Association for Consumer Research Conference 2009: Session Proposal
“Underpinnings of Risky Behavior: Non-health Motives for Health-related Behaviors”
Chairs: Merrie Brucks, The University of Arizona and Paul M. Connell, Stony Brook University
Discussant: Punam Anand Keller, Dartmouth University
Titles of Presentations (Presenting author in bold):
The Use of Hypocrisy to Motivate Health Attitude and Behavior Change
Jeff Stone, The University of Arizona
Nicholas C. Fernandez, The University of Arizona
Children’s Ascribed Motivations for Smoking Elicited by Projective Questioning
Merrie Brucks, The University of Arizona
Paul M. Connell, Stony Brook University
Dan Freeman, University of Delaware
Symbolic Interactionism and Adolescent Reactions to Cigarette Advertisements
Connie Pechmann, University of California, Irvine
Dante Pirouz, University of California, Irvine
Todd Pezzuti, University of California, Irvine
Session Proposal
Consumers often know of risks to themselves, but fail to act in ways to reduce these risks
(Verplanken and Wood 2006; Thaler and Sunstein 2008). In their comprehensive review of the
literature on health risk perceptions literature published in the Handbook of Consumer
Psychology, Menon, Raghubir, and Nidhi Agrawal (2008) persuasively argue that there is a need
to identify “antecedents other than cognitive belief-based ones.” They identify a wealth of
motivational, affective, individual, contextual, and disease factors that have been studied and
how they relate these to consumer outcomes. Within the specific category of motivation, they
discuss theoretical work on self-control, self-positivity, and social desirability. The papers in this
proposed session share a focus on motivational antecedents for behaviors, but are guided by
theoretical frameworks that have been under-explored in the health context: hypocrisy theory,
associative processing of others’ motivations, and symbolic interactionism.
This session brings together rigorous research relevant to understanding consumers’
motivations for engaging in desirable and non-desirable social outcomes. More broadly, the
authors in these studies take novel theoretical and methodological approaches within the context
of motivation, persuasion, and consumer attitudes and behavior. Bundling these papers together
is intended to stimulate discussion and to explore ideas for future research, perhaps beyond the
health context. We will build in time to discuss each paper immediately after it has been
presented. The audience for this session would likely include researchers interested in the self,
motivation, persuasion, transformative consumer research, and public policy topics.
The first paper is co-authored by Jeff Stone (Associate Professor of Social Psychology at
the University of Arizona), who has published numerous studies on cognitive dissonance in highimpact journals such as Journal of Personality and Social Psychology and Journal of
Experimental Social Psychology, and his doctoral student Nicholas Fernandez, also at the
University of Arizona. Their work examines the use of inducing cognitive dissonance via
hypocrisy to achieve desirable health outcomes. The second paper is co-authored by Merrie
Brucks (Professor of Marketing at the University of Arizona), Paul Connell (Assistant Professor
of Marketing at Stony Brook University) and Dan Freeman (Associate Professor of Marketing at
University of Delaware). In this paper, the authors find that children ascribe motivations to
smoke or not to smoke at a very early age, even though they cannot articulate the reasoning
behind these motivations. The final paper is co-authored by Connie Pechmann (Professor of
Marketing at the University of California, Irvine), Dante Pirouz (Doctoral Student at the
University of California, Irvine), and Todd Pezzuti (Doctoral Student at the University of
California, Irvine). Across three experiments, the authors find that when teens are exposed to
advertisements featuring young adult models, they actually have higher intentions to smoke than
when exposed to advertisements featuring other teens, as teens see cigarettes as a means of
communicating an adult identity.
Extended Abstracts
The Use of Hypocrisy to Motivate Consumer Health Behavior Change
Jeff Stone and Nicholas C. Fernandez
This presentation examines the use of the hypocrisy strategy as a social-marketing tool
for changing consumer health behaviors. Feelings of hypocrisy occur when people make a public
statement about the importance of a target health behavior, such as using condoms to prevent
sexually transmitted diseases like AIDS (Stone et al. 1994, 1997), quitting smoking (Peterson et
al 2008), or using sunscreen to reduce the risk for skin cancer (Fernandez et al 2009). By itself,
the advocacy is consistent with prevailing attitudes beliefs about the issue, and does not cause
discomfort. However, when people are then made mindful that they themselves have not
performed the behavior regularly in the past, the discrepancy between their advocacy and past
behavior causes the discomfort associated with cognitive dissonance. To reduce their discomfort,
hypocrites become motivated to "practice what they preach" and take the necessary steps toward
bringing their own health behavior into line with their “preaching” about the importance of the
standards for good health.
