These are my annotated notes from a rapid prototyping workshop EdTechTeam’s Chris Bell ran at the Google Apps for Education summit at Mulgrave School in Vancouver, May 3 & 4.
See also Students drop a bomb at #gafesummit Vancouver: “We don’t like rubrics.”
Our team was asked to rethink assessment and this is what we came up with in the hour:
- Assessment has two functions: it provides feedback to the learner and it provides institutions a way to show accountability for time and money spent (a reflection of the person-individual distinction I wrote in an earlier post.) Summative assessments serve the institution, formative assessments serve the student. Is there a way for schools to mine formative work and so eliminate summative assessments but still have some way for a school to show taxpayers its doing its job? Can we build a tool that let’s good work bubble up and the junk to sink?
- In either case, our sense was that assessment is inherently artificial, a proxy for an actual event or process. So, we asked how could we make assessment float up naturally out of everyday activities? If Facebook, Amazon, Apple and so on can create very sophisticated user profiles from just clicks on a page, could we do the something analagous in school? What would be the equivalent of a thumbs up or thumbs down action in school?
- There is a debilitating amount of friction in the data gathering process in schools. Students have to sit formal exams, and that is logistically complex, and teachers have to manually enter marks, which is tedious. Both contain inherent time delays. In contrast, these tech companies get enormous piles of useful data in near real time because they make the data collection dead simple: click thumbs up or thumbs down, click a few links, or post a link, or upload a photo. Each of these simple actions tells the tech company something about you. Our contention is that tens of thousands of transactions would give schools much more sophisticated understanding of their students and students much more sophisticated understanding of themselves than our existing exams system. How do we eliminate that friction?
- Current assessment models start with an externally supplied standard–IB. AP, state curricula, etc.–and schools strive to get their students to reach those standards. That creates a sausage-making model of education, says Yong Zhao (see sketchnote after the jump.) Can we build a model that starts with the person, the student, and not the standard? (See Benjamin Zander’s thoughts on this at about 5:405 in the video after the jump.)
- If you can see all of a student’s work and all of a teacher’s teaching, you don’t need a proxy for it. Can radical transparency, perhaps drive by social media, eliminate assessment altogether?