UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comprehension in children with hyperlexia Lester, Michele L. 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2003-0502.pdf [ 7.46MB ]
JSON: 831-1.0053829.json
JSON-LD: 831-1.0053829-ld.json
RDF/XML (Pretty): 831-1.0053829-rdf.xml
RDF/JSON: 831-1.0053829-rdf.json
Turtle: 831-1.0053829-turtle.txt
N-Triples: 831-1.0053829-rdf-ntriples.txt
Original Record: 831-1.0053829-source.json
Full Text

Full Text

C O M P R E H E N S I O N I N C H I L D R E N W I T H H Y P E R L E X I A by M I C H E L E L . L E S T E R B . A . , The University of Manitoba, 1997 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F A R T S In T H E F A C U L T Y OF E D U C A T I O N (Department of Educational and Counseling Psychology and Special Education) We accept this thesis as conforming to the required standard. T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A June 10, 2003 © Michele L . Lester, 2003 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada D E - 6 ( 2 / 8 8 ) Comprehension in Hyperlexia i i Abstract The purpose of this study was to investigate the reading comprehension problems associated with hyperlexia in young children. Two 8-year-old children with hyperlexia and two children with mild intellectual delays (MID), matched in terms of age, grade, and 1Q, were evaluated using various decoding and reading comprehension tasks. Children with hyperlexia had excellent decoding abilities, far above what would be expected for their cognitive level, but did not comprehend passages as well as students without hyperlexia who had similar cognitive ability. Prior knowledge aided the reading comprehension for the M I D participants, where as high interest strongly aided the comprehension for the hyperlexia participants. Students with hyperlexia demonstrated less comprehension monitoring than the M I D students. Results are discussed with regards to implications for teaching children with hyperlexia and M I D . Comprehension in Hyperlexia i i i T A B L E O F C O N T E N T S Abstract Table of Contents ". List o f Tables.... Acknowledgements C H A P T E R I Overview and Summary Hyperlexia Comprehension in Children with Hyperlexia Comprehension in Skilled Readers Overview & Specific Research Question 1 C H A P T E R II Literature Review 1 Hyperlexia Comprehension in Hyperlexia Cognition and Memory Long-term Memory and Schema Metacognition and Strategy Use Motivation and Interest Summary : C H A P T E R III Method ... Phase 1 Measures Phase 2 Measures Phase 1 Procedure Phase 2 Procedure Comprehension in Hyperlexia iv C H A P T E R IV Results : 64 Identification and Screening 64 Participants 65 Phase 1 Results 67 Phase 2 Results 75 Summary 86 C H A P T E R V Conclusion/Discussion 88 Limitations 92 Future Directions 94 Conclusions 94 References 96 Appendix A (Informed Consent Form) 105 Appendix B (Error Detection #1 :Hyperlexia) 107 Appendix C (Error Detection #1 :MID) : 109 Appendix D (Error Detection #2:Hyperlexia) 111 Appendix E (Error Detection #2:MID) 113 Appendix F (Metacognitive Rating Scale).... 115 Appendix G (Self-Correction Task) 116 Appendix H (High Prior Knowledge Task:Hyperlexia) 118 Appendix I (High Interest PassagerHyperlexia) 120 Comprehension in Hyperlexia v List o f Tables Table 1 Congruent Responses on Rating Scale Questions 58 Table 2 Personal Information for Participants 65 Table 3 Results for Phase 1: Cognitive Abi l i ty , Wording Reading Measures, & Comprehension .. 69 Table 4 Data from Phase 2 Comprehension & Metacognitive Awareness Measures: Percent of Correct Answers 78 Comprehension in Hyperlexia v i Acknowedgements I would like to thank my thesis advisor, Nancy Perry, for her consistent advice and support. Without her knowledge and patience I would not have been capable of completing this thesis. I would also like to thank my thesis committee members, Pat Mirenda and Linda Siegel. Both of these women brought so much knowledge and insight to this project and I am so grateful to them for agreeing to be a part o f this thesis. I also owe so much thanks to my parents, Phi l and Irene Lester, for without their continuous support, both emotionally and financially, completing this thesis would not have been possible. Finally, I thank my husband for loving me and standing by me, and God for giving me the knowledge and stamina to complete such an arduous task. Comprehension in Hyperlexia 1 C H A P T E R I Comprehension in Children with Hyperlexia Past research about children with hyperlexia has focused primarily on their exceptional decoding abilities compared with their cognitive impairments (Cobrinik, 1982; Mehegan & Dreifuss, 1972; Richman & Kitchell , 1981; Silberberg & Silberberg, 1967, 1968-69, 1971; Siegel, 1984). However, recent literature has begun to focus more on the comprehension impairment that accompanies the advanced decoding skills within this population (Aaron, Frantz, & Manges, 1990; Goldberg & Rothermel, 1984; Snowling & Frith, 1986; Worthy & Invernizzi, 1993). Some research has shown that hyperlexic children have appropriate single-word comprehension (Cobrinik, 1982; Healy et al., 1982; Richman & Kitchel l , 1981; Siegel, 1984); however, there are still debates regarding the precise nature of their comprehension deficits (Aram, 1997). Specifically, research on comprehension o f children with hyperlexia has not resulted in consistent findings with regard to how much of a comprehension deficit there is and how much of it is syntactic or semantic (Aram, 1997). Additionally, only one study has briefly examined the role o f prior knowledge in children with hyperlexia and none to date have looked particularly at the metacognitive abilities in these children. Hence, these issues remain to be examined. Reading is a complex interactive process that involves attention, memory, metacognition, motivation, and strategic action. Reading is not simply decoding. Similarly, comprehending individual words does not by itself guarantee the understanding of a sentence or paragraph. C Comprehension of text requires that readers attend to meaning, and activate and apply background knowledge and experience or schema to text information to understand (Anderson, 1978; Rumelhart & Ortony, 1977). However, prior knowledge is insufficient. Comprehension also requires that readers use metacognition (awareness of one's own thinking) and strategies for reading and learning to understand text (Borkowski & Buchel, 1983; Borkowski & Kurtz, 1986). Metacognition, or thinking about how one learns, has been shown to be an important asset to Comprehension in Hyperlexia 2 comprehension o f text (Baker & Brown, 1984). Similarly, motivation plays an important role in reading comprehension. Research has shown that readers need both skill and wi l l in order to be successful comprehenders (Paris, Lipson, & Wixon , 1983). M y research wi l l investigate comprehension in children with hyperlexia. Specifically, what accounts for the poor comprehension in these children? Do these children use or have access to prior knowledge or schema about various concepts and do they make use o f metacognitive strategies to help them to make sense of what they are reading? In the sections that follow, the concepts of hyperlexia, schema, metacognition, motivation and strategy use wi l l be introduced and defined. The general characteristics of successful reading comprehension wi l l also be described. Finally, an overview of the research I conducted and the specific research questions I addressed wi l l be provided. Hyperlexia Definitions of Hyperlexia The term hyperlexia was first used by Silberberg and Silberberg (1967) to describe a group of 20 children they had observed whose word recognition skills were at least 2 years above their expected achievement level. In a subsequent study (Silberberg & Silberberg, 1968-69) these authors further examined these case studies and arbitrarily classified their subjects as hyperlexic i f their observed reading level was above the expected word-recognition level by 1.5 years for grades 1 and 2, and 2.0 years for grades 3 and up. In another in-depth study, Healy et al. (1982) investigated hyperlexia in 12 children with early and superior reading and word recognition skills. These children were observed and tested with regard to IQ, language, cognition, reading and behaviors. Additionally, their families were interviewed with regard to their developmental history as well as their family history. Through this intense testing and observation, the authors elaborated the definition of hyperlexia proposed by Silberberg and Silberberg (1967,1971). Healy et al. (1982) define hyperlexia as a syndrome that is characterized by: (a) spontaneous and intense early interest in letters and words prior to the age of Comprehension in Hyperlexia 3 5 years, coupled with, (b) significantly disordered language and cognitive development, and (c) word recognition (decoding) well advanced over other cognitive and linguistic abilities. Similarly, Siegel (1984) in a longitudinal case study of a hyperlexic girl, defined hyperlexia as a condition consisting of advanced word recognition skills in spite of both language and cognitive disorders. However, she added that expressive language in these children is usually very poor, and noted that hyperlexic children typically have IQ scores that range in the moderate to severely retarded range on standardized intelligence tests and exhibit disturbances in behavior, such as social withdrawal, rocking, or self-stimulatory behaviors. Operational Definition of Hyperlexia. For purposes of this research paper, hyperlexia w i l l be defined using Healy et al.'s (1982) criteria: (a) early and spontaneous word recognition prior to the age of 5 years, (b) accompanied by both language and cognitive delays, and (c) word recognition far advanced in comparison with other cognitive and linguistic functions. Comprehension in Children with Hyperlexia Although several studies have examined comprehension in hyperlexic children (Aram et al., 1984; Healy et al., 1982; Patti & Lupinetti, 1992; Siegel, 1984; Snowling & Frith, 1986; Temple, 1990), none have offered definitive conclusions about the nature of their comprehension deficits. Many of the studies have had different conclusions about how much of a comprehension deficit these children actually have. Some studies conclude that these children have very severe comprehension deficits, over and above what would be expected given their linguistic and cognitive delays (Healy et al., 1982; Siegel, 1984), while others suggest that these children's comprehension is commensurate with their IQ (Snowling & Frith, 1986; Temple, 1990; Welsh et al., 1987). Healy et al. (1982) examined the reading comprehension of 12 hyperlexic children. Various tests were used to examine both sentence and paragraph comprehension. On the Stanford Diagnostic Reading Test (SDRT) (Karlsen, Madden, & Gardner, 1976), the children scored well below the mean score for their age on comprehension. Even though the decoding ability of these children was, on average, very accelerated, they could not retell or summarize a story. Most of the Comprehension in Hyperlexia 4 children gave responses that were not even related to the text's meaning. Even on single-word comprehension tasks the children scored very poorly. The definitions given by these children on even the simplest concrete nouns, such as "door", were usually associative, and often described a personal sensorimotor experience, e.g., "I close" (Healy et al., 1982). Worthy and Invernizzi (1995) examined the comprehension deficit in a 14-year-old girl with hyperlexia. Despite the fact that this participant decoded words above a grade 12 level, her comprehension skills were severely deficient. Her single-word comprehension was found to be at a grade 2.6 level, while her passage comprehension sat slightly higher at a grade 3.1 level. However, higher passage comprehension compared to single word comprehension is not the norm for these children. Additionally, Worthy and Invernizzi (1995) found that on various reading inventories the girl was completely unable to retell anything about the story nor answer any of the questions asked. Overall, her comprehension level was considered to be below the primer level (Worthy and Invernizzi, 1995). In a similar study, Welsh et al. (1987) examined single-word comprehension skills in five boys with hyperlexia. These authors found unexpectedly precocious single-word reading in all the boys and precocious single-word reading comprehension (Welsh et al., 1987). There are still many unanswered questions regarding the reading comprehension of children with hyperlexia. For example, what accounts for the poor comprehension beyond the word level? A study that specifically investigates comprehension deficits in these children and looks at various aspects o f their comprehension is needed. Comprehension in Skilled Readers What does it take to be a skilled reader? Much research in the area of reading comprehension tells us that skilled readers take a strategic approach to reading (Paris & Jacobs, 1984; Paris & Oka, 1986; Paris, Lipson, & Wixon , 1983; Wong & Wong, 1986). These children engage in a variety of processes that enhance their reading comprehension. These processes or reading strategies occur before, during and after reading and include such things as goal setting, inference-making, identifying the main theme, summarizing, predicting, monitoring and rereading Comprehension in Hyperlexia 5 (Paris, Wasik, & Turner, 1991; Pressley et al., 1994). Readers who preview their text prior to reading it have increased comprehension of both implicit and explicit information (Graves & Cook, 1980). Previewing of text can be done in various ways, such as skimming text, looking at titles or looking at the pictures (Paris et al., 1991). Previewing activates relevant prior knowledge. Lipson (1983) found that children's prior knowledge about religious customs aided their comprehension when reading a passage on taking communion in church. During the reading process there are many tactics a reader can use to increase their comprehension, such as identifying the main topics and making inferences (Paris et al., 1991). Another tactic skilled readers use is called text inspection which includes skills, such as rereading. Garner and Reis (1981) examined rereading of text in eighth grade skilled and less-skilled readers and found that skilled readers used rereading where less-skilled readers did not. Similarly, Baker (1979) suggested that reading for meaning involves the use of comprehension monitoring which involves keeping track of the success of one's comprehension. This is a type o f metacognitive activity that ensures that reading for meaning continues as a smooth process and that remedial action (i.e., self-correction) is taken whenever necessary (Baker, 1979). A study by Paris and Myers (1981) investigated cognitive monitoring in both good and poor readers. Grade 4 children were divided into two groups depending on their reading abilities. The children read passages and answered questions about them. The investigators observed and recorded various aspects of comprehension monitoring. The authors found that poor readers engaged in significantly less comprehension monitoring and that this was correlated with poor comprehension and recall scores as compared with the good readers (Paris & Myers, 1981). After the reading process, skilled readers take time to review and reflect about the text. Brown and Day (1983) found that high-school students summarized information across paragraphs, restating it in their own words, whereas fifth and seventh-grade students tended to report information verbatim from the text. The authors found that the older subjects could more frequently order text by topics or ideas whereas the younger students followed the sequence of the, Comprehension in Hyperlexia 6 text (Brown & Day, 1983). Skilled readers are active processors of textual information. It is possible that children with hyperlexia are poor comprehenders because they do not actively process text but instead simply decode the words without checking their comprehension. However, there is not enough information on the metacognitive abilities in these children to make any conclusive statements. This study wi l l address these important issues with a view to providing some answers to these questions. Cognition and Memory Cognition and learning are both active and constructive processes. Learners must actively pursue knowledge through various cognitive and metacognitive processes in order to be successful. Learning requires attention to the task as well as encoding, storing and retrieving information in long-term memory. Readers comprehend text by combining new and prior knowledge as well as using cognitive and metacognitive strategies to ensure the text is understood. Skilled readers possess both declarative knowledge (knowing what) and procedural knowledge (knowing how) about reading, as well as executive control processes that allow them to regulate their own reading processes. Children who are skilled comprehenders bring together various components of memory and cognition, such as attention, schema, metacognition, strategy use, and motivation and interest in order to make sense of text. Attention In order to read, one must first attend to the task. Perception and attention play a large role in what information is selected for processing and storage. Attention, therefore, plays an important role in reading and learning. Reading involves a combination of automatic processes and control processes. Automatic processes occur effortlessly and make minimal demands on attention and working memory (e.g., driving to an experienced driver). However, many reading tasks require the allocation of conscious attention and working memory. More complex activities, such as reading difficult text and recognizing a breakdown in comprehension, elaborating on text meaning, or bridging inferences to integrate meaning are the types of tasks that require attention. Comprehension in Hyperlexia 7 Long-term memory — Schema Memory is an active process that involves encoding, storing and retrieving information. Information must first be attended to before it is processed into long-term memory. Additionally, the learner must use strategies to actively encode the information before it can be stored in long-term memory. Once information is stored, it can be used to aid learning of new information. Schema theory is one way o f describing how knowledge is stored in long-term memory. It assumes that all knowledge is packaged into units called schemata. For example, a schema for a chair is: an item with four legs, a back, and a seat that is used for sitting on and often made out of wood. Everyone has schema for a variety of generalized objects, and events (Nelson, 1986). Schema have at least four major functions: (a) they categorize your experiences, (b) they help us remember, (c) they aid in comprehension, and (d) they aid in problem-solving (Smith, 1989). Reading comprehension occurs when a reader is able to connect information in the text with knowledge/schema stored in long-term memory. How Schemata Aid in Comprehension. Many studies have shown that prior knowledge or schemata influence a reader's comprehension (Afflerbach, 1990; Anderson, 1996; Langer, 1984; Pearson, Hansen, & Gordon, 1971; Rumelhart, 1977; Smith, 1989). Garner (1987) claims that for text comprehension to occur, schemata (prior knowledge) must be activated to aid in processing any new text information. Old information in memory acts as an anchor for new information in text (Ausubel, 1963). The old information, once activated, assists the processing of the new information and this leads to successful comprehension. Rumelhart (1980) has proposed three ways in which the reading process might fail: (a) by not having appropriate schemata, (b) by not being able to activate the appropriate schemata, and (c) activating schemata that is not appropriate. A study by Pearson, Hansen, and Gordon (1979) investigated schema in second grade students using passages about spiders. The authors presented the class of students with eight pretest questions to determine their background knowledge of spiders. The children were then placed into two groups based on their prior knowledge, weak and strong schemata about spiders. The participants then read a passage about spiders and answered several.wh- questions that Comprehension in Hyperlexia 8 assessed both explicit and implicit knowledge about spiders. The authors found that students with well-developed schemata were better at answering both types of questions as compared to the weak schema group (Pearson et al., 1979). These results were interpreted as supporting schema theory. Taylor (1979) also tested theories on schema and text comprehension. In her study, grade 3 and grade 5 good and poor readers were tested in regard to prior knowledge and comprehension of text. Both groups of children were given passages on a familiar topic (bird nest building) and an unfamiliar topic (bee dancing) and asked to read and recall the content of both passages. The results showed that participants recalled less from the unfamiliar text topic than from the familiar topic (Taylor, 1979). The results of these investigations support schema theory. The poor comprehension of children with hyperlexia may be due in part to a lack of schema for topics read or difficulties activating relevant schema. Operational Definition of Schema. For the purpose of this investigation, schema wi l l be defined as relevant prior knowledge or topic familiarity that readers apply to comprehend text. Metacognition and Strategic Action Metacognition is knowledge about cognitive phenomena or thinking about one's own thought processes (Flavell, 1979). It consists of two important components: (a) awareness of one's thoughts, and (b) regulation of one's thoughts. Flavell and Wellman (1977) identified three variables that make up metacognition: (a) knowledge about self (i.e., my learning strengths and weaknesses), (b) knowledge about tasks, and (c) knowledge about strategies. Metacognitive knowledge serves to coordinate and direct thinking and behavior, while the executive functions regulate thoughts, such as planning, monitoring, testing and revising of one's thinking. Learning strategies exist on both a cognitive and metacognitive level. Cognitive strategies for learning include such things as verbal rehearsal or the use of mnemonics where as metacognitive strategies would include activities such as setting goals for learning and then monitoring progress towards those goals. Reading strategies that promote comprehension can be applied before (e.g., skimming), during (e.g., comprehension monitoring), or after reading the text Comprehension in Hyperlexia 9 (e.g., re-reading the text). Strategic readers iise appropriate strategies that fit the text, purpose and occasion (Paris, Wasik, & Turner, 1991). How Metacognition Aids in Comprehension. Metacognition or, more specifically, "metacomprehension" has been shown to play an important role in the comprehension process. Garner (1987, 1994) suggests that reading, since it is such an interactive and constructive process, logically involves executive control and metacognition. There has been a vast amount of research dedicated to the area of metacognition and reading with regard to how the active reader can optimize their own learning (Brown & Campione, 1977; Garner & Kraus, 1982; Garner & Reis, 1981; Myers & Paris, 1978; Paris & Myers, 1981). Ryan et al. (1982) suggest that the comprehension failure in poor readers who have adequate decoding skills is largely due to their lack of active participation or strategy use while reading. In a study of metacognitive skills, Flavell and Wellman (1977) investigated the recall o f preschool and elementary school children on a set of items. The children were asked to study these items until they could remember them perfectly. What they found was that older subjects, after studying for a while, said they knew the items and showed perfect recall. In contrast, the younger children studied for a while, said they were prepared, and then showed very poor recall. These results were the foundation for further studies into metacognition, or one's knowledge about their own thinking. Many studies (Brown, 1978; Flavell , 1978; Flavell & Wellman, 1977; Myers & Paris, 1978; Paris & Myers, 1981) have examined metacognition in children and found that poor readers do not have adequate metacognitive skills, such as cognitive monitoring. Poor metacognition and comprehension monitoring can lead to comprehension failures. Paris and Myers (1981) compared comprehension and memory skills of fourth grade poor and good readers. The children's comprehension monitoring was measured through self-corrections, directed underlining of incomprehensible words, and study behaviors. The authors found that poor readers engaged in significantly less monitoring on all three measures as compared to the good readers. These findings were also correlated with difficulties in comprehension and recall. Additionally, the poor readers adopted decoding strategies rather than Comprehension in Hyperlexia 10 comprehension strategies during reading and they were less skilled at applying cognitive monitoring during comprehension failures (Paris & Myers, 1981). One method used to assess comprehension monitoring is error detection. In an error detection task, errors are embedded in a text so that it no longer makes sense to the reader. Garner and Kraus (1982) investigated the detection of inconsistencies both in and among sentences by good and poor comprehenders. Thirty children in the 5 t h grade participated in this study, 15 good and 15 poor comprehenders. The children were given two passages, one about a train, and one about a ship. Each passage contained five sentences of similar structure. Inconsistencies were contained in the first and last sentences. The students were also given two additional altered passages that were adapted in order to create intersentence inconsistencies (inconsistencies within the sentence only). These inconsistencies were found in the last sentence of both these passages. The authors reported that only four of the 12 good comprehenders found the intersentence inconsistencies. A l l 12 good comprehenders detected the intrasentence inconsistencies. None of the 15 poor comprehenders detected either type o f inconsistency. Garner and Kraus (1982) concluded that the poor comprehenders had not used any comprehension monitoring strategies. A study by Paris and Oka (1986) clearly demonstrated the importance of metacognitive skills in aiding comprehension. These authors implemented an experimental curriculum that taught strategy use to five hundred children in grades 3 and five hundred in grade 5. A number of pre- and post-test batteries were administered to all the children. Paris and Oka (1986) reported that all o f the children, regardless of reading ability, who received the strategy training increased their awareness about reading and had significant gains in metacognition and comprehension levels. In more recent investigations on metacognition, Turner (1995) found that grade one children, when given open literacy tasks, used more metacognitive behaviors and strategies, such as rehearsal, planning and self-monitoring. Similarly, Neuman (1997) reported that preschoolers engaged in strategic behaviors in a number of problem-solving situations. Finally, Perry (1998) examined self-regulated learning (SRL) in grade 2 and 3 children. She found that in classrooms Comprehension in Hyperlexia 11 that were considered high-SRL (teachers use and promote SRL) , children engaged in more self-regulated learning, such as monitoring and evaluating their own writing (Perry, 1998). In conclusion, the investigations mentioned above clearly show how children with metacognitive skills who use comprehension monitoring strategies have better comprehension skills than those who do not have these metacognitive abilities. Additionally, when children are trained in these metacognitive strategies their awareness of reading and reading comprehension improves significantly. It is apparent that metacognition is necessary for successful comprehension of text. To date, very few studies of children with hyperlexia have examined their metacognitive activities. In the few studies that have briefly examined whether these children self-correct during reading (Healy et al., 1982; Worthy & Invernizzi, 1992), the findings suggest that these children seldom self-correct and hence show little use of cognitive monitoring. For example, in their study of 12 children with hyperlexia, Healy et al. (1982) found that although their participants did self-correct when reading they did so immediately after the word instead o f at the end of the sentence. Conversely, normal readers tend to self-correct errors at the end of a sentence once they realize the sentence made no sense. Healy et al. (1982) therefore suggested that this pattern of self-correction showed that these children caught errors on a word level and perhaps a syntactic level, yet they were not showing comprehension of the sentence. Similarly, Worthy and Invernizzi (1995), in their case study of a 14-year-old girl with hyperlexia, reported that out of 25 miscues made, none were corrected, two were syntactically acceptable but none were semantically acceptable. In light of this information, there is a high probability that children with hyperlexia lack the important metacognitive skills that aid in reading comprehension. Operational Definition of Metacognition. For the purpose of this investigation, metacognition wi l l be defined in terms of the students abilities to: (a) monitor their own comprehension by re-reading anything they missed or did not understand, (b) self-correct when reading errors occur, and (c) evaluate their understanding of the reading process. Motivation and Interest Comprehension in Hyperlexia 12 Numerous studies