[quoting Gunn, 22 Oct 96] That learning is a complex phenomenon I do not doubt for an instant. That parts of that complex equation can be isolated and measured I am not so sure--the ideas of complex interactions and inter-dependencies would worry me. That any one part of the complex equation could be isolated, measured in a contrived, experimental situation and still produce generalizable results I would have to doubt.
Your discrediting of experimental research in education is bothersome to me. The use of "contrived" experiments is helpful to understand any complex phenomena because of the nature of the complexity. Granted you lose some of the big picture when you only look at one variable, and the purpose of a single study is to gain evidence that can be tested and challenged through further investigation, not to provide the end-all answer.
The alternative is to look at the entire learning picture at once. But you cannot look at the entire picture at once and produce generalizable results that tell you anything meaningful about learning. The purpose of isolation of variables is to allow for generalization, and if you isolate nothing, you cannot generalize.
I am one of the camp that believes learning is quite importantly dependent on motivation, perceived relevance and a number of individual factors.
I too believe that there are numerous factors that influence learning. I also believe that the ways that people learn (whether from instruction or otherwise) don't change at a fundamental level. And these fundamentals of learning can be discovered (e.g., Steve's example of the log-log law of practice). In my view of things, Steve and others who conduct research are trying to discover these fundamentals.