Contextualizing survey questions about alcohol consumption may provide memory cues and increase recall precision. This study tested whether questions about drinking contexts preceding alcohol consumption questions would improve recall of drinking, yielding higher quantities (maximum and volume), than unaided recall. We analyzed 2,363 current drinkers form the Year 2000 National Alcohol Survey. Of these, 2,331 (99%) provided data on the most they drank per day using standard questions and, later, the same questions aided by contextual cues. Comparison indicated that 15% (344) gave higher maximum drinking levels in the contextual cue condition. For these respondents, the Graduated Frequencies measure was repeated. Based on the two GF measures (unaided then aided by contextual cues), 209 respondents (9% of all drinkers) gave higher volumes with the cued measure, averaging 83% higher (mean 256 vs 140 drinks/year, p<<0.0001). They also reported more drinking days/year (mean 83.5 vs 55.7) and more 5+ drinks days (14.4 vs 6.8) with the context-cued approach. However, averaged over all respondents, the cue-enhanced measurement produced only a 3.5% change in overall consumption, with similar changes (3–4%) in frequency of drinking and heavy drinking. Thus, although contextual cuing may increase accuracy, this measure technique cannot fully account for survey undercoverage (typically estimated at 50–65%). Despite need for improved self-report methods for public health studies, the maximum quantities elicited by the two methods showed high Spearman's r (.81) and kappa (.55), indicating good cross-method validation for ordering respondents. See www.arg.org
Learning Objectives: Participants will learn more about self-report alcohol measurement techniques. Participants will better understand survey research methods. Participants will learn about drinking distributions and Graduated Frequencies measures.
Keywords: Alcohol Use, Binge Drinking
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: Alcohol Research Group
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.