In the last day or so, a number of people have contributed to the beginning of this discussion. To be honest, I am a little overwhelmed by the response and the great exchange of ideas based on Barney's original inquiry. I would like to further the discussion a little, or perhaps stir up is a better phrase.
Peter [Arashiro, 25 Jul 96] suggested that Merrill's work on Component Display Theory and later Design and ID2 might apply. I agree with that notion, though certainly not by itself.
I don't think it goes far enough. At least my understanding of the theory is that it deals largely with cognitive aspects of learning and cognitive tasks. That is great as far as it goes. But over the last several years, especially after leaving the protection of a University position and venturing into corporate consulting, I have come to the conclusion that our models (even for cognitive learning) fail to deal with a couple of critical aspects of learning that are affective and social in nature.
As Rod [Sims, 25 Jul 96]indicates, one of the major roles of education is transfer and generalization. In order to do that, learners must encounter a variety of ideas and possibilities and through interaction with the content and with other people learn to deal with the complexities of real world situations.
This has lead me to an alternative way of examining interaction. As I see it, there are five kinds of interactions that may be required for effective learning (not necessarily all five for every objective). The interactions are delivery technology independent, though some technologies are probably better for supporting each kind than others. These types (in no hierarchical order) are:
learner-interface (how to manipulate the learning environment),
learner-content (basic knowledge and skills),
learner-learner (enhances understanding, aids generalization, aids transfer, provides emotional support in the learning process),
learner-instructor (provides performance feedback, can provide many of the things the learner-learner interactivity provides, and can provide mentoring), and
learner-outside expert (aids generalization and transfer, may provide mentoring).
If we examine our desired learning outcomes, they require differing amounts of each of these kinds of interaction. Clearly in a Computer Assisted Learning environment, interface is one issue--so is content. But, I suspect the issue of learner-learner and learner-instructor should also be factored in when designing such instruction.
Given this approach to interactivity, I still see value in Barney's idea, as long as the goal is to determine how and under what conditions each of these types of interactivity can be supported to maximize learner interaction with the content and ac
hievement of the objectives.