or another technique)
to perform PCA correctly.


According to the help file for Stas Kolenikov's -polychoricpca-, results are
available as returned matrixes.  

For rotation, if you've already downloaded -polychoricpca-, then you've
downloaded his -polychoric-, as well.  You can use the latter in tandem with
Stata's official -pcamat-, which works with -rotate- (and which also makes the
components available as a returned matrix).

Joseph Coveney

version 11.2

clear *
set more off
sysuse auto

polychoricpca rep78 foreign mpg
matrix list r(eigenvectors)

polychoric rep78 foreign mpg
pcamat r(R), n(`r(N)')
matrix list e(L)
estat loadings


*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index