2013, Hardcover

ISBN 978-1-60918-230-4
7" x 10", 494 Pages, $65.00
DISCOUNT PRICE: $55.25
www.guilford.com/p/hayes3
Introduction to Mediation, Moderation, and Conditional
Process Analysis
A Regression-Based Approach
Andrew F. Hayes
This decidedly readable, informative book is perfectly suited for a range of audiences, from the
novice graduate student not quite ready for SEM to the advanced statistics instructor. Even the
seasoned quantitative methodologist will benefit from Hayes's years of accumulated wisdom as he
expertly navigates this burgeoningand at times inconsistentliterature. This book is particularly
well suited for graduate-level courses. Hayes brings conditional process analysis to life with such
passion that even the most 'stat-o-phobic' will become convinced that they too can master SPSS (or
SAS) process. The thoughtful use of real-life examples, accompanied by SPSS and SAS syntax and
output, makes the book highly accessible.
Shelley Brown, PhD, Department of Psychology, Carleton University, Canada
A welcome contribution. This book's accessible language and diverse set of examples will appeal
to a wide variety of substantive researchers looking to explore how or why, and under what
conditions, relationships among variables exist. Hayes has a unique ability to effectively
communicate technical material to nontechnical audiences. He facilitates application of several
cutting-edge statistical models by providing practical, well-oiled machinery for conducting the
analyses in practice. I can use this book to enhance my graduate-level mediation class by extending
the course to include more coverage on differentiating mediation versus moderation and on
conditional process models that simultaneously evaluate both effects together.
Amanda Jane Fairchild, PhD, Department of Psychology, University of South Carolina
Mediation and moderation are two of the most widely used statistical tools in the social sciences.
Students and experienced researchers have been waiting for a clear, engaging, and comprehensive
book on these topics for years, but the wait has been worth itthis book is an absolute winner.
With his usual clarity, Hayes has written what will become the default resource on mediation and
moderation for many years to come.
Andy Field, PhD, School of Psychology, University of Sussex, United Kingdom
Hayes provides an accessible, thorough introduction to the analysis of models containing
mediators, moderators, or both. The text is easy to follow and written at a level appropriate for an
introductory graduate course on mediation and moderation analysis. The book is also an extremely
useful resource for applied researchers interested in analyzing conditional process models. One
strength is the inclusion of numerous examples using real data, with step-by-step instructions for
analysis of the data and interpretation of the results. This book's largest contribution to the field is
its replacement of the confusing terminology of mediated moderation and moderated mediation
with the clearer and broader term conditional process model.
Matthew Fritz, PhD, Department of Psychology, Virginia Polytechnic Institute and State
University
This engaging book explains the fundamentals of mediation and moderation analysis and their
integration as conditional process analysis. Procedures are described for testing hypotheses about
the mechanisms by which causal effects operate, the conditions under which they occur, and the
moderation of mechanisms. Relying on the principles of ordinary least squares regression, Andrew
Hayes carefully explains the estimation and interpretation of direct and indirect effects, probing and
visualization of interactions, and testing of questions about moderated mediation. Examples using
data from published studies illustrate how to conduct and report the analyses described in the book.
Of special value, the book introduces and documents PROCESS, a macro for SPSS and SAS that
does all the computations described in the book. The author's website (www.afhayes.com) offers free
downloads of PROCESS plus data files for the book's examples.
Find full information about this title online: www.guilford.com/p/hayes3
Guilford Publications, Inc.
72 Spring Street, New York, NY 10012
Order Online: www.guilford.com
Call Toll-Free:
9 AM - 5 PM Eastern
800-365-7006 (or 212-431-9800)
Fax: 212-966-6708 (24 hrs)
Professor Copies: www.guilford.com/professors
Order Introduction to Mediation, Moderation,
and Conditional Process Analysis
Copies in Hardcover
ISBN 978-1-60918-230-4, $65.00, $55.25*
** Many Guilford titles are available as e-books directly from our website or from major e-book vendors, including Amazon, Barnes & Noble, and Google Play. If an e-book is available, you
can find it for sale on Guilford's product page along with the print edition(s). For titles that are coming soon, the e-book format will be published shortly after the print edition(s).
Go to www.guilford.com/ebooks for more information.
SPSS SOBEL Macro Syntax Reference, updated 19 Feb 2013

SOBEL

SOBEL Y = yvar/X = xvar/M = mvar [/BOOT = {n}(0**)]
[/EFFSIZE = {e}(0**)]
[/VARORD = {v}(2**}]
[/ITERATE = {i}{10000**}]
[/CONVERGE = {cc}{.0000001**}].

