Intermediate level
No prior experience required
Learning Curves and Data Mining for Course Improvement
Flexible schedule
1 week, 6 to 8 hours per week
Instructor feedback
Get guidance on your work
Verified certificate
Share on LinkedIn
*Proof of full-time student enrollment required. Acceptable forms of ID include a letter from your university’s registrar office or an unofficial transcript. Email your documents to learnlab-help@lists.andrew.cmu.edu.
What you will learn
- Use learning curves to analyze how learner performance changes with practice.
- Apply educational data mining methods to identify patterns relevant to course improvement.
- Diagnose bottlenecks, inefficiencies, and misconceptions from learner data.
- Translate analytic findings into concrete redesign priorities for courses or tutors.
Course description
Learning data can do more than report outcomes. It can reveal how performance changes with practice, where learning stalls, and which parts of a course or tutor need revision. Learning curves and educational data mining provide methods for identifying these patterns systematically.
In this course, you will learn how to use learning curves and data mining methods to investigate learning patterns and improve course design. You will examine how to interpret practice data, diagnose inefficiencies or bottlenecks, and turn analytic findings into concrete redesign priorities.
Syllabus
Module 1: Quantitative Cognitive Task Analysis via Data Mining
- Explain how student interactions in an e-learning system can be used to perform Cognitive Task Analysis.
- Recognize the features of a good learning curve.
- Use learning curves to do Cognitive Task Analysis.
- Recognize a bad learning curve.
- Analyze tasks to hypothesize new knowledge components to improve a learning curve.
- Evaluate a new knowledge component model in comparison to an old one.
- Use DataShop tools to support these steps.
- Evaluate and redesign a tutor based on an improved knowledge component model and associated insights.
Meet the instructor
Dr. Ken Koedinger
Ken Koedinger is the Hillman University Professor of Computer Science at Carnegie Mellon University, with appointments in Human-Computer Interaction and Psychology. He holds an M.S. in Computer Science and a Ph.D. in Cognitive Psychology and has experience teaching in an urban high school. He has developed data-sharing and analytics infrastructures that support innovations in learning, including DataShop and LearnSphere, and has used them to improve learning as illustrated in his hundreds of publications. He directs LearnLab and co-founded Carnegie Learning in 1998, the first AI in Education company to bring intelligent tutoring technology into widespread use in schools. His PLUS project provides hybrid human-AI tutoring to middle school math students in schools around the country. He is a fellow of the Cognitive Science Society, the Association for Psychological Science, and the Association for Computing Machinery.
Where to go next
Learning Analytics Foundations: Predicting Student Success
Analyze educational data, build regression and classification models, and predict student outcomes to support better learning decisions.
Predictive Modeling and Knowledge Tracing for Adaptive Learning
Build predictive models and use knowledge tracing to model learner progress and support adaptive learning systems.
A/B Testing for Learning Design
Run A/B tests to compare learning design alternatives and make evidence-based decisions about what works best.