Ethnicity & Health

ISSN: 1355-7858 (Print) 1465-3419 (Online) Journal homepage: http://www.tandfonline.com/loi/ceth20

Examining differences in drinking patterns among Jewish and Arab university students in Israel Sharon R. Sznitman, Shiran Bord, Wafa Elias, Anat Gesser-Edelsburg, Yoram Shiftan & Orna Baron-Epel To cite this article: Sharon R. Sznitman, Shiran Bord, Wafa Elias, Anat Gesser-Edelsburg, Yoram Shiftan & Orna Baron-Epel (2015) Examining differences in drinking patterns among Jewish and Arab university students in Israel, Ethnicity & Health, 20:6, 594-610, DOI: 10.1080/13557858.2014.961411 To link to this article: http://dx.doi.org/10.1080/13557858.2014.961411

Published online: 26 Sep 2014.

Submit your article to this journal

Article views: 58

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ceth20 Download by: [University of Pennsylvania]

Date: 05 November 2015, At: 17:56

Ethnicity & Health, 2015 Vol. 20, No. 6, 594–610, http://dx.doi.org/10.1080/13557858.2014.961411

Examining differences in drinking patterns among Jewish and Arab university students in Israel

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Sharon R. Sznitmana*, Shiran Borda, Wafa Eliasb, Anat Gesser-Edelsburga, Yoram Shiftanc and Orna Baron-Epela a School of Public Health, University of Haifa, Haifa, Israel; bShamoon College of Engineering (SCE), Beer Sheva, Israel; cTransportation Research Institute, Technion – Israel Institute of Technology, Haifa, Israel

(Received 1 April 2014; accepted 1 September 2014) Objectives. Worldwide there is a dearth of studies examining drinking patterns in Arabs and how these compare to other populations. The few studies that exist have suggested distinct drinking patterns in Arabs, with not only high rates of abstinence but also high rates of heavy drinking among current drinkers. No studies have yet examined potential socio-cognitive mechanisms that may contribute to this distinct drinking pattern. Israel represents a unique and valuable resource for studying Arab population drinking patterns because Israeli Arabs are nonimmigrants living in areas where exposure to Western lifestyles, including alcohol consumption, is prevalent. The current study was set out to examine differences in alcohol consumption in a convenience sample of 1310 Jewish and Arab students from Israeli universities and colleges and to explore alcohol expectancies as potential mediators of ethno-religious differences. Design. Logistic regressions were used to produce odds ratios and 95% confidence intervals to test differences between Jewish and Arab students on binary outcomes (lifetime, last month, and heavy drinking). Mediation of ethno-religious differences by alcohol expectancies was tested with bootstrapping procedures. Results. Results show that while Israeli Arab students tend to be more likely to abstain from alcohol than Israeli Jewish students, among current drinkers, Israeli Arab students are at a particular high risk of heavy drinking. Results also show that this is partly mediated by the expectancy that alcohol only influences the drinker at high levels of intake. Conclusion. The current study confirms distinct Arab drinking patterns found in previous studies. The present study is the first demonstration that drinking expectations mediate ethno-religious differences in heavy drinking among Israeli Arabs and Jews. This work contributes to the understanding of ethno-religious group differences in harmful drinking, potentially informing future etiologic research and public health interventions aimed at reducing alcohol-related harm. Keywords: university students; Arab and Jewish Israeli populations; ethno-religious groups; alcohol use; alcohol expectancies

Introduction Alcohol consumption and associated risk behaviors are major public health concerns across the world (WHO 2004). From a prevention and treatment perspective, it is *Corresponding author. Email: [email protected] © 2014 Taylor & Francis

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Ethnicity & Health

595

important to have knowledge about basic information pertaining to alcohol use patterns in majority and minority populations. While researchers have in general been interested in ethnic and minority drinking patterns, one overlooked and culturally distinct ethnoreligious group for whom alcohol use is generally unknown is Arab populations (Karam, Kypri, and Salamoun 2007). There are 280 million people in 22 Arab countries in North Africa and the Middle East (UN 2014), and Arabs constitute a significant and growing population in Western countries (ICM 2010; CBS 2012; Statistics Canada 2014; Australia Census Bureau 2000). Arab people’s presence worldwide makes it particularly important to understand the populations’ drinking patterns in order to assess potential treatment and prevention needs. Theoretically it can be expected that Arabs are more likely to abstain from alcohol than Western populations due to the cultural and religious position of alcohol in Islam; in the Arab world, the majority adhere to Islam, and the Quran discusses and stipulates caution against the use of alcohol (Yusuf 1983). This is in contrast to the cultural position of alcohol in the Western world, rooted in Judaism and Christianity, which embed the use of alcohol in religious ceremonies and accept moderate consumption (Michalak, Trocki, and Bond 2007; Room and Mäkelä 2000). Indeed, it has been suggested that early integration of alcohol in religious practices is partly responsible for controlled drinking and relatively low rates of alcoholism among Jews in particular (Cahalan and Cisin 1968; Luczak et al. 2002; Bales 1946; Glad 1947). The expectation of relatively low lifetime and last month drinking rates among Arabs has previously been confirmed in a study that examined US data from the National Survey on Drug Use and Health (NSDUH) and the Behavioral Risk Factor Surveillance System (BRFSS) studies (Arfken et al. 2011). Indeed, Arabs were more likely to abstain from alcohol than the majority population in both these study samples. However, Arfken et al. (2011) also report that among current drinkers, and in the NSDUH sample, Arab Americans were almost equally likely to binge drink as the majority (38 vs. 43%). Furthermore, in the BRFSS sample, and among current drinkers, Arfken et al. (2011) found that Arab Americans were slightly more likely to report binge drinking than the majority population (38 vs. 32%). In a qualitative research study examining binge drinking in Arab/Chaldean respondents in the USA, it was found that participants were particularly aware of binge drinking problems in their community (Arfken, Owens, and Said 2012). Furthermore, a US study recently found that Arab American drinking prevalence rates differed by religion, with Muslims less likely to drink than Christians (Arfken et al. 2013). Echoing findings from the USA, a study from Belarus recently found that Arab students were less likely than Slavs to drink alcohol, but alcohol-related problems were higher among Arabs than among Slavs (Welcome, Razvodovsky, and Pereverzev 2011). Israel represents a valuable resource for studying ethno-religious differences in alcohol consumption, in general, and for studying Arab population drinking patterns in particular, because Israeli Arabs are nonimmigrants living in areas where exposure to Western lifestyles, including alcohol consumption, is prevalent. With its 7.8 million inhabitants, the Israeli population is mainly comprised of two groups: Jews (75.4% of population) and Arabs (20.6% of population). The Israeli Arab population is further divided into Muslims (84.1% of all Israeli Arabs), Christians (7.9% of all Israeli Arabs), and Druze (8.1% of Israeli Arabs) (CBS 2012). The Druze religion is a secret faith originating in the eleventh century (Chopra et al. 2014). It evolved in an Islamic

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

596

S.R. Sznitman et al.