have determined that motivation plays an important role in learning and in the use of metacognitive strategies (Borkowski et al., 1985). Children need to be motivated and interested in order to use strategies for learning. Interest in a topic has been shown to increase a child's comprehension. Garner, Alexander, and Hare (1991) suggest that low interest in a topic or task is one of the variables that can cause failure to comprehend text material. Additionally, researchers have reported that children with high interest in a particular topic have shown better comprehension o f that topic compared to topics of low interest. A recent study by De Sousa and Oakhill (1996) investigated the effect of levels of interest on comprehension monitoring in 8- and 9-year-old good and poor comprehenders. The children were given passages of both high- and low-interest that contained embedded errors. The authors found that the good comprehenders performed well on both tasks, but the poor comprehenders performed significantly better on the high interest than the low interest passages (De Sousa & Oakhil l , 1996). This confirmed the hypothesis that level of interest has a positive influence on comprehension and comprehension monitoring in children with poor comprehension skills. Limitations of Past Research There are some methodological limitations in several of the earlier investigations of hyperlexia. To begin with, few studies identify comprehension as the primary focus o f their investigation. Aram (1997) suggests that, although most investigators agree that there is a comprehension deficit in hyperlexia, there is not enough research in the area to determine the extent or the nature of this deficit. Many of the studies (Aram, Rose, & Horwitz, 1984; Elliot & Needleman, 1976; Patti & Lupinetti, 1993; Siegel, 1984; Seymour & Evans, 1992; Worthy & Invernizzi, 1993) were single subject case studies that do not allow for any comparison or generalizations to be made. Additionally, in the few studies that did have multiple subjects (Goldberg & Rothermel, 1984; Healy et al., 1982; Silberberg & Silberberg, 1967; Snowling & Frith, 1986) there were a wide range of IQ scores and ages reported in the various subjects. In the Healy et al. (1982) study, the ages of the children ranged from 5 through 11 years-of-age. The participants in the Goldberg and Rothermel (1984) study had significant variance in their ages, Comprehension in Hyperlexia 13 ranging from 5 through 17 years-of-age. Snowling and Frith (1986) selected their participants on the basis of reading age. The autistic hyperlexic children ranged in reading age from 8 to 10 years-of-age. However, their matched comparison group o f mildly retarded children ranged in reading age between 7 and 14 years-of-age (Snowling and Frith, 1986). Similarly, the variance in IQ for subjects in the Silberberg and Silberberg (1967) study ranged from nontestable to 126 on the Stanford-Binet (Terman & Merr i l l , 1960). Similarly, Healy et al. (1982) reported IQ ranges on the McCarthy Scales (McCarthy, 1972) from 47 to 91 in their participants. Goldberg and Rothermel (1984) reported only mean scores of their subject's IQ ratings from the W I S C - R (Wechsler, 1974) o f 70. Since the true definition of hyperlexia is one that describes children with definitive cognitive deficits, then many o f the children in the aforementioned studies would not fit into this category. M y investigation wi l l attempt to correct some of these prior limitations by specifically examining comprehension in hyperlexic children with definite cognitive deficits (IQ ranging from 40 - 70) and with a small age-range o f 7 - 11 years-of-age. Overview and Specific Research Question Children with hyperlexia have an excellent ability to decode words. (Healy et al., 1982; Siegel, 1984; Silberberg & Silberberg, 1967,1968,1971); however, their comprehension skills are found to be seriously deficient (Healy et al., 1982; Huttenlocher & Huttenlocher, 1979; Siegel, 1984; Silberberg & Silberberg, 1971). There is also consensus among researchers that hyperlexic children have great difficulty understanding language both written and oral (Aram, 1997). Although much research has been done on the reading process of hyperlexics (Cobrinik, 1982; Elliot & Needleman, 1981; Healy et al., 1982; Siegel, 1984; Seymour & Evans, 1992; Silberberg & Silberberg, 1971), little research has agreed upon the underlying reason for their poor comprehension (Aram, 1997). Some research has investigated the syntactic versus semantic aspects of language and reading comprehension of hyperlexics (Goldberg & Rothermel, 1984; Healy et al., 1982; Siegel, 1984; Seymour & Evans, 1992; Worthy & Invernizzi, 1992) but inconsistent findings have been reported. To date, few studies have specifically investigated the Comprehension in Hyperlexia 14 schematic knowledge, interest level, and metacognitive abilities in these children. Thus the present study is intended to further clarify the research concerning poor comprehension skills o f hyperlexic children arid extend this research, particularly in regard to their schemata or prior knowledge, level o f interest, and metacognition, specifically comprehension monitoring and self-correction. Two children who meet the criteria for hyperlexia and two children with mental retardation, matched according to their mental ages, were tested and compared on general comprehension abilities, schema (prior knowledge) and metacognition, in addition to general tests for attention, memory and single-word receptive vocabulary. The overall question that was answered is what accounts for the poor comprehension in these hyperlexic children? 1. Is their low cognitive language ability sufficient as an explanation for their poor comprehension (i.e., is there low cognitive and language ability sufficient as an explanation for their poor comprehension)? 2 . Do they comprehend better when prior knowledge is relatively high? 3. Do they comprehend better when their interest in a topic is relatively high? 4. Do they monitor their own comprehension? 5. Do they self-correct? Comprehension in Hyperlexia 15 C H A P T E R II Literature Review In this chapter I w i l l review the literature in the area of hyperlexia, particularly the definition, and characteristics of hyperlexia, and the possible effects hyperlexia has on reading comprehension. I w i l l also examine the hyperlexia literature in terms of comprehension, both listening and reading, with specific focus on reading comprehension. Next, I w i l l examine the areas of cognition and memory, particularly, I w i l l review the literature in the area of long-term memory and schema and its effects on comprehension. Then, I w i l l review the literature in the area of metacognition and strategy use and, specifically, look at how they aid reading comprehension. Additionally, I w i l l briefly examine the literature in the area of effects of interest on reading and comprehension. Finally, I w i l l end the chapter with a summary and conclusions regarding hyperlexia and reading comprehension that w i l l provide a strong argument for my thesis question. Hyperlexia was given its name by Silberberg and Silberberg (1967), who studied several cases of children who showed exceptional word decoding ability. However, there had been cases of hyperlexia described in the literature as early as the 1930's. Munroe (1932) described children with word reading ability that resembles what today is known as hyperlexia. He describes these children as "the defective child who reads fluently although he is unable to deal intelligently with the material read" (Munroe, 1932; p. 1). Even Kanner (1943) when observing children with infantile autism, identified particular children with reading ability and good memory for words. Yet it was not until fhel960's when Silberberg and Silberberg defined hyperlexia, "For this phenomenon of specific word recognition skill we have coined the term hyperlexia" (p. 41). Since this time there have been many studies investigating the incredible abilities and surprising cognitive and language impairments in these children (Cobrinik, 1982; Elliot & Needleman, 1981; Goldberg & Rothermel, 1984; Healy et al., 1982; Richman & Kitchel l , 1979; Siegel, 1984; Silberberg & Silberberg, 1971; Temple, 1990; Worthy & Invernizzi, 1992). The literature on hyperlexia since Silberberg and Silberberg (1967) has examined at various aspects of the disorder. Many studies have attempted to understand the precocious word Comprehension in Hyperlexia 16 decoding skills (Cobrinik, 1982; Goldberg & Rothermel, 1984; Richmond & Kitchel l ; Silberberg & Silberberg, 1968-69). Additionally, several attempts have been made by researchers to understand the incredible deficit in language comprehension in the hyperlexic child (Aaron, Frantz, & Manges, 1990; Glosser, Friedman, & Roeltgen, 1996; Healy et al., 1982; Richman & Kitchell , 1981; Seymour & Evans, 1992; Siegel, 1984; Snowling & Frith, 1986). A few studies have even examined hyperlexia occurring in children with autism (Snowling & Frith, 1986; Tirosh & Canby, 1993; Whitehouse & Harris, 1984). However, the research is very limited in the areas of prior knowledge, and no studies with hyperlexic children have been done looking at metacognition. Given the goals of my study, the literature review in this chapter summarizes the research in the areas of hyperlexia, paying close attention to the studies on comprehension. Most of the research has focused on the precocious word decoding skills and, relatively few studies have examined the nature and extent of the comprehension diffulties these children experience. Specifically, the role o f prior knowledge or schema, metacognitive monitoring and strategy use, and interest on comprehension wi l l be examined. Given the limited research in these areas on hyperlexic children, additional research in the area o f schema involving both normal and learning disabled children wi l l be reviewed. Similarly, due to the lack of research in the area of hyperlexia and metacognition, research involving normal and learning disabled children wi l l also be examined. Hyperlexia Definitions of Hyperlexia Although most researchers agree on the major characteristics of hyperlexia, not all clinicians initially agreed upon the definition. The Silberberg and Silberberg (1967) study was truly the first definition of this complex and perplexing disorder. These authors observed and examined over 20 individual cases o f children with "specific word recognition skills" (Silberberg & Silberberg, 1967, p.41). The Silberberg's reported a wide range of intellectual functioning in these children, from "mentally defective to bright normal" (Silberberg and Silberberg, 1967, p.41). The authors noted similarities in these children, such that their ability to recognize words was Comprehension in Hyperlexia 17 significantly higher than their comprehension of the material read or their verbal functioning level. The authors viewed hyperlexia as appearing on a continuum o f reading ability. On one end o f the continuum children are labelled as dyslexic, at the other end, hyperlexic. They additionally suggested that hyperlexia could be a physiological variant that manifested itself in a specific talent which was completely independent of cognitive functioning and IQ. Three main criteria were outlined for the definition of hyperlexia: (a) a discrepancy between word recognition and comprehension of the material such that children with hyperlexia can read words but do not comprehend what they read, (b) a coexistence of a cognitive and developmental delay, and (c) early age of onset. Another variation of the definition comes from Huttenlocher and Huttenlocher (1972). This study reported on three children, ranging in ages from 4.11 to 5.2 years, and evaluated their ability to carry out two- and three-part instructions. These instructions were given in both oral and written forms. Each o f the three children correctly answered only 14, 11, and 7 out o f 20 spoken commands. For the written commands, one child failed to respond, and the other two scored 11 and 13 out of 20 commands. In conclusion, the authors proposed that hyperlexia is a syndrome that should be placed in "a group with other specific cerebral disorders such as developmental aphasias and dyslexia" (Huttenlocher & Huttenlocher, 1972, p. 1112). Elliot and Needleman (1976) proposed a somewhat different view of hyperlexia. They suggested hyperlexia be defined as, "a remarkably accelerated ability to recognize written words, which may or may not occur along with truly pathological conditions" (Elliot & Needleman, 1976, p. 340). These authors were the first to view hyperlexia in terms of giftedness rather than a disorder. They presented a case study of a mute, hyperlexic girl at the age of 5.8 years. A n informal assessment of her abilities determined that she could: (a) identify a label, (b) correctly identify objects according to category, usage or other attributes, (c) match pictures to written labels, (d) respond correctly 100% of the time to one and two-part commands given both orally and written, and (e) answer questions, given both orally and in writing, about a book she was reading and point to the correct answer. The authors felt the evidence in this study led to the conclusion that Comprehension in Hyperlexia 18 hyperlexia is not a disorder, but rather a manifestation of a hyper accelerated cognitive ability. Healy et al. (1982) conducted a study of hyperlexic children in order to have a more complete description of their cognitive, linguistic, and reading abilities. Twelve children participated in the study who all had early interest in words, onset of word recognition before the age of five years, linguistic and cognitive disorders, and a level of word recognition that was significantly above their cognitive or linguistic abilities. Each child was tested individually, and received multiple tests of cognition, language, and reading competence. To investigate i f their abilities were specific to auditory or visual modalities, several tests were given in both oral and written formats. Additionally, the families of these children were interviewed to assess possible commonality in their development. Subsequent to this investigation, Healy et al. raised several issues regarding the precise definition of hyperlexia, particularly concerning the hallmark characteristics of this condition. They concluded that the term hyperlexia should: . . . be reserved for reference to the syndrome characterized by spontaneous and intense early interest in letters and words which results in the development of extensive word recognition prior to age 5, coupled with significantly disordered language and cognitive development (Healy et al., 1982, p. 22). Healy et al.'s construct of hyperlexia is much more concrete and specific in terms o f its onset as compared to the original definition by the Silberbergs. Much of the literature on hyperlexia since this study uses this definition and the criteria Healy and her colleagues specified, as well as the Silberberg's definition. Characteristics of Children with Hyperlexia Hyperlexia is a highly variant disorder consisting of a multitude of characteristics, all o f which could have an effect on comprehension abilities. Past studies have consistently reported precocious word-reading combined with poor comprehension, poor language, multiple behavioral problems, and undeveloped social skills in children with hyperlexia. This disorder does not usually occur on its own, but rather with a variety of concomitant developmental disorders including autism, pervasive developmental disorders, attention-deficit-hyperactivity disorder and learning Comprehension in Hyperlexia 19 disabilities (Aram & Healy, 1988). This section of the chapter wi l l review the literature with regard to particular characteristics and traits that are associated with the disorder known as hyperlexia and consider the implications for comprehension in these children. Cognition and Memory. Many studies on hyperlexia have examined a variety o f cognitive attributes of children whose reading is characterized by it (Elliot & Needleman, 1976; Healy et al., 1982; Huttenlocher & Huttenlocher, 1973; Mehegan & Dreifus, 1972; Richman & Kitchell , 1981; Siegel, 1984; Temple, 1990). Wi th regard to IQ, a large range has been recorded among many cases (Cobrinik, 1974; Goldberg & Rothermel, 1984; Healy et al., 1982; Richman & Kitchell , 1981), with some IQ scores being near, at, or above 100. This creates a problem with regard to the definition of hyperlexia, which includes significant cognitive delays, since individuals with IQ scores near normal should not be considered truly hyperlexic. However, many studies in the literature still reported IQ scores that were near normal. For example, Silberberg and Silberberg (1967) reported IQ's from the Wechsler Intelligence Scale for Children (WISC) (Wechsler, 1949) and the Stanford-Binet (Terman & Merr i l l , 1960) that ranged from nontestable to 126. Fontanelle and Alarcon (1982) reported Wechsler Preschool and Primary Scale of Intelligence (WPPSI) (Wechsler, 1967) and W I S C - R (Wechsler, 1974) Ful l Scale IQ scores from 57 to 118. Cobrinik (1974) was one of the few studies that remained true to the definition of hyperlexia, reporting W I S C Ful l Scale IQ scores from 47 to 71, whereas Richman and Kitchell (1981) had children whose W I S C scores fell in the normal range, between 92 and 116. Despite this wide range in IQ scores, many studies have reported a consistent cognitive profile. Several researchers have reported some definite patterns of cognition in these children. Several studies (Cobrinik, 1974; Healy et al., 1982; Huttenlocher & Huttenlocher, 1973; Richman & Kitchell , 1981) found strengths in repetition memory tasks and concrete categorization. Numerical memory for digits both backward and forward was found to be a particular strength as well as repetitive tapping sequences. However, Healy et al. (1982) reported that when no immediate sensory cues were available, and they needed to utilize abstract thinking, performance in their participants faltered. Weaknesses in memory were commonly seen in their inability to Comprehension in Hyperlexia 20 retain meaningful sentences, whereas strengths were found in recalling unrelated words. Formulating definitions for concrete nouns was extremely difficult, despite the fact that they obviously knew the term. A n example supplied by Healy et al. (1982) describes a child defining door as "1 close" (p. 12). Participants in several studies also had difficulty in organizing relationships or patterns between stimuli, both verbal and numerical (Cossu & Marshall, 1986; Healy et al., 1982; Richman & Kitchell , 1981). This was apparent in such tasks as puzzle solving, word knowledge, and conceptual grouping (Cossu & Marshall, 1986; Healy et al., 1982; Richman & Kitchell , 1981). Additionally, failure on age-appropriate Piagetian operations have been reported in many of these children (Cossu & Marshall, 1986; Healy et al., 1982; Huttenlocher & Huttenlocher, 1973). Other investigators have reported poor memory skills in their participants with hyperlexia (Cossu & Marshall, 1986; Goldberg & Rothermel, 1984). Goldberg and Rothermel reported poor memory functioning in their 8 participants. Digit span scores in these children were found to be consistently subnormal whereas the highest scores were found in block design. The authors concluded that "it is unlikely that exceptional retention abilities for verbal material play a part in hyperlexia" (Goldberg & Rothermel, 1984; p.776). Associative reasoning skills in these children has been consistently reported as very deficient (Cobrinik, 1974; Richman & Kitchell , 1981). Richman and Kitchell reported that all their participants did poorly on associative-reasoning tests of the Hiskey-Nebraska Test of Learning Aptitude ( H N T L A ) (Hiskey, 1966), particularly picture association, picture identification, and block patterns. Similar results were reported by Cobrinik (1974), who found success in his participants with rote memory tasks and most impairment with arithmetic, picture completion, picture arrangement, coding and comprehension. Since memory in these children has not been significantly studied, it is difficult to make strong statements about the link between their memory function and their comprehension deficit. Although it appears that digit span memory tasks are usually quite strong for these children, some studies still report poor memory functioning in these areas. It seems that most of the information Comprehension in Hyperlexia 21 on memory that we do have suggests that these children have adequate encoding and retrieving items from memory. However, since there have been alternative findings, it cannot yet be concluded that memory has no part in the comprehension deficit in these children. A more detailed look at memory functioning in children with hyperlexia is necessary before any conclusions in this area can be determined. Language. Many studies have assessed language abilities in young children with hyperlexia (Aram, Rose, & Horwitz, 1984; Cobrinik, 1974; Elliott & Needleman, 1976; Goldberg & Rothermel, 1984; Healy et al., 1982). However, there has been no definitive link between the language problems and the comprehension failure in these children. Although some authors suggest that these children can only comprehend as well as their language allows (Snowling & Frith, 1986), there is not yet enough information on the comprehension deficit to come to any definite conclusions. It is quite common that these children receive professional attention because of their failure to develop speech and language as expected (Aram, 1997). In Healy et al.'s (1982) study, out of the 12 subjects, four children began using a few simple words around 1-year-of-age, but then stopped talking altogether until the ages of 4 or 5 years. Additionally, one of the children in Healy's study developed single words at 1-year-of-age but did not begin to use sentences until 4.6-years-of-age. In several studies children with hyperlexia have demonstrated word recognition in reading either prior to or concurrent with talking (Aram et al., 1984; Cobrinik, 1974; Elliott & Needleman, 1976; Goldberg & Rothermel, 1984; Healy et al., 1982). Linguistic functioning in children with hyperlexia has been described in terms of a severe impairment of expressive language (Aram et al., 1984; Healy et al., 1984; Huttenlocher & Huttenlocher, 1973; Mehegan & Dreifus, 1972). In a study that examined 12 children with hyperlexia, Mehegan and Dreifus (1972) describe the extent of their language problems: . . . they manifested an unusual and premature talent in reading against a background of generalized failure of development, or marked impairment, of other language functions. For the most part, spontaneous speech was little in evidence, and in only three could it Comprehension in Hyperlexia 22 be said to be truly present. In only one o f these three could it be used in meaningful conversation. The other two showed a senseless and ill-directed logorrhea but would occasionally respond with single words. A l l three, however, betrayed a primary dysarthria, generally lingual in type and echolalia was a prominent characteristic (p. 1106). The children in this study were 11 boys and 1 girl, and ranged in ages from 5 - 9 years. A l l but one of these children, upon psychometric evaluation, was found to be in the moderate to severely . retarded range (Mehegan & Dreifus, 1972). Even more surprising was that in spite of their language dysfunction they all had an extreme talent and ability to read very well . Several authors used the Test of Language Development ( T O L D ; Newcomer & Hammil l , 1977) to assess linguistic functioning in their participants (Healy et al., 1982; Worthy & Invernizzi, 1995). The T O L D consists of various subtests that assess both syntactic and semantic skills through use of picture recognition and expressive language tasks (Newcomer & Hammil l , 1977). Both studies reported that the language performance in their participants fell below the normal mean for their age. For example, Worthy and Invernizzi (1995) reported that their participant's T O L D score was below the 50 t h percentile. Similarly, Healy et al. (1982) found, "Language performance was highly discrepant from chronological age, with most subscores falling below one standard deviation from the mean for age and none above 1 S D " (Healy et al., 1982, p. 14). In summary, there appears to remain many issues regarding memory and cognition in children with hyperlexia. Not all studies have reported consistent findings. With regard to cognition, IQ scores have been highly variant across the studies reviewed. This extreme variance in IQ may be inevitable; however, IQ scores that are close to average (around 100) do not fit the true definition of hyperlexia and thus should not have been included in these studies. In terms of memory, some investigations found strengths in digit span memory and rote memory, other studies reported very poor memory in these areas. Since there appear to be so many different findings in this area, no strong conclusions can be drawn from these studies. Comprehension in Hyperlexia 23 Neurobiology. Many studies over the years have investigated the neurological as well as biological aspects of hyperlexia (Aram, Rose, & Horwitz; Goldberg & Rothermel, 1984; Healy & Aram, 1986; Elliot & Needleman, 1976; Huttenlocher & Huttenlocher, 1973; Mehegan & Dreifus, 1976; Snowling & Frith, 1986; Tirosh & Canby, 1993). A neurobiological deficit could possibly help explain some of the comprehension deficit in these children. Yet despite hyperlexia's complex and marked language and behavioral disorganization, there has not been, to-date, any consistent neurobiological findings (Aram, 1997). However, the gender ratio for hyperlexia is consistently 7:1 males:females (Aram, 1997; Healy & Aram, 1986). Additionally, Healy and Aram (1986) reported some consistent family histories of these children in the areas of language and reading disorders. In a detailed study about how genetics influence hyperlexia, Healy and Aram (1986) examined 12 families with hyperlexic children. Their findings suggest a definitive pattern o f familial disorders in language and reading, as well as other genetic similarities. For example, left-handedness was reported in 67% of the fathers, 54% of the brothers, and 40% of the sisters. Equally as interesting was the finding that 11 of the 12 fathers reported some type of anomaly in their learning history. A n overwhelming number o f fathers, eight, had some dyslexic symptomatology. Similarly, 44% of siblings reported being diagnosed with some type o f language-learning disorder. Additionally, there were two confirmed incidences of hyperlexia in male siblings. These are all very strong indications of a certain degree o f genetic influence in hyperlexia. However, as Aram (1997) commented, ".. .only one study o f 12 children has systematically recorded familial language and reading histories" (p. 4). Many investigations have noted some common clinical presentations in their subjects, such as, mental retardation, autism, hyperkinetic syndrome, overactivity, distractibility and short attention spans (Elliot & Needleman, 1976; Huttenlocher & Huttenlocher, 1973; Mehegan & Dreifus, 1972; Siegel, 1984; Worthy & Invernizzi, 1995). For example, Mehegan and Dreifus reported eight o f their participants had significant delays in motor milestones and all children showed severe retardation. Another common trait frequently reported in children with hyperlexia Comprehension in Hyperlexia 24 is autism or autistic features, such as echolalia and rocking back and forth (Cobrinik, 1982; Huttenlocher & Huttenlocher, 1973; Mehegan & Dreifus, 1972; Siegel, 1984; Tirosh & Canby, 1993). Tirosh and Canby performed thorough physical and neurological exams on five children with hyperlexia and autism and five autistic children without hyperlexia and reported the hyperlexic group had more persistent immediate echolalia and much better splinter skills, such as drawing. Additionally, Siegel (1984) reported autistic characteristics, such.as rocking and echolalia, in a 7-year-old hyperlexic girl. Neurological examinations in many cases were consistently abnormal and findings suggested either unilateral or bilateral hemispheric dysfunction (Elliot & Needleman, 1976; Mehegan & Dreifus, 1972; Tirosh & Canby, 1993). Neurological examination in the Tirosh and Canby study revealed a better status for the hyperlexic group. However, all ten of the children had abnormal exams. Additionally, one of the children with hyperlexia showed a generalized nonspecific slow pattern with no lateralization on his E E C Conversely, some studies have reported normal neurological examinations. For example, Huttenlocher and.Huttenlocher (1973) reported that all their participants had normal electroencephalographic exams and normal skull x-rays. In examining all of these studies it appears there are not yet any consistent neurobiological findings. However, there are many findings that may have an impact on these children's cognition and their comprehension abilities. The multiple findings of hyperactivity and short attention span (Elliot & Needleman, 1976; Healy et al., 1982; Mehegan & Dreifus, 1972; Siegel, 1984) could help explain the inability to focus attention which makes comprehension of reading materials extremely difficult. Additionally, the various developmental delays, particularly mental retardation and autistic-like features that are found in so many of these children could partially explain some cognitive and comprehension problems. Reading Strategies at the Word Level. Different children may use various routes to read words. Some children prefer to read phonologically whereas some prefer to read orthographically. Learning about the strategies that children with hyperlexia use to read words may give some Comprehension in