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]

Re: st: confa

From   Stas Kolenikov <>
Subject   Re: st: confa
Date   Thu, 7 Jul 2011 08:25:53 -0500

-confa- should run with survey data, but it is intended to work with
continuous data. Saying that you need to use WLS with ordinal data is
a strange piece of advice that I keep hearing over and over. I can
come up with arguments that this advice might have been based on, and
I can come up with arguments against WLS and say in favor of -gllamm-
(that will also work with survey data). Can you give the reference you
had and a specific quote?

On Thu, Jul 7, 2011 at 8:15 AM, Aggie Chidlow
<> wrote:
> Dear Stata users,
> I hope somebody can advice I am quite new to "confa".
> I would like to use confa with my survey data.
> I have to say I am familiar with the Stanislav Kolenikow's work :
> Confirmatory Factor Analysis using Confa (2009) Stata Journal, 9(3),
> 329 - 373. I use is as a guide.
> Today I have read that "ml" should not be used with categorical data.
> Instead it is recommended that "wls" (i.e. weighted least squares)
> should be used as proposed by Browne(1984b).

Stas Kolenikov, also found at
Small print: I use this email account for mailing lists only.
*   For searches and help try:

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