Teaching & Mentoring


Teaching. My research has enhanced online courses in statistics, math, and education, by designing new lessons, problems and learning tools like discussion forums. For a MOOC on Applications of Calculus, I taught the section on item response theory, a statistical model for educational assessments. I taught a one-day course on design and analysis of online experiments at CHI (2014), and was an instructor on data mining of educational experiments, for an NSF-funded summer school at Carnegie Mellon. At UC Berkeley I received ratings of 6.3 and 6.1 (out of 7) as a Graduate Student Instructor for introductory courses in Statistics and Cognitive Science.

Mentoring. I mentored 27 undergraduate students at UC Berkeley, receiving the department's Outstanding Mentor Award. At Stanford, WPI, and Harvard I have mentored 5 graduate students in research projects, covering topics like reinforcement learning in education (Sarah Schultz, WPI) to a machine learning course project on dynamic personalization and recommender systems (Louie Hoang, Harvard). I am on the dissertation committee of Merrin Oliver, and a mentor for an NIH doctoral training grant to Samantha Bernecker on online cognitive behavioral therapy. I was a science & technology mentor for startup teams in an "Lean Startup" entrepreneurship class at Berkeley's Business School. I was selected as a mentor by the faculty teaching the course, Steve Blank and Jerry Engels, because they were using the same methodology they developed for the NSF's Innovation Corps program, which teaches faculty to convert their research into marketable products.

Approach & Philosophy

My goal is to empower students to invent new solutions to real-world problems, which requires interdisciplinary training that combines the methods of computer science and social-behavioral science. My approach is to apply research on learning and technology to my own teaching and mentorship. New researchers need a broad range of knowledge and skills to be available at the moment they are developing software, applying algorithms, or analyzing data.

I provide living and continually improving knowledge bases through wikis, Google Docs, and question-answer systems. These systematically capture the informal but essential knowledge formal courses and papers do not cover, so that students do not reinvent the wheel for every project. Since explaining concepts and teaching helps learning, I involve students in adding explanations to refine wiki pages to help the next person, and capturing the knowledge they gain as new question-answer posts to internal forums. I also scaffold collaboration between computer science and behavioral science students, which drives forward research projects while allowing them to teach each other new skills.

I use technology to bridge my teaching and research program, to help students learn and to demonstrate how they can do the same. For example, I use interactive online quizzes to teach statistical and computer science concepts, and automatically record every lecture or talk I give to youtube. One current research project enhances feedback on online quizzes using crowdsourcing and machine learning for experimentation, while another augments passive video watching with reflective question prompts that elicit written and spoken explanations from students.

Teaching Interests

My multidisciplinary background prepares me to teach courses in the following areas.

Educational Applications. Introduction to Educational Psychology; Design of Intelligent Tutoring Systems; Principles of Software Design for Learning (example).

Human-Computer Interaction

– Introduction to HCI: User Interface Design, Prototyping, Evaluation (example course)

– Interaction Design & Design Thinking (example, example)

– Quantitative Methods for HCI and Information Systems

– Crowdsourcing and Human Computation

– Building Interactive Web Applications (example)

– Rapid Prototyping and End-User Programming for Complex Websites (example)

– Lean-Agile Design of Web-Based Products and Services (example, example)

– Fundamentals of Human Behavior for Information Systems (example course)

– User Modeling and Personalization

   Applications of Machine Learning and Statistics. These courses cover statistical methods and machine learning algorithms for data analysis. I will teach general introductory machine learning (example) and practical machine learning courses, although my preference is for statistical methods that focus on active intervention, automated data collection, dynamic optimization (rather than secondary data analysis) like Multi-Armed Bandits and Reinforcement Learning, Causal Graphical Models, Bayesian Optimization.

    Behavioral Science Methods for Design of Experiments, Data Collection, and Modeling. These courses could cover a broad range of statistics used in HCI, psychology, and education, covering both frequentist and Bayesian methods, parametric and nonparametric tests. From basics of hypothesis testing (t-tests, ANOVA, Regression) to more advanced topics like Multi Level or Structural Equation Modeling, Hierarchical Linear Models, and aspects of psychometric and educational assessment like Item Response Theory. These also include developing computational models of behavior, from neural networks to Bayesian inference and symbolic approaches.

Experimental Design and Causal Inference

– Design and Analysis of Laboratory and Online Field Experiments

– Statistical Methods and Scientific Computing for Dynamic, Personalized, Ethical Experiments

– Adaptive Clinical Trials in Health

Statistics and Machine Learning

– Foundations of Data Science (example, example)

– Computer Science for Statistics (example)    – Applied Bayesian Statistics (example)

– Practical/Applied Machine Learning         – Interpretable and Interactive Machine Learning (example)

– Machine Learning for Adaptive User Technologies (see NIPS workshop, tiny.cc/mlaihci)

Cognitive Science and Psychology. Cognitive Science Applications to Designing Technology; Knowledge Representation and Conceptual Change; The Role of Explanation in Learning, Inference and Causal Reasoning; Judgment and Decision Making; Social-Cognitive Psychology Approaches to Motivation; Computational Models of Cognition (Bayesian, Neural Network, and Symbolic models).