ON THE USE OF MASS COMMUl)jlCATlON$ TO PROMOTE
           THE PUBLIC HEALTH

VWiam DeJong, Ph.D., Department of Health and Social Behavior, Harvard School
of Public Health

A key to continuing the improvement in Americans' health is the public's adoption
of healthier lifestyles. Changes in lifestyle can be made only by a well-informed
and motivated public. As a result, the mass media, as never before, have a vital
role to play in advancing the public health.

Even in the case of diseases with a known biological cause, such as AIDS, choices
in lifestyle and behavior can mean the difference between living and dying. Thus,
when confronted in the 1980s by the menace of the AIDS epidemic, U.S. public
health officials turned to the mass media as a means of quickly educating the
American general public about the disease and appropriate prevention measures
(DeJong & Winsten, 1991).

Beyond promoting changes in individual behavior, the mass media have an equally
vital role to play in directing the public agenda to important health issues,
reinforcing community-based programs, and building support for changes in
institutional structures, public policy, or law that will support and sustain the
efforts of individuals to alter their behavior.

Learning from both the successes and failures of past mass communication
campaigns, public health advocates begin the 1990s with a renewed enthusiasm
for using the mass media to promote the public health. Such enthusiasm will be
heard at the Surgeon General's Workshop as participants debate how our nation
can increase the number of organ transplantation donors. To help inform this
debate, this background paper explores in general how public health advocates can
effectively harness the power of the mass media.

TH R L
3
       FMA
EVOLVING VIEW

When the broadcast media emerged after World War II as a major force in
American society, and as commercial advertising brought new sophistication to
satisfying post-war consumer appetites, many public health advocates assumed
that the presentation of factual information through mass communication
campaigns would automatically result in changed attitudes and improved health
behavior (Atkin, 1979; Griffiths & Knutson, 1960).

37


Subsequently, when several mass media campaigns produced disappointing results
(e.g., Hyman & -Wheatsley, 1947; Star & Hughes, 19501, this boundless optimism
was replaced in the `1950s and 1960s by a decidedly more pessimistic view that
the strategic use of mass media is doomed to failure (Atkin, 1979, 1981). It was
recognized that mass communication campaigns could reinforce existing attitudes
and behavior (Alcalay, 1983; Schlinger, 19761, but this outcome was viewed as
being the most that could be achieved (Klapper, 1960). Little more could be done,
it was argued, because the audience will choose to attend to and retain
information that supports existing opinions (Bauer, 1964; cartwright, 1949), and
because it will employ various psychological defenses to fend off ego-threatening
information (Bauer, 1964)`.

Skeptics noted that the apparent success of commercial advertising provided a
misleading example, for the simple reason that influencing people's health-related
behavior presents a far greater challenge than influencing their brand preferences
(Griffiths & Knutson, 1960; Rehony, Frederiksen, & Solomon, 1984). Several
considerations supported this skeptical view.

First, because of the widespread behavioral changes needed to bring about
significant improvements in public health, health promotion campaigns must
establish more ambitious objectives, even while usually having fewer financial
resources. In contrast, commercial marketing campaigns are considered a
tremendous economic success when they result in even a 1 percent increase in
market share (Rosenstock, 1960; Schlinger, 1976).

Second, the behaviors that public health advocates seek to change are often
ingrained habits or have accrued a cultural meaning or emotional sig.nificance that
fortifies a resistance to change (Robertson & Wortzel, 1971). In contrast,
commercial advertising is seldom designed to inculcate new attitudes or patterns of
behavior, but to heighten and give direction to already existing attitudes and
preferences (Lazarsfeld & Merton, 1971; Rosenstock, 1960; Wiebe, 1951).
Moreover, while many of the benefits of health maintenance and disease
prevention are delayed and uncertain, product consumption affords tangible and
immediate gratification (Lefebvre, Harden, & Zompa, 1988; Schlinger, 1976).

Third, the very people who would be most responsive to a public health campaign
may have already made the desired changes, leaving a target group of so-called

   ' For addressing some health problems, however, the reinforcement of existing attitudes and
behavior is precisely what is needed (Bauman, Brown, Bryan et al., 19881. When they are on the
threshold of junior high school, nearly all children express a strong disliking for alcohol, tobacco,
and other drugs and indicate that they do not intend to use them (DeJong, 19871. A key to
prevention strategies, therefore, is to find ways to bolster those behavioral intentions as the
children grow older.

38


"late adopters" that is much more difficult to persuade (Bloom & Novelli, 1981;
Green, Kreuter, Deeds, & Partridge, 1980; Green & McAlister, 1984; Kasl,. 1980).
In many cases, that target group enjoys good health, at least currently, and this
too diminishes their motivation to change, especially when such change is difficult
to accomplish, painful, inconvenient, or expensive (Rosenstock, 1960). In
contrast, commercial marketers try to reach consumers who are likely to want their
product and can afford it (DeJong, 1989).

Finally, Wallack (1981, 1990a) notes that U.S. health campaigns often take place
in a relatively hostile environment created by strong economic interests. For
example, tobacco and alcohol advertisements encourage consumer indulgence,
which undermines the less appealing messages of prudence and restraint offered
by prevention education (Jacobson, Atkins, & Hacker, 1983)*. Beyond..that,
tobacco and alcohol companies, abetted by a financially vulnerable media, have
limited the public's exposure to information about the long-term consequences of
alcohol and tobacco use (Gerbner, 1990; Hacker, Collins, & Jacobson, 1987;
Warner, 1985; Warner 81 Goldenhar, 1989; Weis & Burke, 1986) and sometimes
exert political pressure to disrupt prevention campaigns (e.g., Wallack & Barrows,
1982-83).

This restricted vision of what mass communication campaigns can accomplish was
reluctantly embraced by most public health advocates. One reason is that it
contravenes the common wisdom that the mass media, especially television, exert
enormous influence over our ideas, values, and behavior (Gerbner, 1987; Roberts
& Maccoby, 1985). Moreover, both broadcast and print media are a cost-efficient
way of reaching millions of people with tiealth messages (Gerbner, 1987; Warner,
1987). Indeed, previous studies that were said to support a circumspect view of
mass communication campaigns often focused on proportional success rates with-
out taking into account the large audiences that mass media can have (Flay &
Sobel, 1983; Warner, 1987).

By the 1980s a new view had evolved. The current perspective on mass
communication campaigns is more balanced, based on the growing recognition that
when long-term mass communication campaigns are designed and executed
according to certain principles, they can play a meaningful role in changing health-
related behavior and lifestyles (DeJong & Winsten, 1990a). The power of the
mass media to effect behavior change is less dramatic than once hoped (or feared),

   ' In several foreign countries (e.g., Austria, Norway, Greece), all forms of cigarette advertising
have been banned, end the governments have launched vigorous anti-smoking campaigns (e.g.,
Doxiadis, Trihopoulos, & Phylactou, 19851. This absence of opposing information achieves what
Lazarsfeld and Merton 1197 1) call a condition of "monopolization. n As Flay (1966) has noted, the
chances of achieving a total ban on tobacco promotion in the U.S., coupled with an aggressive
anti-smoking campaign, are extremely remote.

39


but it is real. Mass communicatidn campaigns can keep, health-related problems
and policy options at the top of the public's.agenda and help bind community
resources into a cohesive force for change; inform citizens about their community's
attack on the problem and inspire their full participation; shift the meaning of
health-related behaviors and thereby facilitate a shift in social norms; and motivate
the adoption and maintenance of healthy lifestyle behaviors. The obstacles to
changing health-related behaviors and lifestyles through mass communications are
substantial but not insurmountable.

Public health advocates have a better understanding of what mass communication
campaigns can and cannot do as a result of several important studies conducted
during the past 20 years. Two successful campaigns, both focused on promoting
changes in individual behavior, were especially influential: the Stanford. Three
Community Study, focused on cardiovascular risk,reduction (Farquhar, Maccoby,
Wood, et al., 1977; Maccoby, Farquhar, Wood, & Alexander, 1977; Meyer, Nash,
McAlister, et al., 1980; Stern, Farquhar, Maccoby, & Russell, 1976); and the
North Coast "Quit for Life" smoking cessation campaign in Australia (Egger,
Fitzgerald, & Frape, 1983).

Failures were instructive too (e.g., Robertson, Kelley, O'Neill, et al., 1974; Vdry,
1974). A review of these campaigns made evident that the long period of
skepticism about the strategic use of mass communications was strongly
influenced by findings from short-term studies that were seriously flawed due to
poor planning, inappropriate messages, or deficiencies in research design.

Thus, studying both the successes and failures of past campaigns, public health
advocates have developed a more sophisticated understanding of how mass
communication campaigns can change health-related behaviors and lifestyles. The
remainder of this background paper reviews several important lessons that have
been learned.

LESSON 1: ESTABLISH A LONG-TERM COMMITMENT

The potential of the mass media for stimulating and reinforcing widespread
changes in behavior can typically be realized only over a long period of time (Flay &
Burton, 1990). Seduced by the apparent ease by which commercial marketers
influence consumer brand preferences, public health advocates have frequently
launched short-term advertising campaigns that have little chance of success
(Bandy & President, 1982; Maccoby, 1987). For example, one well-known study
evaluated an anti-drug television and radio campaign that lasted only 8 weeks
(Hanneman, Eisenstock, Hunt, 81 Weinbeck, 1977). Not surprisingly, the campaign
was found to have no effect. While this example is extreme, in general, past

40


efforts have often ignored the fact that attitude and behavior change in response
to mass communications is "usually characterized by a slow process of erosion and
accretion rather than by one of sudden ueheaval and conversion," (Roberts &
Maccoby, 1985, p. 547). Commercial marketers know this too: brand loyalty (or
"goodwill") in fact takes years to nurture and develop (Bovee & Arens, 1986).

The value of a long-term perspective is evident when we'consider the role of mass
media in the evolution of U.S. social norms regarding the acceptability of tobacco
smoking. Once viewed as a sexy, glamorous habit, smoking-today is seen as a
sign of poor self-discipline; to be a smoker now is to carry the weight of stigma
(Cooke, 1989; "All Fired Up," 1988). This evolutionary change, which began with
the Surgeon General's widely publicized report in 1964; has had a dramatic effect
on public health. While 43 percent of the U.S. adult population smoked in 1964,
under 30 percent did so in 1985; between 1964 and 1985, reduced tobacco
consumption resulted in the avoidance of an estimated 789,200 premature
smoking-related deaths (Warner, 1989). Through news reports, anti-smoking
advertising, and fewer portrayals of cigarette smoking in entertainment
programming (Signorielli, 1990), the mass media played a major role in stimulating,
amplifying, and sustaining this gradual, and eventually radical, shift in smoking
norms and behavior (DeJong & Winsten, 1990a).

LESSON 2: APPLY A BEHAVIOR CHANGE MODEL TO IDENTIFY INTERMEDIATE
OBJECTIVES

Very few mass commun.ication campaigns can be expected- to produce an immed-
iate change in behavior (McGuire, 1984, 1985; Roberts 81 Maccoby, 1985;
Skirrow, 1987). Whether a campaign can achieve that objective depends largely
on the specific problem being tackled -- its complexity; the level of public concern
.about it; at what stage in the behavior change process the target audience can be
found; whether early or late adopters are being targeted; and the personal,
interpersonal, institutional, and societal barriers to change (Atkin, 1981; Wallack,
1990a). Hence, rather than focusing on immediate behavior change, it is often
more realistic and appropriate to concentrate on achieving intermediate objectives
that set the stage for or otherwise contribute to behavior change in the long-term
(Farquhar, Maccoby, & Solomon, 1984; `Flora, Maibach, & Maccoby, 1989).

A useful framework for understanding these intermediate objectives is the
communication/behavior change model developed for the Stanford Three
Community Study (Maccoby & Alexander, 1980)3.  This model elaborates a series

   a  This framework was inspired by Cartwright's 119491 description of the three stages of
change that a campaign must achieve in order to influence behavior -- an increase in awareness
and knowledge (cognitive structure); a change in attitudes, leading to e heightened motivational

41


of steps in the behavior change process which,can be addressed through various
mass communications channels. Specifically, mass communication campaigns can
accomplish the following objectives: 1') increase awareness of a health problem
and establish it as a priority concern; 2) increase knowledge and change beliefs
that Impede the adoption of health-promoting attitudes; teach new behavioral
ski!!s; and 4) provide supports for sustaining behavior change.

Increase awareness'of a health orobiem and establish it as a orioritv concern.
Mass communications can be used to get the public thinking about a health issue,
a function commonly referred to as "setting the agenda" (Roberts & Maccoby,
1985). The objective at this stage is to present information that makes an issue
interesting, understandable, and personally meaningful, leading ultimately to self-
appraisal and a consideration of possible action (Bandy & President, 1982; Griffiths
& Knutson, 1960; McGuire, 1984, 1985; Roberts & Maccoby, 1985; Skirrow,
1987).

Increase knowledqe and chanae beliefs that imoede. the adootion of health-
promotina attitudes. Roberts and Maccoby (1985) argue that changes in
cognitions are a necessary precursor to eventual changes in attitudes and behavior:
1,
. . . any influence of mass media content depends on how people interpret mes-
sages relative to previously established conceptualizations of the world --
fundamentally a cognitive effect," (p. 547). Attitude and behavior change, they
assert, can eventually follow, making changes in knowledge and beliefs important
outcomes in and of themselves (Fishbein & Ajzen, 1975).