A recent review of the hypocrisy literature (Stone and Fernandez 2008) shows that there
are over 20 studies of the effect of hypocrisy on motivating consumer behavior change in the
domains of health, the environment and the community. The results of these studies indicate that
following hypocrisy, people are most motivated to perform the target behavior when they
publically advocate the target behavior and then are privately made mindful of past recent
failures to perform the behavior. Studies also indicate that the hypocrisy strategy operates
effectively to modify behavior in non-Western cultures (Takaku 2001, 2006).
Recent empirical research focuses on changing behaviors related to the risk for cancer
(Fernandez et al 2009). A new line of study examines how much "mindfulness" of past failures is
necessary to motivate behavioral change following hypocrisy. According to Festinger (1957),
the magnitude of dissonance is highest when more inconsistent than consistent cognitions are
present in memory. In the case of hypocrisy, this implies that after advocating the target health
behavior, recalling many past failures will cause more dissonance and more behavior change.
However, recent research on the role of self-validation in ease-of-retrieval processes (Tormala et
al 2007) suggests that when advocates are asked to recall past instances of when they failed to
perform the behavior, they may also recruit examples of when they successfully performed the
behavior, especially when they are motivated and have the ability to think carefully about the
past (i.e., high elaboration). The “self-validation” process predicts that if advocates carefully
recall both failures and successes, it could balance the ratio of inconsistent to consistent
cognitions, which would reduce the level of dissonance and need to change behavior following
hypocrisy. This leads to the counter-intuitive prediction that when advocates think carefully as
they recall past failures to perform the target health behavior, recalling fewer past failures may
reduce the number of successes that are also recalled, such that recalling fewer past failures will
cause more dissonance and more behavior change. Thus, it was predicted that under high
elaboration conditions, when advocates were asked to recall many past failures to perform a
health behavior, the self-validation process would reduce the magnitude of dissonance and the
motivation to change behavior. However, carefully recalling few past failures would reduce the
self-validation process and cause more dissonance and behavioral change following hypocrisy.
In contrast, it was hypothesized that when they are not highly motivated to think about
past failures (i.e., low elaboration), advocates will focus primarily on the number of failures
recalled without recruiting other relevant information (e.g., successes). As a result, under low
elaboration, recalling many past failures will induce more dissonance and behavioral change
following hypocrisy than recalling few past failures. In summary, we predicted that under high
elaboration, advocates who think about few past failures will exhibit more behavior change, but
under low elaboration, advocates who think about many past failures will exhibit more behavior
change.
In a 2 (Elaboration: High vs. low) X 2 (Past failures: 2 vs. 8) experimental design, 90
female college students wrote a brief persuasive message for other college students about the
importance of using sunscreen to reduce the risk for skin cancer. All were then asked to report
past failures to use sunscreen. To manipulate high elaboration (Tormala, Brinol, and Petty 2007),
half were told that only a few people were being asked to report information about past failures
to use sunscreen; those in the low elaboration condition were told that thousands of people were
reporting information about past failures to use sunscreen. Then half were asked to recall 2 past
failures to use sunscreen whereas the other half were asked to recall 8 past failures to use
sunscreen. All were then provided an opportunity to order a sample of sunscreen from an
independent national organization, with the percentage that acquired sunscreen as the primary
dependent measure.
The results revealed the predicted elaboration X past recall interaction. As hypothesized,
under conditions of high elaboration, significantly more participants (82%) acquired a sample of
sunscreen when they were asked to recall 2 past failures compared to those asked to recall 8 past
failures. In contrast, under low elaboration, significantly more participants (68%) acquired a
sample of sunscreen when asked to recall 8 past failures compared to those asked to recall 2 past
failures (39%). Overall, the pattern supports the hypothesis that in hypocrisy, the effect of
recalling many past failures on behavior change is a function of how carefully advocates think
about their past behavior. Potential mediators of this finding and other future directions for
research will be discussed.
