Are we marketing the right message: Can kids “just say ‘no’” to smoking?

ARE WE MARKETING THE RIGHT MESSAGE: CAN KIDS"JUST SAY 'NO' " TO SMOKING?

KATHLEEN K. REARDON, STEVE SUSSMAN, AND BRIAN R. FLAY

fiuo hundred sixty-eight adolescents from two Southern California schools partici-pated in a study designed to assess the compliance-resisting strategies children usewhen rejecting pressure to smoke. Participants were asked to indicate how they wouldrespond to eight hypothetical situations varying in relationship (friend vs. acquain-tance), number of people present (dyad vs. group), and amount of pressure (first vs.second exertion of pressure). The influence of risk for future smoking on compliance-resisting strategies was also assessed. Results indicate that contrary to the popularview, children do not always "just say no. " Their strategies vary with the conditionsmentioned above. Moreover, high risk adolescents appear to have difficulty usingstrong forms of rejection. These results suggest that researchers should consider theinfluences of situational variations and risk status for future smoking when developinginterventions to teach children to resist smoking and other forms of substanceabuse.

"dearth of interest in persuasion research" characteristic of the 1970sX. (Roloff & Miller, 1980) has given way to a reemergence of persuasion study

among communication experts characterized, in large part, by the study ofcompliance-gaining and compliance-resisting. These studies have focused on individ-ual, relational, and situational determinants of the strategies people use to persuadeothers to comply and those used to resist such efforts.

The vast majority of compliance studies have utilized college student respondentswith some exceptions such as work by Burgoon et al. (1987) on compliance in thephysician-patient relationship, and by Fitzpatrick et al. (1987) on compliance-gaining choices of marital partners. There appears to be a readiness on the part ofcommunication researchers to extend compliance research to nonstudent populationsso that the generalizability of prior findings might be explored and knowledge ofcompliance processes within specific populations with unique compliance problemsmight be advanced (see Boster, 1988).

Consistent with this emerging path, the primary concerns of this study aretwofold: (1) extension of prior research assessing the effects of situational andrelational factors on compliance-resisting strategy selections, and more specifically,(2) examination of these factors within the context of adolescent resistance topressure by peers to smoke.

Overview of Relevant Compliance-Gaining and Compliance-Resisting Research

Communication researchers have focused more attention on compliance-gainingthan compliance-resisting. While the two differ in terms of goals, certain similaritiesexist. Both involve persuasion and, thus, encourage others to do or accept somethingthey might not do or accept otherwise. Both require the selection of strategies suited

Kathleen K. Reardon is Associate Professor of Business Communication and Preventive Medicine,University of Southern California. Steve Sussman is Assistant Professor of Preventive Medicine,university of Southern California. This investigation was supported by a NIH postdoctoral researchgrant to Kathleen K. Reardon and by a grant from the National Institute of Drug Abuse # DA03468 toBrian R. Flay and C. Anderson fohnson, and a NCI grant §CA44907 to Steve Sussman.

COMMUNICATION MONOGRAPHS, Volume 56, December 1989

308 COMMUNICATION MONOGRAPHS

to the situation at hand, and both can result in success or failure, depending upon theskill of the actors involved.

Compliance-gaining research has advanced beyond compliance-resisting researchas evidenced by the attention it has received in major communication journals.Considerable controversy exists, hov^ever, over the best methods for the study ofcompliance-gaining: the strategy selection procedure in which respondents arepresented with hypothetical situations and asked to select among strategies providedto them by the researcher; the strategy generation method, in which respondentsprovide their own strategies rather than selecting from ones provided by theresearcher; and behavioral observation methods, during which participants areobserved as they actually engage in compliance-gaining or compliance-resistingbehaviors. There is also considerable debate as to which factors, including situation,social appropriateness, and personal predispositions, account for strategy selections,and the best research questions to ask about compliance-gaining. One such question,for example, is "Do reports of strategy selection have any relationship to actualcompliance-gaining behavior and, if not, does that make worthless studies that do notinvolve the observation of actual behavior?" (See Boster, 1988; Burleson & Wilson1988; Dillard, 1988; Hunter, 1988; Seibold, 1988).

While these battles continue to be waged, the findings of several studies providesome insights relevant to the study of compliance-resisting. For example, mostcompliance-gaining research suggests that people prefer prosocial to antisocialstrategies (Boster, Stiff, & Reynolds, 1985; Miller, Boster, Roloff, & Seibold, 1977;Roloff, 1976; Roloff & Barnicott, 1978). As noted by deTurck (1985), most studies ofcompliance-gaining are limited to initial persuasive attempts, so it is impossible todetermine at what point during a sequence of persuasive attempts such positiveapproaches might be abandoned. Other research demonstrates that such factors asgender and the nature of the relationship influence the compliance-gainer's use ofantisocial strategies (deTurck, 1985; Witteman & Fitzpatrick, 1986). Researchusing college students, which typically results in a preference for prosocial strategies,may not be easily generalized to other populations. The nature of the relationship,predispositions of the compliance-gainer, and situational factors are likely toinfluence preferences with respect to prosocial and antisocial strategies (see Boster,1988).

Unlike compliance-gaining research, compliance-resisting research is scarce.There are some exceptions. McLaughlin, Cody and Robey (1980) proposed theconcept "compliance-resisting strategies," which they define as "verbal attempts by atarget to gain the agent's acceptance of the target's unwillingness to comply" (p. 749).McLaughlin et al. proposed a four category typology of compliance-resistingstrategies: (1) non-negotiation, (2) identity management, (3) justifying, and (4)negotiation. Their work suggests that selection of a strategy type depends upon therisk associated with compliance-resisting. Work by McQuillen, Higginbotham, andCummings (1984) assessed how children of different ages (first, fourth, and ninthgraders) resist persuasion. Three communication situations were generated varyingin age/status of the persuader (mother, peer, younger sibling). Using the typologydeveloped by McLaughlin et al. (1980), they found that children of different agesdiffer significantly in their use of compliance-resisting strategies. Non-negotiationdecreased with age. This finding is consistent with the McLaughlin et al. (1980)claim that non-negotiation is a high-risk strategy. As children become more aware of

MARKETING THE RIGHT MESSAGE 309

the feelings of others, it appears that they tend to use more prosocial strategies. Twoprosocial strategies, justification and identity management, increased with age. Thisresearch also indicates that compliance-resisting strategies vary according to the typeof strategy an individual is asked to resist, and the type of compliance-gainingagent.

