Effective Mass Media Strategies for Drug Abuse Prevention Campaigns

Full text

CHAPTER
2
Effective Mass Media Strategies
for Drug Abuse
Prevention Campaigns
PHILIP
PALMGREEN
LEWIS
DONOHEW
INTRODUCTION
Mass communication holds substantial promise as a tool for reaching and persuading people to
adopt
new
and healthier
lifestyles.
This
has
long
been recognized
by
those interested
in
prevention
of drug abuse and in other unhealthy behaviors (Flay & Sobel, 1983; Rogers & Storey, 1987;
Schilling & McAlister, 1990; Wallack, 1989). Prevention efforts, such as the National Institute
on Drug Abuse's
"Cocaine:
The Big
Lie"
campaign and
the
Partnership for
a
Drug-Free America
anti-drug campaign, have heavily relied on the promise. It is also reflected in the launching in
1998 of
a
historic $2 billion,
5-year,
media-based campaign directed at reducing illicit drug use
among
9-
to 18-year-olds. This campaign, directed
by
the Office of National Drug Control Policy
(ONDCP), has many facets but relies primarily on televised anti-dmg public service ads (PSAs)
and is by far
the
largest federally funded drug abuse prevention effort in history.
This strong dependence on the mass media in prevention efforts is not unusual—the mass
media are the primary or leading components
in
a
variety
of
public
health campaigns and frequently
are the only component (Backer, Rogers, & Sopory, 1992; Flay, 1987; Rice & Atkin, 1989). As
Bauman et
al.
(1991)
note,
"This is the most common and practical application of
mass
media in
public health and, unlike multiple-component approaches, is capable of ready distribution on a
PHILIP PALMGREEN AND LEWIS DONOHEW
• Department of Communication, University of Kentucky, Lexington,
Kentucky 40506
27

28 Philip Palmgreen
AND
Lewis Donohew
national level"
(p.
602). At the very least, as Romer (1994) observes, "mass-media communication
campaigns to alter risky behavior are seen increasingly as a critical adjunct to school-based
programs and community-wide interventions" (p. 1073). To what extent is this widespread faith
in the power of the media justified?
Although the eaily history of mass-media campaigns, particularly those involving health,
was lai'gely one of failure (Flay & Sobel, 1983; Rogers & Storey, 1987), the promise of reaching
large audiences has led to continued efforts, a sharpening of design methodologies, and more
realistic campaign expectations. These more sophisticated efforts, combined with more powerful
evaluation methodologies, provide evidence that media health campaigns can be effective in
changing beliefs, attitudes, intentions, and even behaviors, when properly designed (Backer,
1990;
Perloff,
1993; Rogers & Storey, 1987).
Design elements that have contributed to successful campaigns include sophisticated au-
dience segmentation and targeting, the use of formative research in message creation, the de-
velopment of professional-quality messages that compete effectively with product ads and other
features of the communication environment for the attention of the audience, the use of appropriate
channels of communication, and the incorporation of more sophisticated theories of persuasion in
campaign design (Backer,
1990;
Perloff,
1993;
Rogers & Storey,
1987).
More rigorous techniques
of formative, process, and summative evaluation, coupled with more powerful statistical tools,
have detected a variety of campaign effects. Such research generally shows that coupling media
with other kinds of interventions is more successful than either media or nonmedia efforts alone
(Flora, Maibach, & Maccoby,
1989;
Rogers & Storey,
1987).
There is growing evidence, however,
that, when used correctly, media alone can have significant positive impacts on health-related at-
titudes, beliefs, and behaviors (Beck et al., 1990; Flay, 1987; Flora, Maccoby, & Farquhar, 1989;
Zastowny et al., 1993).
So much research has been compiled on successful public health campaigns, either media-
only or media supplemented by other channels, that
a
series of generalizations on the most effective
ways to use the media has disseminated widely through the literature for use by communication
practitioners (see Backer et al., 1992; Flay, 1987; Flay & Sobel, 1983;
Perloff,
1993; Rogers &
Storey, 1987). This chapter highlights three of the most important principles—ones that we have
found to be highly useful in our own approach to media interventions.
1.
Design a campaign that will achieve widespread, frequent, and prolonged exposure to a
message.
In traditional advertising terms this means that the media campaign messages must have
high reach (the proportion of target audience members exposed to a message at least once) and
frequency (the average number of exposures per audience member reached). These goals are
much easier to state than to achieve. To accomplish them means that campaign practitioners must
develop messages that can elicit high levels of attention from the target audience and disseminate
the message through media channels actually used by audience members. It also means that
(1) sufficient financial resources must be available to purchase adequate amounts of time or space
in desired media vehicles (such as
TV
and radio, newspapers, magazines), or (2) considerable
salesmanship and maiiceting skill must be used to persuade media gatekeepers to donate these
precious resources in times or locations that are likely to be seen by the target audiences, or (3) a
combination of both purchased and donated time and/or space should be used. More campaigns
are tuming to option
3,
with an emphasis on purchasing, to achieve campaign
goals.
These include
successful anti-smoking initiatives in California and Massachusetts (Hu, Sung, & Keeler, 1995;
Siegel
&,
Biener, 2000) and the ONDCP antidrug campaign. Still, paid media schedules in health
campaigns are in the minority, and more research is needed to compare the effectiveness of
paid versus donated schedules (Murry, Stam, & Lastovicka, 1996). One recent field experiment

EfTecUve Mass Media Strategies for Drug Abuse Prevention Campaigns 29
investigating this found no difference in effectiveness; however, the donated campaign in this
study emulated the paid campaign closely, something that is rare in practice (Murry et al., 1996).
The targeting advantages of paid campaigns ordinarily are substantial, so we would expect the
trend toward paid media schedules to continue.
2.
Use audience segmentation strategies to target messages to at-risk audiences.
This is the cornerstone of the social marketing approach. Segmentation or targeting can
lead to much more efficient and effective dissemination of campaign messages to those most in
need of prevention information. While demographic data can provide a rudimentary beginning,
any targeting scheme should also be based on psychographic variables (such as attitudes, values,
beliefs,
and
personality characteristics) linked both to the behavior of interest (marijuana or other
substance use) and to the communication channels and message styles most preferred by target
audience members (Backer et al., 1992; Slater, 1996).
3.
Use formative research throughout
the
audience segmentation, message design,
and
chan-
nel selection phases.
Such research, both qualitative and quantitative, is essential in determining the relevant
needs,
beliefs, behaviors, and attitudes of the target audience; in designing messages to attract
the attention of and persuade audience members; and in determining the media channels and
vehicles most used by the audience (Atkin & Freimuth, 1989; Backer et al., 1992; Rogers &
Storey, 1987). The research should involve careful pretesting of prevention message ideas at the
concept stage, the "storyboard" or "rough-cut" stages, and the
final
production stage. Ideally, this
testing should be done with members of
the
target audience, media professionals, and behavioral
scientists knowledgeable in both the behavior of interest and theory-based approaches to message
design.
Despite encouraging
growth in the use
of
these
and other principles, many important questions
remain. A number of techniques have been found
to
be successful, for example, but little is known
about
the
process by which media messages begin to change attitudes and behaviors. What
are
the
causal lag periods involved? Are there more effective ways of designing and placing prevention
messages? What amounts of expensive media time and space are needed to bring about the
desired change? And one of
the
most important and vexing questions concerns the effectiveness
of different channels in the media mix. Many public communications campaigns, in an effort
to maximize effects, have used a variety of media simultaneously, including television, radio,
newspapers, magazines, and billboards, as well as nonmedia interventions. This makes it difficult
to evaluate the separate contributions of
these
different channels on observed changes in outcome
variables.
EFFECTS OF TELEVISED PUBLIC
SERVICE ADS
The
effects of television
are
of particular interest
to those
involved
in drug
abuse prevention because
of this medium's ability to reach a variety of populations, including adolescents (Klein et al.,
1993;
Romer, 1994). Television is by
far the
most widely used means of disseminating prevention
messages, usually
in the
form of PSAs (Backer, 1990). Understanding television's potential effects
on at-risk populations, whether
when used alone or in
conjunction with institutional
or other
media
channels, is vital to campaign designers; so the confounding of television's effects with those of
other channels in many otherwise well-designed campaigns is unfortunate. Studies involving
the use of televised anti-drug PSAs alone, on the other hand, generally suffer from mistakes in

