What Do Dutch College Students Talk about When ...

What Do Dutch College Students Talk about When They Talk about Alcohol? Hanneke Hendriks, PhD Gert-Jan de Bruijn, PhD Objectives: Interpersonal communication about alcohol influences alcohol consumption, yet evidence is scarce about the content and valence of alcohol-related conversations and how these concepts predict alcohol consumption. Methods: By employing a correlational design among Dutch undergraduate students (N = 133), this study measured alcohol consumption predictors and conversational valence and occurrence regarding three topics (personal alcohol-related experiences; alcohol-related experiences of others; and alcohol-related media messages). Results: Results showed that people talk more often and more positively about (personal) alcohol-related experiences than about alcohol-related media messages. In contrast to media messages, whether and how positively people talk about alcohol-related experiences was related to several alcohol consumption determinants. Conclusions: Health promotion attempts should elicit negative conversations about alcohol-related experiences, thereby resulting in more healthy alcohol consumption predictors. Key words: alcohol consumption; interpersonal communication; conversational topics; conversational occurrence and valence Health Behavior & Policy Review. 2015;2(3):232-242 DOI: http://dx.doi.org/10.14485/HBPR.2.3.8

A

lcohol abuse and heavy drinking are prevalent problems in today’s society. Among young adults, and college students in particular, there is frequent, heavy, episodic drinking.1-3 For example, Naimi et al4 demonstrated that one-third of their respondents in the United States 21-25 years of age showed heavy drinking at least once a month. These numbers are alarming given the detrimental consequences of heavy drinking. Excessive alcohol consumption has been linked to severe accidents, injuries, fights and aggression,5-7 sexual risks and abuse,8,9 brain damage,10 multiple diseases such as liver cancer and coronary heart disease,11 and increased odds of alcohol addiction and additional substance abuse.12,13 Given these consequences, excessive alcohol consumption is considered a major cause of preventable death and morbidity14 and is associated with large societal costs.15,16 Many studies have shown that interpersonal communication about health topics, such as alcohol consumption, influences health behaviors.17-21

Dorsey et al22 showed that college students talk relatively often (at least once a week) about the topic of alcohol consumption and related consequences with their peers. Whether people talk about alcohol has been shown to influence alcohol consumption determinants. For example, Real and Rimal23 showed that discussing the topic of drinking alcohol increased subsequent intentions to drink alcohol, Dorsey et al22 demonstrated that the frequency of conversations about alcohol was related to alcohol abuse, and Hendriks et al24 showed that talking versus not talking about alcohol and binge drinking could increase intentions to binge drink. Aside from the relationship between whether people talk about alcohol (conversational occurrence) and alcohol consumption predictors (attitudes, norms, and intentions regarding alcohol consumption), some evidence exists showing that how negatively or positively people talk about alcohol (conversational valence) also affects alcohol consumption predictors. Some studies about conversational valence about alcohol indicate that talking negatively about al-

Hanneke Hendriks, Department of Social and Organisational Psychology, University of Leiden, Leiden, Netherlands. Gert-Jan de Bruijn, School of Communication Research, University of Amsterdam, Amsterdam, Netherlands. Correspondence Dr Hendriks: [email protected]

232

Hendriks & Bruijn

cohol and binge drinking results in more negative attitudes and social norms towards binge drinking, a decrease in intentions to binge drink, and a decrease in actual binge drinking behaviors.25,26 Thus, multiple studies have shown that whether and how negatively or positively people talk about alcohol consumption influences alcohol consumption predictors. However, there is hardly any information about what people talk about specifically when they discuss the topic of alcohol consumption. Moreover, it is unclear how negatively or positively different alcohol-related topics are discussed. Furthermore, whether the discussion of some alcohol-related topics is more strongly related to alcohol consumption predictors as compared to other alcohol-related topics is not yet known. For example, based on self-perception theories,27,28 talking about personal alcohol-related experiences as opposed to other’s alcohol-related experiences may have a stronger effect on personal drinking attitudes and intentions. The idea that the influence of health-related conversations could depend on the particular content that is discussed also has been stressed in other research areas. In the context of political communication, it has been shown that not only how (framing) but also what (content) is being said is important.29 The present study aims to address these 3 gaps in the literature. If we are able to indicate the occurrence and valence of alcoholrelated conversational topics better, as well as the relationships between these topics and alcohol consumption predictors, then interventions that aim to reduce heavy drinking can stimulate conversations about desirable topics instead of others.

For example, Van den Putte et al37 distinguished between talking about anti-smoking messages and talking about the topic of smoking itself. Similarly, Nilsen et al38 distinguished between the topic of alcohol consumption and mediated alcohol-related information and showed that more respondents talked about (their personal) alcohol consumption (87%) as compared to about mediated alcohol-related information (5%). A few studies have further specified the concept of talking about the health topic itself by distingusing between talking about personal experiences and talking about experiences of others. In the context of skin cancer prevention, Morton and Duck39 for example showed that the frequency of interpersonal communication was more strongly related to perceptions of personal risk than to risks of others. However, they also showed that individual differences in mass media dependency played an important role herein. Therefore, Morton and Duck39 proposed that (talking about) health-related media messages, personal risks, and risks to others are part of a complicated interaction. The importance of this difference between self and others’ health-related experiences in relation to health determinants also has been confirmed in the context of alcohol consumption by Leigh.40 Although these studies provide valuable insights into the potential effects of conversations about these 3 topics (topics regarding personal alcohol-related experiences, topics regarding others’ alcohol-related experiences, and topics regarding alcohol-related media messages), none of these studies have compared how often and how negatively or positively people talk about these topics. The present study aims to address this gap.