Other research relevant to an understanding of compliance-resisting strategychoices focuses on rationales or accounts for resistance. Following a proceduredeveloped by Howie-Day {\911), McQuillen and Higginbotham (1986) studiedchildren's rationales for compliance-resisting. They found significant differences inrespondents' use of rationales as a function of age, type of request, and respondent'sgender. Among tenth graders, there were differences between male and femalerationales. For example, males used their highest level of rationales to respond tosimple requests, and their lowest level rationales to resist altruistic requests. Femalesused their highest level rationales in response to incentive requests and their lowestlevel rationales with simple requests. McQuillen and Higginbotham conclude thatmales at this age appear to honor, via increased levels of consideration, thosestrategies that are direct, whereas females are more attentive to strategies thatemphasize social convention over those that stress overt control. They also found thatthe extent to which respondents took the perspective of the persuadée when justifyingtheir resistance varied as a function of the type of compliance-gaining strategyemployed by the agent. They did not, however, find that level of perspective-takingvaried as a function of the type of compliance-gaining agent.

McLaughlin, Cody and O'Hair (1983), McLaughlin, Cody and Rosenstein(1983), and Cody and McLaughlin (1985) have conducted research on accounts.They define accounts as the way in which "failure events" are managed in socialinteraction. Their focus is on how accounts work to affect repair in threatenedcommunication episodes and/or relationships. Since compliance-resistance has thepotential to threaten conversations and relationships, such work is highly relevant.For example, McLaughlin, Cody and O'Hair (1983), and Cody and McLaughlin(1985) found that the nature of accounts, whether mitigating (e.g., excuses,concession and justifications) or aggravating (e.g., refusals), is inñuenced by thenature of the reproach. Cody and McLaughlin reported that aggravated reproacheslead to aggravating accounts more consistently than mitigating reproaches lead tomitigating accounts. They explained the latter as possibly the result of theaccounter's resenting attempts by the reproacher to understand why the accounterengaged in the act in question, since he or she feels falsely accused of it in the firstplace.

Research on rationales and accounts examines the reasoning processes thataccompany communication behaviors. Both compliance-gaining and compliance-resisting often involve encouraging or discouraging the target with reasoning. Thustheory and research regarding how people reason about their efforts to gaincompliance or about their resistance of efforts to gain it is also relevant to thisinvestigation. According to Reardon's (1981, 1987) ACE model, reasoned attemptsto persuade may be classified into three broad categories: appeals to appropriateness,consistency and effectiveness. She also argues that preferences among these threecategories vary with relationship and situational factors. Two studies provideconsensual validation for the ACE Model. In a study of tobacco use, Friedman et al.(1985) found that adolescents' refusal responses fall into three broad categories

^^^ COMMUNICATION MONOGRAPHS

similar to those in Reardon's ACE Model: fear of disapproval (appropriateness)respondent's personal convictions (consistency), and fear of the effects of smokinè(effectiveness). Newman (1984) found that these same categories could be used todescribe the reasons adolescents gave in support of smoking. Neither study providedinformation concerning the effects of relationship and situational variations onpreferences among reasoning styles.

Compliance-Resisting In the Context of Adolescent Smoking

Having earlier made the case that the findings of compliance research usingcollege student respondents may not necessarily be generalizable to nonstudentpopulations, it seems reasonable to introduce the characteristics of the populationand situation under study that render it somewhat distinctive. Early-stage adolescentsmoking is largely a social activity. More than 50% of adolescents report smokingtheir first cigarette with a friend (Bewley et al., 1974; Biglan et al., 1984; Friedmanet al., 1985; Palmer, 1970). Friedman, Lichtenstein and Biglan (1985) found that89% of their adolescent respondents were in the presence of others when pressured tosmoke. Of these others, 50% were friends, 20% were acquaintances, and 11% weresiblings.

Recognition of the social nature of smoking has led to the development of a numberof school-based interventions focusing on skills for resisting social pressure. Theseinterventions are based on the presumption that young adolescents smoke primarilyas a result of social pressures to smoke, especially from peers. Social influenceinterventions have reduced the onset of smoking among young adolescents by anaverage of 50% (Flay et al., 1985). Teaching adolescents that cigarette smoking isharmful to health appears to be less effective than also teaching them what they cansay to peers who invite or pressure them to smoke (Flay, 1985; Thompson 1978-Tobler, 1987). ' '

Despite the obvious importance of training adolescents to resist smoking, littleattention has been given to what they are able to say to peers who pressure them tosmoke. Interventions have been developed on the basis of what researchers believeadolescents should be able to say, rather than on the basis of what adolescents reportthey would say. Moreover, few studies have addressed the influence of relationshipand contextual variations on adolescents' strategies for rejecting smoking.

Hops et al. (1986) used the verbal refusals of adolescents to 26 audiotapedsituations involving social pressure to smoke in an assessment of the effectiveness ofan intervention designed to improve refusal skills. Their coding scheme containednine response categories: refusals, health facts, conciliatory/supportive, excuses/change of subject, withdrawal statements, assertive, aggressive, external conse-quences, and acceptance. They found that adolescents who were favorably disposedto smoking and who were current smokers did worse in terms of use of refusal skillstaught to them. They found a preference across respondents (control group andexperimental group) for excuses. The researchers suggest, however, that this may bedue to the fact that excuses are effective for turning down offers of cigarettes withoutalienating friends. They did not systematically vary the nature of the relationshipbetween the refuser and the person exerting pressure, so their conclusion is tentative.They did examine influences based on the sex of the offender and the level ofpressure exerted by the offer. The analyses revealed no significant effects

Schinke and Gilchrist (1983, 1984) and Gilchrist and Schinke (1984) tested the

MARKETING THE RIGHT MESSAGE 311

effectiveness of a skills training component of their smoking prevention program.They used videotaped role play assessments of refusal skills. Confederate peersoffered student participants a cigarette. Students who received refusal skills trainingperformed better on a composite index of behavioral ratings than did students whoreceived no program or information only without skills training. Using a similarprocedure, Rorhbach et al. (1987) found that training in refusal skills increasesability to resist pressure to smoke and self-efficacy regarding refusal.