30
Philip
Palmgreen
AND
Lewis
Donoliew
campaign execution, including violating two of the principles discussed previously: (1) lack of
widespread, frequent, and prolonged exposure to messages-in several campaigns, PSAs were
aired outside prime time and/or on noncommercial stations, and then only infrequently; and (2)
lack of appropriate segmentation or targeting—many anti-drug PSA campaigns have been directed
at nonidentifiable audience segments (Flay & Sobel, 1983).
Evaluations of such campaigns tell us little about the potential persuasive effects of well-
executed PSA campaigns that
use more
recently developed and proven
techniques.
Many of these
campaigns, too, have had the limited, and perhaps appropriate, primary objective of increasing
knowledge levels or raising the salience of
a
health-related issue and have not been directly con-
cerned with changing attitudes or behaviors. In addition, when campaigns have been correctly
designed and carried out, they are not always evaluated
correctly.
PSA recall and campaign-issue
salience have been the primary measures of effectiveness in many campaigns (especially infor-
mational ones). Simple cross-sectional post-test surveys have been used frequently. Longitudinal
or
panel
studies often have involved simple pre- and post-test designs that fail to account for
pre-
and postcampaign trends in criterion variables. When such longitudinal trends, which yield po-
tentially valuable information about change processes, have been reported, the data usually have
been subjected to "eyeball" inspection rather than to appropriate statistical tests of intervention
effectiveness, such as time-series analysis (Beck et
al.,
1990;
Hammond, Freimuth, & Morrison,
1987;
Krishnamurthi, Narayan, & Raj, 1986; Murry, Stam, & Lastovicka, 1993; Pierce et al.,
1986,
1992;
Ross & Scott,
1993;
Shelley et
al.,
1991).
Another problem is the failure of most studies of
PSA
effectiveness to use control commu-
nities that are free of confounds from other mass-media efforts. A major exception is a well-
controlled,
4-year
longitudinal study on the prevention of cigarette smoking in adolescents that
compared
the
impact of mass-media-plus-school interventions
in two
communities versus school-
only
interventions
in two
matched communities (Flynn et
al.,
1992,1995).
The
media intervention,
which took the form of four approximately
6-month-long
campaigns spaced over
4
years, used a
combination of television and radio spots in purchased
and
donated
time in
popular
teen
program-
ming to ensure high saturation. There were significant reductions (which increased annually) in
reported smoking, with related effects on smoking attitudes and beliefs, in the media-plus-school
compared to the school-only communities. These reductions could be attributed directly to the
addition of
the
PSA components, although it was not possible to distinguish between the effects
of television and radio.
Despite this research, it is still an open question whether televised anti-drug PSA cam-
paigns using more advanced principles of campaign design can go beyond well-designed and
acknowledged informational or agenda-setting effects to produce significant changes in drug-
related
attitudes,
beliefs, and ultimately
behaviors.
Reviews concluding that televised PSAs have
effects only on knowledge or awareness are based primarily on evaluation of either information-
only campaigns, campaigns that were not designed to isolate the effects of televised PSAs, or
campaigns that contained
flaws
in execution or evaluation (Gantz, Fitzmaurice, &
Yoo,
1990).
Research on the large, long-term, and well-designed Partnership for a Drug-Free America
television campaign (supported by more than
$3
billion in donated air time and print space since
1987) provides some evidence of such effects (Black, 1991; Zastowny et al., 1993). Published
evaluations of this campaign, however, have been criticized for being based on a series of an-
nual cross-sectional samples that used
a
controversial mall-intercept design for several
years.
No
satisfactory control population exists for this national campaign. Complicating the assessment
of the campaign's effects is the fact that a number of drugs (such as marijuana, cocaine, and
amphetamines) were already exhibiting downward trends in use prior to the start of the cam-
paign in 1987. Other history and maturational factors, such as media coverage of
drugs,
are also
uncontrolled in the evaluations.

Effective
Mass Media Strategies for Drug Abuse
Prevention Campaigns 31
Still, areas receiving greater partnership PSA saturation have shown much larger changes in
annual
cross-sectional surveys (compared
to
less-saturated
areas) in
drug-related
attitudes,
beliefs,
reported use of a variety of illicit drugs, and intentions to use such drugs (Black,
1991;
Block,
Morwitz, & Sen, 1996). While these latter findings should be inteipreted cautiously because of
their cross-sectional nature, they provide the strongest nonlaboratory evidence available that the
mass media (at least televised PSAs) can successfully discourage the use of illicit drugs.
WHAT WORKS AND WHY
Despite limited empirical evidence on the effectiveness of the mass media in preventing use of
illicit drugs, practitioners can take heart (and guidance) from the much larger body of research
literature dealing
with the
impact of media-based interventions
on
health-related
behaviors in
gen-
eral
(Perloff,
1993;
Rogers & Storey, 1987) and on the use of licit substances, such as cigarettes
(Bums, 1994; Flay,
1987;
Flynn et al.,
1995;
Hu et al., 1995; Siegel & Biener, 2000). This more
general literature, as noted earlier, provides ample evidence that well-planned media campaigns
can influence a wide variety of health-related attitudes, norms, and
behaviors.
These studies also
address an issue on which the sparse media drug abuse prevention literature (with the exception
of the SENTAR approach discussed in the following) is largely silent—what kinds of campaign
strategies, persuasive arguments, and other message characteristics work best and why?
We
have
already discussed three important principles for campaign design that have emerged from the
public communications campaign literature. We should add, however, one very important em-
pirical
finding
from this literature—that theory-based media interventions have been much more
likely to be successful. Ample evidence exists of successful campaigns that used such guiding
frameworks as social learning theory, diffusion of
innovations,
the theory of reasoned action, the
health belief model, the elaboration likelihood model, and protection motivation theory (Flora
et al., 1989; Maibach & Parrott, 1995; McAlister et al., 1989; Petty, Baker, & Gleicher, 1991;
Rogers,
1995;
Rosenstock, 1990; Schilling
&
McAlister, 1990; Zimmerman & Vemberg, 1994).
Other theoretical perspectives, such as peer cluster theory (Getting & Beauvais, 1987), can be
drawn from school- or community-based prevention efforts.
While principles from a number of these theories have been applied, at least implicitly, in
drug abuse prevention media
campaigns,
there has been no systematic evaluation of their relative
(or combined) efficacy in such interventions. Still, their success in other health contexts strongly
suggests that they can be applied effectively to drug abuse prevention. Schilling and McAlister
(1990) offer
a
number of cogent and detailed suggestions for applying several of the more widely
used theories to anti-drug campaigns. The strategic communication plan developed by Porter
Novelli for
the
ONDCP
media
drug abuse
prevention campaign relies heavily
on
principles derived
from social learning theory, peer-cluster theory, and the theory of reasoned action, as well as on
empirical
findings
from a
host
of
media and
nonmedia
interventions.
Evaluation of
the
campaign's
impact will, in effect, be the
first
major evaluation of
the
explicit application of
these
theories in
a media-based illicit-drug-abuse prevention campaign.
A SENSATION-SEEKING APPROACH TO DRUG
ABUSE PREVENTION
Another theoretical approach represented in the ONDCP campaign's strategic communications
plan is one we have been developing over the past 15 years at the University of Kentucky with
the support of a series of grants from the National Institute on Drug Abuse. This approach is,

