Steve, I think the basic point you made in your paper was a good one. I found your experiment interesting. I agree that you certainly demonstrated in one case media make a difference (and I personally think that media is often, if not always, one important component in the learning process). However, I don't think any of your evidence DISPROVED the contention that media shouldn't be the focus of our attention. You demonstrated that media can be important, not that learning processes aren't, or even that media is the thing we should pay the most attention to (my opinion, of course).
I want to thank Ian Hart for clarifying what I believe is THE important issue--why I and others get confused when we hear a statement like "media doesn't matter" as a summary of the famous (and infamous) delivery truck analogy that I encountered this summer.
You raise an issue that I'm quite familiar with from my background in psychology, educational psychology in particular. If I understand you correctly, the question that Clark raised (and many in the education research community, I might add) is whether "scientific objectivism" is a valid way of knowing about the effects of media on learning or, more specifically, that the following is a useful way of improving learning (I think, everyone's ultimate goal):
[quoting Hart, 22 Oct 96] ...(i.e., the idea that one can hold conditions constant, change one variable to observe the result, and the (totally unsupportable) view that the "results" of such an enterprise can be generalized)...
I believe it is. I'm not sure what you mean by totally unsupportable with respect to generalization. The educational literature is full of instructional techniques that have proven effective across a wide range of situations (there are always limitations, but some things certainly generalize to many situations, e.g., cooperative learning--see the Johnsons and Slavin). Even more specifically, I believe that it is reasonable to assume that under generally similar circumstances, in a "real world" classroom, most students will learn better via the written word when teaching information that is generally similar to Steve's.
With all due respect (and I mean this sincerely, recognizing that you have given this a lot of thought, and you state your case articulately) the following statement is most certainly NOT true.
Even in "straight science" this view is no longer generally accepted--viz. the theory of quantum mechanics.
I teach a class in biological psychology, and, believe me, most of the things we know about the brain and nervous system are the direct result of quantitative methodology based on "scientific objectivism." This also holds true for biological and life sciences in general. I've done statistical consulting work with faculty in the biology department and they rely heavily on quantitative methodologies and the scientific method. Most traditional medical treatments and drug therapies are based firmly on quantitative experiments. It's certainly true that weird things happen at the level of the atom. But, from my conversations with physicists, most are firm believers in the scientific method and scientific objectivism--despite Heisenberg and uncertainty. The reason I know this is that, in discussions with me, they often suggest that I am kidding myself that I can apply these methods to behavior.
Which brings us to, what I believe, is your most important point (which is exactly theirs) that we can not apply this method to behavior--in particular the effect of media on learning.
I believe that the most important contribution made by Clark is to establish that quantitative methods of research will never establish if/how media influence learning.
My personal belief is that the introduction of QUALITATIVE research into educational research is great. But, I DO NOT see it as replacing QUANTITATIVE methods. I think we can learn a lot from studies in which variables are manipulated and quantitative methods are applied to criterion variables. I think we can create an even more powerful experiment/study by also utilizing qualitative methods like interviews, protocol. There is a continuum of methodologies from the most artificial and controlled to the more realistic and uncontrolled. The move toward the more realistic is good (in my view) but not fundamentally (qualitatively, if you will) different from the former.
(2) Previous discussions had given me the impression that constructivism was the flavor of the month.
I would suggest that the leading educational psychologists in the U.S. (that I encounter in the American Psychological Association Division 15, and American Education Research Association) are not all head over heals with constructivism. There are many interpretations (again, along a continuum). And, it is certainly true that there is certainly a move in the direction of student centered learning (which I think is good) but not to the exclusion of the instructor. For those who are interested in the constructivist issue, I would recommend the following: Phillips, D. (1995), The good, the bad, and the ugly: The many faces of constructivism, Educational Researcher, October 1995, and an interesting interchange between Phillips and Ernst von Glaserfeld in the Educational Researcher in the 1996 August/September (no 6) issue, including Phillips' quote: "Speaking personally, however, I have the superstition (unsupported by philosophy) that I would rather go under the knife at the hands of a doctor who thought that I 'really existed' rather than one who regarded me as a 'nonobjectively real construction' of his or her own mind. In practice, of course, both would treat me exactly the same, and they could even have the same bodies of knowledge..."
Different learners will enjoy and profit by different means of accessing, constructing and using information. Choice is useful. Media (as any teacher of art, creative writing or Logo will tell you) are great vehicles for student to learn with.
Embrace complexity. (Was it Tom Reeves or Bob Dylan who said that?)
To suggest that individual differences and other variables are important in learning, and that they often interact with medium is not to suggest that they can not be studied quantitatively/objectively/systematically. Sophisticated multivariate techniques and multi-level/factor experimental designs have evolved specifically so the experimenter can "embrace complexity."
Ian, again, I thank you for re-introducing us to this important issue. I'm sorry to go on so, but I'm not about to sit back while some suggest that "scientific objectivism" has no place in research on education (because I think it clearly does--although, there are many very intelligent people who disagree). For any one who is still reading, I offer a final question that I often offer to those who suggest that quantitative and scientific methods are not useful in educational research "If I have a goal of trying to improve the amount that students learn, how do I go about assessing whether I have done this successfully?" Even more specifically, "If I have a classroom full of students that need to learn foreign language vocabulary, and I am the only instructor and I only have the facilities to present the information orally or in a written form, what's the best way to decide which to use?" Surely I don't want to flip a coin. And, how would this method be better than Steve's experiment?