The current absence of a clear understanding of what adolescents are likely to saywhen resisting pressure to smoke and how relationship and situational factorsinfluence their strategy choices weakens our ability to train adolescents adequately inresistance skills. From smoking cessation programs developed for adults, we havelearned that the strategy must suit the individual or type of individual and situation(e.g., Benfari, Okene, & Mclntyre, 1982). Yet, researchers often rely on their ownperceptions of adolescent interactions to determine the types of resistance skills to beincluded in school-based smoking interventions (Sussman et al, 1986). We need toknow more about what adolescents feel comfortable saying when they want to resistpressure to smoke or take drugs. For example, is it likely that adolescents will walkaway or just say no to smoking if their relationships with the people offering acigarette are important? Are high risk adolescents, who often do not adhere toconventional school norms, likely to adopt strategies advocated in that environmentby people who may not understand the pressures they experience daily ? Are differentresistance strategies likely to be used when there are several other people watching asthe adolescent resists smoking? Are there strategies better suited to females thanmales, and vice versa? Are some resistance strategies better for initial invitations tosmoke while others are suited to repeated pressure? These are some of the questionsthat should be answered if we are to develop effective training programs in resistanceskills.

The present study was undertaken to overcome these limitations and to arrive at amore precise understanding of the compliance-resisting strategies that adolescentsactually use when pressured by peers to smoke.

Research Hypotheses

The study had four related purposes: (1) to evaluate the influence of relationshiptype (friend vs. acquaintance) on preferences for compliance-resisting strategies; (2)to examine the role played by the number of people present in the compliance-gaining episode; (3) to evaluate the influence of increases in pressure on preferencesfor compliance-resisting strategies; (4) to assess the influence of risk status onpreferences for compliance-resisting strategies.

Boster (1988) has identified several ways in which compliance-gaining behaviormay be operationalized. Some of these are amount of verbal aggression, extent oflistener-adapted communication, and the qualitative category into which each of thestrategies fall. This study addresses the first and third of these: amount of verbalaggression and qualitative category of each, in this case, compliance-resistingstrategy.

Consistent with past research (deTurck, 1985; McLaughlin, Cody & Robey,1980; Miller, Boster, RolofT & Seibold, 1977; Witteman & Fitzpatrick, 1986), wepropose that the intensity of rejection should differ with variations in the relationshipof the agent of compliance-gaining and the résister and variations in the situation.

312 COMMUNICATION MONOGRAPHS

Specifically,

HI: There will be more intense rejection with acquaintances than friends.H2: There will be more intense rejection with a group than in a dyad.H3: There will be more intense rejection when pressured a second time than when initially

pressured.

With regard to categories into which strategies fall, the following hypotheses werederived from prior research and the proposal by deTurck (1985) that people have athreshold for repeated persuasion attempts beyond which they tend to preferantisocial to prosocial strategies:

H4: Antisocial strategies are more likely to be used with acquaintances than with friends.H5: Antisocial strategies are more likely to be used in groups than dyads.H6: Antisocial strategies are more likely to be used in second pressure rather than initial pressure

situations.

And in terms of the amount and nature of reasoning used in strategies for resistingcompliance, two research questions were advanced:

Rl: Under what conditions (friend vs. acquaintance, dyad vs. group, first pressure vs. second pressure)are resisters likely to use the most reasoning in their strategies?

Consistent with ACE Model predictions, we propose that the reasons used bychildren when resisting pressure to smoke can be categorized according to the ACEModel and thus pose the additional research question:

R2: How do relational and situational factors influence preferences for appropriateness, consistencyand effectiveness forms of reasoning?

One research question was also advanced with regard to the characteristics of therésister.

R3: Does the compliance-resister's risk status for smoking, gender and ethnicity influence the intensityof rejection?

METHODRespondents

Respondents were 268 seventh graders in two schools in the Monrovia SchoolDistrict in the greater Los Angeles area. Sample sizes from the schools wereapproximately equal (136 for school A and 132 for school B). The balance of malesand females was similar for the two schools. They did differ somewhat in ethnicity.School A subjects were 24% Black, 51% White, 20% Hispanic and 5% Other (mostlyAsian). School B subjects were 4% Black, 52% White, 30% Hispanic, and 13% Other(mostly Asian). There were no significant differences between schools in terms of thecurrent smoking among respondents and self-reports of the likelihood that theywould ever smoke in the future. Six percent were weekly smokers, 50% had tried atleast one cigarette, and 10% stated positive intentions to smoke in the future. Therespondents had not as yet received a health education program. Thus, theirresponses reñect what adolescents without training do or say they will do.

Procedure

As part of a larger investigation, respondents received a one-page set of questionsassessing how they would resist smoking in eight different situations (see Table 1).

MARKETING T H E RIGHT MESSAGE 313

TABLE 1

EIGHT SITUATIONS INVOLVING PRESSURE TO SMOKE

To determine how variations in context, relationship, and amount of pressure influence selections of non-compliance strat^ies, we asked the following questions at pretest and posttest:

1. Imagine that you and your best friend are somewhere alone. Your best friend is smoking a cigaretteand is pressuring you to smoke too. You don't want to smoke. What would you say to your bestfriend?

2. What if he/she really wanted you to smoke, and kept pressuring you? What would you say then?3. Imagine that you and an acquaintance are somewhere alone. Your acquaintance is smoking a ciga-

rette and is pressuring you to smoke, too. You don't want to smoke. What would you say to your ac-quaintance?

4. What if he/she really wanted you to smoke, and kept pressuring you. What would you say then?5. Imagine that you and your best friend are somewhere with a lot of people who are smoking. Your best

friend is smoking a cigarette and is pressuring you to smoke too. You don't want to smoke. Whatwould you say to your best friend?

7. Imagine that you and an acquaintance are somewhere with a lot of other people who are smoking.Your acquaintance is smoking a cigarette, and is pressuring you to smoke too. You don't want tosmoke. What would you say to your acquaintance?

8. What if he/she really wanted you to smoke and kept pressuring you? What would you say then?

These situations varied in terms of the nature of relationship (friend or acquain-tance), amount of pressure exerted on the respondent to smoke (first or secondrequest), and number of people present (dyad or group).

The eight situations were presented to all respondents. Four of the eight situationswere second pressure situations. For example, after respondents were asked toimagine what they would say if pressured by a friend to smoke, they were asked whatthey would say if pressured by this friend a second time. Our purpose was todetermine whether increased pressure alters patterns of compliance-resistance withsmoking requests. It was imperative that first and second pressure situations beplaced together, leaving four pairs of situations each with a first pressure and secondpressure condition.

Rather than divide the four pairs of situations among respondent groups so thateach respondent only received one pair of situations (e.g., friend not in a grouppressuring the respondent to smoke, and this friend pressuring him or her a secondtime), we were interested in determining whether the same respondent wouldpresent different answers across situations. Therefore each respondent received allfour pairs of situations. Respondents were not blinded to our purposes of comparingdifferent situations. This approach could have placed demands on them to respond,for example, differently in a given situation (e.g., friend in a group) than they haddone in response to a situation presented earlier (e.g., friend alone). To alleviatepotential concerns about repeated measures placing such demands on respondents,we conducted a small replication study. A sample of 22 seventh-grade studentssimilar to those in the larger study were randomly assigned to two conditions. Halfreceived the friend alone first pressure scenario, and half received the friend in agroup first pressure scenario. Based on our data from the larger study in which eachsituation was presented to each respondent, we expected friends alone to use fewerrejecting responses than in a group context. Using the same coding scheme for level ofrejection as that employed in the larger study, coder agreement was 88 percent. Themean rejection rate for friend alone was lower than for friend in a group (¿(20) =1 98, p < .06); respective means were 4.09, SD .79 and 4.64, SD .48, as in the largerinvestigation. Achieving even a marginally significant result with such a small

314 COMMUNICATION MONOGRAPHS

TABLE 2

CODER INSTRUCTIONS

We are interested in coding a number of statements made by children in response to being offered a ciga-rette. Please read each statement carefully, and give it a number from 1 to 5, corresponding to the followingscale:

1 ~ highly accepting2 - mildly accepting3 - neither accepting nor rejecting4 - mildly rejecting5 - highly rejecting

Catt^ory 1:

Category 2:

Category 3:

Catt^ory 4:

Category 5:

To be categorized as a highly accepting statement, the child must accept the cigarette.Below are some samples of category one statements:

Sure.Give me one.Thanks.I'd say okay I'll smoke.I'd smoke it.

To be in the mildly accepting category, the child must indicate the acceptance may oc-cur or that it will occur to a limited degree. For example:

Maybe later.I might.Just one puff.Well, maybe this time.

To be in the third category of neither accepting nor rejecting, the child must avoid re-sponding to the request. Some examples are:

I wish you wouldn't smoke.Let's go somewhere else.I'll tell your mother.Let's get out of here.

A statement belongs to category 4 if the child rejects the offer politely or with an expla-nation or excuse that softens the rejection. For example:

No, thanks.No, I'm sorry.I can't.I already had one today.Please don't pressure me.

Place a statement in category 5 if it leaves no room for discussion. The child clearly andemphatically rejects the cigarette or person.

Get lost,Bug off.No, and leave.I'd hit him.I'd grab the cigarette and crush it.I'd leave.

If you are unsure about a statement, provide us with the categories in which it might be placed. For exam-pie, if you are unsure about whether a statement belongs in mildly accepting or highly accepting, writedown '/à, meaning one or two. If there is absolutely no category suited to a particular statement, give it aCATEGORY 6 rating.

For practice, please code the following statements:

1. Absolutely not.2. Bye. I'm splitting.

MARKETING T H E RIGHT MESSAGE 315

3. My mom is calling me.4. Bug on".5. Don't be a fool.6. Don't make me smoke, please.7. Forget it.8. I'd go home.9. I don't know what I'd do.

10. I don't smoke, thank you.11. I would change the conversation, or leave.12. I think I hear my mother calling me.13. I smoke some.14. She ain't my kind of girl.15. Stick it in your ear.16. Why do you smoke?17. No, I don't want to destroy my health.

sample size, using a conservative design, suggests that the demand characteristics ofthe repeated measures design were not sufficient, if they existed at all, to bias theresults.

Respondents were also asked three questions concerning their current smokingstatus and their intentions to smoke in the distant future. These questions were usedto determine risk status for smoking.

Instrument Content, Coding and Scale Reliabilities

Two naive coders rated the noncompliance responses according to a scale of level ofrejection (Table 2). These same coders placed the noncompliance responses incategories of noncompliance strategy type (Table 3). This typology was developed bythe investigators. It is a more extensive typology than that developed by Hops et al.(1986) in their analysis of refusal strategies.

Coders were told that the examples provided were to be used as guides during thetraining session. The examples provided were not to be considered exhaustive orrestrictive. There was room for disagreement. In fact there was some initialdisagreement with the examples given which was resolved by discussion among thecoders. They took notes during the training discussion and were urged to use theirjudgment when coding the responses. Intercoder agreement was acceptable for bothlevel of rejection (Pearson r = .85) and strategy type (Scott's Pi = .83).

The third coding procedure was derived from the ACE Model. Coding accordingto the ACE Model required two steps: (1) selection of responses that involvedreasoning, and (2) assignment of each reasoned strategy to one of the three categories.Reasoned responses were those that included some explanation for rejection. Twonaive coders obtained high intercoder agreement (90 percent for both tasks).

To assess level of risk for smoking, three rating scale items were used. Theyassessed how many cigarettes the respondents smoked at the time, if they thoughtthey would ever smoke, and the likelihood that they would ever ask anyone to letthem try a cigarette. Cronbach's alpha was .87 for these three items, indicatingadequate scale reliability. The items were summed to form an index of risk for futuresmoking.

