A Cultural and Social Cognitive Model of Differences in Acculturation
Orientations, Alcohol Expectancies, and Alcohol-Related Risk
Behaviors Among Hispanic College Students
Sabrina E. Des Rosiers,1 Seth J. Schwartz,1 Byron L. Zamboanga,2
Lindsay S. Ham,3 and Shi Huang1
1
University of Miami
Smith College
3
University of Arkansas
2
Objectives:
The present study used a cultural and social cognitive conceptual framework to investigate whether alcohol expectancies and valuations would mediate the associations between specific acculturation orientations and alcohol-related risk behaviors. Design: The sample comprised
1,527 Hispanic students attending colleges and universities in diverse regions of the United States.
Respondents completed self-report measures of Hispanic and American cultural practices; alcohol
expectancies and valuations; and self-reported frequency of hazardous alcohol use, binge drinking,
sexual activity under the influence of alcohol, driving under the influence of alcohol, and riding with
a drunk driver. Latent class analysis was used to classify participants into acculturation orientations.
Results: Results indicated that acculturation orientations were differentially associated with alcoholrelated risk outcomes, with separated bicultural and low bicultural orientations inversely related to all
of the alcohol-related risk behaviors except for riding with a drunk driver. Negative expectancy valuations were positively associated with endorsement of binge drinking and drunk driving and negative
expectancies were negatively associated with binge drinking, drunk driving, and riding with a drunk
driver. With the exception of sexual activity under the influence of alcohol, the associations between
acculturation orientations and alcohol-related risk behaviors were partially mediated by positive alcohol
expectancies. Conclusions: Our findings provided relevant data that are informative for preventing
C 2012 Wiley Periodicals, Inc. J.
alcohol and related risk behaviors among Hispanic college students. Clin. Psychol. 69:319–340, 2013.
Keywords: acculturation; alcohol expectancies; alcohol-related problems; hazardous alcohol use;
Hispanic college students
National surveys have suggested that heavy alcohol use and negative drinking consequences
are prevalent among college students (e.g., Hingson, Heeren, Winter, & Wechsler, 2005). The
availability of alcohol, the concentration of same-age peers, and the relative absence of adult
supervision in college settings make for the “perfect storm” that can increase college students’
risk for heavy and problematic drinking. These risk factors are common among traditionally aged
college students, however, there are additional factors that could exacerbate problem drinking
in the college context.
Specific ethnic groups, such as Hispanics, are especially vulnerable to alcohol-related problems (Chartier & Caetano, 2010). Hispanic college students have the second highest rates of
alcohol consumption compared with their White counterparts (Johnston, O’Malley, Bachman, & Schulenberg, 2009). However, as a group, Hispanics are more likely to experience
negative alcohol-related consequences (Mulia, Ye, Greenfield, & Zemore, 2009). Alcohol use
among Hispanics is associated with faster escalation to maladaptive short-term and long-term
∗ This
research was supported by Grant DA025694 from the National Institute on Drug Abuse.
Please address correspondence to: Sabrina E. Des Rosiers, Center for Family Studies, Department of
Epidemiology and Public Health, Leonard M. Miller School of Medicine, University of Miami, 1425 N.W.
10th Avenue, Miami, FL 33136. e-mail:[email protected]
JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 69(4), 319–340 (2013)
Published online in Wiley Online Library (wileyonlinelibrary.com/journal/jclp).
C 2012 Wiley Periodicals, Inc.
DOI: 10.1002/jclp.21859
320
Journal of Clinical Psychology, April 2013
health outcomes such as alcohol-related traffic fatalities (Caetano, Ramisetty-Mikler, Wallisch,
McGrath, & Spence, 2008) and alcohol use disorders (Chartier & Caetano, 2010).
Alcohol-related consequences are a serious public health concern; however, few studies have
examined factors associated with alcohol-related risk outcomes in Hispanic college students
(e.g., Bourdeau, Saltz, Bersamin, & Grube, 2007). Considering the elevated risk for negative
alcohol-related consequences among Hispanics, such as progression to alcohol use disorders
(Dawson et al., 2005), it is essential to examine how Hispanic college-attending emerging adults
may be at incrementally higher risk for participation in alcohol-related risk behaviors. In fact,
Arnett (2000) has suggested that emerging adulthood is a distinct period of the lifespan—one
characterized by increased freedom and exploration. Research among emerging-adult college
students has suggested that the increased involvement in alcohol-related risk behaviors (Goldman, Greenbaum, Darkes, Brandon, & Del Boca, 2011) may be partly explained by the identity
explorations (Arnett, 2000) associated with emerging adulthood. Consistent with the conceptualization of emerging adulthood as a period of increased identity exploration, the study of
cultural identity processes, such as acculturation, among Hispanic college students provides a
unique opportunity to identify (a) heterogeneity in acculturation among Hispanic college students and (b) how within-group variations in acculturation are associated with involvement in
alcohol-related risk behaviors.
By definition, acculturation is a multidimensional process that includes the identifications,
values, and practices that change through contact with a new culture or that are shaped through
growing up in an immigrant family (Berry, 1997; Lopez-Class, Castro, & Ramirez, 2011;
Schwartz, Unger, Zamboanga, & Szapocznik, 2010). Such a notion of acculturation considers the heritage and receiving cultures separately (Berry, 1997). Within this conceptualization,
specific acculturation orientations represent permutations of Hispanic and American cultural
practices (Schwartz & Zamboanga, 2008). Schwartz and Zamboanga demonstrated that heterogeneity in acculturation orientations among Hispanic emerging adults was related to a number
of cultural variables such as acculturative stress, collectivist values, familism, and discrimination.
Among many ethnic minority and immigrant groups, such as Hispanics, acculturation has
been associated with whether or not—and to what extent—individuals engage in alcohol use
and related risk behaviors (e.g., binge drinking, drunk driving, sex while under the influence of alcohol; Caetano, Ramisetty-Mikler, & Rodriguez 2008; Maldonado-Molina, Reingle,
Jennings, & Prado, 2011; Zamboanga, Raffaelli, & Horton, 2006). Although such research
has been informative in identifying associations between acculturation and alcohol-related risk
outcomes, many of these studies have adopted a unidimensional approach to acculturation—
and reliance on unidimensional approaches may have led to the inconsistencies found in the
acculturation/alcohol-related risk behavior literature. Another important knowledge gap that
remains is the understanding of mechanisms by which the acculturation orientation/alcoholrelated risk behaviors associations occur. Therefore, variables that have been shown to explain the relationship between more distal processes and alcohol-related risk behaviors should
be included as potential mediators of the link between acculturation and alcohol-related risk
behaviors.
To that effect, much work using college student samples has focused on the extent to which
alcohol-related cognitions, such as alcohol expectancies (i.e., a person’s anticipated beliefs about
the effects of alcohol consumption and their evaluations of those effects), can increase or inhibit the likelihood of drinking (Gaher & Simmons, 2007). Alcohol expectancies have been
shown to explain the mechanism by which distal processes such as family history of alcoholism
predict alcohol-related risk behaviors. The purpose of the present study was to build on this
social cognitive framework for alcohol use and on the empirical literature linking acculturation
with alcohol-related risk behaviors. In doing so, we also sought to extend Schwartz and Zamboanga (2008) by empirically deriving acculturation orientation categories from measures of
Hispanic and American cultural practices, using these categories to examine how heterogeneity
in acculturation is associated with alcohol-related risk behaviors, both directly and indirectly
through alcohol expectancies and their valuations. To increase generalizability, we utilized a
sample of Hispanic emerging adults attending 30 colleges and universities around the United
States.
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
321
Acculturation and Alcohol Use and Related Problems
The last decade of research has vastly expanded our understanding of acculturation. An extensive landscape of theoretical and empirical work has demonstrated that acculturation is an
important area of study (Schwartz et al., 2010) and is related to a number of outcomes, including alcohol use (Wagner, Ritt-Olson, Soto, Rodriguez, Baezconde-Garbanati, & Unger, 2008).
As aforementioned, some studies have operationalized acculturation as a unidimensional process whereby individuals are deemed “acculturated” when they replace the practices and values
of their heritage culture with those embraced by the receiving society (e.g., Ramirez, Crano,
Quist, Burgoon, Alvaro, & Grandpre, 2004). For example, Hispanic individuals would become
“Americanized” or “acculturated” by adopting cultural practices and behaviors associated with
the United States while discarding those associated with their heritage cultures. Indeed, many
health studies have used either unidimensional measurement scales or demographic proxies (e.g.,
language spoken at home, place of birth, number of years in the United States) that imply a
unidimensional view of acculturation (e.g., Allen, Elliott, Fugligni, Morales., Hambarsoomian,
& Schuster, 2008).
However, scholars in cultural psychology have recommended bidimensional approaches,
where heritage and receiving cultural practices are considered separately. Bidimensional approaches allow for the possibility of biculturalism, where individuals both embrace the receiving
culture and retain their heritage culture. Schwartz and colleagues (2010) have reviewed literature suggesting that certain orientations to acculturation may be associated differently with a
number of health outcomes, including alcohol and drug use. Consequently, studies examining
how acculturation is related to psychosocial and health outcomes should utilize bidimensional
models of acculturation. Following Berry’s definition (1997; Berry & Sam, 2010), we use the term
“acculturation” to subsume both American culture acquisition and heritage culture retention.