atmosphere where Ishmaili preachers constructed a new dogma, in which much of the Ishmaili teachings were adopted but certain rules were modified. Only two studies, based on the same representative sample of the Israeli adult population, have previously explored drinking differences across Israeli Arabs and Jews. Results of these studies are interestingly similar to what has been found in other studies that compare Arab drinking patterns to majority Western populations. Indeed, Israeli data have found relatively low lifetime alcohol prevalence rates among Israeli Arabs compared to Jews. However, among current drinkers, Arabs have been found to be at greater risk than Jews of heavy drinking (Neumark et al. 2001; Neumark, Rahav, and Jaffe 2003). Consistent with the lack of literature on Arab drinking patterns, there is currently a dearth in the literature investigating underlying socio-cognitive mechanisms that may help explain the observed distinct drinking patterns in Arab populations. Considering that socio-cognitive factors influence alcohol use, that they are possible to modify, and that they can be part of alcohol treatment and prevention, examining these beliefs is crucial. As proposed by social learning theory (Bandura 1977), research has shown that individual differences in alcohol use is affected by alcohol expectancies. Alcohol expectancies are beliefs a person has about the effects of alcohol consumption. Alcohol expectancy theory notes that behavior is explained by an individual’s belief that a particular effect will occur as a result of performing a certain behavior. It is not relevant whether or not the expectations are correct or logical. What is important is that the expectancies a person holds are likely to influence behavior (Jones, Corbin, and Fromme 2001). Research has found that alcohol expectancies are important predictors of alcohol use, in general, (Brown 1985; Jones, Corbin, and Fromme 2001) and the quantity consumed, in particular (Chen, Grube, and Madden 1994; Labbe and Maisto 2011). For instance, after control for demographic factors, expectancies explained an additional 6% of the variance in frequency and 15% of the variance in quantity of drinking in a cross-sectional study (Mooney et al. 1987). Smaller, albeit still significant, associations were found in another cross-sectional study after control for age, gender, and attitudes toward drinking (Leigh 1989). Longitudinal studies have further found that alcohol expectancies predict changes in drinking behavior in college students (Sher et al. 1996), adolescents (Christiansen et al. 1989), and young adults (Kilbey et al. 1998). Cross-cultural differences in alcohol expectancies might explain the differences in drinking behaviors between Arab and other ethno-religious groups observed in previous studies. Yet, research on ethno-religious differences in alcohol expectancies is limited, and none have specifically focused on Arab or Jewish populations. Within the limited literature, some (Hibell et al. 2011; O’Hare 1995; Nagoshi et al. 1994; Antin et al. 2013; Oei and Jardim 2007) but not all (Room and Bullock 2002; Kline 1990) studies find that there tends to be ethnic differences in terms of expectations of behavioral and emotional changes after alcohol use and that this in turn mediates group differences in alcohol consumption.

The current study In sum, there is limited knowledge about Arab drinking patterns and ethno-religious differences in alcohol expectancies. The limited data that exist do seem to indicate that Arab drinkers may be at particular risk of harmful drinking patterns. The current study

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Ethnicity & Health

597

develops the knowledgebase further by examining whether there are ethno-religious differences in moderate and heavy drinking patterns across Jewish, Arab Christian and Arab Muslim, and Druze students in Israel. Furthermore, the study examines whether there are ethno-religious differences in alcohol expectancies that may mediate potential group differences in alcohol consumption. The focus on university students is of significant interest for prevention efforts because heavy drinking tend to be particularly high among university students (Patrick and Schulenberg 2011; Schulenberg and Maggs 2002). Furthermore, among university students, in particular, heavy drinking has been associated with criminal behavior, academic problems, injury to self and others, illness, unwanted and risky sexual experiences, impaired driving, and death (Hingson et al. 2005; Perkins 2002). Studying the Israeli student population is also important as it allows for a unique insight into studying Arabs who are exposed to Western drinking styles. Although Israel is a relatively segregated society, integration between the Israeli Jewish and Arab populations is prevalent among university students (Okun and Friedlander 2005). Indeed, while Arabs are underrepresented in the Israeli university system, there has been an increase in university-eligible matriculation rates among Israeli Arabs in recent years (Myers JDC Brookdale Institute 2010). At Israeli universities, Israeli Arab students are especially exposed to the Jewish community, in which alcohol consumption has increased during the last 20 years (Sznitman, Baron Epel, and Boker-Keinan 2013). In addition to being exposed to Jewish Israeli entertainment establishments in close proximity to Israeli universities, alcohol is commonly sold on Israeli university campuses and is typically part of campus student events, potentially increasing both young Jews’ and Arabs’ exposure to alcohol. Despite these social influences that may contribute to similarities in Israeli Arab and Jewish students’ drinking patterns, ethno-religious differences in drinking may also be expected as Israeli Jews, Arab Christians, Muslims, and Druze have distinct cultures and norms related to alcohol. While alcohol is part of Jewish and Arab Christian culture and norms (Michalak, Trocki, and Bond 2007), drinking is not the norm among Arab Muslims and Druze, as the religion prohibits it (Yusuf 1983). While previous research has failed to distinguish between Arab Christian, Muslim, and Druze Israeli populations (Neumark et al. 2001; Neumark, Rahav, and Jaffe 2003), the current study was set out to explore heterogeneity within the Israeli Arab populations and, thus, to distinguish between these groups.

Methods Data collection and sample design Data were collected between November 2011 and January 2012 from a convenience sample of 1310 students at 9 universities and colleges in Israel. The institutions were selected in order to represent universities and colleges in different locations and of different types across Israel. In Israel, there are six research universities that offer undergraduate, graduate, and doctorate studies, and several dozen colleges that do not offer doctorate degrees (Israel Science and Technology Directory 2014). Data were gathered from all of Israel’s research universities, except from one institution that refused to grant permission to conduct the study. In addition, four colleges were selected based on their large size and geographical spread (data were gathered from two colleges in the

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

598

S.R. Sznitman et al.

south, one in the center, and one in the north). Only one of the participating universities/ colleges had a specific ethno-religious affiliation as it defines itself as ‘an institution that draws upon Jewish heritage.’ The university is, however, open to everyone (including secular Jew and Arab students). Since the size of the student population differs between the participating institutes, and in order to get reasonably large sample sizes from all institutes, it was decided to recruit 0.5% of students at large institutes and 2% of students at small institutes. Additionally, we oversampled at two universities in the north with relatively large proportions of Arab students in order to increase the number of Arab respondents. Eight focus group interviews with 57 Israeli Jewish students and 6 focus group interviews with 49 Israeli Arab students were conducted in order to develop a culturally appropriate and relevant survey instrument. Data were coded to examine salient alcohol expectancies in the study population from which five alcohol expectancy items were formulated for inclusion in a cross-sectional survey along with standard questions related to demographic information and alcohol use. The questionnaires were originally developed in Hebrew; they were then translated to Arabic and back into Hebrew in order to verify translation accuracy. The questionnaire was pilot tested using responses from 30 university students in order to verify the questions’ reliability. Data were collected using trained research assistants who spoke Arabic and/or Hebrew. Jewish respondents answered the questionnaire in Hebrew, and Arab respondents could choose between Arabic and Hebrew versions of questionnaire. Research assistants chose central locations in each of the higher education institutes when asking students to fill in the self-reported paper and pencil questionnaire. More specifically, the recruitment locations were chosen according to the places at each university/college where most students were likely to pass by when changing classes (e.g. the cafeteria and classroom hallways). Thus, although this study used convenience sampling, it is reasonable to believe that the students had approximately equal opportunity of being invited to participate in the study. Before filling in the questionnaire, participants were informed that all responses would remain confidential and that they had the right to withdraw from the study at any time. Once they had indicated that they had understood the instructions, participants provided written voluntary informed consent and responded to the questionnaire. The total response rate was relatively high (73%). The study was approved by the Faculty of Social Welfare and Health Science’s ethics committee at the University of Haifa.