Alcohol-related Topics Several attempts have been made to analyze the content of health-related conversations in general, and more specifically, ones related to alcohol; however, the aspects explored in these attempts have varied widely. Some studies have examined whether discussions focused on specific dangers or benefits related to unhealthy behaviors such as alcohol misuse30,31 or on the role of meaning-making and identity formation in alcohol-related conversations.32 Other studies, however, have distinguished between more broad thematic categories such as talking about health-related media messages17,33,34 and talking about the unhealthy behavior itself.35,36

Topics and Relationships with Alcohol Consumption Aside from describing how often and how negatively or positively people talk about these 3 alcohol-related topics, the present study also attempts to link talking about these topics with predictors of alcohol consumption. To our knowledge, hardly any studies have investigated whether talking about these 3 topics is differentially related to changes in health determinants. A notable exception is a study by Van den Putte et al37 who distinguished between talking about anti-smoking messages and talking about the topic of smoking cessation itself,

Health Behav Policy Rev.TM 2015;2(3):232-242

DOI: http://dx.doi.org/10.14485/HBPR.2.3.8

233

What Do Dutch College Students Talk about When They Talk about Alcohol?

and who showed that talking about the topic of smoking cessation influenced smoking intentions, whereas talking about the anti-smoking media messages hardly had any effects. Thus, talking about alcohol-related media messages is likely to be less influential in terms of alcohol consumption changes as compared to talking about alcohol-related experiences. We therefore hypothesize, H1. Talking about (personal or others’) alcohol-related experiences is more strongly related to drinking determinants (attitudes, norms, feelings of control, and intentions) than talking about alcohol-related media messages. It is as of yet unknown whether the influence of talking about personal alcohol-related experiences differs from the effect of talking about alcoholrelated experiences of others. However, to develop hypotheses on potential differences in effects, we draw from research on self-perception processes. Self-perception theory27,28 asserts that people change their personal attitudes based on the perceptions of their own behaviors. This idea also is stressed by the cognitive dissonance theory41 that emphasizes that people often adjust their private cognitions so that these are in line with executed behaviors.42 This phenomenon was also demonstrated in a classic study by Janis and King43 who showed that people change their attitudes based on a speech given by themselves, but not necessarily based on a speech given by others. In light of these findings, it can be inferred that personal behaviors or experiences are more important for attitude changes than behaviors or experiences of others. Thus, we hypothesize, H2. Talking about personal alcohol-related experiences is more strongly related to drinking determinants (attitudes, norms, feelings of control, and intentions) than talking about alcohol-related experiences of others. The Present Study The goal of this study is 3-fold. We aim to shed light on: (1) which topics are discussed when people talk about alcohol; (2) how negatively or positively people talk about these different alcohol-related topics; and (3) how talking about these different alcohol-related topics relates to changes in alcohol consumption determinants. These aims are addressed through the use of a questionnaire, distrib-

234

uted to college students, measuring conversational occurrence and conversational valence of the 3 topics, as well as drinking determinants. These drinking determinants are based on the Theory of Planned Behavior44 that posits that attitudes (negative or positive evaluations of a behavior), subjective norms (perceived social pressure to show a behavior), and feelings of control (the perception to be in control and able to conduct the behavior) predict behavioral intentions, and these intentions predict subsequent behaviors. This theory has been shown to be applicable to the context of alcohol consumption.45 METHODS Participants and Design Participants were recruited by using flyers distributed at a university in the Netherlands and contained information about the study design and the inclusion criteria. We chose to focus on participants aged 25 years and younger given the high prevalence of heavy drinking in this age group.46,47 In total, the responses of 133 participants were analyzed (35 men, 98 women, Mage = 20.84, SDage = 1.75). Of the 133 participants, 112 indicated that they had consumed alcohol during the past week, with a mean consumption of 4 alcoholic beverages per day (M = 3.97, SD = 2.33). Most participants were undergraduate students who voluntarily took part in the study for research participation credits or money. Procedure and Materials Conversational occurrence and valence regarding alcohol-related topics. Seven items assessed which topics had been discussed during the past month as well as the valence of these topics. Participants could indicate whether during the past month they had talked about: (1) the amount of alcohol you drink; (2) the things you do when you drink a lot; (3) the amount of alcohol your conversation partner drinks; (4) people you know who drink a lot; (5) people you know who drink often; (6) news messages about alcohol consumption (such as binge drinking among youngsters); and (7) governmental campaigns concerning alcohol. The first 2 items reflected talking about personal alcohol-related experiences and the third, fourth, and fifth items measured talking about alcohol-related experiences of others. The last 2 items, measuring talking about

