Collaborative E-Testing Construction Using Prediction Tools To Improve Test Reliability

• Autores: Pokpong Songmuang, Maomi Ueno, Toshio Okamoto
• Localización: International journal of the computer, the internet and management, ISSN 0858-7027, Vol. 14, Nº. 1 (AGO), 2006 (Ejemplar dedicado a: Suplemento 1: Proceedings of the Third International Conference on eLearning for Knowledge-Based Society), págs. 24-24
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • This paper focuses on one of the features of e-testing, which has not been discussed until now, called collaborative test construction by several test-authors in distant places. It is well known that collaborative work has many advantages. The main idea of this paper is to improve test qualities by applying the unique advantages of collaborative work to e-testing construction and using prediction tools while constructing test. The analysis of collaborative e-testing construction identified the number of test-authors as the most important factor in test validity, while test reliability depends more on the participation of an expert. Based on these findings, a collaborative e-testing construction system was developed that using several prediction tools to improve the reliability of tests constructed by novice test-authors. An experiment in which a novice and an expert test-author each constructed tests by using these prediction tools in one case and not using in the other showed that those constructed using them were more reliable, although those constructed by the expert had even higher reliability.