Personalized feedback in digital learning environments: Classification framework and literature review

Digital learning technologies offer many opportunities to personalize instruction and learning in K-12 and higher education. In the last ten years, a growing body of research described personalized feedback implementations and investigated their effects on educational outcomes. Building on personalized education and adaptive learning systems models, this review provides an analytic framework to summarize key features of personalized feedback implementations and main empirical results. The systematic literature search resulted in 39 studies published in the last ten years. We found that scholars developed and investigated personalized feedback on the microscale, mesoscale, and macroscale of digital learning environments. However, the adaptive sources (To what is feedback adapted?) are mainly restricted to the current knowledge level and learning behavior data. Other interesting data sources for feedback adaptation remain underresearched, e.g., emotional state measures, progress measures, learning goals, or personality traits. Only a minority of the reviewed studies provided an empirical or theoretical rationale for assigning feedback messages to different types of students. Most studies report positive or at least mixed or neutral effects of personalized feedback on educational outcomes. This review discusses several implications for future directions in research on digitalized and personalized feedback. This study also adds to previous literature reviews on automatic and adaptive feedback that did not clearly distinguish task-adaptiveness and student-adaptiveness in digital feedback examples.