Recent literature reviews suggest that acculturation is in fact heterogeneous and that individuals may employ different acculturation orientations depending on the receiving context
in which they live (Lopez-Class et al., 2011). For example, Hispanic adolescents who live in
highly Hispanic areas may be more likely to retain Hispanic cultural practices compared with
Hispanic adolescents residing in more monocultural American contexts (Schwartz, Pantin, Sullivan, Prado, & Szapocznik, 2006; Schwartz, Zamboanga, & Jarvis, 2007). However, regardless
of context, a bicultural approach—where one strongly endorses both Hispanic and American
cultural practices—appears to be associated with the lowest levels of risk behavior for both adolescents and emerging adults (Coatsworth, Maldonado-Molina, Pantin, & Szapocznik, 2005).
Alcohol Expectancies and Valuations
An important line of research that evolved from cognitive-affective and social learning conceptualizations of alcohol use has identified alcohol expectancies and valuations as predictors and
correlates of alcohol-related risk behaviors. Alcohol expectancies have been defined as cognitive
representations or anticipated beliefs about the effects of alcohol use, based on relations between the physiological effects of alcohol and an individual’s learning history (Goldman, Reich,
& Darkes, 2006). Valuations have been defined as one’s beliefs about the desirability (good or
bad) of a particular expected outcome associated with drinking (Fromme, Stroot, & Kaplan,
1993). Generally speaking, positive expectations about the effects of alcohol use (e.g., disinhibition, increased sexual prowess) and favorable evaluations of these expectancies are positively
associated with consumption (Fromme et al., 1993; Ham, Stewart, Norton, & Hope, 2005).
Conversely, studies have found that negative expectations about the effects of alcohol use (e.g.,
feeling sick, cognitive impairment) can decrease or even inhibit use (Thush & Wiers, 2007).
An increasing number of researchers have directed their investigations toward understanding
the mechanisms through which alcohol expectancies may play a role in promoting a range of
alcohol-related health-risk behaviors. In general, these studies have used college samples and
have found that expectancies were a significant predictor of not only involvement in alcohol
use but also alcohol-related problems (Hatzenbuehler, Corbin, & Fromme, 2011). Although
positive expectancies have been identified as a more powerful predictor of drinking outcomes,
322
Journal of Clinical Psychology, April 2013
negative expectancies have also been shown to correlate with drinking outcomes. However, the
literature has been inconsistent regarding the direction of the associations between negative
expectancies and alcohol-related outcomes, with some studies finding positive associations (e.g.,
Neighbors, Lee, Lewis, Forros, & Larimer, 2007), others finding negative associations (e.g.,
Kuntsche, Knibbe, Gmel, & Engels, 2005), and others finding no relationship (e.g., Bot, Engels,
& Knibbe, 2005).
Alcohol expectancy valuations, whether or not specific alcohol expectancies are labeled as
“good” or “bad,” are also predictive of involvement in alcohol use. Research with college
students in general has suggested that expectancy valuations may explain variability in alcohol
use and related outcomes beyond that explained by positive and negative expectancies (Ham &
Hope, 2005; Zamboanga & Ham, 2008). Consequently, in the current study we modeled positive
and negative expectancies, along with their valuations, to construct a comprehensive model
regarding the link between acculturation orientations and the ways in which they are related to
alcohol-related risk behaviors among Hispanic college students.
The associations between expectancies and alcohol use have been demonstrated in a number of ethnic groups (e.g., Chartier, Hesselbrook, & Hesselbrook, 2009), including Hispanic
college students (Zamboanga, 2005). Additionally, some work with college students in general has also focused on links between alcohol expectancies and indices of risky alcohol use
such as pregaming (drinking before going to a social event; Zamboanga, Schwartz, Ham,
Borsari, & Van Tyne, 2010), involvement in drinking games (Zamboanga, Bean, Pietras,
& Pabon, 2005), binge drinking (Balodis, Potenza, & Olmstead, 2009), and drunk driving
(LaBrie, Kenney, Mirza, & Lac, 2011). However, the mediating role of alcohol expectancies
and valuations in the association between acculturation orientations and alcohol-related risk
behaviors (e.g., drunk driving and alcohol-related sexual activity) has not been empirically
examined. Acculturation processes can play an important role in determining Hispanics’ vulnerability to health risks (e.g., Alegrı́a et al., 2007; Schwartz, Weisskirch, et al., 2011), and,
as a result, it is essential to integrate cultural and social cognitive mechanisms to provide a
more comprehensive portrayal of vulnerability to alcohol-related risks among Hispanic college
students.
As part of such an integrative model of alcohol-related risk among Hispanic college students, it is important to examine the culturally based predictors of alcohol expectancies and
expectancy valuations. Expectancies and valuations might not only serve as social cognitive
mechanisms that directly predict alcohol use and related behaviors, but also mediate the associations between acculturation and alcohol-related risks. “More assimilated” Hispanics may
be more likely to consume alcohol and to engage in alcohol-related risk behavior compared
with their “less assimilated” counterparts (Epstein, Doyle, & Botvin, 2003). It is also likely
that more assimilated individuals may hold more favorable expectancies because they have
largely relinquished their cultural heritage, whereas individuals who have maintained their
heritage might be less likely to endorse positive alcohol expectancies. The interplay between
Hispanic and American cultural practices may predict the type of expectancies Hispanic college
students endorse. For example, Hispanic students who endorse a separated-bicultural acculturation orientation may endorse characteristics associated with emerging adulthood, such as
being self-focused and indulgent, to a lower extent compared with individuals endorsing other
acculturation orientations. Consequently, such individuals may endorse less favorable alcohol expectancies that would in turn protect them against engagement in alcohol-related risk
behaviors.
On the other hand, Hispanic students who endorse an assimilated acculturation orientation
would be more likely to endorse self-focused and hedonistic characteristics that are typical for
college students in an American context. The types of expectancies (i.e., positive and negative)
that individuals endorse have been shown to be closely related to the social and cultural context
in which they evolve (Goldman, Reich, & Darkes, 2006). The study of alcohol expectancies
as known mediators of distal factors such as acculturation orientations might elucidate how
acculturation orientations are related to alcohol-related risk behaviors among Hispanic college
students. The examination of acculturation and alcohol expectancies together may provide a
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
323
more comprehensive explanatory framework for understanding alcohol-related risks behaviors
among Hispanic college students.
Acculturation, Alcohol Expectancies, and Alcohol-Related Risk Behaviors
The literature on alcohol use and related negative outcomes among Hispanics has included
acculturation (see Zemore, 2007, for a review), as well as alcohol expectancies (Marı́n, 1996;
Mills & Caetano, 2010), as predictors. In general, studies focusing on the role of acculturation
in alcohol use behaviors have found that “assimilation” (i.e., orientation toward American cultural practices and away from Hispanic cultural practices) is positively predictive of alcohol use
(e.g., Caetano, Ramisetty-Mikler, Wallisch, et al., 2008; Resor & Cooper, 2010). In terms of
alcohol expectancies, compared with other ethnic groups, individuals of Hispanic heritage also
report greater levels of positive alcohol expectancies. Individuals with an acculturation orientation toward American cultural practices (i.e., assimilated) are more likely to endorse positive
expectancies and the endorsement of positive expectancies is more consistently a predictor of
higher levels of involvement in alcohol-related behaviors (Marı́n, 1996; Mills & Caetano, 2010).
The Present Study
The present study was designed to evaluate simultaneously the mediating role of four important
social cognitive variables (positive alcohol expectancies, negative alcohol expectancies, positive
alcohol expectancy valuations, and negative alcohol expectancy valuations) in the association
between acculturation orientations and alcohol-related risk behaviors in a sample of Hispanic
college students. The hypothesized model (see Figure 1) is based on a blend of cultural (e.g.,
Lopez-Class et al., 2011) and social cognitive perspectives (e.g., Goldman, Reich, & Darkes, 2006)
to test our main hypothesis that acculturation orientations are indirectly associated (through
alcohol expectancies and valuations) with alcohol-related problems in Hispanic college students.
To achieve this goal, first, we attempted to replicate the acculturation clusters derived by
Schwartz and Zamboanga (2008). These authors used indices of Hispanic and American cultural
practices (Berry, 1997) to empirically extract different acculturation orientations in a sample of
Hispanic students attending a large urban public university, and to evaluate whether differences
in endorsement patterns of cultural identity indices would differ across these classes. Indices of
cultural identity included cultural values and identifications such as collectivism, individualism,
Figure 1. Conceptual model of hypothesized relationships.
324
Journal of Clinical Psychology, April 2013
ethnic and U.S. identity, independence, interdependence, and familism, and six acculturation
classes emerged: (a) an “Undifferentiated” class characterized by low scores on American and
Hispanic cultural practices; (b) an “Assimilated” class characterized by high endorsement of
American cultural practices and low endorsement of Hispanic cultural practices; (c) a “Full
Bicultural” class defined by the highest scores on both Hispanic and American cultural practices; (d) a “Partial-Bicultural” class characterized by moderate scores on both Hispanic and
American cultural practices; (e) an “American-oriented bicultural” class characterized by high
levels of American cultural practices and moderate levels of Hispanic cultural practices; and (f)
a “Separated” class characterized by high scores on Hispanic cultural practices and low scores
on American cultural practices.
