Tracking and two statistical methods (probability weighting and sample selection modeling) were studied as ways to minimize bias attributable to sample attrition in school-based studies. Data on student smoking from 30 middle schools illustrate that sample weighting yields the best results, with estimates superior to sample selection and much less…

Simple methods for detecting attrition (nonresponse) in longitudinal evaluations are reviewed, focusing on regression-based analyses of data from a longitudinal evaluation. The approaches are illustrated with data from the Fort Bragg Evaluation, an evaluation of a major demonstration in children's mental health services. (SLD)

Problems with conducting randomized field experiments in education are explored. Focus is on problems encountered while evaluating a group of dropout prevention projects. Project planners were asked to manipulate the subject eligibility criteria until they identified as eligible three to four times as many students as they could serve. They were…

The National Education Longitudinal Study of 1988 (NELS:88) serves as an example of how three specific problems of representativeness in a longitudinal study may be approached and overcome. NELS:88 was designed to provide longitudinal data about the educational transitions experienced by students as they leave eighth grade and move through high…