A method for analyzing retrospective pretest/posttest designs: II. Application

Abstract

Koele and Hoogstraten (1988) argued that, when using a retrospective pretest/posttest design, a test on the presence of a response-shift bias is necessary before reliable estimates of a treatment effect can be obtained. To that purpose, they developed a hierarchical model fitting procedure aimed at detecting the presence of a treatment effect and/or a response-shift bias. In the present article this procedure was applied in a reanalysis of the results of an earlier study. © 1988, The Psychonomic Soceity, Inc.. All rights reserved.