Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Apr 16.
Published in final edited form as: Psychol Addict Behav. 2015 Oct 12;30(2):209–219. doi: 10.1037/adb0000126

Attitude Ambivalence, Social Norms, and Behavioral Intentions: Developing Effective Antitobacco Persuasive Communications

Zachary P Hohman 1, William D Crano 2, Elizabeth M Niedbala 3
PMCID: PMC5902035  NIHMSID: NIHMS957169  PMID: 26460476

Abstract

This study assessed the moderating effects of attitude ambivalence on the relationship between social norms, attitudes, and behavioral intentions to use tobacco. It was predicted that people would use social norms to reduce attitude ambivalence, and that reduced ambivalence would lead to changes in attitudes and behavioral intentions. To test this hypothesis, participants (N =152) were exposed to persuasive communications designed to influence attitude ambivalence and perceived social norms regarding tobacco use. Analysis indicated that providing a social norm antagonistic to tobacco use significantly reduced ambivalence among participants reading the ambivalence message (p <.001). Examining changes in tobacco attitudes from pre- to postpersuasive communications demonstrated a significant decrease in tobacco attitudes only for participants reading the ambivalence message who were provided with the antitobacco use norm (p <.001). Ambivalent message participants also expressed significantly lower intentions to use tobacco when provided with social norms indicating antitobacco sentiments (p <.02), and this significant decrease in intentions was associated with changes in attitudes toward tobacco. These results point to the important role of social norms in mediating the effects of attitude ambivalence on subsequent behavior in preventative programs targeting tobacco use.

Keywords: attitude ambivalence, social norms, persuasive messages, prevention

Tobacco products increase the risk of developing serious health issues, including lung cancer, coronary heart disease, and stroke—and cause more than 480,000 deaths every year (U.S. Department of Health & Human Services, 2014). Young adults between 18 and 25 represent the cohort with the highest prevalence of tobacco use in the United States. In 2012, the rate of current users in this age group was 38.1% (Substance Abuse and Mental Health Services Administration; SAMHSA, 2013). While preventative efforts have led to considerable success in decreasing young adult tobacco use between 2002 and 2012 (SAMHSA, 2013), there remains a high prevalence of smokers on college campuses (Johnston, O’Malley, Bachman, & Schulenberg, 2004; Lee, Bahreinifar, & Ling, 2014; Rigotti, Lee, & Wechsler, 2000; Thompson et al., 2007). Among college students, 33% are current smokers, and 46% have used tobacco in the past year (Rigotti et al., 2000). Unlike adolescents, college students possess more freedom to make their own decisions about their health, one of which is deciding to smoke (Emmons, Wechsler, Dowdall, & Abraham, 1998; Gfroerer, Greenblatt, & Wright, 1997). Upon transitioning into college, young adults who once smoked only occasionally become heavier smokers (Christie-Smith, 1999; Colder et al., 2006; Patterson, Lerman, Kaufmann, Neuner, & Audrain-McGovern, 2004; Wechsler, Rigotti, Gledhill-Hoyt, & Lee, 1998). Further exacerbating the problem is the fact that the tobacco industry specifically targets college students through events in bars, nightclubs, and campus parties (Katz & Lavack, 2002; Ling & Glantz, 2002; Rigotti, Moran, & Wechsler, 2005). Tobacco use by college students remains an important issue to be addressed.

The effectiveness of antitobacco campaigns is likely responsible for the steady decrease in young adults’ tobacco use since 2002 (SAMHSA, 2013). Antismoking messages can increase negative attitudes toward the tobacco industry (Hersey et al., 2003) and toward smokers (Pechmann & Knight, 2002; Pechmann & Ratneshwar, 1994), and can reduce young adults’ smoking behaviors (Borders, Xu, Bacchi, Cohen, & SoRelle-Miner, 2005; Borland & Balmford, 2003; Goldman & Glantz, 1998; McVey & Stapleton, 2000; Rigotti, Regan, Majchrzak, Knight, & Wechsler, 2002). Despite the relative decline in tobacco use over the past decade, not all campaigns succeed. The overall effectiveness of antitobacco messages is mixed (e.g., Farrelly et al., 2002; Wakefield, Flay, Nichter, & Giovino, 2003; Wolburg, 2006), and tobacco use in college remains high. For example, in a qualitative study investigating college students’ responses to antitobacco campaigns, Wolburg (2006) found that antismoking advertisements had little impact on college student smoking attitudes and in sometime triggered a desire to smoke (the “boomerang” effect; Ringold, 2002). To decrease smoking behavior in young adults, tobacco prevention advertisers should consider their audience’s preexisting attitudes toward smoking.

Many antidrug media campaigns are designed to instill negative attitudes toward substances, assuming that negative attitudes will decrease the likelihood of usage (Kumpeer, 1997). However, smokers’ attitudes are different from those of nonsmokers (Piko, 2001; Siegel, Alvaro, & Burgoon, 2003), and this likely determines the effectiveness of media campaigns. Nonsmokers are more likely to hold negative attitudes toward smoking and perceive few advantages, if any, of the behavior (Chang, 2009; Siegel et al., 2003; Wolburg, 2006). Smokers, in contrast, are more likely to hold some positive attitudes toward smoking and to report that smoking has benefits (Chang, 2009; Dijk, Reubsaet, de Nooijer, & de Vries, 2007; Schleicher, Harris, Catley, Harrar, & Golbeck, 2008; Siegel et al., 2003). As previous substance use literature demonstrates, favorable attitudes toward drug use are positively associated with behavioral intentions and subsequent use (Lacy, 1981; Ritter, 1988; Wolford & Swisher, 1986). One way to improve preventative messages is to investigate the relationship between attitudes toward tobacco and actual tobacco use. Psychologists have long been at odds over the relationship between attitudes and behaviors (Crano & Prislin, 2006; McGuire, 1985). Today, the consensus is that attitudes can affect actions, but a number of variables moderate the attitude-behavior relation (Ajzen & Fishbein, 1977; Cacioppo, Gardner, & Berntson, 1997; Crano, 2012; Fishbein & Ajzen, 1975; Petty & Krosnick, 1995). Of the variables identified as potential moderators of the attitude-behavior link, attitude ambivalence is one of the most promising, but has received perhaps the least attention (Conner & Armitage, 2008), though a resurgence of research on ambivalence has occurred over the last decade (e.g., DeMarree, Wheeler, Briñol, & Petty, 2014).

People often simultaneously hold negative and positive evaluations of an attitude object. For example, one might enjoy the sensation of smoking and at the same time realize that tobacco is a danger to health. Consequences of attitude ambivalence are numerous: ambivalent attitudes are highly susceptible to persuasion (Bassili, 1996), less predictive of behavior (Armitage & Conner, 2000; Sparks, Harris, & Lockwood, 2004), less accessible in memory (Bargh, Chaiken, Govender, & Pratto, 1992; Krosnick, 1989), and associated with a variety of negative outcomes (Brown & Farber, 1951; Kaplan, 1972; Newby-Clark, McGregor, & Zanna, 2002; Priester, 2002; Rydell, McConnell, & Mackie, 2008). For example, Siegel et al. (2003) found that while nonsmokers and smokers held univalent attitudes toward smoking (either positive or negative), those at-risk for smoking (i.e., those considering smoking) perceived both benefits and disadvantages to smoking, and thus were viewed as ambivalent about tobacco use. Additionally, research by Zhao and Cai (2008) demonstrated that college students who were ambivalent about tobacco use sought out information about tobacco use more than college students who were not ambivalent (regardless of the students’ attitudes toward tobacco use). Further, a study by Lipkus, Green, Fesganes, and Sedikides (2001) demonstrated that current college tobacco users who were ambivalent about tobacco use experienced greater desire to quit than users who were not ambivalent. Overall, ambivalence about tobacco products during college seems to be an important construct for both the initiation and cessation of tobacco use, and individuals who hold ambivalent attitudes toward smoking may be highly susceptible to antismoking messages.

A particularly negative outcome of attitude ambivalence is the cognitive conflict and psychological discomfort associated with simultaneously holding positive and negative evaluations toward the same attitude object (e.g., Newby-Clark et al., 2002). For example, if people believe they will have more friends if they smoke, but also fear the negative health consequences of tobacco use, they will experience cognitive conflict. This conflict results in an aversive evaluative tension (Has, Katz, Rizzo, Bailey, & Moore, 1992), of which the individual may or may not be consciously aware (Petty & Briñol, 2009). In either case, almost all drive-reduction theories (e.g., Festinger, 1957) hold that this tension is experienced as undesirable or unpleasant, and that accommodations will be made to reduce it if possible (Priester, 2002).

