Addressing Bias in Email Surveys on College Students’ Alcohol Use: A Comparison of Strategies and Implications for Health Promotion

ABSTRACT

Background: Email surveys of health behaviors are convenient and low cost, but concerns remain regarding data accuracy and implications for health promotion. Purpose: Compare the accuracy of email-collected survey data with classroom-collected data on college students’ alcohol use, and compare strategies for addressing inaccuracies, including demographic weighting and continuum of resistance model. Methods: Data were gathered via email survey of 2,991 community college and research university students in October 2016 and via an in-person survey of 737 students at these same institutions in February 2017. Results: Classroom respondents were significantly more likely to report alcohol use and high-risk alcohol use, and reported more average weekly drinks than email respondents. Demographic weighting and the continuum of resistance model improved estimates but did not fix all inaccuracies. Discussion: Use caution in interpreting results from email surveys on alcohol use if those results do not include information on estimated nonresponse bias. Translation to Health Education practice: Results from email alcohol surveys should be weighted for demographic differences – particularly age and sex – and should use the continuum of resistance modification if indicated. Recommend collecting additional data through anon-email method to improve accuracy of estimates and inform decision-making about interventions and programs.

A AJHE Self-Study quiz is online for this article via the SHAPE America Online Institute (SAOI) http://portal.shapeamerica.org/trn-Webinars

Disclosure statement

No potential conflict of interest was reported by the authors.