A Practical Method for Identifying Significant Change Scores

Publisher Website
Google Scholar
Abstract
The current literature on measuring "change" contains several methods for estimating difference scores for pretest-posttest designs. While the significance of group difference scores obtained with these techniques can be evaluated using standard statistical tests, the researcher in applied settings is frequently more concerned with identifying individuals who are most (or least) influenced by a particular treatment. This paper presents a simple, easy-to-apply test of significance for identifying such individuals. The technique requires as input 3 pieces of information: true difference scores, the reliability of the obtained differences, and their standard error of measurement. Selection can then be done on the basis of visual inspection of the distribution of difference scores.

This publication has 6 references indexed in Scilit: