INFO 4100/5101: Learning Analytics

Technology is transforming the ways we learn and teach. Learning is increasingly occurring on the Internet or through educational software, which has led to an explosion of educational data. How can all of the interaction, discourse, and performance data that are generated from online courses, learning management systems, and student discussion forums be used to improve educational effectiveness and support basic research on learning? This introductory course on Learning Analytics provides a survey of educational data science methods (predictive modeling, classification, regression, causal inference, and more) and learning science theories (active learning, Bloom’s taxonomy, metacognition, self-regulated learning, and more). Students will collect and analyze their own learning trace data as part of the course.

Learning Objectives
1. Explain and demonstrate key insights from research in the learning sciences on how learning works
2. Select and apply methods from educational data mining and learning analytics to analyze different kinds of educational data
3. Evaluate the results of different methods for different applications
4. Compare the strengths and weaknesses of methods for different applications
5. Identify the potential benefits and risks of learning analytics for students, teachers, and institutions

The course is described in more detail in this book chapter on Learning Analytics Education.

COMM/INFO 4800: Behavioral Science Interventions

“There is nothing as practical as a good theory,” Kurt Lewin wrote. Today behavioral scientists build on social scientific theories about human behavior to develop new intervention approaches that address major challenges facing our society: poverty, poor health, educational inequalities, and many more. This course is designed as a senior capstone seminar that equips students with the knowledge and skills to analyze social problems, consider the ethical implications of intervention, design and pilot appropriate interventions, implement and test them online, analyze and interpret the results, and present policy-relevant findings. The course combines applied quantitative research methods and applied social behavioral science theories to prepare students for careers in research, data science, consulting, and policy evaluation.

Learning Outcomes
1. Analyze social problems in order to identify opportunities for improving social conditions
2. Create and evaluate intervention designs and implementations to change social outcomes
3. Apply quantitative research methods to analyze behavioral science interventions
4. Interpret and present the results of behavioral science interventions
5. Explain the ethical ramifications of intervening in people’s lives

Student reviews: Annie (Sp’20), Tanmay (Sp’20).

INFO 6750: Causal Inference and Design of Experiments (Fa’18)

This PhD-level research methods course provides a hands-on introduction to topics in causal inference, design of experiments, and open science. Topics covered: potential outcomes framework for causal inference, sampling distributions and randomization inference, types of random assignment, uses of covariate measures in inference, statistical power, approaches to non-compliance, quasi-experimental designs, replication, replicability, and open science practices. Short lectures, seminar discussion, and hands-on practice are interwoven on a weekly basis to develop applied analytic skills. The course is targeted at quantitative social scientists who strive to draw causal inferences in controlled lab or field settings and uncontrolled real-world settings. This course should be taken after completing an introductory course in statistics and an introductory quantitative research methods course. Ideally, students taking this course are already conducting (quasi-) experimental research and are looking for ways to improve the precision and power of their inferences.