Alcohol use among university students: Considering a positive deviance approach

Abstract

Harmful alcohol consumption among university students continues to be a significant issue. This study examined whether variables identified in the positive deviance literature would predict responsible alcohol consumption among university students. Surveyed students were categorized into three groups: abstainers, responsible drinkers and binge drinkers. Multinomial logistic regression modelling was significant (χ² = 274.49, degrees of freedom = 24, p < .001), with several variables predicting group membership. While the model classification accuracy rate (i.e. 71.2%) exceeded the proportional by chance accuracy rate (i.e. 38.4%), providing further support for the model, the model itself best predicted binge drinker membership over the other two groups.

Keywords

alcohol binge drinking drinking behaviour positive deviance students

Introduction

Alcohol and university students

Alcohol use continues to be a major problem among university students despite prevention and intervention programming (Arbour-Nicitopoulos et al., 2010; Wechsler et al., 2002). Post-secondary education students consume more alcohol in one sitting compared with their non-post-secondary education cohort (Substance Abuse and Mental Health Services Administration, 2006). Binge drinking, defined as five drinks in a row for men or four drinks in a row for women (National Institute of Alcohol Abuse and Alcoholism (NIAAA), 2002), continues to be a major campus problem despite significant physical, social and psychological consequences (American College Health Association, 2008).

Positive deviance

Students engaging in binge drinking behaviour are typically the primary people investigated through research (Wechsler et al., 2002). Limited studies have focused on university students who restrict alcohol consumption (Epler et al., 2009). This results in information on factors associated with the deviant behaviour among the high-risk drinking population but can miss key information about people who avoid the high-risk behaviour, despite still being members of that high-risk population. What contributes to low-risk behaviour despite ties to a high-risk population remains under-studied. Positive deviance (PD) shifts the research towards the segment of the population that is resisting harms typical of the norm group. This study utilized a normative PD framework which is defined as ‘the violation of or lack of conformity to normative expectations’ (Heckert and Heckert, 2004: 76). Those departing from the norms of a group, where that expectation is for negative behaviour such as excessive alcohol use among university students, are considered positive deviants (Spreitzer and Sonenshein, 2004). From this framework, the positive behaviour is more effective and healthy when compared with the normative behaviour of a population (Babalola et al., 2007). The PD model is underlined by an assets-based approach to help identify what is ‘going right in a community in order to amplify it, as opposed to focusing on what is going wrong in a community and fixing it’ (Dura and Singhal, 2009: 4). Using the PD approach, students who are part of the normative group (i.e. those students who engage in the normative behaviour of drinking) but choose to resist engaging in the normative expectations of excessive drinking are the focus of this study and are considered the positive deviants.

The PD approach has been used in several contexts to address serious psychosocial problems such as student retention programmes within educational settings (Dura and Singhal, 2009), prevention of early sexual activity among highly sexualized youth living in West African countries (Babalola et al., 2007), prevention of delinquent behaviour among youth (Wolfzorn et al., 2006), engagement of safer sexual practices among commercial sex workers (Marsh et al., 2004) and the increase in healthy pregnancy outcomes for women in which the norm is unhealthy outcomes (Fowles et al., 2005).

Studies that investigated positive deviants within the context of sexual health have identified variables that have supported positive behaviour, including a high personal commitment to the PD ideal (Babalola et al., 2007; Wolfzorn et al., 2006), perceived self-efficacy to refuse the negative deviant behaviour and a high regard for parental opinion (Babalola et al., 2007). These findings are also consistent with non-PD studies that have reported reasons why some people avoid alcohol or limit their alcohol consumption. Such reasons have included self-control (Slicker, 1997), parental pressure (Epler et al., 2009; Stritzke and Butt, 2001) and attempts at self-reform due to previous negative consequences (Greenfield et al., 1989). This study assessed the impact of these variables on student alcohol consumption in the context of a PD framework. As seen above, these variables have been shown to motivate positive deviant behaviour among this population (i.e. young adults) in relation to other types of social-health-related behaviour (e.g. sexual health) and thus have the potential to impact university student alcohol consumption.

