Massive open online courses (MOOCs) face persistent challenges related to student performance, including high rates of attrition and low student achievement scores. Previous studies that have examined the performance of students in MOOCs have done so using qualitative analysis and the quantitative analysis of small samples. This study is the first to examine general course features of MOOCs on a large scale and to quantify the influences of these course features on student performance. Informed by the theory of web-based online instruction, this study used two-stage K-means clustering to analyze more than 200 MOOCs that had enrolled about 300,000 students, identifying three patterns of course features among the MOOCs. A MANOVA test and follow-up statistical tests revealed that these patterns of course features influenced the MOOCs’ dropout rates and student achievement scores to statistically different degrees. The implications of these findings are discussed.

Keywords

MOOC · course feature · student dropout · student performance · large-scale data analytics