Date of Award


Degree Name



Public Health


Joe Tomaka


Researchers use survey methods in a variety of settings to assess health behaviors, chronic disease, and other health outcomes. Since most surveys rely on self-report from respondents, prevalence rates have been regarded with criticism on the assumption that some respondents are unwilling to divulge personal information about behaviors due to people's desires to maintain a good impression. Accordingly, the purpose of this study was to assess the extent to which self-reported drinking behavior correlates with measures of two known biases that may affect the validity of such report: Impression management and self-deception. The present study consisted of a secondary analysis of existing data from an alcohol risk reduction intervention among college students (N=511) and an intervention project involving a municipal fire department (N=740). Students and firefighters completed the Balanced Inventory of Desirable Responding, the Alcohol Use Disorders Identification Test, the Rutgers Alcohol Problem Index, and the Daily Drinking Questionnaire. Results suggested that among college students, impression management contributed uniquely and significantly to alcohol risk level, alcohol related-problems, and alcohol consumption. Among firefighters, self-deception contributed uniquely and significantly to alcohol risk and alcohol related problems. The relationships between social desirability biases and alcohol use among college students did not differ by gender. Overall, the results suggest that researchers may need to take into account social desirability bias in research and when implementing alcohol intervention programs in different populations. Increased assessment of social desirability bias may also lead to more accurate interpretation of research results when relying on self-reported behaviors and outcomes.




Received from ProQuest

File Size

50 pages

File Format


Rights Holder

Kristen Eileen Hernandez