View article

The measurement of metacognition has gone through four overlapping phases: The first phase began with the insightful and stimulating paper of Kreutzer, Leonard, and Flavell (1975) on introspective reports about memory states and processes, followed by an important theoretical chapter on the nature of metamemory (Flavell & Wellman, 1977). These early contributions documented, and theoretically clarified, the fact that children could accurately report their knowledge about memory events as they related to a variety of tasks, circumstances, and strategies; furthermore, memory knowledge was shown to be age-related. A second phase quickly followed: The intention here was to show interconnections between memory knowledge and memory performance. Although hindsight now reveals that a modest relationship (r=. 42) links metamemory and memory across a wide range of learning contexts (Schneider & Bjorkland in press), an uncomfortable feeling about the" fuzziness" of the concept prevailed during this second stage of research (Wellman, 1983). From our vantage point, three interrelated conceptual and methodological problems surfaced that hindered the search for reliable and valid measures of metacognition-problems that continue to influence contemporary research and theory development: