Cancer Control Research
3R01CA090955-02S2
Fong, Geoffrey T.
EFFECTS OF GRAPHIC WARNING LABELS ON ADOLESCENT SMOKING
Abstract
DESCRIPTION (provided by applicant): The broad objective of the proposed research is to evaluate and understand the impact of a national tobacco control
policy. Specifically, the proposed research will evaluate the psychosocial and
behavioral effects of the new graphic cigarette package warning labels that
were introduced in Canada in December 2000. These labels contain vivid
photographs of the negative health consequences of cigarette smoking, with text
messages inside the package that include efficacy messages for reducing the
threat depicted on the package. In a quasi-experimental design, the
participants will be a total of 12,000 students at 9 high schools in Canada and
6 high schools in the United States. Schools in the United States are matched
so that they are similar to the Canadian schools. Baseline data confirm that
the Canadian and U.S. schools do not differ in smoking behavior. Students will
participate in a 30-minute classroom administered survey 2 times a year for 3
academic years. The first wave of data collection was conducted prior to the
introduction of the new warning labels. There are six specific aims: (1) to
examine whether the graphic warning labels have any effects on smoking
behavior; (2) to examine whether the graphic warning labels enhance
label-relevant psychosocial variables (e.g., label salience); (3) to examine
whether the graphic warning labels diminish general psychosocial variables that
are related to smoking behavior; (4) to examine whether the effects of the
graphic warning labels will vary as a function of exposure to the labels; (5)
to examine whether the effects of the graphic warning labels will vary as a
function of individual difference variables such as sensation seeking and time
perspective; and (6) to examine whether the effects of the graphic warning
labels on smoking behavior are mediated by variables that have been identified
by theory to be important in predicting and understanding smoking behavior.
Data-analytic methods will be employed that take into account the clustered
nature of the data (e.g., random-effects regression models). The proposed
research takes advantage of a "natural experiment" in an important tobacco
control policy, and thus has the potential to gain insight into the
effectiveness of these and other efforts by governments to reduce tobacco use,
and could provide an empirical basis for future tobacco control policies in the
U.S. and other countries. Moreover, the simultaneous focus on policy evaluation
and theory testing may yield insights into the underlying mechanisms that would
explain how and why the policy achieved its desired effects (or why it failed
to achieve its desired effects).
|