An Instructor-based and Course-based Learning Analytics for Identifying At-Risk Students

碩士 === 元智大學 === 資訊工程學系 === 104 === This study applies midterm assessment and grading historical data to develop a prediction mechanism for identifying at-risk students of fail and low credit-earned rate. The existed midterm assessment grades students as A, B, C, and D where D indicates at-risk stude...

Full description

Bibliographic Details
Main Authors: Kung-Chen Peng, 彭恭箴
Other Authors: Chih-Yueh Chou
Format: Others
Language:zh-TW
Published: 2016
Online Access:http://ndltd.ncl.edu.tw/handle/8f29f7
Description
Summary:碩士 === 元智大學 === 資訊工程學系 === 104 === This study applies midterm assessment and grading historical data to develop a prediction mechanism for identifying at-risk students of fail and low credit-earned rate. The existed midterm assessment grades students as A, B, C, and D where D indicates at-risk students of fail. Students who got more than 4 D are currently regarded at-risk students of low credit-earned rate. This study proposes a prediction mechanism to apply midterm assessment and grading historical data to estimate students’ risk indicators of fail and low credit-earned rate. The evaluative results revealed that the proposed identification mechanism of at-risk students of fail had similar accuracy than the existed midterm assessment. The results also showed that the proposed identification mechanism of at-risk students of low credit-earned rate had higher accuracy than the existed 4 D identification mechanism in Accuracy and Precision. The proposed mechanism estimates students’ risk of fail and low credit-earned rate. The risk can be used to cluster students into higher and lower risk students. The evaluative results revealed that students with the higher risk had higher rate of fail and low credit-earned rate. Keyword: Learning analytics, identification of at-risk students, midterm assessment, risk indicator