Greer and Mark’s (2016) paper suggested and reviewed different methods for evaluating the effectiveness of intelligent tutoring systems such as Propensity score matching. The current study aimed at assessing the effectiveness of automated personalized feedback intervention implemented via the Student Advice Recommender Agent (SARA) in a first-year biology class by means of statistical matching and by reviewing and comparing four different statistical matching methods (i.e., exact matching, nearest neighbor matching using the Mahalanobis distance, propensity score matching, and coarsened exact matching). Data from 1026 (73% female and 27% male) students who took a first-year biology course at a Western Canadian university were used. Two different measures for balance assessment of the matched data (i.e., % of balance improvement and standardized bias) were used to choose the best performing statistical matching method. Nearest neighbor matching using the Mahalanobis distance was found to be the most appropriate method for this study and results showed a statistically significant but small treatment effect for the group who received personalized feedback. Research and practical considerations were discussed and suggestions for future research are provided.

The field of learning analytics (LA) is developing rapidly. However, previous empirical studies on LA were largely data-driven. Little attention has been paid to theory-driven LA studies. The present scoping review identified and summarized empirical theory-driven LA studies, aiming to reveal how theories were integrated into LA. The review examined 37 peer-reviewed journal articles published from 2016 to 2020 from six databases. Results show that most studies were guided by the theories of self-regulated learning and social constructivism; most integrated theories into LA for better interpreting the data analysis results; and most linked theoretical constructs and log variables directly. Several studies employed well-developed survey instruments to measure theoretical constructs. The review results indicate that LA studies still need to strive for new theory advances in learning. Recommendations for future study are discussed.

Universal Design for Learning (UDL) and Blended Learning (BL) formats, are widely adopted across K–12 learning environments. Upon graduation, preservice teachers may be expected to implement UDL and BL practices. The present study was motivated by the need to provide preservice teachers with live modeling of UDL and BL concepts. Learning analytics data from 197 preservice teachers was examined for engagement with UDL/BL Access features (location, day-of-the-week, time-of-day, and regularity), Content features (screencasts and quizzes), and to determine if there was a relationship between engagement and achievement. Examination of the learning management system login data revealed regular access to the digital content across differing locations, week days, and time of day. Associations were significant between academic performance and all features. Designing the BL digital course components following UDL principles appears to have served as a self-regulation enabler for preservice teachers themselves while providing exemplars to adopt in their future practice.