[quoting Dorricott, 17 Nov 95] I thought Dr. van Staden's comments about the merits of both C and I, along with his examples, we're right on. I, too, am skeptical when educators (in their enthusiasm and passion for constructivism) suggest that we cast aside anything having a hint of guidance, pre-specified objectives, or repeated practice.
Perhaps constructivism helps us answer many of the questions about motivation, relevance, retention, and transfer (i.e., developing meaningful cognitive connections) that we have long struggled with. However, "instruction" or "instructivist strategies" help ensure we that the places we get to are useful, and it helps us learn from the expertise of those who have already asked and answered the same questions.
While I have neither the references nor the personal experience in education to substantiate some of my general concerns with the widespread use of constructivist approaches in K-12 (or K-16) education, I have both references and personal experience when it comes to training adults. If you're an educator, you may choose to ignore the following, it may not be relevant--if you're a trainer, it's likely quite relevant--if you are just generally interested in learning at all stages and ages, take it for what it's worth, or what it costs you.
Based on adult learning literature (Malcolm Knowles, etc.), adults are motivated most by direct, obvious relevance to a perceived knowledge/skill deficiency that they feel a NEED to overcome. They are intrinsically motivated, very impatient, very self-conscious, and have little time to waste. These motivators have led, in large part, to the growing use of the electronic performance support system (EPSS) paradigm that suggests making whatever information/training that is needed available directly at the point of performance. (Another current training trend or buzzword is Just in Time training, or JIT, that would use perhaps more traditional instructional methods to perform a similar task--wait until just before needed (increasing transfer) and the need is obvious (increasing motivation) to perform the training.) While hypertext or comprehensive indexes MAY be used to support an EPSS, the goal is to use the "intelligence" of the design or that built into the system to navigate and access the needed information and/or training in the least possible time--e.g., minimize the exploration needed to find and fulfill the NEED. Then, present the information or training in the most concise way possible, avoiding misinterpretation, and "wasting" no more time than necessary. Further, most all research that I've seen indicates that expertise in any knowledge domain REQUIRES an extensive and firm foundation in the facts, concepts, and rules of that specific knowledge domain. The more recent research in AI/intelligent tutors (Lesgold, Gott, etc.) suggests that expertise is quite often limited to a "subset" of a general domain where the broadest, most robust foundation exists--i.e., "experts" in one part of a knowledge system may be only "average performers" in another part of the same system.
In my humble opinion, I'd suggest that we should use constructivist approaches only where best suited in training (cognitive skills training such as problem-solving, troubleshooting, decision-making, etc.) but not use the approach to build these robust underlying foundations of facts, concepts, and rules. My rationale:
(a) We can't afford the time required for the learners to "explore" and formulate their own conclusions (either the "learning time or time-on-task") or the time required to design and develop these rich, complex learning environments. Given the fact that even a fairly limited free-pay simulation may cost me three times the design/development effort as a well constructed tutorial, how much content can I afford to cover using these complex interactions? How many constructivist "lesson plans" will really be designed, built and implemented soon, given all the constraints on time. One of the primary reasons that many of today's tutorials are poor is that the designer/developer simply ran out of time--time to pilot test, enrich, and polish the interactions (especially in the area of feedback to a learner's action).
(b) We can't ensure that the learners draw correct conclusions without extensive testing and "reteaching" being required. And, if there's any validity to Gagne's hierarchy of knowledge (which there appears to be), we can't afford many errors in this foundation knowledge. Further, it is much more difficult to change and correct a misconception than to provide the correct interpretation the "first time," regardless of the theory/method/technique used.
(c) We are disregarding the value of any existing "expertise" about what may be the best way to learn and perform these specific knowledges/skills. We're forsaking any value of a mentor or tutor--and substituting "micro-worlds" where the learner "learns the hard way" by the "school of hard knocks," only it's a simulated "hard knock" instead of real-word results of any mistakes.
One final thought: If we don't think our own learning and experience is valuable and could be helpful to others, why is this dialogue even taking place? Though we may be attracted by the "humanism" of a constructivist approach, aren't we also trying to "instruct" each other in some way? Is no one trying to be persuasive? Perhaps the most we can hope to do efficiently and effectively, at least near term, is to try to make "drill and thrill" lessons, by applying the thin veneer of a gaming strategy to the underlying material. And, continue to use the "classic tutorial" where appropriate (likely the majority of the content to be taught), but work diligently at improving the quality of the interactions in those tutorials. Don't forget the simulations either, but acknowledge how difficult/expensive they are!