CURI AI Tutor Voice Agent

Exploring how a voice-based adaptive tutoring agent can provide one-to-one instruction
Role:
UX Research
Team:
Gen Marconette & Grace Kim (Instructional design), Roza Atarod (UX design)
Duration:
3 months (Jan - May 2021)
Tools:
Figma and Voiceflow (for prototyping), Zoom (to conduct usability study)

Brief

This research aimed to explore how a voice-based adaptive tutoring system may close the equity gap for access to human tutors by providing one-to-one instruction. This was a class project from the Human-AI Interaction course in the School of Information at UT Austin taught by Dr. Min Lee.

Problem

Personalized 1-on-1 learning is inaccessible. A lot of accessible, cost-friendly learning platforms right now are 1-to-many, such as videos on Khan Academy. Meanwhile, human tutors that provide 1-on-1 guidance can be very expensive and geographically inaccessible. To address the weaknesses of both methods of learning, we looked to AI as a way that can provide 1-on-1 help anywhere at anytime.

Goal

Improve a learner's motivation, self-efficacy, and metacognition of learning through an AI tutor

Report

Citation

Ko, E., Marconette, G., Atarod, R., Nguyen, K., & Lee, M. K., (2022). AI Voice Tutor Usability Study: Understanding Impact on Math Self-Efficacy and Metacognition Among Middle Schoolers. In de Chinn, C., Tan, E., Chan, C., & Kali, Y. (Eds.). Proceedings of the 16th International Conference of the Learning Sciences - ICLS 2022 (pp). International Society of the Learning Sciences.
https://www.dropbox.com/s/ws5sdcfi72aykj1/ICLS2022%20Proceedings.pdf?dl=0
CURI AI Tutor Voice Agent