This study

Using a PD framework, this study expanded the PD literature to include student health while addressing certain limitations from previous research that have examined university students who limit their drinking behaviour. As noted above, while significant research exists on students who engage in binge drinking behaviour (Wechsler et al., 2002), a relatively limited amount of studies have examined university students who choose to limit their drinking behaviour (Epler et al., 2009). As well, the above-mentioned studies that have examined this PD population did not combine the methodology of PD and student drinking, nor have they utilized the variables which have already been found in PD literature to maintain positive behaviour.

Hypotheses

This study addressed the following research question: do the variables personal commitment, perceived self-efficacy and family contextual factors, identified within the PD literature, predict refraining from binge drinking behaviour among university students who choose to consume alcohol? It is hypothesized that students within the limited drinking category (i.e. students who drink but do not engage in binge drinking behaviour, the positive deviant group) will demonstrate higher instances of personal commitment (H1), self-efficacy (H2) and family contextual factors (H3) than the students within the binge drinking category as defined by the NIAAA (2002). Personal commitment, self-efficacy and family contextual factors will be predictive of limited drinking behaviour (H4).

Methodology

Participants and procedure

This study utilized a cross-sectional survey design to examine student drinking behaviour. A convenience sample of 541 students from an Atlantic Canadian university, who were 17 years of age or older, were recruited. Emails were sent to students’ university accounts through several Listservs. An online survey platform was used.

Measures

The demographic questionnaire measured primarily nominal and ratio variables such as age, sex, marital status, current year in school, living status, hours of work and volunteer per week, membership in university athletics, and factors preventing alcohol use (e.g. medical condition, disability). This data was collected by asking students to indicate their personal demographic information as part of the survey. Students were given response choices within each demographic category. This data was then entered into, and analysed using, Statistical Package for the Social Science (SPSS).

Participants’ frequency and quantity of alcohol consumption were assessed using the Daily Drinking Questionnaire (DDQ; Collins et al., 1985). Participants were asked to report the number of drinks consumed and the number of hours spent consuming those drinks during each day of the week to describe their typical weekly drinking pattern for the past 30 days. The DDQ has acceptable psychometric properties (Collins and Lapp, 1992; Collins et al., 1985; Neighbors et al., 2006). Using the NIAAA’s (2002) definition of binge drinking, students were then classified into one of three alcohol consumption behaviour categories (i.e. abstaining, limited drinking or binge drinking) based on their self-responses. The abstaining category was classified as those reporting never drinking during a typical week in the last 30 days. The limited drinking category (i.e. the positive deviant group) was classified as those reporting drinking less than five drinks in a row for men and less than four drinks in a row for women, on the occasions that they did drink, during a typical week in the last 30 days, and the binge drinking category was classified as those reporting drinking five or more drinks in a row for men and four or more drinks in a row for women, during the occasions that they drank, in a course of a typical week in the last 30 days.

Personal commitment was assessed by the College Drinking Attitude Scale (CDAS; Gonzalez, 1990) which measures a student’s personal commitment to the concept of responsible drinking. The CDAS lists both responsible and irresponsible attitudes towards the use of alcohol. Higher scores on the scale indicate higher degrees of responsibility. The validity and reliability of the instrument have been demonstrated with over 4000 college students (Gonzalez, 1990). The choice to utilize a quantitative measure to investigate personal commitment parallels the quantitative work done in Wolfzorn et al.’s (2006) study which used an index based on the Social Bond Theory to measure PD students’ personal commitment to obtaining life goals through legitimate means as opposed to through unconventional methods (i.e. criminal activities).

Self-efficacy to refuse binge drinking was assessed using the Drinking Refusal Self-Efficacy Questionnaire–Revised Edition (DRSEQ-R; Oei et al., 2005). The DRSEQ-R assesses an individual’s belief in his or her ability to refuse drinking alcohol within three contexts (i.e. social pressure, emotional relief and opportunistic). The revised edition has been shown to have good construct and concurrent validity across several populations, including college students (Oei et al., 2005).

Participants’ family contextual factors were assessed using a modified version of the Motives for Abstaining from Alcohol Questionnaire (MAAQ; Stritzke and Butt, 2001). The MAAQ measures students’ personal reasons for limiting drinking (i.e. fear of negative consequences, dispositional risk, family constraints, religious constraints and indifference). Higher scores reflect higher degrees of motivation to limit their drinking due to corresponding factors. The test has adequate internal consistency for each factor and for the total scale (see Stritzke and Butt, 2001).

