Skip to content

Personal tools
You are here: Home » writing » (2003). Keeping the Baby and the Bath Water.

(2003). Keeping the Baby and the Bath Water.

Document Actions
Keeping the Baby and the Bath Water

A few months ago I identified a coming collision between the two seemingly orthogonal research camps working with learning objects. It has finally occurred to me that there is a very interesting middle ground between full-on automated instructional design (ala Merrill) and full-on self-organization. It’s actually not so much a middle ground as it is a marriage of the two, facilitated by a third wheel.

On the plane to a recent OLN Institute at Ohio State University, I read part of Lessig’s The Future of Ideas again. His description of, and argument for, the principle of end-to-end with dumb network and smart edge devices struck me differently than ever before in the context of our current work on the EduCommons system, which is a peer-to-peer system for developing and exchanging learning objects formatted for use with educational viewers based on watered-down versions of Merrill’s transaction theory. As this soup stewed in my head on the plane (amidst two bags of honey roasted peanuts) it occurred to me that the type of learning object I had described in my dissertation taxonomy as “generative instructional” really had an important property I passed quickly over in that discussion.

Andy Gibbons’ talking about models has helped me realize that generative instructional objects are synthetic or artificial models of a type. We know about building exploratory simulations or models that represent the real world, allow us to interact with the representation, and see the results of that interaction. These are natural models. However, we can design artificial models of content with which “interaction” really doesn’t have any meaning. For example, when one interacts with a set of facts or a list of steps in a procedure, nature doesn’t send any feedback to one of my sensory organs the way it does when I interact with a xylophone, pack of wolves, or buggy computer code. The important point here is that when we actually build natural or synthetic models of content, rather than just describing content with flat text, static images, or canned video, content can actually respond to user interactions or manipulations, alleviating some requirements of teacher feedback. I’m going to call this type of content (either natural or artificial models) responsive content.

While at OSU I repeated a familiar experience. When asking a number of very well educated people to describe the most valuable learning experiences that occurred during their school dayz. They said things like dorm or conference trip arguments with other students, eating lunch with faculty, unexpected discussions that arise while working on labs and other projects with student teams, networking, and a variety of other responses. The common thread throughout literally all the responses was that they involved non-content-related social aspects of the educational experience.

We observed and commented that it was interesting that most projects to “move a course online” actually end up moving course content online, with little or no thought for facilitating any kind of student interaction. Some courses do at least require regular posting to a discussion board. However, as students calculate the simultaneous maximization of grades and minimization of effort, assignments of the type “make x posts per y units of time” let students login, immediately post without reading or thinking about others’ statements, and logout: “1 post to the board per week” does nothing to facilitate interaction. And when faculty do think more carefully about interaction, they require replies as well as posts. To make sure that people are actually replying they will include peer-critiques of relevance to the chapter topic, etc. And while this looks like interaction, conversation limited to the scope of information in the textbook does not provide any of the value enumerated in the listed above. This lead us to affirm that online courses (where students have no other contact with each other) need dedicated social interaction outlets in addition to interactions structured according to discourse grammars focused on the course content (and was marvelously implemented as “knowledge types” within Fle3).

In my eternally abundant zeal, I spent several years researching and working on the automated instructional design paradigm, built on a foundation of learning objects. When I finally saw the relationship between learning objects and Napster (I will always be grateful to Wayne Hodgins who challenged me out of the blue one day to figure out what that relationship was), I completely rejected the automated paradigm and began trying to build a model of learning object use based on Slashdot’s model of massively scalable, massively social informal learning. One NSF CAREER grant later, I’ve finally gotten to the vantage where I can see that the baby needs the bath water.

Perhaps “the best of all possible worlds” for scalable, sociable, an sustainable online learning involves using responsive content as a seed crystal around which to grow student interaction structured by discourse grammars and quality-assured through peer-review. This is a more moderate position that simultaneously embraces Merrill’s behaviorism and congintivism, the benefits of unmoderated (but scaffolded) discourse, and the social constructionism and peer review of online self-organizing social systems. Sounds tolerant, and almost epistemically naughty. I think I like it.
Created by admin
Contributors :
Last modified 2004-06-14 10:07 PM