316 COMMUNICATION MONOGRAPHS

TABLE 3

NONCOMPLIANCE STRATEGIES

1. Simple rejection: rejections without explanations.

No.No. Thank you.

2. Statement of typical, enduring behavior: statements about the usual behavior of the person being of-fered the cigarette.

I don't smoke.I don't smoke. I drink.I don't smoke, and never will.You know I don't smoke.I don't smoke anymore.

3. Statement of attitude or belief without reasons: statements which describe negative beliefs about smok-ing or expressions of want, need or desire.

I don't think you should smoke.I don't want to smoke.I don't feel like it.

4. Excuse: statements that provide a way out of having to be responsible for having to say "no."

I have a headache.I have allergies.I just had one.My mother is calling me.

5. Repetition: repeating the strategy used previously.

I'd say the same thing.I'd tell him again.I'd say it again.

6. Rejection of the person(s): statements rejecting the person offering the cigarette.

This group is not for me.Get away from me.Bug off.Stop bothering me.

7. Insult: statements in this category are meant to be offensive to the person offering the cigarette.

You're dumb.You're stupid.I can't believe you're like this.

8. Request: asking the person offering the cigarette not to do so.

Please don't make me smoke.Please stop pressuring me.Please don't bother me.

9. Accusation of motive: statements indicating why the other person smokes.

You just smoke to be with the group.You're just trying to be cool.You just do it to be in the group.

10. Offer alternative: statements which suggest an activity to replace smoking.

I want to do something else.Let's go to the park.I'd rather drink.

11. Hedging: statements that limit the amount of acceptance or postpone acceptance.

Only one puff.Not today.Maybe later.Not now, thanks.

MARKETING THE RIGHT MESSAGE 317

12. Walk away: statements about actually leaving or indicating that the person being offered the cigaretteis leaving.

I'd leave.I'd walk away.I'd say no and leave.Bye.

13. Warning: statements promising negative outcomes for the person offering the cigarette.

I'll call the police.I'll hit you.I'm getting ready to leave.I'll tell your mom.

14. Question: asking a question of the person offering the cigarette.

Why do you smoke?Why don't you leave me alone?Why are you smoking?Why are you pressuring me?

15. Appeal to relationship: statements about the relationship with the person offering the cigarette.

If you're my friend, you won't do this to me.I can be in this group without smoking.I guess I never knew you.I can't believe you'd do this.You're not my friend if you smoke.

16. Violence or anger.

I'd hit him.I'd break the cigarette.No. Damn it!I'd take it away from them.

17. Declaration or philosophy: statements about the general effects of smoking.

Smoking doesn't solve anything.Smoking is unhealthy.Smoking causes cancer.Smoking stinks.It isn't good for you.

18. Acceptance of cigarette.

Sure.Thanks.Give it to me.

19. Miscellaneous statements: ones that cannot be coded into one of the 18 categories above.

RESULTS

Intensity of Compliance-Resisting

To contrast intensity of compliance-resisting (dependent measure) across levels ofthe situation variables (i.e., friend vs. acquaintance, dyad vs. group, and first vs.second pressure attempt), responses were averaged across sets of four scenarios inwhich one level of the situation variable was present. Each situation level comparisonwas examined by repeated measures analysis of variance. We used this approach tokeep the presentation of the results straightforward and understandable while

3Î8 COMMUNICATION MONOGRAPHS

making sure the same response was represented in no more than one variable or levelof a two-level contrast.

In response to Research Question 3, the effects of school, gender, ethnicity, andsmoking risk were assessed. School, gender, and ethnicity did not have significantmain effects on intensity of compliance-resisting. Further, they did not interact withany of the situational variables (acquaintance vs. friend, dyad vs. group, firstpressure attempt vs. second pressure attempt). The smoking risk variable did have asignificant effect on intensity of compliance-resisting, although it did not interactwith any of the situational variables. Higher-risk respondents were lower inintensity of compliance-resisting than lower-risk respondents in all cases (Fs rangedfrom 5.6 to 6.8, all ps < .001). Therefore, risk was entered as a covariate in theremaining analyses.

The relationship between the person pressuring and the respondent (friend vs.acquaintance) had no significant effect on intensity of compliance-resisting. ThusHypothesis 1 was rejected. Hypothesis 2 was supported. The influence of thenumber of people present when being pressured to smoke was significant(F(l,215) = 5.01,/) < .02). Respondents reported less intense forms of compliance-resisting in a dyad than in a group. The influence of the number of pressure attemptswas also significant (F(l,215) = 4.33,p < .03), providing support for Hypothesis 3.Intensity of compliance-resisting was greater when respondents were pressured tosmoke a second time.

Subsequent analyses indicated no interactions among the situation variables. Forexample, all four friend vs. acquaintance comparisons were nonsignificant whethernested in (a) dyadic, first pressure attempt; (b) dyadic, second pressure attempt; (c)group, first pressure attempt; or (d) group, second pressure attempt.

Preferences for Strategy Types

Eighteen categories of compliance-resisting were used to assess whether situa-tional variations affect strategy preferences. Five strategies emerged as the mostpreferred across the eight situations: simple rejection, statement of typical behavior,expression of attitude/belief, rejection of the person, and walk away, for the fourdyad and four group situations. These accounted for 80% of all strategies used. Thosestrategies were retained for further analysis. Preferences among the five strategieswere compared by Chi square analyses (df = 1), indicating whether respondents didor did not use a strategy in one situation versus another. Since each situation wasrepresented in four of eight scenarios, the mention of a strategy type at least once (of apossible four times) was defined as "presence" of that response in the situation.

Respondents relied more often on simple rejection with acquaintances than withfriends (x = 63.3, p < .0001). Simple rejection was also the preferred strategy ingroup rather than dyadic situations (x^ = 59.1.,/? < .0001), and first pressure versussecond pressure situations (x^ = 65.1,/? < .0001). To the extent that simple rejectionis an antisocial strategy similar to walk away and rejection of the person (describedbelow), these analyses provide support for Hypotheses 4 and 5 but do not supportHypothesis 6.

