Date: Wed, 16 May 2001 21:16:54 -0400
Reply-To: David.Howell@uvm.edu
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: "David C. Howell" <vze2nkpw@VERIZON.NET>
Subject: Re: Testing for moderator effects
In-Reply-To: <200105170019.f4H0JMB15517@int-mail.hirokoku-u.ac.jp>
Content-Type: text/plain; charset="ISO-2022-JP"
The classic paper on this is Baron and Kenny (Baron, R. M., & Kenny, D. A.
(1986). The moderator-mediator variable distinction in social psychological
research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173-1182.) David Kenny has a
excellent web page at http://w3.nai.net/~dakenny/mediate.htm. It focuses on
mediation, but I don't recall whether he talks about moderation or not. Also
see Aiken and West (Aiken, L. S. & West, S. G. (1991). Multiple regression:
Testing and interpreting interactions. Newbury Park, CA: Sage.)
For a graphical presentation, look at a free program by Norman Peladeau
(author of Simstat, which is a great program in itself) that is called
Italassi. It's available at http://www.simstat.com/viewxyz.htm. It is a
great way to visualize your relationships.
David C. Howell
Dept of Psychology
University of Vermont
Burlington, VT 05405
www.uvm.edu/~dhowell/StatPages/StatHomePage.html
-----Original Message-----
From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU]On Behalf Of
andrea streit schreiner
Sent: Wednesday, May 16, 2001 8:19 PM
To: SPSSX-L@LISTSERV.UGA.EDU
Subject: Testing for moderator effects
What do you all recommend for testing for the affects of a
hypothesized moderator?
I am trying to understand the effect
of Y on the relationship between X and Z.
I can not do LISREL as I have only 100 subjects
and am controlling for 7 other variables (already pushing
the limits with regression).
In the stepwise regression I have run the moderator is
significant but that is just evidence of a linear relationship?
My hypoth is that when moderator is low there will be a stronger
rel between X and Z than when it is high.
I can break sample into high and low groups on moderator but
then there are only about 50 in each group and I can not
really do regression then I think.
Any suggestions would be most appreciated.
Thanks as always.
Andrea
--
"""""""""""""""""""""""""""""""""""""""""""""""""""""""""
Andrea Streit Schreiner PhD
Hiroshima International University
Faculty of Nursing
555-36 Gakuen Dai
Kurose-cho
Kamo Gun, Hiroshima ken, Japan 724-0695
FROM OVERSEAS
phone (your international dialer+) 8182 370 4543
fax " 8182 370 4542
FROM WITHIN JAPAN
phone 082 370 4543
fax 082 370 4542
""""""""""""""""""""""""""""""""""""""""""""""""""""""
