A mechanism to support a adaptive assessment system by apply fuzzy measure

• Main Authors: Ming-hung Lin, 林明宏
• Other Authors: Chang-pin Wang
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/98593857952155910228

碩士 === 南華大學 === 資訊管理學系碩士班 === 98 === 　　This research aims at designing computer system to support adaptive assessment with the purpose of tailoring the appropriate question based on the students'' performance. Moreover, the learning performance of students is examined and suggestions are provided for teaching instructions. 　 　　In this research, a set of vector attribute is exploited to represent a single question. This set of vector attribute is categorized into two sections including the theme concept and problem-solving ability. Subsequently, fuzzy statistics testing and attribute-space measuring are performed on these attributes. By determining the implicated concept from the teacher''s explanation for the question and ability required to solve the problem, values are acquired. Furthermore, the value acquired for the implicated concept from the teacher''s explanation of the question is adopted for fuzzy cluster analysis and estimation of magnitude for the attribute value. Accordingly, two groups of questions are respectively established to yield the application for two types of item selection strategy. Hence, personalized testing with assorted difficulty levels or distinct question concept is administered based on the students'' ability. 　 　　Initially, this research employs fraction as the theme and categorizes the dimension of fraction definition into four attributes. Besides, the dimension of mathematical problem-solving ability is divided into three attributes. With the phase-in revision assessment as the fundamental, computer adaptive assessment system which is capable of adding or removing assessment question freely can be constructed.