A 10-Year Retrospective of Research in Health Mass
Media Campaigns: Where Do We Go From Here?
SETH M. NOAR
Department of Communication, University of Kentucky, Lexington,
Kentucky, USA
Mass media campaigns have long been a tool for promoting public health. How
effective are such campaigns in changing health-related attitudes and behaviors,
however, and how has the literature in this area progressed over the past decade?
The purpose of the current article is threefold. First, I discuss the importance of
health mass media campaigns and raise the question of whether they are capable
of effectively impacting public health. Second, I review the literature and discuss
what we have learned about the effectiveness of campaigns over the past 10 years.
Finally, I conclude with a discussion of possible avenues for the health campaign
literature over the next 10 years. The overriding conclusion is the following: The
literature is beginning to amass evidence that targeted, well-executed health mass
media campaigns can have small-to-moderate effects not only on health knowledge,
beliefs, and attitudes, but on behaviors as well, which can translate into major public
health impact given the wide reach of mass media. Such impact can only be achieved,
however, if principles of effective campaign design are carefully followed.
Mass media campaigns ha ve long been a tool for promoting public health. Ear ly
examples in the United States include the Reverend Cotton Mather’s campaign to
promote inoculation for smallpox in the early 1700s and Dr. William Alcott and
Reverend Sylvester Graham’s healthy eating campaign in the early to mid 1800s.
Using pamphlets and personal appeals, Mather was able to demonstrate that those
who were not inocula ted were much more likely to die from smallpox compared with
those who were inoculated (Paisley, 2001). Using magazines, books, and booklets, as
well as a health food store, Alcott and Graham (later known for the ‘‘graham
cracker’’) promoted healthy eating and today hold a place in the history of veg-
etarianism (Iacobbo & Iacobbo, 2005; Perloff, 2003).
More recently, Rogers and Storey (1987) chronicled the history of mass media
campaigns in the United States, pointing out that the 1940s and 1950s were an era
of minimal effects, when many large-scale campaign efforts failed and scholars
hypothesized about why this was the case. The 1960s and 1970s ushered in a cam-
paigns can succeed era, in which campaign successes, most notably the Stanford
3-city Heart Disease Prevention Program (SHDPP), brought new optimism to the
I gratefully acknowledge Phil Palmgreen for his mentorship in the media campaigns area
and the special issue editors and anonymous reviewers for their helpful comments on earlier
drafts of this article. I also thank Ashley Clark and Abby Miller for assistance retrieving arti-
cles and building a reference database for the current project.
Address correspondence to Seth M. Noar, Ph.D., Assistant Professor, Department of
Communication, University of Kentucky, 248 Grehan Building, Lexington, KY 40506-0042,
USA. E-mail: snoar2@uky.edu
Journal of Health Communication, 11:21–42, 2006
Copyright # Taylor & Francis Group, LLC
ISSN: 1081-0730 print/1087-0415 online
DOI: 10.1080/10810730500461059
21
campaign literature. Campaign scholars began to blame ineffective campaigns,
rather than the recipients of those campaigns, for a lack of effects, and began to
uncover and formalize principles of effective campaign design. The success and
impact of the SHDPP has thus been described as ‘‘the most important single turning
point in the rise of the health communication field’’ (Rogers, 1996, p. 16). Finally,
the 1980s and 1990s have been described as a moderate effects era. As more was
learned about campaigns, scholars came to have a new appreciation for how cam-
paigns work, what makes them effective, and what their limit s may be. While we
have witnessed a greater accumulation of campaign successes in this era, campaign
failures have still abounded (Rogers & Storey, 1987).
So, where does that leave us in 2006? As will be argued in the current article, we
currently find ourselves in an era that might be termed a conditional effects era,in
which we have not necessarily discovered new principles of campaign design but
rather have seen many of the principles that were formalized in previous eras effec-
tively and creatively put into action. In fact, it will be argued that as campaign
designers have increasingly attended to principles of effective campaign design, we
have witnessed increasing health mass media campaign successes.
Purpose of Current Article
The purpose of the current article is threefold. First, I discuss the importance of health
mass media campaigns and raise the question of whether they are truly capable of effec-
tively impacting public health. Second, I review the health mass media campaign litera-
ture over the past 10 years and discuss what we have learned about the effectiveness of
mass media campaigns in that period. Finally, I conclude with a discussion of possible
avenues for the health campaign literature over the next 10 years. This article is in many
ways an update to Rogers’ (1996) report on health communication strategies that
appeared in the very first issue of Journal of Health Communication.
Mass Media Campaigns: An Effective Tool for Promoting Public Health?
