Abstract
Background
Many young adults who are enrolled in a university play drinking games (DG), a risky drinking practice that is known to facilitate heavy alcohol consumption. DG participation is associated with increased risk for harm and as such, identifying university students at risk for experiencing DG harms is key. Standardized instruments designed to assess DG behavior and related outcomes are limited by having no established cutoffs to identify those individuals who play DG who are at risk for hazardous alcohol use. In the present study, we developed a short form of the Brief Young Adult Alcohol Consequences Questionnaire-Drinking Games (B-YAACQ-DG) for use among university student samples based on differential item functioning (DIF)-free items across gender, and established cutoffs indicating hazardous use using receiver-operating characteristic curves with AUDIT scores of 6+ as the reference standard.
Method
Students (N = 1,299; ages 18–25; 67.7% women; 71.0% White) from four large public universities in the United States completed a confidential online survey.
Results
Although gender-based DIF on several items on the B-YAACQ-DG emerged, there were 16-items that were DIF-free across men and women. An optimal cutoff score for detection of hazardous use for the full version (23-items) of the B-YAACQ-DG was experiencing 4+ DG consequences in the past month, and 3+ consequences for the 16-item DIF-free version.
Conclusions
Health practitioners and alcohol researchers can use the B-YAACQ-DG alongside the AUDIT, to identify university students who play DGs that are in need of intervention. The B-YAACQ-DG can also be used to assess specific DG harms experienced among students who play DGs.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
Data from this study can be shared but will be contingent upon approval from the ethics review committee at each site.
Notes
1 The B-YAACQ is the short (24 items) version of its parent measure, the YAACQ (48 items) and was developed specifically for use with university-aged young adults. Like the YAACQ, the items on the B-YAACQ measure different domains (e.g., risky behaviors, school/work consequences) of problematic drinking that are likely to be prevalent among university students (Kahler et al. Citation2005). Using Rasch analysis, Kahler et al. (Citation2005) refined the YAACQ and developed the B-YAACQ by eliminating items that (a) did not function similarly for men and women or fit a single-factor structure, and (b) were redundant with other items.
2 We explored AUCs and cut-offs for men and women separately. For women, the results for the 23-item (AUC = .74; cut-off ≥4) and 16-item versions (AUC = .73; cut-off ≥3) were similar to those from the full sample. For men, the results suggested that a cut-off of ≥3 may be appropriate for the 23-item version (AUC = 0.72) and the 16-item version (AUC = 0.70).