While MOOC platforms allow universities to implement various strategies such as brand promotion and student recruitment, the alarmingly low retention rate suggests a need to explore the critical factors that influence students' course retention. So far, studies on MOOC platforms focus either on the students' individual factors (i.e., students' personal factors such as perceived value) or situational factors (i.e., external influences shaping students' behavior, such as system quality) for students' learning, thus lacking a complete view of those determinant factors. This study integrates the TAM model with the TRA model to analyze the roles of three important antecedents (learning goal orientation; LGO, instructor reputation; IR, & parasocial interaction; PI) on university students' perceived value (PU) and learning attitude (LA), two critical predictors of MOOC retention (CR). Using data from an online survey of 449 Chinese university students, the hypothesis model was tested using PLS. We found that LGO, IR, and PI each positively affect PU; LGO, IR, and PI each positively affect LA; PU and LA each positively influence course retention (CR), with each impact enhanced by tutor intervention (TI). The theoretical and practical implications of such findings are presented.