To reduce ambivalence, research suggests that people pay close attention to information that will facilitate resolution of the conflict (Briñol, Petty, & Wheeler, 2006; Clark, Wegener, & Fabrigar, 2008; Hohman, Crano, Siegel, & Alvaro, 2014; Maio, Bell, & Esses, 1996). For example, Maio and associates (1996) showed that highly ambivalent participants elaborated a persuasive message more intensely, and were more persuaded than nonambivalent participants by strong messages. Other research demonstrates that people with highly ambivalent attitudes look for and use consensus information to resolve ambivalence (Hodson, Maio, & Esses, 2001). It seems evident that people seek information in the social context to reduce ambivalence.

A valued source for resolving ambivalence is one’s social group (Hogg, 2006). Groups help define how members should think and act by supplying (social) norms for appropriate behavior that define expected and apposite attitudes and actions of group members (Hogg & Reid, 2006). Because attitudes are important features of group life, individuals look to their groups to determine the correct attitudes in circumstances of high ambivalence. Given that social norms define group related attitudes, they provide a viable means of reducing ambivalence. This idea is supported by considerable research that demonstrates in the absence of concrete information, people look to similar others for information on what to think and how to behave (Festinger, 1954; Suls & Miller, 1977; Suls & Wheeler, 2000). For tobacco use, a strong predictor of smoking behavior is the perception that smoking among peers is normative or acceptable (Alesci, Forster, & Blaine, 2003; Islam & Johnson, 2005; Parkinson et al., 2009; Sussman et al., 1988; Thompson et al., 2007; United States Department of Health & Human Services, 1994). Because of this, antitobacco campaigns that convey smoking as socially unacceptable are effective at reducing smoking (Pechmann & Goldberg, 1998; Pechmann, Zhao, Goldberg, & Reibling, 2003). Thompson et al. (2007) found that college smokers who perceived that less than 50% of their peers smoked were less likely to continue smoking. College students, in particular, demonstrate a strong identification with their university and other students at their school. Following the norm at their university is important for typical college students, so their perception of peer smoking at their school may play a major role in their own smoking behavior.

Taken together, people with ambivalent attitudes, who presumably feel enhanced evaluative tension, should be motivated to reduce ambivalence by adhering to the social norms of important identity groups. Aligning attitudes with the views of the group will transform the ambivalent attitude positively or negatively, depending on group norms, thereby reducing ambivalence. Social norms also should moderate the relation between attitude ambivalence and behavioral intentions: They should be stronger predictors of behavioral intentions for people holding highly ambivalent attitudes than for those whose attitudes are not ambivalent. For most college students, the university they attend is an essential part of their identity, rendering fellow university students as valuable references of behavior when dealing the discomfort associated with attitudinal ambivalence.

There is disagreement about the most appropriate way to measure ambivalence (Conner & Sparks, 2002; Priester & Petty, 1996; Thompson et al., 1995), but most research uses either the felt or the potential ambivalence methods. Felt ambivalence uses bipolar scales, asking respondents to indicate the degree of conflict they feel when evaluating an object. A strength of this approach is that it taps the discomfort experienced when evaluations conflict. However, this method has been criticized because extraneous factors may affect felt ambivalence (Bassili, 1996) and people’s cognitive access to this type of information is far from established (Nisbett & Wilson, 1977; Zajonc, 1980). An alternate approach involves measurements that use separate items to tap positive and negative evaluations of an attitude object (Armitage & Arden, 2007; Kaplan, 1972; Newby-Clark, McGregor, & Zanna, 2008). These positive and negative evaluations are combined to yield a single ambivalence score. The potential ambivalence measure does not suffer from the same criticisms as the felt ambivalence, and for that reason we use a measure of potential ambivalence in the current study. Additionally, a study by Lipkus et al. (2001) demonstrated that felt ambivalence about tobacco use is related to potential ambivalence.

Current Study

The current study was designed to determine if college students with ambivalent attitudes toward tobacco use referenced social norms to reduce ambivalence. Relatively few studies have explored the relationship between social norms and attitude ambivalence. In one, Hodson and colleagues (2001) assessed participants’ ambivalence regarding social welfare. Their participants watched a debate about social welfare that presented both pro- and antiwelfare arguments. Then they provided participants with peer evaluations of the debate that indicated the majority thought the prosocial welfare (or the antisocial welfare) side had won the debate. Results indicated that highly ambivalent participants were more persuaded than those low in ambivalence by peer evaluations (consensus information), suggesting that they looked to others for information to reduce ambivalence. However, the researchers did not manipulate ambivalence experimentally, thus limiting our ability to determine cause-effect relationships. Additionally, consensus information about the winner of a debate is not the same as a social norm, which explicitly or implicitly indicates the beliefs one should hold.

In a more relevant study, Hohman and associates (2014) conducted a secondary analysis of data from a longitudinal study to determine if ambivalence about marijuana use rendered adolescents more persuadable by their friends’ opinions about marijuana. Analysis revealed that friends’ norms significantly predicted behavioral intentions (cross-sectionally) and future behavior (longitudinally) when adolescents held ambivalent marijuana attitudes. These results suggest that adolescents use social norms to reduce ambivalence. However, the study did not manipulate ambivalence or social norms, limiting our ability to determine if adolescents looked to their friends when they were ambivalent or if having friends who strongly favored marijuana use led adolescents to become ambivalent. Despite the research of both Hodson et al. (2001) and Hohman et al. (2014), there remain important questions with respect to the relationship between social norms and attitude ambivalence.

In the current research, attitude ambivalence and the presence or absence of a stated social norm regarding tobacco use were manipulated. Change in attitudes from pre- to postpersuasive communication, along with behavioral intentions to use a tobacco product, were measured. The attitude object for the current study was tobacco use—participants read an ambivalent or univalent persuasive communication about tobacco use, and were provided with either an antitobacco social norm or were told that the social norm was not known. The hypotheses for this study follow:

Hypothesis 1: Ambivalence about tobacco use will be greater when the persuasive message is ambivalent and the social norm is unknown compared to when the message is ambivalent and the social norm is antitobacco.

Hypothesis 2: A decrease in favorability of tobacco attitudes from pre- to postpersuasive communication is predicted among participants provided with the ambivalent message and antitobacco norm.

Hypothesis 3: Intentions to use tobacco will be lowest among those provided with the ambivalent message and antitobacco norm.

Hypothesis 4: The relationship between the persuasive communication and social norms on behavioral intentions will be mediated by changes in attitude.

Overall, this research will inform prevention researchers on one method through which tobacco use can be reduced—messages that increase ambivalence and then provide a social norm antagonistic to usage should be particularly effective at reducing tobacco use.

Method

Participants

Seventy-seven male and 75 female participants (N =152, Mage =19.60, SD =3.34) were recruited from the psychology participant pool at a Midwest University to participate in the study, which was hosted on the Qualtrics website.

Procedure and Measures

This study was a 2 (persuasive communication—ambivalent vs. univalent communication) × 2 (social norms—antitobacco use vs. unknown) factorial design with intentions to use a tobacco product as the main dependent variable, and change in attitude toward tobacco use as the mediating variable. Prior tobacco use was a covariate.

The research was introduced as a study of how people respond to health-related messages. Participants were told that they would read a health-related message, answer a few questions, and provide their opinions about health-related behaviors. To begin, participants indicated their attitude with a single item about the degree to which they favored or opposed tobacco use (1 = strongly oppose, 7 = strongly favor; M = 3.31, SD = 1.49). Then, participants read a health-related persuasive communication that either contained only negative aspects of tobacco use (univalent condition, n = 70) or both positive and negative aspects of tobacco use (ambivalent condition, n = 82), see Appendix. This is a standard manipulation of ambivalence (Conner & Sparks, 2002; van Harreveld, Rutjens, Rotterveel, Nordgren, & van der Plight, 2009).