Given that Schwartz and Zamboanga (2008) used a sample from a single urban university, it
remains to be determined whether similar acculturation orientations would emerge in a diverse
sample of Hispanic emerging adults attending a number of colleges around the United States,
including both schools located in urban areas and schools located in college towns. Based on this
prior research (Schwartz & Zamboanga, 2008) and on acculturation theory (e.g., Berry, 1997),
we anticipated that latent-class procedures conducted on heritage and American cultural practices would yield classes representing three of Berry’s four cultural orientations: biculturalism
(retaining both heritage and receiving culture); separation (retaining the heritage culture and
rejecting the receiving culture); and assimilation (rejecting the heritage culture and acquiring the
receiving culture). We expected that multiple forms of biculturalism would emerge, with a “high
biculturalism” cluster characterized by stronger endorsement of both heritage and receiving
cultural practices compared with a “low biculturalism” cluster (Benet-Martı́nez & Haritatos,
2005).
Second, on the basis of the research reviewed above, we also hypothesized that acculturation
orientations would directly predict alcohol-related risk behaviors. We hypothesized that acculturation orientation class would predict involvement in alcohol-related risk outcomes, such that
classes that are more oriented toward heritage culture would be associated with lower levels of
involvement in alcohol-related risk outcomes. Given that alcohol expectancies (Jones, Corbin, &
Fromme, 2001) and acculturation orientations (Lopez-Classet al., 2011) are both, at least in part,
the results of prior social and cultural experiences, we expected that acculturation orientations
would be associated with alcohol expectancies and valuations. Furthermore, consistent with the
finding that being “less acculturated” (i.e., orientations other than assimilation) is associated
with lower levels of drinking (Zemore, 2007), we also expected that acculturation classes oriented toward heritage culture would be associated with less positive, and more negative, alcohol
expectancies. In turn, we expected that positive expectancies would be most likely to promote
drinking and alcohol-related risk outcomes.
Last, we examined whether the hypothesized model fit equivalently by gender and across university settings. Research using Hispanics in general has reported gender differences in alcohol
use (e.g., Zamboanga, Raffaelli, & Horton, 2006). Consequently, an auxiliary goal of the current study was to explore between-group differences by testing whether the model we posited in
Figure 1 fits equivalently for male and female Hispanic students. Using a unidimensional characterization of acculturation, many studies found that “less acculturated” Hispanic men do
not differ significantly in their mean-level of alcohol use compared with their “more acculturated” counterparts. On the other hand, “more acculturated” Hispanic women tend to report
higher levels of drinking compared with those who are less acculturated (e.g., Pearson, Dube,
Nelson, & Caetano, 2009). In light of these prior findings using unidimensional acculturation models, it is important to evaluate whether the proposed model (based on a bidimensional model of acculturation) operates differently for college-attending Hispanic men and
women.
Moreover, the type of living environments have been found to be associated with alcohol
use and related risk behaviors among college students (e.g., Labrie et al., 2011; e.g., LaBrie,
Kenney, Tehniat, & Lac, 2011; Zamboanga, Olthuis, Horton, McCollum, Lee, & Shaw, 2009).
For example, living arrangements often associated with college towns, such as dormitories and
fraternity/sorority houses, tend to be associated with drinking and related problems (e.g., Labrie
et al., 2011). Such findings raise the question whether university setting (i.e., college town versus
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
325
urban/suburban) may also contribute to differences in the links between acculturation and
alcohol-related risk related outcomes. Consequently, we also evaluated the consistency of the
hypothesized model between college towns and urban/suburban settings).
Method
Sample and Procedures
For the present analyses, we used data from the Multi-Site University Study of Identity and
Culture (MUSIC), a national research collaborative that involved data collection from 30 colleges
and universities in the United States (Schwartz, Waterman, et al., 2011). The present sample
comprises all participants identifying as Hispanic or Latino and who were in the traditional
college age range (18-25 years). The present sample comprises 1,527 undergraduate students
(76% women, 24% men; mean participant age = 20.35 years, standard deviation [SD] = 3.88).
With regard to nativity, 77% of participants were born in the United States, and 23% reported
being born in another country. The most prominent countries of familial origin were Mexico
(19%), Cuba (17%), Colombia (6%), the Dominican Republic (3%), Peru (3%), Puerto Rico
(2%), Nicaragua (2%), Venezuela (2%) and El Salvador (2%).
An additional 8.5% of participants reported being of mixed Hispanic ancestry, where the
mother was born in one Hispanic country and the father was born in a different Hispanic
country. Two percent reported mixed Hispanic/non-Hispanic ancestry, where one parent was
born in a Hispanic country and the other parent was born in the United States or in a nonHispanic country. In terms of socioeconomic status, 28% of participants reported annual family
incomes below $30,000, 26% between $30,000-$50,000, 26% between $50,000-$100,000, and 17%
above $100,000. The remaining 3% of participants did not report their annual family income.
Measures
Acculturation orientations. The Stephenson (2000) Multigroup Acculturation Scale was
used to create the acculturation orientation categories. This instrument includes two subscales:
heritage cultural practices (17 items; e.g., “I think in my heritage language”) and American
cultural practices (15 items; e.g., “I think in English”). All items were responded to using a
5-point scale, ranging from 1 (strongly disagree) to 5 (strongly agree). In the current data set,
Cronbach’s estimates were .91 for heritage cultural practices and .85 for American cultural
practices.
Alcohol expectancies. Alcohol expectancies were measured using the Brief Comprehensive Effects of Alcohol Scale (BCEOA; Ham et al., 2005). The BCEOA is a 15-item version that
measures positive and negative expectancies and their valuations. The positive expectancies subscale includes items that assess positive effects of alcohol such as “I would act sociable,” whereas
the negative expectancies subscale measures anticipated negative effects associated with alcohol
use, e.g., “I would be clumsy.” The BCEOA also asks participants to indicate how desirable or
how undesirable they consider each positive or negative outcome expectancy to be. The positive
and negative drinking expectancy items are rated on a 4-point scale, ranging from 1 (disagree) to
4 (agree), and the valuation items are rated on a 5-point scale, ranging from 1 (bad) to 5 (good).
In the present sample, Cronbach’s α estimates were .81 for positive expectancies; .72 for negative
expectancies; .88 for positive valuations and .81 for negative valuations.
Alcohol-Related Risk Behaviors
Alcohol-related risk behaviors included hazardous alcohol use (e.g., drinking accompanied by
negative consequence such as missing school of work, relationship problems), binge drinking,
drunk driving, riding with a drunk driver, and sex under the influence of alcohol. Hazardous alcohol use was measured using the Alcohol Use Disorders Identification Test (AUDIT; Saunders,
Aasland, Babor, De La Puente, & Grant, 1993). The AUDIT assesses frequency and quantity
326
Journal of Clinical Psychology, April 2013
of drinking in the past year, as well as a number of negative consequences that individuals
experienced as a result of their drinking in the past year. The total AUDIT score (α = .87 in the
present study) was derived by taking the sum of the 10 items. The first AUDIT item measures
the frequency of alcohol use in the past year, with response options of (0) never, (1) monthly or
less, (2) 2-4 times a month, (3) 2 to 3 times a week, (5) 4+ times a week. The subsequent questions
measures problematic alcohol use and negative consequences associated with alcohol use (e.g.,
how often during the last year have you found that you were not able to stop drinking once you
had started?).
Responses to these remaining items were provided on a 5-point scale, where the scale anchors
varied as a function of item content. Binge drinking was measured by asking participants how
many times during the 30 days prior to assessment they had consumed four or more drinks (for
women) or five or more drinks (for men) on a single occasion (cf. Olthuis, Zamboanga, Ham, &
Van Tyne, 2011; Wechsler, Dowdall, Davenport, & Rimm, 1995). Drunk driving, riding with a
drunk driver, and sex under the influence of alcohol were measured by asking participants how
many times in the 30 days prior to assessment they had engaged in each of these behaviors. Each
of these items was answered using a 5-point scale: 0 (never), 1 (once or twice), 2 (3-5 times), 3
(6-10 times), or 4 (11 or more times).
Data Analytic Plan
Analyses for the current study were conducted in four steps. Preliminary analyses were conducted
on the first step to obtain the frequency of alcohol-related risk behaviors in this sample of
Hispanic college students and to evaluate the need to control for the nesting of students within
data collection sites. To provide the frequency of involvement in alcohol-related risk behaviors, we
created a dichotomous variable for each of these behaviors, where (0) represented no involvement
and (1) represented any involvement in the behavior in question. The hazardous alcohol use
variable was also dichotomized. Consistent with empirical evidence regarding the validity of
the AUDIT as a measure of hazardous alcohol use among college students, we used a standard
cut-point of 6 (Reinert & Allen, 2007). Participants with a composite score of 6 or higher
were coded as (1) representing involvement in hazardous drinking. Participants not meeting this
criterion were assigned a score of 0. Also in the first step, as an internal validation, we used latent
class analysis to ascertain whether the same acculturation dimensions reported by Schwartz and
Zamboanga (2008), using a different dataset, would emerge in the present multisite sample of
Hispanic college students.