Hendriks & Bruijn

alcohol-related media messages, were chosen because people can communicate as a result of health campaigns35 but also as a result of news messages.48 When a participants indicated that they had talked about a particular topic, they also could indicate the valence of the conversation regarding that topic on 7-point scales (1= very negative to 7 = very positive). Several subscales were created and the reliability of these subscales were assessed using Cronbach’s alpha (for > 3-item subscales) or the Spearman Brown Coefficient (for 2-item subscales49). Conversational occurrence about personal alcohol-related experiences was coded as “1” when at least one of the personal alcohol-related experiences items (item 1 or 2) was discussed, and coded as “0” when none of these topic-related items were discussed (reliability of occurrence scale = .78). The same method was applied calculating the occurrence of alcohol-related experiences of others (reliability = .78) and the occurrence of alcohol-related media messages (reliability = .70). Furthermore, one conversational valence subscale was created regarding personal alcohol-related experiences (by calculating the mean of item 1 and 2; reliability = .37), one conversational valence subscale addressed alcohol-related experiences of others (by calculating the mean of item 3, 4, and 5; reliability = .75) and one conversational valence subscale was created regarding alcohol-related media messages (by calculating the mean of item 6 and 7; reliability = .34). Thus, whereas conversational occurrence varied between 0 (= not discussed) and 1 (= discussed), conversational valence varied from 1 to 7. Drinking determinants. In line with the Theory of Planned Behavior44 we measured attitude, subjective norm, feelings of control, and intention towards drinking moderately. Drinking moderately was defined at the beginning of the questionnaire as “the consumption of a maximum of 2 (for men) or one (for women) alcoholic beverage(s) per day.” Attitude. Attitude towards drinking moderately was assessed by calculating the mean of the responses to 7 statements measured on 7-point scales (for similar items, see Hendriks et al.).26 Each statement began: “If I would drink moderately during the next 7 days, this would be…” The 7-point scales ranged from very harmless (1) to very harmful (7); very positive (1) to very negative (7); very sociable (1) to very unsociable (7); very wise (1) to very unwise

(7); very good (1) to very bad (7); very pleasant (1) to very unpleasant (7); and very healthy (1) to very unhealthy (7) (M = 3.11, SD = 0.95, reliability = .79). Subjective norm. The subjective norm related to drinking moderately was assessed with the item: “Most people who are important to me think I should drink moderately during the next seven days” answered on a 7-point scale (1 = agree completely to 7 = do not agree at all; M = 4.52, SD = 1.59). Feelings of control. Feelings of control concerning drinking moderately (from now on referred to as controllability) was assessed by calculating the mean score of 8 statements measured on 7-point scales (1 = agree completely to 7 = do not agree at all). Each statement began: “If I would drink moderately during the next 7 days…” The 8 statements were: “I would succeed in doing so,” “This would be very easy for me,” “This would cost me no effort,” “This would cost me little effort,” “I could do that,” “I could control this,” “I could decide this for myself,” and “I am confident that I would succeed in doing so” (M = 2.34, SD = 1.52, reliability = .96). These items capture feelings of internal and external control and draw from the concepts of self-efficacy and perceived behavioral control.44 Intention. The intention to drink moderately, based on Hendriks et al26 and Ajzen,44 was assessed by calculating the mean score of 3 statements: (“I intend to drink moderately during the next 7 days,” “I plan to drink moderately during the next 7 days,” and “I will try to drink moderately during the next 7 days”) answered using 7-point scales (1 = agree completely to 7 = do not agree at all; M = 3.93, SD = 1.81, reliability = .94).



Data Analysis To investigate which alcohol-related topics were most often discussed, the occurrence frequencies (0 = not discussed and 1 = discussed) were examined for the different conversational topics. To explore whether differences in occurrence between topics were significant, we used McNemar tests to contrast the binary scores across different topics. To investigate how negatively or positively people discussed alcohol-related topics, we examined mean conversational valence scores for the different topics and conducted t-tests to contrast these mean scores. To investigate the relationships between

235


Table 1 Occurrence and Valence of Conversational Topics Topic

N

Percentage of participants

M Valence

SD Valence

Personal alcohol-related experiences

106

80%

4.50a

1.19

Alcohol-related experiences of others

109

82%

3.59b

1.03

Alcohol-related media messages

46

35%

2.93

1.31

c

Note. a, b, c = Contrasts regarding valence were compared between the 3 topics with different letters indicating significant differences and identical letters indicating no significant differences.

conversational occurrence and drinking determinants, we conducted multiple hierarchical linear regression analyses separately for all drinking determinants with conversational occurrence (no/yes) about the 3 topics (about personal alcohol-related experiences, alcohol-related experiences of others, and alcohol-related media messages) entered hierarchically (based on the hypotheses) as predictors. This means that personal alcohol-related experiences were entered during the first step, alcohol-related experiences of others during the second step, and alcohol-related media messages during the third step. The relationship between conversational valence of the different topics and drinking determinants was investigated in a similar manner.

