Methodological Issues: Drug Abuse Treatment Research in Prisons and Jails (From Drug Abuse Treatment in Prisons and Jails, P 246-260, 1992, Carl G Leukefeld and Frank M Tims, eds. -- See NCJ-138622)

• NCJ Number: 138638
• Author(s): B W Fletcher; F M Tims
• Date Published: 1992
• Length: 15 pages
• Sale Source: Superintendent of Documents, GPO; National Institute of Justice/; NCJRS Photocopy Services
• Publication Type: Program/Project Evaluation
• Language: English
• Country: United States of America

Annotation

This paper identifies several methodological pitfalls encountered in the design and conduct of drug treatment evaluation research with incarcerated populations.

Abstract

Within the context of experimental and quasi- experimental designs, Campbell and Stanley (1963) describe threats to internal validity (whether the results of the research findings are true) and threats to external validity (whether the research findings are representative or generalizable outside the research context). Types of threats to internal validity are history, maturation, instability, testing, instrumentation, regression artifacts, selection biases, attrition, and selection-maturation interactions. The threats to external validity, which apply equally to experimental and quasi-experimental research, include the interaction effects of testing, the interaction of selection and experimental treatment, the reactive effects of the research, multiple-intervention interference, irrelevant responsiveness of measures, and irrelevant replicability of treatments. The process of randomization, when properly implemented, controls for most threats to internal validity. Threats to external validity may be the most troublesome for treatment evaluation research, because they are not susceptible to remedy through such statistical procedures as randomization or covariance analysis. The paper concludes with a discussion of threats to evaluation of drug treatment in criminal justice settings and possible remedies. 54 references