Differential Responses to Personalized Learning Recommendations Revealed by Event-Related Analysis

Educators are increasingly embracing personalization in online and blended learning programs as a means of focusing students' investment of time and energy into learning plans that are best tailored to their individual needs. When personalized learning tools are deployed into structured learning environments like schools, however, educators and students must consider program provided recommendations alongside potentially immutable factors like set daily schedules, mandated curricula, and student needs in other content areas. These on-the-ground factors make researching the impacts of personalized learning challenging because they are difficult to measure directly, especially for digital programs deployed at scale. Inspired by a widely influential methodology in brain imaging, we tackled this challenge by employing an "event-related" approach that emphasizes changes in student behavior that are time-locked to changes in program provided usage recommendations. Our analysis reveals that while student usage time can often be quite far from the amount recommended, students nevertheless respond to changes in program recommendations by adjusting usage in a corresponding manner. We further extend this general approach to demonstrate that students more often stayed on track toward their end of year goals following a week where they met or exceeded their program provided recommendation. Through these examples, we demonstrate the value of an event-related approach towards understanding how personalized paths can positively influence student learning. [For the full proceedings, see ED607784.]