Study on Computer-adaptive Testing: Proposal of a Scaffolding Tool

Abstract

A multiple-choice method is one of the major methods in Classical Testing Theory used at university, in certification exams, and so on. This method enables faculty to score a test easily while requiring a certain amount of quizzes to estimate examinees' abilities with a certain accuracy. This also requires time to answer for examinees and time to create quizzes for faculty. Then it can be considered to adopt CAT, Computer-Adaptive Testing. CAT is another testing theory that reduces the number of quizzes to theoretically half compared to the existing multiple-choice method while keeping the accuracy of estimating examinees' abilities almost the same; however, it is necessary for each quiz to have its own difficulty. In this study, a programming exercise tool with a scaffolding method is going to be suggested aiming at generating quizzes with various difficulties automatically. A scaffolding method is a method that helps examinees to solve problems by providing some hints gradually. By using this tool, what elements are associated with reduction in difficulty can be examined, and they will be used to a quiz generation. Students will be required to answer half the number of quizzes and faculty will need to create much less ones than before. This tool also supports beginners to avoid bias in a difficulty distribution by providing very basic quizzes so that the beginners will be encouraged to brush up their skills and more detailed classification will be acquired.



Author Information
Hirotaka Azuma, Tokyo Denki University, Japan
Hironobu Takenouchi, Tokyo Denki University, Japan
Tatsuyuki Takano, Kanto Gakuin University, Japan
Takashi Kohama, Tokyo Denki University, Japan
Osamu Miyakawa, Tokyo Denki University, Japan

Paper Information
Conference: ACE2020
Stream: Design

This paper is part of the ACE2020 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window

Video Presentation


Posted by amp21