Yenilikçi Eğitim Kurumları İçin Rasch Metodu Temelli Değerlendirme

Abstract

Generating high-quality questions is one of the most significant things to improve learning process quality. One of the contemporary methods of developing high-quality questions is the Rasch Method because of superiority for scale development. The most important advantage of the Rasch method is that it allows the measurement of item calibration which is independent of a person's ability. At the same time, it allows the measurement of a person's ability which is independent of item difficulty. In this way, instructors can criticize the quality of their questions, and educational institutions can update their own assessment systems for higher quality and innovative education. This study aims to be a guide that will meet this need of instructors based on the Rasch method. This approach makes it possible for well-qualified items can be used throughout the learning process to improve the student’s learning experience. Thus it is possible to realize the most appropriate content and test design which is necessary for personalized education. On the other hand, it can analyze items psychometric properties and can improve the quality of evaluation systems for the continuous improvement system of future educational institutions. This study evaluated the properties of mathematics items within the test which is applied during term using the Rasch Model. The test was developed for use within a math course which used a modular approach based on learning objects. Results have been evaluated with a view to making recommendations for the improvement of the test psychometric properties. This approach not only enables course developers to evaluate the quality of their test in term but also provides useful information about their own process performance during the e-learning process and their own ability. Due to enable analyzing the student's ability while the e-learning process ongoing, it is possible to correct one's deficiencies. In this study, answers given to 20 items, provided by 62 students of the Internet and Network Technologies distance education program at the level of the associate degree in the fall semester of the 2019-2020 academic year, were examined. Measurement reports, fit statistics, reliabilities were calculated by the Winstep package program.