Analysis

Data were entered into SPSS for analysis. Frequencies, means and standard deviations were calculated for the predictors and appropriate demographic variables for each type of drinking behaviour. Spearman’s rho (rs) correlations were conducted to determine any univariate relationships between the three PD variables described above with the criterion variable, students who identified as engaging in limited drinking behaviour. Demographic variables were also assessed with the criterion variable to determine whether potential relationships existed between the type of drinking behaviour a university student engaged in and relevant demographic variables (i.e. Year in University, Living Status and Athletics), and how they may impact the relationship between the PD variables identified above and students who identified as engaging in limited drinking behaviour. The proportional odds assumption (Test of Parallel Lines) was not met with our data. Accordingly, a less restrictive model (i.e. multinomial logit model) was utilized (Borooah, 2002). Therefore, a multinomial logistic regression was conducted to examine the predictive relationship of the PD and demographic variables with the criterion variable, type of drinking. Missing data were dealt with using complete case analysis (i.e. listwise deletion; Allison, 2001).

Results

Participants

Following the NIAAA’s definition of binge drinking, participants were classified into one of three alcohol consumption categories (i.e. abstaining, n = 122; limited drinking, n = 154 and binge drinking, n = 265). Participants’ characteristics are also reported for each category (see Supplementary Table S1).

Correlations between variables

Consistent with H1–H3, personal commitment to responsible drinking (CDAS), self-efficacy to refuse drinking (DRSEQ-R (Emotional, Social, Opportunistic)) and family contextual factors (MAAQ (Family)) were significantly related to student self-reported alcohol consumption, such that higher scores on these variables were associated with less drinking (see Supplementary Table S2).

Multinomial regression analysis

Multinomial logistic regression explored a model using the PD variables and demographic variables to predict student self-reported alcohol consumption (H4). Results from the multinomial regression were supported (χ² = 274.49, degrees of freedom (df) = 24, p < .001) with several of the predictor variables and one of the demographic variables predicting group membership. The overall relationship between the predictor variables and type of drinking behaviour can be seen in Table 1. Within the model, there was a significant relationship between CDAS, DRSEQ-R (Social, Oppor-tunistic) and MAAQ (Family) in relation to alcohol consumption. Alpha coefficients for the PD variables are also provided in Table 1. While the classification accuracy rate (71.2%) exceeded the proportional by chance accuracy rate (38.4%), the model itself best predicted binge drinking behaviour (87.5%) and abstaining behaviour (65.9%), rather than limited drinking behaviour (44.2%). This provides support for H4 but shows that the model predicts binge drinking and abstinence at higher rates than limited drinking.

Table 1.

Overall relationship between the independent variables and the dependent variable and alpha coefficients for the PD variables.

Variables	Chi-square (χ²)	df	Significance	α
CDAS	17.25	2	<.001*	.570
DRSEQ-R (Emotion)	3.17	2	.205	.964
DRSEQ-R (Social)	15.48	2	<.001*	.904
DRSEQ-R (Opportunistic)	11.27	2	.004*	.918
MAAQ (Fear)	4.10	2	.129	.835
MAAQ (Family)	7.16	2	.028*	.853
MAAQ (Religious)	5.83	2	.054	.877
MAAQ (Indifference)	45.25	2	<.001*	.689
MAAQ (Risk)	20.32	2	<.001*	.921
Year in University	2.39	2	.303
Living status	3.96	2	.138
Athletics	8.12	2	.017*

PD: positive deviance; df: degrees of freedom; CDAS: College Drinking Attitude Scale; DRSEQ-R: Drinking Refusal Self-Efficacy Questionnaire–Revised Edition; MAAQ: Motives for Abstaining from Alcohol Questionnaire.

p <.01.

Results from the multinomial regression differentiating students who engaged in abstinence drinking behaviours from those who engaged in limited drinking behaviours (see Table 2) and students who engaged in binge drinking behaviours from those who engaged in limited drinking behaviours (see Table 3) are presented below.