In the second step, we estimated Model 1, in which acculturation dimensions were used
to directly predict alcohol-related risk behaviors. In the third step, we estimated Model 2 to
test the primary hypothesis that alcohol expectancies and their valuations would mediate the
relationship between acculturation dimensions and alcohol-related risks behaviors. In the current
study, with the exception of hazardous alcohol use, which was modeled as a continuous variable,
each of the alcohol-related risk variables was characterized by at least 75% zero responses. In
such cases, zero-inflated Poisson (ZIP) modeling (Atkins & Gallop, 2007) should be used to
model the zeroes separately from the nonzero count data so that associations within the nonzero
data can emerge. In a ZIP model, the count variable is split into two parts: (a) a dichotomous
(endorsement) indicator representing whether the participant engaged in the behavior in question
during the 30 days prior to assessment, and (b) a count (frequency) indicator representing how
often participants reported involvement in the behavior during that same time span. The count
portion of the ZIP model can be used to distinguish experimentation from chronic engagement
in a given behavior (Schwartz, Waterman, et al., 2011).
In a ZIP model, the unstandardized regression coefficient for the dichotomous portion is
interpreted as an odds ratio (OR), and the count portion is interpreted as an incidence rate
ratio (IRR). Both ratios are derived by taking the exponential (inverse natural logarithm) of the
unstandardized regression coefficient. To evaluate mediation, we used the joint significance test
advanced by MacKinnon (2008; MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002). This
approach to mediation analysis creates a 95% confidence interval around the product of the two
unstandardized paths that contain the mediating pathway. If this product does not include zero,
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
327
then partial mediation is assumed at p < .05. The joint significance test has been shown to have
more statistical power and accurate Type I error rates compared with other tests of mediation.
The mediation analyses were conducted using the PRODCLIN software (Fritz & MacKinnon,
2007).
In the final step of analysis, we conducted invariance testing across gender and university
setting, to examine the consistency of the hypothesized model. Of the 1527 students who participated in the study, 1174 students (77%) self-identified as alcohol drinkers. Of these 1174
drinkers, 64% were younger than 21 years of age, and 36% were 21 years of age or older. Given
reported differences by age (i.e., legal versus underage drinkers) regarding alcohol use among
college students (e.g., Zamboanga et al., 2010), we created a dichotomous variable to represent
whether the student was below (younger than 21) or above (21 or older) the legal drinking age.
This dichotomized age variable was then entered into the model as a covariate by including it as
an additional predictor of each mediating and outcome variable.
Results
Preliminary Analyses
Intraclass correlations. Given that the data were collected from 30 different colleges and
universities, we evaluated the magnitude of the intraclass correlations (ICC) prior to model
testing. The ICC indicates the percentage of variance in a given outcome variable that is attributable to between-site differences (Raudenbush & Bryk, 2002). Generally, ICC values of .05
of greater are considered indicative that site-level differences should be controlled in analysis.
Three of the ICC estimates—hazardous drinking (.07), positive expectancy valuations (.05), and
drunk driving (.11)—met this criterion. Consequently, to adjust standard errors for the nesting
of participants within university sites, we used the sandwich estimator (Kauerman & Carroll,
2001). This adjustment was done using the TYPE = COMPLEX command in Mplus (Muthén
& Muthén, 2007). The use of the sandwich estimator is recommended when the objective is
to account for nesting without predicting between-site variability. Robust maximum likelihood
estimation was used to account for missingness and to control for non-normality.
Descriptive statistics. The frequencies of past month involvement in alcohol-related
risk behaviors, including hazardous drinking, binge drinking, sex under the influence of alcohol, driving under the influence of alcohol, and riding with a drunk driver, are presented in
Table 1. The prevalence of past year alcohol use, as assessed using the first item of the AUDIT,
was 77%. The frequencies of past-month involvement in alcohol-related risk behaviors were
lower than those for national samples of college students (Hingson et al., 2009). Hazardous
drinking was the most common risk behavior—26% of the sample was classified as past-year
hazardous drinkers based on their AUDIT scores.
Creation of acculturation orientation classes. To place participants into acculturation
orientation classes, we used latent class analysis. Following classification, we validated the acculturation orientation classes to document significant between-class differences in endorsement
Table 1
Prevalence of Alcohol Use and Alcohol Risk Behaviors (Last 30 Days)
N (%)
Alcohol use
Binge drinking
Sex while drinking
Drunk driving
Riding w/drunk driver
Hazardous drinking
1174 (77%)
254 (17%)
250 (16%)
246 (16%)
890 (18%)
402 (26%)
328
Journal of Clinical Psychology, April 2013
patterns of the clustering variables, as an internal validation of the class solution. The class
solution that we extracted was extremely similar to the solution reported by Schwartz and
Zamboanga (2008), with the exception of the undifferentiated class, which did not emerge in
the present sample. Specifically, LCA results suggested the presence of five empirically distinct
acculturation dimensions: separated bicultural (n = 219), separated (n = 76), assimilated (n =
401), high bicultural (n = 526), and low bicultural (n = 169). Compared with one-class to six-class
solution, the five-class solution provided the best model fit with smallest Bayesian information
criterion (22031.5) and Akaike information criterion (22116.5). The Vuong-Lo-Mendel-Rubin
likelihood ratio value for the five-class solution, compared with a four-class solution, was 53.02,
p < .01. The entropy value (E = .75) suggested that the five-class solution was reliable and stable.
Classification accuracy rates for all five classes were above .70.
Model 1: Direct Associations Between Acculturation Dimensions and Alcohol-Related
Risk Behaviors
In Model 1, we tested the direct relationship between acculturation classes and alcohol-related
risk outcomes (hazardous drinking, binge drinking, sex under the influence of alcohol, drunk
driving, and riding with a drunk driver). The acculturation classes were entered into the model
using dummy-coded variables for separation, separated biculturalism, high biculturalism, and
low biculturalism. The assimilated acculturation class was used as the reference category. Compared with the assimilated class, the three variants of biculturalism—separated bicultural, low
bicultural and high bicultural—were differentially associated with alcohol-related risk behaviors.
Separated biculturalism was negatively related to frequency of binge drinking (IRR = 0.55, 95%
confidence interval [CI] = 0.34 to 0.89, p = .02) and low biculturalism was negatively linked with
prevalence of binge drinking (OR = 0.36, 95% CI = 0.13 to 1.01, p = .05). A parallel association
for the separated bicultural class with prevalence of binge drinking approached significance
(OR = 0.50, 95% CI = 0.23 to 1.08, p = .08).
An inverse relationship was also observed between the low bicultural class and the frequency
of sex while under the influence of alcohol (IRR = 0.63, 95% CI = 0.44 to 0.92, p = .02).
Compared with the assimilated class, separated biculturalism was negatively associated with
the prevalence of drunk driving (OR = 0.40, 95% CI = 0.19 to 0.86, p = .02), but positively
associated with the frequency (IRR = 1.59, 95% CI = 1.12 to 2.25, p < .01) of drunk driving. Low
biculturalism was positively related to the frequency of riding with a drunk driver (IRR = 1.62,
95% CI = 1.21 to 2.17, p < .001), whereas the negative relationship between high biculturalism
and the prevalence of riding with a drunk driver approached significance (OR = 0.59, 95% CI =
0.32 to 1.08, p = .09). Both separated biculturalism β = −.11, p < .01, and high biculturalism,
β = −.09, p < .05, were negatively related to hazardous drinking.
Model 2: Indirect Effects of Alcohol Expectancies and Valuations in the Association
Between Acculturation Dimensions and Alcohol-Related Risk Outcomes
The third step of analysis was to test the proposed cultural and social cognitive model
(Figure 1) for the mediating role of alcohol expectancies and valuations in the association
between acculturation orientations and alcohol-related risk behaviors among Hispanic college
students. Results evaluating our hypothesized model indicated that acculturation orientations
were significantly associated with positive alcohol expectancies (Figure 2). Compared with individuals in the Assimilated class, individuals in the other five classes reported fewer positive
expectancies: Separated Bicultural, ß = −.40, p < .001; Separated, ß = −.23, p < .01; High
Bicultural, ß = −.20, p < .001; and Low Bicultural, ß = −.26, p < .01. On the other hand,
compared with participants in the assimilated class, individuals in the Separated Bicultural, ß =
−.29, p < .001, and High Bicultural, ß = −.13, p < .01, classes endorsed fewer positive expectancy valuations. The acculturation classes were not related to negative alcohol expectancies
or to negative expectancy valuations.
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
329
Figure 2. Cultural and social cognitive model of the association between acculturation orientations, positive alcohol expectancies and valuations and alcohol-related risk behaviors.
Alcohol Expectancies and Valuations as Predictors of Alcohol-Related Risk Behaviors
Odds ratios (ORs) for prevalence of alcohol-related risk behaviors are shown in Table 2 and
Incidence rate ratios (IRRs) for frequency of alcohol-related risk behaviors are reported in Table
3. Positive alcohol expectancies were positively associated with the prevalence of binge drinking,
sex under the influence of alcohol, and riding with a drunk driver (Table 2). Similarly, positive
alcohol expectancies were positively associated with hazardous alcohol use (β = .30, p < .001) and
with negative expectancy valuations (β = .16, p < .001). In terms of prevalence of alcohol-related
risk behaviors, a negative relationship was observed between negative alcohol expectancies and
prevalence of binge drinking, drunk driving, and riding with a drunk driver. Negative expectancy
valuations were positively associated with the prevalence of binge drinking and drunk driving
(Table 2). In terms of frequency of alcohol-related risk behaviors, negative alcohol expectancies
were positively associated with the frequency of drunk driving and approached significance as
positive correlates of the frequency of binge drinking and sex under the influence of alcohol
(Table 3).
