Statalist The Stata Listserver


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

Re: st: mediation analysis


From   "Garth Rauscher" <garthr@uic.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re: st: mediation analysis
Date   Fri, 18 May 2007 07:37:17 -0500

Martin, 

thanks so much for the excellent reference on standardized coefficients,
which gives a very clear description of why Y-standardized coefficients are
sometimes useful in logistic regression, especially in my situation. Also,
thanks for the code, which provides insight about how to conduct a bootstrap
analysis more generally when your parameter of interest is not prepackaaged
but has to be calculated. 

I'd like to modify the program to calculate the differences in the excess
***Y-standardized*** odds ratios but don't know how to tell Stata that.
Instaed of using _b(variable_name) for the unstandardized ORs, do you happen
to know how to ask for the Y-standardized estimate? I tried _b[bStdy] but
Stata did not recognize that.

*----------- partial program starts here -----------

program define difodds, rclass

	logit latestage black hispanic, or
fitstat
listcoef, std help
	tempname totalblack
	scalar `totalblack' = _b[bStdy]

*----------- end of partial program -----------


Thanks, Garth

Garth Rauscher
Assistant Professor
SPH Epidemiology and Biostatistics
University of Illinois at Chicago
1603 W Taylor St (M/C/ 923)
Chicago, IL 60612
(312) 413-4317 phone
(312) 996-0064 fax
garthr@uic.edu email 


Date: Thu, 17 May 2007 16:44:00 +0100 (BST)

From: Maarten buis <maartenbuis@yahoo.co.uk>

Subject: Re: st: mediation analysis

- --- Garth Rauscher <garthr@uic.edu> wrote:

> I have been conducting mediation analyses to determine to what extent
> certain social variables can explain why Black women are diagnosed at
> a later stage of breast cancer compared to white women.
> 
> I am estimating two models...
> 
> logit latestage Black, or
> logit latestage Black mediator, or
> 
> ...then calculating the percent reduction in the odds ratio for
> "Black".
> 
> Putting aside issues of whether this is the best method to do such an
> analysis (probably not),

>To be clear, the answer I just sent is by no means an official
>endorsement of this method (not that many people have ever asked me for
>an official endorsement, let alone need one, but you get the point).
>You are right this is not the best way of doing this. The clearest
>exposition I have found of issues you should consider is this course
>note by Richard Williams: http://www.nd.edu/~rwilliam/xsoc73994/l06.pdf
<http://www.nd.edu/~rwilliam/xsoc73994/l06.pdf> 
>Hope this helps,

>Maarten
 

Garth Rauscher
Assistant Professor
SPH Epidemiology and Biostatistics
University of Illinois at Chicago
1603 W Taylor St (M/C/ 923)
Chicago, IL 60612
(312) 413-4317 phone
(312) 996-0064 fax
garthr@uic.edu <mailto:garthr@uic.edu>  email 

*
*   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