A compelling health communication intervention tool that potentially can address
health attitude and behavioral change across numerous health problems and in
numerous audiences is the mass media campaign. Rogers and Storey (1987) define
a campaign as follows: ‘‘a campaign intends to generate specific outcomes or effects
(2) in a relatively large number of individuals, (3) usually within a specified period of
time, and (4) through an organized set of communication activities’’ (p. 821). What is
the current practice of health mass media campaigns, and have they been found to be
effective such that they are worth the investment of time, resourc es, and further
study by communication scholars? Meta-analyses (research syntheses) of health mass
media campaigns recently have emerged in the literature and two such projects
provide answers to this critical question.
Snyder and Hamilton (2002). Snyder and Hamilton (2002) conducted what
appears to be the first meta-analysis of the U.S. he alth mass media ca mpaign litera-
ture.
1
These researchers meta-analyzed a systematic sample of 48 health campaigns
1
Although the current article encompasses a review of the campaign literature from 1996
forward, these meta-analyses examined literature from differing times frames, including 1974–
1997 (Snyder & Hamilton, 2002) and 1958–1998 (Derzon & Lipsey, 2002).
22 S. M. Noar
that were conducted in the United States and appeared in the published literature.
Overall, they found a mean effect size representing the impact of health mass media
campaigns on behavior to be r ¼ .09. In percentage terms, the average campaign chan-
ged the behavior of about 8% of the population in the expected, positive direction.
In addition, Snyder and Hamilton (2002) found that (1) success varied with differ-
ent behaviors, with seatbelt, oral health, and alcohol campaigns being the most success-
ful; (2) greater effects were found for campaigns focused on adoption of new behaviors
as compared with prevention or cessation of problem behaviors; and (3) greater effects
were found in campaigns with greater reach=exposure. In addition, the strongest
campaign effects were found with regard to campaigns that had a law enforcement
aspect, such as seatbelt campaigns. When these enforcement campaigns are removed
from the analysis, the overall effect of campaigns on behavior drops to r ¼ .05.
Derzon and Lipsey (2002). A second meta-analysis of mass media campaigns
appeared in the drug abuse prevention literature and was conducted by Derzon
and Lipsey (2002). These researchers meta-analyzed 72 published and unpublished
campaign studies that were designed to prevent or reduce youth substance use. Over-
all, they found a mean effect size representing the impact of media campaigns on
behavior to be .04 standar d deviations, and found impact on knowledge gain (.05)
and pos itive attitude change (.02) as well.
In terms of moderators of the effects on behavior, campaigns utilizing radio, video,
and television were found to have greater effects than those utilizing print media, and
campaigns focused on alcohol use had more success than campaigns focused on illicit
drugs or tobacco. Campaigns directed toward youth had little success, whereas those
directed toward parents and retailers had more success. In addition, campaigns includ-
ing certain types of messages were found to be particularly effective (e.g., stressing
alternatives to drug use). Finally, campaigns with certain intervention characteristics
such as supplementing media messages with other campaign components were
found to have stronger effects than those without such characteristics.
Comment on Meta-Analyses. These meta-analytic projects are excellent contri-
butions to the literature, as they are ambitious projects that provide an indication
of the types of effects that average health campaigns produce. Certainly, the good
news is that these meta-analyses suggest that campaigns to date have had actual
impact not only on attitudes and knowledge, but on health behavior as well. Of
course, by most standards these effects have been small.
In addition, we must consider some important caveats when inter preting these
meta-analytic results. First, while the Derzon and Lipsey (2002) project included
unpublished work, the Snyder and Hamilton (2002) project included only published
work. Given academic journals’ bias toward publishing significant findings, there is
likely a publication bias (Rosenthal, 1991) in the group of studies included in the
Snyder and Hamilton (2002) analysis. Thus, one could make the argument that if
a number of failed campaign efforts were included in their analysis, then the actual
effect size would be less than r ¼ .05.
Further, the moderating influences examined in both meta-analyses, such as type
of behavior, target audience, and so forth, are descriptive in nature. While such
analyses generate important knowledge about the state of the literature, such mod-
erators are not theoretical in nature. Thus, a crucial question that we ask here is the
following: Did campaigns that followed effective campaign design principles (Atkin,
2001) have greater influence on knowledge, attitudes, and behavior? While Snyder
Mass Media Campaigns and Health 23
and Hamilton (2002) did find evidence that greater exposure led to greater cam-
paigns effects, other camp aign design principles were not examined. If studies that
follow campaign principles achieve greater effects, then among that group of studies
the impact may be greater than the small overall effects reported here. While these
meta-analyses did not examine this question, and it remains a high priority for futur e
meta-analyses, evidence from selected recent primary studies does suggest that
following effective campaign principles can lead to effects not only on attitudes,
but on health behavior change as well (e.g., Farrelly et al., 2005; Hornik, 2002a;
Palmgreen et al., 2001, 2005; Vaughan et al., 2000). And, these effects appear to
be larger than the very small overall effects reported in the meta-analyses.