In terms of age effects, older Hispanic college students were less likely than younger students
to report involvement in sexual activity under the influence of alcohol. Age group did not
significantly predict positive expectancies and valuations.
Alcohol Expectancies and Expectancy Valuations as Mediators of the Association
Between Acculturation Orientations and Alcohol-Related Risk Outcomes
We next tested whether alcohol expectancies and their valuations mediated the relationship
between acculturation orientations and alcohol-related risk outcomes. Our results indicate that
only positive alcohol expectancies partially mediated the relationship between acculturation
orientations and alcohol-related risk behaviors (Figure 2). Results of these mediation tests are
presented in Table 4. The current findings support our hypothesis with regard to positive expectancies as partial mediators of the relationship between acculturation and alcohol-related
risk behaviors. With the Assimilated class used as the reference group, positive expectancies partially mediated the associations between acculturation classes and prevalence of binge drinking,
drunk driving, and riding with a drunk driver (see Table 4), except for the association between
Separated Biculturalism and drunk driving, in which the mediating effect of positive alcohol
3.48
1.56
1.96
2.37
Positive
expectancies
To
To
To
To
OR (95% CI)
OR (95% CI)
(1.90
(0.75
1.01
(1.71
Positive
valuations
Negative
expectancies
OR = odds ratio; CI = confidence interval.
* p < .05; ** p < .01; *** p < .001.
Binge drinking
Sex while drinking
Drunk driving
Riding w/drunk driver
Alcohol-related
risk behaviors
3.27)
3.80)
3.27)***
6.38)***
OR (95% CI)
0.69
0.77
0.66
0.69
(0.49
(0.57
(0.44
(0.57
Association of Alcohol Expectancies and Valuations With Alcohol-Related Risk Behaviors
Table 2
to
to
to
to
1.05)
0.99)*
0.84)***
0.99)*
1.35
1.90
1.20
1.10
(0.87
(0.80
(0.73
(0.63
to
to
to
to
OR (95% CI)
Negative
valuations
2.10)
4.52)
1.80)
1.94)
1.75
0.83
1.78
0.99
(1.01
(0.52
(1.26
(0.57
to
to
to
to
3.00)*
1.30)
2.53)***
1.72)
330
Journal of Clinical Psychology, April 2013
1.09
1.21
1.15
0.97
Positive
expectancies
to
to
to
to
IRR (95% CI)
IRR (95% CI)
(0.81
(0.81
(0.79
(0.71
Positive
valuations
Negative
expectancies
IRR = Incident Rates Ratio IRR; CI = Confidence Intervals.
§ p < .1; * p < .05.
Binge drinking
Sex while drinking
Drunk driving
Riding w/drunk driver
Alcohol-related
risk behaviors
1.45)
1.81)
1.68)
1.31)
IRR (95% CI)
1.19
1.18
1.29
1.24
(1.00
(0.96
(1.03
(0.97
to
to
to
to
1.44)§
1.61)*
1.57)§
1.42)§
Association of Alcohol Expectancies and Valuations With the Frequency of Alcohol-Related Risks Behaviors
Table 3
1.01
0.88
0.91
1.09
(0.76
(0.58
(0.68
(0.77
to
to
to
to
IRR (95% CI)
Negative
valuations
1.34)
1.36)
1.21)
1.54)
1.75
1.14
1.00
1.12
(0.89
(0.92
(0.86
(0.81
to
to
to
to
1.18)
1.41)
1.18)
1.55)
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
331
0.86
0.97
0.82
1.01
(0.54 to 1.00)*
(0.80 to 1.10)
(0.56 to 0.87)***
(0.89 to 1.15)
(0.59 to 1.12)
(0.77 to 1.09)
0.76
0.94
0.71
1.01
0.83
0.93
0.90
0.96
0.75
0.98
(0.41 to 0.83)***
(0.85 to 1.09)
(0.70 to 1.07)
(0.84 to 1.05)
(0.65 to 0.99)*
(0.93 to 1.10)
(0.65 to 1.01)§
(0.86 to 1.06)
(0.53 to 0.99)*
(0.90 to 1.05)
Yes/No - OR (95% CI)
Count- IRR (95%CI)
0.61
0.97
Yes/No - OR (95% CI)
Count - IRR (95%CI)
OR = odds ratio; CI = confidence interval.
§ p < .1; * p < .05; ** p < .01; *** p < .001.
Binge drinking
Yes/No
Count
Drunk driving
Yes/No
Count
Riding w/drunk driver
Yes/No
Count
Sex while drinking
Yes/No
Count
Alcohol-related risk
behaviors (outcome variables)
Separated
0.91
0.96
0.84
1.01
0.87
0.97
0.78
0.98
(0.77 to 1.06)
(0.88 to 1.04)
(0.74 to.93)***
(0.94 to 1.07)
(0.73 to 1.00)*
(0.89 to 1.05)
(0.64 to 0.91)***
(0.92 to 1.04)
Low bicultural
0.89
0.95
0.80
1.01
0.84
0.96
0.73
0.98
(0.70 to 1.08)
(0.84 to 1.05)
(0.66 to 0.94)***
(0.93 to 1.10)
(0.65 to 1.00)*
(0.86 to 1.06)
(0.53 to 0.92)**
(0.90 to 1.06)
Yes/No - OR (95% CI)
Count - IRR (95%CI)
Separated bicultural
Yes/No - OR (95% CI)
Count - IRR (95%CI)
High bicultural
Acculturation orientations (predictors)
Mediated Effects of Positive Expectancies in the Association Between Acculturation Orientations and Alcohol-Related Risk Behavior
Table 4
332
Journal of Clinical Psychology, April 2013
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
333
expectancies approached significance. With the assimilated class again used as the reference
group, results indicated that positive expectancies partially mediated the association between
the acculturation classes and hazardous drinking: Separation, point estimate = −.615, 95%
CI = −1.01 to - .276, p <.001; Separated biculturalism, point estimate = −.357, 95% CI =
−1.01 to −.276, p < .05; High biculturalism, point estimate = −.309, 95% CI = −.508 to −.137,
p < .001; and Low biculturalism, point estimate = −.394, 95% CI = −.714 to −.110, p < .01.
The hypothesis that negative expectancies, positive valuations, and negative valuations would
mediate the relationship between acculturation orientations and alcohol-related risk outcomes
were not supported.
Equivalence Across Gender and College/University Setting
As the final step of analysis, we examined the consistency of the hypothesized model across male
(n = 266) and female (n = 855) students, as well as across urban/suburban (n = 694) and college
town (n = 423) settings. For each invariance test, we estimated two models: an unconstrained
model with all paths free to vary by gender or setting, and a constrained model with all paths
constrained equally across gender or setting. The difference between the two log-likelihood
values was adjusted according to the non-normality corrections used by the robust maximum
likelihood estimator (Muthén & Muthén, 2011), and this adjusted difference was interpreted as
a chi-square difference value. For gender, the model comparison was not statistically significant
(χ2 (65) = 58.95, p = .70), suggesting that the model fit equivalently for male and female
Hispanic college students. Similarly, the model comparison across settings was not statistically
significant (χ2 (65) = 69.55, p = .56), suggesting that the model fit equivalently between college
towns and urban/suburban locations.
Discussion
The primary purpose of the present study was to evaluate the mediating role of alcohol expectancies and expectancy valuations in the associations of acculturation orientations with
alcohol-related risk behaviors in a sample of Hispanic college students from diverse regions
of the United States. In addition, two auxiliary aims were pursued to achieve this goal: (a)
the replication of the cluster solution obtained by Schwartz and Zamboanga (2008) and
(b) the testing for invariance of the hypothesized model by gender and across university
settings. The present results suggest three main conclusions.
First, we were able to replicate and extend the class solution reported by Schwartz and Zamboanga (2008) in a sample of Hispanic college students from 30 colleges and universities around
the United States. Second, our findings also indicated that heterogeneity in approaches to acculturation was differentially associated with alcohol-related risk behaviors. Third, our primary
hypothesis was partially supported by the finding that positive alcohol expectancies played a
mediating role in the association of acculturation orientations with alcohol-related problems.
This finding highlights that positive alcohol expectancies as relevant targets for prevention and
intervention aimed at reducing alcohol use and related problems on college campuses. Fourth,
the posited model appeared to operate equivalently across gender and across university settings.
Replication and Validation of Acculturation Orientations in a Hispanic College Sample
In the present study, we replicated and extended the acculturation class solution reported by
Schwartz and Zamboanga (2008) in two ways. First, we found a highly similar cluster solution in a
different sample; and second, we found that these classes are differentially associated with alcohol
expectancies and expectancy valuations, and with alcohol-related risk behaviors. This pattern
of findings is consistent with prior evidence that acculturation must be conceptualized as a
multidimensional process, and that heterogeneity in acculturation is important for understanding
involvement in alcohol-related risk outcomes.
