A COGNITIVE DIAGNOSTIC ASSESSMENT OF STOU STUDENTS USING DINA MODEL

Abstract

Objectives of the study were: 1. To assess test quality using the classical test theory and item response theory 2. To perform a cognitive diagnosis of STOU students using DINA Model, 3. To study in-depth data obtained from the cognitive diagnosis relating to students’ non-mastery and 4. To propose approaches to improve instructions of the diagnosed courses. The research was design by using mixed-method methodology. The data consisted of mid-term test scores from the 1^st semester of the academic year 2017 from the selected 6 courses. The data were analyzed by DINA Model and content analysis.

Findings were as follows: 1. The test quality had reliability coefficients ranged from 0.77 to 0.85 and difficulty levels were high and very high by using the item response theory, it was shown that the number of items having high and very high difficulty levels are greater than the results yielded by using the classical test theory. 2. There were 5 courses having students with non-mastery in all units and the probability of mastery based on individual unit showed that the first booklet was easier than the second booklet, 3. Courses involving many formula and symbols were deemed to be difficulty to students, In addition, courses that feature highly-detailed lecturing on complicated matters which content in each unit draws from unrelated topics and 4. The university should allocate extra financial support for aiding students in courses with very high difficulty level; If the contents involve calculation, infographics can be very useful. Units with highly complicated contents should be provided with conclusions at the end of the unit or the section; The courses with difficult contents should assign score-based activities and provide short video clips; Teachers should be assigned based on their expertise and teaching techniques and The assessment should be designed to assess various aspects of student performance.