Measures Background variables included respondents’ age, gender, and the institution where they were interviewed. Ethno-religious group: Respondents were asked to self-identify as Jewish, Arab Muslim, Arab Christian, Druze, or other. Respondents in the ‘other’ category were excluded from the analysis (n = 43). Due to the small number of Druze respondents (n = 70) and their similarity in traditional social control of alcohol to the Muslim tradition, Druze and Arab Muslims were collapsed into one category, leading to the following coding of the ethno-religious group nominal variable: 0 = Jewish, 1 = Arab Christian, and 2 = Arab Muslim/Druze.

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Ethnicity & Health

599

Lifetime alcohol consumption was based on an item that asked respondents whether they had ever drank alcohol aside from ritual alcohol consumption (0 = no, 1 = yes). Last month alcohol use was based on an item that asked respondents how often they drank alcohol aside from ritual consumption. Responses were coded only for respondents with lifetime alcohol consumption (0 = did not drink last month, 1 = drank last month). Recent heavy drinking was based on an item that asked respondents how many drinks on average they drink on typical drinking occasions. ‘Drinks’ was defined according to single standard drink servings in Israel: a glass of wine, half a pint of beer, or a shot of hard liquor. Responses were coded only for those who reported last month alcohol consumption, and response categories were defined as having consumed five or more drinks for men and four or more drinks for women (0 = no, 1 = yes). This definition is in line with the current consensus of the definition of heavy/harmful drinking constituting the lower bound at which harms are significantly increased (Courtney and Polich 2009; Hindmarch, Kerr, and Sherwood 1991; Lane et al. 2004). Alcohol expectancies were only measured in respondents with lifetime alcohol experience. Expectancies were measured using five separate items asking respondents to disagree (= 0) or agree (= 1) to statements about how alcohol influences them: (1) alcohol does not affect me at all, (2) alcohol only has an effect on me when I drink large quantities, (3) alcohol improves my mood, (4) alcohol improves my self-confidence, and (5) alcohol makes me sleepy. Correlations and factor analysis were conducted to test whether the expectancy items formed a scale. Since no acceptable internal consistency was found (alpha < .60), the items were analyzed separately.

Analysis Descriptive and bivariate statistics were conducted to describe the analytic sample. Three multivariate logistic regressions evaluated differences in lifetime, last month, and recent heavy alcohol consumption across the ethno-religious groups after controlling for background variables. Specifically, we entered gender, age, and the data collection site as control variables. Indeed, given that the data were collected from nine different colleges and universities, all multivariate models adjusted for potential differences between institutions by including a categorical variable indicating data collection site. Next, the steps outlined by Baron and Kenny (1986) were conducted to identify potential mediating alcohol expectancy variables contributing to ethno-religious group differences in last month alcohol use and recent heavy drinking. In a first step, five separate logistic regressions were modeled to test whether there were ethno-religious differences in each of the alcohol expectancy measures, while controlling for background variables. Next we tested whether the alcohol expectancies significantly predicted last month alcohol use and recent heavy drinking independent of ethno-religious group. Baron and Kenny (1986) specify that a variable qualifies as a potential mediator only if it reaches statistical significance in both these steps and that the mediator should decrease or completely eliminate the association between the independent variable (ethno-religious group) and the dependent variable (last month alcohol use and heavy drinking). After mediators were identified, a mediation analysis was conducted in Mplus (Muthen and Muthen 2007) using the bias-corrected bootstrap resampling method in order to calculate

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

600

S.R. Sznitman et al.

confidence intervals (CI) of the mediation effects (Mackinnon, Lockwood, and Williams 2004). Overall, there were small rates of missing data on each of the individual variables (>3%), yet when all missing data were added up, analysis showed that 7.9% of respondents had missing data on independent and/or dependent variables in the models predicting lifetime and last month alcohol use, and 18.4% of the sample had missing data on variables for the heavy drinking models. Furthermore, examination of the data showed that the data were not missing at random, but rather that it was related to ethnoreligious group (Arab Muslim/Druze students were less likely than Jewish students to have missing data [odds ratio, OR, = 0.08, p = .014]) and age (older respondents were more likely to have missing data [OR = 1.16, p = .002]). Since data were not missing at random, and since the use of listwise deletion for missing data can result in biased parameter estimates and loss of useful information (Allison 2002; Rubin 1987), all analyses were conducted with five imputed data-sets obtained with multiple imputation by chained equations using ICE in STATA (StataCorp 2011). At the next step, the MIM STATA command was used to average these five sets of parameter estimates for each model and calculate a new p value, taking into account variation of standard errors across the five imputed data-sets.

Results After excluding respondents in the ‘other’ ethno-religious category, the sample consisted of 1310 students (66% Jews, 11% Arab Christians, 23% Arab Muslim/Druze) aged 17–43 (M = 23.8, SD = 2.96), and 45% were female. As shown in Table 1, Israeli Arab Christian and Muslim/Druze students were more likely to be younger than Israeli Jewish students (91% vs. 48%, OR = 0.09, p < .001 and 84% vs. 48%, OR = 0.18, p < .001). This is to be expected as mandatory army service for Jews, but not Arabs, disproportionately delay the Jews university enrolment. Past research shows that females across all ethno-religious groups in Israel tend to be slightly overrepresented at Israeli universities (Fidelman 2009). In the current sample, Arab females were slightly underrepresented, which may be due to the sensitive nature of the research topic that may have led female Arabs to disproportionately refuse participation. The Jewish students were significantly more likely to report lifetime alcohol experience than Arab Christian and Arab Muslim/Druze students (94% vs. 78%, OR = 0.23, p < .001 and 94% vs. 28%, OR = 0.18, p < .001, respectively), and among those with lifetime alcohol experience, Jews were more likely to report last month drinking than Arab Christian (85% vs. 77%, OR = 0.59, p < .05) but not Arab Muslim/Druze students (85% vs. 77%, OR = 0.03, p > .05). Among current drinkers, Arab Christian and Arab Muslim/Druze students were more likely than Jews to report recent drinking (30% vs. 9%, OR = 4.06, p < .001 and 33% vs. 9%, OR = 5.17, p < .001, respectively). Among respondents with lifetime alcohol experience Arab Christians and Arab Muslims/Druze were more likely than Jews to agree that alcohol only influences them at high levels of consumption (63% vs. 38%, OR = 2.61, p < .001 and 74% vs. 38%, OR = 4.70, p < .001, respectively) and to agree with the statement that ‘alcohol does not influence me at all’ (12% vs. 2%, OR = 3.59, p < .05 and 12% vs. 2%, OR = 4.62, p < .01, respectively). Conversely, Arab Christian and Arab Muslim/Druze students were less likely than Jews to agree that alcohol improves their mood (58% vs. 94%, OR = 0.11,

