Do the Critical Success Factors From Learning Analytics Predict Student Outcomes?

Abstract

This article starts with a detailed literature review of recent studies that focused on using learning analytics software or learning management system data to determine the nature of any relationships between online student activity and their academic outcomes within university-level business courses. The article then describes how data was collected from an online course in Moodle learning management system and the student test scores are compared with the engagement learning analytics indicators to measure the strength and predictive nature the hypothesized relationships. The results were less than definitive and in some ways even puzzling. Despite the paucity of positive results, the findings indicated there was very little correlation between student online practices and their academic outcomes in a qualitative business course. On the other hand, several explanations were offered to account for these results, including small sample size, the specialized nature of the course, and the likelihood that students were downloading the material instead of interacting online. The article closes by suggesting directions for future learning analytics research in higher education.

A matter of trust: Higher education institutions as information fiduciaries in an age of educational data mining and learning analytics

“…Significant research and technological development have focused on predicting student success in a given course and providing instructors tools to intervene when success looks unlikely (Macfadyen & Dawson, ). Common learning management systems (LMSs), like Canvas and Blackboard, all have predictive features based on clickstream data, and increasingly LMSs are incorporating advanced affordances that create social network maps to determine students' connections (or lack thereof) with their peers and instructor (Filvà, García‐Peñalvo, & Forment, ; see Strang, ). Some new systems use geolocation data derived from sensors and card swipes to take attendance, while biometric data can inform judgments of in‐class student engagement (Alcorn, ; Schiller, ).…”

Drill-Down Dashboard for Chairing of Online Master Programs in Engineering

“…Only two paper discussed in any way the role of masters program chairs. Strang [12] relates different approaches of employing Learning Analytics (LA) in Moodle, distinguishing between Course-level and Organizational-level information. Uskov et al [13] provide a larger analysis of these various levels of depth, two of which are related to master program chairs: "Concentration/minor program level" and "Departmental/program of study/curriculum level".…”

Learning Analytics: a Metacognitive Tool to Engage Students

“…It is categorized as a hierarchical model technique because from a visual point of view, decision trees are tree-like constructions that demonstrate the categorization of data, where the main node is called the root and other lower nodes referred to as leaves (Kabakchieva, 2013;Strang, 2016). Decision trees usually consist of several branches that indicate various features, and every single leaf node on those branches point to a different class (Kabakchieva, 2013;Strang, 2016). A decision tree as a method is designed to illustrate connections between various features (Kabakchieva, 2013).…”