Demonstration of a campaign's impact on knowledge, beliefs, and attitudes is
often dismissed as unimportant because of the often-cited inconsistency between
attitudes and behavior (Bandy & President, 1982). In fact, attitudes are good
predictors of behavior when the attitude and behavior are measured at corres-
ponding levels of specificity (Ajzen & Fishbein, 1977). Thus, general attitudes
toward drugs will not necessarily predict whether a person will try a marijuana
cigarette at a particular time and place. To be predictive, the attitude
measurement should be equally specific. Attitude-behavior consistency is also
more likely when social norms support the behavior (Fishbein & Ajzen, 1975).

At the same time, it must also be remembered that attitude change is not a
necessary precursor to behavior change (Bern, 19701, in contrast to past
assumptions about the absolute primacy of attitudes in the behavior change

state (motivational structure); and engagement in the actual behavior lbehavioral structure/. The
key to inducing a given behavior, according to Cartwright, is having an appropriate cognitive and
motivational system "gain control of the person's behavior at a particular point in time, W lp. 2641.

42


process. Indeed, a persuasive communication may induce curiosity to engage in
new, low-risk behaviors on a trial basis, with .a change in,attitudes emerging later
in consequence (Flay & Burton, 1990; O'Keefe & Mendelsohn, 1984).

It is important to note here that when detailed, complicated, or politically
controversial information must be communicated, television and radio spots,
billboards, and print advertisements can stimulate further information-seeking by
promoting information hotlines (Pierce, Dwyer, Frape et al., 1986; Stein, 1986) or
the availability of pamphlets and other written materials (Metidelsohn, 1973;
Schlinger, 1976). It is critical that people not be merely exhorted to obtain
additional information but that specific strategies. be explained or modeled
(Solomon, 1983). This has been a central feature of several public health
campaigns, including national AIDS campaigns in the U.S. (DeJong 8; Winsten,
1991).

Teach new behavioral skills. Important behavioral.skills can be taught through
modeling or step-by-step instruction (Alcalay, ,1983; Bandura, 1977). Until
recently, behavior change was most often conceived in terms of altering existing
patterns of behavior; that- is, certain unwanted behaviors were to be suppressed,
and other behaviors, already in the behavioral repertoire were to be brought forth.
As a result, the power of the mass media to expand people's behavioral repertoire
was frequently overlooked (Alcalay, 1983).

The performance of newly acquired behaviors can be enhanced by demonstrating
how various barriers to behavior change can be overcome, thus increasing
perceptions of "self-efficacy" (Bandura, 1984). Self-efficacy is not a global
concept like self-esteem, but deals with a person's perceptions about his ability to'
act out a specific behavior at a particular time and place. Whether a person holds
such a belief is predictive of subsequent behavior change, whether it is the
suppression of existing behaviors or the acquisition of new ones (Strecher,
DeVellis, Becker, & Rosenstock, 1986). Vicarious experience is an important
source of information for efficacy expectations; thus, mass communications can be
employed to change self-efficacy by modeling determined effort that leads to
success. Performance of newly acquired behaviors can also be enhanced
suggesting so-called "cues to action" in the physical and social environment that
can stimulate a person to call up and apply a particular behavioral repertoire in the
appropriate circumstances (Maiman & Becker, 1974).

Provide suooorts for sustainina behavior chanaa. Learning and maintaining a new
pattern of behavior requires that people know how to monitor their behavior, apply
self-reinforcement strategies, and anticipate, eliminate, or cope with environmental
or social stimuli that trigger unwanted or competing behaviors. Mass
communications can be used to teach these self-management techniques (Bandura,
1977; Flay, DiTecco, & Schlegel, 1980).

43


Social support is also important for maintaining, new behavi,ors. Mass
communications can he!p here too by: 1) communicating the fact that others have
also adopted new behaviors.and are facing the same struggles in trying to maintain
them (Mogielnicki, Neslin, Dulac et al. 1986); 2).stimulating the social support of
opinion leaders, family, and peers (Green & McAlister; 1984; Rogers, 1983); and
3) reaching people how to elicit from others the support they need.

Application of the ccmmunicationlbehavior change model first. requires that
campaign planners establish where in the behavior change process the target
audience can presently be found. From there, the campaign can try to move the
audience sequentially through the remaining steps'. Consider the case of AIDS
prevention. Presently, most American adolescents and adults know a .great deal
about what is required to prevent AIDS, thanks in part to national public'awareness
campaigns. At this point, the successful promotion of condoms requires moving
beyond basic factual information about AIDS prevention to motivating messages
that address the various psychological barriers that impede condom use and that
present the personal and social benefits that their use can bring if approached in
the right way (Solomon & DeJong, 1986, 1989; DeJong & Winsten, 1991).

With this model in mind, campaign planners should develop a strategic plan that
divides the campaign into distinct phases,.each with realistic, specific, and
measurable objectives (DeJong & Winsten, 1990a; Green & McAlister, 1984;
Maccoby & Alexander, 1980). Once the planning phase is over and measurable
objectives are set, campaign organizers will have a framework that can be used to
guide media selection, message development, and campaign evaluation (Maccoby
& Solomon, 1981; Solomon, 1982).

While this point seems obvious, reviewers of mass communication campaigns have
f,requently noted that campaigns often have unrealistic goals, because the
objectives are vague or defy valid and reliable measurement, because the use of
mass media to effect the desired change is unfeasible, or because insufficient time
and resources have been committed to the campaign (Bandy & President, 1982;
McGuire, 1984; Mendelsohn, 1973; Schlinger, 1976; Solomon, 1982; Wallack,
1980, 1981).

   * It should be remembered, however, that people do not always go through this sequence of
steps in ex8ct order, especially if the choices among behaviora/ alternatives are unimportant to
them, or if they have somehow been induced to perform the behavior in the absence of prior
aititude Ch8nQe (MeGhe, 1989; Solomon, 1989).

44


JFSSON 3: APPLY THE "CONSUMFR" ORIENTATION OF SOCIAL MARKETING

By the late 197Os, several reviewers of mass communication campaigns had noted
that the public health community knew too little about modern marketing (e.g.,
Schlinger, 1976). In recent years, however, Federal health officials and other
public health advocates have become increasingly sophisticated in designing and
executing campaigns, primarily as a result of the "consumer" orientation that
characterizes Madison Avenue's approach to marketing. This section describes the
central tenets of this so-called "social marketing" approach5.

All campaian messaaes should be directed to a well-defined taraet audience. In
the argot of marketing, a target audience should be "segmented" into subgroups
with similar geographic, demographic, psychological, and problem-relevant
characteristics (Flay & Burton, 1990; Maccoby 81 Alexander, 1980; Solomon,
1983). With this information, campaign planners can develop strategies that are
appropriate for each segment (Simon, 1974). The target audience can be
members of the general public or business and government leaders.

Commercial marketers rely on a variety of market segmentation techniques. At a
minimum, markets are typically divided according to geographic location (e.g.,
urban, suburban, or rural residence) and consumers' demographic characteristics
(e.g+ gender, age, race/ethnicity, education, occupation, income, religion). More
recently, so-called "geo-demographic" techniques have been developed whereby an
individual is classified into one of 40 lifestyle categories defined by the social class
and consumer spending patterns of that person's postal "zip code" area
(Townsend, 1985).

"Psychographics" classifies consumers on the basis of their psychological make-up
and lifestyle. As defined by Mitchell (1983), psychographics describes "the entire
constellation of a person's attitudes, beliefs, opinions, hopes, fears, prejudices,
needs, desires and aspirations that, taken together, govern how one behaves," (p.
vii). The most popularly used system for adults is the Values and Lifestyles (VALS)
typology developed in 1978 by SRI International (Mitchell, 1983). A new version
of this typology, designed to correct deficiencies in the original system, was
released in the mid-1980s.

For.any mass media campaign designed to change health behavior, it is important
to have a "psychographic" system for segmenting the audience. Typically,

   '  Social marketing is defined by Kotler and Zattman (197 1) as "the design, implementation,
and control of programs calculated to influence the acceptability of social ideas and involving
considerations of product planning, pricing, communications, distribution, 8nd marketing research, -
/p. 3). In short, social marketing is the application of commercial marketing and distribution
strategies to the promotion of social goals (Kotler, 1984; Solomon, 19891.

45


however, commercially available systems, with their focus on consumer spending,
will be inadequate to the task. Thus, campaign planners often "customize" their
own typology based on problem-specific research, beginning with focus group or
one-on-one interviews and mov'ing,on to formal survey research (e.g., Lastovicka,
Murry, Joachimsthaler et al., 1987; Slater & Flora, i 991).

U.S. "market segments*' that deserve special consideration are non-English
speakers. Too often, English-language campaign materials are translated for
minority group audiences, with cultural nuances ignored (Farquhai, Fortmann,
Maccoby et al., 1985a; .McAlister, Ramirez, Amezcua et al., 1987). There are
distinct subcultures within these groups, each with its own history, customs,
dialect, and patterns of health-related behavior. Because it will often be
impractical to develop materials for each subgroup, care must be taken to' create
materials that transcend these differences and "ring true" to the largest number of
persons possible.

*
                                        rich
ynderstandina of the taraet audience (Atkin & Freimuth, 1989; OCC, 1989).
Focus groups are the qualitative method most frequently used by advertising
researchers (Basch, 1987; Folch-Lyon & Trost, 1981). With this technique, a
small group of eight to 12 individuals is interviewed in depth by a trained
moderator.. The moderator, following a previously developed agenda, creates a
non-threatening, accepting atmosphere to draw out each participant and strives to
facilitate discussion among the participants.  Success is highly dependent on the
skills of the moderator. Sessions are usually conducted at a facility that permits
observers to watch from behind a one-way niirror.

The focus group technique is often viewed skeptically. The research usually
involves a relatively small number of respondents. They are recruited to meet
certain selection criteria and do not know one another, but they are not selected at
random, so there is no guarantee that they are truly representative of the
population as a whole. Moreover, the technique can be inappropriate for gleaning
information about highly charged or embarrassing topics.

Some marketing researchers claim that the technique is largely ineffective with
adolescents because of their heightened self-consciousness and sensitivity to peer
pressure.. Other researchers disagree, noting that a moderator with sufficient skill
can work successfully with adolescents (Greenbaum, 1988). A type of focus
group that seems especially effective with teens is the so-called phenomenotogical
focus group; the respondents do not respond to direct questions, but are given an
appropriate task to accomplish as a group without the moderator's help (Basch,

46


1987). With elementary school children, be,cause of their more limited
communication skills and attention spans, role-playing, fantasy play, drawing, and
other experiential techniques are usually necessary to elicit useful information
(Greenbaum, 1988).

!n favor of focus group research, it can provide large amounts of rich, detailed
information in a short period of time and at relatively little.cost. Especially
important is information on the target aud-ience's speech patterns, body language,
and style of dress. These observations can provide valuable clues as to how a
product, service, or idea can be "positioned" ahd marketed.

An alternative technique is a series of private, one-on-one interviews conducted by
a trained moderator or clinical psychologist. The interviewer follows a prepared
protocol, but detailed follow-up questions may also be asked. There are several
advantages of this technique. Issues can be probed in greater depth. Respondents
are typically more willing to share highly personal information, and group dynamics
do not bias the subjects' responses. Special facilities are not needed. On the
other hand, this method is more expensive and time-consuming than focus group
interviews. Moreover, clients or other members of the project team do not observe
the interviews so that respondents will speak more freely.

The use of clinical psychologists for these interviews is intriguing. Recognizing
that people's self-reports concerning the reasons for their behavior are frequently
inaccurate (Nisbett & Ross, 1980), those who employ this method note that
clinical psychologists have a range of projective techniques at their disposal to
assess the psychological or emotional benefits that a product offers,. benefits
which are often not consciously'acknowledged. Moreover, psychological methods
or theories can be readily applied to explaining consumer behavior and generating
advertising ideas. For example, products might be seen as answering basic ego
needs, such as a need for self-esteem or dependency needs?

   * Another useful framework for generating campaign ideas is provided by McGuire 119891. By
his scheme, general theories of human motivation are organized into four broad categories.
corinitive stabilitv theories: According to this set of theories, human behavior is motivated by a
need to maintain beliefs and attitudes that are consistent. A strategy suggested here is to focus on
the inconsistency between the target audience's values and their behavior.  Coanitive Drowth
theories: By these theories, human behavior is motivated by certain needs that are cognitive,
rather than affective, such as a need for stimulation, or a need to believe in one's capacity to
maintain control over the environment. Affective stabilitv theories: By these theories, human
behavior is motivated by a need to maintain a stability of affect. Fear appeals, for example, seek to
induce a tension or stress that the target audience would then be motivated to eliminate. Affective
growth theor&: These theories, also emphasizing growth, deal with the realm of emotions, such
as the need for love and acceptance.

47


Qua!itative research is especially valued for the new insights it can generate (Bovee
& Arens, 1986). Marketing researchers typically propose that this research be
followed with surveys and other quantitative methods to test the hypotheses
genera,ted, but commercial clients often skip this step because of time or financial
constrraints or because they find the insights provided through the qualitative
research to be especially compelling or at least consrstent with their past
experience (Szybillo & Berger, 1979). -Public health advocates are likely to do the
same, but whenever resources and time permit, findings from qualitative research
should be validated through surveys or other more rigorous methods (e.g.,
Lastovicka, et al., 1987).