Hyperlexia 25 insight into their reading comprehension deficit. A couple of authors have suggested a direct reading route in children with hyperlexia (Goldberg & Rothermel, 1984; Seymour & Evans, 1992). Both these investigations tested their participants on various formal and informal language and reading tests. Goldberg and Rothermel (1984) reported that when reading aloud, all 8 participants read high imagery words better than low imagery words and high frequency words better than low frequency words. Additionally, all the children were able to read function words with more than 85% accuracy. The authors concluded that this was accomplished via the direct reading route since the children tended towards more visual errors in short words, which suggests attempts to access the words directly, not via phonological routes. Cobrinik (1982) investigated the ability of hyperlexic children to identify visually degraded words. Nine hyperlexic boys, between the ages of 9- through 13-years, were given an incomplete word task in which the cue characteristics on individual letters were made ambiguous. The task consisted of fourteen familiar words between 7 and 9 letters in length that had partially deleted portions of the letters in each word. The task was estimated in difficulty to be at about the 3 r d - to 4 t h-grade level. When normal children and adults were tested on this task, they found it to be very difficult. The author found that the hyperlexic children were far superior in deciphering the incomplete words when compared to the normal control group. These results suggest that hyperlexics use a configurational approach to identifying words whereas normal children use a serial and analytic process. Cobrinik concluded that hyperlexic reading is ideographic. Other studies have reported a phonological reading route (Healy et al., 1982; Siegel, 1984). These authors found that when children with hyperlexia were tested with non-word reading, they excelled compared to normal readers. In 1984, Siegel described the participant in her study, who despite delayed cognitive, language and motor development was able to read at a 2.8 grade level on the Wide Range Achievement Test ( W R A T ) (Jastak & Jastak, 1978). When given the Goldman, Fristoe, and Woodcock (GFW) Sound Symbol test (Goldman, Fritoe, & Woodcock, 1974), which tests the reading of non-words, she scored in the 61 s t percentile. Additionally, out of Comprehension in Hyperlexia 26 36 regular words and 36 irregular words, she read 33 and 31 respectively. From these findings, Siegel concluded that children with hyperlexia, "indicate a reversal of a normal process because of their use of grapheme-phoneme correspondence rules in reading is clearly more advanced than their semantic or syntactic processing" (p. 583). It appears that the findings in this area vary as to which reading route is preferred by children with hyperlexia. Many studies have found phonologic strategies to be preferred (Siegel, 1984) whereas others have reported orthographic strategies to be the route used (Cobrinik, 1982) and some suggest multiple routes (Goldberg & Rothermel, 1984). Whatever route they use to read the words, it appears that these children are definitely reading without any meaning. Comprehension in Hyperlexia Over the years, various studies have examined comprehension in hyperlexic children (Aram, Rose, & Horwitz, 1984; Elliot & Needleman, 1981; Goldberg & Rothermel, 1984; Healy et al., 1982; Huttenlocher & Huttenlocher, 1973; Patti & Lupinetti, 1992; Siegel, 1984; Silberberg & Silberberg, 1968, 1971; Snowling & Frith, 1986; Temple, 1990; Welsh, Pennington, & Rogers, 1987; Worthy & Invernizzi, 1995). Most o f these studies have focused on multiple aspects of hyperlexia (Aram, Rose, & Horwitz, 1984; Elliot & Needleman, 1981; Healy et al., 1982; Patti & Lupinetti, 1992; Siegel, 1984; Silberberg & Silberberg, 1968, 1971). Few studies have strictly examined comprehension in hyperlexic children (Snowling & Frith, 1986; Temple, 1990; Welsh, Pennington, & Rogers, 1987; Worthy & Invernizzi, 1995). Sti l l fewer studies have examined multiple subjects with hyperlexia regarding their comprehension abilities (Snowling & Frith, 1986). Despite this apparent lack of research in this area, it is continually documented as being a large component within hyperlexia (Aram, 1997). Comprehension of Spoken and Written Language Comprehension in children with hyperlexia has been examined in terms of both spoken and written formats. Additionally, there have been several studies that have compared the two modalities of aural and written comprehension (Healy et al., 1982; Huttenlocher & Huttenlocher, 1973; Temple, 1990). In this section of the chapter the literature in the area o f language and Comprehension in Hyperlexia 27 reading comprehension wi l l be thoroughly examined. Auditory Comprehension. Several investigators have examined auditory comprehension in children with hyperlexia and had similar conclusions. In tests where 2 and 3-part instructions that were read aloud to them, most of the children did very poorly. A n example of the directions given would be, "Clap your hands. Then sit on the table." (Huttenlocher & Huttenlocher, 1973, p. 1110). Upon comparing these results with normal children of 4 '/z-years, the authors concluded that, "The basic language defect appeared to be in the association between speech symbols and meaning." (Huttenlocher & Huttenlocher, 1973, p. 1115). Similarly, 8 hyperlexic participants in the Goldberg and Rothermel (1984) study were asked to perform some simple oral commands such as, "Touch the black square after you touch the circle" (p. 769). This test was given to assess comprehension of both syntax and sequencing. Out of 16 items on the test, the participants could only respond correctly to less than half of them. A more recent study by Glosser, Friedman and Roeltgen (1996) examined the patterns of development of reading and spelling in a hyperlexic child. This longitudinal case study involved one child, L A , who at 6 years old was diagnosed with hyperlexia. L A had a full-scale IQ score of 51 on the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) (Wechsler, 1967). A detailed analysis of both oral and written language abilities was completed to better characterize the reading abilities. Among the battery of tests given, a single word auditory comprehension test was administered, using the Peabody Picture and Vocabulary Test - Revised (PPVT-R) (Dunn & Dunn, 1981). From the data collected the authors concluded that L A showed severe impairments in semantic linguistic processing across all modalities o f stimulus presentation and response. Additionally, substantial deficits were found in comprehension of both oral and written words (Glosser etal., 1996). In conclusion, it appears that there is a deficit in auditory comprehension in children with hyperlexia. A finding in these children was summed up by Goldberg and Rothermel (1984) who commented "the most striking facet of the hyperlexic children's performance on metalinguistic reading tasks was their inability to grasp instructions. More than half the sample could not Comprehension in Hyperlexia 28 comprehend the tasks' instructions" (p. 777). Auditory versus Written Comprehension. There have been several studies that have examined and contrasted auditory and written forms of comprehension in children with hyperlexia (Aram et al., 1984; Glosser et al., 1996; Goodman, 1972; Healy et al., 1982; Huttenlocher & Huttenlocher, 1973; Snowling & Frith, 1986; Temple, 1990). Only one of these studies (Temple) found a nonsignificant trend for written comprehension being better in these children as compared to aural comprehension. In this section, I w i l l review these various studies and examine the evidence that has thus far been put forward. Researchers have reported no differences in comprehension found between auditory and written modalities. A l l the participants in both the Healy et al. (1982) study and the Snowling and Frith (1986) study had close to equal scores for auditory and written comprehension tests. The participants in the Healy et al. study were tested for performance in both auditory and visual input modalities on the T O L D and the McCarthy Scales in the auditory mode and not a week later were they tested in the written format. The authors found differences between the modalities on the grammatic completion subtests in which the mean score for auditory modality was higher than for visual modality in 3 subtests of the T O L D (opposite analogies, oral vocabulary, oral commissions). The participants scored lower on auditory and higher on visual on the grammatic understanding, grammatic completion and oral vocabulary subtests o f the McCarthy instrument. Despite these small differences, the authors still concluded that hyperlexia consisted of a generalized cognitive deficit with no preference in modality processing. Similarly, Snowling and Frith (1986) compared auditory and written comprehension in their participants on the Bishop's (1983) Test for the Reception o f Grammar (T.R.O.G.) (Bishop, 1983). The authors reported that the mean correct number of responses were nearly the same in both written and spoken formats (24/40 and 23/40 respectively). Conversely, Temple (1990) found a nonsignificant trend towards a preference for written over auditory comprehension. The author assessed her case study, M S , on both auditory and written comprehension across multiple reading tests. First, M S was given 100 stimuli from the Comprehension in Hyperlexia 29 P P V T in three different formats, the first was read aloud, the second was presented in written form for reading and matching with the correct stimulus, and the third was presented in written form for reading aloud. Both the T . R . O . G . and the Neale Anlaysis of Reading were also given in the same three formats. Temple found "in comparison to reading accuracy both auditory and reading comprehension are poor, with auditory comprehension at the lowest level" (p 304). For all three o f the assessments given, M S had higher comprehension scores for written material rather than the material that was read to her. In conclusion, it appears that the comprehension deficit in children with hyperlexia is not specific to only one modality. Although most findings appear to indicate that the comprehension deficit is equal in both written and aural modalities, it is still somewhat questionable i f reading comprehension is not slightly better than aural comprehension. The present study wi l l focus on reading comprehension in these children and w i l l not examine auditory comprehension separately. Reading Comprehension. Many studies have assessed the comprehension o f the written word at the word, sentence, or paragraph level (Aram et al., 1984; Glosser et al., 1996; Goldberg & Rothermel, 1984; Healy et al., 1982; Richman & Kitchel l , 1981; Seymour & Evans, 1992; Siegel, 1984; Snowling & Frith, 1986; Temple, 1990). Most of these investigators agree that hyperlexia consists of both incredible decoding abilities and extreme deficits in comprehension. However, there is not yet a consensus as to how much of the deficit is semantic, syntactic, or discourse-based (Aram, 1997). The lack o f unity in this arena may be due to two factors: (a) the limited number of studies in the area of language and reading comprehension in hyperlexia, or (b) that different children may have differing degrees o f impairment (Aram, 1997). Most of these studies agree that both semantic and syntactic comprehension is severely defective (Aram et al., 1984; Healy et al., 1982; Siegel, 1984; Snowling & Frith, 1986). There are a few who suggest that the comprehension deficits are more limited (Seymour & Evans, 1992; Temple, 1990). This section wi l l examine the studies that have attempted to understand the comprehension deficit in hyperlexia. Single-word and sentence comprehension has frequently been reported as sufficient in children with hyperlexia (Goldberg & Rothermel, 1984; Welsh et al., 1987). Goldberg and \ Comprehension in Hyperlexia 30 Rothermel (1984) used the Peabody Individual Achievement Test (PIAT, Dunn & Markwardt, 1970) to assess comprehension. The reading comprehension subtest asks the subject to match the appropriate picture one o f four, to the sentence he/she reads aloud. The authors reported that the children were successful at understanding single words and sentences. The severity of the comprehension deficit in children with hyperlexia is commonly documented in many investigations. Several authors have reported similar findings in regard to paragraph and story comprehension (Goldberg & Rothermel, 1984; Healy et al., 1982; Siegel, 1984; Worthy & Invernizzi, 1995). On various tests of paragraph comprehension, children with hyperlexia have been recorded as being unable to answer many i f not most o f the questions asked of them (Goldberg & Rothermel, 1984; Worthy & Invernizzi, 1995). Goldberg and Rothermel (1984) used the Durrell Analysis of Reading Difficulty (Durrell,T955) to test paragraph comprehension in their participants. First the children read the passage out loud, then listened and answered the questions asked by the investigator regarding the passage. The children were allowed to look back into the passages to retrieve the answers. The passage each subject read was at the grade level indicated by their P I A T reading recognition score. After reading the paragraphs, the children were presented questions simultaneously in written and verbal form. The participants, on average, could only respond to 33% of the questions. Findings of a similar comprehension deficit in children with hyperlexia have been reported in several studies (Healy et al., 1982; Siegel, 1984; Worthy & Invernizzi, 1995). For example, Worthy and Invernizzi (1995) used the comprehension subtest of the Woodcock Reading Mastery Test ( W R M T ) (Woodcock, 1973) to assess comprehension in their 14-year-old participant and reported a grade level o f 3.1 on the passage. On both the Qualitative Reading Inventory (QRI) (Leslie & Caldwell , 1990) and an informal reading inventory (IRI) constructed by the authors and their staff, which contain longer passages of text, the participant "could not recall anything or answer questions about the upper level passages that she could read fluently" (Worthy & Invernizzi, 1995; p. 594). The authors commented that, "her comprehension scores on both inventories were below the primer level" (p. 594). Worthy and Invernizzi described how the Comprehension in Hyperlexia 31 subject could not retell any of the stories nor could she discuss even the simplest stories that she read or were read to her. For example, when "R" was asked to describe a 2 n d grade book about George Washington that she was reading in class, she would often add in bits of information about other famous men and presidents. Similarly, Siegel (1984) reported that upon questioning her subject about the paragraphs she had just read, she could not answer most of the questions that were asked of her. Siegel commented, "She did not even appear to understand the questions" (p. 580). Some researchers have used altered paragraphs or reading miscue inventories to examine reading comprehension at the paragraph level more in depth. Goldberg and Rothermel (1984) used an altered paragraph task in which several grade 4 level paragraphs were changed in terms o f period placement and word appearance. Periods were added to the paragraphs as well as put into very inappropriate places. Additionally, some of the words in the paragraphs were altered in their appearance such that a symbol, e.g., "+" was added to the word. The participants received both altered and unaltered paragraph formats. The authors reported that the altering of punctuation did not produce significant differences in reading the altered or unaltered formats. It appeared that the children were using semantics to process the sentence. The authors concluded that hyperlexic children appear to process semantically during reading and that they have a lexicon, although it is a simple one. Other authors have used the Reading Miscue Inventory (RMI) (Goodman & Burke, 1972) to examine reading patterns. For example, Healy et al. (1982) found the pattern of self-correction for hyperlexic children was found to be very different from normal readers. When using the R M I the child reads a passage and then is asked to retell the story. Although the children with hyperlexia did self-correct errors, they did so immediately after the mispronunciation rather than at the end of the phrase. The authors proposed that: "this fact, as well as the finding that the overwhelming majority of errors occurred from close graphic and sound correspondences, suggests that these readers are uniquely responsive to the perceptual attributes o f the text rather than to the meaning"(p. 17). Comprehension in Hyperlexia 32 It appeared that these children did not realize (like normal readers) that they were not "making sense" (Healy et al., 1982, p. 17). Finally, Healy et al. sum up the totality of the comprehension deficit in these children: The most remarkable finding, however, was that not one of these youngsters, having correctly pronounced almost all o f the words in the text, could retell, summarize, or even relate any information from the story. Isolated words and phrases were sometimes recalled, and often the title was repeated in response to the question, "What was the story about?" Most frequently, responses were unrelated to the textual meaning, and, on at least one occasion, drawn from a story read 1 month earlier (p. 17). Beneath the obvious decoding talent these children possess, lies a serious deficit in their comprehension abilities. Snowling and Frith (1986) were the only investigators who studied comprehension in both autistic and nonautistic hyperlexic children and used mentally handicapped children as a comparison group. To assess children's use of context as an aid to comprehension, the authors used 20 sentences with homographs in both their most frequent and less frequent forms. The children were first given a pretest of words to read that included several homographs to see i f they would use the most frequent form of these words. A l l the subjects were told during the pretesting that each word had two possible meanings and pronunciations. Despite the pretest and the warning the subjects continually used the high frequency homographs as compared to the less frequent homographs. The authors concluded that these children with hyperlexia do not use context when reading in order to disambiguate homographs. In a second experiment, the same investigators examined i f these children understood larger chunks of meaning by using story-appropriate and story-nonappropriate words. The children were given a story to read in a modified cloze procedure, in which at varying intervals in the text a blank was given along with three possible words to use for filling-in-the-blank. The children were given instructions to fi l l in the blank by picking the word that made the most sense. Practice was given until the children understood the task. The authors found that both autistic and Comprehension in Hyperlexia 33 nonautistic hyperlexic children did not distinguish between story-appropriate (appropriate for both sentence and story context) and sentence-appropriate words (appropriate only for the sentence context but not for story context). Snowling and Frith (1986) concluded that these children were not using the story context to choose the most appropriate word and therefore did not comprehend the story. In summary, reading comprehension in children with hyperlexia has frequently been found to exist for single words and short concrete sentences. However, when reading paragraphs or short stories and asked to recall them or answer questions about them, these children fail. Instead, when children with hyperlexia are asked to recall a short passage they often relay only concrete, intermittent pieces of the text accompanied with other unrelated, non-text material. With regard to schema or prior knowledge, it thus far appears that children with hyperlexia do not use their own prior knowledge to answer questions about text. Additionally, these children do not correct or point out rather obvious errors that have been embedded in text. A possible conclusion in the area of reading comprehension comes from Snowling and Frith (1986). These authors reported that the ability to understand, in their participants with hyperlexia, was at a 5-year-level (on average) which was vastly different compared to their ability to decode, which was at a 9-year-level (on average). From this finding, Snowling and Frith suggested that "the results raise the possibility that comprehension wi l l simply be as good as verbal ability allows it to be" (p. 396). However, these same authors later concluded, "Hyperlexia implies a particular failure to comprehend in terms of larger units of meaning, not accounted for by poor general language skills or poor word knowledge" (p.410). It is precisely this type o f confusion that underlies the disorder of hyperlexia. The need for more clarity regarding this comprehension deficit is evident, "the question of why the true hyperlexic children's reading comprehension is so poor remains to be answered" (p. 410). It is for this reason that my study wi l l further examine this perplexing comprehension deficit in these children. Comprehension in Hyperlexia 34 Cognition and Memory Long-term Memory and Schema Definitions of Schema. Schema have been defined as knowledge structures stored in memory (Anderson & Pearson, 1984; Rumelhart, 1980). Anderson et al. (1977) suggest that schema is knowledge brought to the text as opposed to in-text knowledge. Furthermore, schema are more important for understanding text than in-text knowledge (Anderson et al., 1977). Boblett (1932) refers to schema as "an active organization of past reactions or past experience" (p. 201). He saw schemata as an active process of organizing information and experiences (Boblett, 1932). Rumelhart (1980) describes schema as "the building blocks of cognition" (p. 33). He believes that all thought processing stems from schema which help organize and interpret information (Rumelhart, 1980). He also describes schema as structures that represent various basic concepts that are stored in memory (Rumelhart, 1980). According to Rumelhart and Ortony (1977) there are four major characteristics of schema: (a) they have variables, (b) they can be embeded within one another, (c) they represent knowledge, and (d) they are found at all levels of abstraction. Furthermore, schema can represent all levels of knowledge; therefore, some schema are quite large, while others are small. Anderson and Pearson (1984) have explained how schema aid comprehension. According to these authors, it is the interaction between new and old information that creates comprehension. The new information is accommodated for in memory by the existing stored knowledge and they combine to create a storage place for both new and old knowledge on a topic. Anderson et al. (1977) described three types of schema that represent objects, events, and actions. A n example of a schema for objects is a dog or a house, something concrete. Schema for events have also been called scripts and describe everyday events, such as eating in a restaurant (Anderson et al., 1977). Anderson (1994) defines schema as "organized knowledge of the world" (p. 469). He explains that schema "provide much of the basis for comprehending, learning, and remembering the ideas in stories and texts" (Anderson, 1994; p. 469). Anderson (1994) gives the example, "The big number 37 smashed the ball over the fence" (p. 470). Most individuals with some knowledge Comprehension in Hyperlexia 35 of baseball would find this statement easy to understand. This baseball player, number 37, really hit the ball hard and it flew over the fence. However, for a person with no prior knowledge o f baseball, this sentence would make little sense to them. Some people might imagine a big, metal number 37 being hit by the baseball causing it to break and fall off the fence. It is quite incredible how our schema or lack of it w i l l completely change the meaning of a passage. How Schemata Aid in Comprehension. Schema or prior knowledge has consistently been reported to aid in the comprehension process (Anderson et al., 1977; Anderson et al.,T978; Bransford & Johnson, 1973; Langer & Nicol ich, 1981). In this section of the chapter, literature in the area of schema or prior knowledge and its effects on the comprehension of text w i l l be reviewed. Schema have several functions (Anderson, 1994; Anderson, 1978; Anderson & Pichert, 1978) that include: (a) providing ideational scaffolding for assimilating text information; (b) facilitating selective allocation of attention; (c) enabling inferential elaboration; (d) allowing orderly searches of memory; (e) facilitating editing and summarizing; and (f) permitting inferential reconstruction (Anderson, 1994; p. 473-474). These various functions come together to aid the individual in comprehension of a text. In order to comprehend material, the reader needs to assimilate the information, give their full attention, be able to make inferences about the text, conduct memory searches, summarize and reconstruct the ideas in order to make sense o f the material. A study by Bransford and Johnson (1973) clearly demonstrated the strong role schema play in comprehension of text. In this study, some of the participants received a title to a very vague reading passage prior to reading it, other participants received it after reading it, and the remainder of subjects did not get a title at all. Bransford and Johnson reported significant improvement in comprehension in the participants who received the title of the passage prior to reading it since it activated their schema for the topic. Conversely, the participants who received the title after reading the passage did not gain anything in comprehension as compared to the no title group. The authors concluded that prior knowledge is important in aiding in comprehension and the relevant Comprehension in Hyperlexia 36 knowledge must be activated prior to processing for comprehension to occur. Schema is also referred to as prior knowledge. In 1981, Langer and Nicol ich conducted a study that examined the relationship between prior knowledge and comprehension. These authors examined 36 high school students enrolled in an advanced English literature course. The participants were tested on two passages for the amount of prior knowledge they had about each topic. One passage was about schizophrenia and the other about parakeets. Prior to reading each passage the students were asked to "write anything that comes to mind when you hear the word (or phrase)..." (Langer & Nicol ich , 1981; p.374). Through using this task the examiners were trying to access any knowledge about these topics from the student's memory. After reading each passage the students were told to write down everything they could remember about the passage. Factor analysis revealed that the level of prior knowledge was related to recall of the text. Correlations between prior knowledge and the principal components for each passage were .75 and .70 for schizophrenia and parakeet passages respectively. The findings suggested that the level of prior knowledge is strongly related to recall of the passage. Some studies have investigated prior knowledge in children with learning disabilities (LD) and reported the importance of prior knowledge on comprehension of text (Carr & Thompson, 1996; Montague, Maddux, & Dereshiwsky, 1990; Snider, 1989). Despite the fact that many studies with L D students report that they have less prior knowledge as compared to normal readers, the studies have found that L D students do use these bits of prior knowledge to aid them in the comprehension of text. Additionally, many researchers have been successful in helping students with L D to activate prior knowledge and this has had a positive effect on their comprehension of the text (Carr & Thompson, 1996; Snider, 1989). In 1989, Snider investigated the effects of levels of prior knowledge and types of reading passages on comprehension in junior-high students with L D . The author found that students with L D who were directly taught the knowledge did better than the control group on the comprehension questions (Snider, 1989). Additionally, Snider (1989) concluded that declarative knowledge had a positive and significant impact on comprehension questions of all types. Comprehension in Hyperlexia 37 Similarly, Carr and Thompson (1996) investigated the effects of prior knowledge and schema activation on reading comprehension in L D children. Forty-eight children in elementary and middle school took part in this study. O f the 48 participants, 16 were L D students in grades 7 and 8, 16 were non-LD students in grade 8 and the remaining 16 were non-LD students in grade 5. The non-LD children were both same age peers and reading level peers o f the L D children. The participants were individually administered a pre-test that assessed decoding ability and prior knowledge of the test topics. Each participant was given 16 passages, over two sessions, to read and then comprehension questions to answer. After reading the passage silently, the examiner asked the participant to "Tel l me what you remember about this story" (Carr & Thompson, 1996; p. 52). After the student recalled the story, the examiner read five inferential comprehension questions to the student who answered with an oral response. O f the 16 passages, 8 were familiar topics and 8 were not familiar topics. In the first session the students were expected to activate prior knowledge spontaneously while in the second session the experimenter prompted the student. Prior knowledge was activated by the examiner asking each student prior to the reading of each passage "Tel l me what you know