Table 1. Demographic and alcohol related characteristics of the study population by ethno-religious group (N = 1310). Israeli Arab Christian N (%)

Gender Female 433 (50) Male 430 (50) Age Younger than mean age (24 years) 417 (48) Older than the mean age (24 years) 443 (52) Lifetime drinking No 50 (6) Yes 810 (94) Last month drinkinga No 121 (15) Yes 689 (85) Recent heavy drinkingb No 624 (91) Yes 65 (9) Alcohol expectanciesa Alcohol only has an effect on me when I drink large quantities Don’t agree 503 (62) Agree 307 (38) Alcohol does not affect me at all Don’t agree 790 (98) Agree 20 (2) Alcohol improves my mood Don’t agree 52 (6) Agree 758 (94) Alcohol improves my self-confidence Don’t agree 208 (26) Agree 602 (74) Alcohol makes me sleepy Don’t agree 291 (36) Agree 519 (64)

UOR

Israeli Arab Muslim/Druze N (%)

UOR

Total N (%)

60 (42) 84 (58)

100 (33) 1.48*

593 (45) 203 (67)

2.14***

717 (55)

131 (91) 13 (9)

254 (84) 0.09***

805 (61) 49 (16)

0.18***

505 (39)

31 (22) 113 (78)

217 (72) 0.23***

300 (23) 86 (28)

0.03***

1010 (77)

26 (23) 87 (77)

20 (23) 0.59*

167 (17) 66 (77)

0.59NS

843 (83)

61 (70) 26 (30)

44 (67) 4.06***

729 (86) 22 (33)

5.17***

114 (14)

42 (37) 71 (63)

22 (26) 2.61***

567 (56) 64 (74)

4.70***

443 (44)

99 (88) 14 (12)

76 (88) 3.59*

964 (96) 10 (12)

4.62**

45 (4)

47 (42) 66 (58)

36 (42) 0.11***

136 (13) 50 (58)

0.09***

874 (87)

75 (66) 33 (34)

48 (56) 0.19***

332 (33) 38 (44)

0.27***

678 (67)

51 (45) 62 (55)

42 (49) 0.66NS

385 (38) 44 (51)

0.60*

625 (62)

601

a Only among lifetime drinkers (N = 1010). bOnly among last month drinkers (N = 843). UOR: unadjusted odds ratio comparing Arab Christians and Arab Muslims/Druze to Jews (referent category). *p < .05; **p < .01; ***p < .001, NS = not significant.

Ethnicity & Health

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Israeli Jew N (%)

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

602

S.R. Sznitman et al.

p < .001 and 58% vs. 94%, OR = 0.09, p < .001, respectively) and agree that alcohol increases their self-confidence (34% vs. 74%, OR = 0.19, p < .001 and 44% vs. 74%, OR = 0.27, p < .001, respectively). Arab Muslim/Druze students were less likely to endorse the statement ‘alcohol makes me sleepy’ (51% vs. 64%, OR = 0.60, p < .05 and OR = 0.66, p < .05), while for this expectancy there was no significant difference between Jews and Arab Christians (64% vs. 55%, OR = 0.66, p > .05). It should also be noted that Arab Muslim/Druze were less likely than Arab Christians to have lifetime alcohol experience (28% vs. 78%, OR = 0.11, p < .001). Among those with lifetime alcohol experience, Arab Muslim/Druze and Christian students were equally likely to be monthly drinkers (both ethno-religious groups had 77% last month prevalence rates, OR = 0.97, p > .05), and among recent drinkers, Arab Muslim/Druze and Christian students were equally likely to report recent heavy drinking (33% vs. 30%, OR = 1.43, p > .05).

Direct association between ethno-religious group and alcohol consumption Model 1 in Table 2 shows that after controlling for age, gender, and higher education institution, Arab Christian and Arab Muslim/Druze students were less likely than Jewish students to have drunk alcohol at least once in their lifetime (OR = 0.16, CI = 0.08, 0.32, OR = 0.01, CI = 0.01, 0.02, respectively). Model 2 shows that among students with lifetime alcohol experience, and after controlling for age, gender, and higher education institution, Arab Muslim/Druze, but not Arab Christian, students were less likely than Jewish students to report last month drinking (OR = 0.56, CI = 0.31, 1.02; OR = 0.36, CI = 0.19, 0.69, respectively). The third multivariate logistic regression model tested ethno-religious differences in recent heavy drinking among last month drinkers while controlling for age, gender, and higher education institution. Results show that Arab Christian and Arab Muslim/Druze students were more likely than their Jewish counterparts to report recent heavy drinking (see Model 3, Table 2, OR = 3.58, CI = 1.76, 7.28; OR = 4.45, CI = 2.09, 9.51, respectively). Table 2. Logistic regression predicting lifetime, last month drinking, and recent heavy drinking. Model 1: lifetime drinking OR Male 4.47 Age 1 Ethno-religious group (Israeli Jew Israeli Arab Christian 0.16 Israeli Arab Muslim/Druze 0.01 N

Model 2: last month drinkinga

Model 3: recent heavy drinkingb

CI

OR

CI

OR

CI

2.78, 7.19*** 0.93, 1.07 = referent) 0.08, 0.32*** 0.01, 0.02*** 1310

2.9 1.03

1.88, 4.36*** 0.95, 1.11

1.12 0.9

0.65, 1.94 0.81, 1.01

0.56 0.36

0.31, 1.02 0.19, 0.69** 1010

3.58 4.45

1.76, 7.28*** 2.09, 9.51*** 843

a Model was performed on lifetime drinkers only. bModel 2 was performed on current drinkers only. OR, odds ratios; CI, confidence interval. **p < .01; ***p < .001. Note: Given that the data were collected from nine different colleges and universities the models adjusted for potential differences between institutions by including a categorical variable indicating data collection site.