Once the develooment of camoaian materials is underwav. extensive oretesting
must be conducted to ensure that thev are aoorooriate for and aooeal to the taraet
grouo (Atkin & Freimuth, 1989; OCC, 1989). Many design options are available to
campaign designers, and each has its strengths and weaknesses, as evidenced by
both the findings of social psychological research and the successes and failures of
past campaigns. Because the research on persuasion has revealed so many
complex interactions among source, message, and channel variables, and because
the results depend in part on the audience and the specific issue at hand, making a
reasoned decision about which option to select is extremely difficult (McGuire,
1985, 1 989)7.

As a result, it is common to find disagreement among the experts about how
campaign messages should be designed. For example, consider the "style" of the
campaign. Atkin (1979) notes that public communication campaigns are often dry
recitations of factual information, with little attention to style or production values,
and he urges that campaigns designers build in "excitement, drama, and humor" to
buttress entertainment value. In contrast, Bauman et al. (1988) deliberately
avoided a "flamboyant" style of advertising in their anti-smoking spots, based on a
consultant's critique of previous anti-smoking messages as "self-conscious

  McGuire 11989) presents a "communication-persuasion matrix w as a framework for reviewing
the possible effects of various message design and implementation options. The matrix emerges
from a cross-tabulation of "input variables" and "output factors. W Input variables include source
variables, such as communicator credibility and attractiveness; message variables, such as type of
appeal (e.g., celebrity `endorsement, humor, fear) or presentation of alternative viewpoints; channel

(medium) variables, including the number and type of media used; receiver variables, including
demographic and psychographic characteristics; and target variables, or specific features of the
problem behavior itself. Output factors are the sequence of processes that mediate the impact of a
persuasive communication on behavior, which A tkin and Freimuth (19891 have categorized in five
stagesi exposure, information-processing, cognitive learning, yielding, and utilization.

48


imitations" of commercials. They also. noted that, while such advertisements
might increase attention and even earn public accolades,, they might fail in their
mission if the audience is .distracted from attending to or remembering the message
(Benn, 1978)*.

As noted, these considerations are also confounded by-the nature of the audience
and the specific issue involved. For example, a "two-sided" approach, which
acknowledges ideas or arguments in opposition to the advocated position, but then
refutes them, is superior -- but only whenmembers of the audience are highly
sophisticated and well-educated, when they are initially opposed to the advocated
position, and when they will later be exposed to information that supports that
opposing perspective (Atkin, 1979).          

Even when there is general agreement about which approach is best, it must be
remembered that no "rule" of advertising works all the time and that widely
successful strategies can be rendered ineffective .through overuse. Cinematic
techniques, formats, and styles that command attention and-have strong audience
appeal today can eventually become part of the background clutter that is ignored
tomorrow. Audience values change over time, too. For example, appeals
developed in the 1980s that drew on baby-boomers' need for self-fulfillment may
not work in the 1990s as that generation moves on to other concerns, such as the
need to build warm and nurturing relationships (Flay & Burton, 1990).

Past campaigns and research findings should be used, therefore, not as a blueprint,
but as.a source of ideas that need to be pretested within the context of a new
campaign (Solomon, 1983). For this reason, with each step of campaign
development, planners should conduct formative research involving representatives
of the target group to double-check the appropriateness and potential effectiveness
of campaign materials (Maccoby & Alexander, 1980). Unfortunately, careful
formative research of this type is often not done (Palmer, 1981).

A formative research plan should also include television and radio managers and
print editors -- the `*gatekeepers" who control access to the media. These
individuals will have their own ideas about what is appropriate and effective
(Hammond, Freimuth, & Morrison, 1987). They also have responsibility for making
sure that all advertising, including public service announcements, meet their
station's standards of propriety and will not alienate important segments of their

   ' A further cokplication: Research suggests that if the distraction from a persuasive message
is only enough to prevent silent rehearsal of counter-arguments to the message but not enough to
interfere with message reception, such advertisements might actually be m effective (Festinger
& Maccoby, 19641.

49


audience (DeJong & Winsten,. 1991; Flay &,Burton, 1990). Therefore,. they `should
be involved early on in planning and be.given the opportunity to react to materials
during each phase of the development process (Flay, 1986).

The principle method used in formative research at this stage is focus group
interviews. One use of this research technique, to learn basic information about
the target audience;was described earlier. This technique is also used to evaluate
message concepts (that is, brief summaries of key benefits or "promises"),
preliminary media executions, and finished products. Issues that can be explored
through such research include the material's ability to command attention;
audience comprehension and recall; aesthetic appeal; and the message's credibility,
relevance, and acceptance by the target group.

How frequently campaign planners conduct this type of formative research
obviously depends on the resources they have at their disposal. At a minimum,
focus groups should be conducted to test preliminary executions, such as scripts,
storyboards, and mock-ups of print advertisements. Tests of finished products are
generally less criticalg. Often, tests of preliminary message concepts can be
incorporated into initial focus group research being conducted to learn about the
target a.udience.

Mass communication messaaes should address the taraet audience's existinq
knowledae and beliefs. In order for a persuasive communication to succeed, it
must anticipate the audience's points of resistance and then .address them
(Lefebvre, Harden, & Zompa, 1988). In the Health Belief Model (HBM) (Maiman &
Becker, 1974; Janz & Becker, 1984),,those resistance points are defined in terms
of the target audience's underlying beliefs about a particular health problem and its
prevention or cure.

These beliefs concern the following: the person's perceived susceptibility to the
disease or health problem; the seriousness of the disease or problem if it is
encountered, including health, social, and financial consequences; the
effectiveness of the prescribed course of action in reducing the threat to health;
and the barriers to executing that behavior, such as financial cost, inconvenience,
pain, and lost time (Bloom & Novelli, 1981; Solomon, 1989). Also important is the
person's perceptions of popular opinion and normative perceptions for the roles

   ' In some cases,. however, this type of testing is essential. One example illustrates this point:
the California Prevention Demonstration Program, a mass communication campaign aimed at
reducing problems related to alcohol abuse. For an interim evaluation, those survey respondents
who had seen each television commercial being aired were asked to say what they remembered
about it and its message. Across the commercials, between 35% and 44% of the respondents
misinterpreted them to be pro-drinking rather than pro-modeiation messages /WaJJack & Barrows,
1982-831.

50


and behaviors that are appropriate to members of the society (Fishbein & Ajzen,
1975). Messages emphasizing that others have adopted a desired behavior can be
especially persuasive (Farquhar, 1984).

The HBM framework is useful for organizing formative research and then guiding
message content (Solomon & DeJong, 1986). For example, in designing messages
for an anti-smoking campaign, Bauman et al. (1988) discovered that adolescents'
expectations of certain positive and negative consequences to smoking was a good
predictor of their later smoking initiation. -They then designed their campaign to
address these beliefs. Messages focused on short-term consequences of smoking,
such as bad breath and loss of concentration and appetite; possible loss of friends;
possible trouble with adults; and the supposed `fun" and "relaxation" of smoking.

Mass communication camoaions should communicate incentives or benefits for
gdootina the desired behavior that build on the existing motives, needs, and values
gf the target arouo (El-Ansary & Kramer, 1973; Lefebvre & Flora, 1988), a process
that Lazarsfeld and Merton (1971) have called "canalization." The failure to do
this is at the heart of many campaign disappointments.

Commercial marketers have long understood that people are more likely to attend
to and remember messages that meet their needs or support values they believe in
(Flay, 1986; Rehony et al., 1984). Accordingly, product advertising often plays on
peoljle's insecurities, desires, and aspirations, and then "positions" the advertised
product or service as a means of meeting those needs immediately (Solomon,
1989). The key, therefore, is to establish the benefits of the product or service
(Lefebvre & Flora, 1988; Ogilvy, 1985).

The experience of commercial advertisers makes clear that, in many cases, a
health promotion campaign should emphasize benefits that are not related to health
per se, but to universal and more salient human desires for beauty, acceptance,
love, security, status, or wealth (Bonaguro & Miaoulis, 1983; McGuire, 1984;
Roberto, 1972; Schlinger, 1976). In the past, health advocates focused too
narrowly on trying to motivate people to change their behavior for the sake of their
health rather than in response to -other motives, a strategy that proved to be
limiting and self-defeating. Instead, as explained by Cartwright (1949), people
must be encouraged to "see the actionas a path to an existing goal," (p. 261 I.

Commercial advertisers often play on young people's insecurity about their place in
the world and their growing desire for. autonomy. Health campaigns can do the
same. To take one example, a campaign to discourage alcohol, tobacco, and other
drug use can: provide accurate information on the actual level of substance use
among preteens and adolescents, thereby showing that far from "everyone" uses
these substances; urge young people to encourage one another to refrain from
experimenting with and using these substances (Bauman et al., 1988); present

51


peer role models -- older adolescents who are independent, mature, and popular
without drinking, smoking, or using other drugs (McAlister, Perry, & Maccoby,
1979); and illustrate peer approval for refusing or stopping substance use (Liepold,
1986; Worden, Flynn, Geller et al., 1988). Moreover, such campaigns can
encourage young people to select a peer group or clique'that. does not use these
substances and helps bring out their best qualities (McLeroy, Bibeau, Steckler,
Glanz, 1988). A good, example is the 1Be Smart. Don't Start!" campaign
sponsored in the mid-1980s by the National Institute on Alcohol -Abuse and
Alcoholism (Atkin, 1989a).

Because of the importance of social influences in substance use, campaigns aimed
at preteens and adolescents should also focus on the important role of social
pressure, both direct and indirect, from peers, family, and the media; ways in
which tobacco and beer advertisers try to persuade people to use their products
(Atkin & Arkin, 1990; Flay & Sobel, 1983); specific counterarguments to refute
arguments made by others in favor of substance use (McGuire, 1985); and
alternative strategies for refusing various offers of alcohol, tobacco, and drugs
from peers or family members (Polich, Ellickson, Rueter, & Kahan, 1984; DeJong,
1987). The underlying message is that refusing to give into pressure to
experiment with and use alcohol, tobacco, and other drugs is a true demonstration
of independence and freedom.

Commercial advertising experience also makes clear that, whenever possible, the
target audience's attention should be drawn to the immediate consequences of
behavior that have a high probability of occurring. This is especially true for
campaigns directed to preteens and adotescents.

Again consider the case of preventing alcohol, tobacco, and other drug use by
youth. Young people need to be taught the facts about the long-term impact that
substance use might have on their health and well-being. This is most important
when there is a cloud of mythology and misinformation about the addictive
potential and long-term effects of a particular drug, as was the case for cocaine
before the National Institute on Drug Abuse (NIDA) began its public awareness
campaign, "Cocaine: The Big Lie" (Johnston, O'Malley, & Bachman, 1988). But

while an awareness of long-term risks will dissuade many young people from sub-
stance use, the deterrent effect will be far from universal. By age 12, for example,
children have been told repeatedly that tobacco use is dangerous, but a large
number still begin to smoke anyway (Evans, Rozelle, Maxwell et al., 1981).

It is important to understand why information about the long-term sequelae of
substance use have limited effect. First, even when the credibility of factual
information is accepted, people might question whether it really applies to them
(Atkin, 1979). Most young people take good health for granted, and many view
the dire, long-term consequences of substance use as too distant and too unlikely

52


to be of concern to them. Second, their own observations may lead young people
to conclude that experimentation does not necessarily result in addiction, and that
regular use does not necessarily result in severe health consequences. Because
many people do not understand the probabilistic nature of risk, this uncertainty is
an opening for denial (DeJong, 1987). Third, for most young people, substance
use is occasional, not daily, and typically occurs in social situations. As a result,
they might overestimate their own capacity to control the extent of their substance
use, believing that they are fundamentally different from those who develop a drug
dependency. In fact, in some cases this,belief might even stimulate curiosity about
a substance and lead to experimentation (Polich et al., 1984).

In trying to dissuade young people from smoking cigarettes, for example, an
emphasis on long-term risks of heart disease and cancer will have little apparent
effect in delaying experimentation or reducing use. Instead, the emphasis should
be placed on immediate consequences that have. a high probability of occurrence
(Flay & Sobel, 1983; Job, 1988). With greater impact, prevention programs now
emphasize readily noticeable effects such as discoloration of the teeth, body and
mouth odor, loss of concentration and appetite,. and the deterioration of physical
performance (Bauman et al., 1988; Evans et al., 1981). This strategy is also
reflected in a recent public service campaign, "Nit, A Teen," by the federal Office
on Smoking and Health, which portrays smoking as being unattractive to desirable
members of the opposite sex.

Health camoaians should exolore the use of "imaae" or "lifestvle" advertisina to
promote an active, healthv lifestvle. Commercial advertising of this sort allows
consumers to envision the transformation of a less desirable past to a more
desirable future by manipulating images of the consumer as a person with
intelligence, power, status, and popularity (Chapman & Egger, 1980; Graham &
Hamdan, 1987). In essence, image advertising associates product consumption
with the "good life," as defined by the target audience. It rarely seeks to elicit an
immediate behavioral response, but instead tries to evoke an emotional response
that can catalyze later changes in behavior (Schwartz, 1983). As a result, such
advertising needs to be repeated more frequently than advertising that uses a more
direct sales pitch (Benn, 1978).

Image advertising is commonly used to promote cigarettes, especially in women's
and youth-oriented publications (Altman, Slater, Albright, & Maccoby, 1987).
Advertising for Virginia Slims cigarettes is illustrative. The launch of this brand
coincided with the woman's movement, and the advertising copy has always
stressed women's emancipation and growing freedom as an appeal to "liberated"
women (Benn, 1978; Rehony et al., 1984). At the same time, the ads have a
distinctly feminine aura, suggesting that Virginia Slims smokers can compete with
men without sacrificing their femininity. This juxtaposition is apparent in the
campaign slogan "You've Come a Long Way, Baby."

