Investigating the Benefits of Student Question Generation in Data Science MOOC Assessments

Abstract

Question generation is an important metacognitive learning activity for students that encourages the development of higher-order thinking skills. Question generation can also serve as an alternative assessment method. Yet, it is rarely incorporated into learning settings. I propose a study to investigate the effects of using Multiple Choice Question (MCQ) generation in data science MOOC assessments in two different settings where - i) MCQ generation is a required task, and ii) students have a choice between creating an MCQ and answering MCQs. The findings from this study are intended to provide evidence of the learning benefits of MCQ generation in MOOC assessments, the impact of giving students a choice on the quality of the MCQs they create, and the factors that influence students' choice.