Talk:Example-rule coordination principle
The way this principle is stated:
"Instruction that combines or helps students' combine learning from examples and learning of or from rules tends to be more effective than instruction that includes the same examples and rules but does not help students combine them."
it is a stretch to consider experiments that compare examples+problems v. problems+problems as being relevant to this principle, in my opinion.
First, these two forms of instruction include the same examples and rules only if you consider the term "example" to mean both example and open problem, which is just not intuitive. Second, there is an assumption that examples+problems provide more support for connecting examples and rules than do problems+problems but that is only so after the fact - we know they do but that is based on post-test results, not on an intuitive notion of what supporting means. I mean, the notion of worked examples as somehow emphasizing rules (especially when the rules are not stated explicitly in the examples) or connections between rules and examples is not at all obvious. (Only after the fact, if you know the literature.)
So I would prefer to reserve this term only for studies that explicitly support the mapping between example/problem and rule/principle, such as the last Butcher & Aleven study in which students were prompted for diagrammatic self-explanations of the rule-step mapping. And maybe there are other examples within the PSLC. Self-explanations more generally should count.
By the way, the theoretical rationale is terrific. But what is it the theoretical rationale for? It seems to be the rationale for why instruction that emphasizes rules and examples is better than instruction that emphasizes one or the other. But many examples given on this page don't fit that mold.
One final comment: would really like to dig into the theoretical rational for why examples may be better than tutored problem solving, but I don't think this is the page for it.