Terry Freedman responds to my suggestion that the future of assessment is in "systems that can 'taught to analyse and assess responses to an open-ended question'." He raises three issues: first, "the program itself doesn't know how it reached the conclusion"; second, the phenomenon of "automation bias", where we are inclined to believe the AI system without question; and third, the concern that the teacher has "no opportunity to see what misconceptions the student has developed."
Today: 4 Total: 110 [Direct link] [Share]
