Measuring change in academic self-concept resulting from curricular and instructional innovations.

The paper begins with a discussion of the psychometric and psychological problems involved in attempts to measure changes in human characteristics. The Rasch psychometric model is proposed as a model which has the potential of alleviating or eliminating many of the problems. The model is applied to a set of responses to an academic self-concept instrument administered to seventh-, eighth-, and ninth-grade students. Items that failed to conform adequately to the model at every grade level were eliminated from the scale. The resulting scale was found to possess several properties which permitted its use in the measurement of school-induced change in academic self-concept.