RESULTS Which Topics Are Discussed? Table 1 shows the occurrence of conversations across the different alcohol-related topics. Results showed that significantly more people talked about personal alcohol-related experiences (N = 106; 80%) and alcohol-related experiences of others (N = 109; 82%) than about alcohol-related media messages (N = 46; 35%), respectively (McNemar χ² = 52.74, p = < .001; McNemar χ² = 57.37, p = < .001). The difference between conversational occurrence about personal alcohol-related experiences versus alcohol-related experiences of others was not statistically significant (McNemar exact p = .508).

Table 2 Correlations among Main Variables Variable

1

2

3

4

5

6

7

8

9

1. Occurrence personal experiences 2. Occurrence experiences others

.784**

3. Occurrence media-messages

.249**

.259**

4. Valence personal experiences

.

-.096

.084 .001

.214*

.

-.157

a

5. Valence experiences others

.251

.

6. Valence media-messages

-.184

-.176

7. Attitude

-.192

*

-.187

*

.044

.018

.118

-.022

8. Controllability

.170*

.187*

-.036

-.021

.014

.063

.175*

9. Norms

-.188*

-.204*

.151

-.040

-.065

-.125

.214*

.078

10. Intention

-.071

.000

.086

-.055

.000

.268

.144

.233**

**

a

a

.135

*p < .05; **p < .01 Note. a = Correlation could not be computed because at least one of the variables was constant.

236

.291**

Hendriks & Bruijn

Table 3 Hierarchical Regression Results with Occurrence of Conversational Topics as Predictors and Alcohol Consumption Measures as Outcomes Attitude

Model 1:

R2

R2 Δ

.037

.037*

Personal experiences Model 2:

Controllability β

R2

R2 Δ

.029

.029*

-.192* .040

.003

Norms Β

R2

R2 Δ

.035

.035*

.170* .036

.008

Intention Β

R2

R2 Δ

.005

.005 -.071

-.188* .044

.008

β

.013

.008

Personal experiences

-.118

.060

-.074

-.185

Experiences others

-.095

.140

-.146

.145

Model 3:

.050

.010

.045

Personal experiences

-.130

Experiences others

-.112

Media-messages

.105

.008

.090 .071

.023

.046*

.010

-.101

-.197

.155

-.183

.128

-.094

.223*

.101

*p < .05; **p < .01 Note. For clarification purposes, models that did not significantly improve R2 appear in bolded red.

How Negatively or Positively are Topics Discussed? As shown in Table 1, participants talked significantly more positively about personal alcohol-related experiences (M = 4.50, SD = 1.19) than about alcohol-related experiences of others (M = 3.59, SD = 1.03, t(1, 102) = 5.99, p < .001) and about alcohol-related media messages (M = 2.93, SD = 1.31, t(1, 42) = 6.46, p < .001). Furthermore, participants talked significantly more positively about alcohol-related experiences of others than about alcohol-related media messages (t(1, 43) = 2.98, p = .005). Topics and Relationships with Alcohol Consumption Conversational occurrence. Table 2 shows the correlations between conversational occurrence and valence and alcohol consumption predictors. Multiple hierarchical linear regression analyses showed that whether people talk about personal alcoholrelated experiences (no/yes) was related to attitude (β = -.192, p = .027) and controllability (β = .170, p = .051). This indicates that talking (as opposed to not talking) about personal alcohol-related experiences is related to more positive attitudes towards drinking moderately and less control over drinking


moderately. Adding alcohol-related experiences of others and alcohol-related media messages to the model did not significantly improve the predictive value of the model for attitude and controllability (all R2Δ < .010, pΔ > .240). When looking at the predictors of subjective norms, however, the addition of alcohol-related media messages improved the model significantly (R2Δ = .046, pΔ = .012) resulting in a statistically significant relationship between norms and talking about media messages (β = .223, p = .012) and non-significant relationships between norms and talking about personal experiences (β =-.101, p = .459) and experiences of others (β = -.183, all p = .182). No statistically significant relations between talking about the 3 topics and intention were found (all β < |.197|, p = .164). Table 3 shows the results of these hierarchical regression models. Conversational valence. Multiple hierarchical linear regression analyses revealed that how negatively or positively people talk about personal alcohol-related experiences was not significantly related to any of the drinking variables (all β < |.137|, all p > .387). Adding alcohol-related experiences of others and media-related messages did not improve the models’ predictions of controllability or subjective norms (all R2Δ < .004, all pΔ > .668). Focusing on


237


attitude, however, the addition of alcohol-related experiences of others improved the model significantly (R2Δ = .099, pΔ = .045) showing a statistically significant relationship between attitude and valence about others’ alcohol-related experiences (β = .315, p = .045) and a non-significant relationship between attitude and valence about personal alcohol-related experiences (β = .060, p = .695). No statistically significant relations between valence about the 3 topics and intention were found (all β < |.097|, all p > .562). DISCUSSION The goal of our study was to investigate what people talk about when they discuss alcohol-related topics and how negatively or positively they talk about these issues. Furthermore, we aimed to shed light on the relationships between talking about these different topics and predictors of alcohol consumption. Three main findings can be distinguished. First, more participants talked about alcohol-related experiences (about themselves and others) than about alcohol-related media messages. Second, participants discussed personal alcoholrelated experiences more positively than others’ experiences or alcohol-related media messages. Third, whether and how positively participants talked about alcohol-related experiences (about themselves and others) was related to several alcohol consumption determinants. In contrast, conversational occurrence and valence about alcohol-related media messages was mostly unrelated to alcohol consumption predictors. Which topics are discussed? First, in line with H1, this study showed that participants talked more often about alcohol-related experiences than about alcohol-related media messages and that participants talked equally frequently about personal as well as others’ alcohol-related experiences. This study is one of the first to shed light on the question of what people specifically talk about when they talk about alcohol, a subject that is frequently discussed by young adults.22 Considering previous research showing that whether people talk about health-related topics influences health behaviors,21,23 the present study indicates that, due to more frequent discussions about alcohol-related experiences, conversations about experiences may have a stronger association with public health than

238

conversations about media-messages. How negatively or positively are topics discussed? Second, although participants talked equally frequently about personal and others’ alcohol-related experiences, they talked differently about themselves than about others. That is, participants talked more positively about personal alcohol-related experiences and more negatively about experiences of others and about media messages. These findings are especially relevant given the fact that conversational valence has been shown to be important for health determinants, and even more so than conversational frequency.50,51 Our findings show that people talk relatively often about personal alcohol-related experiences and, when doing so, they talk relatively positively about this topic. Considering previous evidence that a more positive conversational valence about alcohol leads to more unhealthy drinking attitudes, intentions, and behaviors,26 this finding is alarming because more positive conversations about personal alcohol-related experiences subsequently could lead to more excessive alcohol consumption. On the other hand, although many people discuss alcohol-related experiences of others, they do so in a more negative valence than when talking about personal experiences, potentially resulting in relatively more healthy drinking determinants. It is possible that people are more positive and lenient towards alcohol use of themselves (in line with positive self-biases52,53), but are more critical and less forgiving of drunken acts of others. This explanation is in line with Kane and Green’s54 study which showed that after reading an alcohol-related story about someone else, participants claimed that this person was likely to have an alcohol problem, but when that exact same story focused on themselves, they were less likely to characterize the drinking behavior as problematic. Thus, an implication of the current research can be to stimulate conversations about others’ experiences as opposed to about personal experiences. Topics and relationships with alcohol consumption. Two conclusions can be drawn regarding the relationships between conversational occurrence and valence and alcohol consumption predictors: (1) discussing personal or others’ experiences with alcohol both appear important for alcohol consumption predictors, albeit in potentially different

Hendriks & Bruijn

ways; and (2) talking about alcohol-related experiences seems more important for alcohol consumption predictors than talking about media messages. These findings provide partial support for H2. The fact that whether people talk about personal experiences - as opposed to others’ experiences - is related to alcohol attitudes and controllability is in line with research on self-persuasion and selfperception effects asserting that people change their attitudes based on perceptions of their own behaviors.39,40,42,43 Apparently, whether people talk about personal experiences is more strongly related to alcohol consumption predictors than talking about others’ experiences (or talking about media messages). However, a remarkable finding was that how negatively or positively people talk about others’ alcohol-related experiences was related to attitudes, whereas the valence of conversations regarding personal experiences was not. It is possible that the act of choosing to talk about personal alcohol-related experiences (occurrence) instead of the specific content (valence) is more important for attitude changes, whereas the valence of talking about others’ experiences is more important than the act of talking per se. This possibility needs to be explored in future research. In line with Van den Putte et al35, the results indicate that talking about the health topic itself (by discussing alcohol-related experiences) is related to health determinants whereas talking about healthrelated media messages is not. These findings seem partly in contrast with classic theories such as the diffusion of innovations theory20 and the twostep flow theory,18 which posit that people’s talking about media messages subsequently influences the impact of the media messages. Our findings indicate that: (1) people do not talk about media messages that often; (2) people talk about media messages negatively; and (3) that talking about media messages is mostly unrelated to changes in drinking determinants. Thus, one might conclude that talking about health-related media messages is not necessarily beneficial for health promotion attempts. However, it is important to note that talking about media messages may indirectly influence health determinants. That is, Van den Putte et al35 showed that, after exposure to an anti-smoking campaign, people first talk about the campaign and then talk about the smoking behavior itself. Similarly, it is possible that alcohol-related media

messages stimulate conversations about these messages which, in turn, elicit conversations about alcohol-related experiences and subsequently influence health determinants. This possibility needs to be addressed in future research.



Limitations and Future Research Suggestions Although this study provides several new insights, some limitations need to be taken into account when interpreting the results. The first limitation is related to causality. That is, alcohol consumption measures were assessed at the same time as the conversation measures. Given these cross-sectional data, we cannot be certain that conversational occurrence and valence lead to subsequent changes in attitudes, norms, feelings of control, and intentions, instead of the other way around. Although previous research has shown that conversational valence about alcohol leads to changes in drinking predictors and actual behaviors,24,25 there is also some evidence that predictors of health behaviors affect conversational valence.56 Such reciprocal effects need to be explored further in future studies. A second limitation concerns the use of selfreport measures. Although previous research has shown that people are fairly accurate reporting their drinking behaviors and related constructs,57 and although studies have shown that alcohol consumption intentions can predict actual alcohol intake,49 we cannot be certain that changes in self-reported attitudes, norms, feelings of control, and intentions found in the current study reflect differences in actual alcohol consumption. Moreover, the conversational occurrence and valence measures were also based on self-report. Such selfreport conversational valence measures have been used in previous studies and shown to be associated with more objective measures.58 However, some memory bias regarding previous conversations may still exist. Given research showing that the memory of conversations plays an important role for conversations’ effects,59,60 future research should assess the extent to which the recall of the occurrence and valence of conversational topics accurately reflects conversations that have taken place.61 Third, a notable finding was that we found fewer significant relationships between conversational valence and alcohol consumption predictors than between conversational occurrence and

239


alcohol consumption predictors. This is in contrast with evidence that conversational valence is more important than conversational occurrence or frequency.51 It is possible that the lack of statistically significant relationships between conversational valence and predictors is a result of how our variables were measured. That is, when participants indicated that they had not talked about the topics, they also did not indicate a conversational valence. As a result of this strategy, we had fewer participants for the analyses focusing on conversational valence. Future research might adapt this strategy so that more participants can be included in the analyses. Fourth, given the fact that the current study did not distinguish between conversations taking place shortly after the drinking episode versus conversations taking place more distally in the future, it is not possible to investigate whether the valence of conversations changes over time. That is, drinking heavily may be talked about negatively the following day due to a hangover (“I’ll never drink again”) but more positively after some time has passed (“Remember that fun night out we had”). Future studies should explore this possibility. IMPLICATIONS FOR HEALTH BEHAVIOR OR POLICY Our study results in several possible practical implications. For example, the result that people discuss alcohol-related experiences of others in a more negative valence than personal experiences implies that health promotion attempts should try to stimulate conversations about others’ experiences instead of personal experiences. Health promotion attempts can try to elicit conversations about others’ alcohol-related experiences by using discussion groups in which young adults talk to each other about others’ experiences with alcohol. By talking about others’ experiences with alcohol, the valence is more likely to be negative, and thus, more likely to result in healthy behaviors. An alternative approach would be to have extensive discussions about negative alcohol-related experiences. Additionally, more healthy behaviors can be stimulated by encouraging those who already talk in a desired (negative) way about their own alcohol-related experiences to discuss these experiences frequently with others,55 so that more young people have negative stories about others’ experiences with alcohol.

240

Moreover, as discussed earlier, our findings indicate that people do not talk about media messages that often and that talking about media messages is mostly unrelated to changes in drinking determinants. Therefore, it seems that talking about health-related media messages is not necessarily beneficial for health promotion attempts. Thus, efforts should be directed at changing conversations about the target behavior instead of the media messages. A potential way to achieve this can be to state explicitly in a health promotion campaign that people should talk about the targeted behavior. In the context of alcohol, this could mean adding clear sentences to an anti-alcohol leaflet or anti-alcohol poster, such as “Talk with your peers about their (negative) experiences with alcohol,” “Talk with your peers, not only about this ad, but especially about their (negative) experiences with alcohol,” or “Go beyond this message. Talk with your peers about their (negative) experiences with alcohol.” Lastly, although this study did not focus specifically on online interpersonal communication (such as chatting or social network sites), this does not mean that our findings have no implications for conversations in such forms. That is, it is possible that, in line with the current study and earlier research by Hendriks et al,24-26,58 positive online conversations about alcohol increase alcohol consumption. Given recent studies that many young adults talk about alcohol on social network sites,62 and that they frequently refer positively to alcohol,63 this increases the need to investigate whether and how such references influence drinking behaviors. If there are indeed harmful consequences of alcohol-related online conversations, health promotion attempts should try to decrease the occurrence and harmful effects of such online interpersonal communication. To summarize, our study provides some important new insights. That is, people talk more often about alcohol-related experiences than about alcohol-related media messages. Moreover, personal alcohol-related experiences are more positively discussed than others’ experiences or alcohol-related media messages. Whether and how positively people talk about alcohol-related experiences was related to several alcohol consumption determinants; however, talking about alcohol-related media mes-

Hendriks & Bruijn

sages was generally unrelated to alcohol consumption predictors. Thus, health promotion attempts should aim especially at eliciting conversations about alcohol-related experiences that are negative about alcohol consumption. Human Subjects Approval Statement The study was approved by the ethics committee of the Amsterdam School of Communication, University of Amsterdam (NR=2012-CW-53). Conflict of Interest Declaration The authors declare no conflicts of interest. References

