<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7638.1">
<TITLE>RE: Multilevel analysis for complex surveys</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Thank you very much, Kentaro.<BR>
<BR>
Angel<BR>
<BR>
<BR>
-----Mensaje original-----<BR>
De: Kentaro Hirose [<A HREF="mailto:hirose@Princeton.EDU">mailto:hirose@Princeton.EDU</A>]<BR>
Enviado el: mar 26/08/2014 19:02<BR>
Para: Angel Rodriguez; mediation-information@r-forge.wu-wien.ac.at<BR>
Asunto: RE: Multilevel analysis for complex surveys<BR>
<BR>
Hi Angel,<BR>
<BR>
> I don't understand why in med.fit the term related to the clustering units is (1|SCH_ID), while (1 + attachment|SCH_ID) is the term in the out.fit.<BR>
<BR>
This is just fitting a varying-intercept model for the mediator and a varying-intercept-and-slope model for the outcome. You could also fit two varying-intercept models for both the mediator and outcome. And similarly, you could also fit two varying-intercept-and-slope models for both the mediator and outcome. It's all up to your theory.<BR>
<BR>
<BR>
> My first question is if I should use mediation with glmer.<BR>
<BR>
If your dependent variable is binary and you want to use a multilevel model, then you should use glmer.<BR>
<BR>
<BR>
Best,<BR>
Kentaro Hirose<BR>
<BR>
Postdoctoral Fellow<BR>
Department of Politics<BR>
Princeton University<BR>
<BR>
<BR>
<BR>
________________________________________<BR>
From: mediation-information-bounces@r-forge.wu-wien.ac.at [mediation-information-bounces@r-forge.wu-wien.ac.at] on behalf of Angel Rodriguez [angel.rodriguez@matiainstituto.net]<BR>
Sent: Tuesday, August 26, 2014 12:49 PM<BR>
To: mediation-information@r-forge.wu-wien.ac.at<BR>
Subject: [Mediation-information] Multilevel analysis for complex surveys<BR>
<BR>
Dear subscribers,<BR>
<BR>
I want to carry out a mediation analysis with variables collected in a survey (for more information, see the thread "causally dependent mediators"). It was a complex survey with strata and individuals clustered in census sections. I do not have census section variables, but I want the models to be aware that individuals are not independent because they're grouped in census sections.<BR>
<BR>
My first question is if I should use mediation with glmer.<BR>
<BR>
In that case, my second question would be about the syntax to carry out this. In the example in Tingley's "mediation: R Package for Causal Mediation Analysis", the proposed models are:<BR>
<BR>
<BR>
R> library(lme4)<BR>
R> set.seed(2014)<BR>
R> med.fit <- glmer(attachment ~ catholic + gender + income + pared + (1|SCH_ID),<BR>
+ family = binomial(link = "logit"), data = student)<BR>
R> out.fit <- glmer(fight ~ catholic*attachment +<BR>
+ gender + income + pared + (1 + attachment|SCH_ID),<BR>
+ family = binomial(link = "logit"), data = student)<BR>
<BR>
I don't understand why in med.fit the term related to the clustering units is (1|SCH_ID), while (1 + attachment|SCH_ID) is the term in the out.fit.<BR>
<BR>
Thank you very much,<BR>
<BR>
Angel Rodriguez-Laso<BR>
Research Project Manager<BR>
Matia Instituto Gerontologico<BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>