53


Another example is provided by malt liquor advertising directed to low-income
black male consumers. A great deal of such.advertising emphasizes the extra
"power" of the drink, which does have a higher alcohol content than beer, and ties
its use to promises of sexual power and conquest (Postman, Nystrom, Strate, &
Weirrgartner, 1987). In contrast, ads for cognac and'other liquors, which are
designed to appeal to upwardly mobile and economically successful blacks, portray
the product as a sign.of elegance (Hacker et al., 1987).

Other image advertising associates a product with stories, vignettes, themes, or
characters that tap into cultural myths. In so doing,- these ads evoke tha culture's
ideals and give expression to profound and universal emotions (Schwartz, 1983;
Chapman, 1986). Most important, by evoking particular myths, these ads
communicate the essence of those `people who use the product and their place in
the world (Chapman & Egger, 1980). As a result, consumers see aspects of their
real or desired image portrayed in the advertising, and the product becomes a
vehicle for expressing or gaining that identity. The quintessential example of this
type of advertising is the campaign for Marlboro cigarettes, which exploits the
American myth of the "frontier," a free, open space where a resourceful and self-
sufficient man can start fresh and make his way in the world. The "Marlboro Man"
is a man of inner-strength -- self-confident, tough, straightforward, and
independent; he is portrayed, essentially, as the embodiment of American
individualism and potency.

Chapman and Egger (1980) assert that anti-smoking appeals directed at preteens
and adolescents also need to use "image" advertising. One strategy they suggest
is to present a mythical character with whom this group can identify, a character
who demonstrates his power and independence by choosing not to smoke. A
similar strategy was applied in a drunk-driving prevention campaign just pilot-tested
in California, "The Driver,*' which was designed to promote the designated driver
concept (SYSTAN, 1987). "The Driver" is portrayed in television spots as a classic
American hero, the handsome loner who thinks and acts apart from others, but is
also willing to take charge when others must be rescued from themselves.

Image advertising can be used more broadly to promote a health-enhancing
lifestyle, one that is incompatible with substance use and other behaviors that put
health at risk. For example, the Alberta Alcohol and Drug Abuse Commission
(AADAC) in Canada has operated its "Make the Most of You" campaign since
1981 with a focus on encouraging and assisting teens to adopt a healthy lifestyle.
Through a combination of print and electronic media and school- and community-
based programs, the campaign shows that adolescents, with peer support, must
take an active role `in their own growth and development by involving themselves
in fun yet healthful activities, In this context, substance use and other high-risk
behaviors are portrayed as impediments to teens' achieving their own goals and

54


aspirations. An evaluation incorporating a quasi-experimental research design, with
Manitoba used as a comparison area, has produced encouraging results
(Thompson, 1988).

A variation of this general strategy is to label mem.bers of the target audience as
"winners" and to present the desired behavior as an integral part of that image
(e.g., Wallack & Barrows, 1982-1983). Labels provided by others can influence a
person's subsequent behavior (DeJong, l979), though their impact is far greater
when they are based ,on observations of past behavior (Kraut, 1973)`or are
otherwise personalized (e.g., Strenta & DeJong, 1981), conditions that cannot be
met easily, if at all, through mass communication channels (DeJong; 1988). Still,
this approach might be valuable in certain cases. Alternatively, Schwartz (1983)
suggests that mass communication campaigns can use "shame" to motivate
desired behaviors, essentially by confronting members of the target audience with
any discrepancy between their self-image and their actual behavior (see also, Ball-
Rokeach, Rokeach, 81 Grube, 1984).

Fear aooeals should be used onlv under limited circumstances. A continuing
controversy in the U.S. -concerns the use of fear appeals. Most experts have
concluded that fear campaigns are extremely difficult to execute and rarely
succeed (Atkin, 1981; Bandy 81 President, 1982; Boster & Mongeau, 1984; Job,
1988). Indeed, they argue that there is a real risk that fear appeals will backfire,
making the problem behavior even more resistant to change (Kleinot & Rogers,
1982; Rogers & Mewborn, 1976).

To work, the onset of fear should precede an explication of the simple, concrete
steps that people can take, and .fear onset should then reinforce the desired
behavior to confirm its effectiveness. This is harder to do than it sounds. If the
appeal is too mild, or if the threatened punishment seems too unlikely or remote in
time, people will not be motivated by it. If the threat is too strong, or if the
behavioral prescription being offered as an alternative is inadequate to alleviate the
induced fear, people might tune out the message, deny its validity, derogate the
source's credibility, or adopt a fatalistic attitude (Job, 1988; Solomon & DeJong,
1986).

The central difficulty is the impossibility of anticipating or modulating in advance
the level of fear that will be generated by a particular set of campaign materials or
to judge whether the fear reduction or other benefits resulting from the prescribed
action will be adequate to motivate behavior change (Bandura, 1986). Moreover,
the target audience's response to the fear appeal will also vary according their
perceptions of their own vulnerability. If the audience has low anxiety, a fear
appeal might serve to raise concern about the problem and motivate action. But if
the audience is already anxious about the problem, this type of appeal might
interfere with comprehension or a rational consideration of options (Bandy &

55


President, 1982). A complicating factor is that members of the same target
audience will vary on this dimension, again making the appropriate strength of the
fear message difficult to establish.

Despite these considerations, fear appeals continue to have strong intuitive appeal
and r?re frequently used by advertising professionals in health promotion
campaigns. One reason for this is that focus group participants usually rate strong
fear appeals as highly motivating and effective. But this is true even when
subsequent experimental studies show those appeals to be ineffective (Job, 1988).
The reason for their continuing allure is clear: In general, the threat of punishment
is re!ied upon to control behavior when its causes are insufficiently understood or
those causes are difficult to change (Bandura, 1986).

Compounding the confusion is the lack of definitional clarity about what
constitutes a "fear appeal." In their zeal to promote alternative approaches, some
experts extend their concerns about fear appeals to any message that focuses on
the negative consequences of certain behaviors. Further adding to the confusion,
of course, is that such campaigns do occasionally work (e.g., Advertising Council,
1991; Chu, 1966). As noted before, however, this is usually the case when the
audience has low awareness or anxiety about a problem (Flay & Sobel, 1983). In
1987, for example, an annual survey of high school seniors found a decline in
reported `cocaine use; the investigators attributed this decline to a heightened
awareness of cocaine's potential to kill even first-time users, due to the widely
publicized death of basketball star Len Bias and the mass media campaigns that
followed (Johnston et al., 1988)". On the .other hand, once people are already
aware of a problem, other means of influencing their behavior, such as modeling
appropriate skills and demonstrating the benefits of alternative behaviors; must be
found.

Cuse
A celebrity
is often used in product advertising as a means of drawing attention to the product
or to show that the it is "in fashion" (Graham & Hamdan, 1987). Many public
service campaigns have used this tactic as well. A celebrity's involvement will
typically include public service announcements and other educational materials,

lo

   It should be noted that when school-based programs were started in the 196Os, educators
believed that the key to effective prevention was to warn young people about the dangers of
substance abuse. It is `now recognized that programs which only emphasize consequences are of
limited value. First, investigators found evidence that, among some youth, such programs might
stimulate curiosity about drugs and a rise in experimentation (Bandy & President, 1982; Polich et
al., 19841. Second, such programs often resorted to exaggerated "scare tactics, n which
adolescents detected and discounted.

56


publicity events, promotional tours, and fund-raisers. If the spokesperson's
association with the message becomes strong enough, his or her presence alone
can evoke recall of the message (B&n, 1978).

Even so, there are risks to using this strategy. First, the message may be
overwhelmed by the celebrity's presence and ultimately forgotten (Atkin, 1979;
Ogilvy, 1985) Second, celebrities can lose their lustre; among adolescent fans in
particular, perceptions of entertainment and sports stars often change very quickly
(Graham & Hamdan, 1987). Third, celebrities.can suddenly become newsworthy
in ways that directly undermine the campaign or are otherwise inappropriate".

In selecting a celebrity spokesman, the watchword is caution. A celebrity should
be selected whose public image fits the underlying strategy of the campaign, not
just because he or she is available (Bandy & Pres.ident, 1982). Available data on
the celebrity's popularity among different demographic groups should be examined,
and formative research should be undertaken to test the target audience's
perceptions of the celebrity's trustworthiness'2,-credibility'3, and attractiveness
(McGuire, 1985). Most important, people who know the celebrity, and whose
judgment can be trusted, should be consulted for their advice whenever possible.

Camoaian olanners should seek oooortunities.to oromote a product or service
whdse use is consistent with the camoaian's health theme (Rehony et al., 1984).
Perhaps the classic example of this approach is the promotion of condoms as part
of family planning or AIDS prevention programs (Altman & Piotrow, 1984; DeJong,
1989; Sherris, Lewison, & Fox, 1982); The central benefit of this approach is that
it allows the direct application of commercial marketing and advertising strategies
(Black & Harvey, 1976).         '

Campaign planners can often develop a new product or service that can be
promoted as part of their program (Lefebvre & Flora, 1988). A good example of

   I1 In the case of substance abuse-prevention messages, there is also a fourth consideration:
Adolescents often view celebrity messages skeptically, because they suspect the celebrity was
paid to deliver the message or because they believe that many stars are substance users (Harvard
Business School, 1987).

   I2 Perceptions of trustworthiness will be undermined if the source of the message is seen as
lacking good will or having manipulative intent. As a result, young people, especially those who
are rebellious aQainst adult authority, may view any adult spokesperson with suspicion (Eauman et
al., 19881.

   " Because of the documented importance of credibility, it is tempting to rely on scientific
experts to provide health promotion messages. In some cases this will backfire. Schlegel (19771
cautions, for example, that adolescents who see themselves as relatively sophisticated about drugs
may view all others, including scientific "experts, n as less credible than their friends (see Smart &
Fejer, 19 721.

57


this strategy is provided by the Pawtucket Heart Health Program, which is focused
on cardiovascular risk reduction (Lasater, L8febVr8, & Carleton, 1988; Lefebvre,
Peterson, McGraw et al., 1986). Commodities publicized through the mass media
include a cookbook and a "Four Heart" restaurant program through which eateries
offer low-fat, low-salt menu options. Another possibility is to promote coupon
books that offer free or discounted products and services whose use is consistent
with the public health campaign (Bollier, 1989). It should also be recalled that
community-based intervention programs (e.g.., self-help groups, screening/
counseling events, smoking cessation programs) can also be promoted (Lefebvre et
al., 1988).

A related tactic is contest sponsorship (Graham & Hamdan, 1987). For example,
as part of their radio-based anti-smoking campaign, Bauman et al. (1988) included
a special promotion, the "I Won't Smoke" Sweepstakes, as a means of
encouraging peer involvement and interpersonal communication about the
campaign. Entrants sent in an entry form with their name and address; at their
option, they could sign a pledge not to smoke. Entrants could enroll friends in the
drawing in return for cash awards. During the drawing itself, the prize doubled in
value to $2,000 if the winner's entry form indicated that he or she was a non-
smoker.

Another possibility is for a health promotion campaign to establish links with
commercially available products or services. In this strategy, called "cause-related"
marketing, advertising for the product or service also carries a public health
message (Moses, 1988). While it is ideal'for there to be a direct link between the
commodity and the health message (e.g., life insurance with drunk-driving
prevention, physical fitness equipment with smoking prevention), such a
connection is not strictly necessary. For example, the Frito-Lay Corporation, one
of the country's largest makers of potato chips and other snack foods, agreed to
donate money to the non-profit Just Say-No Foundation with 8aCh customer
purchase, a promotion announced in advertising and on its packaging with a brief
anti-drug message.

Public health advocates can also emulate a strategy used by tobacco and b88r
companies -- the official sponsorship of sporting events, concerts, and other
cultural or youth-oriented activities. For example, Canada's "Break Fr88" campaign
became the official sponsor for the Arctic Winter Games and the Student Regional
Games and distributed "Break Free" toques and headbands as part of those events
(Liepold, 1986). Because such events are often financially dependent on the fees
paid by their sponsors, use of this strategy by public health advocates will be
limited unless they can negotiate with a product or services company to co-
sponsor the event.

58


LESSON 4: STIMULATE INTERPERSONAL COMMUNICATION

Stimulating interpersonal channels. of communication can enhance a mass
communication campaign in several ways (Flay & Burton, 1990; Rogers, 1983).
First, communication through a social network ensures that the message is spread
beyond the original audience. If discussion of a topic is frequent, then any new
information provided through the mass media is more likely to be attended to and
remembered. Second, interpersonal communication is generally very effective for
delivering complex information and persuasive messages. Third, interpersonal
communication can increase the likelihood of new behaviors being tried, adopted,
and maintained (Flay, 1986).

In using a mass communication campaign to stimulate this process, `it is important
to target messages to a community's opinion leaders or "trend-setters" (Black &
Farley, 1977; Rogers & Shoemaker, 1971). Katz and Lazarsfeld (1955) describe
this as a two-step flow of communication, from the sources of mass com-
munications to informal opinion leaders, and then from these leaders to others.

