[Paper Review] Indicators of Good Student Performance in Moodle Activity Data

Ewa Młynarska, Derek Greene+1 more|arXiv (Cornell University)|Jan 12, 2016

Innovative Teaching and Learning Methods16 citations

TL;DR

This study analyzes Moodle activity logs from 60 undergraduate Computer Science assignments to identify early predictors of high student performance. It finds that early submission, high overall activity, and evening activity are strong indicators of good grades, though some assignments show negative correlations between effort and performance, suggesting potential issues with course design or assessment.

ABSTRACT

In this paper we conduct an analysis of Moodle activity data focused on identifying early predictors of good student performance. The analysis shows that three relevant hypotheses are largely supported by the data. These hypotheses are: early submission is a good sign, a high level of activity is predictive of good results and evening activity is even better than daytime activity. We highlight some pathological examples where high levels of activity correlates with bad results.

Motivation & Objective

• To identify early indicators of good student performance using Moodle log data.
• To test three key hypotheses: early submission, high activity, and evening activity as predictors of high grades.
• To detect anomalous cases where high activity correlates with low grades.
• To support timely academic interventions by identifying at-risk students based on behavioral patterns.

Proposed method

• Collected and anonymized Moodle log data from 360 courses, focusing on 60 assignments with complete submission records.
• Calculated Pearson correlation between assignment grades and time-to-deadline for submissions.
• Defined activity events using a curated subset of Moodle events (e.g., view assign, submit for grading, view page).
• Computed correlation between grade and number of activity events in the two weeks before submission.
• Segmented activity into daytime (8am–6pm) and evening (6pm–midnight) intervals to compare predictive power.
• Used time-series analysis to examine activity patterns and identify clusters of high or low performance.

Experimental results

Research questions

• RQ1 Does early submission correlate with higher grades in Moodle-based courses?
• RQ2 Is a high level of Moodle activity before submission predictive of better academic performance?
• RQ3 Does evening activity show stronger correlation with high grades than daytime activity?
• RQ4 Are there cases where high activity is negatively correlated with grades, and what might explain them?

Key findings

• 42 out of 60 assignments showed a positive correlation between time remaining before submission and final grade, supporting early submission as a predictor of success.
• 41 out of 60 assignments showed a positive correlation between overall activity levels and grades, confirming that higher engagement is generally linked to better performance.
• Evening activity had a higher average correlation with grades (0.13) than daytime activity (0.07), indicating that timing of engagement matters.
• Negative correlations between activity and grades were observed in 19 assignments, particularly in Level 1 courses, suggesting possible issues with time management or assessment design.
• Level 3 courses showed the highest mean correlation (0.33) between submission timing and grades, indicating better time management among advanced students.
• Anomalous cases with high activity and low grades were linked to vague deadlines, lack of late penalties, or marking scheme artifacts, highlighting the need for course-level analysis.