After reading the persuasive communication, participants read a message that constituted the manipulation of social norms, using a standard manipulation of social norms (e.g., Ajzen & Fishbein, 1977; Reno et al., 1993). The antitobacco use norm read (n = 79), “Results of a recent study at [Midwest] University found that only 9% of students approve of tobacco use. The goal of this message is to convince students that they should continue to hold a negative view of tobacco use.” In the norm unknown condition, participants read (n = 73), “Results of a recent study at [Midwest] University found that students do not have an opinion about tobacco use. The goal of this message is to convince students that they should hold a negative view of tobacco use.”

Participants answered a few questions about the message to substantiate the purpose of the study, and then rated the message’s persuasiveness (1 = very ineffective, 7 = very effective; M = 4.35, SD = 1.36), how well it was constructed (1 = horrible, 7 = excellent; M = 4.84, SD = .98), and how convincing it was (1 = not at all convincing, 7 = extremely convincing; M = 4.44, SD = 1.08). As a manipulation check for the social norms manipulation, participants were asked, “In your opinion, how much do [Midwest University] students approve or disapprove of tobacco use?” (1 = strongly disapprove, 7 = strongly approve; M = 4.24, SD = 1.31).

To measure ambivalence, participants answered four semantic differential items about tobacco use (i.e., how good [1]- bad [5], pleasant [1]-unpleasant [5], beneficial [1]- harmful [5], and valuable [1]- worthless [5]; α = .87). We computed the SD across the four-items of the semantic differential to determine participants’ level of ambivalence (M = .628, SD = .873). The SD scale can range from 0 (no variability across the items) to 2.31 (highest possible variability across the four items). A similar measure of ambivalence was used in Hohman et al. (2014). This measure is conceptually similar to the potential ambivalence measure of Kaplan (1972), which uses a composite of negative and positive evaluations to measure ambivalence. The SD measure avoids some of the issues with Kaplan’s method of combining positive negative scores into a separate index (for a discussion refer to Ullrich, Schermelleh-Engel, & Böttcher, 2008). The SD approach provides a clear and useful conceptual analog to the typical measure of potential ambivalence and avoids the statistical issues associated with the typical measure. Additionally, we chose to use an indirect measure of ambivalence after manipulating ambivalence because a more direct measure (e.g., measuring felt ambivalence) could impact on the dependent measures that followed.

A central dependent measure, which also is used in a mediation analysis of usage, was participants’ change in attitude toward intentions for tobacco use from pre- to postpersuasive communication. The same one-item measure used in the premessage attitude measure was used for the postmessage measure (M = 3.05, SD = 1.50). Next, participants indicated their intentions to use a tobacco product in the next year (1 = absolutely will not use a tobacco product, 5 = absolutely will use a tobacco product; M = 2.43, SD = 1.56).

The final section of the questionnaire obtained demographic data (e.g., sex, age) and previous tobacco use. To control for prior tobacco use in analyses reported below, participants identified their previous tobacco use (1 = I have never used a tobacco product, 2 = I have used a tobacco product once or twice in the past but I no longer use tobacco products, 3 = I use tobacco products once a month, 4 = I use tobacco products once a week, 5 = I use tobacco products more than once a week but not daily, 6 = I use tobacco products daily; M = 2.43, SD = 1.70).

Results

Background Variables

Participants were college students enrolled in a psychology class. The sample was predominately White (62.5% White, 19.1% Hispanic, 5.3% Multiracial, 3.3% Asian, 6.6% Black, and 3.3% other)—ethnic proportion did not differ as a function of the persuasive communication or social norms manipulation. A two-way persuasive communication by social norms analysis of variance (ANOVA) on age revealed no significant effects. A logistic regression (with persuasive communication and social norms entered in the first step and their interaction entered in the second step) on sex revealed no significant main effects or an interaction.

Persuasive Communication Effectiveness

Effectiveness

A two-way persuasive communication by social norms analysis of covariance (ANCOVA) on perceived effectiveness of the persuasive message was conducted, controlling for previous tobacco use. The effectiveness analysis revealed only a significant main effect for social norms, F(1, 147) = 3.99, p = .047, ηp2=.026. The persuasive communication was perceived to be more effective when the antitobacco norm was provided (M = 4.11, SE = .157) compared with when no norm was provided (M = 4.55, SE = .150).

Constructed

A two-way persuasive communication by social norms ANCOVA on how well constructed the persuasive message was, controlling for previous tobacco use, was conducted. This analysis revealed only a significant main effect for social norms, F(1, 147) = 11.10, p = .001, ηp2=.070. The persuasive communication was perceived to be better constructed when the antitobacco norm was provided (M = 4.57, SE = .112) compared to when no norm was provided (M = 5.09, SE = .108).

Convincing

A two-way persuasive communication by social norms ANCOVA on how convincing the persuasive message was perceived to be, controlling for tobacco use, was conducted. The convincing analysis revealed a significant main effect for social norms, F(1, 147) = 6.88, p = .01, ηp2=.044. The persuasive communication was perceived to be more convincing when the antitobacco norm was provided (M = 4.20, SE = .123) compared with the no norm condition (M = 4.64, SE = .117). As well, there was a nonsignificant marginal trend for the interaction between the persuasive communication and social norms, F(1, 147) = 3.51, p = .063, ηp2=.023. When reading the ambivalent message, participants thought the message was more convincing when provided the antitobacco norm (M = 4.81, SE = .162) compared with no norm (M = 4.03, SE = .171), F(1, 147) = 10.43, p = .002, ηp2=.066. However, for the univalent message, there was no significant difference on convincingness when the norm was unknown (M = 4.37, SE = .176) or known to be antagonistic to tobacco use (M = 4.48, SE = .178), F(1, 147) = .21, p = .651, ηp2=.001. For the other simple effects, when social norms were antitobacco there was no significant difference between the ambivalent (M = 4.81, SE = .162) and univalent messages (M = 4.48, SE = .178), F(1, 147) = 1.71, p = .192, ηp2=.011. When social norms were unknown there was no significant difference between the ambivalent (M = 4.03, SE = .171) and univalent messages (M = 4.37, SE = .176), F(1, 147) = 1.95, p = .165, ηp2=.013.

Social Norm Manipulation Check

A two-way persuasive communication by social norms ANCOVA on the social norm measure, controlling for previous tobacco use, was conducted. The social norms analysis revealed only a nonsignificant marginal trend for social norms, F(1, 147) = 3.80, p = .053, ηp2=.025. Participants tended to believe that students approved of tobacco use more in the norm unknown condition (M = 4.47, SE = .152) than in the antitobacco norm condition (M = 4.05, SE = .147); although not statistically significant, the marginal trend suggests the social norm manipulation operated as planned.

Ambivalence

A two-way persuasive communication by social norms ANCOVA on ambivalence, controlling for previous tobacco use, revealed a significant main effect for social norms, F(1, 147) = 4.77, p = .03, ηp2=.031. Participants had significantly more ambivalence in the norm unknown condition (M = .497, SE = .09) than the antitobacco norm condition (M = .775, SE = .09). However, this main effect must be interpreted in light of a significant interaction between persuasive communication and social norms, F(1, 147) = 5.82, p = .017, ηp2=.038, see Figure 1. As predicted (Hypothesis 1), when reading the ambivalent message, participants had significantly greater ambivalence when the social norm was unknown (M = 1.004, SE = .128) compared with when the norm was clearly antagonistic to tobacco use (M = .406, SE = .120), F(1, 147) = 10.97, p = .001, ηp2=.071. However, for the univalent message there was no significant difference in ambivalence when the norm was unknown (M = .547, SE = .131) or known to be antagonistic to tobacco use (M = .588, SE = .134), F(1, 147) = 0.01, p = .924, ηp2<.001. For the other simple effects, when social norms were antitobacco use there was no significant differences between the ambivalent (M = .406, SE = .120) and univalent conditions (M = .588, SE = .134), F(1, 147) = 0.35, p = .556, ηp2=.002. When social norms were unknown, ambivalence was significantly higher in the ambivalent condition (M = 1.004, SE = .128) compared with univalent condition (M = .547, SE = .131), F(1, 147) = 6.51, p = .012, ηp2=.044.

Figure 1.

Figure 1

Open in a new tab

Ambivalence as a function of the ambivalence and social norms manipulations. Error bars represent 95% confidence interval.

