Metacognitive Development and the Cognitive Internal State Lexicon

This investigation sought to determine the role of cognitive word knowledge in metacognitive development. Subjects were fifth graders, seventh graders, tenth graders, and college undergraduates. Each subject completed a Likert scale self-report qWstiomuNire that measured the frequency of their metacognitive strategy use and a multiple-choice test that measured their cognitive word knowledge. Achievement percentiles were then collected from the subject’s records. Metacognitive strategies representing the production of external strategies were found to decrease with age (e.g., stzdy habits), while metacognitive strategies representing internal processing increased with age (e.g., muterid ehzborutid, Only the metacognitive strategies assessing more internal processing were significantly related to standard&d verbal and quantitative achievement percentiles. Finally, cognitive word knowledge, explained a significant amount of variance in metacognition total for seventh and tenth graders, and undergraduates, but in tenth graders only when metacognition was partialled for verbal achievement percentiles. Cognitive words such as think and know are a subdivisionof the internal state lexicon (Hall & Nagy, 1986) and may be central to accessing, monitoring, and transforming our internal states (Scholmck & Hall, 1991). Cognitive words may enable people to understand and interrelate various aspects of their mental functioning (Scholnick & Hall, 1991). Cognitive words may provide a medium that makes it possible to engage in metacognition, that is, to generate goals for reading, communicate the intended meaning of a text, or evaluate one’s level of understanding. Similarly, cognitive words may equip the reader with a vehicle by which to evaluate comprehension strategies critically or to consciously reflect on the logical organization and interdependence of the components of a text. At least, the correct and independent semantic use of cognitive words represents a conscious act of metacognition 2 James R. Booth and William S. Hall (Hall & Nagy, 1986). Since cognitive words label internal mental states, they may facilitate metacognitive development. More specifically, cognitive word knowledge may enable children to more efficiently derive metacognitive strategies from independent problem solving, social interaction, and instruction, because cognitive words provide pertinent epistemological knowledge. Since many cognitive words have several semantic definitions and pragmatic functions, exposure to this lexicon may facilitate the development of metacognition, that is, the ability to realize that a word is only a symbol for its referent, that context determines the polysemous nature of words, and that language can be an object of thought. Also, children are likely to generalize this knowledge of multiple functions and meanings to other lexical domains, and compare uses within and between lexical domains. The embellishment and enrichment of the different functions and meanings of one cognitive word and/or between cognitive words raises a child’s consciousness of the distinctions these words make and the processes they designate. This leads to greater reflection on, and mastery of, the child’s knowledge system. Cognitive words may provide epistemological knowledge that accelerates this metacognitive development. Despite the commonly accepted relationship between cognitive words and metacognitive strategies (e.g., Olson & Torrance, 1986; Scholnick & Hall, 1991), only one study provides direct, but inconclusive, empirical evidence for their relationship (Olson & Torrance, 1987). The present study sought to remedy this situation by relating cognitive word knowledge, as measured by a multiple-choice task, to metacognitive strategy use as measured by a self-report Likert scale questionnaire. Furthermore, we expected to confirm the traditionally strong relationship found between certain metacognitive strategies and reading comprehension as measured by verbal achievement percentiles. Unfortunately, any investigation of the development of metacognitive strategies is difficult because, depending on their perspective, different researchers have different conceptions of metacognition. According to Flavell (1976), metacognition generally refers to “one’s knowledge concerning one’s own cognitive processes and products or anything related to them” and includes “the active monitoring and consequent regulation and orchestration of these processes” (p. 232). Later, Brown (1979) divided metacognition into two broad categories: (1) activities that involve conscious reflection on one’s cognitive abilities and activities; and (2) activities that are concerned with “self-regulatory mechanisms during an ongoing attempt to learn or solve problems” (p. 244). Other researchers have focused on more specific aspects of metacognition, such as knowledge of the structure of language (Tumner & Bowey, 1984), memory processes (Schneider, 1985), or strategies and self-regulation during reading. For example, Armbruster, Echols, and Brown (1982) divided metacognitive reading into four levels of awareness: (1) Text metacognition refers to the awareness of the logical organization, intended meaning, and inadequacies of written composition. (2) Task metacognition refers to the ability to change reading based on purpose, select appropriate retrieval cues, and anticipate one’s comprehension. (3) Strategy metacognition NATIONAL READING RESEARCH CENTER, READING RESEARCH REPORT NO. 18 Metacognition and Cognitive Words 3 involves the realization that study and comprehension strategies are beneficial to reading. (4) Learner characteristic