Prevention Science, Vol. 7, No. 1, March 2006 (
Assessing the Total Effect of Time-Varying Predictors
in Prevention Research
Bethany Cara Bray,
Rick S. Zimmerman,
and Susan A. Murphy
Published online: 18 February 2006
Observational data are often used to address prevention questions such as, “If alcohol ini-
tiation could be delayed, would that in turn cause a delay in marijuana initiation?” This
question is concerned with the total causal effect of the timing of alcohol initiation on the
timing of marijuana initiation. Unfortunately, when observational data are used to address
a question such as the above, alternative explanations for the observed relationship be-
tween the predictor, here timing of alcohol initiation, and the response abound. These al-
ternative explanations are due to the presence of confounders. Adjusting for confounders
when using observational data is a particularly challenging problem when the predictor
and confounders are time-varying. When time-varying confounders are present, the stan-
dard method of adjusting for confounders may fail to reduce bias and indeed can increase
bias. In this paper, an intuitive and accessible graphical approach is used to illustrate how
the standard method of controlling for confounders may result in biased total causal ef-
fect estimates. The graphical approach also provides an intuitive justiﬁcation for an alter-
nate method proposed by James Robins [Robins, J. M. (1998). 1997 Proceedings of the
American Statistical Association, section on Bayesian statistical science (pp. 1–10). Retrieved
from http://www.biostat.harvard.edu/robins/research.html; Robins, J. M., Hern
an, M., &
Brumback, B. (2000). Epidemiology, 11(5), 550–560]. The above two methods are illustrated
by addressing the motivating question. Implications for prevention researchers who wish to
estimate total causal effects using longitudinal observational data are discussed.
KEY WORDS: confounding; weighting; total effect; time-varying; graphical approach .
Observational data areoften used to address
prevention questions concerning the consequences
of an adolescent’s actions on drug use. Consider the
motivating question, “If alcohol initiation could be
delayed, would that in turn cause a delay in mari-
juana initiation?” The answer to this question could
The Methodology Center, Department of Human Development
and Family Studies, The Pennsylvania State University, Univer-
sity Park, Pennsylvania.
Institute for Social Research, Department of Statistics, The Uni-
versity of Michigan, Ann Arbor, Michigan.
Departments of Communication and Psychology, The University
of Kentucky, Lexington, Kentucky.
Correspondence should be directed to Bethany Cara Bray The
Methodology Center, Pennsylvania State University, 204 E.
Calder Way Suite 400, State College, Pennsylvania 16801; e-mail:
be used to anticipate whether an alcohol use preven-
tion program implemented during adolescence might
also have effects on marijuana use. The answer is pro-
vided by the total causal effect of delaying the timing
of alcohol initiation (predictor) on the timing of mar-
ijuana initiation (response). As is well known, a fun-
damental problem in addressing this question with
observational data is the presence of confounders.
Confounders are common correlates of the
predictor and response, and often provide alternate
explanations for the observed relation between the
two. An example of a common correlate of alcohol
and marijuana initiation is peer pressure resistance.
Adolescents with high levels of peer pressure resis-
tance may be less likely to initiate both alcohol and
marijuana use. In this case, failing to take peer pres-
sure resistance into account would result in a biased
2006 Society for Prevention Research