Attitude

A two-way persuasive communication by social norms repeated measures ANCOVA, with the pre- postpersuasive communication attitude as the repeated factor, controlling for previous tobacco use, revealed a significant repeated measures by social norm interaction, F(1, 147) = 4.19, p = .049, ηp2=.028 (Figure 2). When the antitobacco norm was provided there was a significant decrease in attitudes from pre- (M = 3.16, SE = .114) to postpersuasive communication (M = 2.78, SE = .106), F(1, 147) = 15.89, p = .001, ηp2=.098. However, there was no significant difference in preattitude (M = 3.52, SE = .118) and postattitude (M = 3.42, SE = .110) when no norm was provided, F(1, 147) = .97, p = .326, ηp2=.007.

Figure 2.

Figure 2

Open in a new tab

Change in tobacco attitude from premessage to postmessage as a function of the ambivalence and social norms manipulations. Error bars represent 95% confidence interval.

The three-way interaction between persuasive communication, social norms, and the repeated factor also was statistically significant, F(1, 147) = 5.88, p = .017, ηp2=.038 (Figure 3). Examining the three-way interaction, the change from preattitude to postattitude was only significant for the ambivalent communication when the antitobacco norm was provided, F(1, 147) = 20.31, p = .001, ηp2=.12: participants had a significant decrease in their attitude from pre- (M = 2.72, SE = .158) to postpersuasive communication (M = 2.12, SE = .147). There was no significant change in attitude from pre- (M = 3.56, SE = .165) to postpersuasive communication (M = 3.59, SE = .154) when participants read the ambivalent message with no norm provided (F(1, 147) = .07, p = .80, ηp2<.001), when they read the univalent message with no norm provided (Ms = 3.49 and 3.26, SEs = .169 and 1.57; F(1, 147) = 2.63, p = .107, ηp2=.018), or when they read the univalent message with the antitobacco norm provided (Ms = 3.60 and 3.44, SEs = .172 and .160; F(1, 147) = 1.29, p = .259, ηp2=.009).

Figure 3.

Figure 3

Open in a new tab

Intentions to use tobacco as a function of the ambivalence and social norms manipulations. Error bars represent 95% confidence interval.

Intentions

A two-way, persuasive communication by social norms ANCOVA with behavioral intentions as the dependent measure, controlling for previous tobacco use, revealed only a statistically significant interaction between persuasive communication and social norms, F(1, 147) = 6.67, p = .011, ηp2=.043. When the norm was unknown, there was a nonsignificant marginal trend for the difference between the ambivalent (M = 2.69, SE = .151) and univalent messages (M = 2.31, SE = .155), F(1, 147) = 3.15, p = .078, ηp2=.021. When the antitobacco use norm was present, participants had significantly less intention to use tobacco after reading the ambivalent message (M = 2.17, SE = .143) compared to reading the univalent message (M = 2.60, SE = .160), F(1, 147) = 3.82, p = .05, ηp2=.025. For the other simple effects, for the ambivalent message, participants had significantly lower intentions to use tobacco when the antitobacco norm was present (M = 2.17, SE = .143) than when the norm was unknown (M = 2.69, SE = .151), F(1, 147) = 5.96, p = .016, ηp2=.039. For the univalent message, there was no significant difference in intentions between the antitobacco norm (M = 2.60, SE = .160) and norm unknown (M = 2.31, SE = .155) condition, F(1, 147) = 1.75, p = .188, ηp2=.012.

Mediation Analysis

Following Preacher and colleagues (Preacher & Hayes, 2008; Preacher, Rucker, & Hayes, 2007), moderated mediation analyses were conducted to examine the conditional indirect effect of tobacco attitude on the interaction of persuasive communication and social norms on intentions to use tobacco (see Antonakis, Bendahan, Jacquart, & Lalive, 2010; Hayes, 2009). For this analysis we used Model 8 in SPSS (Hayes, 2013) with 50,000 bootstrapped samples.

The total effect of the interaction between persuasive communication and social norms on intentions was statistically significant (Path c, R2 = .664, ΔR2 = .015, F(1, 147) = 6.67, p = .011, β = −.593, t = −2.58, p = .011). When the postmessage attitude (and the premessage attitude entered as a covariate) was entered into the model, this relationship was no longer statistically significant (Path c,’ β = −.170, t = −.730, p = .467). Results from the moderated mediation analysis indicated that the total indirect effect (interaction between ambivalence and social norms → attitude → intentions) was statistically significant—total indirect effect = −.282, SE = .130, 95% CI bias corrected [−.585, −.071]. Examining this mediation, the conditional indirect effect of attitude on intentions remained significant for social norms for the ambivalent message, indirect effect = −.266, SE = .113, 95% CI bias corrected [−.519, −0.073], whereas it was not significant for the univalent message, indirect effect = .016, SE = .046, 95% CI bias corrected [−0.064, .126]. Thus, change in tobacco attitudes mediated the interactive effect of persuasive communication and social norms on intentions to use tobacco.

Discussion

Young adults show the highest prevalence of tobacco use of all other age groups in the United States. Compounding this problem, college students are specifically targeted by the tobacco-industry through events in bars, nightclubs, and campus parties (Rigotti et al., 2005). Preventative efforts designed to counter tobacco-industry marketing strategies rely on the assumption that increasing negative attitudes toward tobacco will decrease tobacco use (Kumpeer, 1997), and while these techniques show some success (Dennhardt & Murphy, 2013; Larimer, Kilmer, & Lee, 2005), there is more that can be done to improve antitobacco use messages. The current study investigated the relationship between attitudes and behavior with specific focus on attitude ambivalence as key feature in preventing college student tobacco use. At some point, almost everyone will experience the evaluative conflict that arises from attitude ambivalence. The resultant aversive feeling arising from this conflict has important implications for prevention researchers, because people are motivated to seek information in the social environment to reduce it, and its cause—ambivalence. To develop a more complete understanding of how ambivalence affects the attitude-behavior relationship for health behaviors may require researchers to investigate aspects in the social environment that reduce ambivalence. The purpose of the current study was to determine if social norms could be one such factor.

For the first hypothesis (with ambivalence as the outcome), analysis demonstrated that social norms decreased ambivalence—as evidenced by the significant difference in ambivalence when the social norm regarding tobacco use was presented as either unknown or antagonistic to use. These results indicate that social norms reduced attitudinal ambivalence. Research indicates that when people are ambivalent they process information in the environment to reduce it (Briñol et al., 2006; Clark et al., 2008; Hohman et al., 2014; Maio et al., 1996). Our results suggest that social norms may be a crucial feature of the information search.

For attitude change (Hypothesis 2), the analysis disclosed a significant decline in favorability of tobacco use attitudes among participants exposed to the ambivalent message, if they received the antitobacco use norm. For the other experimental conditions, there was no significant change in attitudes. This result suggests that to reduce ambivalence people may change their attitudes to be in accord with social norms. Normative information from important groups lets people know what their attitude should be, and to the extent that the norm is univalent, attitude ambivalence should be reduced. The current study provided either an antitobacco norm or information that the norm was not known. It would have been interesting to include an ambivalent norm condition and determine how that affected attitude change and behavioral intentions; however, it was deemed undesirable, as such a manipulation might have encouraged tobacco use. There is some research that supports this concern: when people are given a reason to attribute the aversive feelings of attitude ambivalence (e.g., as because of a placebo) they do not experience the negative outcomes of ambivalence (Nordgren, van Harreveld, & van der Plight, 2006).

For behavioral intentions (Hypothesis 3), there was a significant difference in ambivalence effects when the antitobacco social norm was present versus unknown—antitobacco use norms led to significantly lower future use intentions. A moderated mediation analysis (Hypothesis 4) was conducted to determine if the significant results for behavioral intentions were linked to a change in attitudes. Results from this analysis indicated that the significant decrease in intentions among those provided the antitobacco norm was associated with a significant decrease in attitudes toward tobacco use. This result makes sense in light of research that demonstrates that ambivalence moderates the attitude-behavior relation (e.g., Conner, Povey, Sparks, & James, 2003; Moore, 1973, 1980; Sparks, Hedderley, & Shepherd, 1992; Sparks, Conner, James, Shepherd, & Povey, 2001). It suggests that attitudes are stronger predictors of behavior when they are univalent. When participants changed their attitude to fall in line with the social norm, the attitude became univalent, and they subsequently aligned their intentions to fall in line with the norm as well.