32
Philip Palmgreen
AND
Lewis
Donoliew
to our knowledge, the only theory-driven approach to media-based illicit-drug-abuse prevention
developed specifically for, and tested in, that important context. It revolves around sensation
seeking, a particularly potent risk factor for
drug
use,
which can be used at three critical stages in
media campaign
design:
(1) segmenting or targeting the at-risk audience, (2) designing messages
that are effective with this audience; and (3) placing these messages in program contexts that are
attractive to the target
audience.
The result is a coherent, parsimonious, and powerful theoretical
framework that guides intervention strategies from inception to delivery and meshes well with a
number of other theoretical approaches to prevention.
Sensation Seeking
Sensation seeking is a personality trait associated with the need for novel, complex, ambigu-
ous,
and emotionally intense stimuli (Zuckerman, 1979, 1994). As measured by Zuckerman's
sensation-seeking scale, the concept has four dimensions:
1.
Thrill and Adventure Seeking:
A
desire to seek sensation through physically
risky
activi-
ties that provide unusual situations and novel experiences, such
as
parachuting and scuba
diving.
2.
Experience
Seeking:
A
desire
to
seek sensation through
a
nonconforming lifestyle, travel,
music, art, drugs, and unconventional friends.
3.
Disinhibition:
A
desire to seek sensation through social stimulation, parties, social drink-
ing, and a variety of
sex
partners.
4.
Boredom Susceptibility: An aversion to boredom produced by unchanging conditions or
persons and great restlessness when things are the same for any period of time.
Describing differences between high and low sensation seekers, Zuckerman (1988) has
observed that:
The high sensation seeker is receptive to novel stimuli; the low tends to reject them, preferring the more
familiar and less complex. The high sensation seeker's optimal level of stimulation may depend on the
levels set by the characteristic level of arousal produced by novel stimuli. Anything producing lower
arousal levels may be considered 'boring.*... Apart from the voluntary avoidance of high intensities of
stimulation, the low sensation seeker may have a type of nervous system that rejects such stimulation or
inhibits cortical reactivity to high intensity stimuli, (pp. 181-182).
Sensation Seeking and Substance Use
Sensation seeking is a consistent predictor of use of a variety of drugs and earlier onset of use
(Kilpatricket
al.,
1976;
Segal,
Huba,
&
Singer,
1980;
Zuckerman, 1979,1983,1994).
In
data from
our recent study of prevention of adolescent marijuana use (Palmgreen et al., 2001), sensation
seeking also correlates positively with the other five risk factors measured (deviance and lack
of opportunity, and peer, family, and community use of marijuana) and negatively with all six
protective factors (self-acceptance, absence of depression, quality of home life, law abidance,
religiosity, perceived sanctions against marijuana
use).
As such, the concept of sensation seeking
offers
an
important
avenue
for targeting at-risk
groups
and designing messages
to
reach
them.
The
construct
is
based
on
psychobiological theory
and has
been shown
to have
a
high
heritability factor
(Fulker, Eysenck,
&
Zuckerman,
1980;
Zuckerman, 1990,1994). It
has
a number of biochemical
correlates, including testosterone, monoamines and their metabolites (particularly monoamine

Effective Mass Media Strategies for Drug Abuse Prevention Campaigns 33
oxidase), and endorphins (Zuckerman, 1979,1986,1994). Research by Bardo and his colleagues
(Bardo, Nieswander, & Pierce, 1989; Bardo & Mueller, 1991; Bardo, Donohew, & Harrington,
1996) strongly suggests that novelty-seeking behavior
and
self-administration of drugs in animals
may involve a common dopamine system in the brain.
A moderate to strong association of sensation seeking with alcohol and illicit drug use has
been demonstrated in a large number of studies in a variety of populations (e.g., Kilpatrick
et al., 1976; Pederson, 1991; Segal et al., 1980; Zuckerman, 1979, 1994). Strong evidence
has emerged in the past decade that the relationship also holds with adolescents (Bamea,
Teichman, & Rahav, 1992; Clayton, Cattarello, & Walden, 1991; Huba, Newcomb, & Bentler,
1981;
Newcomb & McGee, 1989; Pederson, 1991; Teichman, Bamea, & Rahav, 1989; Thombs
et al., 1994), including those from different cultures (Bamea et al., 1992; Pederson, 1991;
Teichman et al., 1989). In a study of
junior
and senior high school students in Fayette County,
Kentucky, high-sensation seekers (HSS), as defined by median splits, were twice as likely as
low-sensation seekers
(LSS)
to report use of beer and alcohol during the prior 30 days and up
to 10 times as likely to report use of other dmgs (Donohew, 1988, 1990). Similar pattems of
HSS versus
LSS
differences in drug use were found among a cohort of Fayette County students
measured at four timepoints from the sixth to the eighth grades as part of an evaluation of
Dmg Abuse Resistance Education (DARE) (Clayton et al., 1991). In addition, HSS adolescents in
our most recent study (Palmgreen, et al., 2001) were up to four times as likely to exhibit past
30-day use of marijuana. Sensation seeking has been related to adolescent alcohol use in several
recent studies (Huba et al., 1981; Newcomb & McGee, 1989; Pederson, 1991; Thombs et al.,
1994;
Webb et al., 1991); and in a cross-sectional study of 1,900 Israeli high school students,
it was strongly associated with use of a number of licit and illicit drugs (Bamea et al., 1992;
Teichman et al., 1989). A Califomia study of 1,068 adolescents found moderate relationships
between various sensation-seeking dimensions and
a
number of illicit and licit substances (Huba,
etal., 1981).
In a 20-month Norwegian longitudinal study of 553 adolescents, sensation seeking was
characterized by
a
relatively high degree of temporal stability and was a consistent and important
predictor of use of cannabis, alcohol, benzodiazepine,
and
cigarettes (Pederson, 1991). Longitudi-
nal studies of variables closely related to the dimensions of
the
sensation-seeking scale also offer
strong evidence of
the
ability of
a
sensation-seeking
"superfactor"
to
predict risk-related behaviors
across long developmental time spans. A study of 1,034 boys measured at ages 6 and 10 years
showed that those high on novelty seeking and low on harm avoidance at age 6 (as measured by
Cloninger's personality scale) exhibited earlier onset of substance use (Masse
&
Tremblay, 1997).
A study
in
New Zealand followed
a cohort
from age
3
(n =
1,037)
to
age21 (n =961)(Caspietal.,
1997).
At age 3, study participants were rated on 22 behavioral characteristics. At age 18 they
were administered the Multidimensional Personality Questionnaire
(MPQ;
Tellegen, 1982), and at
age 21 they were measured on four different health-risk behaviors: alcohol dependence, violent
crime, risky sexual behavior,
and
dangerous driving habits. It was found that those who exhibited
each of these risky behaviors scored much lower (in comparison to those not exhibiting) on the
MPQ
scales of
Harm
Avoidance (the inverse of Thrill
and
Adventure Seeking), Control (roughly the
inverse of Experience Seeking), and Traditionalism (in many ways the inverse of Disinhibition),
and higher on Aggression (and, in some cases, on Alienation). The greatest differences on these
traits were displayed by those involved in multiple risky behaviors. Moreover, those possessing
this "risky
personality"
configuration
at
age
18 had
displayed similar temperament qualities
at
age
3 (Caspi & Silva, 1995). Drawing upon other data gathered on the cohort at ages 5,7, 9, 11, and
13 years, Caspi et al. (1997) suggest that "the origins of
a
personality type at risk for health-risk
behaviors may be found early in life and.. .the type stabilizes during adolescence." (p. 1061).