References
Fernandez, Nicholas C., Jeff Stone, Joel Cooper, Toni Cascio, and Michael Hogg (2009),
“Vicarious Hypocrisy: Using Attitude Bolstering to Restore the Integrity of the Ingroup,”
working paper, The University of Arizona.
Festinger, Leon (1957), A Theory of Cognitive Dissonance, Evanston, IL: Row, Peterson.
Peterson, Alexandra A., Graeme A. Haynes, and James M. Olson (2008), “Self-esteem
Differences in the Effects of Hypocrisy Induction on Behavioral Intentions in the
Health Domain,” Journal of Personality, 76 (2), 305-22.
Stone, Jeff, Andreew W. Wiegand, Joel Cooper, and Elliott Aronson (1997), “When
Exemplification Fails: Hypocrisy and the Motive for Self-integrity, Journal of
Personality and Social Psychology, 72 (1), 54-65.
Stone, Jeff, Elliot Aronson, A. Lauren Crain, Matthew P. Winslow, and Carrie B. Fried,
(1994), “Inducing Hypocrisy as a Means of Encouraging Young Adults to Use Condoms.
Personality and Social Psychology Bulletin, 20 (1), 116-28.
Stone, Jeff and Nicholas C. Fernandez (2008), “How Behavior Influences Attitudes:
Cognitive Dissonance Processes,” in W. Crano & R. Prislin (Eds). Attitudes and Attitude
Change , 314-34, New York: Psychology Press.
Takaku, Seiji (2001), “The Effects of Apology and Perspective Taking on Interpersonal
Forgiveness: a Dissonance-Attribution Model of Interpersonal Forgiveness,” Journal of
Social Psychology, 141 (4), 494-508.
Takaku, Seiji (2006), “Reducing Road Rage: An Application of the Dissonance-Attribution
Model of Interpersonal Forgiveness,” Journal of Applied Social Psychology, 36 (10),
2362-78.
Tormala, Zakary L., Pablo Brinol, and Richard E. Petty (2007), “Multiple Roles for Source
Credibility under High Elaboration: It’s all in the Timing,” Social Cognition, 25 (4), 53652.
Children’s Ascribed Motivations for Smoking Elicited by Projective Questioning
Merrie Brucks, Paul M. Connell, and Dan Freeman
Critics of increased regulation on tobacco advertising and promotion seen by children argue
that such regulation would result in small, if any, effects in reducing initiation of tobacco use
among minors. After all, the argument goes, eight year olds do not smoke, so cigarette
advertising is personally irrelevant to them. But this perspective assumes that advertisements
must be actively processed to be effective. In contrast, we note that considerable research has
documented advertising effects on attitude, even under very low involvement conditions (e.g.,
peripheral processing, mere familiarity, evaluative conditioning). Taking this perspective, we
argue that exposure to cigarette advertisements and media images are likely to be processed and
encoded into memory despite the lack of individual salience of tobacco promotional activity in
childhood.
Our reasoning is consistent with the associative processing model of memory, which is one
of the two memory systems proposed by Smith and DeCoster (2000). Associative processing
operates preconsciously and automatically (Bargh 1994) and is learned over many experiences.
Hence, individuals are typically not aware of the processing itself, but only the results of it.
Because tobacco advertising and media images are not likely to be self-relevant to children, we
argue that they are processed through such an associative mode.
Furthermore, such associative processing of smoking imagery may produce effects that
extend beyond childhood. This is because bias correction is best facilitated when individuals
possess both the ability and motivation to reconsider their attitudes). If positive psychosocial
associations are learned at a nonconscious and automatic level, then the individual will not likely
recognize his or her own biases held in memory, thereby inhibiting the ability to metacognitively
reconsider attitudes (Petty and Briñol 2008).
The goal of this study was to aid in generating a theoretical model for the psychological
processes involved in children’s learning of lifestyle associations with adult-themed products.
Given this objective, we pursued our empirical research in the spirit of discovery-oriented
research (Wells 1993). Because we suspected that children’s lifestyle associations might have
been learned implicitly, and because social desirability biases are a threat to validity in substance
use research, we employed projective interviewing techniques.
