Epistemic practices are the cognitive and discursive activities that we want students to engage in to develop their epistemic understanding. We suggest that the following practices, while not the only valuable epistemic practices we might want students to learn, are key practices for developing the epistemological understanding we have described.
Explicit articulation and evaluation of one's knowledge
Coordinate theory and evidence
Make sense of patterns of data
Develop representational fluency
Hold claims accountable to evidence and criteria
Given that we want students to see knowledge as an object of inquiry and to understand different forms of scientific knowledge, then their science learning should be centered around creating and evaluating knowledge. The argument for knowledge articulation and evaluation is common in current theories of learning. We want to emphasize that this practice is useful for more than deepening students' understanding of science concepts, but is crucial to students' development of sophisticated epistemological conceptions about science.
The central aim of science is to construct theories that explain natural phenomena. This effort requires the coordination of theoretical ideas with the data that provide evidence of their utility as explanations. A key aspect to the practice of coordinating theory and evidence is to be able to distinguish claims from evidence. The practice of coordinating theory and evidence entails using theories to explain data, and using data to evaluate theories. This could be done a number of ways, as demonstrated by the approaches described in the following sections of this paper.
An important aspect of theory building is the development of explanatory frameworks that make sense of disparate sources of data, by imposing patterns on them. Scientists often speaking of seeing patterns in the data, but the patterns that we see are constrained by our own theoretical frameworks. For students, the desired practice of making sense of patterns of data includes the explicit consideration of multiple sources of data. Also, in combination with the practice of coordinating theory and evidence, looking at the same data from alternative perspectives is an important way to make sense of them.
We view representational fluency as being able to interpret and construct various disciplinary representations, and to be able to move between representations appropriately. This includes knowing what particular representations are able to illustrate or explain, and to be able to use representations as justifications for other claims. This also includes an ability to link multiple representations in meaningful ways.
Understanding the criteria to which scientific claims are held, and the essentially social nature of science, requires that students' own knowledge claims be accountable to these criteria. We view discourse as a central means to realizing this practice in classrooms. Science classrooms should be organized as knowledge building communities (Scardamalia & Bereiter, 1993) in which discussion and debates about claims and evidence are central activities. Students should be encouraged to justify their claims, and causal claims should be challenged with respect to available data and consistency with other theories and knowledge.