Table 2.

Multinomial regression predicting self-reported abstaining behaviours.

Predictor	B	SE	Wald	Significance	OR	95.0% CI (OR)
Predictor	B	SE	Wald	Significance	OR	Lower	Upper
DRSEQ-R (Social)	0.16	0.05	12.40	<.001	1.17	1.07	1.28
DRSEQ-R (Opportunistic)	−0.19	0.08	5.86	.015	0.82	0.70	0.96
MAAQ (Indifference)	0.36	0.10	11.82	.001	1.43	1.17	1.75
MAAQ (Risk)	−0.11	0.03	10.20	.001	0.90	0.84	0.96

SE: standard error; OR: odds ratio; CI: confidence interval; DRSEQ-R: Drinking Refusal Self-Efficacy Questionnaire–Revised Edition; MAAQ: Motives for Abstaining from Alcohol Questionnaire.

OR indicates the likelihood of being in the abstaining group in reference to the limited drinking group. Data only shown for significant predictors that were also significant in the overall relationship (p < .05).

Table 3.

Multinomial regression predicting self-reported binge drinking behaviours.

Predictor	B	SE	Wald	Significance	OR	95.0% CI (OR)
Predictor	B	SE	Wald	Significance	OR	Lower	Upper
CDAS	−0.07	0.02	11.35	.001	0.94	0.90	0.97
MAAQ (Family)	0.14	0.05	6.85	.009	1.15	1.04	1.28
MAAQ (Indifference)	−0.31	0.08	14.56	<.001	0.73	0.62	0.86
Athletics (Participation)	0.82	0.35	5.40	.020	2.28	1.14	4.55

SE: standard error; OR: odds ratio; CI: confidence interval; DRSEQ-R: Drinking Refusal Self-Efficacy Questionnaire–Revised Edition; MAAQ: Motives for Abstaining from Alcohol Questionnaire.

OR indicates the likelihood of being in the binge drinking group in reference to the limited drinking group. Data only shown for significant predictors that were also significant in the overall relationship (p < .05).

Self-efficacy (DRSEQ-R (Opportunistic)) was a predictor for students who engaged in limited drinking behaviour, such that students who were certain they could resist drinking during opportunistic situations (e.g. after playing a sport) were more likely to identify with limited drinking alcohol consumption levels rather than abstinence levels. The results show that for each unit increase in certainty that one could resist drinking in opportunistic situations, the odds of being in the abstaining group decreased by 17.6 per cent. As well, students who had a higher degree of responsible drinking (i.e. higher values on CDAS) were more likely to engage in limited drinking, rather than binge drinking, such that for each unit increase in commitment to responsible drinking, the odds of being in the binge drinking group decreased by 6.5 per cent.

However, for self-efficacy within social situations (DRSEQ-R (Social)), students who were certain they could resist drinking during social pressure (e.g. when my friends are drinking) were more likely to identify with abstaining drinking behaviours rather than with limited drinking behaviours. The findings show that for each unit increase in certainty that one could resist drinking in social situations, the odds of being in the abstaining group increased by 17.4 per cent. Also, for MAAQ (Family), it was found that students who had a higher motivation to limit their drinking due to family factors (e.g. my family disapproves of drinking alcohol) were less likely to be in the limited drinking behaviour group, rather than the binge drinking behaviour group, such that for each unit increase in motivation to limit drinking due to family factors, the odds of being in the binge drinking behaviour group increased by 26.9 per cent. Self-efficacy within emotional situations (DRSEQ-R (Emotion)) (i.e. resisting drinking during emotional situations such as when a person is upset or angry) had no predictive value for students who engaged in limited drinking behaviour.

While MAAQ (Indifference) and MAAQ (Risk) did not comprise any of the variables under investigation, it should be noted that both were found to play a predictive role in the model. For all three alcohol consumption levels, MAAQ (Indifference) was a predictive variable. Students who had a higher motivation to limit their drinking behaviour due to indifference (e.g. I have no desire to drink alcohol) were more likely to be in the abstaining behaviour group than the limited drinking behaviour group and less likely to be in the binge drinking behaviour group rather than the limited drinking behaviour group. Therefore, a higher motivation to limit alcohol consumption was associated with lower drinking rates for participants in this study. As well, MAAQ (Risk) was also a predictor for students who engaged in limited drinking behaviour, such that students who were more motivated to limit their drinking due to dispositional risk (e.g. having a genetic condition that predisposes one to have a hard time breaking down alcohol) were more likely to identify with the limited drinking behaviour group, rather than the abstaining behaviour group.

