[quoting Grabrowski, 14 Oct 94] Is there a role for tutorial CBI that is engaging and transactive?
We've been developing the concept of the "instructional simulation." This may have some similarity to what Tim Spannaus mentioned. The basic idea is to first establish the context of a meaningful problem space for the learner, and then conduct three levels of dialog with the learner about the problem space: (1) simulation-specific feedback, (2) underlying skills/knowledge needs which (we hope) are perceived as linked to the problem space, and (3) problem-solving strategy. In this model, the tutorial still plays a major role, in support of the second and third types of dialog. The trick, I think (for now), is to make apparent to the learner the links between what the tutorial is teaching and the context of the problem space.
Is it worth the risk to develop it, if there is a chance that we might fail?
Our situation is somewhat unique, in that we're in a commercial environment but somewhat protected from the harsh requirements of the usual custom development world, such as Tim references. So, we have the liberty to try a few things.
There is a more general issue underlying this question, I think. This field is made up of practitioners who have in common the necessity of dealing with real-world learners all the time. Real-world learners never conform perfectly to any of our theoretical models, so we are always cutting and pasting prescriptions from various models in an effort to find something that will work in the current case. We're never fully successful, so we always fail to some degree. Our goal is simply to improve our batting average over what it was last time, or maybe over that of a purely intuitive practitioner. If you accept this view of what the field is (I'm intentionally avoiding the term "discipline" because it's not appropriate), then it's always worthwhile to try things, and it's always worthwhile to theorize about what we're doing and why it is or isn't working.
As a practitioner in a commercial environment, I always try new things, but on a small scale so as to limit risk. What stimulates me to do so is encountering a problem which isn't neatly solved by an existing model.
Is learner control as Rod was defining it at the lowest end of the transaction continuum, with electronic page turning on the negative side of tat same continuum--and unacceptable to be classified as a transaction?
"How low will they go?" huh? I think the issue of where the continuum ends at the low end, particularly, has to do with what the learner brings to the transaction, and how much covert responding goes on. One can, of course, read a book (the ultimate page turning experience) in a highly interactive way--or not. The point is that the lower one goes on the continuum, the less it's possible to make strong assumptions about the nature and degree of the interaction. If you believe that there is intentionality of the instructor in instruction, then one has to cut off the transaction continuum at whatever point the intentionality of the instructor has diminished too low for you. Below that point, learning will continue to take place, but it's no longer as a result of instruction. It could be as a result of information, or performance, or art, or whatever--but not instruction.
How do we find out if someone is mentally engaged covertly? Does it matter that we know?
Reference the discussion above. It matters only if we think we're instructing--i.e., if we, as instructors, have assumed some responsibility for the learning outcome.
If branching happens in between transactions, what does the computer do or how does it respond within a transaction?
The point here is that many kinds of computer responses are possible within a given context. Simple S-R-S+ models (e.g., present-question-response-feedback) need not limit our thinking.