1. Kypri K, Cronin M, Wright CS. Do university students drink more hazardously than their non-student peers? Addiction. 2005;100:713-714. 2. O’Malley PM, Johnston LD. Epidemiology of alcohol and other drug use among American college students. J Stud Alcohol Suppl. 2002;14:23-39. 3. Wechsler H, Davenport A, Dowdall G, et al. Health and behavioral consequences of binge drinking in college: a national survey of students at 140 campuses. JAMA. 1994;272:1672-1677. 4. Naimi TS, Brewer RD, Mokdad A, et al. Binge drinking among US adults. JAMA. 2003;289:70-75. 5. Hughes K, Anderson Z, Morleo M, Bellis MA. Alcohol, nightlife and violence: the relative contributions of drinking before and during nights out to negative health and criminal justice outcomes. Addiction. 2008;103:60-65. 6. Li G, Keyl PM, Smith GS, Baker SP. Alcohol and injury severity: reappraisal of the continuing controversy. J Trauma. 1997;42:562-569. 7. Steptoe A, Wardle J, Bages N, et al. Drinking and driving in university students: an international study of 23 countries. Psychol Health. 2004;19:527-540. 8. Conner M, Sutherland E, Kennedy F, et al. Impact of alcohol on sexual decision making: intentions to have unprotected sex. Psychol Health. 2008;23: 909-934. 9. Perkins H. Surveying the damage: a review of research on consequences of alcohol misuse in college populations. J Stud Alcohol Suppl. 2002;14:91-131. 10. Crews FT, Braun, CJ, Hoplight B, et al. Binge ethanol consumption causes differential brain damage in young adolescent rats compared with adult rats. Alcohol Clin Exp Res. 2000;24:1712-1723. 11. Rehm J, Room R, Graham K, et al. The relationship of average volume of alcohol consumption and patterns of drinking to burden of disease: an overview. Addiction. 2003;98:1209-1228. 12. Batel P, Pessione F, Maître C, Rueff B. Relationship between alcohol and tobacco dependencies among alcoholics who smoke. Addiction. 1995;90:977-980. 13. Grant BF, Dawson DA. Age at onset of alcohol use and its association with DSM-IV alcohol abuse and dependence:


results from the national longitudinal alcohol epidemiologic survey. J Subst Abuse. 1997;9:103-110. 14. Cherpitel CJ. Alcohol and injuries: a review of international emergency room studies since 1995. Drug Alcohol Rev. 2007;26:201-214. 15. Bouchery EE, Harwood HJ, Sacks JJ, et al. Economic costs of excessive alcohol consumption in the United States. Am J Prev Med. 2011;41:516-524. 16. Harwood HJ. Updating Estimates of the Economic Costs of Alcohol Abuse in the United States: Estimates, Update Methods, and Data. Bethesda, MD: US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute on Alcohol Abuse and Alcoholism; 2000. 17. Hornik R, Yanovitzky I. Using theory to design evaluations of communication campaigns: the case of the National Youth Anti-Drug Media Campaign. Commun Theory. 2003;13:204-224. 18. Katz E. The two-step flow of communication: an up-todate report of a hypothesis. Public Opin Q. 1957;21:6178. 19. Lazarsfeld PF, Berelson B, Gaudet H. The People’s Choice: How the Voter Makes up His Mind in a Presidential Campaign. New York, NY: Columbia University; 1944. 20. Rogers EM. Diffusion of Innovations. 3rd ed. New York, NY: Free Press; 1983. 21. Southwell BG, Yzer MC. The roles of interpersonal communication in mass media campaigns. In Beck CS, ed. Communication Yearbook 31. New York, NY: Erlbaum; 2007:419-462. 22. Dorsey AM, Scherer CW, Real K. The college tradition of “drink ‘til you drop”: the relationship between students’ social networks and engaging in risky behaviors. Health Commun. 1999;11:313-334. 23. Real K, Rimal RN. Friends talk to friends about drinking: exploring the role of peer communication in the theory of normative social behavior. Health Commun. 2007;22:169-180. 24. Hendriks H, Van den Putte B, de Bruijn G-J, de Vreese CH. Predicting health: the interplay between interpersonal communication and health campaigns. J Health Commun. 2014;19:625-636. 25. Hendriks H, De Bruijn G-J, Van den Putte B. Talking about alcohol consumption: health campaigns, conversational valence, and binge drinking intentions. Br J Health Psychol. 2012;17:843-853. 26. Hendriks H, Van den Putte B, De Bruijn G-J. Changing the conversation: the influence of emotions on conversational valence and alcohol consumption. Prev Sci. 2014;15:684-693. 27. Bem DJ. An experimental analysis of self-persuasion. J Exp Soc Psychol. 1965;1:199-218. 28. Wilson TD, Dunn EW. Self-knowledge: its limits, value, and potential for improvement. Annu Rev Psychol. 2004;55:493-518. 29. Semetko HA, Valkenburg PM. Framing European politics: a content analysis of press and television news. J Commun. 2000;50:93-109. 30. Aalto M, Seppä K. Usefulness, length and content of alcohol-related discussions in primary health care: the exit poll survey. Alcohol Alcohol. 2004;39:532-535. 31. Bradizza CM, Collins RL, Vincent PC, Falco DL. It does