The results showed support for H1, H2 and H3 – a relationship did exist between the PD variables under investigation and the type of drinking behaviour a student identified with. As can be seen, higher personal commitment, self-efficacy and regard for family factors were all associated with less alcohol consumption. Findings showed less support for our primary hypothesis, H4. The predictive model was most useful in classifying students who engaged in binge drinking behaviour and abstaining behaviour rather than limited drinking behaviour. Of all the PD variables, CDAS and DRSEQ-R (Opportunistic) were the only two variables found to predict limited alcohol consumption group membership, lending some support for H4. However, these relationships only showed predictive value for the limited alcohol consumption behaviour group in reference to only one of the other dependent variable levels (i.e. limited drinkers in reference to current abstainers or binge drinkers) and not both (i.e. limited drinking behaviour in reference to current abstaining behaviour and binge drinking behaviour). The remaining two self-efficacy variables, as well as the family variable, were not found to be common predictors of students who reported limited drinking behaviour which was inconsistent with H4.

Discussion

Research disproportionately focuses on university students who engage in binge drinking behaviour as compared with those students who limit their alcohol consumption (Epler et al., 2009; Wechsler et al., 2002). The authors were unaware of any PD studies that have focused on university students who have curbed their alcohol consumption. Heavy drinking within post-secondary institutions has long been a norm for this population, with students often overestimating alcohol consumption levels of their fellow students (Arbour-Nicitopoulos et al., 2010). With the serious consequences associated with this normative behaviour and the inefficiency of some past prevention programmes to reduce binge drinking (Dowdall, 2009), it is important to consider how a PD framework can help support healthier alcohol consumption practices.

Personal commitment

This study demonstrated a relationship between personal commitment to responsible drinking and students’ reported drinking behaviour, such that higher personal commitment was associated with less alcohol consumption. Babalola et al. (2007) found that youth not taking part in the group norm of high-risk sexual activity were motivated by their personal commitment to the ideal of waiting to engage in sexual activity. Likewise, Wolfzorn et al.’s (2006) study noted that students’ personal commitment to obtaining personal goals through legitimate and ethical means discouraged them from taking part in unethical or criminal behaviour.

Self-efficacy

The study found that higher self-efficacy was associated with less drinking within three contexts (i.e. social pressure, emotional relief and opportunistic) within univariate analyses. As discussed below, the nature of these relationships varied during multivariate analysis. Youth in the Babalola et al. (2007) study were motivated by their self-efficacy to refuse sex and use condoms. This was also seen in the Wolfzorn et al. (2006) study in which students’ self-efficacy to obtain life goals through ethical and legitimate means was correlated with the absence of delinquent behaviour. Research has identified self-control as a potential motive for abstaining or decreasing alcohol consumption (Greenfield et al., 1989).

Parental influence

The study showed that students who limited their binge drinking indicated a higher motivation to do so based upon family constraints (i.e. ‘my family may get upset if I drink’). Babalola et al. (2007) found that youth abstained from high-risk sexual behaviour due to parental influence, such as a father living in the same household, and youth perceptions of family disapproval. Wolfzorn et al. (2006) also found their positive deviant youth to be motivated by not wanting to disappoint family.

Predictive model of the PD variables

Results of this study showed that higher personal commitment, self-efficacy and regard for parental influence were associated with less alcohol consumption. However, the predictive model was most effective at classifying other levels of alcohol consumption (i.e. binge drinking and abstinence). Personal commitment (CDAS) and self-efficacy for opportunistic contexts (DRSEQ-R (Opportunistic)) were the only two variables found to predict limited drinking group membership. These results provide some support for what has been found in the previous PD literature (Babalola et al., 2007; Wolfzorn et al., 2006) and also in studies that have examined the reasons some people limit their alcohol consumption (Greenfield et al.,1989; Slicker, 1997). These findings are also consistent with emerging literature that focuses on knowledge–attitude–behaviour (KAB) change models and responsible drinking (e.g. Gagnon et al., 2012) in that variables such as personal commitment and self-efficacy have been shown to underlie part of the effectiveness of such approaches. In addition to this, these findings are also consistent with recent research by Conroy and de Visser (2015) which found that self-authenticity was a main theme non-drinking students gave when discussing their decision not to engage in alcohol consumption.