The importance of this approach is underscored by studies on the impact of mass
media on youth at high-risk for drug abuse. In their review of the literature, Bandy
and President (1982) noted that the more often students use drugs, the less likely
th8y.are to believe information conveyed through the mass media. One study they
reviewed found that non-users rely primarily on the mass media to learn about
drugs, whereas drug users rely primarily on information from their friends and their
own experience.

LESSON 5: USE A VARIETY OF MASS COMMUNICATION CHANNELS

A variety of communication channels should be used to provide a clear and
consistent message, thereby reinforcing one another (Atkin & Arkin, 1990;
DeJong, 1989). Particular channels should be selected according to the target
audience's media preferences and the objectives of the campaign (Alcalay &
Taplin, 1989).

Television is an attractive medium because of its large and diverse audience, which
creates the capacity for bringing immediate attention to an important issue or
concern. On the other hand, television programming is passively consumed, which
limits the type of information that can be conveyed (Atkin, 1979). In general,
television is excellent for providing short, uncomplicated messages, evoking
emotional reactions, establishing evidence of new social norms, and modeling
behaviors that can be easily taught.

59


Because each television show's audience varies in its demographic profile,
narrowly defined audiences can be targeted., Current trends toward the
"fractionalization" of television, caused primarily by stronger independent stations,
cable channels, and syndication of first-run shows, also provide opportunities for
targeting specific audiences such as young people (Atkin &. Arkin, 1990; Graham &
Harridan, 1987). On the other hand, use of these media greatly increases the
logistical demands on campaign pIann-ers.

The potential of radio, especially for reaching adolescent audiences, has not been
fully appreciated by public health campaign organizers. Radio shares many of
television's advantages, but obviously does not carry the impact of the Visual
medium. On the other hand, radio calls upon the imagination more than television,
which may be an advantage for certain types of messages. Radio is also much
less expensive, which allows greater repetition of the messages. Moreover, the
variety of stations on the air facilitates the targeting of messages to very narrowly
defined groups, which can be identified through marketing reports used by the
industry. For a national campaign, a major disadvantage of radio is the large
number of independently owned stations that need to be approached, even with a
focus on major markets. For certain audiences, such as adolescents, nationally
syndicated programs with celebrity "disk jockeys" might provide a more feasible
venue. .

The print media are less able to command attention than the broadcast media, but
they can be used to reach finely differentiated target groups. And because print
advertisements, brochures, and feature articles can be reread and invite reflection,
they are better suited for presenting rational arguments or detailed information
(McGuire, 1989).

A media plan should rely on audience rating systems and formative research to
identify which specific stations, programs, or print media are the best vehicles for
reaching the target audience at the lowest cost per contact. Obviously, the right
"media mix" varies according to the particular audience being addressed.

An issue that cannot be satisfactorily answered is what does, in fact, constitute
"adequate" exposure. Public communication specialists have not extensively
studied this issue in health campaigns because, historically, they have relied on
public'service time or have had small budgets to work with. Only two broad
generalizations can be offered as guidance. First, repetition helps draw attention
to the message (Solomon, 1983), facilitates learning, and increases liking for the
message. However, repetition can be Count8rprOdUCtiV8 if it iS excessive and turns

the message into a cloying annoyance (Bornstein, 1989). Second, airing ads in
high-frequency bursts (or "flights") is superior to having the same number of ads

drawn out over a longer period (Atkin, 1979).

60


Precise guidance is lacking, however,.since the right amount of exposure depends
on too many variables -- the nature of the target group, the precise objective, the
complexity and `entertainment value of the message, the nature and extent of
competing messages, and so forth,- Because there is no valid way to measure
advertising effectiveness, even commercial advertising agencies do not always
know which media should be selected or how frequently they should be used to
achieve a particular advertising objective (Sissors & Surmanek, 1987).. Audience
tracking studies can be executed as the campaign is launched to gauge audience
exposure to thecampaign message. In general, such stud/es will not be a priority
for health campaigns operating on limited budgets. But if severe doubts about the
media plan develop, focus groups could be used in lieu of a formal audience survey
to see how the plan might be modified (Solomon, 1989).

While public health advocates typically turn to the traditional media -A television,
radio, and print -- for mass communications campaigns, there is, in fact, a diversity
of media venues to be considered, including billboards, transit cards, newsletters,
videotapes, and booklets. New advertising venues will continue to emerge. The
Whittle Corporation developed several new venues during the 198Os, including
commissioned books distributed free to the nation's opinion leaders; message
boards for school gyms; and, with much controversy, daily news programming
beamed to high schools via satellite. Other new developments include screen
advertising in movie theaters (Atkin & Arkin, 1990) and the use of "dial-in" phone
messages, which are promoted in billboard and print advertisements (OSAP,
1990).

SSO
LE N

Many public service campaigns that use television and radio simply fail to reach
their intended audience (Flay, 1986; Flay & Sobel, 1983; Warner, 1987). Public
service advertising developed by the Advertising Council and others is typically
used by broadcast stations as "filler" and is often aired at off-hours with small
audiences. Of particular concern, television public service ads are infrequently
aired during prime time when preteens and adolescents are viewing (Hanneman,
McEwan, & Coyne, 1973). This has been a long-standing problem, but it has been
worsened by the loosening of broadcasters' public service obligations during the
1980s (Atkin & Arkin, 1 990)14.

l4

   In response to this concern, there were calls at the December 1988 Surgeon General's
Workshop on Drunk Driving for commensurate time for "responsible drinking" messages as a coun-
terweight to beer and wine advertisements (OSG, 19891. A similar lift was given to the anti-
smoking movement in the late 1960s when the Federal Communications Commission (FCC) applied
the fairness doctrine to tobacco advertising on radio and television and required stations to provide
anti-smoking groups with free time to air their views (Flay, 19871.  When tobacco advertising was

61


Because campaign organizers want to ensure audience exposure to their campaign,
the competition for donated time is fierce. A public service campaign's success in
winning donated time depends in part on the issue involved and its perceived
importance to media "gatekeepers" who establish public service priorities for each
broadcast station. The Centers for Disease Controi's AlDS campaign, for example,
received more than $28 million in donated air time from October 1987. through
February 1989. A 1989 survey showed that 22 percent of U.S. adults recalled
seeing or hearing during the past month a.public service announcement with the
slogan "America Responds to AIDS" (Dawson, .1989). It has also been observed
that local gatekeepers will respond more favorably to public service spots that are
"creative" and that have local program or agency taglines or allow a local station to
add their own (Freimuth & Van Nevel, 1981; Hammond et al., 1987; Goodstadt &
Kronitz, 1977).

Public health advocates have responded to the intensified competition for public
service time by trying to build relationships with the media gatekeepers (Maccoby
& Alexander, 1980). For example, in implementing the Asbestos Awareness
Campaign, the project staff made special efforts to persuade the gatekeepers in 16
target communities to air the public service announcements, including personal
contact, press kits, and fact sheets (Freimuth & Van Nevel, 1981). At the national
leveli the Harvard Alcohol Project was able to secure the cooperation of the three
major networks to produce and air prime-time advertising to promote the
"designated driver" concept (DeJong & Winsten, 1990b).

Some recent campaigns have purchased air time, using it exclusively or in
combination with public service time (Bauman et al., 1988; Wallack & Barrows,
1982-83; Worden et al., 1988). This strategy is controversial within the public
health community. While some health advocates focus on the need to get their
messages on the air in the best time slots, others express concern that paying for
media time will cause the broadcasters to demand payment for other public service
advertising (Bloom & Novelli, 1981; Dessart, 1990).

Clearly, however, the restricted financial resources of nearly all public health
agencies limit the extent to which this approach can be used, especially with
television. Even the U.S. Department of Health and Human Services, which
sponsors numerous health promotion campaigns through the Public Health Service,
generally does not have funding for buying airtime (Warner, 1987). Some local
stations have recently received public.ity for selling air time for public service
campaigns at greatly reduced rates (Meyers, 1989), but the cost still remains too

banned from the broadcast media in 1971, the mandate for an?i-smoking advertisements ended as
well (Wallack, 198 II. Similar action by the FCC on alcohol advertising now appears to be
extremely unlikely (Colford, 1989).

62


high for most public health programs. Radio, billboards, and transit cards represent
more realistic alternatives for paid public health advertising (Arkin, Denniston, &
Romano, 1990).

Corporate sponsorship might provide a partial answer to this funding problem. For
example, anti-drug spots developed by the Media-Advertising Partnership for a
Drug-Free America are frequently aired through a combination of donated time,
both nationally and locally, and corporate sponsorship.

Another possibility is to work directly with broadcasters to develop a new
campaign. In local communities, a broadcast station might want to develop an
exclusive, station-sponsored campaign, one that focuses the station's energy on a
single problem of public concern (Wishnow, 1983). Occasionally, a major network
will undertake such a campaign as well, as evidenced by CBS's involvement in the
"Be Smart. Don't Start!" campaign (Atkin, 1989a).

In station-sponsored campaigns, newscasts, documentaries, talk shows, public
service announcements, and editorials work in sync to inform and motivate the
public and to offer solutions, either in the form of individual/collective action or a
change in public policy. For some campaigns, the station or network will work in
partnership with business, government, and voluntary organizations to create a
direct intervention, such as print materials, smoking cessation kits (Sallis, Flora,
Fortmann et al., 1985), and special events. Station managers find that such
campaigns serve the station's needs by bringing community recognition and
acclaim, which in turn increase ratings for local programs. Corporate sponsors can
be asked to provide funding in return for on-air mention of their involvement.

For certain issues, such campaigns provide a unique opportunity for public health
advocates to achieve their health promotion objectives. For example, in
Washington, D.C., WRC-TV sponsored a prenatal health care campaign, "Beautiful
Babies: Right from the Start," which featured a series of news reports,
documentaries, public service announcements, and editorials. Most notably, local
obstetricians participated in distributing coupon books that offered a variety of
products and services at reduced COSti many of them directly related to prenatal
and child care. As a result of the campaign, more than 60,000 coupon books
w.ere distributed. After 18 months, prenatal visits to public health clinics had
increased 22 percent and infant deaths had declined 6 percent (Bollier, 1989),
although campaign organizers acknowledge that the campaign's contribution to
these changes cannot be proven.

63


ESON 7: INCORPORATE THE STRATEGIC US.E OF TELEVISION
ENTERTAINMENT AND NEWS PROGRAMMING

Mass communication campaigns have traditionally involved the use of public
service advertising, including radio and television spots, print ads, billboards,
posters, and printed literature. More recently, .several public health advocates have
argued that campaign planners should think more expansively about how the mass
media can be used to reach the public, moving beyond the realm of public service
announcements to include news and entertainment programming as part of a
single, unified campaign (Atkin & Arkin, 1990; Men.delsohn, 1973). The failure to
use such programming as vehicles for health education represents a missed
opportunity. Beyond that, however, this failure frequently results in the public
being exposed to conflicting messages that undermine the campaign (Atkin, 1979).
The advantages and limitations of using entertainment and news programming are
reviewed here.

Television Entertainment Proaramminq

As the "most effective purveyor of language, image, and narrative in American
culture,1' (Marc, 1984), television is an important influence in shaping cultural
norms, public opinion, and behavior. Television programming not only mirrors
social reality but helps shape it, by communicating what constitutes popular
opinion and by influencing people's perceptions of the roles and behaviors that are
appropriate to members of a culture (Roberts & Maccoby, 1985; Signorielli, 1990).
For many people, and for children and adolescents especially, television.pro-
gramming is a key source of information about how the world works and how one
should behave in that world (Atkin, 1989b; Roberts, 1989).

Unfortunately, because plots and script lines are selected primarily for their
entertainment and artistic value, a great deal of the health-related information that
is presented on television is inaccurate, misleading, or antithetical to good health
(Atkin, 1979; NIMH, 1982). For example, recent data show that, while drug and
tobacco use is infrequently shown on television, the portrayal of alcohol use
averages several "acts" per hour, far more often than several other more popular
beverages, including water (Signorielli, 1990). In combination with beer and wine
advertising, this barrage overwhelms the message from infrequently aired public
service advertising that urges preteens and young adolescents to abstain (Wallack,
1990a).

As a result, public health advocates have pursued two strategies for working with
the entertainment industry to promote lifestyle changes and healthier behavior: 1)
consulting with producers, writers, and directors concerning the content of regular
entertainment programming (Breed & DeFoe, 1982; Montgomery, 1989, 19901,

64


and 2) collaborating in the production of special programming around health
promotion themes (Mendelsohn, 1973).

Consultation with the Entertainment Industry. While lobbying the U.S.
entertainment industry has been common over the last 20 years, this strategy has
been given new impetus by the highly publicized Harvard Alcohol Project, a
university-based campaign designed to prevent drunk driving through promotion of
the designated driver concept via entertainment programming and network-
sponsored public service announcements (DeJong & Winsten, 1990a, 1990b;
Winsten, 1990). After three television seasons, messages consistent with the
project's agenda have appeared in over 100 shows, including both dialogue and
the display of a designated driver poster developed by the project. While several
would-be imitators have been inspired by the Harvard project, several facts about it
point to its limitations as a model for other health promotion efforts.

The director of the Harvard project was afforded access to top executives of major
Hollywood production studios through the direct intervention of Dr. Frank Stanton,
former president of CBS and, at the time, a member of Harvard's Board of
Overseers. As Montgomery (1989) notes, producers in general have sought to
keep outside lobbyists at arms length. As a result, "Hollywood lobbyists" usually
devoted a number of years to bui1ding.a small network of supporters in the
industry, largely through hosting seminars, panel discussions, or lectures. In more
recent years, however, the most successful lobbying efforts have been set up by
industry insiders or greatly assisted by them and have used those personal
connections as a basis for private discussions with individual producers
(Stevenson, 1990). Obviously, few public health advocates can rely on that kind
of insider support and will find access extremely difficult to achieve.

