On the reliability and validity of children’s metamemory

Abstract

Test-retest reliability, internal consistency, and construct validity of a metamemory battery were assessed. Fifty-five second-graders were given an extensive metamemory battery on two occasions separated by a 6-week interval. Test-retest correlations ranged from .29 to .49 for individual subtests, with a correlation of .67 for the composite battery. The children were then given training on the use of an interrogative-elaboration strategy to aid paired associate learning. Following three training sessions, tests for strategy maintenance and generalization were administered. Correlations relating strategy use with metamemory were significant for the transfer tasks, even when general knowledge was partialled out. The concept of metamemory appears useful in explaining production deficiencies and insightful strategy invention in young children.Metamemory refers to self-knowledge about memory processes (Brown, 1978;Flavell & Wellman, 1977).Recent studies have examined the construct validity of metamemory in terms of its relationship to a variety of strategic behaviors Kendall, Borkowski, & Cavanaugh, 1980; Beuhring, Note 1). While significant correlations between memory knowledge and memory processes or recall accuracy have been found, the modest size of these relationships is not strongly supportive of extant metamemory theory. The low to moderate correlations may be due as much to weaknesses in the instruments measuring meta· memory as to the theoretical invalidity of the concept . The present studies attempted to measure the testretest reliability, internal consistency, and construct validity of a new metamemory battery. In Experiment 1, 55 second-graders were given a seven-item metamemory battery on two occasions, with testing sessions separated by a 6-week interval. The metamemory test was composed of four items from the Kreutzer, Leonard, and Flavell (1975) scale, a memory monitoring task (Levin, Yussen, DeRose, & Pressley, 1977), and two new sub tests assessing children's knowledge about input/output processing strategies. Experi- ment 2 examined the role of metamemorial knowledge as a predictor of strategy transfer. That is, the secondgraders, used originally to determine reliability, were trained on a four-step interrogative-elaboration strategy as an aid to paired associate learning. The elaboration strategy was not explicitly associated with any of the metamemory subtests. Kestner and Borkowski (1979) employed similar task and training procedures and found recall to be significantly related to the quantity and quality of elaborations used at transfer.Borkowski (in press) has argued that metamemory/ strategy-use relationships will be found only in limited contexts, such as on novel tasks or transfer tests. We hypothesized that metamemory, as indexed by our battery, would predict strategic behavior on paired associate tasks following the training of the elaboration strategy and would show stronger relationships than a measure of general intelligence. That is, young children given prolonged, thorough strategy...

Memory Development Between 2 and 20

“…Other studies (e.g., Borkowski, Peck, Reid, & Kurtz, 1983;Brown, 1978;Schneider, 1986;Schneideret al, 1986;Weinert et al, 1984) have incorporated some of the items, however, with most of the data consistent with Kreutzer et al's (1975) original fmdings. lt has been established that the items used in the study are generally reliable (Kurtz, Reid, Borkowski, & Cavanaugh, 1982). More recent interview studies are even better from a psychometric perspective, however.…”

“…There is convincing evidence of relationships between metamemory, Strategie behaviors and memor from studies that u s w1t atent varia les Hasselhorn, 1986;Kurtz et al, 1982;Schneider, Körkel, & Weinert, 1987b;Weinert et al, 1984;Weinert, Schneider, & Knopf, in press). These analyses permit the conclusions that even relatively young elementary-school children possess knowledge about organizational strategies that has a direct influence on strategic behaviors, which increases the tendency to organize to-be-learned materials taxonomically.…”

“…Task-specific metamemory is usually a better predictor than generat metarnemory measures. Metamemory predicts recall even when other factors are controlled statistically (e.g., Kurtz et al, 1982).…”

Metamemory and metalinguistic development: Correlates of children’s intelligence and achievement

“…In short, the first four subtests consisted of the assessment of verbalized, declarative knowledge regarding the actions of a hypothetical child, whereas the other two subtests encouraged the child to engage actively in tasks requiring metamemorial judgments. Details of the administration and scoring of these subtests can be found in Kurtz et al (1982). For purposes of the present research, raw scores for each subtest were transformed into standard (Le., "z") scores and summed across subtests for each child.…”

“…This article outlines empirical contact points between metamemory and metalinguistic development and their association with general intelligence and achievement. Kurtz, Reid, Borkowski, and Cavanaugh (1982) developed a six-item metamemory battery, assessed its reliability, and demonstrated its validity by showing its relationship with the transfer of an acquired elaborative strategy. Similarly, Ryan and Ledger (1982) showed that metalinguistic knowledge of phonological and syntactic features and of the functions of reading correlate with early reading and sentence-memory performance.…”