Ethnicity & Health

603

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Direct association between ethno-religious group and alcohol expectancies In order to examine whether alcohol expectancies mediate the ethno-religious differences in last month alcohol use and heavy drinking, we ran five different logistic regressions that tested the association between ethno-religious group and each of the alcohol expectancies while controlling for background variables. Results presented in Table 3 show that among students with lifetime alcohol experience, Arab Christian, and Arab Muslim/Druze students were more likely than Jews to agree to the statement that ‘alcohol has an effect on me only when I drink large quantities’ (Table 3, Model 1, OR = 2.24, CI = 1.36, 3.71; OR = 3.38, CI = 1.94, 5.88, respectively). Muslim Arab/Druze were also more likely than Jews to endorse the statement that ‘alcohol has no effect on me at all’ (OR = 3.29, CI = 1.01, 10.62), whereas there was no significant difference between Christian Arabs and Jews for this outcome (Table 3, Model 2, p > .05). Arab Christian and Arab Muslim/Druze students were less likely than Jewish students to agree to the statement ‘alcohol improves my mood’ (OR = 0.13, CI = 0.07, 0.29; OR = 0.11, CI = 0.06, 0.21, respectively) and the statement ‘alcohol increases my self-confidence’ (OR = 0.17, CI = 0.10, 0.29; OR = 0.24, CI = 0.14, 0.39, respectively, Table 3, Models 3 and 4). No ethno-religious group differences were found for agreeing to the statement ‘alcohol makes me sleepy’ (p > .05, Table 2, Model 5).

Mediating effect of alcohol expectancies in the association between ethno-religious group and (1) recent alcohol use and (2) recent heavy drinking A logistic regression first estimated last month use among students with lifetime alcohol experience. Ethno-religious group, the four alcohol expectancy variables that were found to be different for Arab and Jewish students in the previous analysis step, and control variables were entered as predictors. Results show that three of the four moderator variables were significantly related to last month alcohol use, and that their inclusion moved the OR for the Arab Muslim/Druze indicator toward 1. Indeed, as shown in Table 2, the OR for the Arab Muslim/Druze indicator was 0.36 when the mediator variables were excluded from the model, but including them (Table 4), results show that the corresponding OR is 0.48. To estimate the significance of this mediating effect a bootstrap mediation analysis was conducted. Results showed that the mediation effect for each of the mediators were significant (alcohol only has an effect on me when I drink large quantities, coefficient = 0.04, CI = 0.01, 0.07; alcohol improves my mood, coefficient = −0.06, CI = −0.10, −0.01; alcohol improves my self-confidence, coefficient = −.03, CI = −0.06, −0.01). A second logistic regression estimated recent heavy drinking among last month drinkers. Ethno-religious group, the four alcohol expectancy variables that were found to be different for Arab and Jewish students in the previous analysis step (Table 3), and control variables were entered as predictors. As shown in Table 4, only the statement ‘alcohol only has an effect on me when I drink large quantities’ was a significant predictor of heavy drinking once ethno-religious group and control variables were entered into the model. Results also show that the alcohol expectancy is a partial and not a full mediator in that the ORs for heavy drinking for both Arab Christians and Muslims/Druze remained significant, but reduced, when alcohol expectancies were included in the model (for Arab Christians the OR was reduced from 3.58 to 3.26 and for Arab Muslim/Druze it was reduced from 4.45 to 3.92). To estimate the significance of this mediating effect a

S.R. Sznitman et al.

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

604

Table 3. Logistic regression predicting alcohol expectancies among students who reported lifetime drinking. Model 1: alcohol only has an effect on me when I drink large quantities OR Male Age Ethno-religious group (Israeli Israeli Arab Christian Israeli Arab Muslim/Druze N

CI

2.24 1.67, 3.02*** 0.96 0.91, 1.02 Jew = referent) 2.24 1.36, 3.71** 3.38 1.94, 5.88*** 1010

Model 4: alcohol increases my self-security

Model 2: alcohol has no influence on me at all

Model 3: alcohol improves my mood

OR

CI

OR

2.9 0.95

0.77, 6.57* 0.81, 1.10

0.89 1.02

0.54, 1.44 0.96, 1.15

1.45 0.97

3.38 3.29

0.77, 14.77 1.01, 10.62* 1010

0.13 0.11

0.07, 0.24*** 0.06, 0.21*** 1010

0.17 0.24

CI

OR

CI

Model 5: alcohol makes me sleepy OR

CI

1.03, 2.04* 0.92, 1.03

0.67 1.02

0.48, 0.94* 0.97, 1.08

0.10, 0.29*** 0.14, 0.39*** 1010

0.75 0.76

0.46, 1.22 0.44, 1.30 1010

OR, odds ratios; CI, confidence interval. *p < .05; **p < .01; ***p < .001. Note: Given that the data were collected from nine different colleges and universities the models adjusted for potential differences between institutions by including a categorical variable indicating data collection site.

Ethnicity & Health

605

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Table 4. Full logistic regression model with independent and moderating variables predicting last month alcohol use (Model 1) and recent heavy drinking (Model 2).

Male Age Ethno-religious group (Israeli Jew = referent) Israeli Arab Christian Israeli Arab Muslim/Druze Alcohol only has an effect on me when I drink large quantities Alcohol has no influence on me at all Alcohol improves my mood Alcohol increases my self-confidence N

Model 1: last month alcohol use

Model 2: recent heavy alcohol use

OR

OR

2.58 1.03

CI

1.67, 4.00*** 1.01 0.95, 1.12 0.91

CI 0.58, 1.76 0.81, 1.01

0.83 0.42, 1.65 0.48 0.23, 0.98* 1.83 1.23, 2.73**

3.26 1.60, 6.63** 3.92 1.71, 9.00** 1.91 1.15, 3.16*

0.61 2.33 1.92

0.72 0.74 1.33

0.26, 1.44 0.37, 3.97** 1.28, 2.89** 1010

0.22, 2.39 0.34, 1.61 0.79, 2.26 843

OR, odds ratios; CI, confidence interval. *p < .05; **p < .01; ***p < .001. Note: Given that the data were collected from nine different colleges and universities the models adjusted for potential differences between institutions by including a categorical variable indicating data collection site.

bootstrap mediation analysis was conducted. Results showed that the mediation effect was significant for both the Arab Christian group (coefficient = 0.03, CI = 0.01, 0.24) and the Arab Muslim/Druze group (coefficient = 0.04, CI = 0.02, 0.24).

Discussion To the best of our knowledge, this is the first study to examine differences in drinking patterns between Arab and Jewish students in Israel and the potential mediating role of alcohol expectancies. The current study shows that Israeli Arab Muslim/Druze and Christian students were less likely to have lifetime alcohol experience than their Jewish counterparts. Among students with lifetime alcohol experience, Arab Muslim/Druze but not Arab Christian students were less likely than Jewish students to report last month alcohol use. Conversely, among current drinkers, the risk of heavy drinking was shown to be greater for Israeli Arab Muslim/Druze and Christian students than for Israeli Jewish students. These results are similar to ethno-religious drinking patterns found in research from over a decade ago in the Israeli adult population (Neumark et al. 2001; Neumark, Rahav, and Jaffe 2003). Thus, added assurance of the validity of both sets of findings was provided. The distinct drinking patterns are also similar to what has been found in Arab populations in the USA and Belarus (Arfken et al. 2011; Welcome, Razvodovsky, and Pereverzev 2011). The fact that similar distinct Arab drinking patterns are found across diverse geographical locations where alcohol is readily available is particularly interesting and warrants further investigation. Indeed, little is known about underlying socio-cognitive mechanisms for the drinking patterns observed. The current study signifies a first step for exploring this further as it is the first to examine alcohol expectancies as a potential mediator for relatively high heavy drinking rates among Arabs. Results show that there are ethno-religious differences in the cognitive expectation of alcohol’s effect. This