Statement of typical behavior followed the same pattern of results as simplerejection (x^ = 37.2 for friend versus acquaintance, and 24.1 for first versus secondpressure).

Expression of attitude or belief was used more frequently in acquaintance than

MARKETING T H E RIGHT MESSAGE 319

TABLE 4

ACE CODING SCHEME (Reardon, 1981, 1986)

Only applies to strategies involving reasoning.

AppropriatenessI don't think you should smoke.It's wrong.My parents don't want me to.If you're my friend, you won't do this.I guess you're not my friend.It isn't cool to smoke.It's dumb.

ConsistencyI don't smoke.I never smoke.I don't need to be cool.I don't smoke and never will.I don't smoke just because others do.You may feel good about smoking, but I don't.

EffectivenessI don't want to die.My life is precious to me.I'd tell him what it does to you.They're bad for you.Don't you know it's bad for your health?Smoking isn't good for you.You can get cancer.

friend situations (x^ = 16.4, p < .0001), but not more frequently in groups thandyads. It followed a pattern similar to statement of typical behavior in that greateruse of expression of attitude occurred when pressured to smoke the first time thanwhen asked a second time (x^ = 22.3, p < .0001).

Both rejection of the person and walk away (two antisocial strategies that do notinvolve reasoning, and which threaten the future of the relationship) were used morefrequently with acquaintances than friends (x^=16.1 and 32.7, respectively;p < .0001) and more frequently when respondents were asked a second time toaccept a cigarette (x^ = 12.6 and 12.7 respectively;/? < .0001). Thus Hypotheses 4and 6 were supported. However, rejection of the person did not differ in amount ofuse in group versus dyad situations. Walk away occurred more frequently in a groupthan in a dyad (x^ = 14.5,/? < .0001), providing partial support for Hypothesis 5.

Reasoning Patterns

Using the ACE Model procedure outlined in Table 4, analyses were conducted todetermine whether preferences for appropriateness appeals, consistency appeals, andeffectiveness appeals are influenced by relationship and situational variations(Research Question 1). Table 5 provides percentages of reasoned responses for eachof the eight situations. Clearly, respondents relied more on reasoned rejection insituations in which they had not been pressured a second time (x^ = 15.34,/> < .01).The greatest amount of reasoning occurred in situation one—the first offer of acigarette in a dyadic interaction between friends. Overall, however, there was nosignificant difference between amount of reasoning used with friends and with

320 COMMUNICATION MONOGRAPHS

TABLE 5

PERCENTAGES OF REASONING BY SITUATION

Situation 1 (friend, dyad, first pressure) 42.9%Situation 2 (friend, dyad, seœnd pressure) 17^4%Situation 3 (acquaintance, dyad, first pressure) 28.6%Situation 4 (acquaintance, dyad, second pressure) 13.6%Situation 5 (friend, group, first pressure) 25.4%Situation 6 (friend, group, second pressure) 12.9%Situation? (acquaintance, group, first pressure) 21.6%Situation 8 (acquaintance, group, second pressure) 12.5%

acquaintances. Comparison of dyad versus group situations indicated greaterreliance on reasoning in dyads (x^ = 3.96,/? < .05).

Assessments of preferences for appropriateness, consistency, and effectivenessrejection strategies revealed patterns across and within situation types. Comparisonof friend versus acquaintance situations revealed a clear preference for consistencyand effectiveness appeals over appropriateness. Within category types, a significantdifference in preference was found for effectiveness. This approach is more preferredwith friends than acquaintances (x^ = 8.59, p < .01). Appropriateness was usedsomewhat more often with friends than with acquaintances (x^ = 3.48, ji? < .10). Nodifferences were obtained for consistency (see Table 6).

Comparisons between first pressure and second pressure situations producedsignificant differences across situations. Appropriateness was used significantly moreoften in first pressure situations than when respondents were pressured a second time(X = 10.13, d( = I, p < .01). The same pattern held for consistency (x^ = 41.0,p < .0001), and effectiveness (x^ = 12.5, p < .01).

Comparisons of dyad versus group situations indicated no significant differencesin reasoning pattern for appropriateness and consistency, although the latterapproached significance with a preference for consistency in dyads over groups. Foreffectiveness, a significant preference was observed. Effectiveness was favored as astrategy with friends rather than acquaintances (x^ = 7.73,/? < .01).

In all eight situations, consistency was the preferred strategy (all Chi squaressignificant at the .01 level or better).

DISCUSSION

The findings of this study indicate that adolescents' rejections of peer pressure tosmoke vary according to the nature of their relationship with the peer or peers,situational characteristics and the rejector's own smoking risk status. These findingshave important implications for the development of future adolescent smoking anddrug abuse school-based interventions.

Teaching cross-situational strategies such as the widely advocated "Just say no"may be insufficient to the needs of adolescents in a variety of relationships andsituations which they experience daily. For example, in terms of the intensity withwhich adolescents are willing to express rejection of peer pressure to smoke,situational factors appear to be important. This study reveals that rejection of greaterintensity occurs when adolescents are pressured a second time during the sameinteraction, when there is more than one person doing the pressuring, and when therejector is at low risk to become a smoker in the future.

MARKETING THE RIGHT MESSAGE 321

TABLE 6

ACE CATEGORY FREQUENCIES

FriendAcquaintance

First PressureSecond Pressure

DyadGroup

App.

5033

App.

5627

App.*

4934

Con.

125114

Con.

16970

Con.*

138101

Eff.*

10869

Eff.*

11265

Eff.*

10770

•Significant Chi Square at p < .01 or better.

"Just say no" emerged as a preferred strategy when adolescents were dealing withacquaintances rather than friends, and when they were in groups rather than dyadicsituations. "Just say no" appears to be a strategy reserved for less intimaterelationships and contexts. Adolescents favor statements about their typical behavior(e.g., I never smoke), and attitudinal statements (Smoking is disgusting) in initialpressure situations. They tend to reserve attitudinal statements for dyadic ratherthan group encounters. Overall, it appears that among the five most frequently usedstrategies, the two that involve some rationale for rejection—statement of typicalbehavior and attitudinal statements—are less frequently used in groups and whenpressured a second time.

Strategies that are threatening to relationships, such as walk away and rejection ofthe person, are used most often when pressure to smoke has been imposed on theadolescent a second time. Tolerance appears to drop when additional pressure isexerted and the need for a rationale for one's rejection is lessened.

The ACE Model analyses provided some insights into the use of strategycategories, specifically appropriateness (what is right or expected), consistency (whatI typically do or what I am typically like), and effectiveness (what outcomes arelikely). The findings of this investigation were consistent with those of Newman(1984) and Friedman, Lichtenstein, and Biglan (1985). Adolescent rejection strate-gies involving reasoning fall into three categories. This study also indicates thatreasoning is most frequent with friends in dyadic situations.

When reasoned responses were used, the major pattern that emerged was a clearpreference for consistency strategies. It appears that when adolescents express theirreasons for rejection of a cigarette, they generally prefer to appeal to their ownpersonal views. Their second preference is effectiveness, with appropriatenessappeals being used least, and primarily in situations in which only one attempt hasbeen made to obtain compliance. Although adolescents are frequently accused ofbeing concerned about the opinions of others, this study suggests that it is not to theseopinions that they generally turn when defending their decision to reject smoking.They appear to prefer expressions of personal beliefs and attitudes followed byknowledge of the negative consequences of smoking.

We might reasonably conclude from these findings that adolescent reasoning aboutsmoking is based more on personal convictions and knowledge of outcomes than on

322 COMMUNICATION MONOGRAPHS

appropriateness. Another, perhaps more plausible, explanation is that while appro-priateness may enter into their reasoning, adolescents do not express their views ofwhat is right or expected as a means of persuading peers. They prefer to be seen asfollowing their own personal rules rather than pleasing others. The fact thatappropriateness was not used as frequently as consistency or effectiveness may notindicate a disregard for what is right or expected, so much as a reluctance byadolescents to express such rules as reasons for their action. Telling peers aboutappropriate actions may be less socially acceptable than expressing one's ownconvictions or asserting desires to avoid illness and death. This explanation of thefindings is consistent with the widely held view that adolescents are attracted tosmoking in part because of its departure from what is appropriate for their aeegroup.

This study explored what teenagers say in simulated cases of compliance-resistance to peer pressure to smoke a cigarette. It did not identify which strategiesare relatively more or less effective in refusing tobacco offers. Future studies whichexamine how the offerer and résister interpret the meaning of various refusalstrategies may better assess the relationship between strategies that teenagers use andthose they should use to refuse tobacco offers most effectively.

This research extends earlier work on compliance-resisting to a situation impor-tant to improving the health of children. The findings suggest that adolescents,especially those at high risk, often do not "just say no." Like adults, they often wish topreserve relationships by providing justifications for their rejection. And like adults,they know that different situations require different responses. The more we knowabout their strategy preferences and the factors that influence them, the more capablewe will be of developing realistic and effective school-based interventions.

REFERENCES

Benfari, R.C., Ockene, J.K., & Mclntyre, K.M. (1982). Control of cigarette smoking from a psychologicalperspective. Annual Review of Public Health, i , 101 -128.

Bewley, B.R., Bland, J.M., & Harris, R. (1974). Factors associated with the starting of cigarette smoking byprimary school children. British foumal of Preventive Medicine, 28, 37-44.

Biglan, A., McConnell, S., Severson, H.H., Barry, J., & Ary, D. (1984). A situational analysis of adolescentsmoking, foumal of Behavioral Medicine, 7, 109-114.

J (g f f , , 109114.

Boster, F.J. (1988). Comments on the utility of compliance-gaining message selection tasks. Human CommunicationResearch, 15, 169-177.

Boster, F.J., Stiff, J.B., & Reynolds, R.A. (1985). Do persons respond differently to inductively-derived anddeductively-derived lists of compliance-gaining message strategies? A reply to Wiseman and Schenck-Hamlin.Western foumal of Speech Communication, 49, 177-187.

Burgoon, J.K., Pfau, M., Parrott, R., Birk, T., Cocker, R., & Burgoon, M. (1987). Relational communication,satisfaction, compliance-gaining strategies, and compliance in communication between physicians and patients.Communication Monographs, 54, 307-324.

Burleson, B.R., & Wilson, S.R. (1988). On the continued undesirability of item desirability: A reply to Boster,Hunter and Seibold. Human Communication Research, 15, 178-191.

DeTurck, M.A. (1985). A transactional analysis of compliance-gaining behavior: Effects of noncompliance,relational contexts and actors' gender. Human Communication Research, 12, 54-78.

Cody, M.J., & McLaughlin, M.L. (1985). Models for the sequential construction of accounting episodes:Situational and interactional constraints on message selection and evaluation. In R.L. Street, Jr. & J. Cappella(Eds.) Sequence and pattern in communicative behavior (pp. 50-69). London: Edward Arnold.

Dillard, J.P. (1988). Compliance-gaining message-selection: What is our dependent variable? CommunicationMonographs, 55, 162-183.

Flay, B.R. (1985). Psychosocial approaches to smoking prevention: A review of findings. Unpublished manuscript.University of Southern California.

Fitzpatrick, M.A. (1988) Between husbands and wives: Communications in marriage. Newbury Park: Sage.Friedman, L.S., Lichtenstein, E., & Biglan, A. (1985). Smoking onset among teens: An empirical analysis of initial

situations. Addictive Behaviors, 10, 1-13.

MARKETING T H E RIGHT MESSAGE 323

Gilchrist, L.D., & Schinke, S.P. (1984). Self-control skills for smoking prevention. In Engstrom, P.F., Anderson,P.N., & Mortenson, L.E. (Eds.). Advances in cancer œntrol, 1983 (pp. 125-130). New York: Allen R. Liss.

Gilchrist, L.D., Snow, W.H., Lodish, D., & Schinke, S.P. (1985). The relationship of cognitive and behavioral skillsto adolescent tobacco smoking. Journal of School Health, 55,132-134.

Green, D.E. (1975). Teenage cigarette smoking in the United States: 1968, 1970, 1972, and 1974. In Smoking andhealth: Proceedings of the Third World Conference (pp. 375-380). Washington, D.C.

Hops, H., Weissman, W., Bigian, A., Thompson, R., Faller, C , & Severson, H.H. (1986). A taped situation test ofcigarette refusal skill among adolescents. Behavioral Assessment, 8,145-154.

Howie-Day, A.M. (1977). Me taper suasion: The development of reasoning about persuasive strategies. Unpublisheddoctoral dissertation. University of Minnesota.

Hunter, J.E. (1988). Failure of the social desirability response set hypothesis. Human Communication Research, 15,162-168.

Hurd, D.D., Johnson, C.A., Pechacek, T., Bast, L.P., Jacobs, D.R., & Leupker, R.V. (1980). Prevention ofcigarette smoking in seventh grade students. Journal of Behavioral Medicine, 3,15-28.

Kniskern, J., Bigian, A., Lichtenstein, E., Ary, D., & Bavry, J. (1983). Paper modeling effects in the smokingbehavior of teenage smokers. Addictive Behaviors, 8,129-132.

Luepker, R.V., Johnson, C.A., Murray, D.M., & Pechacek, T.F. (1983). Prevention of cigarette smoking: Threeyears follow-up of an educational program for youth. Journal of Behavioral Medicine, 6, 53-62.

McLaughlin, M.L., Cody, M.J., & O'Hair, H.D. (1983). The management of failure events: Some contextualdeterminants of accounting behavior. Human Communication Research, 9, 208-24.

McLaughlin, M.L., Cody, M.J., & Robey, C.S. (1980). Situational influences on the selection of strategies to resistcompliance-gaining attempts. Human Communication Research, 7, 14-36.

McLaughlin, M.L., Cody, M.J., & Rosenstein, N.E. (1983). Account sequences in conversations between strangers.Communication Monographs, 50, 102-25.

McQuillen, J.S., Higginbotham, D.C, & Cummings, M.C. (1984). Compliance-resisting behaviors: The effects ofage, agent, and types of requests. In R.N. Bostrom (Ed.), Communication yearbook 8 (pp. 747-762). BeverlyHills: Sage.

McQuillen, J.S., & Higginbotham, D.C. (1986). Children's reasoning about compliance-resisting behavior. In M.McLaughlin (Ed.), Communication yearbook 9 (pp. 673-690). Beverly Hills: Sage.

Miller, G.R., Boster, F.J., Roloff, M.E., & Seibold, D.R. (1977). Compliance-gaining message strategies: Atypology and some findings concerning efl"ects of situational differences. Communication Monographs, 44,37-51.

Newman, I.M. (1984). Capturing the energy of peer pressure: Insights from a longitudinal study of adolescentcigarette smoking. Journal of School Health, 54, 146-148.

Palmer, A.B. (1970). Some variables contributing to the onset of cigarette smoking in junior high school students.Social Science Medicine, 4, 359-366.

Reardon, K.K. (1981). Persuasion: Theory and context. Beverly Hills: Sage.Reardon, K.K. (1987). Interpersonal communication: Where minds meet. Belmont, CA: Wadsworth.Rohrbach, L.A., Graham, J.W., Hansen, W.B., Flay, B.R., & Johnson, CA. (1987). Evaluation of resistance skills

training using multitrait-multimethod role play skill assessments. Health Education Research, 2, 401-407.Roloff, M.E. (1976). Communication strategies, relationships and relational change. In G.R. Miller (Ed.),

Explorations in interpersonal communication (pp. 173-195). Beverly Hills: Sage.Roloff, M.E., & Barnicott, E.E. (1978). The situational use of pro- and anti-social compliance-gaining strategies by

high and low Machiavellians. In B.D. Ruben (Ed.), Communication yearbook 2. New Brunswick, NJ:Transaction Books.

Roloff, M.E., & Miller, G.R. (1980). Persuasion: New directions in theory and research. Beverly Hills: Sage.Salomon, G., Stein, Y., Eisenberg, S., & Klein, L. (1984). Adolescent smokers and nonsmokers: Profiles and their

changing structure. Preventive Medicine, 13, 446-461.Schinke, S.P., & Blythe, B.J. (1981). Cognitive-behavioral prevention of children's smoking. Child Behavior

Therapy, 3, 25-42.Schinke, S.P., & Gilchrist, L.D. (1983). Primary prevention of tobacco smoking. Journal of School Health, 53,

416-419.Schinke, S.P., & Gilchrist, L.D. (1984). Preventing cigarette smoking with youth. Journal of Primary Prevention, 5,

48-56.Seibold, D.R. (1988). A response to "Item desirability in compliance-gaining research." Human Communication

Research, 75,152-161.Syme, S.L., & Alcalay, R. (1982). Control of cigarette smoking from a social perspective. Annual Review of Public

Health, 3, \19-\99.Sussman, S., Brannon, B.R., Flay, B.R., Gleason, L., Senor, S., Sobel, D.F., Hansen, W.B., & Johnson, C.A_

(1986). The television, school and family smoking prevention/cessation project II. Formative evaluation oftelevision segments by teenagers and parents—implications for parental involvement in drug education. HealthEducation Research: Theory and Practice, 1,185-194.

Sussman, S., Dent, C.W., Mestel-Rauch, J., Johnson, CA., Hensen, W.B., & Flay, B.R. (In press). Adolescent

COMMUNICATION MONOGRAPHS

^ ' ^ programs, 1969-1976. American Journal of Public Health, 68,

« ^ 5 f P^vention programs: Quantitative outcome results

CoZm^ f t Ï ' p ? ; ^ r " " ' '"^^'"•' '^''^'" «Z^«*'«^- Paper presented at the 87th AnnualConvention of the American Psychol(^ical Association, New York 'annualn r ^ / " ' - ^" i " ^•'"P^*"^''' ^ ^ <^986)- Compliance-gaining in marital interaction: Power bases powerprocesses, and outcomes. Communication Monographs, 53,130-143. ^

Copyright of Communication Monographs is the property of National Communication Association and its

content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.