PRELIMINARY DEVELOPMENT OF LEARNER SUPPORT PREDICTION MODEL FOR E-LEARNING BASED ON SELF-REGULATED LEARNING FACTORS

The purpose of this research is to develop learner support prediction model for e-learning courses based on self-regulated learning (SRL) factors. The model aims to predict learners’ needs and typical behavior to help them complete e-learning courses without any face-to-face instruction. There found four SRL factors including (1)cognitive, (2)affective, (3)help-seek, and (4)self-directed factors to determine learner types at the forethought phase of Zimmerman’s three-phase model for SRL (1998). The standardized questionnaire for SRL skills developed by Wolter, Pintrich, & Karabenic (2003) were modified and shorten based on the exploratory and confirmatory factor analysis results and it was utilized as an instrument. It consisted of 40 questions and had high reliability (Chronbach alpha = .931). The subjects included 305 students from three e-learning classes without any face-to-face instruction. In order to develop a preliminary model, 149 students’ data were analyzed and the model validation was done with other data of 156 students. The learner types were categorized as a result of a combination of three levels (positive, neutral, negative) for each SRL factor. Among 81 learner types (3*3*3*3), the types with the upper 27% and lower 27% on a completion rate of a course were selected. Then using the selected data, the multiple regression correlation and path analysis were performed for SRL factors on the completion rate. The significance of this research is to provide information on learners’ needs priori to a semester, which helps instructors and mentors to plan learner support and estimate necessary efforts to help students complete a course.