about..." (Carr & Thompson, 1996; p. 53). The authors reported that all three groups of children benefited from the activation of prior knowledge by the examiner, especially when the passages were unfamiliar. However, the L D students benefited most by the activation of prior knowledge. Finally, Snowling and Frith (1986) have studied prior knowledge in children with hyperlexia. These authors examined the use o f general knowledge and text knowledge in groups of autistic and nonautistic hyperlexic children, normal children and mentally handicapped children. They asked 32 comprehension questions, 16 in regard to a "Beaver story" and the other 16 about a "Hedgehog story." The questions were of two types: (a) questions requiring the subject to remember factual detail from the text, and (b) questions requiring general knowledge about these topics. In the previous study, the investigators had informal discussions with the children about animals and explained about a beaver and hedgehog for those who did not know. When examining the results, the authors found that these retarded autistic and nonautistic hyperlexic children did Comprehension in Hyperlexia 38 equally poorly on both types of questions. After comparing this group of children to 10 young normal readers, the authors concluded that the autistic and nonautistic hyperlexic children processed the text superficially and remembered very few isolated details. This section of the chapter has reviewed the literature in the area of schema or prior knowledge and its effects on comprehension. In most situations prior knowledge is found to enhance comprehension of text. The strongest conclusion that comes from this section is that children who have or use schema when reading text have improved comprehension for that text. These children combine both prior knowledge and text knowledge in order to have a complete understanding of what they are reading. Many of the investigations above have strongly shown that without prior knowledge o f the topic they are reading, children do poorly on comprehension questions compared to those children who have a strong prior knowledge of the topic. Even studies on children with L D have shown that prior knowledge for a topic increases their reading comprehension for that topic. The studies above suggest that activation of that prior knowledge in children with L D strongly increased their comprehension for the task. This strongly suggests that children with hyperlexia may also have some schemata for particular topics that when activated would increase their comprehension of the topic. This is a question that w i l l be addressed in my research study. Metacognition and Strategy Use Definitions. Metacognition is defined as cognition about cognition. It involves thinking about the various processes o f cognition such as perceiving, understanding, and remembering (Garner, 1987). First, metacognition refers to awareness of one's cognitive processes, your strengths and weakness, and how well the fit is between the cognitive task and your cognitive resources. Second, it refers to self-regulation, or regulating your own comprehension (Flavell, 1978). Metacognition has been defined by Flavell (1979) as "knowledge and cognition about cognitive phenomena" (p. 906). He further described that metacognition plays important roles in oral communication, oral and reading comprehension, writing, language acquisition, memory and attention. Metacognition is broken into various components, such as metacognitive knowledge, Comprehension in Hyperlexia 39 metacognitive experiences, and strategies (Flavell, 1981). Each of these processes is a subordinate component of metacognition. Metacognitive knowledge is knowledge about people, tasks and strategies. Strategies are the employment o f cognitive and metacognitive resources in order to remedy any possible cognitive failure (Garner, 1994). In their 1982 article, Paris and Lindauer described the concept of metacognition as "knowledge we have about people's mental states, abilities, and processes of behavioral regulation. It includes our understanding of task goals and the strategies that are useful for accomplishing different purposes" (p. 334). Metacognition involves information that is conscious and deliberate. For example, "I had better reread this paragraph to understand the author's point" (Paris & Lindauer, 1982; p. 334). This example clearly shows the executive control process o f metacognition and the conscious strategizing that is a part of it. Executive control processes direct activities at various stages in processing and make certain that the system functions as a whole throughout all the processing stages (Garner, 1987). In her 1977 article, Brown stressed the importance o f executive control processes "in the domain o f deliberate learning and problem-solving situations, conscious executive control o f the routines available to the system is the essence of intelligent activity" (p. 4). Additionally, Brown (1980) investigated metacognition during reading tasks and how using metacognitive strategies can improve learning and comprehension. The author discussed several important strategic components for effective reading: (a) making predictions about the text, (b) checking predictions, (c) monitoring comprehension, (d) reality testing, and (e) controlling attempts to learn a task (Brown, 1980). These strategies have been vastly researched over the years and have brought much insight into important differences between good and poor readers metacognitive abilities (Garner, 1980; Garner & Reis, 1981; Garner & Taylor, 1982; Paris & Myers, 1981; Wong & Jones, 1982). Strategies such as the ones described above often differ with the age of the individual. Unti l recently, researchers assumed that older children and adults are more likely to use self-regulatory strategies when learning as compared to younger children (Baker & Brown, 1984). However, recent studies have shown that preschool children and children Comprehension in Hyperlexia 40 in grade one do use metacognitive strategies when learning (Neuman, 1997; Perry, 1998; Turner, 1995; Winne, 1997). Skilled readers use strategies, such as comprehension monitoring, when reading, to ensure comprehension of the text. Paris, Lipson and Wixon (1983) discussed metacognition in terms of these strategies for reading. According to these authors knowledge about reading is categorized in three ways: (a) declarative knowledge, (b) procedural knowledge, and (c) conditional knowledge. Declarative knowledge is understanding "what" factors, for example, help reading comprehension. This type of knowledge includes beliefs about one's abilities and beliefs about the task. A s an example, a child may see himself as a slow reader and may consider a particular reading task difficult or boring. Procedural knowledge is knowing "how." For example, knowing how to use particular skills to support learning. This type o f knowledge includes knowing how to perform certain strategies, such as skimming, scanning, and summarizing. Finally, conditional knowledge is knowing "when" and "why." For example, knowing when to use certain strategies and why they are affective. Conditional knowledge is linked to motivation (e.g., Knowing why I should use a particular strategy - because it w i l l improve my score on a test - motivating me to use it). Paris et al. suggest that "strategies combine components o f both skill and w i l l " (p. 798). This means that motivational aspects of metacognition are important for the child to accomplish a goal. Children need to have intent to pursue knowledge and to behave in a particular manner. Motivational aspects of learning are important incentives that give energy to the system that then direct the children's behavior. A model of metacognition for L D children comes from Pressley, Borkowski, and O'Sullivan (1985). This model consists of several components: specific strategy knowledge, general strategy knowledge, relational strategy knowledge, and metamemory acquisition procedures. Specific strategy knowledge includes an understanding of (a) a strategy's goals and objectives, (b) the appropriate tasks to use this strategy for, (c) its range of applicability, (d) the learning gains expected from consistent use of the strategy, (e) the amount of effort associated with it's use, and (f) i f the strategy is enjoyable or difficult to use (Borkowski, Johnston, & Reid, 1987). Comprehension in Hyperlexia 41 Specific strategy knowledge is needed for efficient strategy use. Once the child has a number o f strategies and knowledge of how to use them, they can make informed decisions on which strategy is best for which task. Specific strategy knowledge is not stagnant, it changes over time and accumulates slowly as the child matures. General strategy knowledge is a child's understanding that it requires effort to apply these strategies and that this effort usually leads to success (Borkowski et al., 1987). Motivation is linked to general strategy knowledge because as the child experiences the success from strategy use, they are more motivated to learn new and challenging tasks. Relational strategy knowledge is a type of classification system that highlights various aspects o f specific strategies and helps the child to select which strategy would be the best for the task at hand (Pressley et al., 1985). According to Borkowski et al., there is not much known about relational strategy knowledge in normal or L D children. The final component of the model, metamemory acquisition procedures ( M A P S ) , are the higher order components of metamemory and are important in the self-regulation of human behavior (Brown, 1978). They produce self-control in young and L D children (Pressley et al., 1985). Finally, the authors of the model suggest that all these characteristics are required for a normal, L D , or mentally retarded child to be able to transfer strategies successfully (Pressley et al., 1985). In summary, metacognition is an executive control process that consists o f higher order thinking. It is described best as thinking about our thinking. Metacognition consists of person, task, and strategy variables that describe what the child knows about his/her own learning abilities, what is known about the task (is it hard or easy?), and what strategies would be best to help him/her learn most efficiently. Finally, metacognitive thinkers are strategic learners who employ various strategies to aid their comprehension processes. How Metacognition Aids Comprehension. A s discussed above, metacognition involves knowing how and when to use certain strategies in order to be successful at a reading task. Wong (1991) explained that successful learners need to have an awareness of their own cognitive strengths and weaknesses in order to select appropriate strategies for learning. Metacognition is, Comprehension in Hyperlexia 42 therefore, necessary not only to effectively choose the appropriate strategy, but also to deploy it and to evaluate it's usefulness (Wong, 1996). Many studies have examined comprehension in good and poor readers and have reported that good comprehenders are more likely to use comprehension monitoring strategies when they are reading than poor comprehenders (Garner, 1980; Garner & Taylor, 1982; Paris & Myers, 1981). Children who read for meaning are essentially trying to comprehend the material. However, in order to comprehend the material, children need to use metacognitive strategies, such as comprehension monitoring (Baker & Brown, 1984). Wagoner (1983) describes comprehension monitoring as "an executive function, essential for competent reading, which directs the readers cognitive process as he/she strives to make sense of the incoming information" (p. 328). Nonstrategic readers have been found to read as i f on auto-pilot and are not even aware of any comprehension difficulties (Paris, Wasik, & Turner, 1991). This section of Chapter 2 wi l l examine the literature on metacognition and learning strategies in poor and good comprehenders. Comprehension monitoring has been investigated in both good and poor readers (Garner, 1980; Garner & Taylor, 1982). In both studies, good readers and poor readers from grades 4 through 7 were given altered passages from expository texts to read. These passages were changed so that they contained several inconsistencies and no longer made any sense. After reading each section, the child had to choose i f the section was "very easy to understand", i f the section "was okay", or i f the section "was difficult to understand." Garner (1980) reported that good readers rated most of the consistent-information sections as "very easy to understand" and the inconsistent-information as either "okay" or "difficult to understand." In contrast, Garner and Taylor (1982) reported that many o f their poor comprehenders never mentioned the internal inconsistencies in the texts. Instead, the poor readers gave comments such as, "The story makes good sense" and "There is no problem with this story" (Garner & Taylor, 1982; p. 4). Similarly, Paris and Myers (1981) examined comprehension monitoring in good and poor fourth grade children. The authors examined the frequency o f monitoring in both spontaneous and directed conditions. To test for the spontaneous monitoring the authors asked the children to read Comprehension in Hyperlexia 43 two stories that contained nonsense words and nonsense phrases and their hesitations, repetitions and self-corrections were recorded. In the directed condition, the children were told that parts of the story may not make sense and that they should underline and words or phrases that they did not understand. After the children read the stories they were asked to recall as much of the story as they could. The authors prompted the recall of the story by giving the children the title o f the story. Paris and Myers (1981) found that in comparison to good readers, the poor readers did not engage in accurate monitoring as frequently and they did not evaluate information that they did not understand. Finally, the poor readers showed less accurate recall and less comprehension of the stories compared to the good readers (Paris & Myers, 1981). In a second experiment, the authors examined the strategic behaviors o f both poor and good readers with regard to obtaining meaning from difficult vocabulary words. The.participants were 14 good and poor fourth grade readers. Grade four level reading passages were transformed to contain several difficult vocabulary words that were judged to be above fourth grade readers level ("e.g., anther, papaya, meandered, menageria") (Paris & Myers, 1981; p. 14). The children were given the story and told to read it so they could remember it later. They were also given pencil, paper, and dictionary and told that they could use to help them remember the story. After the children completed the story they were given a short addition task and then they were asked to verbally recall the story. Results of the study indicated that 79% o f good readers and 29% of poor readers used a study strategy for understanding the difficult words (Paris & Myers, 1981). Additionally, six o f the good readers used the dictionary to look up difficult words whereas none of the poor readers did (Paris & Myers, 1981). Five good readers and only two poor readers used the paper and pencil to write things down to help them remember the story. Good readers were found to use more spontaneous monitoring strategies than poor readers (Paris & Myers, 1981). In terms of recall, poor readers were able to remember less of the stories compared to the good readers and showed more forgetting and less organized recall (Paris & Myers, 1981). The authors concluded that the difference in the reading performance and recall of the students was significantly correlated to the use of successful comprehension monitoring strategies (Paris & Comprehension in Hyperlexia 44 Myers, 1981). Metacognition has been assessed in various ways, including comprehension monitoring (Garner, 1980; Garner & Taylor, 1982; Paris & Myers, 1981), lookback strategies (Garner et al., 1983; Kinnunen & Vauras, 1995), and interview formats (Garner & Kraus, 1982). In a more recent investigation, Kinnunen and Vauras (1995) examined comprehension monitoring in good and poor fourth grade readers. In this study, the authors assessed the changes in reading time and the number of lookbacks children had when confronted with comprehension obstacles. The children were given six reading passages in which one of the three paragraphs was always consistent while the other two included one word (either a nonsense word, or a word that made no sense) designed to provoke comprehension monitoring. Therefore, the children read 18 paragraphs, six being consistent and 12 having comprehension obstacles embedded in them. During the experiment the participants were instructed to read and learn the passages in the way they normally do when they are doing their homework and that they would be asked questions about the passages after their studying. The authors reported that high achieving readers read the passages faster than low achieving readers and used more comprehension monitoring strategies as compared to the low achieving readers. A s expected the good readers scored higher on the comprehension questions as compared to the poor readers. The authors concluded that both level o f monitoring and level of comprehension are related, which explains the differences between good and poor comprehenders (Kinnunen & Vauras, 1995). Metacognition has also been examined in children with Learning Disabilities (LD) . Wong and Wong (1986) investigated metacognition in above average, average, and L D readers. These students were examined for their awareness of difficult vocabulary and disorganized prose. Four reading passages were used in which two of them contained difficult vocabulary and two of them contained relatively easy vocabulary. Two other passages were used in which the text was either organized and logically sequenced or disorganized and not logical. The children were asked questions about hypothetical students who read the passages in order to assess their metacognitive knowledge. Several weeks later, the children were randomly assigned to receive either the hard or Comprehension in Hyperlexia 45 easy vocabulary passage-pairs first, followed by the organized or disorganized passage-pair. The children were told to study the passages for subsequent recall. The children's study times were recorded and their recall was tape-recorded and later transcribed. The authors found that above-average readers were substantially more aware that the level of vocabulary difficulty and passage organization affect the ease of studying a passage (Wong & Wong, 1986). L D readers showed the same metacognitive awareness as the above-average readers but at a substantially lower level (Wong & Wong, 1986). The L D children were able to explain that the disorganized passage would take longer to study because it was harder to learn and remember. The authors concluded that these findings challenged the assumption that L D children are metacognitively deficient (Wong, 1991). In summary, the studies above have shown that metacognition aids comprehension through giving children the higher level thinking needed to evaluate and monitor their own reading comprehension. The propbnderance of studies have shown that both young children and children with L D do have and use metacognitive skills while reading. However, these children are not as adaptive at using these skills compared to older readers and non-LD students. A s we have seen, children who are good comprehenders tend to apply more metacognitive strategies, such as monitoring their comprehension or looking back at the text when they don't understand or when they need to find an answer to a question. We have also seen that children who are good comprehenders focus more on meaning and comprehension and use various learning strategies to achieve their reading goals. The studies above clearly provide evidence that without metacognition and strategies, children have lower levels of reading comprehension and hence are less successful learners. Strategy Use and Metacognition. Many studies over the years have shown the connection between metacognition and strategy use and that fostering awareness of this leads to improved strategy use and improved reading comprehension (Cross & Paris, 1988; Paris, Cross, & Lipson, 1984; Paris & Jacobs, 1984; Paris, Sarnio, & Cross, 1986; Paris & Oka, 1986). Strategic reading requires ongoing monitoring and evaluation of one's comprehension in order to achieve the goals Comprehension in Hyperlexia 46 of reading (Paris, Lipson, & Wixon , 1983). Children's awareness about the reading process and strategy use is a very important part of reading (Baker & Brown, 1984; Paris & Oka, 1986). Therefore, it is a logical assumption that an increase in children's awareness of and use o f strategies would improve reading comprehension. Paris, Cross, and Lipson (1984) designed an instruction package called "Informed Strategies for Learning" or ISL. The goal of ISL is to stimulate and increase children's awareness of declarative, procedural, and conditional knowledge, as well as teaching them ways to evaluate, plan and regulate their own comprehension (Paris et al., 1984). ISL includes various features of direct instruction, such as "(a) directing children's attention to the material to be learned, (b) generating high levels o f student involvement, and (c) providing frequent practice and immediate feedback" (Paris et al., 1984; p. 1241). ISL teaches a wide variety of comprehension strategies that have been tested in previous reading studies. Some fundamental comprehension strategies, such as "understanding the purpose of reading, activating relevant background knowledge, allocating attention to main ideas, critical evaluation, monitoring comprehension, and drawing inferences" are taught as a part of this package (Paris et al., 1984; p. 1241). In a study that examined strategy use and metacognition in children, Paris and Oka (1986) compared 500 third-graders and 500 fifth-graders that received strategy instruction to 600 same-aged peers who received no instruction. The authors gave numerous pre-tests to assess the children's comprehension level and their use o f strategies in learning. Cloze tasks, error detection tasks and reading awareness measures were given to children prior to the experimental treatment (in the fall) and subsequent to the treatment (in the spring). The authors reported that the strategic training facilitated the children's use of comprehension strategies as well as promoted comprehension (Paris & Oka, 1986). Children who had the strategy training scored significantly higher on comprehension tests as well as error detection tasks compared to the control group (Paris & Oka, 1986). The children in the experimental group had significantly increased awareness about reading and were able to describe various strategies and cognitive processes that influence reading comprehension (Paris & Oka, 1986). Additionally, children who were considered skilled readers Comprehension in Hyperlexia 47 were clearly distinguishable from their less skilled counterparts on the basis of greater metacognitive knowledge about reading and higher motivation (Paris & Oka, 1986). However, all children, regardless of skill level improved in their metacognitive abilities and comprehension skills (Paris & Oka, 1986). Recent investigations (Neuman, 1997; Turner, 1995; Perry, 1998; Winne, 1997) have shown that even very young children, preschoolers and grade one children, use metacognitive strategies in order to reach a specified goal. In her study o f grade one children, Turner (1995) examined literacy instruction and its effects on children's motivation for literacy. The study examined 84 grade one children in either basal classroom or whole language classrooms. Observers recorded both the type of literacy activity (opened and child-directed, or closed and teacher-directed) and the children's voluntary use o f motivational behaviors (strategies, persistence, and attention control). Additionally, the children were interviewed about the purposes of literacy in a four-question, open-ended interview. Turner (1995) found that open activities promoted the use of learning strategies, such as rehearsal, self-monitoring, elaboration and planning. Interviews provided evidence that children in whole-language classrooms using open activities had a better understanding of the purposes of literacy compared to students in classrooms where literacy instruction isolated skills for instruction (Turner, 1995). This study clearly shows that very young children do use strategies when reading, particularly when they are in classrooms that encourage open, child-directed learning. Successful learners are metacognitive thinkers, who use strategies to learn and who have self-regulated learning behaviors. Self-regulated learning (SRL) is used to describe children who are intrinsically motivated and regulate their own learning via metacognition and strategy use (Zimmerman, 1990). Children high in S R L classrooms are active, persistent learners who are aware of what they know and i f they know how to achieve a goal. Recent studies have shown that children can be self-regulated at a very young age (Perry, 1998; Winne, 1997). In her study of 94 grade 2 and 3 students, Perry (1998) examined children's writing arid portfolio activities in both high- and low-SRL classrooms. Perry used observation and questionnaires to assess children's Comprehension in Hyperlexia 48 beliefs and expectations about writing, their perception of control and support, and their regulation of writing behavior (Perry, 1998). The results of her study indicated that children in high-SRL classrooms were more likely to approach tasks metacognitively and strategically. In comparison, children in the l o w - S R L classrooms were more dependent on the teacher for learning and some adopted defensive and self-handicapping approaches to learning (i.e., procrastinating, giving up) (Perry, 1998). The studies above have shown that children, at a very young age, can be strategic, active learners. Very young children have been found to use various strategies, such as rehearsal, self-monitoring, elaboration, and planning. It is also clear, contrary to what researchers found in the past (Myers & Paris, 1978), that young children do have metacognitive abilities and do use them in various literacy tasks. Additionally, children, even very young children, do use strategies for learning. It has also been proven that children at a very young age can be self-regulated learners in a classroom environment that nurtures S R L . Motivation and Interest There have been a number of studies over the years that have addressed how motivation and interest lead to improved strategy use and therefore, increased comprehension and learning (Hidi , 1990; Schiefele, 1991; Tobias, 1994). Paris et al. (1991) have suggested that, in reading, cognitive tactics (planning, monitoring, elaborating) are combined with executive control strategies for managing time and attention. Motivation plays an important role in learning because it mediates both the amount of effort put into a task and the amount of satisfaction the child gets from reading. Therefore, readers need both knowledge and motivation in order to use various cognitive strategies. Paris et al. (1983) describe reading strategies as consisting of "both skill and w i l l " (p. 304). Increasing a child's feelings of self-efficacy, such as their feeling competent in using a strategy, leads to an increase in motivation (Palmer & Goetz, 1988). High interest in a topic has also been shown to increase motivation for learning (Hidi , 1990). Clearly, motivation is an important variable that is needed for a child to behave strategically and hence improve their comprehension of the material being read. Comprehension in Hyperlexia 49 Interest is one way of increasing motivation for a task. High interest in an activity creates intrinsic motivation that leads to more effortful behavior and increased strategy use. Various studies have reported that high interest in a topic leads to better comprehension and recall (Asher, 1979; De Sousa & Oakhil l , 1996; Hid i , 1990; Shirey & Reynolds, 1988). In a study that investigated the effects of interest and prior knowledge on reading comprehension, Baldwin, Peleg-Bruckner, and McClintock (1985) tested 41 seventh- and eighth-grade students. Prior to the experiment, the students completed a 10-item interest inventory and took a 100-item prior knowledge test. After completing these, each participant read passages that contained various combinations of high and low prior knowledge and high and low topic interest. The authors found interesting results in that boys scored higher in comprehension when interest or prior knowledge was high, as well as when both interest and prior knowledge were high (Baldwin et al., 1985). However, for the girls, comprehension was significantly higher on passages with high prior knowledge, but not in passages with high interest topic (Baldwin et al., 1985). The authors concluded that the impact of topic interest on comprehension was greater for boys while the impact of prior knowledge was greater for girls (Baldwin et al., 1985). In conclusion, motivation plays a very important role in reading comprehension and using strategic behaviors while reading. Motivating students to use strategies that w i l l help increase their reading comprehension can be achieved in various ways, such as helping the child to feel successful at a task. Additionally, interest in the reading material has been found to increase children's motivation for learning. Prior studies have clearly demonstrated that high-interest in a text can improve reading comprehension for that text. It is for this reason that this investigation wi l l examine the effects of high-interest o f reading material on reading comprehension