Learning Analytics: in the Case of Learning Check Testing, LCT
147 , 2017-02 , 広島工業大学
In response to the extensive use of computer systems in learning, we are now in the age that we can do the learning analytics using the huge-size of database. We have recently established the follow-up program system (FP system) aimed at helping students who need basic learning and aimed at assisting teachers who have to engage in teaching a variety of educational students. The follow-up system consists of the learning check testing (LCT), follow-up program testing (FPT), and collaborative work testing (CWT). Since the FP system was first introduced, four months have passed, and we have accumulated the large scale of testing results. In the LCT we use the item response theory (IRT), which provides more accurate and fairer evaluations of individual abilities than classical test theory does, and thus the IRT has gradually been recognized as one of the proper evaluation methodologies in many testing fields. In this paper, we show the results of learning analytics focused on the LCT results. From the analytics, we have found that 1) five or six items are too small to evaluate the students' abilities, at least ten items are required, 2) although use of only one time testing results do not provide sufficient reliability to ability values, combination of multiple testing results provide ability values with a better reliability, which will give us a new measure to students' abilities, 3) the EMtype IRT worked in estimating the empty elements of item-user response matrix, 4) the learning analytics to LCT results showed us valuable information in leading us to a new direction.