metacognition refers to the awareness that an individual’s past knowledge is valuable for comprehension. This taxonomy informs the current investigation because it is concerned primarily with strategies employed during reading. Many investigations have examined the development of metacognitive reading and study strategies, and these may be classified into several general categories: alteration of learning based on purpose, elaboration of text, monitoring of learning, use of study habits, detection of main ideas, and organization of knowledge. For example, younger and poor readers do not distribute their study time effectively -they do not use a successful text reinspection strategy (Gamer, Macready, & Wagoner, 1984; Gamer & Reiss, 1981; Garner, Wagoner, & Smith, 1983; Forrest & Waller, 1980). In contrast, older and good readers distribute their study time appropriately. They spend more time studying unlearned words (Masur, McIntyre, & Flavell, 1973), more difficult texts (Owings, Petersen, Bransford, Morris, & Stein, 1980), and the main ideas of the text (Brown & Smiley, 1978). Many studies also show that younger and poor readers are deficient in main idea detection (e.g., Yussen, Matthews, Buss, & Kane, 1980; Brown & Smiley, 1977; Brown, Smiley, & Lawton, 1978), which appears to be at least partially dependent on the students’ knowledge of how organization affects learning. For example, older and good readers realize that memory varies with the individual and the task demands, that category structure aids memorization, that gist and verbatim recall tasks are different (Kreutzer, Leonard, & Flavell, 1975), that the first and last sentences of a story are of particular significance, and that skimming involves only processing the words that contain the most information (Myers & Paris, 1978). Finally, older and good readers can explain how well-organized and disorganized stories differ, whereas younger children are unaware of the relationship between organization and recall (Danner, 1976). Taken together, this research reveals that younger and poor readers are deficient in their text organization knowledge and main idea detection (Elliot, 1980). Ability and age-related differences also appear frequently in “strategies for understanding’ or text elaboration, but not in “strategies for remembering” or study habits (Baker, 1989). Good, not poor, college readers use internal and high-level text elaboration, such as identifying important ideas, identifying logical relationships in the text, relating information to prior knowledge, and reacting emotionally and critically to the text. In contrast, good and poor college readers have similar proficiency in study habits, such as summarizing, paraphrasing, and underlining words in text (Spring, 1985; Kaufman, Randelett, & Price, 1985). Younger and poor readers use observable low-level strategies for remembering and are often limited to decoding abilities such as pronouncing words correctly, knowing the meaning of words, reading loud fluently, determining whether there were long words, and concentrating on the size of the print (Garner & Kraus, 1981-1982; Myers & Paris, 1978). Many studies also show that younger and poor readers are deficient in monitor learning (Markman, 1977, 1979); they spend less time NATIONAL READING RESEARCH CENTER, READING RESEARCH REPORT NO. 18 4 James R. Booth and William S. Hall on anomalous sentences than older and good readers (Harris, Kruithof, Terwogt, & Visser, 1981; August, Flavell, & Clift, 1984). These deficiencies may be the result of the readers’ lack of awareness of their comprehension failure or may come from focusing on the empirical truth, intrasentence consistency, and individual word meaning at the expense of detecting intersentence consistency (Gamer, 1981; Markman & Gorin, 1981; Wagoner, 1983). Even high-school children have difficulty identifying inconsistencies (Otero & Campinario, 1988), and undergraduate college students can only detect about one-third of the errors in a written text, even after being directed to find them (Baker, 1979). However, this error detection seems to depend on content; adults spend more time resolving inconsistencies involving the main point than resolving trivial inconsistencies in detail (Baker & Anderson, 1982). In summary, younger and poor readers seem to be limited to decoding words and detecting intrasentence consistencies. They tend to use strategies for remembering, such as study habits and monitor learning. These are the more productive and external strategies that are often used to alleviate comprehension failure. Furthermore, since these strategies do not guarantee higher levels of reading comprehension, their frequency of use declines with age. In contrast, older and good readers use strategies for understanding, such as material elaboration and personal elaboration. Older and good readers also know that organization affects learning, and can alter learning based on purpose, as well as detect main ideas. So far, we have reviewed some established metacognitive taxonomies and the developmental course of these metacognitive strategies. However, the illumination of the mechanisms involved in the development of metacognitive strategies will allow an understanding of human development at a deeper level (see Siegler, 1989a). We suggest that metacognitive strategies develop through direct instruction, independent problem solving, and social interaction, as well as through exposure to and use of cognitive words. Most researchers agree that acquisition of metacognitive strategies is dependent on learning experiences (e .g . Pressley , Borkowski, & O’Sullivan, 1985). For example, certain metacognitive strategies