Different Settings, Different Results? A Comparison of School and Home Responses

Abstract

DIFFERENCES in the amount of reported drug use among adolescent respondents in studies conducted in school and at home by our research group at Columbia University raised the question of possible response bias because of the setting of the interview. In a 1971 study among students in 18 high schools, the average level of reported use of marijuana and other illicit drugs was greater in every school studied than the reported use in a small-scale national representative household sample survey of youngsters twelve to seventeen years old completed the same year. Among the 18 high schools, the proportion of students reporting any use of marijuana ranged from a low of 23 percent in one school to a high of 58 percent in another. But in the national sample survey, only 15 percent reported such use (Elinson et al. 1977: Sections 3 and 7). Several possible explanations may account for the difference in results. One is that the 18 schools were not selected to represent all United States high schools. Indeed, some were purposely chosen because drug use was thought to be especially prevalent among their students. Another explanation involves age differences in the two samples. In the study of high schools, approximately 5 percent of the students were over the age of seventeen. Since older youth report using more drugs than those younger (Elinson et al.), this may account for some of the differences in results.