Suchergebnisse

A solution to the problem of specification error due to excluded variables in statistical models of treatment effects in nonrandomized (nonequivalent) control group designs is presented. The solution relies upon longitudinal observation involving at least two pre tests. The assumptions of a single equation probit model for selection bias, structural invariance, and a first order autoregressive pattern are sufficient to identify the model and treatment effect—with one overidentifying restriction. The estimate of the treatment effect is adjusted for an unmeasured variable that influences both the dependent variable and the selection of cases into treatment and control groups. A small simulation suggests that the use of a ML estimation program such as LISREL, which assumes multivariate normality and linearity, may provide reasonable estimates of treatment effects.

The purpose of this article is to describe statistical procedures to assess how prevention and intervention programs achieve their effects. The analyses require the measurement of intervening or mediating variables hypothesized to represent the causal mechanism by which the prevention program achieves its effects. Methods to estimate mediation are illustrated in the evaluation of a health promotion program designed to reduce dietary cholesterol and a school-based drug prevention program. The methods are relatively easy to apply and the information gained from such analyses should add to our understanding of prevention.

The National Institutes of Health sponsored a conference on the Design and Analysis Issues in Community Trials on March 23,1992, in Bethesda, MD. This article presents a synopsis of each of the seven presentations given at that conference and summarizes the discussion that followed.

This study evaluated the relationship between level of program implementation and change in adolescent drug use behavior in the Midwestern Prevention Project (MPP), a school- and community-based program for drug abuse prevention. Trained teachers implemented the pro gram with transition year students. Implementation was measured by teacher self-report and validated by research staff reports. Adolescent drug use was measured by student self-report; an expired air measure of smoking was used to increase the accuracy of self-reported drug use. Regression analyses were used to evaluate adherence; exposure, or amount of implementation; and reinvention. Results showed that all schools assigned to the program condition adhered to the research by implementing the program. Exposure had a significant effect on minimizing the increase in drug use from baseline to one year. Exposure also had a larger magnitude of intervention effect than experimental group assignment. Reinvention did not affect drug use. Results are discussed in terms of research assumptions about quality of program implementation, and possible school-level predictors of implementation.