Subcommands in brackets are optional.
** Default if subcommand is omitted



Overview

SOBEL estimates the total, direct, and indirect effects of causal variable xvar on outcome variable yvar
through a proposed mediator variable mvar. In also calculates the Sobel test for the indirect effect as
well as a percentile-based bootstrap confidence interval for estimating the indirect effect, as described in
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in
simple mediation models. Behavior Research Methods, Instruments, and Computers, 36, 717-731.
Since the publication of Preacher and Hayes (2004), the macro has been updated to allow for a binary
outcome (Y only) and can also now use either a first or second order standard error estimator in the
Sobel test.

Instructions for Use

The SOBEL.sps file should be opened as a syntax file in SPSS. Once it has been opened, execute the
entire file exactly as is. Do not modify the code at all. Once the program is executed, the SOBEL
program window can be closed. You then have access to the SOBEL command until you quit SPSS.
See below for some example commands. The SOBEL.sps file must be loaded and reexecuted each time
to open SPSS. To install SOBEL permanently in SPSS, install the custom dialog version (see below).

Example

SOBEL Y = know/X = educ/M = attn/BOOT = 5000/effsize = 1.

Estimates the total and direct effects of educ on know, as well as the indirect effect of educ on
know through attn
Produces the Sobel test for the indirect effect using the second order standard error estimator.
Generates a bootstrap confidence interval for the indirect effect using 5000 bootstrap samples.
Produces effect size estimates for the indirect effect.
SPSS SOBEL Macro Syntax Reference, updated 19 Feb 2013


Bootstrapping

As discussed in Preacher and Hayes (2004, 2008), Hayes (2009, 2013) and Scharkow and Hayes (in
press), bootstrap confidence intervals are preferred over the Sobel test because of the assumption the
Sobel tests makes about the shape of the sampling distribution of the indirect effect. The /BOOT
subcommand implements bootstrap estimation of a confidence intervals for the indirect effect. If it is
not used, it defaults to 0 bootstrap resamples, meaning bootstrapping of the indirect effect is disabled.
The user can enter the desired number of bootstrap samples as the argument for n, in intervals of 1000,
for generating percentile-based bootstrap 95% and 99% confidence intervals. Only percentile-based
bootstrap confidence intervals are estimated. For bias-corrected and accelerated confidence intervals,
use the INDIRECT macro described in Preacher and Hayes (2008) or the PROCESS procedure (Hayes,
2013).

Because bootstrapping is based on random resampling of the data, bootstrap confidence intervals will
differ slightly each time the macro is run as a result of the random sampling process. The more
bootstrap samples that are requested, the less this variation between runs.

Standard Error Estimator in the Sobel Test

By default, SOBEL uses the second order standard error estimator in the Sobel test for the indirect effect.
This standard error is defined as



The user can request the first order standard error estimator, defined as



by setting v in the /VARORD subcommand to 1. If the /VARORD subcommand is not used, v defaults to
2 for the second order estimator.

Binary Outcome

By default, all paths in the model as estimated using OLS regression. But if yvar is binary, the c, c and
b paths are estimated using logistic regression. The coefficients the macro prints for b(YX), b(YX.M),
and b(YM.X) are logistic regression coefficients, which estimate the effect of a one unit difference on
the predictor on the log odds of the outcome. The macro automatically detects whether the outcome is
binary and estimates the paths accordingly. The indirect effect is still estimated as the product of the a
and b paths, and this product can be tested against the null of zero using the Sobel test or a bootstrap
confidence interval, as in OLS.

In the event of nonconvergence during iteration toward the maximum likelihood solution for model
coefficients, the macro allows the user to change the default number of iterations (i) and convergence
criteria (cc) using the /ITERATE and /CONVERGE subcommands, which default to 10000 and .0000001,
respectively. Iteration is accomplished using the Newton-Raphson method.

