Statalist


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

Re: st: constructing and implementing logistic regression


From   Maarten buis <maartenbuis@yahoo.co.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: constructing and implementing logistic regression
Date   Fri, 5 Feb 2010 00:14:53 -0800 (PST)

--- On Thu, 4/2/10, Mike Smith wrote:
> I am trying to do propensity score matching, but first need
> to do logistic regression and that's what I am have trouble 
> with. suppose I have a model as follows: gpa (the dependent
> variable) and sex and race being the independent variables.
<snip>

This suggests that you think you can make propensity weights
by estimating a logit on a continuous variable. logistic
regression is a model for a binary depedent variable, so this
way you make propensity scores when your treatment variable
is binary. Estimating a causal effects for continous 
explanatory variables is much harder. See for instance Stephen 
L. Morgan and Christopher Winship (2007) Counterfactuals and 
Causal Inference: Methods and Principles for Social Research. 
Cambridge University Press. I am not recomending that you 
turn gpa in a binary variable, in most cases that just doesn't
make substantive sense. The book I refered to earlier does
provide some options for continous explanatory variables.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   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