34 Philip Palmgreen
AND
Lewis Donohew
They go on to say that, in public health interventions:
Individual
differences in personality may influence (different) steps in the
persuasion
process (Cacioppo,
1986).
Thus,
different types of
individuals
may attend to,
comprehend,
accept, and
retain
different types
of
messages. Our research shows that young adults who engage in health-risk behaviors are different
psychologically
from their
peers.
If we know the personality characteristics of a target audience, it may
be
possible to tailor
campaigns
to zero-in on the
characteristic
motivations,
attitudes, and feelings of
that
audience
(Plant & Plant,
1992).
Knowledge of the psychological characteristics that motivate youth to
engage
in
health-risk behaviors
may
thus help public health
officials
choose more
effective
campaigns
that
would motivate
risk
takers
to
minimize
harm.
(p. 1061)
Message Sensation Value
and
SENTAR
We have followed the path described previously in designing our own approach to drug abuse
prevention—^SENTAR
(for SENsation-seeking TARgeting). It is well established that high-sensation
seekers, including the important target group of
HSS
adolescents, are particularly drawn to the
stimulation and/or mood-altering effects of a variety of
drugs.
What is especially important from
a prevention perspective, however, is that they also have distinct and consistent preferences for
particular kinds of messages based on their needs for the novel, the unusual, and the intense
(Donohew, Lorch, & Palmgreen, 1991; Zuckerman, 1979, 1990, 1994). High-sensation seek-
ers (usually defined as those above the median on the sensation-seeking scale) strongly prefer
messages that are high in sensation value, that is, the degree to which the content and formal
features of a message elicit sensory, affective, and arousal responses. These same individuals
dislike messages low in sensation value; low-sensation seekers generally display the opposite
pattern of message preferences. Our own extensive program of focus-group research involving
HSS
and
LSS
adolescents and young adults has confirmed that these preferences extend to televised
commercials and PSAs (Donohew et al., 1991). This research shows that
HSS
prefer messages
that have higher levels of the following attributes: (1) novel, creative, or unusual; (2) complex;
(3) intense stimuli that are emotionally powerful or physically arousing; (4) graphic or explicit;
(5) somewhat ambiguous; (6) unconventional; (7) fast paced; and (8) suspenseful. Of course, it
is not necessary for a message to have all of these characteristics at high levels to be attractive to
high-sensation seekers; but we can say with some confidence that the greater the number of these
characteristics a message has, the more attractive it will be to high-sensation seekers. If any one
of these characteristics is of primary importance though, it is high levels of novelty. Zuckerman
(1990) has reviewed research sho\ying that
HSS
"tend to give stronger physiological orienting
responses than lows to novel stimuli of moderate intensity, particularly when such stimuli are of
specific interest" (p. 313).
High-sensation-value messages thus may elicit more favorable evaluations and greater at-
tention from
HSS,
but are they more persuasive? In one laboratory experiment we designed and
produced two versions of a televised antidrug PSA—one high in sensation value
(HSV)
and one low
in sensation value
(LSV).
With high-sensation-seeking young adults, the HSV message produced
greater intent to call a hotline featured in each PSA than did an otherwise comparable
LSV
mes-
sage.
The opposite pattern was observed for
LSS
(Donohew et
al., 1991;
Palmgreen et
al.,
1991). In
another experiment a perceived message sensation value scale was developed and used to classify
(based on the responses of 50 subjects in a pilot study) 13 existing
TV
anticocaine PSAs as either
HSV
or
LSV
PSAs.
The
HSV
PSAS were much more effective than the
LSV
spots with high-sensation-
seeking young adults on the dependent variables of free and cued recall of message content, attitude
toward cocaine, and behavioral intention to use cocaine. Low-sensation seekers displayed the

Effective Mass Media Strategies for Drug Abuse Prevention Campaigns 35
opposite pattern for both free and cued recall but showed no significant
HSV
versus
LSV
dif-
ference on the attitude and behavioral intention measures (Everett & Palmgreen, 1995). The
amount of variance accounted for by these interactions was high, particularly for free and cued
recall.
Evidence for the persuasive impact of high-sensation-value anti-drug PSAs was also found
in a field study involving an actual televised PSA campaign conducted in Lexington, Kentucky
(Palmgreen et al., 1995). The campaign targeted young adults and older teens and included
five PSAs developed through formative research with focus groups consisting of high-sensation
seekers. The high-sensation-value spots concluded with an appeal to call a hotline for more
information about exciting alternatives to drug use. The PSAs were the sole source of information
about
the
hotline.
More
than
2,100 calls
to the
hotline
were
received
over the
course of the 5-month
campaign, with 98% calling to get information for themselves (as opposed to calling to get
information for friends, children, etc.). This is a relatively large number of calls from a small
market and a narrowly defined target audience. More than 73% of the callers were above the
population median on the sensation-seeking scale, as determined by a survey of hotline callers
and by
a
probability survey of
the
general population of 18- to 25-year-oids in Lexington (the age
range in which most of the callers fell). Within-campaign surveys indicated that high-sensation
seekers were indeed reached frequently by the PSAs, more so than low-sensation seekers. A
postcampaign probability survey also revealed the combined influence of sensation seeking and
drug use on exposure to the two most-aired PSAs. Both PSAs displayed the same recall pattern,
with Hss users of illicit drugs in the past 30 days displaying the highest recall certainty, followed
closely by the small group of
LSS
users (whose use status apparently rendered the PSAs salient to
them).
Close behind this group were the
HSS
nonusers, another very important group to reach in
a prevention campaign. Trailing these groups by a substantial margin (but still manifesting good
recall certainty levels) was the large group of
LSS
nonusers, the segment least at-risk for use of
illicit substances. Reported frequency of exposure was related to sensation seeking and drug use
in a similar fashion.
Evidence for the impact of a SENTAR campaign on actual illicit drug use stems from a
recent study (Palmgreen et al., 2001) that involved an innovative controlled interrupted time-
series design to evaluate the effectiveness of televised antimarijuana PSA campaigns targeted at
high-sensation-seeking adolescents
in
two matched
cities:
Lexington (Fayette
County),
Kentucky,
and Knoxville (Knox County), Tennessee. Specifically, televised antimarijuana PSAs, designed
and developed through formative research, were shown (using a combination of paid and do-
nated time) from January through April 1997 in Lexington. Similar campaigns were conducted
from January through April 1998 in both Lexington and Knoxville (see Figure 2.1). Beginning
8 months prior to the first Lexington campaign and ending 8 months after the 1998 campaigns,
personal interviews (computer assisted, self-administered) were conducted with 100 randomly
selected (without replacement) students in each county during each month (total
n
= 6,400). The
Fayette baseline campaign 1 post-campaign campaign 2 post-campaign
County
O1...O8
O9...O12
0,3... O20
O21...O34
O25...O32
Knox baseline baseline baseline campaign 1 post-campaign
County Oi...Og O9...O12
0,3...
Ojo On...
O24
O25...O32
Note: 0| corresponds to the /th observation. Observations are separated by one month and are based on
means of
100
participants each. Total
£{
~
3,200
per county.
FIGURE
2.1. Overview of Controlled Interrupted Time-Series Design with Switching Replications.

