Newswise — How young adults perceive their own drinking habits may distort their self-reported alcohol use, according to a new study in Alcoholism: Clinical and Experimental Research. The study scrutinizes the accuracy of participants’ self-reported drinking — a frequent component of alcohol research. Self-reports are prone to inaccuracies, especially in recalling past use. To improve accuracy, researchers often incorporate both “real time” self-reports and retrospective assessments. When these two reports diverge, however, the implications for research are not well understood. For this study, investigators assessed how these two types of self-report differ and what factors may predict inaccurate self-reporting. Unraveling these influences has the potential to improve the accuracy of some alcohol research — and, ultimately, better support people experiencing hazardous drinking.

Investigators worked with data from 310 college students aged 18–24. Participants were surveyed about their alcohol use via a custom smartphone app five times a day (real time reports). The first assessment of the day included questions about yesterday’s drinking (“end-of-day” report). Researchers analyzed discrepancies between those reports and explored the possible impact of certain factors. Potential influences included individuals’ established drinking patterns. In addition, researchers considered the effects of how much participants drank that day (e.g., their number of drinks), whether they consumed liquor and/or cannabis, and how intoxicated participants thought they were. Investigators analyzed data collected over 54 days.

Using statistical analysis, researchers found that real time and end-of-day assessments were reasonably consistent in seven out of ten self-reports. In divergent assessments, real time reported alcohol use could be either below or above that of end-of-day reports. Discrepancies were relatively frequent among participants who had alcohol problems or who drank heavily that day. As participants’ number of drinks increased, their self-reports were more likely to diverge. Discrepancies were also associated with weekends, the use of cannabis or liquor, and participants’ sense of their own intoxication levels.

The study suggested that participants whose alcohol use is problematic may retrospectively overestimate their consumption, perhaps because they identify as problematic drinkers. But participants whose drinking habits are variable, or who perceived themselves as intoxicated on the day in question, may later underestimate how much they consumed. The researchers recommend further studies to establish the optimal use of survey types.Heaviness of alcohol use, alcohol problems, and subjective intoxication predict discrepant drinking reports in daily life. A. Stevens, A. Sokolovsky, P. Treloar, H. White, K. Jackson (pages xxx).

ACER-19-4236.R2