SPSS SOBEL Macro Syntax Reference, updated 19 Feb 2013

Note that unlike in OLS, when the binary is outcome, ab typically is not equal to c c. The coefficients
and standard errors printed are not standardized, a method sometimes advocated for placing the direct,
indirect, and total effects on comparable scales. However, the macro will also print an estimate of the
total effect that places it on the same scale as the indirect effect, using Formula 11.6 in MacKinnon
(2008). This rescaled total effect then yields a difference between the total and direct effect that is much
closer to (but still not the same as) the indirect effect. This rescaling does not affect the computation of
the indirect effect or inferential tests.

Effect Size

By setting e to 1 in the /effsize subcommand (i.e., /effsize = 1), SOBEL generates 5 point
estimates of the size of the indirect effect. The estimators, outlined in the table below, are discussed in
Preacher and Kelley (2011).


Output

Formula


P_m

ab / c
R_m
ab / c'
R2_45
r
2
YM
(R
2
Y.MX
r
2
YX
)
ab_ps
ab / s
Y

ab_cs
(s
X
ab) / s
Y


When used in conjunction with the bootstrapping option, percentile bootstrap 95% and 99% confidence
interval estimates for these effect size measures are generated.

Multicategorical Independent Variables

For estimation of direct and indirect effects of a categorical IV with more than two levels, use
MEDIATE (Hayes and Preacher, in review).

Nonlinear Effects

If you believe or have evidence that one or more of your effects is not well represented with a
linear function, consider the use of MEDCURVE (Hayes & Preacher, 2010).

SOBEL Custom Dialog Box

If you use SOBEL frequently, you might find it convenient to install a version of the SOBEL macro into
your SPSS menus. To do so, download the sobel_spss.spd (UI Dialog Builder) file from
http://www.afhayes.com/ and double click on it. If you have administrative access to your machine, this
should install a new option under your SPSS AnalyzeRegression menu titled Preacher and Hayes
(2004) Simple Mediation Procedure (SOBEL). You may have to run SPSS as administrator by right-
clicking on the SPSS icon and selecting Run as Administrator. If you do not have administrative
access, you will have to contact your local information technology specialist for assistance in setting up
administrative access to your computer.



SPSS SOBEL Macro Syntax Reference, updated 19 Feb 2013



Notes

The proposed mediator variable, mvar, must be a quantitative variables and is assumed to have at
least interval-level measurement properties. The independent variable, xvar, and outcome
variable, yvar, can be quantitative with interval-level properties, or binary. SOBEL should not be
used with a dichotomous mediator.

Covariates are not allowed in SOBEL. If youd like to control for the effect of variables outside the
causal system on mvar and yvar, use the INDIRECT macro, as described in Preacher and Hayes
(2008).

When bootstrapping is enabled, the bootstrap samples are saved to an SPSS data file called
bootstrp.sav, typically in the SPSS root directory, although the exact location will vary from
machine to machine depending on how the SPSS program was installed.

A case will be deleted from the analysis if missing on any of the variables in the model.

In the output, the following notation is used for the a, b, c, and c paths in the diagram above:
a = b(MX); b = b(YM.X); c = b(YX); c = b(YX.M). All path coefficients are unstandardized.

Do not use STRING formatted variables in any of your models. Doing so will produce errors. All
variables should be NUMERIC format.

References

Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A
regression-based approach. New York: The Guilford Press.

Hayes, A. F. & Preacher, K. J. (2010). Estimating and testing indirect effects in simple mediation
models when the constituent paths are nonlinear. Multivariate Behavioral Research, 45, 627-660.

Hayes, A. F. & Scharkow, M. (in press). The relative trustworthiness of tests of indirect effects in
statistical mediation analysis: Does method really matter? Psychological Science.

Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium.
Communication Monographs, 76, 408-420.

MacKinnon, D. P. (2008). An introduction to statistical mediation analysis. Mahwah, NJ: Lawrence
Erlbaum Associates.

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

Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and
comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879-891.

Preacher, K. J., & Kelley, K. (2011). Effect size measures for mediation models: Quantitative strategies
for communicating indirect effects. Psychological Methods, 16, 93-115.