The utility of collateral informants in substance use research involving college students

Abstract

Collateral informants have been routinely included in substance abuse treatment research to corroborate subject self-reported alcohol and other drug use. However, only a few studies to date have examined subject–collateral correspondence with respect to non-clinical populations (e.g., college students). The purpose of the present study was to examine the associations between college students self-reported substance use and corresponding collateral (i.e., friends') reports. A total of 100 subject–collateral pairs were recruited from psychology courses at a large public university located in the Southeastern, United States. Subjects and collaterals provided information specific to their own, as well as their friend's, recent (i.e., last 90-days) substance use. Study data yielded moderate to good, statistically significant, correlations between subject–friend pairs for each type of substance use. Discrepancy analyses revealed that the majority of subjects reported greater substance use relative to their collateral reports. This pattern of response (i.e., subject reporting greater use) is consistent with the extant literature. In addition, the friend's personal substance use appeared to influence his/her report of the subject's alcohol and other drug use. It appears that college student self-reports regarding alcohol and other drug use are reasonably accurate.

Introduction

To date, there have been only a few studies that have examined correspondence between subject and collateral estimates of subject drinking/drug use among college students (Borsari and Carey, 2005, Laforge et al., 2005, Marlatt et al., 1998, Stacy et al., 1985). Each of these studies demonstrated moderate to good, statistically significant, correlations between subject and collateral reports for specific alcohol use variables. Laforge et al. (2005) and Baer et al. (2001) examined specific contextual variables associated with agreement scores between student self-reports and collateral estimates and found that collaterals who reported a greater frequency of drinking with their subject counterpart, higher levels of confidence in their reports of the subject's use, and a closer relationship with the subject (e.g., boyfriend/girlfriend) provided estimates that were more consistent with the subject's self-reported alcohol use.

Substance use research involving college students typically does not include collateral informants. At this time, little is known about subject–collateral correspondence specific to drug use other than alcohol. The purpose of the present study was three-fold: 1) to examine the associations between college student self-reported substance use and corresponding friends’ reports; 2) to evaluate the utility of friends as a collateral source; and 3) to assess the extent to which the collateral's own self-reported substance use influenced his/her estimate of the subject's use.