Study Example:

Yağcı, M. (2022). “Predicting Student Performance Using Educational Data Mining Techniques.”

This study uses EDM to analyze student performance data from an online learning platform. The researcher applied machine learning models to identify which factors—such as quiz scores, time spent on tasks, and participation patterns—best predicted final course outcomes.

✅ What Educational Data Mining Can Tell Us

Predictive insights: EDM can identify students who are likely to struggle or drop out based on early behavior patterns.

Key performance indicators: It can reveal which activities or behaviors (e.g., frequent logins, timely submissions) are most strongly associated with success.

Learning pathways: EDM can map how students navigate content and which sequences lead to better outcomes.

Instructional effectiveness: It can show which materials or tasks are engaging or confusing for learners.

⚠️ What Educational Data Mining Cannot Tell Us

Motivation or emotion: EDM cannot explain why a student is disengaged or struggling.

Contextual factors: It cannot capture home environment, personal challenges, or cultural influences.

Creativity or deep understanding: EDM measures patterns, not the richness of thought or originality.

Causation: It can show correlations, but it cannot prove that one behavior caused a particular outcome.

✅ Overall Insight

The study demonstrates that EDM is a powerful tool for identifying patterns and predicting performance, but it must be paired with human judgment. Data can guide decisions, but it cannot replace the nuanced understanding that teachers bring to the learning process.