36
Philip Palmgreen
AND
Lewis
Donohew
population cohort followed was in the 7th through 10th grades initially and in the 10th grade
through 9 months after high school graduation upon completion.
During
the
first
8
months of the study (Oi
to
Og),
participants provided
data on
marijuana use
patterns in each county prior
to
the
first
Lexington campaign. During the next
4
months, students
in Fayette County were exposed to a televised anti-drug ad campaign employing high-sensation
value messages developed
by the
research
team.
Data collection continued
in the two counties
(O9
to O12), permitting comparisons of marijuana use with and without a campaign. Data gathered
over the next 8 months
(O13
to
O20)
established marijuana use trends after the
first
campaign in
Lexington and extended the baseline trend in Knox County prior to that county's
first
campaign.
During the ensuing 4 months
(O21
to
O24),
students in both counties were exposed to campaigns
identical to the
first
Fayette campaign, except a few new PSAs were introduced in both counties
to add novelty in Fayette. Data collection then continued
(O24
to
O32)
to measure postcampaign
trends.
The design controlled for trends in marijuana use prior to the campaigns and allowed esti-
mation of postcampaign trajectories. It also partially controlled for history, because any national
events affecting drug use should have affected both counties. In addition, contacts with school
drug prevention staff and daily monitoring of
the
major newspapers in each county revealed no
local or regional events or prevention efforts threatening comparability.
Because the cohorts in each county aged as the study progressed, marijuana use tended
to increase due to sociodevelopmental or maturational factors. However, because teens in both
counties reflected this secular trend, each county served as an appropriate control for the other.
Because each monthly sample was independent, sensitization, testing, and attrition were mini-
mized. External validity was enhanced by campaign replication at different sites and times, and
the design allowed both within- and between-county evaluations of campaign impact.
Full sample medians were used to separate the Knox and Fayette monthly samples into
groups of high- and low-sensation seekers. Time-series regression analyses indicated that all
three campaigns not
only
arrested but also actually reversed upward changes in 30-day marijuana
use among
HSS
adolescents. For
example,
30-day use among Knoxville
HSS
rose in linear fashion
from 16.6% initially to 33% over the 20-month precampaign period, then fell to 24% from
the start of the campaign to the completion of data gathering 12 months later. The drop in the
proportion of
HSS
using marijuana
was
26.7%.
The Lexington campaign results were similar. The
first campaign also reversed a strong upward trend in 30-day use among
HSS.
Perhaps because
Lexington
HSS
were higher than their Knoxville counterparts on most risk factors and lower on
most protective factors, the effects of the first campaign appeared to wear off about 6 months
after the campaign, as indicated by the resumption of
an
upward trend. This trend, however, was
also reversed by the second or "booster" Lexington campaign, and marijuana use continued to
fall until the completion of data gathering. The time-series regression models indicated that all
changes in slopes were statistically significant (p < .003).
Thirty-day use levels among
LSS
in both cities were less than one-third of
HSS
levels,
LSS
also
exhibited
no
upward trends
in use
during the
32
months of the study in either community. Because
of the "floor effect" of low use levels, and because
LSS
were not targeted by the campaign,
LSS
displayed no indication of campaign effects. These patterns give further emphasis to the impor-
tance of targeting high-sensation seekers with prevention messages and illustrate the strengths of
an interrupted time-series design with a control community in detecting campaign effects.
PROGRAM CONTEXT, We also applied the concept of message sensation value to the
TV program context of antidrug messages. Viewers ordinarily tune in to watch programs, not
commercials and
PSAs.
It follows that to reach high-sensation seekers at risk for drug use, PSAs

Effective Mass Media Strategies for Drug Abuse Prevention Campaigns 37
should be placed, if at all possible, in programs preferred by high-sensation seekers. Our research
shows that such programs have characteristics associated with high sensation value PSAs and
commercials (Lorch et
al.,
1994).
In a large laboratory experiment involving 328 young adults, we
found that antidrug PSAs embedded in Hsv programming received considerably higher attention
levels from high-sensation seekers than did those placed in
LSV
programming. Exposure to the
programming and ads took place with subjects placed individually in a naturalistic living room
setting with various reading options available if they chose not to watch television. Results from
this experiment were applied in the two campaign studies described previously by purchasing
PSA time in programming preferred by Hss audience members in precampaign audience surveys.
This use of
HSV
program contexts undoubtedly contributed to the success of these campaigns in
reaching the target audience of high-sensation seekers.
Summary of SENTAR Principles
The
SENTAR
approach to the prevention of substance use and abuse (as well as risky sex, reckless
driving, and other risk behaviors) can be summarized in the following principles.
1.
Use the sensation-seeking trait as one major segmentation variable. While sensation
seeking certainly is not the only risk factor in substance use and abuse, it is positively
correlated with most other risk factors identified in the literature and is moderately to
strongly related to use of a wide variety of substances. It also is longitudinally stable
and predictive of drug initiation and use over long developmental time spans. While the
trait should not be the only segmentation variable, it should play a major role in any
psychographic/demographic or other multivariate targeting scheme.
2.
Design prevention messages high in sensation value to reach high-sensation seekers.
Designing messages that have as many high-sensation-value characteristics as possible
(especially novelty) is essential to gain the attention of at-risk audience members in the
highly cluttered context in which most media exposure takes place. Messages too low
in sensation value are very likely to be ignored by those whom prevention practitioners
would most like to reach, especially when such messages run counter to audience attitudes
and behavior
patterns.
Breaking through the clutter is critical, and
HSV
messages are most
likely to accomplish this and go on to effect desired changes in attitudes, beliefs, and
behaviors.
3.
Use formative research with high- (and sometimes low-) sensation-seeking members of
the target audience. Such research is invaluable in determining the informational, social,
and other needs of the target audience regarding the behavior of interest, in designing
effective messages, and in choosing appropriate media channels and program contexts.
Such research at the message-design stage is especially important since there are many
ways to blend
HSV
message characteristics in novel and effective (or ineffective) ways.
The SENTAR approach offers no rigid prescriptions for message design, but rather is an
overarching theoretical framework in which there is much room for creative talent to
operate. Such freedom is essential if one is to succeed at the task of constantly generating
novel messages for an easily bored or habituated audience.
4.
Place prevention messages in high-sensation-value contexts. The most elegant message
consigned to a media channel or program context that no one in the target audience pays
attention to is like the beautiful hemlock falling in the forest—except that in a prevention
campaign it clearly makes no sound of consequence, nor is anyone there to appreciate its

38
Philip Palmgreen
AND
Lewis Donohew
beauty. Social
marketers,
of
course,
have
long
been aware of this important
but
often over-
looked maxim. Our research, however, has identified clear differences in the
TV
channel
and program preferences of high- and low-sensation
seekers,
based
on the
presence
or
ab-
sence
of
Hsv
attributes.
Because
this
research
also
indicates that
HSS
older teens
and
young
adults watch considerably
less
television than
do
LSS
(as
much
as 45
minutes
less
per
day),
information about
HSS
program preferences obtained through audience surveys prior to
(and during) a campaign provides valuable guidance for placement decisions. Such
HSS
versus
LSS
media consumption differences probably also extend to media channels other
than television, although research has not addressed this issue.
Sensation seeking, then, can be extremely useful in drug abuse prevention campaigns from
the social-marketing perspective of audience segmentation. As Slater
(1996)
observed in
a
highly
sophisticated treatment of health-audience segmentation, "It is essential that segments be pre-
dictive of the targeted behavior—if there is no association between segment membership and
the behavior of interest, the segment will have little or no value to the campaign designer or
health educator." (p. 272). But "to better guide channel selection and intervention decisions, the
segments should also be predictive of distinctive patterns of media use or reliance on different
organizational, community, or interpersonal channels." (Slater,
1996,
p.
272).
In other
words,
the
ideal segmentation variable should also predict use of communication channels through which
the target audience might be reached. We would add one more provision: that it also specify the
characteristics of
messages
most preferred
by
target audience members. Most risk and protective
factors associated with drug use can satisfy criterion
No.
1— they singly or multiply predict use
of
a
variety of illicit
substances.
Most, however, can provide little theoretical guidance
in
fulfilling
criterion No.
2—use
of communication channels (although formative research can describe the
channels used by those high or low on a particular factor). And almost none can also meet the
third criterion of specifying characteristics of effective
messages.
Sensation seeking, on the other
hand, satisfies all three criteria defining an ideal segmentation variable for
drug
abuse prevention.
Moreover, the SENTAR approach provides both a theoretical basis and empirical evidence for the
connections between sensation seeking and each segmentation criterion.
The Flexibility of
SENTAR
The SENTAR approach to media drug abuse prevention does not compete with other successful
theoretically based approaches, such as social cognitive theory, the theory of reasoned action,
or peer cluster theory. Rather, it can and should be used in conjunction with one or more of
these established
approaches.
For example, the televised PSAs we developed for our recent anti-
marijuana campaign study primarily follow a theory of reasoned action approach in presenting
a number of negatively valued social, physical, and aspirational consequences of adolescent
marijuana
use.
Certain positively valued consequences of
nonuse
also poitrayed, particularly by
drawing upon principles from social-learning and peer cluster theories. One PSA, which used
principles from all three theories, shows a group of five White and African-American teenage
girls interacting socially in a family room with a television on in the background. On the screen
is a white male authority figure ranting (in "Reefer
Madness"
style) about the evils of marijuana.
One girl, in exasperation, clicks off
the
television with the remote and
says,
"Are they still trying
to feed us that junk about weed?" She then proceeds to roll a
joint.
One of her friends replies,
"Maybe you need to listen, girl. Maybe you won't become an addict on the streets; but like my
cousin Derek, he sure got hung up on it." She goes on to say that "you can stop caring about