in children with hyperlexia. Summary 3 A review o f the literature describing hyperlexia and comprehension in children with hyperlexia along with aspects of cognition and memory, such as attention, motivation, interest, schema, metacognition and strategy use has been presented. Prior research has not been conclusive Comprehension in Hyperlexia 50 regarding the intensity and nature of the comprehension deficit in children with hyperlexia. Although some investigations have found that children with hyperlexia have adequate reading comprehension as could be expected from their cognitive ability, others describe a much more severe deficiency. Prior studies have not thoroughly examined reading comprehension in regard to cognition and memory, attention, motivation, interest, schema, metacognition and strategy use. Since these aspects of memory and cognition are so important to comprehension and learning, it is therefore important to examine these traits in children with hyperlexia. Studies in the area of cognition and memory have clearly shown the effects that prior knowledge/schema can have on reading comprehension. Children who have prior knowledge about a topic perform better on comprehension tests than children with no prior knowledge. Additionally, studies on metacognition in children have shown that children as young as preschool use metacognitive thinking when performing tasks. Similarly, children in gradel have shown use of strategies, such as rehearsal and cognitive monitoring. It is clear that children both have and use metacognitive strategies and thinking. A s well , it is well documented that children who use metacognitive strategies perform better on comprehension tasks than children who use no strategies. Additionally, attention, motivation and interest have strong effects on reading comprehension. Children, when motivated to learn, use more effective strategies, such as comprehension monitoring. The increase in strategy use leads to an increase in comprehension of the text. High-interest in a text has also been found to increase motivation and hence increase strategy use and comprehension. Children, when given high-interest texts and low-interest texts, perform better on recall and comprehension tasks for the text of high-interest. In conclusion, the nature and severity of the comprehension deficit in children with hyperlexia is yet to be determined. In order to understand the deficit, one must examine all aspects of reading comprehension including schema, interest, metacognition and strategy use. A n in-depth examination of all these areas of reading comprehension should give us more insight into the true nature of the comprehension deficit in children with hyperlexia. Comprehension in Hyperlexia 51 CHAPTER III Method Design This study used a multiple and contrasting case study design and was conducted in two phases. There were four participants, two with hyperlexia and two with mild intellectual delays (MID) and each was assessed using multiple measures for testing word recognition and decoding, reading comprehension, prior knowledge, metacognition, strategy use and interest. The purpose of Phase 1 was to confirm the participants met the classification criteria for selection in this study. For the students with hyperlexia these criteria included: (a) early and spontaneous word recognition prior to the age of 5 years, (b) both language and cognitive delays, and (c) word recognition that exceeds what would be expected given other cognitive and linguistic functions. Students with M I D were selected to match students with hyperlexia on the basis of age, grade and cognitive ability. In Phase 1, the children completed the PPVT-III in order to assess receptive language; the Wide Range Achievement Test-Ill ( W R A T 3) to determine word-reading ability; a pseudoword reading task, the Woodcock Johnson Word Attack (Goldman, Fristoe, & Woodcock, 1974) to determine decoding abilities; and the reading subtest of the brief form of the Kaufman Test of Educational Achievement ( K - T E A ) as a standardized measure of reading comprehension. The purpose o f Phase 2 was to answer the questions posed in this study. In Phase 2, the children completed several tasks that assessed how prior knowledge, metacognition, strategy use, and interest effect reading comprehension. The multiple case study is a much stronger design compared to a single-case study. Often, single-case studies are criticized because their findings are difficult to generalize or replicate. However, a multiple case study allows for both generalization and replication over several participants. Additionally, by employing a contrasting case design, stronger conclusions can be drawn with regard to the uniqueness of the comprehension deficit in children with hyperlexia. This study wi l l suggest possible interventions for children with hyperlexia that may help them to improve their reading comprehension and learning. Comprehension in Hyperlexia 52 Participants Two children with hyperlexia between the ages of 7 to 9 with full-scale IQ scores between 50 through 70 and two children with M I D matched with the hyperlexic children for age, grade and IQ participated in this study. These children were attending a large suburban school district in Western Canada. A l l participants were required to have previously recorded intelligence testing on file at the school, since the experimenter was not qualified to administer intelligence tests. The participants were selected through district/teacher nomination. The participants with hyperlexia were chosen according to the operational definition of hyperlexia given in the first chapter of this thesis (i.e., early and spontaneous word recognition prior to the age of 5 years, as reported by the parents; both language and cognitive delays; IQ in the mild to moderately retarded range; and word recognition far in advance of their other cognitive and linguistic functions). The participants with M I D were selected as matches for the two children with hyperlexia based on their age, IQ, and grade. Preliminary discussions with school district representatives confirmed that all criteria could be met. Informed consent was obtained from the parents/guardians of each child who participated in this study. A parental consent form (see Appendix A ) was sent to each family, describing: (a) what was involved in participating in the study, (b) any potential benefits or concerns, (c) voluntary participation, and (d) confidentiality. Additionally, an informal interview conducted by the investigator was performed with each set of parents, either by phone or in person, to find out special interests o f each child and knowledge about particular reading topics. Phase 1 Measures Two measures were used to assess children's word recognition and decoding skills, the W R A T 3 and the Woodcock Johnson Word Attack, a pseudoword reading task. Receptive language was assessed with the Peabody Picture Vocabulary Test. Additionally, the reading Comprehension in Hyperlexia 53 subtest of the Kaufman Test of Educational Achievement, brief form, was administered as a standardized measure of reading comprehension. Wide Range Achievement Test 3 (WRA T3). The W R A T 3 (Wide Range Achievement Test) (Jastak & Jastak, 1978) is a test of word reading ability that consists of letter recognition, letter naming, and reading words in isolation. The entire subtest is displayed on one sheet of paper o f which there are two alternate forms. Form A was used for all participants in this study. There are two subsections; one is letter reading and the other word reading. The words at the beginning of the list are quite simple and gradually get more difficult as the test progresses. The W R A T 3 is used specifically for assessing decoding skills. According to the W R A T 3 Testing Manual (Wilkinson, 1993), the W R A T 3 has excellent internal consistency and test-retest reliability. The corrected stability coefficients range from .91 to .98 for all subtests. Additionally, the coefficient alpha for the combined reading test for ages 7 to 11 years ranges from .94 to .95. For Form A of the reading test the coefficient alpha for the same age range is .89 to .91. The W R A T 3 has excellent content as well as construct validity (Wilkinson, 1993). Each test of the W R A T 3 scored a 1.00 on the Rasch statistic of item separation, which gives strong evidence for content validity on each measure. O f the two possible subtests for the reading section, letter reading and word reading, the former section consists of 15 letters with a maximum of 15 points and the latter has a maximum of 57 points that can be earned on either Form A or B . One point is given for each correct letter and/or word read. The combined form has a maximum of 99 points. Raw scores were converted to age and grade equivalent scores. Woodcock Johnson Word Attack (WJWA). A l l participants completed the Woodcock Johnson pseudo-word reading task (Woodcock, 1973). This task consisted of reading non-words of increasing complexity and provided another example of the participants decoding and phonological skills. The initial items required participants to produce the sounds for single letters. The remaining items required participants to read aloud letter combinations that were phonetically consistent (i.e., regular) patterns in English orthography but were pseudowords. The test retest Comprehension in Hyperlexia 54 reliability for the age range of 5 to 19 years is .77. Each correct response given was scored as 1 and each incorrect response as 0. Words that were not read fluently were scored as incorrect. Participants were not penalized for any mispronunciations resulting from articulation errors, dialect variations, or regional speech patterns. The number of items answered correctly were recorded for the raw score. Raw scores were then translated into age or grade equivalent scores. Peabody Picture Vocabulary Test, Third Edition (PPVT-III). A l l participants completed the PPVT-III (Dunn and Dunn, 1997) as a standardized measure o f their receptive language abilities. The PPVT-III is a widely used test that is easy to use and requires no oral or written responses. The PPVT-III has two parallel forms, Form III A and Form IIIB. Form IIIA was used for this study. Each form contains four training items and 204 test items grouped into 17 sets of 12 items each. The items are arranged in order of increasing difficulty. Each item consists of four black-and-white pictures arranged on a page. The test taker is asked to select the picture that best represents the meaning o f the stimulus word presented orally by the examiner. Testing time averages only 11 to 12 minutes since most individuals only respond to five sets, or 60 items. The PPVT-III was standardized nationally on a stratified sample of 2,725 persons in the U.S . , 2,000 children and adolescents and 725 adults. Raw scores can be converted to age-referenced normative scores, such as standard scores, percentiles, stanines, normal curve equivalents, and age equivalents. Four types of reliability were computed for the PPVT-III . Both alpha reliability and split-half reliability were used for internal consistency. Additionally, alternate-forms reliability and test-retest reliability data was also collected. The alternate-forms reliability coefficient computed from standard scores range from .88 to .99 with a median value of .94, and the coefficients computed from raw scores range from .89 to .99 with a median value of .95. Alpha coefficients were computed as an additional measure o f internal consistency. The alpha reliability coefficients ranged from .92 to .98 with a median value of .95 for both forms. Split-half reliability coefficients ranged from .86 to .97 with a median value o f .94 for both forms. Comprehension in Hyperlexia 55 Finally, test-retest reliability scores ranged in the .90s. Scoring is accomplished by adding up the number of errors for a total number o f errors score and then calculating the raw score. A raw score is calculated by taking the total number o f errors and subtracting them from the ceiling item. The ceiling item occurs when the participant has 5 consecutive errors within a set. Kaufman Test of Educational Achievement (K-TEA). The K - T E A is an achievement test used frequently by the participants' school district to assess achievement in reading, mathematics, and spelling. There are two forms of the K - T E A , the brief form and the comprehensive form. For the purpose of this study, only the reading subtest of the brief form was used. The reading subtest starts at the grade 1 level and goes to grade 12. The first number items on the K - T E A requires the participant to read aloud the letters presented to them. The next number items are followed by words. Finally, the items require the participant to read directions and act out particular directions, such as, "drink water." The K - T E A has a split-half reliability coefficients for grades 2 and 3 of .97 and .90 respectively, for the brief form. Additionally, the test-retest reliability for grades 1 through 6 was .84 for the reading section. The standard error for grades 2 and 3 are 2.7 and 4.8, respectively. Phase 2 Measures Children completed error detection and self-correction tasks as measures of their ability to monitor their comprehension. They also read passages that contained relatively low knowledge and high knowledge content, as well as relatively low and high interest content. They answered questions to assess how prior knowledge and interest influenced their comprehension performance. Finally, for all passages, children were asked to rate their understanding, as a measure o f their ability to evaluate their comprehension, self-monitoring and self-evaluation are important aspects of metacognition. A l l passages were similar in length and difficulty for each pair, and had similar questions asked regarding the passage content. Error detection. Comprehension monitoring is a form of metacognition, thus, the ability to detect errors provides evidence o f comprehension monitoring (Garner, 1987). Therefore, this Comprehension in Hyperlexia 56 study used an error detection task to measure students' monitoring of reading comprehension. The children were asked to read two passages with errors embedded in them. One of the passages contained more obvious errors, such as symbols (ex., *%$) embedded in the middle of words (see Appendix B & C) . The second passage contained pseudowords, making the passage impossible to comprehend (see Appendix D & E). The two types of errors were used in order to determine the level of comprehension monitoring in both groups of children. It was predicted that children with M I D would find more errors overall and that participants with hyperlexia would recognize the symbol errors more easily than pseudoword errors, since it would be impossible for them to decode these errors. In previous studies, children with hyperlexia have not treated pseudo-words any differently than real words. Therefore, this task was used to force their recognition of the errors to observe how they coped with them. Upon completion of the passage, the children were given several comprehension questions consisting of passage dependent and independent questions. Passage dependent questions contained answers found directly in the text. Passage independent questions did not contain answers found directly in the text but rather required the use of prior knowledge and inference. Scoring for the error detection task consisted of the number of errors recognized over total number of embedded errors. Following the comprehension questions, the children were asked to rate the passage as (a) "easy", (b) "a bit difficult, but ok", or (c) "hard, it didn't make sense" (see Appendix F). These choices were printed on a sheet of paper with pictures of bunnies acting out each level o f understanding. If, after reading a passage, the participant decided that it was easy to understand, the experimenter asked whether there was anything difficult or confusing about it. If the participant answered "no", the experimenter directed the child's attention to an error and asked "What does this word mean?" and recorded the child's answers. Participants were trained to use the simple rating scale aimed at assessing their self-evaluations of reading comprehension. Each child was given a sheet that asked practice questions with congruent answers, such as "Is riding your bike up a very big h i l l . . .(a) easy, (b) a Comprehension in Hyperlexia 57 bit difficult, okay, or (c) hard?" and were then asked to point to the multiple-choice answer. A measure of congruency between the questions asked and the children's response was computed for the rating scale as a measure of reliability (see Appendix F). This congruency measure was used to demonstrate the students' ability to self-evaluate (i.e., how accurate is their assessment o f their performance?) Congruency is a construct borrowed from informal reading inventories, such as Manzo, Manzo, and McKenna (1995). It is a measure o f the degree to which students' responses to a question are congruent with the information asked in the question. A n incongruent response supplies information that is not relevant and suggests that the student has failed to understand or attend to the question topic. The congruency measure was scored as 1 for congruent or 0 for non-congruent responses. Table 1 shows congruent and non-congruent responses for the metacognitive rating scale and practice questions across all participants. • A l l participants' responses to the practice questions for the error detection ratings were congruent. Students with M I D ' s responses to the self-evaluation questions for the error detection tasks were congruent. In contrast, students with hyperlexia achieved a low level of congruency with regard to their judgements about the difficulty of the error detection passages. They judged that the passages were not difficult when a more reasonable response would be that the passages were difficult because they contained words that were either undecodable, not meaningful, or very difficult to read and understand. Self-Correction Task. The participants with hyperlexia were given a passage containing very difficult vocabulary and irregular words so they would have opportunities to self-regulate their reading and comprehension (see Appendix G). The students with M I D were not given this passage since it was much too difficult for them to read. The passage was taken from a text and was altered by adding extremely challenging vocabulary and irregular words into the paragraphs. After the participant read the passage, he was given both comprehension questions assessing passage dependent and independent knowledge, as well as the same rating scale described in the error detection task (see Appendix F). The same measure of reliability was used to indicate i f the responses to the questions were congruent or non-congruent. Comprehension in Hyperlexia 58 Table 1 Congruent vs. Non-congruent Responses for Rating Scale Hyperlexia M I D Hyperlexia M I D Assessment Bob Harvey Myles Nelson Practice 1 1 1 1 E D #1 (symbols) 0 1 0 1 r E D #2 (nonwords) 0 1 0 1 Self-Correction 0 0 L o w P . K . 0 1 0 1 Low Interest 0 1 0 1 High P . K . 0 1 0 1 High Interest 1 1 1 1 Note. ED = error detection, P.K. = prior knowledge, 1 = congruent response, 0 = non-congruent response. Bob's responses to the rating scale questions for the self-correction task were non-congruent. Myles also had non-congruent responses to the rating scale for the self-correction task. Both students with hyperlexia judged the passage to be easy when in fact it was very difficult for them to read and understand (see Table 1). Scoring for the self-correction followed standard scoring procedures for scoring informal reading inventories. A n accuracy score was calculated, percentage o f all words read accurately. Additionally, an error analysis was completed to look for patterns such as decoding errors and non-meaningful substitutions. Tests for Schema/Prior Knowledge. This task was designed to assess whether children with hyperlexia use prior knowledge spontaneously to aid in comprehending text, and whether their comprehension could be improved by the activation of prior knowledge. Two reading passages to test for prior knowledge were given to each participant (see Appendix H). Information obtained Comprehension in Hyperlexia 59 from an interview of the parents prior to testing of the participants was used to choose topics for each passage. The first task was to read a passage that contained information the children did not know a lot about. The second reading passage was one of relatively (i.e., relative to the first passage) high knowledge but of no particular interest to the child. Prior to reading these passages, several questions were asked with regard to the passage topic, such as "what color is a frog?" (see Appendix H). This was done in order to assess the children's level of prior knowledge on the topic. After the participants finished reading each passage, several passage dependent and independent comprehension questions were administered as well as the rating scale. The same measure of congruency that was described previously was used to assess each student's understanding of the questions. Each passage was accompanied with a title and a picture that related to the passage in order to activate schema. Each passage was compared as to how successful it was in activating the child's schema and the level of help each student required for activation. The number of correct responses to the comprehension questions was the dependent variable for this prior knowledge task. Therefore, the participants received a score for the number of questions they answered correctly out of the total number of questions asked. On the measure of congruency, Bob had a non-congruent response for both low and high knowledge passages. Similarly, Myles gave a non-congruent response to the rating scale for both the high and low knowledge passages. However, the students with M I D evaluated the level of difficulty of the passage correctly and gave congruent responses to the rating scale. Tests for Interest. Following the test for prior knowledge, a layer of interest was added to the previous reading task. This allowed for comparison between high knowledge, low interest, and high knowledge, high interest passages. Each participant read a passage on a topic that had high personal interest. Then each participant was asked several passage dependent and passage independent comprehension questions (see Appendix I). A s in the prior knowledge task, the children were given the rating scale for self-evaluation of their reading. Reliability was assessed through the use of this congruency measure. Scoring for this task was similar to the prior knowledge task. The dependent variable was the scores on the comprehension questions. Comprehension in Hyperlexia 60 Therefore, a score was calculated from the total number o f correct responses over the total number of questions asked. Performance across passages was also compared. Both participants with hyperlexia gave non-congruent responses to the rating scale for the low interest passages and congruent responses to the rating scale on the high interest passages (see Table 1). Both participants with M I D responded correctly to the rating scale and had congruent responses for both low and high interest passages. For the first time (besides the practice questions) the hyperlexic students had congruent responses for a passage. It appeared that the high interest passage motivated the students with hyperlexia to attend more to the text, to comprehend better, and to evaluate their understanding appropriately. Phase 1 Procedure Initially, participants were selected for the study through teacher nomination. Then, the investigator followed up with a brief meeting, either on the phone or in person, with each set of parents prior to the start of the investigation. The meeting lasted approximately 30 - 45 minutes. In the meeting, the experimenter asked the parents to supply information regarding their children's interests and knowledge about reading topics. Additionally, the investigator visited the classrooms o f each participant, met their teachers, and had some interaction with the children prior to beginning data collection. The Phase 1 measures were administered at the participants' schools. For the students with hyperlexia, this phase took place over 2 sessions. However, the participants with M I D took 4 sessions in order to complete this phase. During each experimental session, the participant and the experimenter met at the child's school, in a quiet room, and sat across from one another at a table. The length of the sessions varied across participants from 15 to 45 minutes. In the first and second session, each participant in the hyperlexia group completed the Phase 1 measures of word reading ability (on the W R A T 3), pseudoword reading (on the Woodcock Word Attack), receptive language (PPVT-III) and reading comprehension (on the K - T E A ) . The student's in the M I D group needed 4 sessions in order to complete all the Phase 1 measures. To begin the first session, the investigator sat across the table from the participant and told him that she has a few tasks that she wanted them to do. The participants with hyperlexia Comprehension in Hyperlexia 61 completed all phase 1 assessments in two sessions. However, the participants with M I D required 4 sessions to complete all phase 1 tasks. Therefore, the students with M I D did complete one task per session. The procedure for each student remained the same regardless o f the number of sessions needed to complete the assessments. For the first task, the experimenter showed the student a word list from the W R A T 3 and asked him to read the words aloud. The investigator simultaneously observed and scored the child. After the child completed the word list, he was asked to read a list of pseudo-words taken from the Woodcock Word Attack list. The following directions were given to the participant, "Read each of these words out loud and tell me i f any o f the words are too difficult to read." The PPVT-III was administered to assess the participant's receptive language. The participant was presented with a set of four pictures and was asked to point to the picture that matched the stimulus word given by the investigator. Next, each participant completed the brief form of the K - T E A Reading Comprehension Test. For the comprehension portion of the test, the students were asked to read the word or words and complete the desired action. The investigator remained seated with the child for the duration o f the test. If the child had difficulty completing the task, the experimenter supported the child by prompting them to stay on task. Additionally, the experimenter encouraged each child to ask questions i f they did not understand something. Phase 2 Procedure In Phase 2 o f the experiment, each child received measures for metacognition, prior knowledge, and interest. In the third session, the students with hyperlexia completed two of the Phase 2 tasks: the error detection task and the self-correction task. Since the M I D group required more time to complete each assessment, these same tasks were completed in the 5 t h and 6th sessions. However, the procedure remained the same for each participant regardless of the number of sessions. First, each child was asked to read the passage. Upon completion of the passage, the children were given several comprehension questions, consisting of passage dependent and independent knowledge. Following the comprehension questions, the children were asked to rate the passage as (a) easy, (b) O K , a bit hard, or (c) hard, it didn't make sense. These choices were Comprehension in Hyperlexia 62 printed on a sheet of paper with pictures of bunnies acting out each level of understanding. If, after reading a passage, the participant decided that it was easy to understand, the experimenter asked whether there was anything difficult or confusing about it. If the participant answered "no" the experimenter directed the child's attention to an error and asked "What about this word? Can you read it?" and then recorded the child's answers. A running record of each participant's reading was completed in order to check for monitoring of comprehension while reading. Each child was asked to read the passages out loud to the experimenter. The investigator recorded the children's reading and metacognitive skills, such as re-reading and self-correction. Body language and facial expressions were observed and recorded as indicators of perceived difficulty, and frustration. Additionally, both participants with hyperlexia received a challenging passage with difficult vocabulary so they had the opportunity to self-correct. This particular task was too difficult for the children with M I D who could not read any difficult vocabulary. Therefore, they were not given this particular task. This self-correction task was used to examine the children's metacognitive abilities. After the child finished reading the passage, the experimenter gave the children several passage dependent and passage independent comprehension questions as well as the rating scale described previously. In the next session, each participant completed a test for prior knowledge. First, in order to assess their prior knowledge, the participant was asked a few questions regarding their knowledge on the passage topic. This was performed in order to see what prior knowledge they display as well as to activate their schema of the topic. Next, the child was given two reading passages with a title and picture to help activate their schema. The first passage was low interest/low knowledge and the second, no particular interest/relatively high knowledge. After reading each passage, several comprehension questions, both passage dependent and independent, were given to them to read and answer. Again the same rating