The majority of research on adolescent drug use is school-based and assumes that data collected in the classroom on a single day are representative of the adolescent population. However, in inner-city schools, absenteeism is a recalcitrant problem. When such school-based surveying is performed, a substantial number of students will be missed. In fact, it may be the adolescent out of the classroom who is at greater risk. Of particular concern is the potential distinction between the drug-related risk behaviors of those students in the classroom and those who are not. Thus, to reach the adolescents who are at greater need for preventive programming, efforts first need to be made to assess the bias introduced in one-shot school-based sampling. This research proposes to determine if classroom-based sampling in urban settings introduces measurable bias when the population of interest is adolescents. It will measure the bias associated with: 1) limiting the sample of adolescents to those students who are in the classroom at the time of survey administration, 2) basing any adjustments for absenteeism on the self-identified chronically absent student who happens to be in the class at the time of survey administration. Although the study is geographically limited to New York City, its findings will be relevant to many inner-city school systems with high rates of absenteeism.
| Weitzman, B C; Guttmacher, S; Weinberg, S et al. (2003) Low response rate schools in surveys of adolescent risk taking behaviours: possible biases, possible solutions. J Epidemiol Community Health 57:63-7 |
| Guttmacher, Sally; Weitzman, Beth C; Kapadia, Farzana et al. (2002) Classroom-based surveys of adolescent risk-taking behaviors: reducing the bias of absenteeism. Am J Public Health 92:235-7 |