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

606

S.R. Sznitman et al.

finding is in line with the expectation that drinking patterns are tied to cultural social learning regarding alcohol expectancies and the valued and expected outcomes of alcohol consumption. The study further shows that these differences in alcohol expectancies partly mediates the relatively high Arab heavy drinking rates among Israeli students, and the relatively low rate of Arab Muslim/Druze last month drinking rate. The current study was not designed in a way that enables a deeper understanding of the development, usages, and meaning of alcohol expectancies among Israeli students. However, the current study suggests that alcohol expectancies are specific to different ethno-religious groups in Israel and that these expectancies are associated with ethnoreligious differences in alcohol consumption. Since the expectancy that alcohol only influences at high levels of intake is related to heavy drinking and partly explains the relative high risk of heavy drinking among Israeli Arab students, the current study suggests that one important avenue for alcohol prevention in the Israeli Arab student population in particular may be interventions designed to change alcohol expectancies associated with heavy drinking. It is important to note that the alcohol expectancy mediation found to be significant in this study was partial, implying that additional unmeasured variables are likely to explain the observed ethno-religious differences in drinking patterns. It is, for instance, possible that the army service, which the Jewish students are much more likely than the Arab students to do prior to entering higher education programs, makes it more likely for Jewish than Arab students to assume adult social roles by the time they enter college. This may in turn influence distinctive views on alcohol and alcohol use patterns (Chilcoat and Breslau 1996). Research is needed to examine this and other possibilities further.

Limitations This study relies on self-reported alcohol use data, which can be influenced by memory or motivational biases. It is possible that Arab students, particularly Muslims, underreport drinking as they may be more reluctant to report religiously prohibited behaviors. However, in the present study, the rate of self-reported recent heavy drinking was found to be higher among Arab students than Jewish students. Thus, any underreporting of this measure by Arab students would only serve to strengthen the observed differences. It should also be mentioned that research has generally found self-reported alcohol use to be reliable (Harrison et al. 2007). Furthermore, the underlying assumption in this study is that alcohol expectancies predict drinking behavior. However, the cross-sectional nature of the study precludes confirmation of this assumption, and it is possible that previous drinking experiences influence alcohol expectancies. Nevertheless, studies with appropriate design to detect causal relations have found that alcohol expectancies affect consumption over and above any pharmacological effects of alcohol (Christiansen et al. 1989). For instance, research has found that well-developed expectancies exists prior to any personal alcohol consumption (Gustafson 1992; Miller, Smith, and Goldman 1990) and, thus, that the primary source of alcohol expectancies appears to be based on social learning rather than pharmacological experience. Nevertheless, future research with longitudinal designs are needed to confirm casual and temporal relations between alcohol expectancies and ethnoreligious differences in drinking patterns in Israel and elsewhere.

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Ethnicity & Health

607

It also needs to be mentioned that the study is based on a convenience sample, and it is, thus, not representative of the Israeli university population. Nevertheless, the recruitment locations were chosen according to the places at each university/college where most students were likely to pass by when changing classes (e.g. the cafeteria and classroom hallways). Thus, despite the convenience sampling, it is reasonable to believe that students had approximately equal opportunity of being invited to participate in the study. Finally, the five alcohol expectancy items used in this study differ from commonly used alcohol expectancy scales (see, for instance, Nicolai, Demmel, and Moshagen 2010; Brown et al. 1980). Hitherto, population inventories of different alcohol expectancies that might be conceivably held have most commonly been examined for white Caucasian non-Arab, non-Jewish samples. Since there is no previous research on alcohol expectancies in the Arab and Jewish Israeli student population, it was appropriate to first consider expectancy themes salient in the study population rather than to use already existing scales. The expectancies that can conceivably be held in any given population can be derived using a variety of procedures (Jones, Corbin, and Fromme 2001). In the current study, the items developed were grounded in a qualitative formative research phase where individual interviews conducted with members of the target population provided information about personal views on alcohol expectancies, which in turn was developed into quantitative questionnaire items. Since this is the first study to examine alcohol expectancies in Arab and Jewish student populations, and in order to increase confidence in alcohol expectancies as an important component of ethno-religious differences in drinking patters in Israel and elsewhere, further research is needed to prove consistency in findings across different projects (Jones, Corbin, and Fromme 2001).

Conclusion Similar to what has been found in the general Israeli adult population (Neumark et al. 2001; Neumark, Rahav, and Jaffe 2003) and elsewhere (Arfken et al. 2011; Welcome, Razvodovsky, and Pereverzev 2011), Israeli Arab students tend to be more likely to abstain from alcohol when compared to Jewish students. However, among current drinkers, Israeli Arab students are at particular high risk of heavy drinking. The current study shows that this is partly mediated by the expectancy that alcohol only influences the drinker at high levels of intake. The current study makes important contributions to the public health literature. Knowledge about drinking expectancies can offer insight into reasons why people drink the way they do and inform intervention strategies. Studies based on US samples have found that interventions aimed at changing alcohol expectancies reduce alcohol consumption, particularly among male college students (Labbe and Maisto 2011). Interventions aimed at reducing university student drinking patterns through altering alcohol expectancies are needed in Israel and should be developed to address differences in alcohol expectancies across ethno-religious groups.

Funding This work was supported by the Israel National Road Safety Authority and by the Research Fund on Insurance Matters (affiliated with the Israel Insurance Association), and a Marie Curie Reintegration grant [grant number 293591].

608

S.R. Sznitman et al.

Key messages (1) The current study adds to limited evidence-base on Arab drinking patterns. (2) The current study finds distinct and potentially harmful Arab drinking patterns. (3) Drinking expectations mediate ethno-religious differences in heavy drinking among Israeli Arabs and Jews.

