Investigating Relevance of Prior Learning Data Connected through the Blockchain

Abstract

Learners often change their learning environment over the course of their education. This makes it difficult to measure their engagement across different contexts due to a lack of seamless connection and shared analytics across heterogenous learning systems. Previous research has shown that access to prior engagement information of learners can be useful in enabling personalization, learning content design and early identification of problematic prerequisite topics. In this paper, we connect learning systems at two different schools through the blockchain to enable the transfer of learning footprints across both schools. Our primary aim is to investigate the relevance of students’ prior engagement behaviour and provide stakeholders with actionable insights on dashboards. Specifically, we analyze the engagement behaviour in Junior High School grade 3 Math course of students who are currently in High School grade 1. Engagement in this context is defined based on five metrics: self-evaluation, cognitive behaviour, backtracking behaviour, time commitment and content completion rate. We further validate relevance by measuring the correlation between students’ engagement level and their final score. Our analysis shows a significant difference in mean scores of very high and very low engagement students. Also, for each of the courses and scores, we provide stakeholders access to the learning materials used, assessments taken and the solutions by the students. Finally, we present implications for the field and present potential directions on how to use decentralized learner data to improve learning outcomes.