The Measurement of Change in Groups and Individuals, with Particular Reference to the Value of Gain Scores: A New IRT-Based Methodology for the Assessment of Treatment Effects and Utilizing Gain Scores.
Jorg Prieler & John Raven
Part I of this paper reviews problems in the measurement of change and, in particular, in the calculation of change scores (such as before vs. after difference scores purporting to index variance in individuals responsiveness to such things as instruction, stress, therapy and drugs). Part II describes a new approach to the differential measurement of change among respondents having different levels of ability and, in particular, the calculation and use of change scores.
Editor's note: This version differs from the 10/2/02 version only in layout corrections.