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

References Allison, P. 2002. Missing Data, Series Quantitative Applications in the Social Sciences. Thousand Oaks, CA: Sage. Antin, T. M., S. Lipperman-Kreda, M. J. Paschall, M. Marzell, and R. Battle. 2013. “The Role of Drinking Beliefs to Explain Ethnic Variation in Drinking Practices among U.S. College Students.” Substance Use & Misuse. doi:10.3109/10826084.2013.821659. Arfken, C. L., B. B. Arnetz, M. Fakhouri, M. J. Ventimiglia, and H. Jamil. 2011. “Alcohol Use among Arab Americans: What is the Prevalence?” Journal of Immigrant and Minority Health 13 (4): 713–718. doi:10.1007/s10903-011-9447-8. Arfken, C. L., C. L. Broadbridge, H. Jamil, and B. B. Arnetz. 2013. “Immigrant Arab Americans and Alcohol Use: Longitudinal Study.” Journal of Immigrant and Minority Health 10: 10. Arfken, C. L., D. Owens, and M. Said. 2012. “Binge Drinking among Arab/Chaldeans: An Exploratory Study.” Journal of Ethnicity in Substance Abuse 11 (4): 277–293. doi:10.1080/ 15332640.2012.735163. Australia Census Bureau. 2000. Population by Selected Ancestry Group and Region: 2000. Australia Census. Accessed April 20, 2014. http://www.censusdata.abs.gov.au/ABSNavigation/ prenav/ViewData?action=404&documentproductno=0&documenttype=Details&order=1&tabname= Details&areacode=0&issue=2006&producttype=Census%20Tables&javascript=true&textversion= false&navmapdisplayed=true&breadcrumb=POLTD&&collection=Census&period=2006&product label=Ancestry%20%28Region%29%20by%20Country%20of%20Birth%20of%20Parents&product type=Census%20Tables&method=Place%20of%20Usual%20Residence&topic=Ancestry&=. Bales, R .F. 1946. “Cultural Differences in Rates of Alcoholism.” Quarterly Journal of Studies on Alcohol Use and Misuse 6: 480–499. Bandura, A. 1977. Social Learning Theory. Englewood Cliffs, NJ: Prentice Hall. Baron, R., and D. A. Kenny. 1986. “The Moderator-mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations.” Journal of Personality and Social Psychology 5: 1173–1182. doi:10.1037/0022-3514.51.6.1173. Brown, S. A. 1985. “Expectancies versus Background in the Prediction of College Drinking Patterns.” Journal of Consulting and Clinical Psychology 53 (1): 123–130. doi:10.1037/0022-006X.53.1.123. Brown, S. A., M. S. Goldman, A. Inn, and L. R. Anderson. 1980. “Expectations of Reinforcement from Alcohol: Their Domain and Relation to Drinking Patterns.” Journal of Personality and Social Psychology 48 (4): 419–426. Cahalan, D., and I. H. Cisin. 1968. “American Drinking Practices: Summary of Finding from a National Probability Sample.” Quarterly Journal of Studies on Alcohol 29: 142. CBS. 2012. Statistical Abstract of Israel 2012. Jerusalem: Central Bureau of Statistics. Chen, M., J. W. Grube, and P. A. Madden. 1994. “Alcohol Expectancies and Adolescent Drinking: Differential Prediction of Frequency, Quantity, and Intoxication.” Addictive Behaviors 19: 521– 529. doi:10.1016/0306-4603(94)90007-8. Chilcoat, H. D., and N. Breslau. 1996. “Alcohol Disorders in Young Adulthood: Effects of Transitions into Adult Roles.” Journal of Health and Social Behavior 37 (4): 339–349. doi:10.2307/2137261. Chopra, S., D. Das, G. Lotha, S. Sinha, and G. Young. 2014. “Druze.” Encyclopaedia Britannica. http://www.britannica.com/EBchecked/topic/172195/Druze. Christiansen, B. A., G. T. Smith, P. V. Roehling, and M. S. Goldman. 1989. “Using Alcohol Expectancies to Predict Adolescent Drinking Behavior after One Year.” Journal of Consulting and Clinical Psychology 57 (1): 93–99. doi:10.1037/0022-006x.57.1.93. Courtney, K. E., and J. Polich. 2009. “Binge Drinking in Young Adults: Data, Definitions, and Determinants.” Psychological Bulletin 135 (1): 142–156. doi:10.1037/a0014414.

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

Ethnicity & Health

609

Fidelman, I. 2009. Data on Students from the Arab Sector at Higher Education Institutions Jerusalem. Israel: The Government’s Institute of Research. Glad, D. D. 1947. “Attitudes and Experiences of American-Jewish and American-Irish Male Youth as Related to Differences in Adult Rates of Inebriety.” Journal of Studies on Alcohol 8: 406–472. Gustafson, R. 1992. “The Development of Alcohol-related Expectancies from the Age of 12 to the Age of 15 for Two Swedish Adolescent Samples.” Alcoholism, Clinical and Experimental Research 16 (4): 700–704. doi:10.1111/j.1530-0277.1992.tb00664.x. Harrison, L. D., S. S. Martin, T. Enev, and D. Harrington. 2007. “Comparing Drug Testing and Selfreport of Drug Use among Youths and Young Adults in the General Population.” DHHS Publication No. SMA 07–4249, Methodology Series No. M-7. Rockville, MD: Office of Applied Studies, Substance Use and Mental Health Service Administration. Hibell, B., U. Guttormsson, S. Ahlström, O. Balakireva, T. Bjarnason, A. Kokkevi, and L. Kraus. 2011. The 2011 ESPAD Report – Substance Use among Students in 36 European Countries. Stockholm: The Swedish Council for Information on Alcohol and Other Drugs (CAN). Hindmarch, I., J. S. Kerr, and N. Sherwood. 1991. “The Effects of Alcohol and Other Drugs on Psychomotor Performance and Cognitive Function.” Alcohol and Alcoholism 26 (1): 71–79. Hingson, R., T. Heeren, M. Winter, and H. Wechsler. 2005. “Magnitude of Alcohol-related Mortality and Morbidity among U.S. College Students Ages 18–24: Changes from 1998 to 2001.” Annual Review of Public Health Communication 26: 259–279. doi:10.1146/annurev. publhealth.26.021304.144652. ICM. 2010. Intra-regional Labour Mobility in the Arab World, Facts and Figures. Cairo: International Organization for Migration. Accessed February 15, 2014. http://publications.iom. int/bookstore/free/ALO-IOM_intra-regional_labour_mobility_EN.pdf. Israel Science and Technology Directory. 2014. Academic and Regional Colleges 2014. Accessed February 12. http://www.science.co.il/Colleges.asp#Academic. Jones, B. T., W. Corbin, and K. Fromme. 2001. “A Review of Expectancy Theory and Alcohol Consumption.” Addiction 96: 57–72. doi:10.1046/j.1360-0443.2001.961575.x. Karam, E., K. Kypri, and M. Salamoun. 2007. “Alcohol Use among College Students: An International Perspective.” Current Opinion in Psychiatry 20 (3): 213–221. doi:10.1097/ YCO.0b013e3280fa836c. Kilbey, M. M., K. Downey, and N. Breslau. 1998. “Predicting the Emergence and Persistence of Alcohol Dependence in Young Adults: the Role of Expectancy and Other Risk Factors.” Experimental and Clinical Psychopharmacology 6 (2): 149–566. doi:10.1037/1064-1297.6.2.149. Kline, R. 1990. “The Relation of Alcohol Expectancies to Drinking Patterns among Alcoholics: Generalization across Gender and Race.” Journal of Studies on Alcohol 51: 175–182. Labbe, A. K., and S. A. Maisto. 2011. “Alcohol Expectancy Challenges for College Students: A Narrative Review.” Clinical Psychology Review 31 (4): 673–683. doi:http://dx.doi.org/10.1016/j. cpr.2011.02.007. Lane, S. D., D. R. Cherek, C. J. Pietras, and O. V. Tcheremissine. 2004. “Alcohol Effects on Human Risk Taking.” Psychopharmacology 172 (1): 68–77. doi:10.1007/s00213-003-1628-2. Leigh, B. C. 1989. “Confirmatory Factor Analysis of Alcohol Expectancy Scales.” Journal of Studies on Alcohol 50 (3): 268–277. Luczak, S. E., S. H. Shea, L. G. Carr, T.-K. Li, and T. L. Wall. 2002. “Binge Drinking in Jewish and Non-Jewish White College Students.” Alcoholism: Clinical and Experimental Research 26 (12): 1773–1778. doi:10.1111/j.1530-0277.2002.tb02483.x. MacKinnon, D. P., C. M. Lockwood, and J. Williams. 2004. “Confidence Limits for the Indirect Effect: Distribution of the Product and Resampling Methods.” Multivariate Behavioral Research 39: 99–128. doi:10.1207/s15327906mbr3901_4. Michalak, L., K. Trocki, and J. Bond. 2007. “Religion and Alcohol in the U.S. National Alcohol Survey: How Important Is Religion for Abstention and Drinking?” Drug and Alcohol Dependence 87 (2–3): 268–280. doi:10.1016/j.drugalcdep.2006.07.013. Miller, P. M., G. T. Smith, and M. S. Goldman. 1990. “Emergence of Alcohol Expectancies in Childhood: A Possible Critical Period.” Journal of Studies on Alcohol 51 (4): 343–349. Mooney, D. K., K. Fromme, D. R. Kivlahan, and G. A. Marlatt. 1987. “Correlates of Alcohol Consumption: Sex, Age, and Expectancies Relate Differentially to Quantity and Frequency.” Addictive Behaviors 12 (3): 235–240. doi:10.1016/0306-4603(87)90033-5. Muthen, B., and L. Muthen. 2007. Mplus (Version 5.21). Los Angeles, CA: Muthén & Muthén.

Downloaded by [University of Pennsylvania] at 17:56 05 November 2015

610

S.R. Sznitman et al.

Myers JDC Brookdale Institute. 2010. Matriculation Exams Data 2009, Special Analysis of Data from the Ministry of Education. Jerusalem: Examinations Division and Data Processing Administration. Nagoshi, C. T., T. Nakata, K. Sasano, and M. D. Wood. 1994. “Alcohol Norms, Expectancies, and Reasons for Drinking and Alcohol Use in a U.S. versus a Japanese College Sample.” Alcoholism: Clinical and Experimental Research 18 (3): 671–678. doi:10.1111/j.1530-0277.1994.tb00929.x. Neumark, Y., G. Rahav, and D. H. Jaffe. 2003. “Socio-economic Status and Binge Drinking in Israel.” Drug and Alcohol Dependence 69: 15–21. doi:10.1016/S0376-8716(02)00248-X. Neumark, Y., G. Rahav, M. Teichman, and D. Hasin. 2001. “Alcohol Drinking Patterns among Jewish and Arab Men and Women in Israel.” Journal of Studies on Alcohol 62: 443–447. Nicolai, J., R. Demmel, and M. Moshagen. 2010. “The Comprehensive Alcohol Expectancy Questionnaire: Confirmatory Factor Analysis, Scale Refinement, and Further Validation.” Journal of Personality Assessment 92 (5): 400–409. doi:10.1080/00223891.2010.497396. Oei, T. P., and C. L. Jardim. 2007. “Alcohol Expectancies, Drinking Refusal Self-efficacy and Drinking Behaviour in Asian and Australian Students.” Drug and Alcohol Dependence 87 (2–3): 281–287. doi:10.1016/j.drugalcdep.2006.08.019. O’Hare, T. 1995. “Differences in Asian and White Drinking: Consumption Level, Drinking Contexts, and Expectancies.” Addictive Behaviors 20 (2): 261–266. doi:http://dx.doi.org/ 10.1016/0306-4603(94)00061-1. Okun, B. S., and D. Friedlander. 2005. “Educational Stratification among Arabs and Jews in Israel: Historical Disadvantage, Discrimination, and Opportunity.” Population Studies 59 (2): 163–180. doi:10.1080/00324720500099405. Patrick, M. E., and J. E. Schulenberg. 2011. “How Trajectories of Reasons for Alcohol Use Relate to Trajectories of Binge Drinking: National Panel Data Spanning Late Adolescence to Early Adulthood.” Developmental Psychology 47 (2): 311–317. doi:10.1037/a0021939. Perkins, H. W. 2002. “Surveying the Damage: A Review of Research on Consequences of Alcohol Misuse in College Populations.” Journal of Studies on Alcohol. Supplement 14: 91–100. Room, R., and S. Bullock. 2002. “Can Alcohol Expectancies and Attributions Explain Western Europe’s North-south Gradient in Alcohol’s Role in Violence?” Contemporary Drug Problems 29: 619–648. Room, R., and K. Mäkelä. 2000. “Typologies of the Cultural Position of Drinking.” Journal of Studies on Alcohol 61: 475–483. Rubin, D. B. 1987. Multiple Imputation for Nonresponse in Surveys. New York: Wiley. Schulenberg, J. E., and J. L. Maggs. 2002. “A Developmental Perspective on Alcohol use and Heavy Drinking during Adolescence and the Transition to Young Adulthood.” Journal of Studies on Alcohol. Supplement 14: 54–70. Sher, K. J., M. D. Wood, P. K. Wood, and G. Raskin. 1996. “Alcohol Outcome Expectancies and Alcohol Use: A Latent Variable Cross-lagged Panel Study.” Journal of Abnormal Psychology 105 (4): 561–574. doi:10.1037/0021-843X.105.4.561. StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP. Statistics Canada. 2014. The Arab Community in Canada 2006. Accessed February 15. http://www. statcan.gc.ca/pub/89-621-x/2007009/4123260-eng.htm. Sznitman, R. S., O. Baron Epel, and L. Boker-Keinan. 2013. “Convergence of Drinking Patterns of Young Russian Immigrants and Veteran Israelis Decades after Mass Immigration: Results from a Bidirectional Acculturation Model.” Journal of Studies on Alcohol and Drugs 3: 437–446. UN. 2014. Demographic Yearbook, Population by Age, Sex and Urban/Rural Residence: Latest Available Year, 2002–2011. United Nations 2011. Accessed February 15. http://unstats.un.org/ unsd/demographic/products/dyb/dyb2011.htm. Welcome, M. O., Y. E. Razvodovsky, and V. A. Pereverzev. 2011. “Prevalence of Alcohol-related Problems among the Slavs and Arabs in Belarus: A University Survey.” The American Journal of Drug and Alcohol Abuse 37 (3): 189–195. doi:10.3109/00952990.2011.563334. WHO. 2004. Global Status Report on Alcohol 2004. Geneva: Department of Mental Health and Substance Abuse, World Health Organization. Yusuf, A. A. 1983. The Holy Qur’an: Text Translation and Commentary. Brentwood, MD: Amana Corp.