Thanks for the kind words, Alan [Carr, 24 Feb 97]. I'm hoping to get someone really riled up, though.
With some of the "old hands" around who wrote Tutor or one of its successors (there's an interesting family tree, if you think about it), we've been close to the answer analysis issue for a long time. We still wind up writing our own proprietary answer analysis utilities.
For anyone who is interested, the Chemistry course Alan referenced is available through the TRO Web site: http://www.tro.com
The isolation of designers from learners certainly was never "part of the plan" when ISD was conceived, but it certainly seems to have turned out that way, hasn't it? As I said elsewhere, I think the reason has to do with the "waterfall" structure of the ISD methodology. At TRO, we're trying to get as close as we can to a "rapid prototyping" methodology (Tennyson's fourth generation ISD), but the tools aren't really there to do it the way we'd like. I agree that the lesson is clear: stay close to your customer (learner), or you will get what you deserve when judgment day (or assessment day? market day?) comes.
Amen, brother!