24 Mar 2015

I hadn't thought about collecting my own version of Teaching Machines - another slip of the fedwiki mindset. Mike recommends doing so here, with details.

An FYI: from an email from Mike, 24 Mar 2015

THE ASSIGNMENT First, Audrey has her sentences up and documents for idea mining/summary. You can find them here . So, as before, use those as starting points to build out your site with new pages. Define terms, write up short histories, detail broad themes through idea-mining. Second, go out and look at what others have written. If you've been stingy about forking, it's time to get unstingy. Treat forking at this point like social bookmarking. If you see a page that you think you'd want to find again -- one where if you couldn't find it you would be frustrated, fork it. If you see a page that you'd like to share with others, fork it. Third, use conversation clubs to find pages with more than one version. For each page, review the page versions in the conversation clubs viewer and either: Fork the best version to your site Make a better version hosted on your site Do not fork any of them if you believe they are not helpful to your understanding of learning machines, or if you feel the quality or focus is not suitable for your site. A more moderate version of this could be to review a random 20 articles with two versions, or to only look at pages with three versions, etc. The point is to get some eyes on multiple versions so we can see whther comminity consensus on the most useful documents emerges (or doesn't, which is fine too). WHY COMPLETING THIS IS IMPORTANT This experience has been amazing for me, but also a bit frustrating at times. On the amazing side, I am continually stunned by what I learn in these events, and I know many of you feel the same. There's a deepness and a richness to what you walk away with that transcends ordinary online experience. And the people we have are fascinating people. We truly get some of the best thinkers out there to participate in these things. On the frustrating side, I feel like people came into this experience with 18 different ideas of what they wanted it to be, and many of those ideas just weren't compatible. What we really wanted to do here by setting this up (and what I hope I communicated to participants that we were doing) is test the use of this in a traditional educational scenario, where participants would learn about stuff that was sometimes difficult and challenging, but with the wiki-driven option of coming at that material in ways that made sense to them and allowed dense interconnection with other material. And a number of people joined this event (and put in a lot of effort) for the same reasons. And I feel like the contract we made with those people is we would see this experiment out. So on to why completing the course is important: one crucial test of whether this works is whether the structure we have adopted here will float better versions of things to the top. Sarah even gave us a theoretical framework for this in one of her pages -- the process we are looking at here is something like "Adaptive Comparative Judgement". If through individual acts of judgment good and interesting work floats to the top, that can form a starting point for a professor trying to make sense of a class's work. That's not to say it would be the be-all, end-all or sole basis for evaluating the class. But it's a significantly better starting point than looking at 50 separate wiki sites. The resulting intersections would show what the class saw as an accurate or useful map of the domain. That in turn could help the professor assess the success or failure of the class (or opportunities for further instruction). So this is the final hypothesis. You obviously don't have to do anything you don't want to do. But I'm hoping you'll help us test this by completing the final assignment.