Across all results in this study, we demonstrated that social norms play an important role in the reduction of ambivalence. However, the extent to which social norms reduce ambivalence depends whether or not the norm provides a univalent attitudinal prescription—that is, what the person’s attitude should be—and the extent to which the person identifies with the group that provides the norm (Hogg, 2006). The extent to which people follow social norms depends on how strongly they identify with the group—the more they identify with the group the more they follow the group’s norms (Abrams & Hogg, 2010). This suggests that identification may be the underlying process that accounts for social norms reducing ambivalence. Future research should explore the relationship between attitude ambivalence, social norms, and group identification.

That we could compare only an antisocial norm with an unknown norm condition is limitation to this study, but as noted, because of ethical concerns we could not include a protobacco use norm in the study design. A future study might use a balanced design using an attitude object in which it would be ethical to include both pro- and antisocial norm conditions. Another limitation to our design was that we included a measure of behavioral intentions but no measure of actual behavior. Therefore, whereas we can confidently conclude that social norms provided to people who are ambivalent will lead to changes in behavioral intentions, we cannot be certain that those changes will lead to behavior change. However, considerable research suggests that intentions are powerfully predictive of subsequent behavior (e.g., Webb & Sheeran, 2006).

Another limitation to this study was that tobacco attitudes were measured with only a single item. It would have been better to include a multiitem measure of tobacco attitudes and future research should explore how ambivalence and social norms impact multiitem scales of tobacco attitudes. Additionally, the single-item measure of tobacco attitudes was a global attitude measure, rather than a specific measure of tobacco use by oneself. Therefore, when answering this question it is possible that participants were not thinking about their attitude about tobacco use, but rather for other people. However, past research suggests that specific attitudes are linked with global attitudes in memory (Petty, Briñol, & DeMarree, 2007). Nonetheless, future research should include more specific attitude questions when measure attitudes about tobacco use.

This research indicates that social norms can reduce the cognitive conflict that occurs as a result of attitude ambivalence. When people hold both positive and negative opinions about an attitude object, they will look to social norms of important in-groups to reduce that conflict. Antitobacco campaigns should consider the social groups that are most important to the targeted respondents’ identity when designing prevention messages, because these groups serve as normative references that foster resolution of attitude ambivalence. The prevention messages in the current study focused on attitudes toward tobacco use, but future research should expand to other problematic health behaviors on college campuses such as binge drinking, abuse of prescription drugs, and engaging in risky sexual behavior. Adding to the utility of this research, the use of ambivalence and social norms in prevention messages could be applied to many different populations (not just college students) and many different health behaviors. Overall, prevention messages that increase ambivalence about a health issue and then provide a social norm on the issue should be particularly effective in changing people’s behavioral intentions.

Appendix. Persuasive Communications

Univalent Message: Tobacco use is one of the most deadly behaviors. In the United States alone, 1,200 people die each day from a tobacco related disease. Use of tobacco products lead to an increased risk of heart disease, stroke, lung cancer, and several other cancers (e.g., bladder, mouth, kidney, and stomach). Other health consequences of tobacco use include higher rates of infertility, pregnancy problems, and general fitness issues. Therefore, the World Health Organization, Food and Drug Administration, the Centers for Disease Control and Prevention, and almost every medical board and association on the face of the Earth holds an antitobacco platform. Even though the negative health consequences of tobacco use are well documented and known, people continue to begin using and continue using tobacco products. If you or someone you know would like more information on tobacco use or help to quit using please go to thetruth.com, smoke-free.gov, or call 1-877-937-7848 (a toll-free quit line).

Ambivalent Message: Tobacco use is one of the most deadly behaviors. In the United States alone, 1,200 people die each day from a tobacco related disease. Use of tobacco products lead to an increased risk of heart disease, stroke, lung cancer, and several other cancers (e.g., bladder, mouth, kidney, and stomach). Other health consequences of tobacco use include higher rates of infertility, pregnancy problems, and general fitness issues. Therefore, the World Health Organization, Food and Drug Administration, the Centers for Disease Control and Prevention, and almost every medical board and association on the face of the Earth holds an antitobacco platform. Even though the negative health consequences of tobacco use are well documented and known, people continue to begin using and continue using tobacco products. Reasons that people provide for using tobacco products include: enjoying the taste, to deal with stress, it makes them to feel good, because it is cool, and for social reasons. If you or someone you know would like more information on tobacco use or help to quit using please go to thetruth.com, smokefree.gov, or call 1-877-937-7848 (a toll-free quit line).

Contributor Information

Zachary P. Hohman, Texas Tech University

William D. Crano, Claremont Graduate University

Elizabeth M. Niedbala, Texas Tech University

References

  1. Abrams D, Hogg MA. Social identity and self-categorization. In: Dovidio JF, Hewstone M, Glick P, Esses VM, editors. The SAGE handbook of prejudice, stereotyping and discrimination. London: SAGE; 2010. pp. 179–193. http://dx.doi.org/10.4135/9781446200919.n11. [Google Scholar]
  2. Ajzen I, Fishbein M. Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin. 1977;84:888–918. http://dx.doi.org/10.1037/0033-2909.84.5.888. [Google Scholar]
  3. Alesci NL, Forster JL, Blaine T. Smoking visibility, perceived acceptability, and frequency in various locations among youth and adults. Preventive Medicine. 2003;36:272–281. doi: 10.1016/s0091-7435(02)00029-4. http://dx.doi.org/10.1016/S0091-7435(02)00029-4. [DOI] [PubMed] [Google Scholar]
  4. Antonakis J, Bendahan S, Jacquart P, Lalive R. On making causal claims: A review and recommendations. The Leadership Quarterly. 2010;21:1086–1120. http://dx.doi.org/10.1016/j.leaqua.2010.10.010. [Google Scholar]
  5. Armitage CJ, Arden MA. Felt and potential ambivalence across the stages of change. Journal of Health Psychology. 2007;12:149–158. doi: 10.1177/1359105307071749. http://dx.doi.org/10.1177/1359105307071749. [DOI] [PubMed] [Google Scholar]
  6. Armitage CJ, Conner M. Attitudinal ambivalence: A test of three key hypotheses. Personality and Social Psychology Bulletin. 2000;26:1421–1432. http://dx.doi.org/10.1177/0146167200263009. [Google Scholar]
  7. Bargh JA, Chaiken S, Govender R, Pratto F. The generality of the automatic attitude activation effect. Journal of Personality and Social Psychology. 1992;62:893–912. doi: 10.1037//0022-3514.62.6.893. http://dx.doi.org/10.1037/0022-3514.62.6.893. [DOI] [PubMed] [Google Scholar]
  8. Bassili JN. Meta-judgmental versus operative indexes of psychological attributes: The case of measures of attitude strength. Journal of Personality and Social Psychology. 1996;71:637–653. http://dx.doi.org/10.1037/0022-3514.71.4.637. [Google Scholar]
  9. Borders TF, Xu KT, Bacchi D, Cohen L, SoRelle-Miner D. College campus smoking policies and programs and students’ smoking behaviors. BMC Public Health. 2005;5:74. doi: 10.1186/1471-2458-5-74. http://dx.doi.org/10.1186/1471-2458-5-74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Borland R, Balmford J. Understanding how mass media campaigns impact on smokers. Tobacco Control. 2003;12(Suppl 2):ii45–ii52. doi: 10.1136/tc.12.suppl_2.ii45. http://dx.doi.org/10.1136/tc.12.suppl_2.ii45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Briñol P, Petty RE, Wheeler SC. Discrepancies between explicit and implicit self-concepts: Consequences for information processing. Journal of Personality and Social Psychology. 2006;91:154–170. doi: 10.1037/0022-3514.91.1.154. http://dx.doi.org/10.1037/0022-3514.91.1.154. [DOI] [PubMed] [Google Scholar]
  12. Brown JS, Farber IE. Emotions conceptualized as intervening variables—With suggestions toward a theory of frustration. Psychological Bulletin. 1951;48:465–495. doi: 10.1037/h0058839. http://dx.doi.org/10.1037/h0058839. [DOI] [PubMed] [Google Scholar]
  13. Cacioppo JT, Gardner WL, Berntson GG. Beyond bipolar conceptualizations and measures: The case of attitudes and evaluative space. Personality and Social Psychology Review. 1997;1:3–25. doi: 10.1207/s15327957pspr0101_2. http://dx.doi.org/10.1207/s15327957pspr0101_2. [DOI] [PubMed] [Google Scholar]
  14. Chang C. Psychological motives versus health concerns: Predicting smoking attitudes and promoting antismoking attitudes. Health Communication. 2009;24:1–11. doi: 10.1080/10410230802465241. http://dx.doi.org/10.1080/10410230802465241. [DOI] [PubMed] [Google Scholar]
  15. Christie-Smith D. Smoking-cessation programs need to target college students. American Journal of Health-System Pharmacy: AJHP: Official Journal of the American Society of Health-System Pharmacists. 1999;56:416. doi: 10.1093/ajhp/56.5.416. [DOI] [PubMed] [Google Scholar]
  16. Clark JK, Wegener DT, Fabrigar LR. Attitudinal ambivalence and message-based persuasion: Motivated processing of proattitudinal information and avoidance of counterattitudinal information. Personality and Social Psychology Bulletin. 2008;34:565–577. doi: 10.1177/0146167207312527. http://dx.doi.org/10.1177/0146167207312527. [DOI] [PubMed] [Google Scholar]
  17. Colder CR, Lloyd-Richardson EE, Flaherty BP, Hedeker D, Segawa E, Flay BR the Tobacco Etiology Research Network. The natural history of college smoking: Trajectories of daily smoking during the freshman year. Addictive Behaviors. 2006;31:2212–2222. doi: 10.1016/j.addbeh.2006.02.011. http://dx.doi.org/10.1016/j.addbeh.2006.02.011. [DOI] [PubMed] [Google Scholar]
  18. Conner M, Armitage CJ. Attitude ambivalence. In: Crano WD, Prislin R, editors. Attitudes and attitude change. New York, NY: Psychology Press; 2008. pp. 261–288. [Google Scholar]
  19. Conner M, Povey R, Sparks P, James R. Moderating role of attitude ambivalence within the theory of planned behavior. British Journal of Social Psychology. 2003;42:75–94. doi: 10.1348/014466603763276135. [DOI] [PubMed] [Google Scholar]
  20. Conner M, Sparks P. Ambivalence and attitudes. European Review of Social Psychology. 2002;12:37–70. http://dx.doi.org/10.1080/14792772143000012. [Google Scholar]
  21. Crano WD. The rules of influence. New York, NY: St. Martin’s Press; 2012. [Google Scholar]
  22. Crano WD, Prislin R. Attitudes and persuasion. Annual Review of Psychology. 2006;57:345–374. doi: 10.1146/annurev.psych.57.102904.190034. http://dx.doi.org/10.1146/annurev.psych.57.102904.190034. [DOI] [PubMed] [Google Scholar]
  23. DeMarree KG, Wheeler SC, Briñol P, Petty RE. Wanting other attitudes: Actual-desired attitude discrepancies predict feelings of ambivalence and ambivalence consequences. Journal of Experimental Social Psychology. 2014;53:5–18. http://dx.doi.org/10.1016/j.jesp.2014.02.001. [Google Scholar]
  24. Dennhardt AA, Murphy JG. Prevention and treatment of college student drug use: A review of the literature. Addictive Behaviors. 2013;38:2607–2618. doi: 10.1016/j.addbeh.2013.06.006. http://dx.doi.org/10.1016/j.addbeh.2013.06.006. [DOI] [PubMed] [Google Scholar]
  25. Dijk F, Reubsaet A, de Nooijer J, de Vries H. Smoking status and peer support as the main predictors of smoking cessation in adolescents from six European countries. Nicotine & Tobacco Research. 2007;9(Suppl 3):S495–S504. doi: 10.1080/14622200701587060. http://dx.doi.org/10.1080/14622200701587060. [DOI] [PubMed] [Google Scholar]
  26. Eagly AH, Chaiken S. Attitude strength, attitude structure, and resistance to change. In: Petty RE, Krosnick JA, editors. Attitude strength: Antecedents and consequences. Hillsdale, NJ: Erlbaum; 1995. pp. 413–432. [Google Scholar]; Experimental Social Psychology. 14:398–408. [Google Scholar]
  27. Emmons KM, Wechsler H, Dowdall G, Abraham M. Predictors of smoking among US college students. American Journal of Public Health. 1998;88:104–107. doi: 10.2105/ajph.88.1.104. http://dx.doi.org/10.2105/AJPH.88.1.104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Farrelly MC, Healton CG, Davis KC, Messeri P, Hersey JC, Haviland ML. Getting to the truth: Evaluating national tobacco countermarketing campaigns. American Journal of Public Health. 2002;92:901–907. doi: 10.2105/ajph.92.6.901. http://dx.doi.org/10.2105/AJPH.92.6.901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Festinger L. A theory of social comparison processes. Human Relations. 1954;7:117–140. http://dx.doi.org/10.1177/001872675400700202. [Google Scholar]
  30. Festinger L. A theory of cognitive dissonance. Evanston, IL: Row, Peterson; 1957. [Google Scholar]
  31. Fishbein M, Ajzen I. Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison Wesley; 1975. [Google Scholar]
  32. Gfroerer JC, Greenblatt JC, Wright DA. Substance use in the US college-age population: Differences according to educational status and living arrangement. American Journal of Public Health. 1997;87:62–65. doi: 10.2105/ajph.87.1.62. http://dx.doi.org/10.2105/AJPH.87.1.62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Goldman LK, Glantz SA. Evaluation of antismoking advertising campaigns. Journal of the American Medical Association. 1998;279:772–777. doi: 10.1001/jama.279.10.772. http://dx.doi.org/10.1001/jama.279.10.772. [DOI] [PubMed] [Google Scholar]
  34. Has RG, Katz I, Rizzo N, Bailey J, Moore L. When racial ambivalence evokes negative affect, using a disguised measure of mood. Personality and Social Psychology Bulletin. 1992;18:786–797. http://dx.doi.org/10.1177/0146167292186015. [Google Scholar]
  35. Hayes AF. Beyond Barron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs. 2009;76:408–420. http://dx.doi.org/10.1080/03637750903310360. [Google Scholar]
  36. Hayes AF. Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press; 2013. [Google Scholar]
  37. Hersey JC, Niederdeppe J, Evans WD, Nonnemaker J, Blahut S, Farrelly MC, … Haviland ML. The effects of state counterindustry media campaigns on beliefs, attitudes, and smoking status among teens and young adults. Preventive Medicine. 2003;37:544–552. doi: 10.1016/j.ypmed.2003.07.002. http://dx.doi.org/10.1016/j.ypmed.2003.07.002. [DOI] [PubMed] [Google Scholar]
  38. Hodson G, Maio GR, Esses VM. The role of attitudinal ambivalence in susceptibility to consensus information. Basic and Applied Social Psychology. 2001;23:197–205. http://dx.doi.org/10.1207/S15324834BASP2303_6. [Google Scholar]
  39. Hogg MA. Social identity theory. In: Burke PJ, editor. Contemporary social psychological theories. Palo Alto, CA: Stanford University Press; 2006. pp. 111–136. [Google Scholar]
  40. Hogg MA, Reid SA. Social identity, self-categorization, and the communication of group norms. Communication Theory. 2006;16:7–30. http://dx.doi.org/10.1111/j.1468-2885.2006.00003.x. [Google Scholar]
  41. Hohman ZP, Crano WD, Siegel JT, Alvaro EM. Attitude ambivalence, friend norms, and adolescent drug use. Prevention Science. 2014;15:65–74. doi: 10.1007/s11121-013-0368-8. http://dx.doi.org/10.1007/s11121-013-0368-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Islam SMS, Johnson CA. Influence of known psychosocial smoking risk factors on Egyptian adolescents’ cigarette smoking behavior. Health Promotion International. 2005;20:135–145. doi: 10.1093/heapro/dah604. http://dx.doi.org/10.1093/heapro/dah604. [DOI] [PubMed] [Google Scholar]
  43. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring the future: National survey results on drug use 1975–2003. Bethesda, MD: National Institute on Drug Abuse; 2004. Vol. 2: College students and adults ages 19–45. (NIH Publication No. 04–5508) [Google Scholar]
  44. Kaplan KJ. On the ambivalence-indifference problem in attitude theory and measurement: A suggested modification of the semantic differential technique. Psychological Bulletin. 1972;77:361–372. http://dx.doi.org/10.1037/h0032590. [Google Scholar]
  45. Katz SK, Lavack AM. Tobacco related bar promotions: Insights from tobacco industry documents. Tobacco Control. 2002;11(Suppl 1):i92–i101. doi: 10.1136/tc.11.suppl_1.i92. http://dx.doi.org/10.1136/tc.11.suppl_1.i92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Krosnick JA. Attitude importance and attitude accessibility. Personality and Social Psychology Bulletin. 1989;15:297–308. http://dx.doi.org/10.1177/0146167289153002. [Google Scholar]
  47. Kumpeer KL. Drug abuse prevention: What works. Darby, PA: DIANE publishing; 1997. [Google Scholar]
  48. Lacy WB. The influence of attitudes and current friends on drug use intentions. The Journal of Social Psychology. 1981;113:65–76. doi: 10.1080/00224545.1981.9924350. http://dx.doi.org/10.1080/00224545.1981.9924350. [DOI] [PubMed] [Google Scholar]
  49. Larimer ME, Kilmer JR, Lee CM. College student drug prevention: A review of individually-oriented prevention strategies. Journal of Drug Issues. 2005;35:431–456. http://dx.doi.org/10.1177/002204260503500210. [Google Scholar]
  50. Lee YO, Bahreinifar S, Ling PM. Understanding tobacco-related attitudes among college and noncollege young adult hookah and cigarette users. Journal of American College Health. 2014;62:10–18. doi: 10.1080/07448481.2013.842171. http://dx.doi.org/10.1080/07448481.2013.842171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Ling PM, Glantz SA. Why and how the tobacco industry sells cigarettes to young adults: Evidence from industry documents. American Journal of Public Health. 2002;92:908–916. doi: 10.2105/ajph.92.6.908. http://dx.doi.org/10.2105/AJPH.92.6.908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Lipkus IM, Green JD, Fesganes JR, Sedikides C. The relationship between attitudinal ambivalence and desire to quit smoking among college smokers. Journal of Applied Social Psychology. 2001;31:113–133. http://dx.doi.org/10.1111/j.1559-1816.2001.tb02485.x. [Google Scholar]
  53. Maio GR, Bell DW, Esses VM. Ambivalence and persuasion: The processing of messages about immigrant groups. Journal of Experimental Social Psychology. 1996;32:513–536. doi: 10.1006/jesp.1996.0023. http://dx.doi.org/10.1006/jesp.1996.0023. [DOI] [PubMed] [Google Scholar]
  54. McGuire WJ. Attitudes and attitude change. In: Lindzey G, Aronson E, editors. Handbook of social psychology. 3. Vol. 2. New York, NY: Random House; 1985. pp. 233–346. [Google Scholar]
  55. McVey D, Stapleton J. Can anti-smoking television advertising affect smoking behaviour? Controlled trial of the Health Education Authority for England’s anti-smoking TV campaign. Tobacco Control. 2000;9:273–282. doi: 10.1136/tc.9.3.273. http://dx.doi.org/10.1136/tc.9.3.273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Moore M. Ambivalence in attitude measurement. Educational and Psychological Measurement. 1973;33:481–483. [Google Scholar]
  57. Moore M. Validation of attitude toward any practice scale through use of ambivalence as a moderator. Educational and Psychological Measurement. 1980;40:205–208. [Google Scholar]
  58. Newby-Clark IR, McGregor I, Zanna MP. Thinking and caring about cognitive inconsistency: When and for whom does attitudinal ambivalence feel uncomfortable? In: Fazio RH, Petty RE, editors. Attitudes: Their structure, function, and consequences. New York: Psychology Press; 2008. pp. 199–213. [PubMed] [Google Scholar]
  59. Newby-Clark IR, McGregor I, Zanna MP. Thinking and caring about cognitive inconsistency: When and for whom does attitudinal ambivalence feel uncomfortable? Journal of Personality and Social Psychology. 2002;82:157–166. http://dx.doi.org/10.1037/0022-3514.82.2.157. [PubMed] [Google Scholar]
  60. Nordgren LF, van Harreveld F, van der Plight J. Ambivalence, discomfort, and motivated information processing. Journal of Experimental Social Psychology. 2006;42:252–258. http://dx.doi.org/10.1016/j.jesp.2005.04.004. [Google Scholar]
  61. Nisbett RE, Wilson TD. Telling more than we can know: Verbal reports on mental processes. Psychological Review. 1977;84:231–259. [Google Scholar]
  62. Parkinson CM, Hammond D, Fong GT, Borland R, Omar M, Sirirassamee B, … Thompson M. Smoking beliefs and behavior among youth in Malaysia and Thailand. American Journal of Health Behavior. 2009;33:366–375. http://dx.doi.org/10.5993/AJHB.33.4.3. [PMC free article] [PubMed] [Google Scholar]
  63. Patterson F, Lerman C, Kaufmann VG, Neuner GA, Audrain-McGovern J. Cigarette smoking practices among American college students: Review and future directions. Journal of American College Health. 2004;52:203–212. doi: 10.3200/JACH.52.5.203-212. http://dx.doi.org/10.3200/JACH.52.5.203-212. [DOI] [PubMed] [Google Scholar]
  64. Pechmann C, Goldberg M. Evaluation of ad strategies for preventing youth tobacco use. Sacramento, CA: California Tobacco-Related Disease Research Program; 1998. [Google Scholar]
  65. Pechmann C, Knight SJ. An experimental investigation of the joint effects of advertising and peers on adolescents’ beliefs and intentions about cigarette consumption. Journal of Consumer Research. 2002;29:5–19. http://dx.doi.org/10.1086/339918. [Google Scholar]
  66. Pechmann C, Ratneshwar S. The effects of antismoking and cigarette advertising on young adolescents’ perceptions of peers who smoke. Journal of Consumer Research. 1994;21:236–251. http://dx.doi.org/10.1086/209395. [Google Scholar]
  67. Pechmann C, Zhao G, Goldberg ME, Reibling ET. What to convey in antismoking advertisements for adolescents: The use of protection motivation theory to identify effective message themes. Journal of Marketing. 2003;67:1–18. http://dx.doi.org/10.1509/jmkg.67.2.1.18607. [Google Scholar]
  68. Petty RE, Brinol P. Implicit ambivalence: A meta-cognitive approach. In: Petty RE, Fazio RH, Brinol P, editors. Attitudes: Insights from the new implicit measures. New York, NY: Psychology Press; 2009. pp. 119–164. [Google Scholar]
  69. Petty RE, Briñol P, DeMarree KG. The meta-cognitive model (MCM) of attitudes: Implications for attitude measurement, change, and strength. Social Cognition. 2007;25:657–686. http://dx.doi.org/10.1521/soco.2007.25.5.657. [Google Scholar]
  70. Petty RE, Krosnick JA. Attitude strength: Antecedents and consequences. Mahwah, NJ: Erlbaum; 1995. [Google Scholar]