We conducted 271 projective interviews with second and fifth grade children from three
different elementary schools. Two varieties of projective stimuli were used to elicit participant
responses: print advertisements and pictures of people who have various personal and lifestyle
characteristics. Each child saw two ads for cigarettes, which were embedded in a series of five
ads (including three unrelated products). For each ad, children were asked to choose select
pictures of specific people who might be likely or unlikely to use that product. Each child was
probed with follow up questions to reveal the motivations he or she attributed to these people.
In the presentation, we will show: (1) the three images that were most strongly associated
with smoking, as these images were attributed with multiple motives for smoking; (2) the four
images that were also associated with smoking, and were attributed with one or two motives for
smoking; (3) the six images that were strongly associated with non-smoking, and were attributed
with motives for non-smoking; and (4) four images that were inconsistently associated with
smoking, and were attributed with motives for both smoking and non-smoking. Typically, the
second graders had difficulty in articulating these motives, often with responses such as “s/he
just looks like s/he would smoke.” Therefore, the emergent themes were drawn largely from the
fifth grade interview dataset. Nevertheless, the second graders often made many of the same
lifestyle associations as the fifth graders. Chi-square analysis of the pattern of picture selections
indicated that they associated the same images with smoking and non-smoking as the fifth
graders did.
Qualitative data analysis revealed three broad areas of motives attributed to the characters
in the images: social motives, esteem motives, and relaxation motives. Themes within the social
motive included smoking for fun in social situations and smoking to impress others, whereas
themes within the esteem motive primarily included issues with weight and thinness. Themes
within the relaxation motive primarily included needing to escape one’s troubles or smoking for
leisure. Finally, for one of the images, general inactivity or lack of motivation in general was
associated with smoking.
We argue that the similarity of lifestyle picture selections and attributed motivations between
the second and fifth graders, in combination with the non-verbal nature of the second graders’
associations, suggests that children did not purposefully and thoughtfully develop them. This
pattern of data is consistent with the associative processing model of memory, in which
advertisements, media images, and personal observations are encoded into memory
preconsciously through associative processing.
References
Bargh, John A. (1994), “The Four Horsemen of Automaticity: Awareness, Intention, Efficiency,
and Control in Social Cognition,” in Handbook of Social Cognition: Basic Processes,
eds. Robert S. Wyer and Thomas K. Srull, Hillsdale, NY: Erlbaum.
Brinol, Pablo and Richard E. Petty (2008), “Persuasion: From Single to Multiple to
Metacognitive Processes,” Perspectives on Psychological Science, 3 (2), 137-47.
Smith, Elliott R. and Jamie DeCoster (2000), “Dual-process Models in Social and Cognitive
Psychology: Conceptual Integration and Links to Underlying Memory Systems,”
Personality and Social Psychology Review, 4 (2), 108-31.
Wells, William D. (1993), “Discovery-oriented Consumer Research,” Journal of Consumer
Research, 19 (4), 489-504.
Symbolic Interactionism and Adolescent Reactions to Cigarette Advertisements
Connie Pechmann, Dante Pirouz, and Todd Pezzuti
Does restricting the age of models in cigarette advertising to 25 years or older really work
in deterring adolescents from smoking? Health researchers and advocates have stressed the
importance of using only adult models in cigarette ads because younger models might entice
adolescents to believe that smoking is for them. As a result, there have been a number of
regulatory attempts to specify the minimum age of models in cigarette ads (Richards, Tye, and
Fischer 1996). For example, the Voluntary Cigarette Advertising and Promotion Code, which
tobacco manufacturers claim to use as a guideline, states that "any models used in advertising be
and appear to be older than 25” (The Tobacco Institute 1990, p. 1). Previous academic literature
has focused on the perception of model age in cigarette advertising and has shown that people
perceive many cigarette ad models to be younger than 25 (Arnett 2005; Barbeau et al. 1998;
Mazis et al. 1992). However, these studies have not linked model age to persuasion nor have
they studied the effect of manipulating model age.
Similarity-based theories, such as social comparison (Festinger 1954) and social identity
(Tajfel and Turner 1986), predict that consumers are more influenced by similar others (White
and Dahl 2006). A major implication is that ad models should mirror the target audience and
teens are more likely to be persuaded by models their own age than by older or younger ones.
However, these theories have focused on normative influence and have not addressed
aspirational or symbolic factors. On the other hand, symbolic interactionism focuses on how
people interpret, act towards, and give meaning to symbolic objects, events and situations around
them (Blumer 1969; Sandstrom, Martin, and Fine 2003; Solomon 1983; Stryker 1980). Symbols
acquire meaning through socialization and products that function as symbols can carry potent
information about an individual’s social role and status (Reynolds and Herman-Kinney 2003).