The designated driver concept itself has several important features that distinguish
it from other public health/topics and make it a more attractive theme for
producers and network executives. First, Mothers Against Drunk Driving (MADD)
and other advocacy groups had already aroused public concern about the drunk
driving problem. Because shifts in social norms regarding the acceptability of
driving after drinking had already begun, the role of the television industry could be
accurately described as one of reinforcing an emerging trend rather than
"engineering" a new one. This fact helped deflect criticism that a systematic effort
to influence public opinion and behavior through entertainment programming was
an abuse of concentrated media power.

Second, the designated driver message could be easily incorporated into
programming on a routine basis, especially since scenes involving alcohol use are a
staple of television programming (Breed & DeFoe, 1981; Gerbner, 1990). The
introduction of other public health topics -- drug abuse, AIDS, teenage pregnancy,
organ transplantation -- represents a more radical departure from standard


programming fare. Many of these health issues lend themselves to dramatic
treatment, but only as a special episode or program. Such programs might be
worthwhile for drawing attention to an issue, but they should not be seen a tool
for bringing about substantial. change over the long-term.

Third, the designated driver message, by emphasizing individual responsibility to
prevent the drunk driving, meets the television industry's need to do something
positive while not. alienating the alcohol industry, on whom broadcasters depend
for a significant portion of their advertising revenue ("Advertiser Report," 1990;
Wallack, 1990a). Indeed, many public health experts worry that the Harvard
campaign has taken the pressure off the television networks to modify alcohol
advertising practices that appeal to minors. In general, the television industry is
more likely to focus on politically non-controversial subjects or solutions -- an
important limitation (Montgomery, 1990).

A restriction commonly shared by any effort to use television entertainment
programming to communicate a public health message is the inability of outside
organizations to control programming content. Fbr example, guidelines developed
by the Harvard Alcohol Project emphasize that promotions of the designated driver
concept should not imply that it is socially acceptable or even expected that the
driver's companions will drink to excess. (DeJong & Winsten, 1990b). Even so, a
few television episodes that mentioned the designated driver concept did so in a
way that directly violated this important guideline.

Develooment of Soecial Proaramming. A second option for working with the
entertainment industry is to collaborate'in developing special health-related
programming. While programs aired nationally are an occasional possibility
(Montgomery, 1989), true col!aboration is more likely to occur at the local level.
As one example, the ongoing Stanford Five Community Project initiated its
education program with the airing of a locally produced hour-long documentary
("Heart Health Test") that provided general information on cardiovascular risk
reduction (Farquhar et al., 1985a). Development of special programming is par-
ticularly useful for audiences that are not well-served by the major networks, such
as Latinos (McAlister et al., 1987; Mendelsohn, 1973).

News Proaramminq

Public health advocates have long recognized the importance of obtaining favorable
news coverage to support the objectives of a health campaign designed to promote
behavior change. Especially valued is the greater visibility and the "third-party
endorsement" of campaign messages that news stories can bring (Alcalay &
Taplin, 1989; Goldman, 1984).

66


In recent years, this strategy has been given renewed prominence by public health
advocates wanting to promote changes in public policy. What they call the "media
advocacy" approach features a variety of public relations strategies to stimulate
media coverage that will serve to reframe ho.w the general public and opinion
leaders conceptualize a public health problem, and to promote a consideration of
public policy options (NCI, 1988; Wallack, 1990b). This strategy is especially
important when the policy changes being urged threaten commercial advertisers
who financially support the media.

A number of guidelines should be followed in Seeking news coverage. First, it is
important to have clear objectives for what 4s to be achieved through news
coverage. Publicity for its own sake is an insufficient reason. In 1.988, for
example, the Harvard Alcohol Project issued a press release announcing that the
major television production studios had agreed to introduce the designated driver
concept in entertainment programs. The resulting page-one story in The New York
Times (Rothenberg, 1988) served several purpo$es: giving additional visibility to
the issue of alcohol-impaired driving, directly promoting the designated driver
concept, and establishing the institutional identity and stature of the Harvard
Alcohol Project itself. But the principal rationale for issuing the press release was
to solidify the production studios' commitments through public exposure, an ob-
jective which was achieved.

There are other considerations to using this strategy successfully. Foremost, the
press announcement must present information that has real news value15. Public
health advocates too often assume that all of their activities, because they are
important, are also "newsworthy" (DeJong & Winsten, 1990a). Instead, public
health advocates must understand how news institutions define the "news," the
incentives to which reporters and `editors respond, and their preferences for how
news information is packaged and presented (Atkin & Arkin, 1990; Bantz,
McCorkie, & Baade, 1980; Stuyck, 1990).

Generally, for health-related information to be defined as news it must be more
than new. The story must also -have aspects that make it especially attention-
grabbing. Meyer (1990) lists several factors that contribute to a story's appeal:
timeliness, geographical proximity, prominence of the people involved, human
(emotional) interest, controversy, novelty, and potential impact on people's lives.
Stuyck (1990) adds another important factor: the story's potential for creating
more informed health consumers. Thus, campaign planners must think creatively
about how they can get reporters interested in their message or story.

   As explained by Alcalay and Taplin (79891, press releases can announce an upcoming event
and invite the press to cover it; issue a public statement on a news development or issue; or
provide background information that gives perspective on late-breaking news.

67


There are several standard ways of achieving this. One is to issue a survey report
or other research findings that give new insight into a health problem and its
solutions (Wallack, 1990b). For local news coverage, national data have to be
given a local perspective (NCI, 1988). Another is to,announce upcoming program
events. Indeed, when deciding what steps to take next in a program or-campaign,
planners should take into consideration the public relations value of the various
alternatives (Goldman, 1984). The American Cancer Society., for example, found a
powerful idea in its Great American Smokeout, which continues to generate news
coverage year after year (Flay, 1987). Yet another means of generating news
coverage is to create "media events" such as-receptions, speeches, policy debates,
awards ceremonies, or special fund-raising events (Player, 1986; Wallack, 1990b;
Warner, 1987). Staging events for television is es'pecially important, as stories
with a strong visual element are given higher priority (Klaidman, 1990).

Another consideration is that the press announcement must be timed properly so
that it does not compete with other stories (DeJong & Winsten, 1990a). While
campaign planners cannot fully anticipate the flow of world and national events,
certain days are typically -"slower" news days than others. Knowing this, staff for
the Harvard Alcohol Project issued its aforementioned press release in late August,
when Congress is typically in recess and many news-making organizations are
enjoying a vacation lull.

To maximize its serious consideration, a press release is best addressed to
reporters with whom the campaign planners have established credibility over time,
perhaps by serving as a background source for other stories. On occasion,
relationships with the press can be nurtured by offering periodic exclusives, al-
though it must be cautioned that overusing this strategy might eventually alienate
those relationships. Relationships with print reporters are especially critical, since
local broadcast media often rely on area newspapers when determining what
stories to cover.

A relatively new means of generating press coverage is the distribution of so-called
"video news releases," which are professionally produced program segments
suitable for use on evening news programs, TV magazine shows, or
documentaries. Unfortunately, production expenses do not make this a reasonable
alternative for many public health agencies (Kleiner, 1989; Davis, 1988). On the
other hand, "audio news releases" for use on radio might be an effective
alternative, especially if only a limited number of radio stations are targeted.

In considering the advantages and disadvantages of trying to generate supportive
news coverage, public health advocates must recognize that the news media are
an imperfect instrument for communicating health messages (Stuyck, 1990). First,
because broadcast news programming thrives on political controversy, it provides
limited opportunities for providing clear and conSistent messages health messages.

68




This fact was amply demonstrated by, press coverage of New York City's
announcement of its 1987 AIDS campaign aimed at promoting condom use among
heterosexuals. Several television reports aired the campaign's spots as part of
their news coverage, thereby furthering the objectives of the campaign. On the
other hand, these reports also featured interviews with a clerical spokesman for
the Catholic Archdiocese of New York, who objected strongly to the "graphic"
portrayals in the campaign's television spots and their implicit endorsement of
"promiscuity." In further support of his attack on the campaign, this spokesman
asserted that the condom's high failure rate made it a poor choice in AIDS
prevention efforts. This claim, which failed to account for consumers' sometimes
improper or inconsistent use of condoms (Sherris et al., 1982), was not challenged
in any of the news reports.

Second, the competitive pressures on news reporters frequently resulted in
incomplete investigative reporting and the sensationalized treatment of tentative
scientific findings, both of which contributed to public confusion and panic. In one
especially important case, a study reported in a- 1983 issue of the Journal of the
American Medical Asociation hinted that "routine contact" among household
members might be sufficient for AIDS transmission to occur. Although subsequent
studies showed that the disease was not spread through casual transmission, and
although these studies -were covered in turn by the news media, the initial stories
nevertheless had a residual effect (Milavsky, 1988). This residual effect was
sustained in the mid-1980s by continuing coverage of public hysteria about AIDS,
as manifested in the boycotts of schools attended by AIDS victims, public arrests
of AIDS patients by police wearing protective gloves, and so forth.

Recognizing the inherent limitations in' using news coverage to carry forth a health
promotion message, some health advocates have begun to create their own
publications, a strategy also used.in the commercial sector. The Philip Morris
tobacco company, for example, publishes Philio Morris, a "lifestyle" magazine that
is distributed free to self-identified smokers. At one level, the magazine serves as
a company-controlled venue for its standard print advertising, but the magazine
also reinforces the reader's self-definition as a smoker by publishing articles,
features, and consumer advertising that associate this identity with an active,
upscale lifestyle. The magazines also presents a political agenda that portrays
smokers as an "oppressed" minority; direct mail solicitations sent to subscribers
encourage them to put pressure on State and local politicians to defeat anti-
smoking ordinances and legislation (Sylvester, 1989). A public health counterpart
to Philio Morris is 2001, a slick and highly popular "lifestyle" magazine distributed
free to teens by the Alberta Alcohol and Drug Abuse Commission (AADAC) in
Canada as part of its general campaign to promote healthy lifestyles that exclude
alcohol, tobacco, and drug use ("Unique Magazine," 1988-89).

69


SYNC WITH COMMUNITY-BASED PROGRAMS

Both the Stanford and North Coast studies suggest that mass communications
alone, when properly designed and executed, can produce behavior change. Even
so, both sets of investigators later turned to demonstration `projects that used
mass media in tandem with community-based programs -- the Stanford Five
Community Study (Farquhar et al., 1985a), and the "Quit for Life" campaign in
Sydney, Australia (Dwyer, Pierce, Hannam, & Burke, 1986; Pierce, et al., 1986).

It has long been recognized that a media-driven health promotion campai'gn is more
likely to succeed in changing behavior when other program components are in
place, a condition that Lazarsfeld and Merton (1971) have called          i
"supplementation." Accordingly, many researchers emphasize that media
approaches should work in sync with community-based program components that
involve long-term, face-to-face education and community reorganization (Alcalay &
Taplin, 1989; Flay, 1986; Solomon, 1983; Wallack,. 1980). Media can support
these local programs in several ways. In turn, these programs can reinforce a
media effort by giving shape to an audience that is primed to attend to the media
campaign (Flay, Hansen, Johnson et al., 1988; LaRose, 1989) and by helping link
the media campaign to local concerns and issues (O'Keefe & Mendelsohn, 1984).

Those who favor using mass communications with community-based programs
note that many problems, such as cancer prevention and cardiovascular risk
reduction, involve altering long-standing habits and complex behavior patterns.
The Stanford study described earlier showed that a media campaign alone, while
producing an reduction in overall risk,. had a demonstrable impact only on fat
consumption but no long-lasting effect on smoking or leisure-time physical activity.
The same issue arises with substance abuse. While some have endorsed using
mass communication campaigns to reach. high-risk youth whose profile makes their
abuse of drugs more likely (Worden et al., 1988; Wallack, 1986), most
investigators do not endorse this approach, since a person's psychological make-up
is a product of life experience and is extremely difficult to change even with direct,
intensive intervention. As a result, many programs try to induce members of
target audience to participate in programs with face-to-face intervention or to use
specially crafted self-help programs (Lefebvre et al., 1988).

In other cases, standard mass communications approaches will be thwarted
because of political controversy. For example, to encourage condom use for AIDS
prevention, there is a need for motivating messages that focus on how condom
use can bring sexual pleasure; failure to do so means ignoring the principal
psychological barrier that impedes condom use by U.S. consumers (Solomon &
DeJong, 1986, 1989). Reacting to widespread public sentiment against
contraception, and wanting to avoid the political or economic pressures that would

70


ensue, television broadcasters have established standards that make it impossible
to use paid or public service advertising for'this purpose (DeJong & Winsten,
1991)`? With these restrictions in. view, health advocates would better focus on
community-based education programs that stress condom promotion or local social
marketing programs (DeJong, 1989; DeJong & Winsten, 1991). Local broadcast
media could support these programs by publicizing them and directing people to
information hotlines.