EfTective Mass Media Strategies for Drug Abuse Prevention Campaigns 39
things—like school, like your friends." Another friend chimes in with "and girls let me tell ya—it
can sure mess up your lungs." This is followed by a pregnant pause in which the first girl stops
rolling the joint, then huffs, "Well, I can see Vm in the wrong place," and gets up and leaves
the room. During a closing message board containing the words, "When you know more about
marijuana, you learn not to use it," we hear a buzz of conversation from the remaining members
of the group, unintelligible except for "She's headed for a whole lot of trouble."
This spot combines theory of reasoned action and peer cluster theory perspectives by pre-
senting certain negatively valued consequences, such as physical and psychological dependence,
loss of focus on schoolwork, and the social consequences of coming under fire from your peer
group (including social isolation). Social learning theory is incorporated as the members of the
group model ways in which to argue against marijuana use with peers. From a SENTAR perspec-
tive,
the PSA also contributes novelty and creativity (through the bizarre and attention-getting
"Reefer Madness" harangue by the authority figure on the
TV
screen, the rarely portrayed frank
peer-group discussion of marijuana, and unusual camera angles and movement); drama (in the
realistic "slice-of-life" social interaction); complexity (there is a great deal going on visually and
auditorily in the room); ambiguity/suspense (concerning what action the girl rolling
the
joint will
take at the end of the PSA, after the suspenseful pause); and stimulus intensity (through amplified
natural sounds from the room—dropped remote control striking the coffee table, the rustle of
paper as the joint is rolled, and other distinctive sounds—in the absence of a music track).
As this spot shows, incorporating
HSV
characteristics into a prevention message often has
more to do with how message arguments are presented rather than with the content of the persuasive
elements themselves (although content can also be involved, as in depicting graphic physical
consequences of heroin use). The important principle here is that the drug prevention practitioner
should use those theoretical schemes deemed most appropriate and effective for a particular
audience when developing persuasive messages, but should target the messages at high-sensation
seekers using the principles we have described. In other words, we should pay attention to the
communication needs of our audiences, particularly when those needs are tied so closely to the
behaviors we wish to prevent.
Extensions to Nonmedia Settings
Although our primary focus here is on media campaigns, the flexibility of the SENTAR approach
allows it to be extended to nonmedia settings as well, such as school-based prevention programs
or to multiple-modality interventions involving media, school, and community channels. It may
be extended as well to risky behaviors other than ding use. An example is provided by a SENTAR
school-based intervention project directed at
HIV
prevention and alcohol abuse among adolescents,
currently being conducted in two midwestem cities with funding from the National Institute on
Alcohol Abuse and Alcoholism. This study draws on theory and data from our media research on
sensation seeking and message design to adapt
a
nationally respected classroom-based curriculum,
Reducing the Risk, to make it more appealing to higher sensation seekers and impulsive decision
makers.
Although the content of the curriculum was left essentially unchanged, its format was altered
to add trigger films to enhance interest in topics for discussion. Talk-show formats were used for
other discussions, with video cameras placed in the hands of student participants and proceedings
videotaped and played back for further discussion. Contests were held for best role-plays, and
small prizes were awarded. The intervention also included greater participation in instruction by
group leaders chosen from the classroom and trained for their roles and training of teachers for

40 Philip Palmgreen
AND
Lewis Donohew
the revised format of classroom instruction. All programs used formative research in their devel-
opment, with focus and reaction groups chosen from students similar to those in the intervention
classes participating in development and testing of the classroom programs. In one of the cities,
a radio campaign was developed following SENTAR principles and used to prime audiences for
the classroom instruction. The radio campaign used paid and unpaid spots placed in programs
popular with high sensation seekers. Results showed significantly greater gains in knowledge
and on a number of the efficacy variables and significantly lower onset of sexual activity among
members of the primary target group (high-sensation seekers) receiving the curriculum than in
groups receiving other or no organized curricula.
This offers further evidence that it is vital in any prevention intervention aimed at risky
behaviors to pay close attention to the sensation-related communication needs and preferences of
the target population, no matter what delivery channels are used. The design of the
HIV
study also
illustrates how a prevention framework based on these sensation needs can be used to integrate
both media and nonmedia strategies in a theoretically coherent fashion. Such coherence, rarely
achieved in prevention practice, is greatly facilitated on one very important level by following the
approaches we have recommended in this chapter.
REFERENCES
Atkin,
C. K., & Freimuth, V. (1989). Formative evaluation research in campaign design. In R. E. Rice & C. K. Atkin
(Eds.),
Public
communication
campaigns (pp. 131-150). Newberry Park, CA: Sage.
Backer,
T.
E. (1990). Comparative synthesis of
mass
media health behavior campaigns. Knowledge: Creation, Diffusion,
Utilization, 110), 315-329.
Backer, T. E., Rogers, E. M., & Sopory, R (1992). Designing health communication campaigns:
What
works? Newbury
Park, CA: Sage.
Bardo,
M.
T., Donohew, R. L., & Harrington, N. G., (1996). Psychobiology of novelty seeking
and
drug seeking behavior.
Behavioural Brain
Research,
77, 23-43.
Bardo,
M. T., & Mueller, C. W. (1991). Sensation seeking and drug abuse prevention from a biological perspective.
In L. Donohew, H. E. Sypher, & W. J. Bukoski (Eds.), Persuasive communication and drug abuse prevention
(pp.
209-226). Hillsdale, NJ: Lawrence Erlbaum.
Bardo,
M. T., Nieswander, J. L., & Pierce, R. C. (1989). Novelty-induced place preference behavior in rats: Effects of
opiate and dopaminergic drugs. Pharmacology, Biochemistry, and
Behavior,
32, 683-689.
Bamea, Z., Teichman, M., & Rahav, G. (1992). Personality, cognitive, and interpersonal factors in adolescent substance
use:
A longitudinal test of
an
integrative model. Journal of
Youth
and
Adolescence,
27(2), 187-201.
Bauman, K. E., LaPrelle. J., Brown, J. D., Koch, G. C, & Padgett, C. A. (1991). The influence of three mass media
campaigns on variables related to adolescent cigarette smoking: Results of
a
field experiment. American Journal of
Public
Health,
81(5), 597-604.
Beck, E. J., Donegan, C, Kenny, C, Cohen, C. S., Moss, V., Terry, R, Underbill, G S.. Jefferies, D. J., Pinching, A. J.,
Miller, D. L.,
Harris,
J.
R. W.,
&
Cunningham, D. G. (1990). Update on HlV-testing at
a
London sexually transmitted
disease clinic: Long-term impact of the AIDS media campaigns. Genitourinary Medicine, 66, 142-147.
Black, G. S. (1991). Changing attitudes toward drug use: The effects of advertising. In L. Donohew, H. E. Sypher, &
W.
J.
Bukoski (Eds.), Persuasive
communication
and drug abuse prevention (pp. 157-191). Hillsdale,
NJ:
Lawrence
Erlbaum.
Block, L. G., Morwitz, V. G., & Sen, S. K. (1996). Does anti-drug advertising work? Proceedings of the marketing and
public policy
conference,
Washington, DC.
Bums,
D. M. (1994). Use of
media
in tobacco control
programs.
American
Journal
of Preventive
Medicine,
/(?(Suppl. 1),
3-7.
Cacioppo,
J.
X, Petty, R. E., Kao,
C.
F.,
&
Rodriguez, R. (1986). Central
and
peripheral
routes
to
persuasion:
An individual
difference perspective. Journal of personality and social psychology, 51, 1032-1043.
Caspi,
A.,
Dickson,
D.,
Dickson, N.,
Harrington,
H.,
Langley,
J.,
Moffitt,
T.
E.,
&
Silva, P A. (1997). Personality differences
predict health-risk behaviors in young adulthood: Evidence from a longitudinal study. Journal of Personality and
Social
Psychology,
73, 1052-1063.

