hypothesis, a limited framework theory, among others (Chinn and Brewer, 1993). More explicit metacognitive knowledge (that allows one to identify and describe different sources of problems) can help direct the learner’s attention to determine the likely source of the problem given further information. For example, the learner might check to see if the result is reproducible, reexamine the data collection methods, compare results with other groups, etc. Thus, a second function of metacognition is a more directive and reflective one: to consider possible reasons for the incongruity and gathering or selecting further information that helps refine one’s understanding of the problem. Adding a reflective component to learning not only speeds up the time it takes to learn, but also makes it possible to learn things that one might never figure out through trial and error (Case, 1997).
Evidence that metacognition plays a key role in conceptual change learning comes from a variety of types of studies. The instructional techniques that have been shown to be effective in producing conceptual understanding of new science content all have a strong metacognitive component (Minstrell, 1982, 1984; Nussbaum and Novick, 1982; Chinn and Brewer, 1993; Roth, 1984; Brown and Campione, 1994; Hennessey, 2003; Beeth and Hewson, 1999; National Research Council, 2000). Typically, activities are introduced to make students aware of their initial ideas and that there may be a conceptual problem that needs to be solved. Students may be asked to make a prediction about an event and give reasons for their prediction, a technique that activates their initial ideas and makes students aware of them. Class discussion of the range of student predictions foregrounds alternative ways of thinking about the event, further highlighting the conceptual level of analysis and creating a need to resolve the discrepancy. Gathering data that expose students to unexpected discrepant events or posing challenging problems that they cannot immediately solve are other ways of sending signals to students that they need to stop and think, step outside the normal “apply” conceptual framework mode, to a more metaconceptual “question, generate and examine alternatives, and evaluate” mode. In addition, experimental manipulation of the amount of reflective inquiry and self-assessment in two identical versions of a carefully designed inquiry unit on motion for sixth graders produced greater gain scores for the students in the classroom with enhanced opportunities for reflective self-assessment (White and Frederickson, 1998). Particularly important, the gains were evident for all ability levels; indeed, they were highest for students with lower ability levels.
Elementary schoolchildren have much more capacity for metacognitively guided learning than has been commonly supposed or taken advantage of by existing science curricula (see Hennessey, 2003, for a detailed analysis of the subtle and diverse expressions of metacognitive understandings shown