Age is a powerful indicator of social role and status (Holstein and Gubrium 2003).
Across life stages, products are used as symbols to signify to the self and others the transition to
a new developmental level (Peterson and Peters 1983; Piacentini and Mailer 2004). Many
adolescents have a fragile and unstable self-concept which leads them to aspire to look and act
older (Barker and Galambos 2005; Galambos, Turner, and Tilton-Weaver 2005). Thus, products
with specific social meaning can be used to facilitate the transition but the imagery and
symbolism surrounding the product must appropriately signal adulthood (Berger and Rand 2008;
Solomon 1983).
Cigarettes belong to a unique class of products that are a key identity signal of adulthood
(Eckert 1983). Unlike other publicly consumed products, such as music and clothing, smoking
symbolizes the adult world and is taboo for adolescents and children in most cultures (Rugkasa
et al. 2001). Thus many adolescents seek out conspicuous symbols of adulthood, such as
cigarettes, to reinforce their pseudomature identity and help them bridge the gap to adulthood
(Eckert 2007; Galambos and Tilton-Weaver 2000; Noble and Walker 1997). For adolescents, this
may have an unexpected influence on their response to cigarette advertising featuring models
their own age.
Our theoretical framework predicts that for an adult-signaling product category like
cigarettes, young adult models would influence adolescents more positively than same age
models. It also predicts that teenage or same age models might, in fact, boomerang. However, for
other product categories such as clothing, teen models would be more influential than young
adults. These effects would be mediated by what the model’s age communicates about the
product, for example, age appropriateness. Theory would also predict that because adolescence is
a period of transition, only teens would be susceptible to the teenage model boomerang effect but
young adults would not be.
In a series of experiments, we investigated how altering the age of models used in
cigarette advertising affects whether adolescents are drawn to -- or deterred from -- smoking. We
pretested over 1000 model images to choose a set of 16 advertising models that were equally
attractive but varied on gender (2 levels), ethnicity (2 levels) and age (4 levels: child, teen, young
adult and middle age). We then conducted the first study with 221 ninth grade students who rated
the model images. We manipulated whether the models held cigarettes or not. As hypothesized,
we found that adolescent participants identified most with teen and young adult models,
suggesting that they are most likely to be influenced by these age groups.
A second study investigated how cigarette model age might affect persuasion.
Participants were 479 ninth grade students who viewed a mock-up magazine containing cigarette
ads or matched control ads. The ad models’ age was manipulated. The teenage cigarette models
boomeranged, lowering intent to smoke. The young adult cigarette models increased intent to
smoke, while child and middle-aged cigarette models had no effect on intent. We repeated this
experiment using 284 college students who were 18-21 years old. As predicted, the cigarette
models had null effects because young adults do not need cigarettes to signal an adult identity.
A third study examined whether the age effects were moderated by product category and also
explored the underlying process by identifying mediators. Participants were 278 ninth graders
who were shown a magazine with cigarette ads or designer t-shirt ads. The ad models’ age was
manipulated. As posited, participants perceived cigarettes (vs. designer t-shirts) as more for
public display, more reflective of “who the person is,” and more “for adults.” The cigarette ads
also affected both product category (primary) and brand specific (secondary) demand.
Specifically, teen versus young adult cigarette ad models lowered cigarette product and brand
intent, as hypothesized. The effect of ad model age on intent to use the cigarette product and
brand was at least partially mediated by age appropriateness perceptions. Teen (versus young
adult) cigarette models decreased perceptions that cigarettes were an appropriate teenage
product, which then lowered intent to use cigarettes.
References
Arnett, Jeffrey Jensen (2005), “Talk Is Cheap: The Tobacco Companies’ Violations of Their
Own Cigarette Advertising Code,” Journal of Health Communication: International
Perspectives, 10 (5), 419-31.
Barbeau, Elizabeth M, William DeJong, Douglas M. Brugge, and William M. Rand (1998),
“Does Cigarette Print Advertising Adhere to the Tobacco Institute’s Voluntary
Advertising and Promotion Code? An Assessment,” Journal of Public Health Policy, 19
(4), 473-88.
