Friday, July 25, 2014
I have a tremendous amount of respect for Woodie Flowers and all he has done for FIRST Robotics. He has a unique and interesting perspective on education , which I think deserves a view here. Some of what he says though isn't consistent. He says we should have automated and adaptive software to teach subjects - fair enough, but he also says we shouldn't teach to a test or evaluate students based on multiple choice. Most of the adaptive software I've seen needs an assessment to determine if the learner has learned before they move on to the next subject. What does he propose to replace that? Not much really.
Similarly he indicates we should use big data to evaluate teacher performance and teaching best practices and apply them across the board. Again though - to use big data to make these sorts of projections and assumptions the data must be gathered in the first place - benchmark data, progress data, effectivity data and demographic data. All things that require some sort of testing and evaluation to determine how well people are learning, and what is actually working.
I would love to see his proposed non-test way of gathering the data required to make these kinds of analysis - truly - because if he can do that he's invented something VERY valuable.