Second, as has been reported in many samples of adolescents and emerging adults within many
different receiving contexts (e.g., Chia & Costigan, 2006; Matsunaga, Hecht, Elek, & Ndiaye,
334
Journal of Clinical Psychology, April 2013
2010; Schwartz & Zamboanga, 2008), many of the Hispanic emerging adults in the present
sample were placed into a class associated with biculturalism. Exposure to the heritage culture
in the home, and to American culture through their college peers, media, and other external
influences increases the likelihood that young people from immigrant families will endorse both
their heritage and receiving cultural streams. However, as has been speculated and documented
in other studies (e.g., Benet-Martı́nez & Haritatos, 2005; Rudmin, 2003), there appear to be
multiple types of biculturalism. In the present sample, three variants of biculturalism emerged:
a separated version, where Hispanic orientations were elevated; a “low” version, where both
Hispanic and American cultural streams were endorsed moderately; and a “high” version,
where both Hispanic and American cultural streams were endorsed strongly.
Heterogeneity in Acculturation Orientations and Association With Alcohol-Related
Risk Behaviors
Our findings indicate that, relative to assimilation, separated biculturalism was negatively associated with all of the alcohol-related risk behaviors measured in this study. The only exception was
frequency of drunk driving, where separated bicultural participants engaged in greater degrees
of drunk driving compared with assimilated participants. This finding suggests that separated
bicultural Hispanic college students are comparatively unlikely to report driving under the influence of alcohol, but when they do, they are likely to drive drunk more often compared with
Hispanic college students who are assimilated. A similar pattern emerged in the association
between low biculturalism and riding with a drunk driver: endorsement of riding with drunk
drivers was lower in the low bicultural class, but for those low-bicultural participants who did
report riding with a drunk driver, they tended to do so more often compared with assimilated
participants. Among the other two acculturation classes, low biculturalism was negatively related
to endorsement of binge drinking and alcohol-related sexual activity, whereas high biculturalism
was negatively related only to hazardous drinking.
Taken together, these findings suggest that “biculturalism” is not a single entity that protects
against all alcohol-related risk behaviors. Rather, different types of biculturalism were linked
with different risky behaviors. The fact that each type of bicultural orientation was inversely
related to the endorsement of at least one alcohol-related risk behavior is consistent with research
suggesting that, for Hispanics, retention of one’s heritage culture is negatively related to healthrisk outcomes (e.g., Szaflarski, Cubbins, & Ying, 2010). However, the positive associations with
frequency of drunk driving (for separated biculturalism) and with frequency of riding with a
drunk driver (for low biculturalism) suggest that retention of Hispanic cultural practices may,
in some cases, no longer be protective against the severity of the behavior once it has been
initiated. However, the high bicultural class, which is characterized by high endorsement of both
Hispanic and American cultural practices, appeared to be protective against alcohol-related
negative consequences (i.e., hazardous drinking).
Mediation of Positive Alcohol Expectancies in the Association between Acculturation
Orientations and Alcohol-Related Risk Outcomes
The present results lend support to the proposition that alcohol expectancies may mediate the
relationship between acculturation orientations and alcohol-related risk behavior outcomes.
Specifically, positive, but not negative, alcohol expectancies mediated the relationship between
acculturation orientations and alcohol-related problems. Positive expectancies partially mediated the relationships of both the low and high variants of biculturalism with hazardous drinking,
binge drinking, drunk driving, and riding with a drunk driver, and they partially mediated the
associations of the separated variant of biculturalism with hazardous drinking, binge drinking,
and riding with a drunk driver. Positive expectancies also partially mediated the inverse association between separation and hazardous drinking, binge drinking, drunk driving, and riding with
a drunk driver. It appears that heritage-culture retention, which is absent in the assimilated class,
may help to protect against positive alcohol-related beliefs. Curiously, however, the acculturation
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
335
classes were not associated with negative alcohol expectancies, suggesting that “aversive” effects
of alcohol may be perceived similarly regardless of one’s degree of heritage culture retention.
As has been documented in previous research (Jones et al., 2001; Labrie et al., 2011), in
the current study we found that alcohol expectancies served as strong correlates of several
alcohol-related risk behaviors among Hispanic college students. Moreover, we found that these
associations generalized across male and female college students and across types of college
settings, suggesting that the link between expectancies and behavior is not moderated by gender
or whether the university is located in a college town or a suburban/urban area. Further, negative
alcohol expectancies also appeared to serve as risk factors for a number of alcohol-related risk
behaviors, suggesting that beliefs that alcohol is harmful or induces undesirable mood states
is nonetheless associated with drunk driving, riding with impaired drivers, and sex under the
influence of alcohol. The fact that specific alcohol-related beliefs are labeled by researchers as
“negative” does not necessarily indicate that students will perceive these outcomes as sufficiently
aversive to prevent them from drinking alcohol and engaging in risky behaviors linked with
alcohol consumption. Surprisingly, however, expectancy valuations were not significantly related
to the risky behaviors we studied. Labeling a specific alcohol-related outcome as “good” or “bad”
may not influence the likelihood that, or extent to which, an individual will drink heavily, drive
while drunk, ride with an impaired driver, or engage in sexual activity under the influence of
alcohol. For Hispanic college students, alcohol expectancies therefore appear to be closely related
to drinking behaviors, whereas alcohol expectancy valuations do not (Zamboanga, 2005).
Clinical Implications of the Current Findings
As found in previous studies, the current findings suggest that alcohol expectancies represent
a common pathway (e.g., Darkes, Greenbaum, & Goldman, 2004) through which more distal
variables, including acculturation orientations, may influence alcohol use and related risk behaviors. In light of the prevalence of alcohol problems and associated risks among Hispanic
college students, and among emerging adults in general (Corbin, Iwamoto, & Fromme, 2011),
intervening to modify positive alcohol expectancies among Hispanic college students (particularly those with an assimilated acculturation orientation who may be more responsive to positive
alcohol expectancy challenges) is essential as a way of preventing or decreasing participation in
alcohol-related risk behaviors.
Such interventions may be promising given that expectancies can be challenged and modified
(e.g., Jones et al., 2001; Weirs & Rummeling, 2004). The current results affirm the role of
alcohol expectancies as key processes that can be targeted effectively as a strategy for reducing
alcohol-related problems among Hispanics on college campuses. Encouraging heritage culture
retention (Kulis, Nieri, Yabiku, Stromwall, & Marsiglia, 2007) may also represent an important
(but perhaps more distal) way to prevent alcohol-related risk behaviors among Hispanic college
students. The increased freedom and exploration that characterize emerging adulthood may be
utilized more prudently among those who retain Hispanic culture, which emphasizes behaviors
and practices that facilitate family relations and interdependence.
Moreover, college counseling programs may systematically include, as part of their alcoholrelated risk prevention efforts, sessions that discuss heritage cultural practices with Hispanic
students. In particular, given the power of language as a transmitter of cultural knowledge,
Hispanic students who have stopped speaking Spanish (or who may have never learned it) may
be encouraged to reacquaint themselves with the language and to use it to reconnect with their
cultural heritage, as well as with family members who may be more comfortable in Spanish than
in English. Such heritage cultural retention may decrease positive alcohol expectancies, which in
turn may lower the likelihood of hazardous drinking, drunk driving, riding with a drunk driver,
and sex under the influence of alcohol.
Second, the finding that the pattern of results generalized across gender and college setting
suggests that similar cultural and social cognitive mechanisms are likely associated with alcoholrelated risk taking in male and female Hispanic college students, and regardless of whether they
go away to college or stay home and commute to school. The present findings are therefore
336
Journal of Clinical Psychology, April 2013
relevant for college educational and intervention programs that address alcohol use, in Hispanic
students, across university settings and for students of both genders.
Third, the application of a model that integrates person-centered Latent Class Analysis
Latent Class Analysis (LCA) and variable-centered Structural Equation Modeling (SEM) analytic techniques to data analysis (e.g., Muthén & Muthén, 2007) allowed us to take into
account unobserved heterogeneity in acculturation patterns and allowed us to model this heterogeneity as a predictor of alcohol-related risks through alcohol expectancies and expectancy
valuations. Moreover, risk behaviors were modeled in their natural metrics (e.g., normal,
zero-inflated Poisson) rather than attempting to transform or rescale them to meet the assumptions of more commonly used analyses. This analytic approach allowed us to most validly
examine the structural relations between cultural and social cognitive risk/protective factors
that may influence alcohol use and related risk outcomes.
Last, the findings reported here both build upon and expand the current literature on acculturation, alcohol expectancies, and health-risk behaviors among Hispanics college students.
First, as previously noted, especially in relation to health outcomes, acculturation has often been
studied as a unidimensional process whereby receiving culture acquisition and heritage culture
retention are cast as opposing ends of a continuum (Abraido-Lanza, Armbrister, Florez, &
Aguirre, 2006; Hunt, Schneider, & Comer, 2004). The present results suggest that bidimensional
models of acculturation may hold greater explanatory power because they examine orientations
toward one’s heritage separately from orientations toward the receiving culture (e.g., Szapocznik et al., 1980). The identification of distinct acculturation orientations in the present study
allows for exploration of the interplay between Hispanic and American cultural practices—
most of which would be indistinguishable from one another within unidimensional models of
acculturation—and may be differentially linked with alcohol-related risks.
Study Limitations
The present results should be interpreted in light of these important limitations. Perhaps the
most important limitation is the use of a cross-sectional design. Although this design permitted
us to map the associations among the study constructs, it does not permit inferences regarding
causal or directional effects. Longitudinal measurement of heterogeneity in acculturation would
yield information about individual differences, as well as the mechanisms underlying the relations among acculturation orientations, alcohol expectancies, and alcohol-related risk behaviors
(Matsunaga et al., 2010). In addition, alcohol use and alcohol-related risk behaviors should be
examined over the college years as a function of acculturation.
