[quoting Alessi in different thread, 16 Oct 95] Tom [Reeves] takes the position that eventually one end of the continuum will be proven the "better" one, namely, the end where students do the creating. A very constructivist point of view, of course. While we would probably all agree that research will determine who is right, I rather suspect that there is no "better" end of the continuum.
This, in conjunctions with Steve's posting on the article, reminds me of a point which seems to get lost in the constructivism debate: the common distinction (Anderson, et. al) between declarative and procedural knowledge. They are complementary. It seems to me that simulations are well suited for helping learners make the transition (recompilation, in Anderson's terms) from the declarative to the procedural. Specifically, activities such as predicting the behavior of a simulation are very useful for synthesizing concepts and principles into "working" mental models. Playing (reflectively) with the simulation is useful for developing procedural knowledge based on the mental model.
But the other side of the coin is that such "constructivist" activities tend to be very inefficient in initial acquisition of declarative knowledge. That's where all the rest that we know about instruction applies very nicely.
It seems to me that the more "purist" advocates of constructivism have only weak methods for declarative knowledge, while they criticize infidels for weakness in dealing with procedural knowledge. Don't we need both?
Which brings us to SD. Wouldn't the same criticism apply to an SD model: it's great for showing (and learning) the (procedural) relationships between system components, very inefficient for learning (declaratively) what the components are, in the first place.
I'd also like to focus attention on Steve's comment about how hard it is for learners to do modeling. I agree, and I think it's because we tend to do a poor job of helping the learner learn the declarative knowledge and "recompile" it into mental models and procedural knowledge. Our analyses and designs--even those done by the cognitive and constructivist types--rarely deal rigorously with how the learner is going to do this. In retrospect, I think it is precisely this issue which defeated the "inquiry" curriculum movement of the 1960's. Unless we can figure out a way to do it now, I fear that the current constructivist movement will meet a similar fate.