UROP Openings

Have a UROP opening you would like to submit?

Please fill out the form.

Submit your UROP opening

(Remote) Analyzing text/audio communication for emotion and conversational strategies in practice spaces (digital clinical simu


Term:

Fall

Department:

CMS: Comparative Media Studies

Faculty Supervisor:

Justin Reich, EdD

Faculty email:

jreich@mit.edu

Apply by:

9/11

Contact:

Garron Hillaire: garron@mit.edu

Project Description

Our multidisciplinary laboratory - the MIT Teaching Systems Lab (TSL) - is comprised of engineers, learning designers, learning scientists, and social science researchers. We are looking for students with an interest in education, learning science, and quantitative and AI research. The project focuses on how participants communicate during difficult conversations, use of conversational strategies, and how to support reflection by novice teachers, using AI and Natural Language Processing (NLP) and Deep Learning with Audio data (all within TSL’s Teacher Moments platform - teachermoments.mit.edu). Students will work closely with TSL researchers to collect dispositional information and psychological measures about participants to examine convergent validity of text/audio measures with validated measures. Possible student tasks include: Developing text/audio classifiers that measure emotional expression and/or conversational strategies Analyzing how text analytics relate to validated measures Supporting implementations of Teacher Moments Participating in product development for Teacher Moments Supporting TSL researchers in drafting analysis plans, writing project documents, and other writing as needed. Students will work closely with TSL researchers familiar with the detailed goals of the project and will gain hands-on experience in qualitative and mixed-methods research. Sponsored research funding is available.

Pre-requisites

Some previous experience using natural language processing (NLP) and or Audio Analysis tools (e.g., NLTK, Python). Some previous experience training text classifiers (e.g., scikit-learn in Python). Some previous experience doing regression analysis (e.g., logistic regressions, linear regression).