Barker, Erin T. and Nancy L. Galambos (2005), "Adolescents' Implicit Theories of Maturity,"
Journal of Adolescent Research, 20 (5), 557-76.
Berger, Jonah and Lindsay Rand (2008), “Shifting Signals to Help Health: Using Identity
Signaling to Reduce Risky Health Behaviors,” Journal of Consumer Research, 35 (3),
509-18.
Blumer, Herbert (1969), Symbolic Interactionism: Perspective and Method, Berkeley: University
of California Press.
Eckert, Penelope (1983), “Beyond the Statistics of Adolescent Smoking,” American Journal of
Public Health, 73 (4), 439-41.
Festinger, Leon (1954), “A Theory of Social Comparison Processes,” Human Relations, 7 (2),
117-40.
Galambos, Nancy L. and Lauree C. Tilton-Weaver (2000), "Adolescents' Psychosocial Maturity,
Problem Behavior, and Subjective Age: In Search of the Adultoid," Applied
Developmental Science, 4 (4), 178-92.
Galambos, Nancy L., Pamela K. Turner, and Lauree C. Tilton-Weaver (2005), “Chronological
and Subjective Age in Emerging Adulthood,” Journal of Adolescent Research, 20 (5),
538-56.
Holstein, James A. and Jaber F. Gubrium (2003), “A Constructionist Analytics for Social
Problems,” in James A. Holdstein and Gale Miller (eds.) Challenges and Choices,
Hawthorne, NY: De Gruyter.
Mazis, Michael B., Debra Jones Ringold, Elgin S. Perry, and Daniel W. Denman (1992),
“Perceived Age and Attractiveness of Models in Cigarette Advertisements,” Journal of
Marketing, 56 (1), 22-37.
Noble, Charles H. and Beth a. Walker (1997), “Exploring the Relationships Among Liminal
Transitions, Symbolic Consumption, and the Extended Self,” Psychology and Marketing,
14 (1), 29-47.
Peterson, Gary W. and David F. Peters (1983), “Adolescents’ Construction of Social Reality:
The Impact of Television and Peers,” Youth & Society, 15 (1), 67-85.
Piacentini, Maria and Greg Mailer (2006), “Symbolic Consumption in Teenagers’ Clothing
Choices,” Journal of Consumer Behaviour, 3 (3), 251-62.
Reynolds, Larry and Nancy J. Herman-Kinney (2003), “Taking Stock: A Handbook for
Symbolic Interactionsists,” in Larry Reynolds and Nancy J. Herman-Kinney (eds.)
Handbook of Symbolic Interactionsim, Lanham, MD: Rowman and Littlefield.
Richards, J.W., J.B. Tye, and P.M. Fischer (1996), “The Tobacco Industry’s Code of Advertising
the United States: Myth and Reality,” Tobacco Control, 5 (4), 295-391.
Rugkasa, Jorun, Barbara Knox, Julie Sittlington, Orla Kennedy, Margaret P. Treacy, and Pilar
Santos (2001), ”Anxious Adults Vs. Cool Children: Children’s Views on Smoking and
Addiction,” Social Science & Medicine, 53 (5), 593-602.
Sandstrom, Kent L., Daniel D. Martin, and Gary Alan Fine (2003), Symbols, Selves, and Social
Reality: A Symbolic Interactionist Approach to Social Psychology and Sociology, Los
Angeles, CA: Roxbury.
Solomon, Michael R. (1983), "The Role of Products as Social Stimuli: A Symbolic Interaction
Perspective," Journal of Consumer Research, 10 (3), 319-29.
Stryker, Sheldon (1980), Symbolic Interactionism: A Social Structural Version, Upper Saddle
River, NJ:Benjamin Cummings.
The Tobacco Institute 1990, Voluntary Cigarette Advertising and Promotion Code,
www.tobaccoinstitute.com.
Tajfel, Henri and Turner, John C. (1986), “The Social Identity Theory of Inter-group Behavior,”
In S. Worchel and L. W. Austin (eds.), Psychology of Intergroup Relations, Chicago:
Nelson-Hall
White, Katherine and Darren W. Dahl (2006), “To Be or Not Be: The Influence of Dissociative
Reference Groups on Consumer Preferences,” Journal of Consumer Psychology, 16 (4),
404-14.