Effective Mass Media Strategies for Drug Abuse Prevention Campaigns 41
Caspi, A., & Silva, P. A. (1995). Temperamental qualities at age 3 predict personality traits in young adulthood: Longi-
tudinal evidence from a birth cohort. Child
Development,
66, 486-498.
Clayton, R. R., Cattarello, A., & Walden, K. R (1991). Sensation seeking
as a
potential mediating variable for school-based
prevention intervention: A two-year follow-up of
DARE.
Health
Communication,
3(4), 229-239.
Donohew, L. (1988). Effects of
drug
abuse message styles: Final
report.
(Grant DA03462). Rockville, MD: National
Institute on Drug Abuse.
Donohew, L. (1990). Public health campaigns: Individual message strategies and a model. In E. B. Ray & L. Donohew
(Eds.),
Communication
and
health:
Systems and applications. Hillsdale, NJ: Lawrence Erlbaum.
Donohew, L., Lorch, E. P., & Palmgreen, P. (1991). Sensation seeking and targeting of televised anti-drug PSAS.
In L. Donohew, H. E. Sypher, & W. J. Bukoski (Eds.), Persuasive communication and drug abuse prevention
(pp.
209-226). Hillsdale, NJ: Lawrence Erlbaum.
Everett, M. W., & Palmgreen, P. (1995). Influences of sensation seeking, message sensation value, and program context
on effectiveness of anticocaine public service announcements. Health
Communication,
1, 225-248.
Flay, B. R. (1987). Mass media and smoking cessation: A critical review. American Journal of
Public
Health, 77(2),
153-160.
Flay, B. R., & Sobel, J. L. (1983). The role of mass media in preventing adolescent substance abuse. NIDA Research
Monograph
47.
In T.
J.
Glynn,
C.
G. Leukefeld,
&
J.
P.
Ludford
(Eds.),
Preventing
adolescent
drug
abuse:
Intervention
strategies (pp. 5-35). Rockville, MD: National Institute on Drug Abuse.
Flora, J. A., Maccoby, N., & Farquhar, J. W. (1989). Communication campaigns to prevent cardiovascular disease: The
Stanford conununity studies. In R. E. Rice & C. K. Atkins (Eds.), Public
communication
campaigns (pp. 233-252).
Newbury Park, CA: Sage.
Flora,
J.
A., Maibach,
E.
W.,
&
Maccoby, N. (1989). The role of media across four levels of health promotion intervention.
Annual
Review of Public
Health,
10, 181-201.
Flynn, B. S., Worden, J. K., Seeker-Walker, R. H., Badger, M. S., Geller, B. M., & Costanza, M. C. (1992). Prevention of
cigarette smoking through mass media intervention and school
programs.
American Journal
of Public
Health,
82(6),
827-834.
Flynn, B. S., Worden, J. K., Seeker-Walker, R. H., Badger, G. J., & Geller, B. M. (1995). Cigarette smoking prevention
effects of mass media and school interventions targeted to gender and age groups. Journal of Health
Education,
26,
45-51.
Fulker, D. W., Eysenck, H. J., & Zuckerman, M. (1980). A genetic and environmental analysis of sensation seeking.
Journal
of Research on
Personality,
14, 261-281.
Gantz, W., Fitzmaurice, M., & Yoo, E. (1990). Seat belt campaigns and buckling up: Do the media make a difference?
Health
Communication,
2, 1-12.
Hanmiond, S. L., Freimuth,
V.
S., & Morrison,
W.
(1987). The gatekeeping funnel: Tracking a major PSA campaign from
distribution through gatekeepers to target audience. Health Education
Quarterly,
14(2), 153-166.
Hu,
T., Sung, H., & Keeler,
T.
E.
(1995). Reducing cigarette consumption in California: Tobacco taxes vs.
an
anti-smoking
media campaign. American
Journal
of Public
Health,
85, 1218-1222.
Huba, G. J., Newcomb, M. D., & Bentler, P. M. (1981). Comparison of canonical correlation and interbattery factor
analysis on sensation seeking and drug use domains. Applied Psychological
Measurement,
5, 291-306.
Kilpatrick, D. G., Sutker, P. B., & Smith, A. D. (1976). Deviant drug and alcohol use: The role of anxiety, sensation
seeking, and other personality variables. In M. Zuckerman & C. D. Spielberger (Eds.),
Emotions
and
anxiety:
New
concepts, methods, and applications (pp. 247-278). Hillsdale, NJ: Lawrence Eribaum.
Klein, J. D., Brown, J. D., Walsh-Childers, K., Oliveri, J., Porter, C, & Dykers, C. (1993). Adolescents' risky behavior
and mass media use. Pediatrics, 92(1),
24-31.
Krishnamurthi, L., Narayan, J., &
Raj,
S.
P.
(1986). Intervention analysis of
a
field
experiment to assess the buildup effect
of advertising. Journal of
Marketing
Research,
23, 337-45.
Lorch, E. P., Palmgreen, P., Donohew, L., Helm, D., Baer, S. A., & Dsilva, M. U. (1994). Program context, sensation
seeking, and attention to televised anti-drug public service announcements.
Human Communication
Research,
20(3),
390-412.
Maibach, E., & Parrott, R. L. (1995). Designing health messages: Approaches from communication theory and public
health
practice. Thousand Oaks, CA: Sage.
Masse, L. C, & Tremblay, R. E. (1997). Behavior of boys in kindergarten and the onset of substance abuse during
adolescence. Archives of General
Psychiatry,
54, 62-68.
McAlister, A., Ramirez, A. G.. Galavotti, C,
&
Gallion, K.
J.
(1989). Anti-smoking
campaigns:
Progress in the application
of social learning theory. In R. E. Rice & C. K. Atkin (Eds.), Public communication campaigns (pp. 291-307),
Newbury Park, CA: Sage.