Contrary to previous research, parental influence was not found to predict limited drinking. Gagnon et al. (2012) found that perceived approval of parents held no bearing over students’ intention to limit their alcohol consumption. Another study that investigated factors that influenced students’ intentions to consume alcohol showed that students with higher intention to drink were more likely to mention family as disapproving of their alcohol consumption behaviour versus students with lower intention to drink (French and Cooke, 2012), which is consistent with present findings.

Implications

Enhanced knowledge of the factors explaining why students do not engage in binge drinking can support effective prevention and inter-vention policies and practices in this area. Interventions that focus on the variables that are found to differentiate students who are able to limit their alcohol consumption from those students who cannot (i.e. self-efficacy and personal commitment factors, from this study) may be able to be used to effectively reduce alcohol use among students who engage in binge drinking behaviour.

Study limitations

The major limitation to this study was the exploratory nature of this investigation. While a plethora of research exists on university students’ drinking behaviour, most studies focus on students who are engaged in the behaviour. Only a small amount of research has focused on students who are actively involved in limiting their drinking behaviour, and those studies have not investigated this sub-population from a PD framework. In addition to this, while the PD literature is growing, it too is still a relatively small body of research, with no particular studies within the field focusing on examining students’ drinking behaviours. Therefore, this study was relatively novel from both the university drinking behaviour and PD standpoint.

A second limitation in this study was the sampling method used. While the sample size for this study was ample (541 participants), the students were gathered through a non-random, convenience sampling strategy. Students were recruited through campus Listservs. As well, all students attended one Atlantic Canadian university; research has shown that university students engage in alcohol consumption above the national average (Paradis et al., 2010). In addition to this, a significant portion of our sample size consisted of female participants (76.2%), which is not consistent with the average percentage of females enrolled in post-secondary institutions in Canada in the past decade (56.5%) (Statistic Canada, 2013). Finally, we also noted that two of our reliability coefficients fell below an ideal level. Specifically, the following two scales are noteworthy in this regard: CDAS (α = .570) and MAAQ (Indifferent) (α = .689).

Concluding comments

Alcohol consumption is a major issue in university campuses. The PD framework offers a unique perspective to address this issue and is in line with a harm reduction approach. This study involved a survey of university students to assess their level of alcohol consumption along with various correlates of alcohol consumption. A prediction model was developed to predict non-binge-drinking alcohol consumption among university students. Results from this study can be used to inform existing practices such as motivational intervention strategies and campus alcohol policies. The results also contribute to the emerging PD literature.

Footnotes

Acknowledgements

We would like to thank the participants who took part in the study as well as Dr. Gerry White (Faculty of Education, Memorial University) for his contibutions to the statistical analysis of the study.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Allison

(2001) Missing Data. Thousand Oaks, CA: SAGE.

American College Health Association (2008) American College Health Association-National College health Assessment: Memorial University of Newfoundland Executive Summary Spring 2008. Baltimore, MD: American College Health Association.

Arbour-Nicitopoulos

Kwan

Lowe

. (2010) Social norms of alcohol, smoking, marijuana use within a Canadian university setting. Journal of American College Health 59: 191–196.

Babalola

Ouedraogo

Vondrasek

(2007) Motivation for late sexual debut in Côte d’Ivoire and Burkina Faso. Journal of HIV/AIDS Prevention in Children and Youth 7: 65–87.

Borooah

(ed.) (2002) Logit and Probit: Ordered and Multinomial Models (No. 138). Thousand Oaks, CA: SAGE.

Collins

Lapp

(1992) The Temptation and Restraint Inventory for measuring drinking restraint. British Journal of Addiction 87: 625–633.

Collins

Parks

Marlatt

(1985) Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psychology 53: 189–200.

