If you are programming the parser yourself, you can even set up "permissive" rules. Using the above example, say 4 of the responses are the ones sought after, but there are 2 others that might be viewed as correct in certain circumstances. You could then allow all of the following to proceed to the card for "correct answer": correct 4, (correct 4 + one other), (correct 4 + the other possibly correct), (correct 4 + both others). The answer card then tells the student what the sought after answer was and why, then explains the "permissive" answers.
The other thing which can be done is to simulate real-world tasks. I've put together a module that simulates the way a bench tech goes about interpreting an antibody identification panel. The programming for this was actually quite easy. What was difficult was figuring out the graphical display--once I created the graphics for a hardcopy book I wrote, the programming solution made itself apparent.
For those who are curious, I'll describe what I did (ignore if you like). If you envision a grid consisting of 10 rows and 25 columns, the first step the student needs to do is to highlight the correct rows (on-off toggle using unhighlighted graphic vs highlighted one (both created in Photoshop). Then the student needs to toggle one of the following 3 (no image, slash, or X) at the top of each column. Eventually, the parser needs to see whether the correct rows are highlighted and whether the correct mark appears at the top of each column. Again, there are "permissive" rules covering circumstances in which both a slash or an X could be correct.
M
On Aug 11, 2004, at 12:53 PM, Richard Gaskin wrote:
What sorts of enhanced question models do you think would be ideal for computer-based learning?
_______________________________________________ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