scale questions (as described above) were given to assess self-awareness. Finally, all participants received a passage that was of specific high-interest to each of Comprehension in Hyperlexia 63 them. Again they were asked several passage dependent and independent comprehension questions with regard to the passage. The children were also given the same rating scale questions (as described above) in order to assess metacognition. After all the testing was completed, each child was thanked for helping the experimenter to learn more about reading. Comprehension in Hyperlexia 64 C H A P T E R IV Results The purpose of this study was to examine reading comprehension in children with hyperlexia and compare them with children with M I D . The participants were matched for age, grade, and IQ. Two children with hyperlexia and two matched participants with M I D were assessed on word reading, pseudo-word reading, general comprehension, and the specific aspects of reading comprehension that are listed above. The primary research question was: What accounts for the poor comprehension in these hyperlexic children? Is it their low language or cognitive ability? Is it their failure to monitor or self-correct? Is it their failure to activate relevant schema? In this chapter, the results from the various assessments used are presented for each child. A s well , some comparisons are made between the pairs of students with hyperlexia and children with M I D . Initially the screening process is presented, followed by thorough descriptions of each case. The results are divided into three main sections: identification and screening,.Phase 1 results, and Phase 2 results. Identification and Screening Elementary and middle schools in the school district were notified about the study and its criteria. Through teacher and school psychologist nomination, 24 students were considered for the study. After consulting with each of the student's teachers, only 12 o f the 24 students were considered to fit the study's criteria and were screened for the study. After initial discussions with teachers and examination o f their files, only four children who fit the criteria for the study and had parents who gave permission. The eight children not chosen for the study were found either to have IQ scores that were not within the target range or had no appropriate age and IQ match. One child did fit the criteria very well , but his parents did not give consent for him to participate. Participants The four participants chosen for the study were selected on the basis of either the criteria for hyperlexia or M I D described in Chapter 3. A l l participants were matched in pairs according to Comprehension in Hyperlexia 65 their age, IQ, and grade level. Bob and Myles were the participants with hyperlexia. Harvey and Nelson were matches for Bob and Myles respectively. Table 2 shows the demographic information for each student. Table 2 Participant Information Hyperlexia M I D Hyperlexia M I D Matching Variable Bob Harvey Myles Nelson Age (years/months) 8^ 2 8.8 7^8 Grade 3 2 2 2 Ful l Scale IQ 54 68 69 69 Note. FSIQ data (full-scale IQ) was taken from the WISC-III. : Bob. Bob was an 8-year-old boy in the third grade, who was described by his teachers as an extremely good decoder with mild to moderate cognitive delays. Bob had an FSIQ on the WISC-III of 54 (in the mi ld to moderately intellectually delayed range). He was reading at least 2 grade levels above his own grade, at a grade 5 level, but had very poor comprehension for what was read. His teachers described him as having difficulty following and remembering directions given to him in class. For example, at the beginning of each class the children were supposed to put away their coats and bags and take out their books and homework from the previous day. Bob was unable to perform this task on a daily basis without consistent prompting from one o f his teachers. Bob was also considered to be somewhat of an enigma to both his parents and teachers. Neither his teachers nor his parents knew how to help Bob improve his learning skills, as he seemed somewhat difficult to teach. Bob did not appear to grasp any of the typical grade 3 lessons. Although he did not have a classroom aide assigned to him, the assistant in the class was consistently involved in Comprehension in Hyperlexia 66 helping Bob to keep on track and complete his work for the day. Bob was usually pleasant and well behaved in the classroom, However, teachers did comment on his temper and difficult behavior that arose from time to time. Bob was described by his parents as a precocious reader who began reading before the age of 5 years and who would read anything he could get his hands on. Healy et al. (1982) described their participants with hyperlexia with similar characteristics, such as early and intense interest in words and reading, combined with delayed cognition and language. With regard to reading interests, Bob, his parents, and teachers agreed that his main interest was learning about Orcas (killer whales). Harvey. Harvey was 7 years and 10 months old when the study took place. He was the age, IQ match for Bob. Harvey was in grade 2 and had a WISC-III FSIQ of 68. He was described by his teachers as mildly to moderately intellectually delayed and a very poor reader. Harvey needed help with many tasks in his class and often needed an assistant to aid him. Harvey was well behaved in class but had a very short attention span. Harvey was reading at a kindergarten level and had many problems with decoding and pronunciation. According to his teacher, Harvey's comprehension ability was better than his decoding ability, and during the study he was able to answer grade 1 level comprehension questions from a passage that was read to him. His mother described him as a "slow" learner and wanted to learn how to help him to do better in school. According to his mother, Harvey was not always comfortable working with new people and she preferred that i f he did not want to do the necessary reading for the study that he should not be made to do it. Harvey and his mother both agreed that his main interest was different types of vehicles. My les. Myles was an 8-year-old boy in grade 2. His mother and teachers commented on his exceptional decoding ability and intense interest in reading. Myles was able to decode at a grade 5 level or higher. According to his mother, he began reading at a very early age, 3 or 4 years, and would read anything he could get his hands on. His biggest interest was hockey and he knew a lot of information about different teams and how the game was played. Children with hyperlexia have Comprehension in Hyperlexia 67 been described in previous studies as having autism or autistic like behaviors (Snowling & Firth, 1986; Whitehouse & Harris, 1984). Similarly, although not officially diagnosed, Myles had an / / almost autistic-like demeanor, presenting with definite sensory and behavior problems such as sensitivity to touch, textures, sound, and temper tantrums. In addition, Myles was not successful with social relationships with other children and had little interaction with them. Often, Myles would become over-stimulated in the classroom and, to calm him, his sister would accompany him to a piano room where Myles would play the piano for 20 to 30 minutes until he felt calm. Changes in schedule were very difficult for Myles . In one incidence, the investigator came early for a testing session and Myles was in his computer class. He threw an intense temper tantrum when asked to leave the computer and go with the investigator. In order to calm him, the investigator told him that she would wait to meet with him until his computer time was over. This appeared to sooth Myles and he calmed down quickly. Myles also displayed a very short attention span, especially when he was not interested in something or i f the task was too difficult for him. During error detection tasks, Myles was very inattentive and would try to escape the task by running around the room or playing with a toy plane. Nelson. Nelson was described by his teachers as mild to moderately cognitively delayed with very poor reading skills. He was the age and IQ match for Myles . Nelson was reading at a kindergarten level and had considerable difficulty with pronunciation and decoding. His reading comprehension was much better than his decoding. Nelson really enjoyed school and liked his teachers. He was generally well behaved, although he was quite hyperactive with a very short attention span. Nelson's parents and teachers reported that he was interested in anything about cars or vehicles. Phase 1 Results The purpose of this phase was to confirm the status of students as hyperlexic or mildly intellectually delayed group. For example, the PPVT-III gave a measure of each child's receptive language ability. A l l the measures, except the K - T E A , were selected because they were used in previous investigations to characterize students with hyperlexia. The PPVT-III was chosen since it Comprehension in Hyperlexia 68 has proven to be well correlated with IQ and comprehension. In addition, the PPVT-III has been well utilized in past research on hyperlexia (Glosser et el., 1996; Goldberg & Rothermel, 1984; Siegel, 1984; Temple, 1990; Worthy & Invernizzi, 1995). The W R A T 3 was employed to assess single word reading ability, which is generally very good in children with hyperlexia. The W R A T 3 has also been well utilized in past literature (Healy et al., 1982; Patti & Lupinetti, 1993) and was employed by Siegel (1984) to assess word reading in a young girl with hyperlexia. The Woodcock-Johnson Word Attack was used to test each participant's pseudo-word reading (or decoding) ability, which has been shown to be very strong for the children with hyperlexia. Finally, to get a standardized measure of each child's reading and comprehension skills, as well as to demonstrate the discrepancy between word reading and comprehension in children with hyperlexia, the reading subtest of the brief form of the K - T E A was administered. The K - T E A , although not utilized in past research, was the standardized measure of achievement that was used in the participating school district. The thesis question addressed in Phase 1 was: D id the participants with hyperlexia have cognitive or IQ levels that are commensurate with their language or P P V T levels? Was their word reading ability far above what was expected, given their age and IQ ability? Table 3 shows the student's performance on these measures of IQ, receptive language, word-reading ability, and reading comprehension. Comprehension in Hyperlexia 69 Table 3 Results for Phase 1: Cognitive Ability, Word Reading Measures and Comprehension Hyperlexia M I D Hyperlexia M I D Assessment Score Type Bob Harvey Myles Nelson Measure of Cognitive Abi l i ty WISC-III FSIQ 54 68 69 69 V I Q 63 55 73 72 PIQ 52 84 70 70 Measure o f Language Abi l i ty PPVT-III SS 66 60 84 76 RS 63 49 93 84 A E 5.1 3.11 7 6.4 G E PreK PreK 1.5 K Measure o f Word Reading W R A T SS 129 69 102 67 RS 40 17 29 16 A E 14 6 9 6 G E 8 K 3 K W O R D SS 113 0 102 67 RS 26 1 21 3 A E 13.5 5.1 10 6.3 G E 7.5 PreK 4 K Comprehension in Hyperlexia 70 Assessment Score Type Hyperlexia M I D Hyperlexia M I D Bob Harvey Myles Nelson Measure of Comprehension K - T E A SS 101 . 0 93 0 RS 24 0 23 5 A E 10 8 G E 4 2.5 Note. A E = age equivalent, G E = grade equivalent, RS = raw score, SS = standard score, %ile = percentile. A l l IQ scores are reported as: FS = full scale IQ, V = verbal IQ, P = performance IQ, PreK = pre-kindergarten, K = kindergarten. Peabody Picture Vocabulary Test III (PPVT-III) Bob. Bob's standard score (SS) of 66 corresponded to an age equivalent (AE) of 5.1, compared to his chronological age o f 8 years. This score was consistent with his Ful l Scale IQ (FSIQ) of 54, which was in the moderately intellectually delayed range. Thus, Bob's cognitive and language ability were significantly below average. During testing, Bob was very quick and reckless when making his choices on the P P V T . When he did not know something he would simply pick an item at random. He understood the task, but did not appear to take any extra time or put any effort to think through each answer. Harvey. Harvey achieved an SS of 60 on the P P V T , which corresponded to an A E score of 3.11, compared to his biological age of 8 years. Harvey's FSIQ score o f 68 was in the mildly intellectually delayed range. It appeared that Harvey's IQ score was higher than his receptive language level of 3.11 years of age. However, because his verbal IQ score was 55 and his performance IQ score was 84, this greatly inflated his full-scale IQ. Therefore, his verbal IQ was very close to his receptive language level. Harvey understood the task, but found it extremely difficult. Frequently, he was unfamiliar with the receptive vocabulary for his age level, as well as Comprehension in Hyperlexia 71 vocabulary years below his age level. Harvey often did not want to complete the task. Therefore, he had to be tested on two different days in order to complete the assessment. Myles. Myles achieved an SS on the P P V T of 84, which placed him only 1 year behind his actual age of 8 years. This score suggested that Myles ' s receptive language resembled that of a 7-year-old child. His FSIQ score o f 69 was in the mildly intellectually delayed range. Therefore, Myles 's language and cognition levels did appear close to equivalent, however, they were not as close as Bob's scores. In past research, it has been frequently reported that children with hyperlexia have higher P P V T scores compared to their IQ level (Temple, 1990; Worthy & Invernizzi, 1995). Myles was very quick to give answers, whether he actually knew the answer or not. He would often look to the investigator for reassurance of his choice. He also would try to look at the investigator's answer sheet to see i f he had given the correct answer. Myles needed lots of reassurance that he was doing the task well and appropriately. Nelson. Nelson achieved a SS of 76, which corresponded to an A E of 6.4 years. His V I Q score on the WISC-III was 72. This illustrated his low receptive language compared to age and grade level. Nelson became easily bored and often needed a lot of prompting to keep him on task. He had difficulty sitting still and would frequently try to escape the task at hand by telling the investigator stories. Comparison. The two children with hyperlexia appeared to have receptive language scores that were below their age but at or around their IQ level. The boys with M I D had receptive language scores that were well below their age level. However, Harvey's receptive language score was much lower than his IQ level. Wide Range Achievement Test 3 (WRA T 3) Bob. Bob's SS of 129 on the W R A T 3 placed him at the 97 t h percentile and a grade 8 word-reading level. His A E score was 14 years. Bob's scores on this measure demonstrated his excellent single word reading ability, which was far above what was expected for his cognitive level (FSIQ = 54), current age (8 years) and grade level (grade 3). These findings paralleled results from past research in which children with hyperlexia were found to have excellent word reading Comprehension in Hyperlexia 72 abilities (Healy et al., 1982; Siegel, 1984; Temple, 1990). Bob appeared to enjoy this word-reading task, which also paralleled children with hyperlexia in the literature. Bob seemed to use a phonological approach to word reading as the words became more challenging to decode. He was very successful at sounding out the words. This appeared to be relatively easy for him, which mimicked many children in the literature with hyperlexia, who were reported as successful phonological readers (Healy et al., 1982; Siegel, 1984). Harvey. Harvey found this task very difficult. He could only read two words on the W R A T 3. Harvey's SS of 69 on this measure placed him at the 2 n d percentile and at a kindergarten reading level, far below what was expected for his age and grade level. His score was eight G E ' s below Bob's . Harvey attempted to use phonetics to sound out the difficult words on this measure. Myles. Myles found this task quite simple. He enjoyed reading the words, but became frustrated as the words grew more difficult than he could decode, like "bibliography." Myles appeared to use a sight word or orthrographic approach to reading, since he did not appear to be very good at sounding out the words. Similarly, previous studies that examined reading routes in children with hyperlexia reported a sight word or othrographic approach to reading in their participants (Goldberg & Rothermel, 1984; Seymour & Evans, 1992). Myles had a SS of 96 on the W R A T 3, which placed him at the 39 t h percentile, with an A E of 9 years and a grade 3 decoding level. This meant that Myles was reading one grade ahead of his actual age and grade. However, upon examination of Myles ' s cognitive level (FSIQ = 69), his word-reading score is much higher than would be expected. Nelson. Nelson found this task very difficult. He could only read two words on the W R A T 3 and 14 out of 15 letters. His SS was 66, which placed him at the 1 s t percentile and a kindergarten grade level. Nelson used a phonological approach to reading, as he attempted to sound out each word he read. Comparison. The two children with hyperlexia clearly demonstrated their excellent decoding skills on this task. A s expected, Bob scored much higher than his cognitive, age or grade level. Myles also scored much better than expected for his cognitive level, where as Harvey and Comprehension in Hyperlexia 73 Nelson scored much lower than expected for their age or grade level. The performance of Bob and Myles on this measure was far above what is expected, given their low cognitive ability, whereas Harvey and Nelson's scores were consistent with their cognitive level. Word Attack Bob. Bob had a SS of 113 on the Word Attack, which placed him at the 80 t h percentile. He achieved an A E score of 13.5 years and a G E score of 8. This task clearly illustrated Bob's superior decoding and phonetic skills. Once again, Bob scored much higher than would be expected for his cognitive level. Siegel (1984) reported similar results for a young girl with hyperlexia. Although her comprehension was extremely limited, her pseudo word reading was far above her age/grade level. Harvey. Harvey's raw score (RS) of 1 on the Word Attack placed him below the 1 s t percentile and at an age equivalent of 3.11 years. He was only able to point to the letter "r." Harvey's score was too low to be transformed to an SS or to compute a G E . This clearly demonstrated Harvey's weak decoding skills. Myles. Myles achieved a SS of 102 on the Word Attack. This corresponded to an A E score of 10 years, a G E score of 4, and the 55 l h percentile. These scores were much higher than would be expected for someone with an IQ of 69. A s predicted, Myles ' s decoding skills were very strong for this pseudo-word reading assessment. It appears that Myle ' s orthographic approach to reading may have affected his performance on this task. Nelson. Nelson scored 67, which was equal to an A E score o f 6.3 years and placed him at the 1 s t percentile. He was unable to read any of the pseudo-words. Therefore, he was only able to read and sound out three single letters correctly. Comparisons. Bob and Myles 's performance on the Word Attack was consistent with the criteria for hyperlexia. Bob scored much higher than his age, language or cognitive level should allow. Myles also performed better than his cognitive or language level should allow, but did not do as well as Bob. Conversely, the boys with M I D had scores that were more consistent with their cognitive level, and their scores were well below what would be expected of typical readers in Comprehension in Hyperlexia 74 their age/grade range. Harvey scored very low on this task and was 4 - 5 years below his chronological age. Nelson was 2-years below his age-level on this task. Kaufman Test of Achievement (K-TEA) Bob. Bob achieved a SS of 101 on the K - T E A , which placed him at the 53 r d percentile. Bob 's A E score was 10 years and his G E was 4. This score was much better than expected for a child with hyperlexia, and due to the various limitations o f this assessment, may not present a very accurate estimate of reading comprehension ability. The brief form of the K - T E A does not provide separate assessments of students decoding and reading comprehension. Since Bob was very successful at decoding, he was able to do well at the start o f the test. A s the test progressed, Bob was only able to answer a few of the complex two word directions. This trait has been frequently reported in studies of children with hyperlexia. Siegel (1984) and Healy et al. (1982) reported that participants could not seem to understand the directions given to them, or even simple commands. Harvey. Harvey was unable to read any of the words for this test and therefore achieved a score o f 0 for this task. He did not get beyond the strictly decoding portion of the comprehension section of this assessment. Therefore, there was no measure of reading comprehension for Harvey. Myles. Myles had a SS of 93 on the K - T E A , which falls at the 34 t h percentile and was a good score considering his IQ level. His A E score was 8-years-of-age and his G E was 2.5. Once again, these results may indicate a much higher comprehension score than was true to Myles ' s actual comprehension abilities. Since Myles had a very recent K - T E A performed by the school psychologist, his score from that assessment was used for this investigation. However, since this previously administered test was used, more specific information about Myles ' s answers were not available. Nelson. Nelson was only able to read one word, the word "at." Therefore, he did not even get a chance to demonstrate his comprehension abilities on this task. Comparisons. A comparison o f performances on this task did not do justice to the actual reading comprehension skills in these children. The brief form of the K - T E A was a convenient standardized measure of participant's reading comprehension. Unfortunately, due to the Comprehension in Hyperlexia 75 confounding of decoding and comprehension, it was a poor gauge of reading comprehension in these students. What it did appear to confirm was that word reading ability was much higher in both children with hyperlexia compared to their M I D counterparts. Both Bob and Myles achieved higher scores on the reading comprehension section compared to Harvey or Nelson. The scores for the students with hyperlexia were inflated due to their excellent decoding abilities. Conversely, the scores for the students with M I D were deflated due to their difficulty in decoding any of the words on this measure of "comprehension." Phase 2 Results The purpose for this phase in the study was to examine the comprehension and metacognitive awareness of children with hyperlexia compared to the performance o f the children with M I D . A s previously stated in Chapter 2, reading comprehension is made up of many important components. Key components of successful reading comprehension are: prior knowledge, metacognition, motivation and interest. Motivation, interest, and prior knowledge have been found to have a tremendous impact on children's reading comprehension. For example, when interest is high, children have much higher reading comprehension (De Sousa & Oakhil l , 1996). Previous studies have concluded that successful readers are metacognitive thinkers (Garner, 1987; Perry, 1998). This means they monitor their own comprehension while they read, recognizing and self-correcting problems in their reading. The current study examined the effects o f these key components on reading comprehension. The error detection tasks were used to assess the participant's metacognitive abilities through insertion o f errors in the reading passage. The first error detection passage had symbols inserted in various places in the passage and the second error detection passage contained pseudowords. The self-correction task used very difficult vocabulary in order to force students' recognition of a comprehension breakdown and use of problem solving strategies. Prior knowledge tasks examined both high and low (comparatively) prior knowledge about a topic and the impact on comprehension. The interest passages were similar to the prior knowledge tasks as they assessed the difference in comprehension level for both high and low interest passages. The Comprehension in Hyperlexia 76 data collected for phase 2 measures on reading comprehension can be found in Table 4. It was predicted that the children with hyperlexia would perform more poorly on error detection tasks compared to the M I D children. It was also predicted that the two children with hyperlexia would detect more errors in the first error detection with embedded symbols compared to the second error detection with pseudo words. For the self-correction task, which consisted of some very difficult vocabulary, it was predicted that the children with hyperlexia would decode the words without comprehending them and that the students with M I D would apply more self-correction strategies. Prior knowledge was predicted to have a positive impact on comprehension for both groups of children. Interest was also expected to have a positive influence on reading comprehension for both the hyperlexic group and M I D group. However, it was predicted that the hyperlexic students would not use metacognitive tactics to aid their reading comprehension. Conversely, the students with M I D were predicted to display some simple metacognitive tactics when reading. Error Detection #1 Bob. Bob's error detection score on the 1 s t task was 25%, which means he only detected 2 out of 8 errors in the passage. Bob appeared very confused by these words with embedded symbols in them. He tried to read these words one letter and symbol at a time. For example, "r$#k", he would read as "r, dollar, number, k." Although he did, in fact, read the pseudoword correctly, his reactions to these errors illustrated his view that reading is decoding and not comprehending. Frequently Bob would make non-meaningful substitutions for words, such as "slow" for "s%*#y", which made no sense in the context of the sentence. When asked i f he found the passage "easy" to read, "ok, a little hard", or "really hard" to read, Bob answered, "easy, because the words were easy to read." Consequently, Bob did not exhibit any metacognitive thinking, since he did not recognize the errors. In addition, Bob demonstrated no understanding that the passage made no sense as written. Harvey. Harvey detected all 5 out of 5 errors in this reading passage. Harvey was more cautious than the other children and although he noticed all 5 errors, he was hesitant to say that the Comprehension in Hyperlexia 77 first error was not a real word. He hesitated at the first error then said, "what's this?" The investigator responded with, "I 'm not sure, what do you think it is?" "It's not a word", Harvey responded. Even Harvey, with his very poor decoding was able to acknowledge that those "weird words" were not "real" words. When given the metacognitive rating scale, Harvey replied that the passage was "really hard to read, because there were weird words in it." It was apparent from Harvey's responses that he understood reading was more than decoding. He tried to understand the passage. Myles. Myles achieved a score o f 12% (1 out of 8 errors) for this error detection task. Myles had a tendency to ignore the symbol words and trying to skip by them to read the next words in the passage. When asked what the word " w * A $ f ' meant, his answer was "I don't know." Myles often made non-meaningful substitutions, such as " w o l f for "w* A $f ' , which did not make any sense in the context of the sentence or the passage. Despite the trouble with reading this passage, when asked i f the passage was "easy" to read, "a little hard, but ok", or "really hard" to read, he answered, "easy, because it was a story." This answer indicated that Myles was not applying metacognitive thinking to help his comprehension of the passage. Nelson. Nelson scored 100% (5 out of the 5 errors) in the first error detection passage. It appeared very easy for Nelson to inform the investigator, "Those aren't real words." When Nelson received the rating scale, he answered that the passage was, "a little hard, but ok, because of the silly words". Since Nelson was aware that the passage did not make sense, it appeared that he was monitoring his comprehension of the passage. Comparison. Upon examination o f the results from this task, it was quite clear that the children with hyperlexia did not monitor comprehension or evaluate their understanding when reading these error detection passages. Bob and Myles found fewer errors compared to their M I D matches. Similarly, Goldberg and Rothermel (1984) reported no difference in reading ability or comprehension of their participants with hyperlexia when reading passages that contained errors or no errors. Comprehension in Hyperlexia 78 Table 4 Phase 2 Comprehension & Metacognitive Awareness Measures: Percent of Correct Answers Hyperl exia M I D Hyperlexia M I D Assessment Bob Harvey Myles Nelson Metacognitive Awareness E D #1 (symbols) 25% b 100% a 13% b 100% a E D #2 (nonwords) 8% b 80% a 8% b 80% a Self-correction 30% b 30% b Low P. K . 1 6 % b 83% a 50% b 8 3 % a Low Interest 33% b 100% a 66% b 100% a High P. K . 