Waterfall-like design spaces are inadequate for AIEd because adaptable systems have untestable states¶

Claim · All claims
Evidence · 1 study · theoretical r3 · q1
Currency · newest study 2022 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? The article argues that waterfall-like design perspectives may work for constrained systems, but AIEd systems have numerous states that cannot in principle be tested in constrained design periods. → Bhimdiwala 2022

Evidence¶

Bhimdiwala 2022¶

Bhimdiwala, A., Neri, R. C., & Gomez, L. M. (2022). Advancing the Design and Implementation of Artificial Intelligence in Education through Continuous Improvement. International Journal of Artificial Intelligence in Education. https://doi.org/10.1007/s40593-021-00278-8

q1 · i? · theoretical · r3

A theoretical argument in the Pursuing FATE through Improvement Science section. The authors draw three implications: designers cannot discern all user needs early; design tasks become longer-term and shared; and each school shapes use and utility of an AIEd application.

"Adaptable systems like AIEd systems have numerous states that may not be, in principle, testable and study-able in constrained design periods."

Discussion¶

Related Claims¶

• FATE issues in AIEd present themselves when users gain ownership and adapt systems to their local contexts — related
• A student model that is easily ported to another ITS may still fail there, because student behaviour depends on domain, interface, and the system that generated training data, so ported models must be re-tested. — related
• Inquiry environments are hard to scale because they are ambitious learning activity systems requiring changes in roles, participation, and system coherence — related
• User-centered and participatory AIEd approaches engage users mainly in early design, limiting recognition of FATE issues arising in use — related
• User experience challenges include poor matching, inadequate scaffolding, and suboptimal interface design — related