By Shizue Izumi and Masahiko Sue (Shiga University, Japan)
Objective: We examine the characteristics of two introductory statistical courses in the Japanese massive open online course (JMOOC) to clarify what affects the completion.
Background: The Japan Statistical Society has created Statistical 1: basic data analysis in the JMOOC in 2014. The Japan Statistical Society and Biometric Society of Japan have created Statistics 2: fundamentals of statistical inference in the JMOOC in 2015. Each free course is taught by five teachers. A face-to-face classroom session is held with a small charge for each course before the course ends. A survey is conducted before and after the course. Weekly and final quiz consists of seventy questions in total with multiple answer choices, and the completion of a course is for their scores beyond 60%.
Method: The data of their quiz scores is merged with pre- and after survey data with a key of user ID for each course to extract the target students that completed a course and answered both surveys. Relation of selected survey questions like enrolled motivation with the score is examined with a regression tree in R software. Survey data from the face-to-face session for Statistics 2 is separately analyzed.
Results: The completion rate among the enrolled students is 13.7% for Statistics 1 and 22.2% for Statistics 2. Among 852 target students in Statistics 1, 74% has a job in full time or part time. Among 558 target students in Statistics 2, 78% has a job in full time or part time. Prior education of statistics, study hours, and use of e-bulletin board are found to be related to their higher scores in Statistics 2. Having a group discussion is also expected among the students of the face-to-face classroom session.
Conclusion: Results may imply that a student-oriented learning community created through activities in e-bulletin board and a face-to-face session helps some students achieve higher scores in a course. Some tips for learning community will be shown in a poster.