Detecting Student Perceptions
Research Assistant @ Delta Lab
Fall 2020 - Present
Algorithm Design
Software Development
Testing & Analysis
User Interviews
Sprint Planning
Read our paper published to AIED2021 here.
How can we detect student perceptions of the programming experience from log data?
As a research assistant at Delta Lab my project is focused on detecting moments where introductory CS students may negatively self-assess during the programming process using interaction log data.
Mentored by Eleanor O'Rourke and Jamie Gorson, I assisted with designing and developing a detection algorithm, interviewing users to collect data, and analyzing the performance of the algorithm. I also assisted with writing a paper that was accepted into AIED2021.
A representation of manually labelling moments where students may negatively self-access from interview recordings. These moments include when a student stops to think or uses resources for syntax. Our algorithm modelled this manual human detection using interaction log data, seen in the figure below.