Contrasting Cases - Guiding Principles

I was reading an article on learning theory:

Rethinking Transfer: A Simple Proposal With Multiple Implications, Review of Research in Education, Ch. 3, Vol.24 pp. 61-100. Authors: John Bransford and Daniel Schwartz

and was intrigued by a particular section starting on page 11 on the role of contrasting cases in perception and learning

"Research on perceptual learning provides a good illustration of what it means to “know with” our experiences (e.g., Garner, 1974; Gibson & Gibson, 1955). Perceptual learning theorists point toward the importance of contrasting cases, like glasses of wine side by side, as guides to noticing and differentiation. One is unlikely to be able to remember each of the contrasting cases, and experience with a set of cases will not necessarily let one induce principles that guide unaided problem solving. Nevertheless, experiences with contrasting cases can affect what one notices about subsequent events and how one interprets them, and this in turn can affect the formulation of new hypothesis and learning goals. " ibid p. 11

This resonated with me in two ways

• It supported one of the principles that I was using as a design criterion in my approach, i.e. whenever possible use direct comparisons between two systems (as opposed to sequential quantitative measurements) to guide students attention to the relevant distinctions.

• That through the process of venturing away from the established principles of modeling instruction, one may perhaps come to understand those principles more deeply. It is through the contrast that we often come to see what previously we could not fully perceive.

So as I try to define the principles and design criteria of my own approach through this forum, I am also hoping to clarify for myself the principles of modeling instruction and understand the pedagogical distinctions that I am playing with. One of the great strengths of the modeling curriculum is the way that it "chunks" the conceptual content into models. Each chuck is carefully and fully developed before moving onto the next. I find that I am messing with that successful formula and that chunkiness, therefore perhaps wandering into deep water.

Another aspect of the success of the modeling curriculum is the affective dimensions and the sense of confidence that develops as the students develop a model for the scientific process that they can use over and over. 

At the moment and I am writing this at the end of December (and editing it a reposting it in January since it got burried somewhere), and I am at a very different place conceptually with my students than I have ever been before and it is not exactly comfortable, we have accomplished some big things but in a crazy order. The tradeoffs are hard to fully fathom.

For those of you who are modelers or just thoughtful educators, I value your opinion about what makes a curriculum approach pedagogically coherent and successful. What are the principles that are sancrosanct, which principles can we play around with? Is it possible to design a curriculum in such a way that it works continuously on the big picture although of necessity it comes into focus gradually like a google earth image or a polaroid image, adding clarity and detail gradually as it cycles through each application? Or would such an approach likely overwhelm our students and their tolerance for ambiguity and incomplete knowledge? These are the kinds of questions that I am wrestling with and would value you contributions to this forum discussion.