  71. Piko B. Smoking in adolescence do attitudes matter? Addictive Behaviors. 2001;26:201–217. doi: 10.1016/s0306-4603(00)00101-5. http://dx.doi.org/10.1016/S0306-4603(00)00101-5. [DOI] [PubMed] [Google Scholar]
  72. Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods. 2008;40:879–891. doi: 10.3758/brm.40.3.879. http://dx.doi.org/10.3758/BRM.40.3.879. [DOI] [PubMed] [Google Scholar]
  73. Preacher KJ, Rucker DD, Hayes AF. Assessing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research. 2007;42:185–227. doi: 10.1080/00273170701341316. http://dx.doi.org/10.1080/00273170701341316. [DOI] [PubMed] [Google Scholar]
  74. Priester JR. Sex, drugs, and attitude ambivalence: How feelings of evaluative tension influence alcohol use and safe sex behaviors. In: Crano WD, Burgoon M, editors. Mass media and drug prevention: Classic and contemporary theories and research. Mahwah, NJ: L. Erlbaum; 2002. pp. 145–162. [Google Scholar]
  75. Priester JR, Petty RE. The gradual threshold model of ambivalence: Relating the positive and negative bases of attitudes to subjective ambivalence. Journal of Personality and Social Psychology. 1996;71:431–449. doi: 10.1037//0022-3514.71.3.431. http://dx.doi.org/10.1037/0022-3514.71.3.431. [DOI] [PubMed] [Google Scholar]
  76. Reno RR, Cialdini RB, Kallgren CA. The transsituational influence of social norms. Journal of Personality and Social Psychology. 1993;64:104–112. http://dx.doi.org/10.1037/0022-3514.64.1.104. [Google Scholar]
  77. Rigotti NA, Lee JE, Wechsler H. US college students’ use of tobacco products: Results of a national survey. Journal of the American Medical Association. 2000;284:699–705. doi: 10.1001/jama.284.6.699. http://dx.doi.org/10.1001/jama.284.6.699. [DOI] [PubMed] [Google Scholar]
  78. Rigotti NA, Moran SE, Wechsler H. US college students’ exposure to tobacco promotions: Prevalence and association with tobacco use. American Journal of Public Health. 2005;95:138–144. doi: 10.2105/AJPH.2003.026054. http://dx.doi.org/10.2105/AJPH.2003.026054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Rigotti NA, Regan S, Majchrzak NE, Knight JR, Wechsler H. Tobacco use by Massachusetts public college students: Long term effect of the Massachusetts Tobacco Control Program. Tobacco Control. 2002;11(Suppl 2):ii20–ii24. doi: 10.1136/tc.11.suppl_2.ii20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Ringold DJ. Boomerang effects in response to public health interventions: Some unintended consequences in the alcoholic beverage market. Journal of Consumer Policy. 2002;25:27–63. http://dx.doi.org/10.1023/A:1014588126336. [Google Scholar]
  81. Ritter C. Resources, behavior intentions, and drug use: A ten-year national panel analysis. Social Psychology Quarterly. 1988;51:250–264. http://dx.doi.org/10.2307/2786923. [Google Scholar]
  82. Rydell RJ, McConnell AR, Mackie DM. Consequences of discrepant explicit and implicit attitudes: Cognitive dissonance and increased information processing. Journal of Experimental Social Psychology. 2008;44:1526–1532. http://dx.doi.org/10.1016/j.jesp.2008.07.006. [Google Scholar]
  83. Schleicher HE, Harris KJ, Catley D, Harrar SW, Golbeck AL. Examination of a brief Smoking Consequences Questionnaire for college students. Nicotine & Tobacco Research. 2008;10:1503–1509. doi: 10.1080/14622200802323175. http://dx.doi.org/10.1080/14622200802323175. [DOI] [PubMed] [Google Scholar]
  84. Siegel JT, Alvaro EM, Burgoon M. Perceptions of the at-risk nonsmoker: Are potential intervention topics being overlooked? Journal of Adolescent Health. 2003;33:458–461. doi: 10.1016/s1054-139x(03)00139-3. http://dx.doi.org/10.1016/S1054-139X(03)00139-3. [DOI] [PubMed] [Google Scholar]
  85. Sparks P, Conner M, James R, Shepherd R, Povey R. Ambivalence about health-related behaviours: An exploration in the domain of food choice. British Journal of Health Psychology. 2001;6:53–68. doi: 10.1348/135910701169052. http://dx.doi.org/10.1348/135910701169052. [DOI] [PubMed] [Google Scholar]
  86. Sparks P, Harris PR, Lockwood N. Predictors and predictive effects of ambivalence. British Journal of Social Psychology. 2004;43:371–383. doi: 10.1348/0144666042037980. http://dx.doi.org/10.1348/0144666042037980. [DOI] [PubMed] [Google Scholar]
  87. Sparks P, Hedderley D, Shepherd R. An investigation into the relationship between perceived control, attitude variability and the consumption of two common foods. European Journal of Social Psychology. 1992;22:55–71. http://dx.doi.org/10.1002/ejsp.2420220107. [Google Scholar]
  88. Substance Abuse and Mental Health Services Administration. Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2013. NSDUH Series H-46, HHS Publication No. (SMA) 13–4795. [Google Scholar]
  89. Suls J, Miller RL. Social comparison processes: Theoretical and empirical perspectives. Oxford: Hemisphere; 1977. [Google Scholar]
  90. Suls J, Wheeler L. Handbook of social comparison: Theory and research. Dordrecht, Netherlands: Kluwer Academic; 2000. http://dx.doi.org/10.1007/978-1-4615-4237-7. [Google Scholar]
  91. Sussman S, Dent CW, Mestel-Rauch J, Johnson CA, Hansen WB, Flay BR. Adolescent nonsmokers, triers, and regular smokers’ estimates of cigarette smoking prevalence: When do overestimations occur and by whom? Journal of Applied Social Psychology. 1988;18:537–551. http://dx.doi.org/10.1111/j.1559-1816.1988.tb00035.x. [Google Scholar]
  92. Thompson B, Thompson LA, Hymer J, Zbikowsi S, Halperin A, Jaffe R. A qualitative study of attitudes, beliefs, and practices among 40 undergraduate smokers. Journal of American College Health. 2007;56:23–28. doi: 10.3200/JACH.56.1.23-28. http://dx.doi.org/10.3200/JACH.56.1.23-28. [DOI] [PubMed] [Google Scholar]
  93. Thompson MM, Zanna MP, Griffin DW. Let’s not be indifferent about (attitude) ambivalence. In: Petty RE, Krosnick JA, editors. Attitude strength: Antecedents and consequences. Mahwah, NJ: Erlbaum; 1995. pp. 361–386. [Google Scholar]
  94. Ullrich J, Schermelleh-Engel K, Böttcher B. The moderator effect that wasn’t there: Statistical problems in ambivalence research. Journal of Personality and Social Psychology. 2008;95:774–794. doi: 10.1037/a0012709. http://dx.doi.org/10.1037/a0012709. [DOI] [PubMed] [Google Scholar]
  95. United States Department of Health and Human Services. Preventing tobacco use among young people: A report of the Surgeon-General. Atlanta, GA: USDHHS, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 1994. [Google Scholar]
  96. U.S. Department of Health and Human Services. The Health Consequences of Smoking: 50 Years of Progress. A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2014. [Google Scholar]
  97. van Harreveld F, Rutjens BT, Rotterveel M, Nordgren LF, van der Plight J. Ambivalence and decisional conflict as a cause of psychological discomfort: Feeling tense before jumping off the fence. Journal of Experimental Social Psychology. 2009;45:167–173. http://dx.doi.org/10.1016/j.jesp.2008.08.015. [Google Scholar]
  98. Wakefield M, Flay B, Nichter M, Giovino G. Effects of anti-smoking advertising on youth smoking: A review. Journal of Health Communication. 2003;8:229–247. doi: 10.1080/10810730305686. http://dx.doi.org/10.1080/10810730305686. [DOI] [PubMed] [Google Scholar]
  99. Webb TL, Sheeran P. Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychological Bulletin. 2006;132:249–268. doi: 10.1037/0033-2909.132.2.249. http://dx.doi.org/10.1037/0033-2909.132.2.249. [DOI] [PubMed] [Google Scholar]
  100. Wechsler H, Rigotti NA, Gledhill-Hoyt J, Lee H. Increased levels of cigarette use among college students: A cause for national concern. Journal of the American Medical Association. 1998;280:1673–1678. doi: 10.1001/jama.280.19.1673. http://dx.doi.org/10.1001/jama.280.19.1673. [DOI] [PubMed] [Google Scholar]
  101. Wolburg JM. College students’ responses to antismoking messages: Denial, defiance, and other boomerang effects. The Journal of Consumer Affairs. 2006;40:294–323. http://dx.doi.org/10.1111/j.1745-6606.2006.00059.x. [Google Scholar]
  102. Wolford C, Swisher JD. Behavioral intention as an indicator of drug and alcohol use. Journal of Drug Education. 1986;16:305–326. doi: 10.2190/J2AB-GAH8-PHEG-0FQU. http://dx.doi.org/10.2190/J2AB-GAH8-PHEG-0FQU. [DOI] [PubMed] [Google Scholar]
  103. Zajonc RB. On the primacy of affect. American Psychologist. 1980;39:117–123. [Google Scholar]
  104. Zhao X, Cai X. The role of ambivalence in college nonsmokers’ information seeking and information processing. Communication Research. 2008;35:298–318. http://dx.doi.org/10.1177/0093650208315959. [Google Scholar]

RESOURCES