ELSEVIER DEPENDENCE Drug and Alcohol Dependence 60 (2000) 19-27 www.elsevier.com/locate/drugalcdep Self-report stability of adolescent substance use: are there differences for gender, ethnicity and age? A.M. Shillington * , J.D. Clapp School of Social W ork, College of Health and Human Services, San Diego State University, 5500 Campanile Drive, San Diego, CA 92182-4119, USA Received 6 April 1999; received in revised form 15 October 1999; accepted 18 October 1999 Abstract This study used the National Longitudinal Survey of Youth and is the first to examine a 2-year report stability of substance use among adolescents while stratifying for gender, ethnicity, and age. This study examined lifetime use and age at onset report stability, and the internal consistency of reports while excluding nonusers and incident cases (respondents who may have initiated substance use between the two reporting periods) from the analyses. Report agreement of lifetime use for each substance was over 80% and was highest among alcohol users and lowest for cigarette and marijuana users. Report agreement was higher for female compared to male cigarette users. External consistency of lifetime use of cigarettes and marijuana was higher for whites compared to Hispanic or African American adolescents. Internal consistency was high but lifetime use reports were more stable than age at onset reports. 0 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Substance use reports; Agreement rates; Reliability; Adolescents; Surveys 1. Introduction The reliability, validity and stability of self-report measures of alcohol and other drugs (AOD) are of great importance to researchers investigating adolescent AOD use patterns. As many researchers have noted, self-reports of AOD use are the most common source of information concerning adolescents’ AOD use (O’Malley et al., 1983; Bailey et al., 1992). Yet the measurement problems of reliability and validity could be greater for AOD than other behaviors because all substance use behaviors are illegal among adolescents. In recent years, a handful of studies have examined the reliability and validity of self-reports among adolescents using general population samples (O’Malley et al., 1983; Barnea et al., 1987; Bailey et al., 1992; Smith et al., 1995). These studies have yielded mixed findings, but generally support the validity and reliability of self-reports of AOD use given by adolescents. Such findings are consistent with studies examining this issue * Corresponding author. Tel.: + l-619-5946860; fax: + 1-619- 5945991. E-mail address: ashillin@mail.sdsu.edu (A.M. Shillington) within the adult general population (Midanik, 1988; Shillington et al., 1995). Other studies have examined the psychometric prop- erties of self-report substance abuse inventories among treatment or pre-treatment adolescent populations (Winters et al., 1991; Stinchfield, 1997). Although this approach has utility for treatment professionals and researchers concerned with examining the efficacy of treatment, there are fundamental differences in treat- ment and general population adolescents. For instance, some adolescents often minimize their AOD use when queried before treatment of their addiction (Aiken, 1986). In contrast, others might exaggerate the level of their AOD use to gain access to treatment or avoid incarceration (Winters et al., 1991). As such, studies examining the reliability of self-reports using treatment or pre-treatment populations are less applicable to re- searchers interested in AOD epidemiology at the popu- lation level. Reliability and validity of self-reports of AOD use have been conceptualized in several ways. Several stud- ies have utilized external validation of self-reports in which self-reported data are validated through biologi- cal measures or eyewitnesses (Single et al., 1975; Smith 0376-8716/00/$ - see front matter 0 2000 Elsevier Science Ireland Ltd. All rights reserved PII:SO376-8716(99)00137-4