Mass media campaigns can be used in several ways to enhance the effectiveness
of community-based programs (Flay, 1986.; Wallack, 1980, 1981). First, mass
communications can be used to enroll new program participants, recruit volunteers,
or win financial support. Second, mass communications can announce the
availability of self-help materials, ongoing program activities, and special events
(Flay, 1986; Mendelsohn, 1973). Third, mass communications can reinforce the
instruction provided by school-, worksite-, or other community-based programs,
including skill development, self-monitoring, and other maintenance activities (Flay,
1986). In this case, the mass media materials should also work on their own,
since many of those who are most in need of prevention education attend school
irregularly (Worden et al., 1988) or are not reached by community programs
working under limited budgets. At a minimum, messages from the mass com-
munications campaign and these other components should be consistent; ideally,
their activities should be guided by an integrated campaign strategy.

While having mass media and community-based programs work together is, in
principal, a sound idea, the issue of cost-effectiveness and other practical concerns
should not be forgotten. Hornik (1989, p. 312) expresses strong reservations
about the widely held view that interpersonal channels should predominate: ". . .
from the point of view of cost, of:feasibility, and of sustainability, organizing a
face-to-face network is rarely possible." In his own study of a Swaziland program
to promote oral rehydration therapy (ORT), Hornik found that a series of radio
programs, given their broad reach among the target audience, were a much more
effective strategy than using interpersonal channels such as clinic-based health
care professionals, health extension agents, and trained community volunteers.

Warner (1987) makes a similar point, noting that a mass communication campaign
needs only a small success rate to produce change among a far group of people
than can be reached through conventional community-based programs that have
much higher success rates. Cost-effectiveness is therefore a more suitable

"

   Television entertainment programming offers greater leeway for introducing motivating
messages, but they cannot be incorporated into programming often enough to make a critical
difference and must be "balanced" by the inclusion of opposing viewpoints (Montgomery, 1990).
News programming, because of its focus on political controversy, is similarly limiting.

71


criterion
projects
1984).

for judging health promotion strategies. Even so, major health promotion
have devoted very little attention to assessing per unit costs (McGuire,

LESSON 9: USE MASS COMMUNICATIONS TO FOCUS ATTENTION ON THE
SOCIAL, ECONOMIC, AND CULTURAL FACTORS THAT IMPACT HEALTH-
RELATED BEHAVIOR

So far we have examined campaigns from the standpoint of trying to effect change
in the knowledge, attitudes, and behavior of individuals. An increasing number of
public health experts have argued that a number of health problems, especially
those resulting from lifestyle behaviors, can only he understood and dealt with by
focusing on the social, economic, and cultural context in which they occur (Atkin
& Arkin, 1990; Alcalay, 1983; Gerbner, 1990; McLeroy et al., 1988; Wallack,
1984, 1986, 1990a).

In their view, it is inappropriate and self-defeating for prevention campaigns to put
the onus on the individual for changing his behavior while failing to change the
business and legal environment in which that behavior occurs. Ultimately, changes
in institutional structures, public policy, or law can lead to greater resources being
allocated to a problem and to environmental and social changes that will support
and sustain the efforts of individuals to alter their behavior. In the case of
adolescent substance abuse, for example, there is a need for increased public
debate on issues related to the cost, availability, and promotion of alcohol and
tobacco (Mosher & Jernigan, 1989), and on policy changes that might discourage
substance use, such as stricter enforcement of laws prohibiting alcohol sales to
minors, the elimination of cigarette vending machines, and increased excise taxes
to provide funding for school- or community-based programs that provide direct,
high-intensity interventions for high-risk youth (Tobacco Education Oversight
Committee, 199 1).

In this context, the agenda-setting function of mass communications can be more
broadly conceived to include targeted messages to official policymakers (Wallack,
1986) or community opinion leaders (Black & Farley, 1977; Rogers & Shoemaker,
1971). Directing awareness campaigns to business and civic leaders, what
Schwartz (1973, 1983) calls "narrowcasting," can be an especially powerful way
of moving a problem to the top of a community or national agenda. As one
example, the first phase of a Swedish campaign to promote condoms was to sell
national opinion leaders on the importance of combatting venereal disease and to
improve the condom's image as an acceptable, even respectable contraceptive
(Ajax, 1974). Having established the need for concerted action, campaign
organizers were able to move ahead with a marketing effort that included

72


introducing new condom brands and specialty retail outlets17. Importantly,
business and civic leaders can be reached through radio and newspapers whose
lower advertising rates make a pa/d campaign feasible (Alcalay & Taplin, 1989;
NCI, 1988).

An important role for social marketing approaches is to promote citizen
involvement in community-based efforts to promote social, institutional,
environmental change. As one example, parents of preteens and adolescents can
recruited to participate in local parents' groups that sponsor alcohol- and drug-free
parties, cooperate with police in monitoring local conditions, and advocate changes
in public policy, such as the development and enforcement of stricter school
policies, more conscientious enforcement of existing laws on the sale and
distribution of alcohol to minors, and new legislation setting a legal limit of .OO
percent BAC for underage drivers or imposing dr,iver's license sanctions for the
purchase or possession of alcohol by minors.

In the end, once legislation or other policy chan.ges have been implemented, mass
communications provide a superb vehicle for publicizing them -- not just through
news coverage, but also through special advertising and promotions. In drunk
driving prevention, for example, widespread publicity of "sobriety checkpoints" and
other law enforcement measures has proven to be essential. In the absence of
such. publicity, such efforts have no discernible impact (Ross, 1988).

LESSON 10: INCLUDE PROCESS AND OUTCOME EVALUATIONS AS PART OF
THE CAMPAIGN

Early program planning should incorporate both process and outcome evaluation
activities to monitor progress and demonstrate project impact (Lau, Kane, Berry, et
al., 1980; Wallack, 1984). Unfortunately, the common failure to evaluate mass
communication campaigns has impeded progress in understanding what works and
what does not (Arkin et al., 199.0).

The behaviorist tradition of most campaign evaluations (Gitlin, 1978; Wallack,
1984) has failed to capture the diversity and complexity of mass media effects,

l7

   On occasion, a mass communications campaign will need to focus on convincing opinion
leaders and the public at large that a particular institution should take responsibility for addressing a
health problem. A substance abuse prevention campaign by the Alberta Alcohol and Drug Abuse
Commission (AADAC) began in this way. Because the AADAC had been associated in the public
mind with efforts to control and treat derelict alcoholics, the first step of the campaign was to use
advertising to position the AADAC as an innovative, forward-looking agency that was concerned
about youth and alcohol while not being "neo-Prohibitionist" Gkirrow, 19871.

73


many of which, as noted before,. set the stage for or otherwise promote behavior
change in the long term., Evaluations should be designed to capture these effects.
As one example, if a campaign's objective is to move an issue to the top of the
public agenda, the project's success can be established by tracking the number of
news stories before and during the campaign, the, number of legislative proposals
submitted and passed, the number and size of grassrobts advocacy groups, and so
forth (Roberts & Maccoby, 1985; Wallack, 1980).

In most cases, evaluators will rely on the collection of survey data. Researchers
often express concern about the honesty of self-reports, especially when they
concern socially undesirable or criminal behavior such as illicit drug use. Even with
respondents remaining anonymous, however; their answers to survey questions are
known to be influenced by a desire to give socially acceptable answers. Moreover,
the data's validity might change over time, perhaps in response to the very
campaign being investigated, if admission to the behavior in question becomes
more or less socially acceptable (DeJong & Winsten, 1991; Wallack, 1986).

These concerns are important, but certain prec&tions can be taken to minimize
them. Respondents are likely to provide valid and reliable information if they are
reassured that their answers will be kept in confidence (Rouse, Kozel, & Richards,
1985). In addition, the survey instrument can be carefully constructed to build in
reliability cross-checks. Also, questions that measure respondents' propensity to
give socially desirable answers, such as the Marlowe-Crowne scale (Crowne &
Marlowe, 1964) can be included to factor out that source of bias.

To complement these survey data, program evaluations can also cqllect
community-level data on product sales or other objective indicators that would be
expected to change in response to the campaign. Doing so only makes sense, of
course, when the campaign is of sufficient intensity and duration to reach a sizable
proportion of the community.

Local or Reaional Camoaians

To test the impact of a long-term local or regional campaign, the research design of
choice is a "quasi-experimental" design in which the knowledge, attitudes,
behaviors, and health outcomes of people in experimental communities exposed to
the campaign are compared to those in similar control communities (Farquhar et
al., 1984; Flay & Cook, 1989). Fiscal limitations often restrict the use of each
experimental condition to just one community (Flay & Cook, 1989). The design is
called "quasi-experimental" because the communities are not randomly assigned to
experimental conditions.

A pre-campaign survey, conducted either by telephone or face-to-face, is typically
administered to establish baselines in both experimental and control communities.

74


Subsequent surveys are conducted during and after the campaign with the same
respondents to.assess change (these are chilled "cohort" or "panel" surveys).
Subsequent data analyses often focus .on those persons at greatest health risk,
when they can be identified (Flay -& Cook, 1.989). To control for the effects of
repeated testing, companion surveys are often done that involve a new,
independently drawn sample ("cross-sectional" surveys). Ideally, that sample
should be drawn at random from the target population". Depending on the
project, physiological measures might also be recorded.

The later surveys typically include questions that assess each individual's exposure
to the campaign. The most conservative approach to analyzing the. data is to
include all respondents, regardless of their exposure.  But it is useful in some cases
to conduct post hoc analyses that compare those who have been exposed to the
campaign with those who have not, recognizing that the lack of random
assignment to these subgroups makes the meaning of such differences ambiguous
(Lau et al., 1980). Brown, Bauman, and Padgett (1990) have identified another
significant drawback: Respondents often cannot report accurately whether they
have been exposed to a campaign, especially when the campaign's materials do
not stand out from those of similar campaigns.

The principal difficulty with this quasi-experimental design is that the communities
will differ in important ways before or during the study (Flay & Cook, 1989; Lau et
al., 1980). Even when efforts are made to match the two sets of communities on
a score of relevant variables, there is always the possibility that other unmeasured
differences between the communities .will contribute to or masking any differences
in outcome. Only the random assignment of very large numbers of .communities to
treatment conditions can obviate this problem, but that is fiscally impossible in
most cases.

Another threat to the validity of this design is the occurrence of events that affect
one community more than another (C&mpbell & Stanley, 1983; Cook & Campbell,
1979). Concern over this potential problem has promoted organizers of the
Pawtucket Heart Health Program to conduct content analyses of the major
newspapers in each of its test communities and to conduct yearly interviews with
health agency officials to learn about other interventions and their participation
rates (Assaf, Banspach, Lasater et al., 1987; Lefebvre, Lasatar, Carleton, &
Peterson, 1987).

   The highly touted Media-Advertising Partnersha for a Drug-Free America (Black, 1988)
conducted both a benchmark and one-year follow-up studies using matched samples of children
ages 9 to 12, teenagers ages 14 to 17, college students, and adults. So that respondents would
remain anonymous, the evaluators used a "mall-intercept" procedure to recruit questionnaire
respondents. Unfortunately, this procedure results in the selection of non-random samples.

75


A final limitation is that the mass media interventions themselves are complex,
which makes it impossible to derive definitive inferences ,about what each
individual component.adds.to the campaign. For this reason, the quasi-
experimental research design employed by Bauman et al. (1991) is especially
noteworthy. In this study, three different campaign variations were tested in
different communities: radio spots alone; radio spots, plus a sweepstakes; and
radio spots, a sweepstakes, and television spots.

National Camoaian$

Testing the impact of a national campaign is extremely difficult, for with use of the
national media, it is usually impossible to create meaningful comparison groups.
An exception is when the intensity of the campaign can be varied acrdss media
markets. The Media-Advertising Partnership for a Drug-Free America reported, for
example, that greater change in negative attitudes toward drugs was found in ten
U.S. media markets that received a 50 percent greater exposure to the campaign
materials (Black, 1988). Unfortunately, the high-exposure areas were not selected
at random and may have differed in important ways from other areas of the
country.

A good alternative is to collect time-series data on attitudes and self-reported
behavior. A good example is the work by Warner (1989) on the impact of the anti-
smoking crusade that began with the Surgeon General's report in 1964 and that
included extended news coverage, anti-smoking public service announcements,
and other media elements. Warner's strategy was to develop mathematical models
that described trends in per capita cigarette consumption in the United States
before 1964, to project post-1964 consumption based on those models (that is, as
if the intensive anti-smoking campaign had never occurred), and then to compare
those projections to actual consumption". Warner (1977) notes several
deficiencies in this commonly used measure of tobacco consumption, particularly
that it masks changes in the composition of the smoking population and that it
does not distinguish changes in smoking status from a mere reduction in smoking.

There are limitations to the time-series method. First, time-series modeling requires
that reliable and valid data are available for an extended period of time. In many
cases, the only data available will be broad indicators, such as statistics on
alcohol-related traffic fatalities, rather than specific indicators of project objectives,

Is

   The predictive mathematical models include the following variables for each year studied
beginning in 194 7: the value of the cigarette price index; the natural logarithm of the last two
digits of the year, which reflects a natural increase over time in the smoking population, primarily
due to more widespread use by women and increased consumption by existing smokers; and the
per capita consumption for the previous year, which reflects the contribution of "habit. w

76


such as greater use of designated drivers.  Such measures must often come from
survey data. Unfortunately, many projects will find that detailed surveys on their
topic area have not been done before.

A second limitation of the time-series method is the extreme difficulty of
disentangling a mass communication campaign's contribution from that of other
programs or even from broader historical trends or events. For the purposes of
program evaluat.ion, the time-series technique requires that there be a signal event
whose occurrence can be precisely defined in time, such as the beginning of a new
anti-smoking campaign in a government-controlled media environment (e.g.,
Doxiadis et al., 1985) or the introduction of. a new concept such as .designated
driver (DeJong & Winsten, 1990b). In the field of public health, however, the
occurrence of such events at the national level is rare.

