DIMINISHED LIFETIME SUBSTANCE USE OVER TIME: AN INQUIRY INTO DIFFERENTIAL UNDERREPORTING

This study investigates underreporting of lifetime marijuanaand cocaine use in the National Longitudinal Survey of Youth.Comparing responses provided in 1988 with responses providedin 1984, this study evaluates the prevalence and correlatesof two indicators of underreporting, including use denial anduse reduction. At follow-up in 1988, a high rate of underreportingwas evident for both marijuana and cocaine, with rates of usereduction observed for just under one-third of all 1984 marijuanausers and just over one-third of all 1984 cocaine users. Outrightdenial of use in 1988 was almost twice as prevalent among cocaineusers as it was among marijuana users; nearly 19 percent ofall cocaine users denied use at follow-up, compared with nearly12 percent of all marijuana users. Correlates of under reportingvaried by substance and by measure of underreporting. The mostconsistent correlates of underreporting were interview mode,race/ethnicity, and educational status. Those interviewed bytelephone, minority respondents, and those with lower levelsof education were more likely to underreport. Race/ethnicityef fects were particularly striking in our analyses. Black respondents had at least twice the odds of underreporting comparedwith white/other respondents for every indicator of underreporting.Findings are compared with other recent research on underreportingand are discussed in the context of recent substance use prevalencefindings.