42 Philip Palmgreen
AND
Lewis Donoliew
Murry, J. P., Jr., Stam, A., & Lastovicka, J. L. (1993). Evaluating an anti-drinking and driving advertising campaign with
a sample survey and time series intervention analysis. Journal of
the
American Statistical Association,88, 50-56.
Murry,
J.
P.«
Jr.,
Stam
A.,
&
Lastovicka,
J.
L.
(1996). Paid-versus donated-media strategies for public service announcement
campaigns. Public Opinion Quarterly, 60, 1-29.
Newcomb, M. D., & McGee, L. (1989). Adolescent alcohol use and other delinquent behaviors. Criminal Justice and
Behavior, 76(3), 345-369.
Getting, E. R., & Beauvais, R. (1987). Peer cluster theory, socialization characteristics, and adolescent drug use: A path
analysis. Journal of
Consulting
Psychology, 34, 205-213.
Palmgreen, P., Donohew, L., Lorch,
E.
P., Hoyle, R. H., & Stephenson, M. T. (2001). Television campaigns and adolescent
marijuana use: Tests of sensation seeking targeting. American Journal of Public
Health,
91, 292-296.
Palmgreen, P., Donohew, L., Lorch, E. P., Rogus, M., Helm, D., & Grant N. (1991). Sensation seeking, message sensation
value, and drug use as mediators of PSA effectiveness. Health
Communication,
5(4), 217-227.
Palmgreen, R, Lorch, E. P, Donohew, R. L., Harrington, N. G., Dsilva, M., & Helm, D. (1995). Reaching at-risk
populations
in a mass
media
drug
abuse prevention
campaign:
Sensation seeking
as a
targeting variable. Co-published
simultaneously in Drugs and
Society,
8, 29-45, and in C. G. Leukefeld & R. R. Clayton (Eds.), Prevention practice
in substance abuse (pp. 29-45). Binghamton, NY: The Haworth Press.
Pedersen, W. (1991). Mental health, sensation seeking and drug use patterns: A longitudinal study. British Journal of
Addiction, 86, 195-204.
Perloff,
R. M. (1993).
The
dynamics of persuasion. Hillsdale, NJ: Lawrence Erlbaum.
Petty, R. E., Baker, S. M., & Gleicher, F. (1991). Attitudes and drug abuse prevention: Implications of the elaboration
likelihood model of persuasion. In L. Donohew, H. E. Sypher, and
W.
J. Bukoski (Eds.), Persuasive Communication
and Drug Abuse Prevention (pp. 71-90). Hillsdale, NJ: Lawrence Erlbaum.
Pierce,
J.
P.,
Anderson, D. M., Romano, R. M., Meissner, H. I.,
&
Odenkirchen,
J.
C. (1992). Promoting smoking cessation
in the United
States:
Effect of public service announcements on the cancer information service telephone
line.
Journal
of the National Cancer
Institute,
84(9), 677-683.
Pierce, J. P., Dwyer, T, Frape, G., Chapman, S., Chamberlain, A., & Burke, N. (1986). Evaluation of the Sydney "Quit
for Life" anti-smoking campaign.
The
Medical
Journal
of Australia, 144, 341-347.
Plant, M., & Plant, M. X. (1992). Risk-takers:
Alcohol,
drugs, sex and
youth.
London: Tavistock/Routledge.
Rice, R. E., & Atkin, C. K. (Eds., 1989). Public
communication
campaigns (2nd ed.) Newbury Park, CA: Sage.
Rogers, E. M. (1995).
Diffusion
of innovations. New York: Free Press.
Rogers, E. M., & Storey, J. D. (1987). Communication campaigns. In C. R. Berger & S. H. Chaffee (Eds.), Handbook of
communication
science (pp. 817-846). Newbury Park, CA: Sage.
Romer, D. (1994). Using mass media to reduce adolescent involvement in drug trafhcking. Pediatrics, 93, 1073-1077.
Rosenstock, I. M. (1990). The health belief model: Explaining health behavior through expectancies. In K. Glanz, F. M.
Lewis, & B. K. Rimer (Eds.), Health behavior and health education: Theory, research, and practice (pp. 39-62).
San Francisco: Jossey-Bass.
Ross,
J. D. C, & Scott, G. R. (1993). The association between HIV media campaigns and number of patients coming
forward for HIV antibody testing. Genitourinary
Medicine,
69, 193-195.
Schilling, R. R, & McAlister, A. L. (1990). Preventing drug use in adolescents through media interventions. Journal of
Consulting
and Clinical
Psychology,
58, 416-424.
Segal, B., Huba, G. J., & Singer, J. L. (1980). Drugs, daydreaming, and personality: A study of college youth. Hillsdale,
NJ: Lawrence Erlbaum.
Shelley,
J.
M., Irwig, L. M., Simpson,
J.
M., & Macaskill,
P.
(1991). Evaluation of a mass-media-led campaign to increase
Pap smear screening. Health Education
Research,
6(3), 267-277.
Siegel, M., & Biener, L. (2000). The impact of an antismoking media campaign on progression to established smoking:
Results of
a
longitudinal youth study. American Journal of Public
Health,
90(3), 380-86.
Slater, M. D. (1996). Theory
and
method in health audience segmentation.
Journal
of Health
Communication,
1,267-283.
Teichman,
M.,
Bamea, Z.,
&
Rahav,
G.
(1989). Sensation seeking,
state and
trait anxiety, and depressive mood
in
adolescent
substance users.
The
International
Journal
of
the
Addictions, 24(2), 87-99.
Tellegen, A. (1982). Brief manual for
the
Multidimensional Personality
Questionnaire.
University of Minnesota.
Thombs, D. L., Beck, K. H., Mahoney, C. A., Bromley, M. D., & Bezon, K. M. (1994). Social context, sensation seeking,
and teen-age alcohol abuse. Journal of School
Health,
64, 73-79.
Wallack,
L.
(1989). Mass communication and health
promotion:
A critical perspective. In R. E. Rice
&
C.
K. Atkin (Eds.),
Public communications campaigns (pp. 353-367). Newbury Park, CA: Sage.
Webb,
J. A., Baer, ?. E., McLaughlin, R. J., McKelvey, R. S., & Caid, C. D. (1991). Risk factors and their relation
to initiation of alcohol use among early adolescents. Journal of
the
American Academy of Child and Adolescent
Psychiatry, 30(4), 563-568.

Effective Mass Media Strategies for Drug Abuse Prevention Campaigns 43
Zastowny, T. R.* Adams« E. H., Black, G. S., Lawton, K. B., & Wilder, A. L. (1993). Sociodemographic and attitudinal
correlates of alcohol and other drug use among children and adolescents: Analysis of
a
large-scale attitude tracking
study. Journal of Psychoactive Drugs, 25(3), 223-237.
Zinrnierman,
R.
S., & Vemberg,
D.
(1994). Models of preventive health
behavior:
Comparison, critique,
and
meta-analysis.
Advances in Medical
Sociology,
4, 45-^7,
Zuckerman, M. (1979).
Sensation
seeking:
Beyond
the
optimal level of
arousal.
Hillsdale,
NJ:
Lawrence Erlbaum.
Zuckerman, M. (Ed., 1983). Biological bases of sensation
seeking,
impulsivity, and
anxiety.
Hillsdale, NJ: Lawrence
Erlbaum.
Zuckerman, M. (1986).
Sensation seeking
and
the endogenous
deficit
theory
of drug
abuse.
Monograph 74 NIDA Research
Monograph Series (pp. 59-70). Rockville, MD: National Institute on Drug Abuse.
Zuckerman, M. (1988). Behavior and biology: research in sensation seeking and reactions to the media. In L. Donohew,
H. Sypher,
& T.
Higgins (Eds.),
Communication,
social
cognition,
and
affect
(pp.
173-194). Hillsdale,
NJ:
Lawrence
Erlbaum.
Zuckerman, M. (1990). The psychobiology of sensation seeking. Journal of Personality, 58(1), 313-345.
Zuckerman, M. (1994). Behavioral
expression
and biosocial bases of sensation
seeking.
New
York:
Cambridge University
Press.