The Revision of the Student Alcohol Questionnaire: A Validation Study
Faculty Sponsor’s Department
Psychology
Type
Oral Competitive
Project's Category
Psychology
Abstract or Artist's Statement
College campuses are a common location for individuals to experience alcohol consequences. Those consequences: health, legal, and/or academic, could cause short- or long-term repercussions on the student. Students likely have been provided education about these consequences from a range of sources (e.g., parents, friends, health courses). It is important to understand what knowledge of alcohol-related information students retain and how that information may impact their decisions around risk behavior engagement.
The Student Alcohol Questionnaire (SAQ; Engs &Hanson, 1973), contains 4 subscales: drinking patterns, problems related to alcohol, knowledge of alcohol, and alcohol attitudes. This measure provided an important model for assessment of students' understanding of alcohol use consequences, however the questionnaire the language had not been updated since it was created. The current study had two aims; 1) to revise the SAQ with language that would be more accessible to current students and reflective of modern drinking trends, and 2) to examine the factor structure of the drinking problems subscale.
The revision process was completed in four steps. First, the SAQ was checked overall on what needed to be rewritten or removed. After the first researcher made these changes, it was looked at by multiple raters of different levels of education (i.e., undergraduates, graduates, and faculty). These raters gave a new perspective and new ideas of what could be added for the drinking problems section. The scale was reconfigured to be shorter and to reflect today’s language. This study (i.e., the factor analysis) is the fourth step in the process to validate the questionnaire.
Participants (N=255) were 18- to 58-year-olds (Mage = 20.2, SD = 4.152). Participants were directed to the SONA research site to complete the SAQ and, upon completion, were given credit to use in classes that either required or had extra credit opportunities for the student. Data was collected through RedCap and analyzed on SPSS.
The Kaiser-Meyer-Olkin measure of sample adequacy was .91, and Bartlett’s test of sphericity was significant (X 2 (351) = 5428.14, p < .001). Since the KMO and Bartlett’s test were both significant, this meant that it was appropriate to run an exploratory factor analysis. Factor analysis extracted two factors, high-risk drinking consequences, and hangover/blackouts, both with eigenvalues above one. A reliability analysis reported that factor one had a Cronbach's Alpha of .96, and factor two had a Cronbach's Alpha of .87.
These results that the revisions to the SAQ resulted in the drinking problems subscale to now reflect two separate factors. All drinking problems questions were retained as meaningful to the model, with a hangover and blackout-related questions forming their own new factor separate from general drinking consequences. The next step in our study will be to examine the correspondence between the scales other three factors and these two factors of the drinking problems subscale. Future research should also be conducted on a larger sample size to examine the stability of factor analytic findings.
The Revision of the Student Alcohol Questionnaire: A Validation Study
College campuses are a common location for individuals to experience alcohol consequences. Those consequences: health, legal, and/or academic, could cause short- or long-term repercussions on the student. Students likely have been provided education about these consequences from a range of sources (e.g., parents, friends, health courses). It is important to understand what knowledge of alcohol-related information students retain and how that information may impact their decisions around risk behavior engagement.
The Student Alcohol Questionnaire (SAQ; Engs &Hanson, 1973), contains 4 subscales: drinking patterns, problems related to alcohol, knowledge of alcohol, and alcohol attitudes. This measure provided an important model for assessment of students' understanding of alcohol use consequences, however the questionnaire the language had not been updated since it was created. The current study had two aims; 1) to revise the SAQ with language that would be more accessible to current students and reflective of modern drinking trends, and 2) to examine the factor structure of the drinking problems subscale.
The revision process was completed in four steps. First, the SAQ was checked overall on what needed to be rewritten or removed. After the first researcher made these changes, it was looked at by multiple raters of different levels of education (i.e., undergraduates, graduates, and faculty). These raters gave a new perspective and new ideas of what could be added for the drinking problems section. The scale was reconfigured to be shorter and to reflect today’s language. This study (i.e., the factor analysis) is the fourth step in the process to validate the questionnaire.
Participants (N=255) were 18- to 58-year-olds (Mage = 20.2, SD = 4.152). Participants were directed to the SONA research site to complete the SAQ and, upon completion, were given credit to use in classes that either required or had extra credit opportunities for the student. Data was collected through RedCap and analyzed on SPSS.
The Kaiser-Meyer-Olkin measure of sample adequacy was .91, and Bartlett’s test of sphericity was significant (X 2 (351) = 5428.14, p < .001). Since the KMO and Bartlett’s test were both significant, this meant that it was appropriate to run an exploratory factor analysis. Factor analysis extracted two factors, high-risk drinking consequences, and hangover/blackouts, both with eigenvalues above one. A reliability analysis reported that factor one had a Cronbach's Alpha of .96, and factor two had a Cronbach's Alpha of .87.
These results that the revisions to the SAQ resulted in the drinking problems subscale to now reflect two separate factors. All drinking problems questions were retained as meaningful to the model, with a hangover and blackout-related questions forming their own new factor separate from general drinking consequences. The next step in our study will be to examine the correspondence between the scales other three factors and these two factors of the drinking problems subscale. Future research should also be conducted on a larger sample size to examine the stability of factor analytic findings.