Ethnic/Racial Differences in Alcohol Use: Does Drinking Refusal Self-Efficacy Matter?

Abstract

Introduction: College students—including those of Hispanic backgrounds—are at risk for hazardous drinking. Research has shown robust group differences between Hispanic and White individuals in alcohol use outcomes. The ability to resist alcohol consumption can be leveraged to reduce hazardous drinking; however, little research has examined Hispanic-White differences and whether drinking refusal self-efficacy accounts for group differences in hazardous drinking. Considering Hispanic individuals make up the largest ethnic/racial minority group in the United States, it is important to identify malleable psychological factors that prevent and reduce drinking problems. Method: Hispanic and White college students at two predominantly White institutions (N = 389; 58.6% women, M_age= 20.22) completed measures assessing drinking refusal self-efficacy, hazardous drinking, and negative drinking consequences. Results: Hispanic students reported lower levels of hazardous drinking, alcohol-related problems, and drinking refusal self-efficacy than White students. Drinking refusal self-efficacy was found to partially explain Hispanic-White differences in the levels of hazardous drinking and drinking-related problems. Specifically, drinking refusal self-efficacy was associated with alcohol use outcomes only among White students and not Hispanic students. Conclusion: The correspondence between drinking refusal self-efficacy and actual behaviors to turn down drinks, ethnic/racial distinctiveness in ratings of self-efficacy and cultural orientations, and situational contexts that surround drinking should be examined in future research.

Keywords:

• Disparity
• drug refusal
• harm reduction
• Latino/a
• protective

Acknowledgment

The authors thank Tian Zhu and Savannah Pham for assistance in data management.

Data availability

Study design and data analytic plan was preregistered and archived in the Open Science Framework (OSF) repository at https://osf.io/rk4zn. Data and analytic codes are available at https://osf.io/9fna2/.

Disclosure statement

The authors report no conflicts of interest.

CRediT statement

Lui led the study conceptualization, design of methods, data collection, conducted and supervised data analyses, led manuscript writing and revisions, and provided resources for the project. Krantz contributed to data analyses, writing of the initial draft, and revisions. Madson contributed to data collection, manuscript writing, and revision.

Positionality statement

Lui is an Asian American clinical psychologist with expertise in ethnic minority alcohol use and health disparities. Krantz is a White post-baccalaureate researcher. Madson is a White licensed counseling psychologist with research expertise on alcohol misuse and protective behavioral strategies.

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Notes

1 We focus on United States individuals with Spanish-speaking and Latin American heritages, including Mexican, Cuban, and Central American. Relatively more individuals of Mexican American background and residing in Texas prefer the broad ethnic/racial label of “Hispanic” to “Latino/a/x,” thus we use “Hispanic” as a term to characterize our sample and this population throughout this manuscript.

2 Preregistration of study design and analytic strategies was archived in the Open Science Framework repository.

3 There were no statistically significant differences in age and years in college by universities. There were cross-site differences in the distributions of gender and proportion of students who were affiliated with the Greek system; hence gender and Greek affiliation were considered as covariates in multivariate analyses.

4 We performed data screening in SPSS to identify potential univariate and multivariate outliers using Z-scores and Mahalanobis distance, respectively. We identified 15 cases as univariate outliers and no multivariate outliers. We found no differences in our univariate and multivariate results with and without these outliers; thus, we used all available data in subsequent analyses.

5 Gender also included transgender and gender nonconforming/fluid; only two research participants indicated that they identified as transgender or gender fluid. Given the small sample size, gender as a covariate included individuals who identified as women or men.