improve with direct instruction, and others, such as text reinspection and text summarization are only learned with comprehensive instruction(Gamer & Alexander, 1989). However, the efficacy of these instructional programs depends on the age of the child, the strategy taught, and the domain of strategy application (Palincsar & Brown, 1981, 1984; Paris, Cross, & Lipson, 1984; Paris & Jacobs, 1984; Paris & Oka, 1986; Paris, Saamio, & Cross, 1986). Other metacognitive strategies arise from independent problem solving experiences (Brown & Deloache, 1978). A child’s confrontation with and solution of increasingly complex problems is an important source of metacognitive development. A child learns through trial, error, and insight about the most efficient way to study or to solve a problem. Garner (1987) describes the probable intuitive experiences that many children go through during independent problem solving: “He or she may experience relief about the format of the next day’s quiz; might realize that dates are becoming confused in his or her mind and that a NATIONAL READING RESEARCH CENTER, READING RESEARCH REPORT NO. 18 Metacognition and Cognitive Words 5 rehearsal strategy should be employed; and might have an ‘aha’ that boldface topic sentences in the textbook highlight particularly important dates” (p. 39). Finally, social interaction plays a definitive role in metacognitive development (Day, French, & Hall, 1985; Gardner & Rogoff, 1982; Wertsch, 1979, 1985; Wood, 1980). According to Vygotsky (198l), all higher psychological functions develop in social interactions. Children must first use and practice their mental processes -voluntary attention, logical memory, the formation of concepts, and planning -in this social context before they can achieve awareness and control over them. Day et al. (1985), after Vygotsky (1962, 1978), summarizes the role that social interaction may play in cognitive development: First, cognitiveabilitiesare socially transmitted. Adults and older peers pass on knowledge and skills required in their culture to children. Second, cognitive abilities are socially constrained. Children employ skills in social interaction that they cannot use when working in isolation. Third, nascent cognitive abilities are socially nurtured. Adult and more expert peers allow children to practice new skills by assuming responsibility for other aspects of the children’s activities.. . . Fourth, independent use of new cognitive abilities is socially encouraged. Adults and more expert peers take on less of the work load as children demonstrate increasing competence. (pp. 33-34) In conclusion, the efficiency of learning from independent problem solving, social interaction, and direct instruction is likely to interact with learner characteristics, such as cognitive word knowledge. We argue that a child high in cognitive word knowledge may benefit more from these independent learning experiences and from metacognitive skills training than a child low in cognitive word knowledge, because cognitive words label all aspects of metacognitive processes and therefore may facilitate insightful problem solving and strategy discovery. Since cognitive words allow people to access, monitor, and transform their internal states, they may enable people to understand, interrelate, and evaluate various aspects of their mental functioning (Scholnick & Hall, 1991). Cognitive words are an episternological content knowledge that may expedite metacognitive development. Garner and Alexander (1989) have encouraged researchers of metacognition to concentrate more on the relationship between content knowledge and metacognition. Three findings may provide insight into the relationship between cognitive word knowledge and metacognition. First, content knowledge may be essential for learning effectively in a domain. For example, pertinent prior knowledge enhances the magnitude, organization, and accessibility of information within that domain (Robinowitz & Glaser, 1985), and allows the reader to detect, process, and retain the important ideas in a text in an organized way. (McDaniel & Einstein, 1989; Recht & Leslie, 1988). Second, content knowledge in a domain is necessary for efficient strategy use (e.g., detecting inconsistencies) and may be necessary for strategy instruction in that domain (Alexander, Pate, Kulikowich, Farrell, & Wright, 1988; Borkowski, Carr, & Pressley, 1987; Bransford & Heldmeyer, 1983; Brown & Palincsar, 1985; Bymes & Wasik, 1991; Hasselhom & Korkel, 1986; Siegler, 1989b). NATIONALREADING RESEARCH CENTER,READINGRJZSEARCH REPORT NO. 18 6 James R. Booth and William S. Hall Third, strategic processing changes as knowledge in a domain increases (Alexander & Judy, 1988). High content knowledge as opposed to low content knowledge is generally accompanied by more abstract problem representations (Voss, Blais, Means, Greene, & Ahwesh, 1986) and a tendency to categorize or classify problems on the basis of underlying concepts or principles (Adelson, 1981; Shoenfeld & Herrmann, 1982). Since cognitive words label all types of metacognitive acts and much of learning occurs through the oral mode, these words may provide a necessary epistemological knowledge allowing more efficient acquisition of metacognitive strategies from learning experiences. The present study had three aims: (1) to investigate the developmental trajectory of different metacognitive strategies, (2) to determine how these metacognitive strategies related to verbal achievement percentiles, and (3) to determine whether cognitive word knowledge explained a significant amount of variance in the metacognitive total over and above the variance explained by the verbal achievement percentiles.