REFERENCES

Advertiser Report: Tough Year Ahead for Brewers. &&&a&$~: 66, 69, December 31, 1990.

Advertising Council: News Release- Accountabilitv of Advertisina Under Close Watch. New York:
Advertising Council, April 8, 199 1. '

Ajax 1: How to Market a Nonmedical Contraceptive: A Case Study from Sweden, in Redford MH,
Duncan GW, Prager OJ feds), The Condom: Increasino Utilization in the United States. San
Francisco: San Francisco Press, 1974.

Aizen I, Fishbein M: Attitude-Behavior Relations: A Theoretical Analysis and Review of Empirical
Research. Psvcholooical Bulletin 84: 888-9 18, 19 77.

Alcalay R: The Impact of Mass Communication Campaigns in the Health Field. Social Science and
Medicine 17: 87-94, 1983.

Alcala y R, Taplin S: Community Health Campaigns: From Theory to Action, in Rice R, A tkin C
feds), Public Communication Camoaions (Second Edition). Newbury Park, CA: Sage, 1989.

All Fired Up Over Smoking: New Laws and Attitudes Spark a War. m: 64-69, 71, April 18,
1988.

Altman DL, Piotrow PT: Social Marketing: Does It Work? Population Reoorts l.lZ I[. Baltimore:
Population Information Program, Johns Hopkins University, 1984.

Altman 0, Slater M, Albright C, Maccoby N: How an Unhealthy Product is Sold: Cigarette
Advertising in Magazines, 1960- 1985. Journal of Communication 37: 95- 106, 1987.

Arkin EB, Denniston R, Roman0 RM: The Government Perspective, in Atkin C, Wallack L feds),
&gs Communication and Public Health: Comolexities and Conflicts. Newbwy Park, CA: Sage,
1990.

77


Assaf A, Banspach S, Lasater T, McKinlay S, Carleton R: The Pawtucket Heart Health Program: 8.
Evaluation Strategies Rhode Island Medical Journal 70: 54 l-546, 6987.

A tkin C: Research Evidence on Mass Med&ed Health Communication Campaigns, in Nimmo D
led), Communication Yearbook (Volume 3). New Brunswick, NJ: Transaction, 1979.

A tkin C: Mass Media Information Campaign Effectiveness, in Rice R, Paisley W feds), Public
Communication Camoaions. Beverly Hills, CA: Sage, 198 1.

Atkin C: Be Smart. Don `t Start!, in Rice R, Atkin .C ledsl, Public Communication Camoaions
(Second Edition). Newbury Park, CA: Sage, 1989a.

Atkin C: Television Socialization and Risky Driving by Teenagers. Alcohol, Druos and Driving 5: l-
11, 1989b.

Atkin C, Arkin E: Issues and Initiatives in Communicating Health Information, in Atkin C, Wallack L
feds.), M J
         mmuni                        nflicts. Newbwy Park, CA:
Sage, 1990.

Atkin C, Freimuth V: Formative Evaluation Research in Campaign Design, in Rice R, Atkin C feds),
Public Communication Campaians (Second Editionl. Newbury Park, CA: Sage, 1989.

Ball-Rokeach SJ, Rokeach M, Grube J: The Great American Values Test: lnfluencino Behavior and
Beliefs Throuoh Television. New York: Free Press, 1984.

Bandura A: Social Learnina Theorv. Englewood Cliffs, NJ: Prentice-Hall, 1977.

Bandura A: Self-Efficacy: Toward a Unifying Theory of Behavioral Change. Psvcholooical Review
84: 191-215, 1984.

Bandura A: Social Foundations of Thouoht and Action: A Social Coanitive Theory. Englewood
Cliffs, NJ: Prentice-Hall, 1986.    '

Bandy P, President P: Review of the Literature in Druo Abuse Prevention and Mass Medig.
Rockville, MD: National Institute on Drug Abuse, U.S. Department of Health and Human Services,
1982.

Bantz C, McCorkie S, Baade R: The News Factory. Communication Research 7: 45-68, 1980.

Basch C: Focus Group Interview: An Underutilized Research Technique for Improving Theory and
Practice in Health Education. my 14: 4 1 l-448, 1987.

Bauer R: The Obstinate Audience: The Influence Process from the Point of View of Social
Communication. American Psvcholooist 19: 319-328, 1964.

Bauman K, Brown J, Bryan E, Fisher L, Padgett C, Sweeney J: Three Mass Media Campaigns to
Prevent Adolescent Cigarette Smoking. Preventive Medicine 17: 5 1 O-530, 1988.

Bauman KE, LaPrelle J, Brown JO, et al: The Influence of Three Mass Media Campaigns on
Variables Related to Adolescent Cigarette Smoking: Results of a Field Experiment. American
Journal of Public Health 8 1: 597-604, 199 1.

70


Bem DJ: Beliefs. Attitudes. and Human Affairs. Belmont, CA: Brooks/Cole, 1970.

Benn A: The 27 Most Common Mistakes in Advertising. New York: American Management
Association, 1978.

Black G: ch n in A i d
   ;Q
                 T war Dr      :
Partnershio for A &Jo-Free America. Inc.. Executive Summarv and Statistical Report. Rochester,
NY: Gordan S. Black Corporation, 1988.

Black T, Far/e y J: Retailers in Social Program Strategy: The Case of Family Planning. ' Columbia
Journal of World Business: 33-41, Winter 1977.

Black TRL, Harvey PO: A Report on a Contraceptive Social Marketing Experiment in Rural Kenya.
Studies in Familv Planning 7: 10 I- 108, 19 76.

Bloom P, Novelli W: Problems and Challenges in Social Marketing. Journal of Marketing 45: 79-
88, 1981.

Bollier 0: Raise the Halo High, Channels: 32-34, 36, 38, April 1989.

Bonaguro J, Miaoulis G: Marketing: A Tool for Health Education Planning. Health Education: 6-
11, January/February 1983.

Bornstein RF: Exposure and Affect: Overview and Meta-Analysis of Research, 1968- 1987.
Psvoholoaical Bulletin 106: 265-289, 1989.

Boster F, Mongeau P: Fear-Arousing Persuasive Messages, in Bostrom R led), Communication
Yearbook (Volume 81. Beverly Hills, CA: Sage, 1984.

Bovee C, Arens W: Contemoorarv Advertising. Homewood, IL: Irwin, 1986.

Breed W, DeFoe JR: The Portrayal of the Drinking Process on Prime- Time Television. Journal of
Communication 3 1: 58-67, 198 1.

Breed W, DeFoe JR: Effecting Media Change: The Role of Cooperative Consultation on Alcohol
Topics. Journal of Communication 32: 88-99, 1982.

Brown JO, Bauman KE, Padgett CA: A Validity Problem in Measuring Exposure to Mass Media
Campaigns. Health Education Quarterly 17: 299-306, 1990.

Campbell 0, Stanley J: Ex rim n I n
          t
                                        rch. Chicago:
Rand-McNally, 1963.

Cartwright 0: Some Principles of Communication: Selected Findings of Research on the Sale of
U.S. War Bonds. Human Relations 11: 253-267, 1949.

Chapman S: Advertising as Myth: A Reevatuation of the Relationship of Cigarette Advertising and
Smoking, in Leathar 0, Hastings G, O'Reilly K, Davies J ledsl, Health Education and the Media II.
Oxford: Pergamon Press, 1986.

Chapman S, Egger G: Forging an Identity for the Non-Smoker: The Use of Myth in Health
Promotion. International Journal of Health Education 23: I- 16, 1980.

79


Chu G: Fear Arousal, Efficacy, and Immanency. Journal of Personalitv and Sot ial Psvcholooy 5:
5 17-524, 1966.

Cofford S: Koop's Swan Song: Celebrity Beer Ads Could Feel the Heat. Advertisino A@: 3, 68,
May 22, 1989.

Cook T, Campbell D: Quasi-Exoerimentation: Desion and Analvsis Issues for Field Settinos.
Chicago: Rand-McNally, 19 79.

Cooke P: Kick Me:, I Smoke. Hiooocrates: 66-72, July/August 1989.

Creamer E: An Advertising Story: Refining the New England Telephone Campaign. New Enaland
Point of View: 15- 17, May 1989.

Crowne D, Marlowe 0: The Aoproval Motive: Studies in Evaluative Oeoendence. New York:
Wiley, 1964.

Davis R: Health Education on the Six O'Clock News: Motivating Television Coverage of News in
Medicine. Journal of the American Medical Association 259: 1036- 1038, 1988.

Dawson DA: AIDS Knowledge and Attitudes for January-March 1989: Provisional Data from the
National Health Interview Survey. A)6. Hyatts-
ville, MD: National Center for Health Statistics, Public Health Service, 1989.

DeJong W: An Examination of Self-Perception Mediation of the Foot-in-the-Door Effect. Journal of
Personatitv and Social Psvcholoay 37: 222 l-2239, 1979.

OeJong W: Arrestina the Demand for Druos: Police and School Partnershi& to Prevent Drug
Abuse. Washington, DC: National Institute of Justice, U.S. Department of Justice, 1987.

DeJong W: wq
m. Unpublished manuscript. Boston:
"H I er "  n R s ons
Harvard School of Public Health, 1988.

DeJong W: Condom Promotion: The Need for a Social Marketing Program in America's Inner
Cities. American Journal of Health Promotion 3: 5- 10, 16, Spring 1989.

DeJong W, Winsten JA: The Use of Mass Media in Substance Abuse Prevention. Health Affairs:
30-46, Summer 1990%.

DeJong W, Winsten JA: The Harvard Alcohol Project: A Demonstration Project to Promote the
Use of the "Designated Driver. * Proceed' os'of the 11 th International Conference on Alcohol.
Druos and Traffic Safety. Chicago: Natirnal Safety Council, 199Ob.

DeJong W, Winsten JA: Responding to AIDS: Limits on the
Strategic Use of the Broadcast Media to Effect Behavior Change, in Sepulveda J, Fineberg H, Mann
J ledsl, AIDS Education and Communication. New York: Oxford University Press, 199 1.

Dessart G: The .Media Industry Perspective, in Atkin C, Wallack L feds}, Mass Communication and
Public Health: Comolexities and Conflicts. Newbury Park, CA: Sage, 1990.

80


Doxiadis SA, Trihopoulos DV, Ph ylactou HD: Impact of a Nationwide Anti-Smoking Campaign.
Lancet: 712-713, 1985.

Dwyer T, Pierce JP, Hannam CD, Burke -N: Evaluation of the Sydney "Quit. For Life" Anti-Smoking
Campaign: Il. Changes in Smoking Prevalence. Medical Journal of Australia 144: 344-347, 1986.

Egger G, Fitzgerald W, Frape G, et al: Results of a Large Scale `Media Anti-Smoking Campaign in
Australia: North Coast "Quit For Life" Programme. British Medical Journal 287: 1125- 1128,
1983.

El-Ansary Al, Kramer OE: Social Marketing: The Family Planning Experience. Journal of Marketinq
37: l-7, 1973.

Evans R, Rozelle R, Maxwell S, et al: Social Modeling Films to Deter Smoking in Adolescents:
Results of a Three Year Field Investigation. Journal of Aoolied Psvcholoav 66. 399-4 14, 1981.

Farquhar J: The Potential Role of the Media in Public Health, Education, and Health Policy, in
Hamner J, Jacobs S reds), The Media. Communication, and Health Policv. Memphis: Center for
Health Sciences, University of Tennessee, 1984.

Farquhar J, Fortmann S, Maccoby N, et al: The Stanfoid Five City Project: Design and Methods.
American Journal of Eoidemioloav 122: 323-334, 1985%.

Farquhar J, Maccoby N, Solomon 0: Community Applications of Behavioral Medicine, in Gentry W
fed), Handbook of Behavioral Medicine. New York: Guilford Press, 1984.

Farquhar J, Maccoby N, Wood P, et al: Community Education for Cardiovascular Health. Lancet:
1192-l 195, 1977.

Festinger L, Maccob y N: Resistance to Persuasive Communications. Journal of Personalitv and
Abnormal Psvcholoav 68: 359-366, 1964.

Fishbein M, Ajzen I: Belief. Attitude, Intention. and Behavior: An Introduction to Theorv and
Research. Reading, MA: Addison-Wesley, 19 75.

Flay B: Mass Media Linkages with School-Based Programs for Drug Abuse Prevention.  Journal of
School Health 56: 402-406, 1986.

Flay B: Sellina the Smokeless Societv: 56 Evaluated Mass Media Proorams and Camoaions
Worldwide. Washington, DC: American Public Health Association, 1987.

Flay B, Burton 0: Effective Mass Communication Strategies for Health Campaigns, in Atkin C,
Wallack L feds), M s
        a.
        Newbury
Park, CA: Sage, 1990.

Flay B, Cook T: Three Models for Summative Evaluation of Prevention Campaigns with a Mass
Media Component, in Rice R, Atkin C feds), P C (Second Edition).
                              Ii
Newbury Park, .CA: Sage, 1989.

Flay B, DiTecco 0, Schlegel R: Mass Media in Health Promotion: An Analysis Using an Extended
Information-Processing Model.  Health Education Quarterly 7: 12 7- 14 7, 1980.

81


Flay B, Hansen W, Johnson C, et al: Implementation Effectiveness T