Measure school learning through Rasch Analysis: The interpretation of results

Abstract

From the beginning of the docimology (1922), the assessing of the learning has become increasingly important in the interpretation of study and, in general, in evaluation of the quality in education. Several measurement instruments have been offered such as: distribution pentenaria, t scores, item analysis. Recently, in the context of item response theory (IRT) it is emerging a new measurement technique based on Rasch Analysis in order to obtain an objective measure that does not dependent on either the characteristics of the measuring instrument and does not depend on the skills of individuals. Although Rasch Analysis allows of offering some interesting research, it remains open issues for the effective implementation of the instrument. The main problems are: the construction of a test and the interpretation of results. In this paper we address the second issue: the interpretation of results. At this regard it is proposed a conversion of logit scores into an evaluation scale more comprehensible and more closed to the measurements of assessing adopted by teachers typically expressed on the scale from 0 (very bad) to 10 (very good). In detail, after having briefly presented the technique of Rasch Analysis, it will be presented an application of the method on 34 students with the purpose to analyze the ability in learning of mathematics. A conversion of the logit output in to a new measurement scale from 0 to 10 has been proposed to make more easy the interpretation of the score © 2010 Published by Elsevier Ltd.