1 6 % b 83% a 83% b 100% a High Interest 6 6 % c 100% a 100% d 83% a Note. P.K. = prior knowledge; a = kindergarten level, = grade 2 level , c = grade 6 level, = grade 5 level Comprehension in Hyperlexia 79 Error Detection #2 Bob. On the second error detection task, Bob detected 1 out of 12 or 8% o f the pseudo word errors embedded in the passage. The word that Bob detected was "dute". He asked, "What word is this?" Through the remainder of the passage, Bob either skipped over the pseudo-words or attempted to decode them. When asked i f the passage was "easy", "okay", or "really hard", Bob answered "easy, because the words were easy to read". This answer demonstrated Bob's lack of monitoring his own comprehension while reading. Harvey. Harvey detected 80% or 4 out of 5 errors. He missed the 1 s t error, "keek". Harvey found this task more difficult than the first error detection passage. Unless the errors were clearly non-words, it was difficult for Harvey to catch them since his decoding skills were so poor. When asked how he found the passage, easy, okay, or very hard, Harvey answered, "really hard, because of the silly words". Therefore, Harvey realized that these non-words were "s i l ly" words and that the passage did not make sense. Once again, Harvey showed evidence of applying metacognitive thinking while reading the passage. Myles. On this error detection task, Myles was able to detect only 1 out of 12 or 8% of the pseudo words embedded in the passage. The word that Myles asked about was "oolum". He questioned the investigator, "What's this word?" Through the rest of the passage Myles skipped over all the pseudowords or he would make up words in their place, such as "open" for "oolum", which made no sense in the context of the passage. Myles said that this passage was, "a bit difficult" to read "because a girl went up the h i l l " . Therefore, Myles did recognize that the passage was difficult, but did not know why it was difficult. This statement conveyed that Myles did not monitor his own comprehension; that he had no metacognitive awareness or reasoning why the passage was a bit difficult to read. However, he did understand that going up a hi l l was hard. Therefore, he did attend to the meaning at the sentence level. Nelson. Nelson achieved a score of 80%, which means he detected 4 out of 5 errors in the passage. He missed the 1 s t error, "keek". Nelson did have some problems on this task that were specifically related to his low decoding abilities. However, he was quite accurate in determining Comprehension in Hyperlexia 80 which words were the pseudowords. When asked how he found the passage, he replied, "Okay, some of the words were hard," and he pointed to the pseudo-words. Nelson did understand that the passage was difficult to understand due to the pseudo-words and, therefore, he illustrated metacognitive thinking. Comparison. It was hypothesized that the children with hyperlexia would do much better on the first error detection task with the embedded symbols compared to the second error detection task with the difficult vocabulary. Bob actually detected slightly more errors in the first passage, however there was not a large discrepancy in errors found between passage 1 and 2. Harvey and Nelson were quite adept at picking out the non-words. They both missed the same error, "keek". It's possible that this word was too similar to a "real" word for the M I D children to detect. These findings suggest that the children with M I D did monitor their comprehension whereas the children with hyperlexia did not appear to monitor their comprehension. Self-Correction Only Bob and Myles were able to complete the self-correction task containing difficult vocabulary. Harvey and Nelson found it too difficult, even though a simple passage was used, rated at a K to grade 1 reading level. They could not read any words from the original passage and, therefore, this task was too difficult for them to complete. Bob. On the self-correction task using difficult vocabulary words, Bob detected 30% or 3 out of 10 of the difficult words in the passage. The words that Bob asked about were "adjudicated", "impulsively", and "sasquatch". He was unable to decode those words and therefore stopped and asked, "What's this word say?" Bob's behavior in this task illustrated some ability to focus on self-correction and comprehension at the sentence level. However, it appeared to be only when he was forced to his decoding limit that he would ask for help with his comprehension. When asked how he found the passage, easy, okay, or very hard, Bob replied "easy, because it's about beavers". Obviously Bob did not consider the complexity o f the words or comprehension, nor did he recall the difficulty he had reading some words. Instead he relied on his prior knowledge of beavers to conclude the text was easy to read. This suggests that Bob Comprehension in Hyperlexia 81 was not reading for meaning or monitoring his ability to comprehend the passage. However, the fact that he recognized some difficult words and asked about them is evidence of some monitoring of his reading. Myles. Myles had a score of 30% detection of the difficult words. This means he detected 3 out of 10 of the difficult vocabulary embedded in this passage. Myles found three of the words hard to decode, "adjudicated", "impulsively", and "resonated". He often skipped over the difficult words when he could not decode them. When asked what "veering" meant, he said, "I don't know". When asked what a "chasm" was, he said, "it's a den", which did imply some use of context to understand the sentence since the sentence implied that the beavers lived in a chasm. In Addition, he was able to understand that a "sasquatch" was some type of a monster. This answer showed that Myles was using the context of the sentence to understand the word "sasquatch". When asked i f he found the passage, easy, okay, or very hard, he answered, "Okay". When asked why it was "okay" for him, he pointed to two of the difficult words, "impulsively" and "resonated" and said, "these were hard to read". In this situation, Myles did show some ability to evaluate his understanding of the passage he was reading. However, he did not attempt to ask for help to read the most difficult words, or ask what they meant. Myles would often demonstrate a lack of self-correction without the prompt of another individual. Low Prior Knowledge Bob. Bob scored 16% on this task, which means he answered 1 comprehension question correctly out of 6 questions. The correct answer was from a passage independent question, which suggests he may have known the answer prior to reading the passage. This grade 2 level passage was easily read by Bob with 100% accuracy. However, he did not seem to retain any information that he read nor be able to look back to the text for answers. This finding is similar to that of Healy et al. (1982) where the children with hyperlexia were unable to retain meaningful sentences or unable to utilize any abstract forms o f thinking. When asked about the level of difficulty of the passage using the rating scale, Bob replied, "easy, because the words were easy to read". Comprehension in Hyperlexia 82 Harvey. Harvey scored 83% on this passage. He answered 5 out of 6 of the questions correctly. The incorrect answer from the comprehension questions was a passage dependent question. Harvey read with 60% accuracy. Decoding errors, included " i f for "sip" and "snail" for "small." When asked how difficult the passage was using the rating scale, Harvey replied, "really hard, because the words were hard to read." Since the passages had to be simple for Harvey to read them (at an early K-level), it is not surprising that he scored quite high. Myles. Myles achieved a score of 50%, which was equal to 3 out of^6 correct answers. The passage was written at a grade 2 level that was easily read by Myles who achieved 100% accuracy. He was a little more adept at remembering what he had read than Bob, but Myles also limited his use of metacognitive strategies and rarely looked back to the text for answers. Myles answered 2 out of 3 passage independent questions correctly and 1 out of 3 passage dependent questions correctly. When asked to rate the level o f difficulty of this passage, he answered, "easy, because the words were easy". Nelson. Nelson scored 83% on this task. He answered 5 out of 6 of the questions correctly. The only error he made was with a passage dependent question. Nelson's reading accuracy was 64% on this passage. Decoding errors included "beak" for " b i l l " and "stop" for "slow". Despite Nelson's low attention span, he was able to answer the comprehension questions with ease. He had no problem when it came to looking back to the text for answers. When asked to rate the level of difficulty of this passage, Nelson replied, "a little hard, because o f some words I can't read". Comparison. The students with M I D did quite well (83%) on this task. A s predicted, students with hyperlexia did poorly on these passages. The question now becomes can we boost comprehension in students with hyperlexia by having them read passages for which they have high knowledge of the topic? High Prior Knowledge Bob. Bob achieved a score of 16% or 1 out of 6 for the grade 2 reading passage, and 60% or 3 out of 5 for the K-level reading passage that all the participants received. Bob read this Comprehension in Hyperlexia 83 passage with 100% accuracy. He appeared to perform better on the simpler passage that was at a much lower reading level. On the grade 1 level passage for high prior knowledge, Bob scored much better compared with the grade 2 passage. He achieved a 75%, or 6 out of 8 correct answers on this passage. One possible explanation for the higher score on the grade 1 passage would be that Bob was capable of comprehending more efficiently at a lower reading level. Harvey. Harvey achieved a perfect score or 5 out of 5 on the K-level passage. This was an easy passage for Harvey to read and comprehend, as he was very familiar with the book. Harvey decoded with 80% accuracy. His errors were all decoding errors. Myles. Myles scored 50% or 3 out of 6 on the K-level passage and 83% or 5 out of 6 on the grade 2 level passage. He decoded these passages with 100% accuracy. Myles read the same grade 1 passage about frogs as Bob. However, Myles found this task easy and achieved 100% comprehension or 8 out of 8 on this passage. It could be possible that Myles had a fairly high interest in frogs, which would explain his higher comprehension rate for the second passage. Nelson. Nelson achieved a perfect score, 100% or 5 out of 5. This passage was quite easy for Nelson as he was familiar with both the book and topic. He read with 84% accuracy on this passage. His errors were all decoding errors. Comparison. There was a lot of difference between the scores on the high and low prior knowledge tasks for Myles (50% on low prior knowledge and 83% on high prior knowledge). Bob did not have any difference in scores between the low and high prior knowledge passages (16% on both passages). However, the children with M I D achieved a score of 100% on the high prior knowledge passage versus 83% on the low prior knowledge task. This pattern of scores suggested that Bob was not using prior knowledge to aid comprehension compared to the M I D children. Similarly, Snowling and Frith (1986) reported that children with hyperlexia did not use prior knowledge when answering questions about the text. Low Interest Bob. Bob scored 33% on the low interest passage, which meant he answered 2 of the 6 comprehension questions correctly, one passage independent and one passage dependent Comprehension in Hyperlexia 84 question. He read this passage with 100% accuracy. Bob had difficulty answering most of the comprehension questions, particularly those that were passage dependent. Harvey. Harvey achieved a perfect score of 6 out o f 6. His decoding for this passage was at 54% accuracy. Decoding errors included "same" for "sam" and "black" for "ball" . He had no problem answering the questions or looking back to the text for answers. Myles. Myles scored 66% or 4 out of 6 correct responses. He decoded with 100% accuracy. Although kites were not a high interest for Myles , he appeared to know a lot about them and was able to answer all passage independent questions as well as one passage dependent question. Nelson. Nelson achieved 100% or 6 out of 6 on this passage. His accuracy score for this passage was 54%. A l l errors were decoding errors. Although he had problems decoding some of the words in the passage, he easily answered all comprehension questions. Comparison. For this task, both children with M I D demonstrated good reading comprehension regardless of interest in the topic. Myles actually did quite well on this passage. A possible explanation for this high score could be that he had quite a bit o f knowledge about kites and was using his prior knowledge to improve his comprehension. However, Bob continued his pattern o f low reading comprehension for this passage. High Interest Bob. Bob achieved a score of 66% comprehension or 4 out of 6 on this grade 6-7-level, high interest passage. He read this passage at 85% accuracy even though it contained many difficult words, such as, "conical". This was the highest comprehension score that Bob achieved despite the fact that this was the most difficult o f all the passages he had read. Bob was very interested in the topic of killer whales and highly motivated to comprehend what he was reading. He even attempted to employ some metacognitive skills, such as re-reading and looking back to the text in order to help with comprehension of the passage. Bob achieved all three passage dependent questions correctly, and 2 out of 3 of the passage independent questions correctly. Harvey. Harvey achieved 100% comprehension or 6 out of 6 on this task. He read this Comprehension in Hyperlexia 85 passage with 67% accuracy, with errors that included "red" for "green" and "blue" for "black". Again, the passage was quite simple, at K-level , and was also about cars, which highly motivated Harvey. He answered all passage dependent and independent questions correctly. Myles. Myles achieved 100% or 6 out of 6 on his grade 5-level passage. He read this passage with 100% accuracy. The passage about hockey clearly captured his full attention and increased his motivation for learning. He found the comprehension questions easy to answer. Myles was able to answer all the passage dependent and independent questions correctly. A s Hid i (1990) suggested, high interest in a topic leads to an increase in motivation for learning. Myles was far more motivated to answer the comprehension questions for this passage than any other passage he read. Additionally, as described in Baldwin et al. (1985), Myles scored better when interest and prior knowledge were high. Nelson. Nelson scored 83% or 5 out of 6 on his high interest passage. He read this passage with 77% accuracy. His decoding errors included "blue" for "black". Nelson had some difficulty decoding even the simple words in this passage. He was also quite distracted and restless this day, which could explain why his score was less than 100% on this passage. Nelson answered all three passage independent questions correctly, and only 1 of the 3 passage dependent questions correctly. Comparison. Reading comprehension for the high interest passage greatly improved for the two children with hyperlexia who scored much higher on the comprehension questions for this high interest versus the low interest passage. These findings are of significance since the reading levels of the high interest passages for the students with hyperlexia were much higher compared to the prior knowledge passages (i.e., Levels 4 and 6 vs. Level 2). Higher grade level passages were used for the pair of students with hyperlexia to test their decoding limit and demonstrate how much of an impact interest had for them on reading comprehension. Somehow their interest for the subject forced their attention to the task and hence improved their comprehension levels. There was no difference in comprehension on the low and high interest passages for the s.tudents with M I D . However, the passages for the M I D children had to be quite Comprehension in Hyperlexia 86 simple in order for them to be able to decode the words. Therefore, the simplicity of the passages could have made it easy for them regardless o f their interest level. These results were similar to the findings reported by De Sousa and Oakhill (1996), who found that good comprehenders performed well on both high and low interest tasks, where the poor comprehenders performed significantly better on the high interest passage. Summary Upon examination of the results, a more detailed picture of reading comprehension in children with hyperlexia and children with M I D is more apparent. In Phase 1 of the investigation, the two students with hyperlexia illustrated their word reading abilities despite their low cognitive levels. In comparison, the two participants with M I D appeared to have great difficulty with word reading. However, in Phase 2, the M I D students demonstrated higher comprehension than the students with hyperlexia. During Phase 2 of the study, Bob and Myles demonstrated little or no understanding of the purpose for reading. Neither boys employed self-correction techniques, nor did they stop for errors unless their decoding limits were challenged. High prior knowledge did not have a positive effect on the reading comprehension for Bob, but did show some effect for Myles ' comprehension. However, high interest had a positive effect on the comprehension of both these students. Both Bob and Myles scored much higher on their reading comprehension questions for the high interest passages. Neither students with hyperlexia illustrated self-awareness o f their reading or comprehension abilities. These students consistently reported the passages they read as "easy", yet they could not answer simple questions about the passages. These findings are important since one would expect all four of the participants would have similar reading comprehension abilities because their cognitive and language levels are all equivalent. However, this was not the case at all . The students with M I D had good comprehension abilities for their cognitive level, whereas the students with hyperlexia did not have good comprehension abilities for their cognitive level. In Phase 2, the boys with M I D displayed very good comprehension for their decoding level. Both Harvey and Nelson were consistent in catching the errors embedded in the passages. Comprehension in Hyperlexia 87 Furthermore, they had quite good comprehension despite their decoding problems. Both boys applied metacognitive thinking and appeared to understand the purpose for reading. They were consistent in rating the passages accurately according to their ability to understand them. Comprehension in Hyperlexia 88 C H A P T E R V Conclusion/Discussion Despite the number of studies over the years that have examined the reading processes of children with hyperlexia, there has been little agreement concerning the underlying comprehension problems in these children (Aram, 1997). Additionally, there has been much lacking in the literature regarding several important aspects of reading comprehension, such as prior knowledge, interest, and metacognition. Therefore, the purpose of this study was to examine these areas of reading comprehension in order to develop a more complete picture o f the comprehension deficit in children with hyperlexia. This study compared these aspects of reading comprehension across cases of children with hyperlexia and children with M I D . Few studies have matched students with hyperlexia with students of the same age, IQ, and grade but who do not have hyperlexia. Therefore, this study, with its matched pairs of participants, further develops an understanding of the unique strengths and weakness in reading for children with hyperlexia. Matched participants allowed for examination and comparison of distinct aspects of reading comprehension in hyperlexia versus M I D . This permitted a much stronger conclusion to be drawn regarding the unique comprehension deficit in children with hyperlexia. Additionally, due to the number of studies with such large variance of ages and IQs in their participants, this study focused on students between the ages of 7 - 9 years, all in grades 2 and 3, and all within the mild to moderately intellectually delayed range o f IQ as measured by the WISC-III. In short, in this study, participants with hyperlexia met Healy et al.'s (1982) criteria of: (a) spontaneous and intense early interest in letters and words prior to the age o f 5 years, coupled with, (b) significantly disordered language and cognitive development, and (c) word recognition (decoding) well advanced over other cognitive and linguistic abilities. In Phase 1 of this study, the participants were assessed on their word reading and pseudo word reading. This was done to confirm the status o f the children as either hyperlexic or M I D , as well as to provide a standardized measure of reading comprehension in all four participants. For Comprehension in Hyperlexia 89 the second phase of the study, reading comprehension was further examined. In particular, assessments of students' ability to detect errors, self-correct, use prior knowledge and benefit from high interest were administered. The findings are summarized below according to each thesis question posed in the first chapter. 1) Is comprehension commensurate with cognition or language abilities (i.e., is low cognition and low language ability sufficient as an explanation for hyperlexics poor comprehension)? This question is more easily addressed in two parts. First, are the verbal IQ scores and PPVT-III scores of both pairs of students equivalent? Second, are low cognition and low language levels sufficient to explain the low comprehension in students with hyperlexia? To answer the first part of the question, IQ and PPVT-III scores must be evaluated with caution since IQ and receptive vocabulary scores are not directly comparable. However, when evaluating standard scores for both the WISC-III and PPVT-III , it would be expected that a child with low verbal IQ would also have low PPVT-III scores. This appears to be the case for all four students. For example, Bob scored a 66 (SS) or an age-equivalent of 5.1 years on the P P V T . On the WISC-III his V I Q was 63, which places him in the moderately intellectually delayed range. Myles scored 84 (SS) on the P P V T , which gives him an A E o f 7 years o f age. On the WISC-III, his V I Q was 73, which places him in the mildly intellectually delayed category. Harvey's score on the P P V T (SS of 60) corresponded to an age equivalent score o f 3.11. His V I Q on the W I S C -III was 55, which places him in the moderately intellectually delayed category. Nelson had a V I Q score of 72 (SS), which places him in the mildly intellectually delayed category. His P P V T score of 76 (SS) gives him an A E score of 6.4. These standard scores are very similar and close to equivalency in all four students. Part 2 of the question examines low cognition and language as a sufficient explanation for low comprehension in children with hyperlexia. It would be expected that, since all four students were matched according to their cognitive ability, and all o f them had language levels equivalent to their cognitive levels, that their level of reading comprehension would also be the Comprehension in Hyperlexia 90 same. However, this was not the case. The two students with M I D achieved much higher comprehension levels compared to the students with hyperlexia. Both Bob and Myles had poor reading comprehension on all passages except the high interest passage. Conversely, Harvey and Nelson had excellent reading comprehension on all reading passages. Therefore, although it would be expected that all four of these students would all perform equally well on reading comprehension, this did not happen. These conclusions suggest that low cognition and low language levels are not a sufficient explanation for the comprehension problems in children with hyperlexia. Snowling and Firth (1986) similarly proposed that low language levels on their own could not account for the poor reading comprehension in children with hyperlexia. 2) Do children with hyperlexia comprehend better when prior knowledge is high? Bob did not show improvement in reading comprehension when prior knowledge was high. Myles did improve reading comprehension for the high prior knowledge passage compared to the low prior knowledge passage. Myles appeared to utilize his prior knowledge to answer the comprehension questions, whereas Bob did not. The students with M I D slightly improved their comprehension with high prior knowledge. Both Harvey and Nelson utilized their prior knowledge to answer the comprehension questions. In regards to the students with M I D , these findings mimic Carr and Thompson (1996), who reported that children with L D benefited from having high prior knowledge when reading passages as it improved their reading comprehension. 3) Do they comprehend better when interest is high? There was an effect of high interest on the comprehension of both students with hyperlexia. The students with hyperlexia had better comprehension on more difficult reading passages with high interest compared to their low interest passages. For example, Bob's comprehension improved greatly on a high interest passage where he scored 4/6, compared to only 2/6 on the low interest passage. Myles ' s scores also improved. He achieved 6/6 on his high interest passage and 4/6 on the low interest passage. This suggests that, when motivation to comprehend is very high, these students can focus enough to understand what they are reading. Paris et al. (1983) described the need for both knowledge and motivation in order to use various Comprehension in Hyperlexia 91 cognitive strategies to improve comprehension. Baldwin et al. (1985) found that when interest in the reading topic was high, the comprehension of males in their study was significantly higher compared to low interest topics. In contrast, interest did not benefit students with M I D compared to the hyperlexic participants. Harvey scored 6/6 on both high and low interest passages, and Nelson scored 5/6 and 6/6, on these passages, respectively. The reading comprehension o f the two students appeared to be quite good regardless of their interest in the topic. 4) Do children with hyperlexia monitor their own comprehension? There was a disparity between the students with hyperlexia with regard to comprehension monitoring. Myles was inconsistent in monitoring his reading comprehension. There were several instances in which he stopped at a difficult word and asked, "What does this say?" However, he preferred to skip the words he did not know or make up words to replace them. Bob would only monitor his own comprehension when he was prompted to look back in the passage. In the error detection #1 task, Bob stopped for only 1 of the 8 errors. He read, made non-meaningful substitutions, or skipped over all the other errors in the passage. On the self-evaluation scale, Bob commented that the passage was "easy" to read and said that it was due to the easy words. However, when asked what the word "r#+k" meant, he could not give an answer. For the error detection #2 passage, Bob stopped for only 2 out of 10 errors. He did not ask about the first two most difficult words, he just skipped over them. He did ask what the third very difficult word meant. Bob did not pay attention to punctuation. For example, he would read right through periods and create one long run-on sentence. A lack o f comprehension monitoring in children with hyperlexia has been consistently reported in the literature. For example, Goldberg and Rothermel (1984) found that their participants with hyperlexia did not even notice the change in punctuation in altered sentences and paragraphs. In contrast, both students with M I D demonstrated comprehension monitoring when reading. For example, Harvey did show signs of monitoring his own comprehension. He was aware of when he could not read the words and he would slow down and sound out the word. He Comprehension in Hyperlexia 92 would rarely skip over a word. In both error detection tasks, Harvey was able to catch all but one error. The only error he missed was the non-word "keek." Nelson also employed comprehension monitoring tactics. He would slow down and strive to sound out the words he could not read or he would ask what a word meant when he did not know it. Nelson was able to discover all errors except one in both of the error detection tasks. 