Note: I’ve pivoted my research focus since this was written in January 2021; stay tuned for an update!
My research experience so far has focused on the educational sector. On the student side, I have worked with the Virtual Learning Lab project for the last two years. In this project, I have developed machine learning models to predict student affect while they interact with an online mathematics learning platform. Besides aiming to improve accuracy, I have investigated how these models generalize between different groups of students. My other work with student affect is through the EMERGE project led by Dr. June Gruber. In this study, I helped develop and implement an experience sampling method to measure first-year students’ affect and its context over several weeks. This has given me experience with the practical side of designing an experiment, managing data collection, and later analysis.
In addition to affect, I have also begun to study the context through which students are using the platform in order to understand their motivations; that is, are students logging in because they want to improve their math ability, or does their grade depend on them completing a certain assignment? Finally, I am currently mentoring a team which is working to prospectively predict student performance on quizzes using their activity patterns on the platform. This work is drawing on theoretical modeling of Item Response Theory as well as time-series analysis and modeling through machine learning. Through this work, I have gained experience managing large and messy datasets. This has primarily involved working with interaction log and student self-report data modalities.
My work in education has also focused on improving teacher outcomes. Over the last year, I have worked with the Cyber-Enabled Teacher Discourse project to develop an automated approach to teacher feedback and professional development. In this project, I have developed machine learning models to identify key elements of teacher speech that are associated with greater student engagement and learning. I am currently integrating these models into an automated framework where teachers can record audio of their teaching and receive feedback in the form of a smartphone application. This project has given me experience working with audio processing as well as natural language processing techniques.
Currently, my research interests more broadly consider how to integrate effective behavioral interventions into technology that users are already using. In my work with teacher feedback described above, I am helping design a user study to investigate how teachers interact with the smartphone feedback application. Does their behavior in the classroom change as a result of the feedback they receive? Do they feel that the given feedback is accurate when compared with their memory of the lesson?
In my work with the Virtual Learning Lab, our measure of student affect is currently integrated into a reinforcement learning system which is used to give students personalized recommendations following a quiz. After the study, I would like to investigate under what conditions students chose to follow the recommendation and whether there is a difference in learning outcomes. In addition, my goal is to use our models of future quiz performance to give interventions regarding the best time to take a quiz. If a student is not prepared to take a quiz, a low score may negatively impact their motivation and future effort. On the other hand, students who are prepared for the quiz can gain confidence after a high score and also gain valuable retrieval practice, which is important for long-term retention.
Finally, I am joining a project studying collaborative problem solving in small teams. In this project, I will be designing and implementing an intervention study to improve key behaviors of collaborative problem solving as teams work to solve problems in an educational physics game. These interventions must not only orient participants to a somewhat complex set of key behaviors, but also provide enough scaffolding for participants to recognize and improve their own behaviors.