I can't imagine there are too many teachers who would welcome the opportunity to be corrected by an artificial intelligence. Yet we move one step closer with "a computer model that came close to an expert educator's ability to discern when a teacher was asking students an 'authentic' question." The model doesn't tell you what makes a question authentic - that would take the human evaluators who trained the model. But it does classify them in roughtly the same categories as human observers. Jill Barshay writes, "Hopefully, decision makers will use automated observer robots to improve instruction and not as a weapon against teachers, repeating the mistakes made using student test scores to pay and punish teachers." Sure. That would never happen. Right?
Today: 5 Total: 5 [Direct link] [Share]