5) Do children with hyperlexia self-correct? The students with hyperlexia were not consistent with using self-correction strategies. They did some self-correction at the word level when they were not able to pronounce the word. However, self-correction never occurred at the sentence level or paragraph level. These findings are comparable to those of Healy et al. (1982) who reported that none of their participants with hyperlexia employed self-correction at the sentence level. They did self-correct, but it was only at the word-level, when pronunciation of a word was not possible that self-correction would occur. Similarly, Bob only demonstrated self-correction at the word level. However, he ignored most of the errors in the passages. If he could not decode a word he would either make up a word or skip the word. Myles did show some use o f self-correction at the word level. However, he frequently skipped words he could not decode or would make up his own word as a replacement. It was evident that the students with hyperlexia did not have well developed metacognitive skills. Conversely, the two M I D students frequently made use of self-correction tactics. Harvey would read over words he could not understand and would correct words that he had read incorrectly immediately after making the error. Nelson was not as consistent as Harvey at using self-correction techniques. However, he always attempted to sound out the words he could not read and would ask what for help i f he could not read it after trying. Clearly, Harvey and Nelson employed metacognitive tactics when reading for comprehension. Both students with M I D knew the purpose for reading was to get meaning from the reading passage. Limitations A s in all research, this study had several limitations. Due to the small sample, the results are not generalizable to all children with hyperlexia or all children with M I D . In order for any Comprehension in Hyperlexia 93 generalizations to be made across the hyperlexic population, a larger scale study or more replications would be necessary. However, this could be difficult to achieve since children with hyperlexia are not easily identified or diagnosed. Another limitation of this study was the use of the reading subtest of the brief form of the K - T E A to assess the participant's reading comprehension. The brief form of the reading subtest of the K - T E A is weak in its ability to truly assess a child's comprehension. The K - T E A confounds decoding and comprehension. The test is organized in a way that a child who is good at decoding words can achieve a high "reading" score, which includes decoding and comprehension. However, there is no real proof o f the children's actual level of comprehension. Children who are not good decoders do not get past the decoding of single words on the test to the comprehension items. A much better assessment of reading comprehension is found in the Gates MacGinit ie Reading Test, which separates decoding from reading comprehension immediately. However, it is quite expensive to acquire and has not been widely used in other studies with hyperlexic children. It is also possible that students with M I D might not be able to read some of the words on the Gates MacGinit ie Reading Test. Perhaps, the next investigation into this area should use multiple reading and listening comprehension assessments, in order to get a more accurate description of reading comprehension in these children. Additionally, since the investigator designed many o f the comprehension measures there are potential weaknesses, such as the variance in length for passages between the pairs o f participants. The M I D pair had very weak decoding skills and therefore the passages for them were shorter in length compared to the passages for the hyperlexia pair. This could have made it easier for the M I D group to achieve higher comprehension levels compared to the hyperlexia group. Similarly, choosing graded passages for the participants at times called for a judgment to be made by the investigator. However, most o f the passages were borrowed from informal reading inventories then changed slightly to suit the particular assessment and were most likely at the appropriate grade level. Comprehension in Hyperlexia 94 Future Directions This investigation points to several avenues for future research in the area of hyperlexia and reading comprehension. Initially, a replication and extension o f this study should be completed with more participants and careful attention paid to the selection and design of assessments. It is important to be confident that the results reported in this study are not due to use o f particular assessment instruments or investigator bias. Furthermore, there is need for a more detailed examination of the semantic processing of children with hyperlexia in order to fully understand the severity of their comprehension deficit. Additionally, a study is required that wi l l utilize the present information about the comprehension deficits in these children and implement a program using metacognitive learning techniques. In addition, the study should also consider each child's interests and what motivates them to learn. Finally, it would behoove educational research to examine the improvements or lack of improvements in reading comprehension when individualized lesson plans are created and used for each child based on their interests and choices. Conclusions In review of the thesis questions in this study, children with hyperlexia appear to have a distinct comprehension problem that is different from children with mild intellectual difficulties. A s this study illustrated, comprehension in the two children with hyperlexia is not as skilled as their matched participants with developmental delays. This suggests that the comprehension deficit in children with hyperlexia is a phenomenon of the disorder and not of the cognitive delay. To answer the main thesis question, "what accounts for the poor comprehension in children with hyperlexia?" it is most probable that a multitude of variables, such as poor language abilities, delayed cognition, lack of ability to monitor their own comprehension, and difficulty activating schema combine to produce the reading comprehension problems children with hyperlexia experience. Snowling and Frith (1986) proposed that poor comprehension in children with hyperlexia is not due to their low language levels or poor word knowledge. The results from this study Comprehension in Hyperlexia 95 appear to confirm these author's conclusions. The direction for future studies in the area of hyperlexia should further examine how much reading comprehension is affected by language level. Although no definitive statements can be made regarding the specific underlying deficits in comprehension in these children, it is apparent that studies such as this one help bring us closer to a more complete understanding of the phenomenon o f hyperlexia. Comprehension in Hyperlexia 96 References Aaron, P. G . , Frantz, S. S., & Manges, A . R. (1990). Dissociation between comprehension and pronunciation in dyslexic and hyperlexic children. Reading and Writing: An Interdisciplinary Journal, 2, 243-264. Afflerbach, P. P. ((1990)). The influence of prior knowledge on expert readers' main idea construction strategies. Reading Research Quarterly, 25, 31-46. Alvermann, D . E . , Smith, L . C , & Readence, J. E . (1985). Prior knowledge activation and the comprehension of compatible and incompatible text. Reading Research Quarterly, 20, 420-436. Anderson, R. C. (1977). The notion of schemata and the education enterprise. In R. C . Anderson, F. J. Spiro, & W . E . Montague (Eds.), Schooling and the acquisition of knowledge, (pp. 415-431). Hillsdale, N J : Erlbaum. Anderson, R. C. (1994). Role of readers schema in comprehension, learning, and memory. In R. B . Ruddell, M . R. Ruddell, and H . Singer (Eds.), Theoretical Models and Processes of Reading (4 t h edition), (pp. 469-482), Newark, D W . Anderson, J. R. (1990). Cognitive psychology and its implications. (2 n d edition). New York: Freeman. Anderson, R. C , & Pearson, P. D. (1984). A schema-theoretic view of basic processes in reading comprehension. In P. D . Pearson, R. Barr, M . L . Kami l , & P. Mosenthal (Eds.), Handbook of Research on Reading, (pp. 255-291). New York: Longman. Anderson, R. C , & Pichert, J. W . (1978). Recall o f previously unrecallable information following a shift in perspective. Journal of Verbal Learning and Verbal Behavior, 17, 1 -12. Anderson, R. C , Spiro, R. J., & Anderson, M . C. (1978). Schemata as scaffolding for the representation of information in connected discourse. American Educational Research Journal, 15 (3), 433-440. Anderson, R. C , Reynolds, R. E . , Schallert, D . E . , & Goetz, E . T. (1977). Frameworks for Comprehension in Hyperlexia 97 comprehending discourse. American Educational Research Journal, 14 (4), 367-381. Aram, D . (1997). Hyperlexia: Reading without meaning in young children. Topics in Language Disorders, 17 (3), 1-1,3. Aram, D. , & Healy, J. M . (1988). Hyperlexia: A review of extraordinary word recognition. In L . Obler & D . Fein (Eds.), Neuropsychology of Talent, (pp. 70-102). New York : Gilford Press. Aram, D. , Rose, D . F. , & Horwitz, S. J. (1984). Hyperlexia: Developmental reading without meaning. In R. N . Malatesha & H . A . Whitaker (Eds.), Dyslexia: A Global Issue. The Hague, The Netherlands: Martinus Nijhoff Publishers. Asher, S. R. (1979). Referential Communication. In G.J . Whitehurst, & B . J . Zimmerman (Eds.), The Functions of Language and Communication, (pp. ). New York. Academic Press. Baker, L . , & Brown, A . L . (1984). Metacognitive Skills and Reading. In P. B . Pearson, R. Barr, M . L . K a m i l , & P. Mosenthal (Eds.), Handbook of Reading Research, (pp. 353-394). White Plains, N Y : Longman. Baldwin, R. S., Peleg-Bruckner, Z . , & McClintock, A . H . (1985). Effects of topic interest and prior knowledge on reading comprehension. Reading Research Quarterly, 20, 497-504. Bishop, D . V . M . (1983). Test for Reception of Grammar. Manchester: University of Manchester. Boblett, F. C. (1932). Remembering. Cambridge: The Cambridge University Press. Borkowski, J. G . , Johnston, M . B . , & Reid, M . K . (1987). Metacognition, motivation, and controlled performance. In S. J. Ceci (Ed.), Handbook of Cognitive, Social, and Neuropsychological Aspects of Learning Disabilities (pp. 147-173). Hillsdale, N J : Lawrence Earlbaum Associates. Bransford, J. D. , & Johnson, M . K . (1972). Contextual prerequisites for understanding. Some investigations of comprehension and recall. Journal of Verbal Learning and Verbal Behavior, 11, 717-726. Brown, A . L . (1978). Knowing when, where, and how to remember: A problem of metacognition. In R. Glaser (Ed.), Advances in Instructional Psychology (Vol.1), (pp. 77-Comprehension in Hyperlexia 98 165). Hillsdale, N J : Lawrence Erlbaum Associates. Brown, A . L . (1980). Metacognitive development and reading. In R. J. Spiro, B . C. Bruce, & W . F. Brewer (Eds.), Theoretical Issues in Reading Comprehension: From Research to Practice, (pp. 453-481). Hillsdale, N J : Lawrence Erlbaum Associates. Carr, S. C , & Thompson, B . (1996). The effects of prior knowledge and schema activation strategies on the inferential reading comprehension o f children with and without learning disabilities. Learning Disability Quarterly, 19, 48-61. Cobrinik, L . (1974). Unusual reading ability in severely disturbed children. Journal of Autism and Childhood Schizophrenia, 4 (2), 163-175. Cobrinik, L . (1982). The performance of hyperlexic children on an "incomplete words" task. Neuropsychologia, 20 (5), 569-577'. Cossu, G . & Marshall, J. C . (1986). Theoretical implications of the hyperlexia syndrome. Two new Italian cases. Cortex, 22, 579-589. Cross, D . R., & Paris, S. G . (1988). Developmental and instructional analyses of children's metacognition and reading comprehension. Journal of Educational Psychology, 80, 131-142. De Sousa, I., & Oakhill , J. (1996). Do levels of interest have an effect on children's comprehension monitoring performance? British Journal of Educational Psychology, 66, 471-482. Durrell, D . D . (1955). Durr ell Analysis of Reading Difficulty, New Edition. New York: Harcourt, Brace. Elliot, D . E. , & Needleman, R. M . (1976). The syndrome of hyperlexia. Brain and Language, 3, 339-349. Flavell, J. H . (1978). Metacognitive development. In J. M . Scandura & C. J. Brainerd (Eds.), Structural/process theories of complex human behavior, (pp. ) Alphen a. d. Rijn, The Netherlands: Sijthoff & Noordhoff. Flavell, J. H . (1979). Metacognition and cognitive monitoring: A new'area of cognitive-Comprehension in Hyperlexia 99 developmental inquiry. American Psychologist, 34 (10), 906-911. Fontanelle, S., & Alarcon, M . (1982). Hyperlexia: Precocious word recognition in developmentally delayed children. Perceptual and Motor Skills, 55, 247-252. Garner, R. (1980). Monitoring of understanding: A n investigation of good and poor readers' awareness of induced miscomprehension of text. Journal of Reading Behavior, 12, 55-63. Garner, R. (1987). Metacognition and Reading Comprehension. Norwood, N J : Ablex Publishing. Garner, R. (1994). Metacognition and executive control. In R. B . Ruddell, M . R. Ruddell, & H . Singer (Eds.), Theoretical Models and Processes of Reading. (4 t h ed.). (pp. 715 - 732). Newark, D E : International Reading Association. Garner, R., & Kraus, C . (1982). Good and poor comprehender differences in knowing and regulating reading behaviors. Educational Research Quarterly, 6, 5-12. Garner, R., & Reis, R. (1981). Monitoring and resolving comprehension obstacles: A n investigation of spontaneous text lookbacks among upper-grade and poor comprehenders. Reading Research Quarterly, 16, 569-582. Garner, R., & Taylor, N . (1982). Monitoring o f understanding: A n investigation of attentional assistance heeds at different grade and reading proficiency levels. Reading Psychology, 3, 1-6. Garner, R., Wagoner, S., & Smith, T. (1983). Externalizing question-answering strategies of good and poor comprehenders. Reading Research Quarterly, 18, 439-457. Glosser, G . , Friedman, R. B . , & Roeltgen, D . P. (1996). Clues to cognitive organization of reading and writing from developmental hyperlexia. Neuropsychology, 10 (2), 168-175. Goldberg, T. E . , & Rothermel, R. D . (1984). Hyperlexic children reading. Brain, 107, 759-785. Goldman, R., Fristoe, M . , & Woodcock, R. W . (1974). GFWSound-Symbol Tests. Circle Pines, M N : American Guidance Service. Goodman, J. (1972). A case study o f an "autistic-savant": Mental function in the psychotic child with markedly discrepant abilities. Journal of Child Psychology & Psychiatry, 13, 267-278. Comprehension in Hyperlexia 100 Goodman, Y . A . , & Burke, C . L . (1972). Reading miscue inventory. New York : MacMi l l an . Healy, J. M . , Aram, D . M . , Horwitz, S. J., & Kessler, J. W . (1982). A study of hyperlexia. Brain and Language, 17, 1-17. H id i , S. (1990). Interest and its contribution as a mental resource for learning. Review of Educational Research, 60, 549-571. Hiskey, M . S. (1966). Hiskey-Nebraska Test of Learning Aptitude: Manual. Lincoln, N E : Union College Press. Huttenlocher, R. R., & Huttenlocher, J. A study o f children with hyperlexia. Neurology, 23, 1107-1116. Jastak, J. F., & Jastak, S. R. (1978). Wide Range Achievement Tests. Wilmington, D E : Guidance Associates. Kanner, L . (1943). Autistic disturbances of affective contact. Nervous Child, 2, 217-250. Karlsen, B . , Madden, R., & Gardner, E . F. (1976). Stanford Diagnostic Reading Test. New York : Harcourt Brace Jovanovich. Kinnunen, R., & Vauras, M . (1995). Comprehension monitoring and the level of comprehension in high and low achieving primary school children's reading. Learning and Instruction, 5, 143-165. Langer, J. A . , & Nicol ich , M . (1981). Prior knowledge and its relationship to comprehension. Journal of Reading Behavior, 13, 373-379. Leslie, L . , & Caldwell , J. (1990). Qualitative Reading Inventory. Glenview, IL: Scott, Foresman. Mehegan, C. C , & Dreifus, R. E . (1972). Hyperlexia: Exceptional reading ability in brain-damaged children. Neurology, 22 (11), 1105-1 111. Montague, M . , Maddux, C. D . , & Dereshiwsky, M . I. (1990). Story grammar and comprehension and production o f narrative prose by students with learning disabilities. Journal of Learning Disabilities, 23 (3), 190-197. Munroe, M . (1932). Children who cannot read. Chicago: University o f Chicago Press. Myers, M . , & Paris, S. G . (1978). Children's metacognitive.knowledge about reading. Journal of Comprehension in Hyperlexia 101 Educational Psychology, 70 (5), 680-690. Neuman, R. S. (1990). Children's help-seeking in the classroom: The role of motivational factors and attitudes. Journal of Educational Psychology, 82, 71 -80. Newcomer, P. L . , & Hammil l , D . O. (1977). The test of language development. Austin, T X : Empiric Press. Palmer, D . J . , & Goetz, E.T. (1988). Selection and the use of study strategies: The role of the studier's beliefs about self and strategies. In C E . Weinstein, E . T. Goetz, & P. A . Alexander (Eds.), Learning and study strategies: Issues in assessment, instruction, and evaluation, (pp.41-61). San Diego, C A : Academic Press. Paris, S. G . , & Jacobs, J. E . (1984). The benefits of informed instruction for children's reading awareness and comprehension skills. Child Development, 55, 2083-2093. Paris, S. G . , & Lindauer, B . K . (1982). The development of cognitive skills during childhood. In Benjamin B . Wolman (Ed.), Handbook of Developmental Psychology. Englewood Cliffs, N J : Prentice-Hall, Inc. Paris, S. G . , & Myers, M . (1981). Comprehension monitoring, memory, and study strategies in good and poor readers. Journal of Reading Behavior, 13 (1), 5-22. Paris, S. G . , & Oka, E . R. (1986). Children's reading strategies, metacognition, and motivation. Developmental Review, 6, 25-56. Paris, S. G . , Cross, D . R., & Lipson, M . Y . (1984). Informed strategies for learning: A program to improve children's reading awareness and comprehension. Journal of Educational Psychology, 76(6), 1239-1252. Paris, S. G . , Lipson, M . Y . , & Wixson, K . K . (1983). Becoming a strategic reader. Contemporary Educational Psychology, 8, 293-316. Paris, S. G . , Saarnio, D . A . , & Cross, D . R. (1986). A metacognitive curriculum to promote children's reading and learning. Australian Journal of Psychology, 38, 107-123. Paris, S. G. , Wasik, B . A . , & Turner, J . C . (1991). The development of strategic readers. In R. Barr, M . L . Kami l , P. Mosenthal, & P. D- Pearson (Eds.), Handbook of Reading Comprehension in Hyperlexia 102 Research. New York, N Y : Longman. Patti, P. J., & Lupinetti, L . (1993). Br ie f report: Implications of hyperlexia in an autistic savant. Journal of Autism and Developmental Disorders, 23 (2), 397-405. Perry, N . (1998). Young children's self regulated learning and contexts that support it. Journal of Educational Psychology, 90(4), 715-729. Pressley, M . , Borkowski, J. G . , & O'Sullivan, J. (1985). Children's metamemory and the teaching of memory strategies. In D . L . Forrest-Pressley, D . MacKinnon, and T. G . Waller (Eds.), Metacognition, Cognition and Human Performance (pp. 111-153). New York: Academic Press. Richman, L . C , & Kitchell , M . M . (1981). Hyperlexia as a variant of a developmental language disorder. Brain and Language, 12, 203-212. Rumelhart, D . E . (1981). Schemata: The building blocks of cognition. In J. T. Guthrie (Ed.), Comprehension and Teaching: Research Reviews, (pp. 3-26). Newark, D E : International Reading Association. Rumelhart, D . E . , & Ortony, A . (1977). The representation of knowledge in memory. In R. C. Anderson, R. J. Spiro, & W . E . Montague (Eds.), Schooling and the acquisition of knowledge, (pp. 99-136). Hillsdale, N J : Lawrence Erlbaum Associates. Seymour, P. H . K . , & Evans, H . M . (1992). Beginning reading without semantics. A cognitive study of hyperlexia. Cognitive Neuropsychology, 9 (2), 89-122. Shirey, L . L . , & Reynolds, R. E . (1988). Effects of interest on attention and learning. Journal of Educational Psychology, 80(2), 159-166. Siegel, L . S. (1984). A longitudinal study of a hyperlexic child: Hyperlexia as a language disorder. Neuropsychologia, 22 (5), 577-585. Silberberg, N . E . , & Silberberg, M . C. (1967). Hyperlexia: Specific word recognition skills in young children. Exceptional Children, 34 (1), 41-42. Silberberg, N . E . , & Silberberg, M . C. (1968-69). Case histories in hyperlexia. Journal of School Psychology, 7 (1), 3-7. Comprehension in Hyperlexia 103 Silberberg, N . E., & Silberberg, M . C. (1971). Hyperlexia; The other end of the continuum. The Journal of Special Education, 5 (3), 233-242. Snider, V . E. (1989). Reading comprehension performance of adolescents with learning disabilities. Learning Disability Quarterly, 12, 87-96. Snowling, M , & Frith, U . (1986). Comprehension in "hyperlexic" readers. Journal of Experimental Child Psychology, 42, 392-415. Temple, C. M . (1990). Auditory and reading comprehension in hyperlexia. Semantic and syntactic skills. Reading and Writing. An Interdisciplinary Journal, 2, 297-306. Terman, L . M . , & Merr i l l , M . A : (1960). Stanford-Binet intelligence scale: Manual for the third revision, for L-M. Boston: Houghton Mif f l in Co. Tirosh, E., & Canby, J. (1993). Autism with hyperlexia. A distinct syndrome? American Journal on Mental Retardation, 98 (1), 84-92. Tobias, S. (1994). Interest, prior knowledge, and learning. Review of Educational Research, 64, 37-54. Turner, J. C. (1995). The influence of classroom contexts on young children's motivation for literacy. Reading Research Quarterly, 30, 410-441. Wagoner, S. A . (1983). Comprehension monitoring: What is it and what we know about it. Reading Research Quarterly, 28, 328-346. Wechsler, D . (1949). Wechsler intelligence scale for children. New York : The Psychological Corporation. Wechsler, D . (1967). Wechsler preschool and primary scale of intelligence. New York : The Psychological Corporation. Wechsler, D . (1974). Wechsler intelligence scale for children - revised. New York: The Psychological Corporation. Welsh, M . C , Pennington, B . F., & Rogers, S. (1987). Word recognition and comprehension skills in hyperlexic children. Brain and Language, 32, 76-96. Whitehouse, D . , & Harris, J. C . (1984). Hyperlexia in infantile autism. Journal of Autism and Comprehension in Hyperlexia 104 Developmental Disorders, 14 (3), 281-289. Winne, P. H . (1998). Experimenting to bootstrap self-regulated learning. Journal of Experimental Psychology, 89 (3), 397-410. Wong, B . Y . L . (1991). The relevance o f metacognition to learning disabilities. In B . Y . L . Wong (Ed.), Learning About Learning Disabilities, (pp. 231-258). San Diego, C A : Academic Press. Wong, B . Y . L . , & Jones, W . (1982). Increasing metacomprehension in learning-disabled and normally-achieving students through self-questioning training. Learning Disability Quarterly, 5, 228-240. Wong, B . Y . L . , & Wong, R. (1986). Study behavior as a function o f metacognitve knowledge about critical task variables: A n investigation of above average, average and learning disabled readers. Learning Disabilities Research, 1 (12), 101-111. Woodcock, R. W . (1973). Woodcock Reading Mastery Tests. Circle Pines, M N : American Guidance Service. Worthy, J., & Invernizzi, M . A . (1992). Linking reading with meaning: A case study o f a hyperlexic reader. Journal of Reading Behavior, 27 (4), 585-603. Zimmerman, B . J. (1990). Self-regulated learning and academic achievement: A n overview. Educational Psychologist, 25, 3-17. Comprehension in Hyperlexia 106 Appendix A C O N S E N T F O R M I have read the letter describing Michele Lester's research for her Master's thesis titled, "Comprehension in Children with Hyperlexia," and the nature of my child's participation in it. 1 have discussed the project with my child and we understand that my child's participation is voluntary and confidential, and that s/he may withdraw from the project at any time without negative consequences. M y signature indicates my desire for to participate in the project. Signature: Date: I do not wish to participate in this project. Signature: Date: In providing my signature above, I acknowledge that I have read the consent form and have kept a copy for my personal records. I would like to receive a summary o f the results of Michele Lester's project. Name: Address: Comprehension in Hyperlexia 107 Appendix B Error Detection #1 (Hyperlexia) The Desert: What Lives There Grd 2 Passage The desert is a place that gets very little rainfall. The ground is often s%*#y and rocky. When the sun beats down, the sand and r+#k grow hot and dry. It is hard to imagine that a place like this is f*%l of living things. A l l living things need food, w*A$r, and some kind of shelter to survive. Some plants and animals are well suited to survive in the desert. They can l%A*e off the food, water, and shelter that are there. The cactus is one kind of plant that is s*A#d to survive in the desert. The cactus has a special way of g*%$Ag water in the dry desert soil. It spreads its roots out close to the top of the ground. When rain comes, the cactus r*%ts soak up the water quickly before it drains deep into the sand. Comprehension in Hyperlexia 108 Appendix B Questions for the Desert Passage Passage Independent Questions 1) Is the desert a hot place? 2) Does it snow in the desert? 3) Is a cactus soft or prickly to touch? Passage Dependent Questions 1) What is the ground like in the desert? 2) What do all living things need to survive? 3) Do many living things live in the desert? Comprehension in Hyperlexia 109 Appendix C Error Detection #1 (MID) M y Cat Can Where can Cat run? Cat can %$&* to the mat. Where can Cat #$**? Cat #%&* run to the sac. Cat can ~*%* to the cap. Cat can run up, up, up to the *#$*. Comprehension in Hyperlexia 1 1 0 Appendix C Questions for Error Detection #1 Passage Independent Questions 1. How many legs do cats have? 2. What sounds do cats make? 3. Can cats run fast? Passage Dependent Questions 1. What did cat run to the first time in the story? 2. Where did cat run to the second time in the story?: 3. D i d cat run up, up, up or down, down, down? Comprehension in Hyperlexia 111 Appendix D Error Detection #2 (Hyperlexia) Where is my Friend? Harriet was looking for her friend. She climbed up a toog. She came dute again. Harriet looked burpole two trees. So Harriet went toolum a gate. She looked umkoop a rock. She flew oolum a hill. She went into a ceer. Harriet climbed on a camsuch to look for her fiplid. Then she goor off amelop. Then Harriet looked again, and tigop was her friend, right in front of her nose. Comprehension in Hyperlexia 11 Appendix D Comprehension Questions for Error Detection #1 Questions for Harriet Passage Passage Independent Questions 1) What is a friend? 2) If you had to look for you friend, does that mean the friend was lost? 3) If you had to look outside for your friend, where would you look? Passage Dependent Questions 1) Who was Harriet looking for? 2) Where did Harriet climb? 3) What did Harriet get on top of? Comprehension in Hyperlexia 113 Appendix E Error Detection #2 (MID) Where is it? Where is the x ant? Why can't we keek it? The tant toop in the hill . Where is the fox? Why can't we geek it? The tox went in the moog. Comprehension in Hyperlexia 114 Appendix E Comprehension Questions for Error Detection #2: M I D Passage Independent Questions: 1) Where do ants live? 2) Where do foxes live? 3) What are logs made from? Passage Dependent Questions: 1) What did the ant go in? 2 ) Where did the fox go? 3) Can we see the ant or the fox? Comprehension in Hyperlexia 115 Appendix F Metacognitive Rating Scale (Hyperlexia and M I D ) (a)easy f (b)a bit difficult, but o.k. (c) really hard Comprehension in Hyperlexia 116 Appendix G Self-Correction Task: Difficult Vocabulary (Hyperlexia) f w c % The Three Beavers Brown Once there were three Beavers Brown who lived in a chasm in the high mountains. A n unfathomable creek ran through the valley. One day the three Beavers Brown adjudicated to build a new lodge for the winter. They could discern some beautiful birch trees on the other side of the creek, just perfect for building a lodge. First Little Beaver Brown gamboled into the valley. Slip, slip, slap, slap went his tiny tail as he began to swim. Impetuously a voice reverberated like thunder. "Who's that swimming across my creek?" "It's only I, Little Beaver Brown. I want to get to the birch trees on the other facet o f the creek." Out from his cave in the rocks emerged the Sasquatch. Comprehension in Hyperlexia 117 Appendix G Comprehension Questions for Difficult Reading Task Passage Independent Questions 1. What color is a beaver? 2. Where do beavers live? 3. Do Beavers have big or small teeth? Passage Dependent Questions 1. The Beavers Brown lived in a chasm in the mountains, what does that look like? 2. Tell me what the Beavers saw across the creek? 3. What happened at the end of the story? o i Comprehension in Hyperlexia 11 Appendix H Prior Knowledge Reading Task (Hyperlexia) Freddie the Frog Freddie is a frog. He lives in the water. He likes to swim. He also likes to make loud noises - croak, croak! He likes to sit on a rock in the sunshine. He likes to eat flies. Freddie swims in the water. When he is hungry he eats a fly or bug. At night he goes to sleep on a log. When Freddie was a baby frog he was called a tadpole. Tadpoles are baby frogs that grow up to be grown-up frogs, like Freddy. Comprehension in Hyperlexia 119 Appendix H Prior Knowledge Task Comprehension Questions Questions asked before reading the passage - passage independent 1) What color are frogs? 2) What do frogs do? 3) Where do frogs live? Questions asked after reading the passage - passage dependent. 1) What is the frog's name in the story? 2) What sound did he make? 3) Where did he live? 4) Where did he like to sit and get some sunshine? 5) What are tadpoles? Comprehension in Hyperlexia 120 Appendix I High Interest Passage: Myles Passage (Hyperlexia) Ice Hockey Ice Hockey is a game played between two teams. It is played indoors or outdoors on an ice rink. The players wear special uniforms and wear pads and helmets to protect their heads. Jack is on a hockey team named the Sharks. The Sharks are going to play against another team called the Lions. The game is played with a rubber puck and a special stick. The players use their sticks to get the puck into the other team's net. The sticks are made o f wood or medal and have a curved blade at the bottom. The goaltender on the Sharks team, Sam, has a special stick that is heavier and has a wider blade than the other player's sticks. The puck is made of a hard black rubber and is shaped like a disc. The game is started with a face-off at center ice and Jack fights for the puck. The game is a close one. After the second period the Sharks are tied with the Lions, 2 to 2. If the two teams are tied after the three periods then the teams play an overtime period. Jack skates very fast and gets the puck away from the other team. Before the third period ends - H E S C O R E S ! ! The Sharks won the game! Comprehension in Hyperlexia Appendix I Comprehension Questions for High Interest Passage Passage Independent Questions: 1. What is your favorite hockey team? 2. How many periods are in a game? 3. What do hockey players wear on their feet? ; Passage Dependent Questions: 1. What is the name of the player who has a special hockey stick? 2. What are the names of the two teams? 3. When were the two teams tied? 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items