Second, the acculturation orientations were empirically derived using cultural practices. Because acculturation is a multidimensional process that includes identifications and values as
well as practices (Schwartz et al., 2010), it is possible that a different set of orientations may
have emerged if all three components were considered. Third, the present findings were based
on participants’ self-reports. Therefore, the results may be confounded by the effects of social desirability—individuals may overreport or underreport risk behavior participation for any
number of reasons. The use of collateral reports, daily diary studies, or other methods that do
not rely on recall of past risk taking behavior may be important for future research.
Additionally, all variables were measured using reports from a single informant (i.e., the
student her/himself), and this shared method variability may artificially increase the correlations
among constructs (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Future studies should also
examine involvement in illicit drug use, given findings suggesting that the construct expectancy
(i.e., an individual belief about the anticipated outcomes associated with future experiences) and
expectancy valuations apply to substances other than alcohol (e.g., Lundahl & Lukas, 2007)
have been shown to play a role in treatment outcomes (Witkiewitz & Marlatt, 2004).
Conclusions
Despite these limitations, the present findings have demonstrated the utility of a cultural and
social cognitive perspective for examining links among acculturation orientations, alcohol ex-
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
337
pectancies and expectancy valuations, and alcohol-related risk behaviors in a national sample
of Hispanic college students. Our results have replicated and validated acculturation orientation
categories and have documented both direct and indirect relationships of acculturation orientations to alcohol-related risk behaviors. Specifically, the indirect associations of acculturation
with alcohol-related risk behaviors through positive alcohol expectancies have brought together
two disparate literatures in the service of creating a larger understanding of the mechanisms
through which Hispanic college students may come to engage in risky behaviors associated with
alcohol use. Our model not only sheds light on the cultural underpinnings of Hispanic health
disparities in alcohol-related consequences, but it also provides avenues for intervention. We
hope that the model will find use in college counseling centers and other clinical settings that
provide services to Hispanic students.
References
Arnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties.
American Psychologist, 55, 469–480.
Abraido-Lanza, A. F., Armbrister, A. N., Florez, K. R., & Aguirre, A. N. (2006). Toward a theorydriven model of acculturation in public health research. American Journal of Public Health, 96,
1342–1346.
Alegria, M., Shrout, P., Woo, M., Guarnaccia, P., Sribney, W., Vila, D., ... ... ... Canino, G. (2007). Understanding differences in past year mental health disorders for Latinos living in the U.S. Social Science and
Medicine, 65, 214–230.
Allen, M. L., Elliott, M. N., Fugligni, A. J., Morales, L. S., Hambarsoomian, K., & Schuster, M. A. (2008).
The relationship between Spanish language use and substance use behaviors among Latino youth: A
social network approach. Journal of Adolescent Health, 43, 372–379.
Atkins, D. C., & Gallop, R. J. (2007). Rethinking how family researchers model infrequent outcomes: A
tutorial on count regression and zero-inflated models. Journal of Family Psychology, 21, 726–735.
Balodis, I. M., Potenza, M. N., & Olmstead, M. C. (2009). Binge drinking in undergraduates: relationship
with Sex, drinking behaviors, impulsivity, and the perceived effects of alcohol. Behavioural Pharmacology, 20, 518–526.
Benet-Martı́nez, V., & Haritatos, J. (2005). Bicultural identity integration (BII): Components and psychosocial antecedents. Journal of Personality, 73, 1015–1050.
Berry, J. W. (1997). Immigration, acculturation and adaptation. Applied Psychology: An International
Review, 46, 5–68.
Bot, S. M., Engels, R. C. M. E., Knibbe, R. A. (2005). The effects of alcohol expectancies on drinking
behavior in peer groups: Observations in a naturalistic setting. Addiction, 100, 1270–1279.
Bourdeau, B., Saltz, R., Bersamin, M., & Grube, J. (2007). Understanding the relationship between alcohol
and sex: Latino and White college students and problematic sexual experiences while drinking. Journal
of American College Health, 56, 299–306.
Caetano, R., Ramisetty-Mikler, S., & Rodriguez, L. A. (2008). The Hispanic American Baseline Alcohol
Survey (HABLAS): DUI rates, birthplace, and acculturation across Hispanic national groups. Journal
of Studies on Alcohol and Drugs, 69, 259−265.
Caetano, R., Ramisetty-Mikler, S., Wallisch, L. S., McGrath, C., & Spence, R. T. (2008). Acculturation,
drinking, and alcohol abuse and dependence among Hispanics in the Texas–Mexico Border. Alcoholism:
Clinical and Experimental Research, 32, 314–321.
Chartier, K. G., & Caetano, R. (2010). Ethnicity and health disparities in alcohol research. Alcohol Research
and Health, 33, 152–160.
Chartier, K. G., Hesselbrock, M. N., & Hesselbrock, V. M. (2009). Ethnicity and adolescent pathways to
alcohol use. Journal of Studies on Alcohol and Drugs, 70, 337–345.
Chia, A-L., & Costigan, C. L. (2006). A person-centered approach to identifying acculturation groups
among Chinese Canadians. International Journal of Psychology, 41, 397–412.
Coatsworth, J. D., Maldonado-Molina, M., Pantin, H., & Szapocznik, J. (2005). A person-centered and
ecological investigation of acculturation strategies in Hispanic immigrant youth. Journal of Community
Psychology, 33, 157–174.
Corbin, W. R., Iwamoto, D. K., & Fromme, K. (2011). Broad social motives, alcohol use, and related
problems: Mechanisms of risk from high school through college. Addictive Behaviors, 36, 222–230.
338
Journal of Clinical Psychology, April 2013
Darkes, J., Greenbaum, P. E., & Goldman, M. S. (2004). Alcohol expectancy mediation of biopsychosocial
risk: Complex patterns of mediation. Experimental and Clinical Psychopharmacology, 12, 27–38.
Dawson, D. A., Grant, B. F., Stinson, F. S., Stinson, F. S., Chou, P. S., Huang, B., & Ruan, W. J. (2005).
Recovery from DSM-IV alcohol dependence: United States, 2001–2002. Addiction, 3, 281–292.
Epstein, J. A., Margaret, D., & Botvin, G. J. (2003). A mediational model of the relationship between linguistic
acculturation and polydrug use among Hispanic adolescents. Psychological Reports, 93, 859–866.
Fritz, M., & MacKinnon, D. P. (2007). Power to detect the mediated effect. Psychological Science, 18,
233–239.
Fromme, K., Stroot, E. A., & Kaplan, D. (1993). Comprehensive effects of alcohol: Development and
psychometric assessment of a new expectancy questionnaire. Psychological Assessment, 5, 19–26.
Goldman, M. S., Greenbaum, P. E., Darkes, J., Brandon, K. O., & Del Boca, K. (2011). How many versus
how much: 52 weeks of alcohol consumption in emerging adults. Psychology of Addictive Behaviors,
25(1), 16–27.
Goldman, M. S., Reich, R. R., & Darkes, J. (2006). Expectancy as a unifying construct in alcohol-related
cognition. In R. W. Wiers & A. W. Stacy (Eds.), Handbook of Implicit Cognition and Addiction (pp.
105–119). Thousand Oaks, CA: Sage.
Ham, L. S., & Hope, D. A. (2006). Incorporating social anxiety into a model of college problem drinking:
Replication and extension. Psychology of Addictive Behaviors, 3, 348–355.
Ham, L. S., Stewart, S. H., Norton, P. J., & Hope, D. A. (2005). Psychometric assessment of the comprehensive effects of alcohol questionnaire: Comparing a brief version to the original full scale. Journal of
Psychopathology and Behavioral Assessment, 27, 141–158.
Hatzenbuehler, M. L., Corbin, W. R., & Fromme, K. (2011). Discrimination and alcohol-related problems
amongcollege students: A prospective examination of mediating effects. Drug and Alcohol Dependence,
115, 213–220.
Hingson, R., Heeren, T., Winter, M., & Wechsler, H. (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, 26, 259–279.
Hunt, L., Schneider, S., & Comer, B. (2004). Should “acculturation” be a variable in health research? A
critical review of research on U.S. Hispanics. Social Science and Medicine, 59, 973–986.
Johnston, L. D., O’Malley, P. M., Bachman, J. G., & Schulenberg, J. E. (2009). Monitoring the Future
national survey results on drug use, 1975–2008: Volume II, College students and adults ages 19-50 (NIH
Publication No. 09-7403). Bethesda, MD: National Institute on Drug Abuse.
Jones, B. T., Corbin, W., & Formme, K. (2001). Conceptualizing addiction: A review of expectancy theory
and alcohol consumption. Addiction, 96, 57–72.
Kauerman, G., & Carroll, R. J. (2001). A note on the efficiency of sandwich covariance matrix estimation.
Journal of the American Statistical Association, 96, 1387–1396.
Kulis, S., Nieri, T., Yabiku, S., Stromwall, L. K., & Marsiglia, F. F. (2007). Promoting reduced and discontinued substance use among adolescent substance users: Effectiveness of a universal prevention program.
Prevention Science, 8, 35–49.
Kuntsche, E. N., Knibbe, R., Gmel, G., & Engels, R. (2005). Why do young people drink? A review of
drinking motives. Clinical Psychology Review, 25, 841–861.
