Academic journal article Memory & Cognition

Individual Differences, Rereading, and Self-Explanation: Concurrent Processing and Cue Validity as Constraints on Metacomprehension Accuracy

Academic journal article Memory & Cognition

Individual Differences, Rereading, and Self-Explanation: Concurrent Processing and Cue Validity as Constraints on Metacomprehension Accuracy

Article excerpt

The typical finding of metacomprehension studies is that accuracy in monitoring one's own level of understanding is quite poor. In the present experiments, monitoring accuracy was constrained by individual differences in both reading comprehension ability and working memory capacity (WMC), but rereading particularly benefited low-ability and low-WMC readers, effectively eliminating the relationship between monitoring accuracy and these reader characteristics. In addition, introducing a self-explanation reading strategy improved the accuracy of all the readers above mere rereading. The observed interaction between individual differences and rereading is interpreted in terms of concurrent-processing constraints involved in monitoring while text is processed, whereas the more general self-explanation effect is interpreted in terms of accessibility of valid, performance-predicting cues.

The typical finding from studies of metacomprehension is that monitoring accuracy is quite poor. Readers' judgments of their comprehension and their actual performance on tests correlate only at about .27 (for reviews, see Lin & Zabrucky, 1998; Maki, 1998; Weaver, Bryant, & Burns, 1995). These results show that readers have poor relative accuracy, meaning that they are unable to judge how their understanding varies from one text to the next. Without accurate metacognitive monitoring, readers will be ineffective in regulating their studying, strategies, and cognitive efforts that produce increased learning (e.g., Nelson & Narens, 1990; Thiede & Dunlosky, 1999).

In the present investigations, we explore two factors that may contribute to poor relative metacomprehension accuracy: (1) the requirement to carry out both text processing and monitoring processes concurrently and (2) limited accessibility of valid cues that predict comprehension. In the first experiment, we investigated whether individual differences assumed to relate to executing the concurrent tasks would predict monitoring accuracy but would become less relevant after rereading. The second experiment replicated Experiment 1 while also testing whether a self-explain task designed to shift readers to an inference level of text processing would improve monitoring accuracy for all readers, presumably by increasing access to inferencelevel monitoring cues.

Successful metacomprehension monitoring requires the execution of two separate tasks. Readers are typically given the goal of reading for comprehension, either explicitly or implicitly via the existence of comprehension tests. In fact, it makes little sense to ask readers to monitor their comprehension if they do not already have the goal of comprehension. However, monitoring of comprehension is a separate, additional task above and beyond reading for comprehension. Just because readers understand a text's meaning (i.e., comprehension) does not mean they have also correctly assessed how complete their understanding of the text actually is (i.e., metacomprehension).

The implications of metacomprehension's being a secondary process to text comprehension have been largely ignored, but they follow directly from what have been widely presumed to be critical features of the metacognitive system. The initial and current models of metacognition presume that different cognitive processes are occurring at two separate levels, referred to as procedural versus metaprocedural, cognitive versus metacognitive, or object level versus metalevel (e.g., Fischer & Mandl, 1984; Flavell, 1979; Nelson & Narens, 1990). Comprehension involves attention to and processing at the object level, where the object is the text, whereas metacomprehension involves attention to and processing at the metalevel, where the focus is upon one's own mental processes and representation of the text. In addition, the act of monitoring consists of an unbalanced "dominance relation" in which the metalevel is secondary to and must "be informed by the object-level" (Nelson & Narens, 1994, pp. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.