testing mediation with categorical IV

[Sara House]

Good morning List:

I'm trying to test for mediation (following Baron and Kenny) with a categorical IV and continuous mediator and DV.  The IV is a three-level and I have two dummy variables.  So when I regress the mediator on to the IV, I have two betas and once again, when I regress the DV onto the IV, I have two betas.  When I perform those analyses, this is what I get:

IV ---> Mediator (Beta1 = -0.479 and Beta2 = -0.454)
IV ---> DV (Beta1 = 0.240 and Beta2 = 0.157)
Mediator ---> DV (Beta = -0.006)

So I conclude that there is no mediation taking place, because the relationship between the mediator and DV is not significant.

Is that how I should be testing mediation in this case?  B&K don't really talk about categorical IVs for mediation (just moderation) so I'm unsure of how to be doing this.  And when I see path diagrams in journal articles, there are only three betas, not five.  Obviously, putting two betas on the line is not a big deal, but what if I had a design with more than three categories?  That could make for some messy path diagrams!  Any thoughts?  Or recommendations for references I should track down?  Thanks in advance for your help on this.

Sara

Sara M. House, M.A.
Adjunct Faculty
Loyola University Chicago, Psychology Department
Email: [hidden email]
Teaching: Research Methods, Psychology & Law

AND

Data Analyst
Chicago Public Schools, Department of Program Evaluation
Email: [hidden email]

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Swank, Paul R]

I like Kris Preacher's bootstrap approach to mediation analysis. He has
a paper (see below) on this including SPSS macros. They are available on
his website.

Preacher, K. & Hayes, A. (2004). SPSS & SAS procedures for estimating
indirect effects in simple mediation models. Behavior Research Methods,
Instruments, & Computers, 36(4), 717-731.

Paul R. Swank, Ph.D.
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center - Houston


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Sara House
Sent: Monday, April 07, 2008 10:19 AM
To: [hidden email]
Subject: testing mediation with categorical IV

Good morning List:

I'm trying to test for mediation (following Baron and Kenny) with a
categorical IV and continuous mediator and DV.  The IV is a three-level
and I have two dummy variables.  So when I regress the mediator on to
the IV, I have two betas and once again, when I regress the DV onto the
IV, I have two betas.  When I perform those analyses, this is what I
get:

IV ---> Mediator (Beta1 = -0.479 and Beta2 = -0.454)
IV ---> DV (Beta1 = 0.240 and Beta2 = 0.157)
Mediator ---> DV (Beta = -0.006)

So I conclude that there is no mediation taking place, because the
relationship between the mediator and DV is not significant.

Is that how I should be testing mediation in this case?  B&K don't
really talk about categorical IVs for mediation (just moderation) so I'm
unsure of how to be doing this.  And when I see path diagrams in journal
articles, there are only three betas, not five.  Obviously, putting two
betas on the line is not a big deal, but what if I had a design with
more than three categories?  That could make for some messy path
diagrams!  Any thoughts?  Or recommendations for references I should
track down?  Thanks in advance for your help on this.

Sara

Sara M. House, M.A.
Adjunct Faculty
Loyola University Chicago, Psychology Department
Email: [hidden email]
Teaching: Research Methods, Psychology & Law

AND

Data Analyst
Chicago Public Schools, Department of Program Evaluation
Email: [hidden email]

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD