Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: statalist-digest V4 #2695


From   "Alan Acock" <acock@comcast.net>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: statalist-digest V4 #2695
Date   Sun, 13 May 2007 08:58:56 -0700

Yan Cheng wrote--
I have a a dichotomous variable as endogenous variable in the path
analysis, sex(0/1), I wrote the following syntax, is it right? Could
you please provide me some advances?
pathreg (knowledge exogenous_vars) /*
*/ (attitude knowledge exegenous_vars) /*
*/ (sex attitude knowledge exogenous_vars)

There are a could problems. An endogenous variable is dependent on the
predictors in a path model and I don't understand how you think a person's
sex depends on their attitude or knowledge--i.e., increasing knowledge
doesn't change a person's sex. Second, if you have an endogenous variable
that is binary it would make more sense to do a logistic regression for that
variable. Path models report standardized coefficients (beta weights) and
these indicate that as you go up one standard deviation on a predictor you
go up (beta weight) standard deviations on the endogenous variable. It
usually doesn't make sense to think of a binary variable such as sex going
up .2, or some other number, standard deviations.
The pathreg command is useful and an easy way to keep the sample size
consistent for each prediction. Unfortunately, I don't think it reports the
R-square for each equation and that is usually important information.

Alan Acock
acock@comcast.net

No virus found in this outgoing message.
Checked by AVG Free Edition. 
Version: 7.5.467 / Virus Database: 269.7.0/801 - Release Date: 5/12/2007
6:40 PM
 


*
*   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