Review of Campaign Literature: 1996–2005
In the current article I suggest that the past 10 years of campaign research have not
necessarily resulted in new campaign design principles. Rather, we have seen cam-
paign design principles developed in previous eras effectively put into action in a
number of impressive health mass media efforts. A number of scholars have sum-
marized in detail the principles of effective campaign design, implementation, and
evaluation (e.g., Atkin, 2001; Perloff, 2003; Rando lph & Viswanath, 2004; Rogers
& Storey, 1987; Salmon & Atkin, 2003; Valente, 2001). Although an exhaustive list
of all campaign principles is beyond the scope of the current article, Table 1 lists and
defines some of the major principles of effective campaigns.
2
Further, the literature review provided here will focus on how recent campaign
efforts have used these principles effectively, and though the review is not meant to
be exhaustive, it is meant to be relatively representative of the literature over the past
10 years. The literature collected for this review was published in a great variety of
journals by individuals in numerous disciplines, and was about evenly split between
U.S. and international campaigns. Evidence is indeed accumulating to support the
proposition that mass media campaigns can be effective on the condition that princi-
ples of campaign design are attended to. Thus, I refer to the recent campaign litera-
ture as an era of conditional effects.
Formative Research. Formative research is extremely important to the design
and implementation of a successful mass media campaign. Such research can enable
campaign planners to truly understand their target audience in terms of the problem
behavior at hand, their message preferences, and the most promising channels
through which they can be reached (Atkin & Freimuth, 2001; Valente, 2001). Mass
media campaigns over the past 10 years have taken advantage of a variety of types of
formative research to fulfill many of these purposes (see Table 2) using a variety of
methods including analysis of archival data (e.g., Glik et al., 1998; Storey et al.,
1999), survey work (e.g., Larsson et al., 2004; McDivitt et al., 1997), focus groups
(e.g., Agha, 2003; Stead et al., 2005), and qualitative interviews (e.g., Witte et al.,
1998; Wray et al., 2004). Few studies, however, reported using formative research
to select campaign channels (e.g., Storey et al., 1999).
2
Different authors highlight slightly different sets of principles of effective campaign
design, although there is much overlap in this area as well. The current list of campaign
principles is meant to be a representation of the major principles of effective campaign design,
although sometimes others are discussed as well (see Randolph & Viswanath, 2004; Rogers &
Storey, 1987).
24 S. M. Noar
As an example, Witte and colleagues (1998) conducted focus groups and quali-
tative interviews with a variety of at-risk groups in Kenya as part of the development
of an HIV=AIDS prevention campaign, and used the extended parallel process
model as a theoretical guide. These resear chers found that formative research with
campaign pamphlets and posters indicated that while perceived threat of HIV=AIDS
AIDS was emphasized, important skills such as negotiating condom use as well as
self and response efficacy for safer sex were neglected within the campaign materials.
Thus, such feedback subsequently can be integrated into campaign messages and can
greatly strengthen campaign materials.
Use of Theory. Use of theory as a guide to campaigns may be vital to campaign
success. Theory can serve as a conceptual foundation for a campaign and can suggest
important determinants upon which campaign messages might focus (e.g., Cappella
et al., 2001; Noar, in press; Slater, 1999). Although one campaign review conducted
through 1998 found mass media campaigns to generally not be theory based (Myhre
& Flora, 2000), Table 2 suggests that newer campaigns are increasingly utilizing
theory. In addition, there has been great diversity in the theories being applied in this
area, and many of the theories being used most often, including the theory of rea-
soned action , social cognitive theory, and the transtheoretical ‘‘stages of change’’
model, also are widely studied in the health behavior change literature (e.g., Noar
& Zimmerman, 2005).
Table 1. Summary of major principles of effective campaign design applied to health
mass media campaigns
Success is more likely when campaign designe rs...
Conduct formative research with the target audience to clearly understand the
behavior and the problem area; pretest messages with target audience to be sure
they are both appropriate and effective
Use theor y as a conceptual foundation to the campaign; theory will suggest
important determinants around which to develop messages, and will help ensure
that campaign messages guide individuals through the process of attitude
and=or behavior change
Segment audience into meaningful subgroups based on important characteristics
such as demographic variables, risk characteristics, experience with the
behavior, personality characteristics, and so forth
Use a message design approach that is targeted to and likely to be effective with the
audience segment; develop novel an d creative messages; design messages that will
spark interpersonal discussions and may persuade individuals important to the
target audience (e.g., influencers)
Place messages in channels widely viewed by the target audience; strategically
position campaign messages within the selected channels
Conduct process evaluation including monitoring and collecting of data on
implementation of campaign activities; ensure high message exposure among
members of the target audience, including both reach and frequency
Use a sensitive outcome evaluation design that reduces threats to internal validity
and permits firm causal conclusions about the campaign’s influence on attitudes
and behaviors to be made
Mass Media Campaigns and Health 25