241

What Do Dutch College Students Talk about When They Talk about Alcohol? the job: young adults discuss their malt liquor consumption. Addict Behav. 2006;31:1559-1577. 32. Sheehan M, Ridge D. “You become really close... You talk about the silly things you did, and we laugh”: the role of binge drinking in female secondary students’ lives. Subst Use Misuse. 2001;36:347-372. 33. Cremeens JL, Usdan SL, Brock-Martin A, et al. Parent-child communication to reduce heavy alcohol use among first-year college students. College Student Journal. 2008;42:152-163. 34. Ennett ST, Bauman KE, Foshee VA, et al. Parent‐child communication about adolescent tobacco and alcohol use: what do parents say and does it affect youth behavior? J Marriage Fam. 2001;63:48-62. 35. Geary CW, Burke HM, Castelnau L, et al. MTV’s “Staying Alive” global campaign promoted interpersonal communication about HIV and positive beliefs about HIV prevention. AIDS Educ Prev. 2007;19:51-67. 36. McCormick KA, Cochran NE, Back AL, et al. How primary care providers talk to patients about alcohol: a qualitative study. J Gen Intern Med. 2006;21:966-972. 37. Van den Putte B, Yzer MC, Southwell BG, et al. Interpersonal communication as an indirect pathway for the impact of antismoking media content on smoking cessation. J Health Commun. 2011;16:470-485. 38. Nilsen P, McCambridge J, Karlsson N, Bendtsen P. Brief interventions in routine health care: a population‐ based study of conversations about alcohol in Sweden. Addiction. 2011;106:1748-1756. 39. Morton TA, Duck JM. Communication and health beliefs mass and interpersonal influences on perceptions of risk to self and others. Communication Research. 2001;28:602-626. 40. Leigh BC. Beliefs about the effects of alcohol on self and others. J Stud Alcohol Drugs, 1987;48:467. 41. Festinger L. A Theory of Cognitive Dissonance (Vol. 2). Palo Alto, CA: Stanford University Press; 1962. 42. Freijy T, Kothe EJ. Dissonance‐based interventions for health behaviour change: a systematic review. Br J Health Psychol. 2013;18:310-337. 43. Janis IL, King BT. The influence of role playing on opinion change. J Abnorm Soc Psychol. 1954;49:211-218. 44. Ajzen I. The theory of planned behavior. Organ Behav Hum Decis Process. 1991;50:179-211. 45. Marcoux BC, Shope JT. Application of the theory of planned behavior to adolescent use and misuse of alcohol. Health Educ Res. 1997;12:323-331. 46. Casswell S, Pledger M, Pratap S. Trajectories of drinking from 18 to 26 years: identification and prediction. Addiction. 2002;97:1427-1437. 47. Poelen EA, Scholte RH, Engels RC, et al. Prevalence and trends of alcohol use and misuse among adolescents and young adults in the Netherlands from 1993 to 2000. Drug Alcohol Depend. 2005;79:413-421.

242

48. Nagelhout G, Willemsen M, Van den Putte B, et al. International Tobacco Control (ITC) Policy Evaluation Project. Evaluatie “in ieder roker zit een stopper” campagne, tweede nameting. The Hague, Netherlands: Stivoro; 2009. 49. Eisinga R, Grotenhuis MT, Pelzer B. The reliability of a two-item scale: Pearson, Cronbach, or Spearman-Brown? Int J Public Health. 2013;1-6. 50. Dunlop SM, Kashima Y, Wakefield M. Predictors and consequences of conversations about health promoting media messages. Commun Monogr. 2010;77: 518-539. 51. Frank LB, Chatteejee JS, Chaudhuri S, et al. Conversation and compliance: role of interpersonal discussion and social norms in public communication campaigns. J Health Comm. 2012;17:1050-1067. 52. Delhomme P. Comparing one’s driving with others’: assessment of abilities and frequency of offences. Evidence for a superior conformity of self-bias? Accid Anal Prev. 1991;23:493-508. 53. Headey B, Wearing A. The sense of relative superiority: central to well-being. Soc Indic Res. 1988;20:497-516. 54. Kane MN, Green D. Substance abuse by elders and selfenhancement bias. Educ Gerontol. 2009;35:95-120. 55. Boster FJ, Carpenter CJ, Andrews KR, Mongeau PA. Employing interpersonal influence to promote multivitamin use. Health Commun. 2012;27:399-407. 56. David C, Cappella JN, Fishbein M. The social diffusion of influence among adolescents: group interaction in a chat room environment about antidrug advertisements. Commun Theory. 2006;16:118-140. 57. Del Boca FK, Darkes J. The validity of self-reports of alcohol consumption: state of the science and challenges for research. Addiction. 2003;98:1-12. 58. Hendriks H, Van den Putte B, de Bruijn G-J. Subjective reality: the influence of perceived and objective conversational valence on binge drinking determinants. J Health Commun. In press. 59. Stafford L, Burggraf CS, Sharkey WF. Conversational memory: the effects of time, recall, mode, and memory expectancies on remembrances of natural conversations. Hum Commun Res. 1987;14:203-229. 60. Stafford L, Waldron VR, Infield LL. Actor-observer differences in conversational memory. Hum Commun Res. 1989;15:590-611. 61. Kiviniemi MT, Rothman AJ. Selective memory biases in individuals’ memory for health-related information and behavior recommendations. Psychol Health. 2006;21:247272. 62. Moreno MA, Briner LR, Williams A, Brockman L, Walker L, Christakis DA. A content analysis of displayed alcohol references on a social networking web site. J Adolesc Health. 2010;47:168-175. 63. Zonfrillo MR, Osterhoudt KC. NekNominate: a deadly, social media–based drinking dare. Clin Pediatr. 2014;53:1215.