The past few years has seen the rapid growth of data mining approaches for the analysis of data obtained from Massive Open Online Courses (MOOCs). The objectives of this study are to develop approaches to predict the scores a student may achieve on a given grade-related assessment based on information, considered as prior performance or prior activity in the course. We develop a personalized linear multiple regression (PLMR) model to predict the grade for a student, prior to attempting the assessment activity. The developed model is real-time and tracks the participation of a student within a MOOC (via click-stream server logs) and predicts the performance of a student on the next assessment within the course offering. We perform a comprehensive set of experiments on data obtained from two openEdX MOOCs via a Stanford University initiative. Our experimental results show the promise of the proposed approach in comparison to baseline approaches and also helps in identification of key features that are associated with the study habits and learning behaviors of students.