#Expert-facilitated Crowdsourcing, #Teachersourcing
#Teacher customization, #Social Computing
Duration: Jan 2019 - Oct 2019
Mentors: Dr. Ken Holstein, Prof. Vincent Aleven and Prof. Joseph Jay Williams
Motivation: AI-based tutoring systems are increasingly common in K-12 classrooms and are proved to be effective for increasing students’ outcome. However, as teachers using the intelligent tutoring systems in classroom sometimes expressed not perfectly satisfied some instructional materials in the systems, as not helpful enough for assisting students’ learning. Teachers expressed desire to customize the materials, yet are short of the time to do so, thus, we seek to improve teachers’ satisfaction without imposing too much on their time, and explored a combination of teachersourcing and crowdsourcing for the instructional materials improvement.
Two studies were conducted to test the respective effectiveness of the teacher-crowd pipeline. Follow-up data analysis inform some design recommendations for future teacher-crowd pipeline for materials improvement.
Paper available upon request.