Design, Efficacy, and Methodology Considerations for Brief Interventions: Intervention Delivery and Incentives

This research evaluated the effects of two methodological factors (i.e., delivery modality and incentives) on attrition, data quality, depth of processing, and perceived value of a personalized normative feedback (PNF) intervention targeting drinking reduction in college students. We expected in lab (vs. remote) participation would be associated with lower attrition, better data quality, and greater depth of processing and intervention value. We further expected that being offered an incentive (vs. not) would be associated with less attrition and better data quality, but lower depth of processing and intervention value. Finally, we expected depth of processing and intervention value to be related to reductions in drinking among PNF participants. Heavy drinking college students (N = 498) participated in a 2 (in-person vs. remote delivery) x 2 (incentive [$30 gift card] vs. no incentive) x 2 (PNF vs. attention control) design. Follow-ups occurred remotely 3 and 6 months later; all participants were compensated with a $15 giftcard per completed follow-up. In-lab participants and those offered an incentive were less likely to drop out of the study. In-lab participants gave higher quality data at baseline and reported greater depth of processing and higher intervention value. PNF was related to reductions in drinking, but depth of processing and intervention value were not, nor was the interaction with PNF. Results suggest several benefits for motivating students to come into the lab and a few for offering an incentive but suggest that this is not a necessary requirement for PNF brief interventions to work. Public Health Significance Statement This research investigates how interventions are delivered (remotely or in-lab) and whether participants are paid as two important methodological factors in brief alcohol interventions. Participants who came into the lab were less likely to drop out, offered higher quality data, and processed the intervention material more deeply and with greater value; however, the intervention was not more efficacious for these participants and incentives were only associated with lower drop out. This work suggests particular benefits to data quality for researchers who ask participants to come into the lab, despite this benefit not spilling over to reductions in drinking.