Prevention Science [PREV] pp1182-prev-485270 April 1, 2004 3:25 Style ﬁle version Nov. 04, 2000
Prevention Science, Vol. 5, No. 2, June 2004 (
How Do Outcomes in a Specified Parent Training
Intervention Maintain or Wane Over Time?
David S. DeGarmo,
Gerald R. Patterson,
and Marion S. Forgatch
In a randomized prevention trial, 238 recently separated mothers and their young sons were
assigned to either Parent Management Training (PMT) or a comparison group. Families were
intensively assessed at baseline and at each 6-month interval through 30 months. To under-
stand the effects of PMT, we ﬁrst evaluated effect sizes among family variables over time.
Second, because observed parenting was the target of PMT, we hypothesized a sequential pat-
tern of structured changes within and between individuals. Using constructs with mismatched
sources of data, we conducted a set of latent growth mediational analyses to test hypothe-
sized mechanisms explaining change. Effect sizes indicated that parenting changed ﬁrst within
12 months, followed by changes in boy behaviors and ﬁnally changes in maternal depression
within 30 months. Unique follow-up ﬁndings indicated that intervention effects on reductions
in maternal depression were mediated by reductions in boy externalizing; intervention effects
on externalizing were mediated by reductions in boy depression. As expected, increases in
effective parenting predicted reductions in child behavior problems. PMT effects on internal-
izing were direct and indirect, partially mediated by parenting practices. Results are discussed
from a system’s perspective on PMT ampliﬁers.
KEY WORDS: parent training; child adjustment; maternal adjustment; divorce; growth curve.
In the last decade the ﬁeld of prevention science
rapidly grew in scope and speciﬁcity. As an increasing
number of intervention programs are demonstrated
to be efﬁcacious and disseminated in multiple con-
texts, the question emerges as to why intervention
processes work and how they maturate over time.
Prevention scientists, for example, have taken the po-
sition that prevention trials should be used to an-
swer questions reﬁning moderating and mediating
mechanisms explaining effectiveness (Brown, 1993;
Brown & Lia, 1999; West et al., 1993). With the cur-
rent data we conducted a series of latent growth
models to examine mediators that deﬁne a preven-
tion process, including a focus on both normative and
Oregon Social Learning Center, Eugene, Oregon.
Correspondence should be directed to David S. DeGarmo, Oregon
Social Learning Center, 160 East 4th Avenue, Eugene, Oregon
97401; e-mail: email@example.com.
pathogenic development. We addressed questions of
change that accrued during the follow-up stages of the
Parent Management Training (PMT)-based interven-
tion that was tailored for single mothers in the Oregon
Divorce Study II (ODS-II, Forgatch & DeGarmo,
In general, applications of PMT to clinical
samples of families referred for treatment of out-
of-control children have generally produced posi-
tive outcomes in both efﬁcacy (Bien & Bry, 1980;
Eisenstadt et al., 1993; Serketich & Dumas, 1996) and
effectiveness trials (Tynan et al., 1999). For some clini-
cal samples, some of these improvements are lost dur-
ing follow-up as shown in the studies by Bank et al.
(1991), Baum and Forehand (1981), and Patterson
and Fleischman (1979). On the other hand, less dis-
turbed samples have maintained improvements for
at least a 12-month follow-up (Webster-Stratton &
Hammond, 1997). Such results call for the under-
standing of why PMT results persist or decline.
In addition to clinical samples, recent efforts
have been made to understand PMT effectiveness
2004 Society for Prevention Research