LaBrie, J. W., Kenney, S. R., Tehniat, M., & Lac, A. (2011). Identifying factors that increase the
likelihood of driving after drinking among college students. Accident Analysis and Prevention, 43,
1371–1377.
Lopez-Class, M., Castro, F. G., & Ramirez, A. G. (2011). Conceptions of acculturation: A review and
statement of critical issues. Social Science and Medicine, 72, 1555–1562.
Lundahl, L. H., & Lukas, S. E. (2007). Negative cocaine effect expectancies are associated with subjective
response to cocaine challenge in recreational cocaine users. Addictive Behaviors, 32, 1262–1271.
MacKinnon, D. P. (2008). An introduction to statistical mediation analysis. New York, NY: Lawrence
Erlbaum.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of
methods to test mediation and other intervening variable effects. Psychological Methods, 7, 83–104.
Maldonado-Molina, M., Reingle, J. M., Jennings, W. G., & Prado, G. (2011). Drinking and driving among
immigrant and US-born Hispanic young adults: Results from a longitudinal and nationally representative
study. Addictive Behaviors, 36, 381–388.
Alcohol Expectancies, Acculturation, and Alcohol-Related Risk Behaviors
339
Marı́n, G. (1996). Expectancies for drinking and excessive drinking among Mexican Americans and nonHispanic Whites. Addictive Behaviors, 21, 491–507.
Matsunaga, M., Hecht, M. L., Elek, E., & Ndiaye, K. (2010). Ethnic identity development and acculturation:
A longitudinal analysis of Mexican-Heritage youth in the Southwest United States. Journal of CrossCultural Psychology, 41, 410–427.
Mills, B. A., & Caetano, R. (2010). The Hispanic Americans Baseline Alcohol Survey (HABLAS): Predictors
of alcohol attitudes and expectancies in Hispanic national groups. Alcoholism: Clinical and Experimental
Research, 34, 790–799. doi:10.1111/j.1530-0277.2010.01151.x
Mulia, N., Ye, Y., Greenfield, T. K., & Zemore, S. E. (2009). Disparities in alcohol-related problems among
White, Black, and Hispanic Americans. Alcoholism: Clinical and Experimental Research, 33, 654–662.
Muthén, L. K., & Muthén, B. O. (1998–2007). Mplus user’s guide (version 5). Los Angeles, CA: Author.
Muthén, L. K., & Muthén, B. O. (2011). In Mplus user’s guide. Retrieved from http://www.statmodel.
com/chidiff.shtml
Neighbors, C., Lee, C. L., Lewis, M. A., Forros, B., & Larimer, M. E. (2007). Are social norms the best
predictor of outcomes among heavy- drinking college students? Journal of Studies on Alcohol and
Drugs, 68, 556–565.
Olthuis, J. V., Zamboanga, B. L., Ham, L. S., & Van Tyne, K. (2011). The utility of a gender-specific
definition of binge drinking on the AUDIT. Journal of American College Health, 59(4), 239–245.
Pearson, W. S., Dube, S. R., Nelson, D. E., & Caetano, R. (2009). Differences in patterns of alcohol
consumption among Hispanics in the United States, by survey language preference, Behavioral Risk
Factor Surveillance System, 2005. Preventing Chronic Disease, 6(2). Available at: http://www.cdc.
gov/pcd/issues/2009/apr/08_0039.htm
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P.(2003). Common method biases in
behavioral research: A critical review of the literature and recommended remedies. Journal of Applied
Psychology, 88, 879–903.
Ramirez, J. R., Crano, W. D., Quist, R., Burgoon, M., Alvaro, E. M., & Grandpre, J. (2004). Acculturation,
familism, parental monitoring, and knowledge as predictors of marijuana and inhalant use in adolescents.
Psychology of Addictive Behaviors, 18, 3–11.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models. Newbury Park, CA: Sage.
Reinert, D. F., & Allen, J. P. (2007). The alcohol use disorders identification test: An update of research
findings. Alcoholism, Clinical and Experimental Research, 31, 185–199.
Resor, M. R., & Cooper, T. V. (2010). Club drug use in Hispanic college students. The American Journal
on Addictions, 19, 319–324.
Rudmin, F. W. (2003). Critical history of the acculturation psychology of assimilation, separation, marginalization, and integration. Review of General Psychology, 7, 3–37.
Saunders, J. B., Babor, T. F., De La Puente, J. R., & Grant, M. (1993). Development of the alcohol use
disorders identification test (AUDIT): WHO collaborative project on early detection of persons with
harmful alcohol consumption–II. Addiction, 88(6), 791–804.
Schwartz, S. J., Pantin, H., Sullivan, S., Prado, G., & Szapocznik, J. (2006). Nativity and years in the
receiving culture as markers of acculturation in ethnic enclaves. Journal of Cross-Cultural Psychology,
37, 345–353.
Schwartz, S. J., Unger, Zamboanga, & Szapocznik, J. (2010). Rethinking the concept of acculturation:
Implications for theory and research. American Psychologist, 4, 237–251.
Schwartz, S. J., Waterman, A. S., Vazsonyi, A. T., Zamboanga, B. J., Whitbourne, S. K., Weisskirch, R.
S., . . . Ham, L. S. (2011). The association of well-Being with health risk behaviors in college-attending
young adults. Applied Developmental Science, 15, 20–36.
Schwartz, S. J., Weisskirch, R. S., Zamboanga, B. L., Castillo, L. G., Ham, L. S., Huynh, Q-L, . . . Cano,
M. A. (2011). Dimensions of acculturation: Associations with health risk behaviors among college
students from immigrant families. Journal of Counseling Psychology, 58, 27–41.
Schwartz, S. J., & Zamboanga, B. L. (2008). Testing Berry’s model of acculturation: A confirmatory latent
class approach. Cultural Diversity and Ethnic Minority Psychology, 14, 275–285.
Schwartz, S. J., Zamboanga, B. L., & Jarvis, L. H. (2007). Ethnic identity and acculturation in Hispanic early
adolescents: Mediated relationships to academic grades, prosocial behavior, and externalizing symptoms.
Cultural Diversity and Ethnic Minority Psychology, 13, 364–373.
Stephenson, M. (2000). Development and validation of the Stephenson Multigroup Acculturation Scale
(SMAS). Psychological Assessment, 12, 77–88.
340
Journal of Clinical Psychology, April 2013
Szaflarski, M., Cubbins, L. A., & Ying, J. (2010). Epidemiology of alcohol abuse among US immigrant
populations. Journal of Immigrant Minority Health, 13(4), 647–658.
Szapocznik, J., Kurtines, W., & Fernandez, T. (1980). Biculturalism and adjustment among Hispanic youths.
International Journal of Intercultural Relations, 4, 353–375.
Thush, C., & Wiers, R. W. (2007). Explicit and implicit alcohol-related cognitions and the prediction of
future drinking in adolescents. Addictive Behaviors, 32, 1367–1383.
Wagner, K. D., Ritt-Olson, A., Soto, D. W., Rodriguez, Y. L., Baezconde-Garbanati, L., & Unger, J. B.
(2008). The role of acculturation, parenting, and family in Hispanic/Latino adolescent substance use:
Findings from a qualitative analysis. Journal of Ethnicity in Substance Abuse, 7, 304–327.
Wechsler, H., Dowdall, G. W., Davenport, A., & Rimm, E. B. (1995). A gender-specific measure
of binge drinking among college students. American Journal of Public Health, 85(7), 982–985.
doi:10.2105/AJPH.85.7.982
Witkiewitz, K., & Marlatt, G. (2004). Relapse prevention for alcohol and drug problems: That was Zen, this
is Tao. American Psychologist, 59, 224–235.
Zamboanga, B. L. (2005). Alcohol expectancies and drinking related behaviors in Mexican American college
students. Addictive Behaviors, 30, 673–684.
Zamboanga, B. L., Bean, J. L., Pietras, A. C., & Pabon, L. C. (2005). Subjective evaluations of alcohol
expectancies and their relevance to drinking games involvement in female college students. Journal of
Adolescent Health, 37, 77–80.
Zamboanga, B. L., & Ham, L. S. (2008). Alcohol expectancies and context-specific drinking behaviors
among female college athletes. Behavior Therapy, 39, 162–170.
Zamboanga, B. L., Olthuis, J. V., Horton, N. J., McCollum, E. C., Lee, J., & Shaw, R. (2009). Where’s the
house party? A preliminary investigation of hazardous drinking behaviors and related risk factors across
campus housing. Journal of Psychology: Interdisciplinary and Applied, 143, 228–244.
Zamboanga, B. L., Raffaelli, M., & Horton, N. J. (2006). Acculturation status and heavy alcohol use
among Mexican American college students: An investigation of the moderating role of gender. Addictive
Behaviors, 31, 2188–2198.
Zamboanga, B. L., Schwartz, S. J., Ham, L. S., Borsari, B., Tyne, K. V. (2010). Alcohol expectancies,
pregaming, drinking games, and hazardous alcohol use in a multiethnic sample of college students.
Cognitive Therapy and Research, 34, 124–133. doi:10.1007/s10608-009-9234-1
Zemore, S. E. (2007). Acculturation and alcohol among Latino adults in the United States: A comprehensive
review. Alcoholism: Clinical and Experimental Research, 31(12), 1968–1990.