Conroy

de Visser

(2015) The importance of authenticity for student non-drinkers: An interpretative phenomenological analysis. Journal of Health Psychology 20(11): 1483-1493.

Dowdall

(2009) College Drinking: Reforming a Social Problem. London: Praeger.

10.

Dura

Singhal

(2009) Will Ramon finish sixth grade? Positive deviance for student retention in Rural Argentina. Positive Deviance Wisdom Series, no. 2, pp. 1–8. Available at: www.positivedeviance.org/pdf/wisdom%20series/PDArgentina07112010.pdf (accessed 3 November 2010).

11.

Epler

Sher

Piasecki

(2009) Reasons for abstaining or limiting drinking: A developmental perspective psychology of addictive behaviors. Psychology of Addictive Behaviors 23: 428–442.

12.

Fowles

Hendricks

Walker

(2005) Identifying healthy eating strategies in low-income pregnant women: Applying a positive deviance model. Health Care for Women International 26: 807–820.

13.

French

Cooke

(2012) Using the theory of planned behaviour to understand binge drinking: The importance of beliefs for developing interventions. British Journal of Health Psychology 17: 1–17.

14.

Gagnon

Tessier

Côté

. (2012) Psychosocial factors and beliefs related to intention to not binge drink among young adults. Alcohol and Alcoholism 47: 525–532.

15.

Gonzalez

(1990) The college drinking attitude Scale: A tool for alcohol education program assessment. The International Journal of the Addictions 25: 121–131.

16.

Greenfield

Guydish

Temple

(1989) Reasons students give for limiting drinking: A factor analysis with implications for research and practice. Journal of Studies on Alcohol 50: 108–115.

17.

Heckert

(2004) Using an integrated typology of deviance to expand Merton’s anomie theory. Criminal Justice Studies 17: 75–90.

18.

Marsh

Schroeder

Dearden

. (2004) The power of positive deviance. British Medical Journal 329: 1177–1179.

19.

National Institute of Alcohol Abuse and Alcoholism (2002) What peer educators and resident advisors (RAs) need to know about college drinking. Available at: www.collegedrinkingprevention.gov/NIAAACollegeMaterials/peerBrochure.aspx (accessed 21 October 2009).

20.

Neighbors

Dillard

Lewis

. (2006) Normative misperceptions and temporal precedence of perceived norms and drinking. Journal of Studies of Alcohol 67: 290–299.

21.

Oei

TPS

Hasking

Young

(2005) Drinking Refusal Self Efficacy Questionnaire-Revised (DRSEQ-R): A new factor structure with confirmatory factor analysis. Drug and Alcohol Dependence 78: 297–307.

22.

Paradis

Demers

Picard

(2010) Alcohol consumption: A different kind of Canadian mosaic. Canadian Journal of Public Health 10: 275–280.

23.

Slicker

(1997) University students’ reasons for not drinking: Relationship to alcohol consumption level. Journal of Alcohol and Drug Education 42: 83–102.

24.

Spreitzer

Sonenshein

(2004) Toward the construct definition of positive deviance. American Behavioral Scientist 47: 828–847.

25.

Statistic Canada (2013) Public postsecondary enrolments and graduates, 2010/2011. Available at: http://www.statcan.gc.ca/daily-quotidien/130-123/dq130123a-eng.htm (accessed 28 September 2013).

26.

Stritzke

Butt

(2001) Motives for not drinking alcohol among Australian adolescents: Development and initial validation of a five-factor scale. Addictive Behaviours 26: 633–649.

27.

Substance Abuse and Mental Health Services Administration (2006) Results from the 2005 national survey on drug use and health: National findings (DHHS Publication no. SMA 06-4194). Rockville, MD: Substance Abuse and Mental Health Services Administration.

28.

Wechsler

Lee

Kuo

. (2002) Trends in college binge drinking during a period of increased prevention efforts: Findings from 4 Harvard School of Public Health College Alcohol Study Surveys: 1993-2001. Journal of American College Health 50: 203–217.

29.

Wolfzorn

Heckert

(2006) Positive deviance and social bond theory. Free Inquiry in Creative Sociology 34: 107–121.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB