in. 1, 3331;: ' .333 .‘3 E ‘ J 3‘ ' ‘E 3 , C" I ' 3 1‘11: ' ,4 E ,1“ . H93 1:, W'kf, 1&41. 1“ E 1 t EE'E" \S-v SJ; xh’ '3' 3 ' E 1 4 ‘33’ ’u . 33:3 3 ’3' SEE ‘ 3'313‘E1EEE'C‘E-33G3EEEEE3 E w—‘wéfi' 3E"? .1333 3 E 3 “"3“" ['3 ' 3‘ ‘ ‘ ."33"3'l"‘£-'|' ‘-3'\" '1‘. -3' ' Ui‘ta.:»3‘Er . 33 1m,”V%“ {WEWMM ‘ W dehnéwV%$&§. L 3 3333‘1'3'3 3E'3E'3E'3AEE‘3E 3 ' 3 ‘ «'xfl‘EEEE‘KbE 1E, \k'E'wM $23-$25??? ' 3 . « EEng .. , « 95E \3». I ‘3'E'1‘E'k‘ 3,1,“, A“ :13" 331E336??? EEEEEEEEEBH 1 \E3":‘S‘ ”hm!“ "S" ”SEL 3:“: “"33“ ”‘3, “; .E .1 “L1,, ‘ ,6,“ E-EhE ' NE” 'E“ \ f ,3" '16. 1,“ 32S: “SEES?" SUV? 33!: £537 1 LEE. .~ 'fim, “Nun. LN u’gSnE §3ESEE3ELQ 163:}, SE!“ NW “SSS, fl m W'QE 3 E111 [KW I U "‘1 .‘ tfl'qfié SSSSS, ‘1‘ “.4 'fihfifififl Emil .37.: [DE 1 a“ iS{3 . .A‘1 "3i: 1 7. . 43;: :&§;?: «3.. :9“ 2.54 #59" 1., =5: .F‘Je’; 5"“) N‘I'E‘au 3 . 3' VHS SELL?» imagaaw‘t: .4 .:.-‘ 3:2?” 3.: E: “g {Stu E‘fEfiéfiwE 4.3 [(4. 3&3; " ”f,” . £6815} lg;- , 4%??? 1 l ‘I Aflé“, ki '333333‘ ‘SEEISWE' \. . ”n" . a, ,. l , “E: EEEL ff‘l‘SlE ,- 1:252 135W? 3 " L ‘ 3 1: ~ -.3-. $1,,“ R§Mw ‘fifi hiwn €~:” 3. «WE E'Eé‘EESSS :23, HEEL SS .‘S-JNSE: "\E “ , 'E EELfELEt :"3I3‘; --,;i-t<“ ‘ RES!) 1? 3 9, my,“ ,, T.” V”. 1:1 -._... E, "'3“ z " 3345‘ 3' "NJ {, EvE'fV- , 1 (E 3EESE,S g? u...,., 2 ..u'. ‘M.<‘ :‘gas. A.‘ “'33" ‘ J- -, E233“; WE ”ENS { 3",‘EN'HEIHES ,J‘fi 53’- -'£ :- C 3" A, .r ‘ ' ' ~ '4'- i Kiri—L. & ,r. gt: ’ r‘gfi .IEEEEH .‘s! ‘lfi‘fi'lulfi 3:. ESEHL 9: S _ 3 :ENE‘ EEW%WEEmWWs Eumflfiw, ,t 'E‘ * x“ 5'ng , 6 “W WE ‘tw,~',.=sz:,‘:; . 317'33333 3E, E5, r3“: . ‘1'“, In .‘ 3‘ SEWSE‘E, {EEEEE'E _ ,, "SE“! “#93st 3‘3" ,"Sy,,a,v- LS?) (VS , 316‘ E" I,f,,,..LE3' ‘1 "E'EE'EEEE‘EEEE ‘ "3' 3 "l 3E E3'33 E9 Sfi'EMSNEN . '.‘." ES‘E#"R{§,3 "z‘E , .“3‘ 33'33HSE3E'3 ., SEE EEEEEUEES'EE‘EE'LE3‘E ‘3 E13SE333EE3EE33'E'3336L3EWEE'B‘E' S . ME 353313 '3 S33E33333' {SEVEI‘EEEEE‘ EEEESEE’SE'LEEJES EEE'EEEEJ‘ SE SEE" .‘EEE’E‘EEEESSEEEF‘ " “NW“: “ES, ,. ,- 3, EMILESEE‘. EE‘ESJEEEWEEE":“E“E'EEE'W“1SEEEE” ‘ SS 13103;“. WEE}; “ 3ESESESSEWEWWMWSES ,Wy” ~. ,,~4,SEASE' ‘ ' |,;f“ . ,-1, ’ ' I, '_,‘ . _ 3,*1.€,“I,i'!4‘ .f‘bE‘ SEES!“'EE‘EE'EE'E’JE'EE'n. ‘ E EEEEH‘EMME‘EEEEE’SEEE“ : E x, H . ”WEEVEMWEEWEE wfi'mfiMEv‘a” E“‘ ,, Ewl E A" SEEMS.” EEE3SSEE3SEEESS ,1. SEE'E'EEEEE" IIIEE3EE'3'EE3ES; 1 EEES ”vaqu 'EE‘EE3E3 ”(If S ,~'3."EE"3 i [I SE1, ’g‘fiEJ 333'. ”No.1“? ,.SSS ES «E {SE36} El" HS’3,]'3I‘ " EMS!“ ‘E‘E E'EE‘EE’EE' , , SSS“ E3'333EE3‘EE'EE33S3S'E. ‘SSJE EE' . a E 3, ""'t :EM“ v-,AWEEWEH EMKSE3E'EES ‘Emwwimfifihfime* ‘, 33 '3' EU" E3’3"I E" ' {EUL'E‘ ‘rE,.3 3 319”}: "‘13 amz‘E?’ H {.13 JESS» S ,, ES EAEE II: ' .r , ,IS'MS‘ .,, SS, My {E . uj'SS ‘Ew 3,3911 SESrE“ ‘ 3 SEE. EEWE ' 1A.} VIE-'5 EETE E E S ,SS SS ’M’W‘FJ ,0, SSS'SSS ,L‘,S,, , t E‘ . E'EEE, l/S‘SSS'E S, SEE‘E" SESSSJEES 'SEEESSS‘, “EEES'MWE EEE'E ESEES SEES EISE SSS}, ESSfiE'ESiEE3EE-E3S."331S3E33‘33'E3EEE'3EHSSE [,4va {"2 1‘3“, '\'S "ESSEE " H. ”'4“ EWSE S'HWSXE‘I'E (u' "S'ES‘ “3:33am ' ' EELEE'E 33E'E3 3 EE3‘EEE3'EE3EE3E1‘E3ErEEiflbL: EEK E’EEEEE ‘EEEEE EE'EEEE 33' $141 SE 33 W “WEE EEEEE33E333E1§ W a “ll \\\\\\\\\| 301 .r; This is to certify that the thesis entitled THE EFFECT OF READER-AUTHORED MATERIALS ON THE PERFORMANCE OF BEGINNING READERS presented by SHARON KAY THOMAS has been accepted towards fulfillment of the requirements for /Mfua4( Major professor DateM 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY WATER; Place in bow rec-m charge from 5'- s i i L -l \llllellll \lflle “3' \E Ell : “:7 4...“: 7‘ i - This is to certify that the thesis entitled THE EFFECT OF READER-AUTHORED MATERIALS ON THE PERFORMANCE OF BEGINNING READERS presented by SHARON KAY THOMAS has been accepted towards fulfillment of the requirements for Ph . D . degree in Education 2ia<4¢fi§7k¢4< Major professor DateM— 0-7639 OVERDUE FINES: 25¢ per day per_ ‘itm RETURNING LIBRARY MATE}? Place “in bod: return 3.:- charge from a" ‘ 1’27“ A _-_ gums 3. “mm. THE EFFECT OF READER—AUTHORED MATERIALS ON THE PERFORMANCE OF BEGINNING READERS By Sharon Kay Thomas A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education T980 ABSTRACT THE EFFECT OF READER-AUTHORED MATERIALS ON THE PERFORMANCE OF BEGINNING READERS by Sharon Kay Thomas The purpose of this study was to investigate the influence of reader-authored materials on reader performance. According to the lit- erature surveyed, measures of the effect of the language experience approach have generally been favorable, especially for beginning reading instruction. Past research in the field has usually been conducted using standardized tests as criteria. The effect of language experience mater- ials as measured by actual student performance had not been investigated. In this study, eight first grade students were asked to dictate a story to accompany a wordless picture book. They were also asked to read a previously selected story from their basal reader. In both cases, a- reading performance was measured through the use of the Evaluation Form of the Reading Miscue Inventory (Y. Goodman & Burke, 1976). In addition to reading performance, the reader's perception of reading and the struc- ture of the two texts was also investigated. Perception of reading was measured through prior instruction, an interview, and asides made during the reading process. The structure of the text was measured through word, syntactic, and semantic measures. On all measures of the Reading Miscue Inventory, the students, as a group, performed more proficiently on the dictated story than on the Sharon Kay Thomas text story. The measures of perception of reading indicated that the students in this study viewed reading as a decoding-to-sound process and their Strategy selection generally focused on letters and words. Word and syntactic level measures of the two texts indicated that the dictated stories were more difficult in both vocabulary and syntactic complexity. The semantic measure of the text showed that the dictated stories were more well-formed than the text story. Given the theoretical orientation of the readers and the increased difficulty of their dictated stories, performance on their own texts should have been less proficient. The fact that, as a group, the use of proficient strategies increased on the dictated stories was cited as evidence of the positive influence of reader-authored texts. Finally, the most well-formed stories were produced by the more proficient read- ers, and the least well-formed stories were produced by the less pro- ficient readers, indicating that a relationship exists between produc- tion and reception of written language. This dissertation is dedicated to my husband, F. Richard Thomas, whose understanding and generous support continue to make all things possible. ii ' ACKNOWLEDGMENTS I would like to express sincere appreciation to a number of people whose help made the completion of this work possible. Sheila Fitzgerald has been my advisor and friend from the first day of graduate school. She gave me the inspiration to begin and the support to complete what I had begun. James Stalker introduced me to the field of linguistics and psycholinguistics and expertly provided access to that world. Stanley Wronski opened my eyes to the world beyond my own culture and enabled me to assess my experiences there. James Snoddy provided a continual challenge to my beliefs and, in the process, allowed me to come to.know what I gjg_believe. To all these members of my committee, I wish to convey my gratitude for sharing their expertise and giving their advice throughout my graduate career. To the faculty and graduate students of the Reading Department at Indiana University, I owe a special debt of gratitude. Although I spent only one year there as a visiting student through the CIC Traveling Scholar Program, they made me most welcome and have continued to treat me as one of their own. Carolyn Burke saw the seed for this disserta- tion germinate in one of her courses and nurtured it through many stages to final completion. Without her inestimable knowledge and willingness to share that knowledge, this dissertation would never have reached fruition. Jerome Harste also provided major support for this project. He, too, spent innumerable hours helping sift through data and organize results. His willingness to share insights and his own research has been a constant source of encouragement. Lynn Rhodes' dissertation and her personal interest in mine have also added immeasurably to the success of this project. Several Indiana University graduate students contribu— ted their time and expertise: Karen Feathers evaluated all of my many attempts to analyze the two texts; Margaret Siegel did an RMI on one of the texts as a comparison measure, kept me abreast of new developments in the field, located references available only at Indiana University, and generally kept my spirits up; and Mary Hill accepted all my late night phone calls, shared invaluable insights, and forced me to take a needed break now and then. I would also like to offer my thanks to the many friends, colleagues, and neighbors who have offered moral support along the way, especially Etta Abrahams who dedicated much of her spare time to the last minute crunch. Last, but definitely not least, I would like to acknowledge the dedicated efforts of my family: To my husband, Dick, who did everything from the laundry to typing the bibliography; to my son, Severn, for cook- ing all those delicious meals and not being offended when I didn't apprec- iate them; to my daughter, Caeri, who was the first subject of the first pilot project and, in addition, brought home all her friends to be subjects as well; and to Troels Hansen, our exchange student from Den- mark, who probably never did understand the furor in our household, but gave his best anyway. iv TABLE OF CONTENTS Page LIST OF TABLES ........................ viii LIST OF FIGURES ........................ ix LIST OF APPENDICES ...................... x Chapter I. THE PROBLEM ...................... I Need for the Study ................. I Background of the Study ............... l A Socio-Psycholinguistic Model of Reading ...... 3 Assumptions ..................... 4 Purpose of the Study ................ 5 Expected Outcomes .................. 5 Measures of Strategy Selection .......... 6 Miscues of Reader's Perception of Reading ..... 6 Measures of Text Structure ............ 7 Limitations of the Study .............. 7 Overview ...................... 8 II. LITERATURE AND THEORY ................. 10 Language Experience ................. l0 Wrightstone Review ................ ll Hildreth Review . . , ............... 12 National First Grade Studies . . . . ....... l4 Hall Review .................... l5 Reading as a Socio-Psycholinguistic Process ..... 24 Assumption One: Reading is a Socio-Psycho— linguistic Process .............. 24- Assumption Two: Readers Make Use of Three Cue Systems in Language ............ 28 Assumption Three: Readers Make Use of Their Knowledge of the World ............ 35 Miscue Analysis ................... 40 Strategy Selection ................. 4l Young Readers ................... 4l Mature Readers .................. 4l Chapter Page Less Proficient Readers .............. 42 Proficient Readers ................ 42 Influence of Instruction ............. 43 Reader/Text Interaction .............. 46 Similar Studies .................. 49 Summary ....................... 52 III. METHODOLOGY ...................... 54 Procedures ..................... 54 Design ...................... 54 Sample ...................... 61 Data Collection .................. 62 Analysis . . . . ' ................. 65 Coding Procedures . . . .............. 66 Expected Outcomes .................. 81 Measures of Strategy Selection .......... 81 Measures of Reader's Perception of Reading . . . . 82 Measures of Text Structure ............ 83 Summary ....................... 84 IV. RESULTS ........................ 87 Measures of Strategy Selection ........... 89 Sentence Level Measures .............. 89 MPHN and Correction Measures ........... 91 Word Level Measures ................ 93 Omission and Substitution Measures ........ 96 Summary of Measures of Strategy Selection ..... 102 Measures of Reader's Perception of Reading ..... 103 Prior Instruction ................. 104 Theoretical Model ................. 106 Verbalizations .................. 109 Summary of Measures of Reader's Perception . . . . 110 Measures of Text Structure ............. 111 Word Level Measures ................ lll Syntactic Level Measure .............. 118 Semantic Level Measure .............. 120 Retellings .................... 125 Summary of Measures of Text Structure ....... 126 Summary of Results ................. 127 V. CONCLUSIONS ...................... 130 Results ....................... 130 Measures of Strategy Selection .......... l3O Measures of Reader's Perception of Reading . . . . 131 Measures of Text Structure ............ 131 vi Chapter Page Discussion ...................... 131 Implications ..................... 134 Teacher Preparation ................ 134 Reading and Composition .............. 135 Texts ....................... 136 Future Research .................. 137 BIBLIOGRAPHY .......................... 140 BIBLIOGRAPHY OF INSTRUCTIONAL MATERIALS ............. 150 APPENDICES ........................... 151 vii LIST OF TABLES Table Page 4.1 Reading Miscue Inventory Results, Sentence Level, Individual Scores ................... 90 4.2 Reading Miscue Inventory Results, Miscues per Hundred Words and Correction Score, Individual Scores . _ .................. 92 4.3 Reading Miscue Inventory Results, Word Level Substitutions in Context, Individual Scores ................... 94 4.4 Reading Miscue Inventory Results, Mean Percentage Scores ................. 97 4.5 Percentage of Omissions and Substitutions, Individual Scores ................... 99 4.6 Focus of Verbalizations ................. 110 4.7 Readability, Unfamiliar Words Correctly Identified, Individual Scores ................... . 116 4.8 Untaught Words Correctly Identified, Individual Scores ................... 117 4.9 Use of Setting Statements ................ 122 4.10 Use of Closure Statements ................ 122 4.11 Percentage of Complete Episodes ............. 123 4.12 Relationship Between Episodes .............. 125‘ 4.13 Retelling Scores .................... 126 viii LIST OF FIGURES Figure Page 2.1 A Model of Proficient Silent Reading ........... 30 2.2 Three Models of Reading .................. 32 2.3 A Reader Encounters a Text ................ 38 4.1 A Reader Encounters a Text ................ 88 4.2 A Comparison of Omissions and Substitutions Made While Reading the Text Story .............. 100 4.3 A Comparison of Omissions and Substitutions Made While Reading the Dictated Story ............. 101 4.4 A Comparison of Readability Levels of the Text Story and the Dictated Story .............. 112 4.5 A Comparison of the Number of Untaught Words in the Text Story and the Dictated Stories .......... 114 4.6 A Comparison of Words per T-unit in the Text Story and the Dictated Stories .......... 119 ix LIST OF APPENDICES Appendix Page A._ ELICITATION METHOD ................... 151 B. PARENT PERMISSION LETTER AND CONSENT FORM ........ 152 C. READING MISCUE INVENTORY CODING SHEET .......... 155 D. READING INTERVIEW. . x ................. 156 E. SPACHE READABILITY FORMULA WORKSHEET .......... 159 F. RETELLING SCORING SHEETS ................ 160 G. VERBALIZATIONS ..................... 162 H. STORY GRAMMAR RESULTS .................. 164 I. DICTATED STORIES .................... 177 a. TEXT STORY ....................... 194 K. THE BEAR AND THE FLY .................. 196 L. INTERVIEWS ....................... 213 M. RETELLINGS ....................... 242 CHAPTER ONE: THE PROBLEM The language experience approach for teaching reading is a method in which instruction is built upon the use of reading materials created by writing down children's spo- ken language. The student-created reading materials rep- resent both the experiences and the language patterns of the learner. . (Hall, 1977, p. 3) Need fer the Study Although the language experience approach is often recommended for beginning reading instruction (Lee and Allen, 1963; Ashton-Warner, 1963; Stauffer, 1970; C. Chomsky, 1971; Clay, 1975; and McKenzie, 1978), to date, research documentation has centered on general achieve- ment, readiness, oral language, vocabulary, word analysis, creative writing, Spelling, and comprehension measures. Little research in the language experience approach has investigated the reading process. In her comprehensive survey of the literature, Hall reported that no study was located that investigated reading performance in conjunction with the language experience approach and suggested: Future language experience research also could explore di- mensions of the reading process. Investigations of chil- dren's oral reading performance, using miscue analysis pro- cedures, could be conducted with language experience materials. (Hall, 1977, p. 40) Background of the Study Since the 1920's, scores of investigations into the various as- pects of the language experience approach have been conducted. It would seem, given the plentitude of studies, that no more research is needed. This conclusion, however, is a hasty one and does not take into account the increase in knowledge about language in the past ten to fifteen years and the resultant changes in reading theory. The discovery, in the 1960's, that children acquired language not through imitation, but through rule governed behavior that could be documented and was universa1,1ed educators to change their thinking about reading. The assumption that mastery of the lower levels of processing (letters and words) was a prerequisite to the understanding of sentences and whole texts was challenged. It was discovered that children both knew and were able to use their knowledge of syntax and meaning in their encounters with language-—oral and written. Thus, the psycholinguistic model of reading emerged. In this model, meaning is the pervasive factor in all language encounters and it is the reader's expectation that print is meaningful that allows the reader to make proficient use of all of the cue systems Of language (the graphopho— nemic, the syntactic, and the semantic). The psycholinguistic model of reading lends credence to the lan- guage experience approach--especially for beginning or troubled readers. One way to make print meaningful to readers is to give those readers stories that are based on experiences similar to their own. The lan- guage eXperience approach does precisely that. "The student-created reading materials represent both the experiences and the language pat- terns Of the learner" (Hall, 1977, p. 3). In the 1960's, proponents of Kenneth Goodman's psycholinguistic model of reading developed the Reading Miscue Inventory (Y. Goodman & Burke, 1971). This instrument allows both researchers and teachers to determine a given reader's ability to use the cue systems of language proficiently. Thus, the suggestion that an investigation "of chil- dren's oral reading performance, using miscue analysis procedures, could be conducted with language experience materials" (Hall, 1977, p. 40), is in reality, a suggestion that the language experience ap- proach needs to be investigated in light of the newer, psycholinguis- tic model of reading. A Socio-Psycholinguistic Model of Reading The model Of reading used in this study is taken from Rhodes U979a)as adapted from Harste (in press), and is composed of three parts: The Print Setting, which is the text structure based on the author's theoretical model of reading instruction; The Mental Setting, which is the reader's perception of reading based on his/her theoreti- cal orientation; and Strategy Selection. The main focus of this study is strategy selection, which is described in terms of the reader's more or less proficient use of the three cue systems of language as meas- ured by the Evaluation Form of the Reading Miscue Inventory (Y. Goodman & Burke, 1976); however, the structure of the texts (Print Setting) and the reader's perception of reading (Mental Setting) are also im- portant to this study as they represent a means of explaining strategy selection. The model of reading presented here and explained in more detail in chapter two is termed "socio" as well as psycholinguistic in order to include both situational and linguistic constraints found in the print setting. While the term "socio" could be construed to include a wide range of factors that may have some bearing on reading perform- ance (level of income, number of books in the home, educational level of parents, and so on) in this study, "socio" refers to the constraints of the particular context in which the reading is taking place. For example, the type of reading instruction Offered in a particular class- room results in both linguistic and situational constraints: During reading instruction emphasizing a phonetic approach to reading, both the material and the setting may Signal readers to access graphophonemic information. Students, in this setting, demonstrate the characteristic behaviors of decoders by producing many highly graphophonemically similar nonsense words (Barr, 1975; Harste & Burke, 1977). On a shopping trip these same children may process print more naturally because other schemata are signaled and accessed in this context. (Harste & Burke, 1978, p. 13) In this example, the material, a linguistic constraint, and the ‘ setting, a situational constraint, modify both the reader's perception Of reading and the way that reader samples print. Further, the process is a cyclical one in that each encounter with print serves to further modify the reader's perception of reading and the reader's sampling of print for the next text encountered (Harste & Carey, 1979, p. 4). The cyclic nature of this model is important to the present study be- cause, although the focus is on strategy selection across two differ- ent types Of texts, a description of the particular strategies utilized by the reader does not explain why_those strategies were utilized. Knowledge of the structure of the texts and the reader's perception of reading is essential to fully understand the differences in strategy selection across the two texts. Assumptions Reading is a socio-psycholinguistic process in which readers use their knowledge of language and Of the world to construct the meaning of a text through predicting, confirming, and comprehending strategies. Readers make use of three cue systems in language: the semantic, the syntactic, and the grapho-phonic. Ideally, the proficient reader uses all three systems in an integral way. Readers also make use of their knowledge of the world. This knowledge includes their view of the world in general as well as their perception of reading in particular which is, in turn, influenced by their past experiences with print, the instruction they have received, the theoretical models embodied in that instruction, and the structure of the text itself. Purpose of the Study; The purpose of this study is to investigate reading performance in one dimension of the language experience approach--the dictation of a story to accompany a wordless picture book. A miscue analysis pro- cedure (Y. Goodman & Burke, 1971), was used to describe the strategies and miscues evidenced by beginning readers while reading a basal text story and while reading a dictated story. The oral readings of both texts were taped and coded using the Evaluation Form of the Reading Miscue Inventory by Y. Goodman & Burke, (1976). The data collected was assessed to determine strategy utilization of the readers on the two texts. In addition, both the structure of the texts and the reader's perception of reading were investigated in order to better explain strategy utilization. Expected Outcomes As measured by a miscue analysis procedure, beginning readers will exhibit more proficient use of the cue systems of language while reading the dictated story than while reading the basal text story as determined by both sentence level measures and the word level measures. The Specific measures used are listed below and represent the major focus of this study. In addition, in an effort to explain the strate- gies selected by the readers during their encounters with the two dif- ferent types of texts, an attempt was also made to describe both the reader's perception of reading as well as the structure of the texts. Measures of Strategy Selection More sentences, as finally produced by the reader, will be judged both syntactically and semantically acceptable on the dictated story than on the text story. Sentences, as finally produced by the reader, will less often result in a meaning change for the dictated story than for the text story. Readers will produce fewer miscues per hundred words (MPHW) on the dictated story than on the text story. Of the miscues produced, readers will more often correct those miscues on the dictated story than on the text story. Miscues produced on the dictated story will be less similar on a phoneme—grapheme measure than those produced on the text story. Miscues produced on the dictated story will more often be of the same grammatical function than those produced on the text story. Miscues of Reader‘s Perception of.Reading,_ The readerfis perception of reading will also help to explain performance (strategy selection) on the two types of texts. Percep— tion of reading was measured by prior instruction, theoretical model (both Students and teacher) as elicited by The Reading Interview (Burke, 1978) and verbalizations made during the reading process. Measures of Text Structure The structure of the text will also help to explain the reader's performance (strategy selection) on the two types of texts. The struc- ture of the text was determined by a readability measure (Spache, 1953), a syntactic measure (Hunt, 1966), and a semantic measure (Stein & Glenn, 1977b). Further, the reader's comprehension of the text was measured using a retelling procedure (Y. Goodman & Burke, 1972). Limitations of the Study This study is based on the reading performance of eight children enrolled in the same first grade classroom, whose reading of the text story could be characterized as "beginning" reading. For the purposes of this study, "beginning" reading was defined as a minimum of 20 mis- cues on the basal story but not so many miscues as to make the reading unacceptable as language. Generalization to larger populations can only be made with caution. The readers were given only one opportunity to engage in a language experience activity, and no effort was made to give any other instruc- tion. As far as was possible, the usual classroom procedures as deter- mined by the teacher were followed. This study is not concerned with the influence of language instruction; rather, it is concerned with the influence of a language experience text on the performance of a selected group of beginning readers. The language experience activity gave the readers the opportunity to produce a story for one wordless picture book, The Bear and the Fly by Paula Winter, which was chosen for its brevity, action, and clear illustrations. Other wordless picture books, or objects, or pictures would have elicited different kinds of stories. The intent of this study was not to examine the effect of wordless picture books in gen- eral, but to focus on the comparison of strategies employed by begin- ning readers in one instance of a language experience activity as com- pared to their usual basal reader activity. Overview The review of the literature and theory in chapter two is divided into three sections: Language Experience; Reading as a Socio- Psycholinguistic Process; and Miscue Analysis. While the first and third sections focus on previous research as it relates to this study, the middle section explores the reading theory upon which this study is built. Chapter three presents the design of the study, including a de- scription of the evolution of this study, and a description of the sample and the analysis of the data collected. The analysis of the data is based on the socio-psycholinguistic model of reading presented in chapter two. In chapter four, the results of the study are presented. Again, the socio-psycholinguistic model described in chapter two is used as the framework for reporting the results. Chapter five reports the conclusions of this study as well as im- plications for further research. CHAPTER TWO: LITERATURE AND THEORY This chapter is divided into three parts. In the first part, four major reviews of language experience research are discussed. In the second section, a model of reading as a socio-psycholinguistic process is developed. Finally, the last section consists of a brief summary of trends in miscue analysis research followed by a thorough discussion Of those studies that are most pertinent to this research. Language Experience Although not always under the same label, the language experience approach to the teaching of beginning reading has been around in one guise or another since about 1900 (Hildreth, 1965). More recent pro- ponents of the method include Dorris Lee and Roach Van Allen, Sylvia Ashton-Warner (1963), Russell Stauffer (1970), Carol Chomsky (1971), Marie Clay (1976) and Moira McKenzie (1978). While these authors may not agree on the particular use to be made of the materials produced, all advocate reader-authored materials based on the Children's own language and experience as the most appropriate for beginning reading instruction. Traditionally, in a language experience approach (LEA), children initially dictate their ideas to another person who acts as a scribe. AS children become more adept at using pencil and paper, they produce their own compositions. Usually, the Children's writings are based on direct experiences: something that happened at home or on the play- ground; a field trip; a response to a classroom visitor; or directions 10 11 to their home in anticipation of a friend's visit. More recently, language experience activities have been expanded to include the use of books as the experience, and the term has been altered accordingly. Harste and Burke, for example, "have adopted the term 'Language Based Activities' to get around some of the direct experience features asso- ciated with language experience per se" (Harste, personal communica— tion, 9/21/79). Since the 1950'S,interest in the language experience approach has burgeoned as evidenced by increased research in that area (Hall, 1977). The major reviews of the literature are Wrightstone (1951), Hildreth (1965), the National First Grade Studies as reviewed by Vilscek (1968), and Hall (1977). Although Hall's review includes all of the earlier reviews, and is therefore the most complete, in order to obtain a clear historical perspective, it is important to look at all four reviews of the literature. Wrightstone Review Concern before the 1950's centered on whether or not a language experience approach to beginning reading could be considered adequate. In order to judge the effectiveness of language experience studies conducted in the 1930's and 40's, Wrightstone (1951) used studies by Betts (1943), Kopel (1942), and Peck and MCGlothlin (1940) to identify a variety of factors that influence initial reading performance. For example, it was suggested by Peck and MCGlothlin that "teachers should consider especially intelligence, informal reading performance, health and physical status, emotional and social growth, language usage and the general breadth of experience of the pupils as Showing the closest 12 relationship to reading achievement" (Wrightstone, 1951, p. 5). Using these factors as criteria, Wrightstone evaluated eleven studies that compared the success of students in experience curriculums as compared with students in basal reader curriculums. He concluded that, given the many problems associated with initial reading, no one approach could be considered a solution and recommended that although the eleven studies reviewed showed a favorable trend towards the experience pro- grams, a total reading program should take advantage of both approaches. Wrightstone's review of the research is interesting in that it suggests an early awareness Of sociological concerns. Thus,the rele- Vance of this review of early studies is not so much in the studies themselves as in the focus on factors influencing early reading instruc- tion. Wrightstone's concern centered on the ability of language ex- perience programs to address those factors. A concern with language and general breadth of experience, for example, highlights two of the most significant aspects of language experience programs. Given Wrightstone's focus, it is understandable that the studies reviewed showed a favorable trend towards the experience programs. Unfortu- nately, the prevailing models of reading, at the time, dictated a more controlled approach. Hildreth Review I By the mid-sixties, the focus had shifted to contrastive studies. Hildreth's (1965) review of the literature includes both an historical overview as well as the new focus on comparison. Hildreth begins with the origin of the method by Flora J. Cooke (c. 1900) and traces its development through a variety of experimental schools until the 1920's, 13 when it blossomed with modifications in the public schools. Use of the method in the public sector brought not only modifications, but objec- tions as well, and by the 1960's numerous studies had been conducted to determine the value of the contrasting methods. 0f the twelve studies cited that compared the teaching of initial reading through experience-related material in contrast to the use of standard tradi- tional textbook methods, only two suggested negative results for the experience method group. In both cases, the experimental groups re- ceived instruction that was largely unplanned and undirected in con- trast to the other studies in which the experimental groups were taught by "activity related methods," that is, methods that based initial reading experiences on child life and interests, integrated reading with other language arts, delayed use of basal texts, did not control vocabulary, used no artificial reading readiness materials, used a variety of experiences with print, made extensive use of children's literature, and recognized the link between reading and writing. By the mid-60's advantages of the language experience approach had crystallized into a well defined method and those studies in which the method was adhered to favored the approach. Hildreth's review points to the necessity of a well defined method, especially in con- trastive studies like the present study. It is clear in the method described by Hildreth that advocates of the language experience ap- proach had begun to pinpoint precisely those ways in which their ap— proach differed from the usual basal approach. Again, in contrastive studies like the present one, it is vitally important to not only keep the two approaches methodologically pure, but to understand the 14 underlying premises as well. National First Grade Studies In the early 1960's, the U. S. Office of Education sponsored a research program designed to investigate first grade reading programs. A number of language experience programs were included in the project giving the first indication that language experience programs had be- come worthy of investigation on a national level by a federally funded project. In 1968, Vilscek reported on the language experience prOjects included in that endeavor. Although the seven studies reviewed were selected because all employed the Stanford Achievement Test and the San Diego Pupil Attitude Test for evaluation, Vilscek reported that comparison of the results was still difficult. One major problem concerned the use of differing populations. 0f the two studies that Showed the greatest achievement for the control groups (basal reader approach) one group came from a Spanish speaking population and the other from a Black dialect popula- tion. Of the five remaining studies, Bond and Dykstra (1967) chose four (Hahn, 1965; Kendrick, 1966; Stauffer & Hammond, 1965; and Vilscek & Cleland, 1966) to analyze further. When controlled for population differences, the results were as follows: significant differences favoring the language experience approach were found for Word Meaning (three of the four studies), Paragraph Meaning (two Of the four studies), and Vocabulary and Word Study (one of the four studies). The only significant difference found for the basal approach was Para- graph Meaning (one study). Vilscek concluded that in those project 15 comparisons in which a difference was apparent, the results showed “a predominant distribution of Significant effects favoring language ex- perience instruction" (Vilscek, 1968, p. 14). Because the National First Grade Studies were the largest and most well controlled investigations ever made into the language experience approach, the favorable findings from these studies are important. For the first time, the approach was given not only national recognition, but credence as well. Hall Review In a recent International Reading Association monograph, Hall (1977) summarized over sixty language experience research projects. Her review includes the Wrightstone, Hildreth, and National First Grade Studies in addition to most Of the research conducted in the 1960's and 70's. In her discussion of the studies, Hall grouped the investi- gations into the following categories: Achievement, Readiness, Oral Language, Vocabulary, Word Analysis, Creative Writing, Spelling, Com- prehension, Special Populations, Affective Factors, Teacher Education, and Language Differences. As not all categories were deemed relevant to the present study, some are not reviewed here. A summary of her findings for Achievement, Readiness, Oral Language and Language Differ- ences, Vocabulary, Comprehension, and Teacher Education, follows.h Achievement. Research emphasizing achievement as the result of comparing one method with another is not as prevalent now as it was from 1920 to the mid-1960's. The early studies cited by Hildreth (1965), the three-year study in San Diego (Allen, 1961; San Diego Department of 16 Education, 1961), two of the Communication Skills through Authorship project studies (Schomer, 1972; G. Harris, 1972), and the language experience projects in the National First Grade Studies (Hahn, A. Harris, Kendrick, McCanne, R. Stauffer, Vilscek, and others) along with the M. Stauffer replication, were the major studies to emphasize reading achievement. (Hall, 1977, p. 24) The early studies cited by Hildreth have already been reviewed here as have the National First Grade Studies. In both cases, the majority of the results favored the language experience approach. The first large scale project, the three-year study in San Diego, included twelve elementary school districts and Sixty-seven teachers. Using standardized reading tests to measure achievement, the results showed that children in the experimental groups (language experience) made as much, or more, progress in reading skills as the control groups. The Communication Skills Through Authorship Program (CSTA) em- ployed the use of a tape recorder in a booth. Children talked to "Homer" and then mailed the tapes to a scribe who transcribed the tapes and returned the text to the child. Schomer (1972) found this approach "effective" and G. Harris (1972) reported significantly higher achieve- ment scores on the Stanford Reading Test for students in the CSTA pro- gram than those in the control group. M. Stauffer (1973) used the same schools and the same basic re- search design to follow the original Delaware project (Stauffer & Hammond, 1969). A major change in this replication was the attempt to determine the teacher's knowledge of language experience procedures. M. Stauffer concluded that first grade children taught by teachers who are knowledgeable about language experience teaching procedures exhibit better achievement scores on standardized tests than those taught by 17 the basal reader method even if those teachers have years of experience with the method. With the exception of M. Stauffer, a major problem in making com- parisons among the studies Cited resulted from the fact that more dif- ferences Often existed within methods than between methods (Bond & Dykstra, 1967). Even though an approach is labelled language experi- ence and children are given opportunities to produce their own reading materials based on their own eXperiences and language, radically dif- ferent uses can be made of these materials. Some proponents of lan- guage experience tightly control the vocabulary and sentence structure the children are allowed to use,or advocate using the child-produced materials as Skills lessons in which students are asked to find all the words with the same initial consonant or underline all the color words and so on. As a result, such programs come to closely resemble a basal program. As was stated in the summary of the Hildreth (1965) review, the necessity of a closely adhered to, well defined method is always important in contrastive studies. Readiness. Despite the frequent use of language experience at beginning levels of reading instruction, little research is available in this area (Hall, 1977). Of the studies cited, types of programs varied and included the use of experience charts (Brazziel & Terrell, 1962), a conceptual-language program that did not stress child pro- duced materials but emphasized basic concepts of language (O'Donnell & Raymond, 1972), an informal language experience program (Reichback, 1973), and a parent administered program (Weber, 1975), as well as a year long researcher-developed language experience program (Hall, 1965). Results of all studies cited were analyzed using either l8 standardized readiness tests or achievement tests. With the exception of Reichback (1973) who found that achievement test scores favored the DISTAR group, the other studies favored language experience instruction. Because the language experience approach is so Often recommended for beginning reading instruction especially by the proponents of the newer psycholinguistic models of reading, more research is clearly needed in this area. Oral language and lapgugge differences. A few studies focused on development of oral language as a result of language experience pro- grams although, again, the number of studies is limited and the dis- crepancies between them large. Giles (1966), Wells (1975) and Christensen (1972) studied oral language growth as measured by gains in such measures as total number of words, number of T-units, number Of words per T-unit, or mean T-unit length. Only Christensen concluded that the language experience children did not exhibit greater growth. Cox (1971) developed her own language analysis scale and found that language skills used in spontaneous expression, dictation, and personal authorship were interrelated. The only two studies with dialect Speakers reported by Hall were concerned with instruction in the learners' dialect and, as such, are not relevant to this study. Two of the populations in the National First Grade Studies, were all dialect Speakers (Harris & Serwer, 1966; McCanne, 1966). In both cases, the control groups obtained higher achievement test scores than the language experience groups. The present study is not concerned with either dialect Speakers or the development of oral language as a result of language experience l9 instruction, but the research is included here to document the impor- tance of oral language to a language experience approach. Unfortu- nately, the over-riding tone of these studies is the intent to measure the oral language of students against some kind of standard which is not the purpose of the language experience approach. Instruction, in this approach, is meant to be built on the oral language the child brings. Fortunately, the vast majority Of children enter school with a vocabulary of around 7,500 words and the knowledge of practically all the essential grammatical structures of their language (Stauffer, 1970, p. 5). Undoubtedly, most Children have a level of linguistic expertise that is sufficient for a language experience approach. Further, oral language continues to develop because it is constantly used as a means of functional communication. A more fruitful approach to the relationship between oral language and the language experience approach might be to investigate h9w_children make use of oral language in their endeavors to produce written language. Vocabulary. Because a measure of vocabulary used by children in their dictated stories is a part of the present study, the research cited by Hall in this area is of particular importance. Language experience instruction is Often criticized for lack of systematic introduction and repetition Of vocabulary. As a result, critics claim, children do not develop the vocabulary necessary for successful reading of basal and other materials. The research cited by Hall does not sub— stantiate this claim. Three studies focused on vocabulary learning of children in lan- guage experience programs compared to children in basal programs: 20 Henderson, Estes, and Stonecash (1972) found that the vocabulary of first graders in a language experience program compared favorably with the vocabulary of those in the basal program; Kelly (1975) found that the vocabulary of the language experience group exceeded that of the basal group; and Hall (1965), using the word recognition portion of a standardized test, reported a Significant positive difference for the language experience group. Vocabulary retention was the concern of Bennett's (1971) study and Shears' (1970) study. They both found that children retained words they had requested or actually used, better than words encoun- tered in basals. It was suggested that the meaningfulness of one's own words accounted for the superior recall. McC. Gallager (1975) did not find a difference in vocabulary retention between the two groups. However, meaningfulness may not have been a factor in this study, since the iNVestigator chose the tOpics used for the language experience lessons. Three researchers investigated the meaningfulness of vocabulary found in basals. Packer (1970) found that the vocabulary in basals is Often not meaningful to culturally disadvantaged children while Cohen and Kornfield (1970) found little divergence between the vocabu- lary of basals and that of Black, urban kindergarten children. In conclusion, the research cited by Hall (1977) suggests that vocabulary found in basals may not be meaningful to some culturally diverse groups. However, vocabulary generated by children in a lan- guage experience program is almost always meaningful and, therefore, better retained. Further, since the vocabulary of the language 21 experience groups was found to compare favorably or to exceed that of the basal groups, the assertion that the vocabulary developed by a lan- guage experience approach is not sufficient for successful reading of basals is refuted by these studies. Comprehension. Although comprehension measures are sometimes in- cluded by researchers investigating reading achievement, the results are usually based on test scores from achievement tests. "Specific questions about comprehension in language experience programs seem not yet to have been asked for reSearch purposes" (Hall, 1977, p. 30). The major reason for the lack of this type of research stems from dis- agreement as to the nature of reading comprehension. As Farr (1969) has suggested, “The most pressing need in measuring comprehension is for a clear understanding of the nature of reading comprehension" (Farr, 1969, p. 64). Currently, a debate exists over not only the exact sub-skills of comprehension, but even the notion that compre- hension is a separate measurable skill. Certainly, as Farr points out, comprehension must, at least, be composed of a variety of skills and, in all probability, these Skills are also dependent on a particular set of conditions (Farr, 1969, p. 64). In the present study, compre- hension was not a major focus although an attempt was made to measure comprehension through the use of retellings and a measure of how Closely the sentences produced by the readers resembled the author's intent. Teacher education. Since 1960, language experience has received increased attention in teacher education (Hoover, 1971). Teachers with knowledge of language experience instructional procedures and 22 some expertise in using the approach rate it positively (Lane, 1963; Hall, 1965; and R. Stauffer, 1966), but, in at least one state surveyed, the method is not widely adopted by school districts (Carstensen, 1973). In the sociO-psycholinguistic model of reading used in this study, the instruction a student receives is believed to have a Significant effect on that student's perception of reading and resultant ability to interact proficiently with texts. In the present study, an attempt was made to measure both the teacher's model of reading and the stu- dent's model of reading as a means of explaining text interaction of the students. If, however, a teacher's education has not included in- struction in the language experience approach, the approach will not likely be used. The teacher's methods, therefore, should affect both the reader's perception of reading and his/her performance when en- countering texts. In conclusion, Hall suggested that while much of the research cited concentrated on comparison of methods and was, therefore, subject to flaws common to any methods research, "the existing body of research does substantiate that language experience is an effective way to teach reading" (Hall, 1977, p. 38). Specifically, in many of the studies, researchers failed to describe adequately programs or procedures so that great differences existed in programs given the same label. Fur- ther, much of the research was reported in terms of statistical treat- ment of achievement scores. The use of achievement scores to make comparisons Of methods is, perhaps, the greatest problem in all of the research surveyed. To begin with, the use of any kind of standardized test is antithetical to the rationale of the language experience 23 approach. Proponents of the language experience approach View their method as a holistic approach in which all the modes of language-- listening, speaking, reading, and writing--form an integrated whole. Standardized reading tests, for the most part, are based on the theory that reading is composed of a variety of sub-Skills that such tests attempt to isolate and measure. Further, the use of standardized tests for any kind of measurement of reading Often results in information that can only be used with extreme caution (Farr, 1969). According to Farr, the elements that psycholinguistic theorists are finding central to the reading process offer some hope of better testing procedures in the future: It may well be that research will find, as the proponents of psycholinguistic theory have suggested, that attempts to define reading sub-skills on a group basis are fruitless. In that case, measurement in reading would have to be based on whether a reader has a strategy for decoding written mes- sages and whether he understands reading as a communication process rather than whether he can simply decode written symbols, supply the meanings of words in isolation, or answer multiple-choice questions based on a literal understanding of a selection. (Farr, 1969, p. 9) Hall also recommends that research needs to go further than gen- eral statistical investigations. There is a need for studies with Specific research questions grounded in a strong theoretical rationale that examine the features of learning, teaching, and the learner. AS an example, Hall suggests using miscue analysis procedures as one method of investigating children's reading performance in conjunc- tion with language experience. Hall reported that no study of this type was located in her survey of the literature and posed the ques- tion, "What strategies are actually employed by children using the 24 language experience approach“ (Hall, 1977, p. 40). In order to begin to answer Hall's question, a description of reading as a sociO-psycholinguistic process is given in the next sec- tion. Reading as a Socio-Psycholinguistic Process In chapter one, three assumptions were stated. Because these assumptions form the theoretical base of this study, they will be dis- cussed in detail in this section. The first assumption requires a general definition of a sociO-psycholinguistic reading model with an emphasis on the role that both knowledge Of the language and of the world play in reading. The second assumption requires a description of the specific ways readers use their knowledge of language when they interact with a text. Finally, the third assumption focuses on Spe- cific kinds of world knowledge that contribute to the complex event called reading. Because this section is necessarily theoretical in nature, few studies will be cited. Instead, the constructs of theorists in reading will be described. In the final section Of this chapter, a general survey of miscue analysis research will be given as well as an in depth review of several studies that are especially pertinent to this study. Assumption One: Reading is a Socio- Esycholinguistic Process Reading is a socio-psycholinguistic process in which readers use their knowledge of language and of the world to construct the meaning of a text through predicting, confirming, and comprehending strategies. 25 With the publication of Kenneth Goodman's classic article, "Read- ing: A Psycholinguistic Guessing Game" (1967), the psycholinguistic or whole language model of reading made its debut. Goodman suggested that reading was not an exact process, but, rather, a selection process in which readers made use of minimal language cues selected on the basis of their expectations and processed through tentative decisions that were confirmed, rejected, or refined as reading progressed (Goodman, 1967). Because learning to read is a highly complex task, emergent readers must learn to use their knowledge about language and about the world to develop strategies to predict, confirm, and comprehend as they interact with a text. Learning to read is not a Special kind of learn- ing. People learn how to read in the same way that they learn anything else. According to Frank Smith, "all human beings endeavor to make sense of the world, to comprehend and to learn in the same fundamental man- ner, from birth through adulthood" (Smith, 1975, p. 4). Learning to read is part of the "making sense of the world" process. Smith, a cognitive psychologist, bases his beliefs on both information proces- sing theory and psycholinguistics. Information processing research rests on the theory that human beings are not passive but are active learners. Their learning is not directed by habits but by intention and is always directed toward increasing understanding. Thus, the learner can be Characterized as a scientist who "constructs theories" and "conducts experiments“ in order to test those theories (Smith, 1975). Psycholinguists are concerned with how people learn language and strive to understand language learning as an example of learning in general. 26 Viewing learners as active participants in their own learning is a radical departure from an earlier, behvaioristic view of learning in which learners are characterized as learning by habit formation. A common view still held by some behaviorists is that learning to read is the result of habit formation and meaning is the linear sum of the words in a sentence (Cooper & Petrosky, l976). In other words, readers identify individual words (possibly even individual letters as a pre- cursor to identifying the words) before they can obtain meaning of the whole sentence. In contrast, reading as viewed by psycholinguists is always an active, robust activity in which "readers employ meaning to assist in the identification of individual words rather than laboring to identify words in order to obtain meaning" (Smith, l978, p. l53). Related to the active role of the learner in acquiring knowledge is the question of how that knowledge is understood. With the excep- tion of Piaget, the concept of comprehension as a constructive process is a relatively new idea. "The notion that everyone's brain contains a structure of knowledge concerning the world, into which all incoming information is assimilated, is a central aspect of Piaget's theorizing” Smith, l978, p. 2l0). Smith calls such a structure of knowledge "the theory of the world in our head" and explains its function as follows: What we have in our heads is a theory of what the world is like, a theory that is the basis of all our perceptions and understanding of the world, the root of all learning, the source of all hopes and fears, motives and expect- ancies, reasoning and creativity. And this theory is all we have. If we can make sense of the world at all, it is by interpreting our interactions with the world in the light of our theory. The theory is our shield against be- wilderment. (Smith, l978, p. 57) 27 In other words, as learners (readers) actively acquire knowledge they come to understand such knowledge only as it relates to what they al- ready know. What they can come to know is limited by the knowledge they already possess. The nature of the structure of knowledge in our heads or our theory of the world has been explored recently by researchers interested in memory as well as those interested in the structure of human knowledge. This knowledge can also be called "behind the eyeball" information (Smith, l97l) or "story grammar" (Rumelhart, l975; Stein & Glenn, l977b; Thorndyke, l977) or "scripts" (Schank, l973) or ”frames" (Minsky, l975) or "schemata" (Anderson, Reynolds, Schallert, and Goetz, l977; Rumelhart, Ortony, and Montague, l977). All of these terms are roughly synonymous. "All of them are presumably cognitive structures that can be used to account for how we organize information in long term memory" (Pearson, l978, p. 5). Further, these cognitive structures control our ability to comprehend. In terms of visual perception, they even control what we can see ”because we can see only what we know how to look for” (Neisser, l976, p. 20). For Smith, then, and for many people working in the field of education, "the basis of comprehension, whether of language or of the world in general, must be some internal organiza- tion of knowledge (or beliefs) about the world" (Smith, l978, p. 2ll). The proficient reader, at any level including beginning reading, is the reader who comprehends (constructs meaning). The fact that this particular type of comprehension comes about through an interac- tion with print (language) does not change the basic process of com- prehension. Learners use their knowledge about the world to com- prehend the world. Readers use their knowledge aboUt language and 28 the world to comprehend language. Assumption Two: Readers Make Use of Three Cue Systems in Language Readers make use of three cue systems in language: the semantic, the syntactic, and the graphophonic. Ideally, proficient readers use all three systems in an integral way. On a general level, the knowledge structures in their heads allow readers to predict and, therefore, eliminate many possibilities before they even come to the text.‘ For example, the fact that they are sit- ting in a reading group in a first grade classroom allows most students to eliminate the possibility that they will be handed a comic book to read. The ability to make use of the three cue systems constitutes one language specific example of the way in which knowledge is organized internally. Even though this knowledge is implicit, readers are able to use it to make predictions and to comprehend based on their pre- dictions. On a semantic level, if the title of the story indicates that the text is about farms, the readers will expect further informa- tion about farms as they proceed through the text. On a syntactic level, if the first word in the first sentence is “the,” readers will expect the next word to be a noun or an adjective. If they are using the cue systems proficiently, they will not, for example, predict that the next word is "of.“ On a graphic level, if the first letter of a word is "b” proficient readers will not expect the next letter to be ”w" or "c" or any other letter that represents an unallowable combina- tion in English. “This ability to predict is both pervasive and 29 profound, because it is the basis of our comprehension of the world” (Smith, l978, p. 64). People predict because they must eliminate unlikely alternatives. If they could not eliminate unlikely alternatives, they would be over- whelmed with possibilities and become confused. Young children, for example, generally assume that they will see their teacher at school, but not in the grocery store. When they encounter their teacher in the local supermarket they are often genuinely confused because they do not expect (predict) such an unlikely alternative. When encountering a text, readers use their knowledge of the three cue systems to eliminate unlikely alternatives and to predict likely alternatives. If the prediction is confirmed (see figure 2.l) the readers comprehend. If not, they must rethink or reread in order to make another prediction. Or, the readers may choose to go on read- ing on the assumption that further reading will give needed informa- tion for comprehending. In the act of comprehending, readers inte- grate the meaning from the text with their stored world knowledge. This integration process enables them to remodel their existing knowl- edge, and, thus, to anticipate the subsequent text with greater certi- tude. Readers make use of the three cue systems in language in more or less proficient ways. Ideally, the proficient reader uses all three systems in an integral way. That is, the most proficient strategy a reader can adopt is to use all three cue systems as they are important. However, some readers, those who are less proficient, often rely too heavily on one of the systems, as the following example reveals: 30 Reader selects cues and predicts ( material The author's ' thought and PREDICTING l language AG\(( 6 . .Q¥ 67 5K Reader checks ~\ semantic and . s ntactic acce ta ii ------- rereading -------------------- y ability p CONFIRMING The reader's thought and language Reader integrates meaning gained with his or her world view INTEGRATING Figure 2.l. A Model of Proficient Silent Reading. (From Y. Goodman and Burke, l980) 31 Text: The boys ran through the dark forest. Reader One: "The boys ran through the dark woods." Reader Two: "The boys ran though the dark frest." (from Y. Goodman & Burke, l97l, p. 4) Apparently, the first reader's strategy is to rely heavily on the syntactic and semantic systems and not attend to the knowledge or sound/ symbol relationships available as the substitution of ”went" for "ran," and ”wood" for “forest" illustrates. The reader did produce a verb for a verb and a noun for a noun (syntactic) and, further, the reader's choices were similar in meaning to the words in the text (semantic), but the reader's choices in no way either look or sound like the words on the page. The second reader chooses a strategy that relys too much on graphophonemic information and doesn't attend to syntactic ("though" and "through" do not have the same grammatical function) or semantic (”frest" is a nonsense word) information. 0f the two readers, the first one would appear to be more profi- cient even though his/her miscues have no sound or graphic similarity to the words in the text. Because this reader's sentence, as finally produced, is very close in meaning to the author's intended meaning, one can assume that the reader probably did comprehend what was read. In contrast, the second reader produced a sentence that has little meaning. His/Her miscues show an ability to "sound out” but, by rely- ing too heavily on that single strategy, the reader probably failed to comprehend. When readers use their knowledge about language and the world to predict, they produce acceptable language as the first reader did. 32 When readers are not competent predictors, they fail to produce accept- able language because they are not using all the available information to predict. In other words, the second reader primarily used grapho- phonic information and he/she made reasonable predictions based on that information. However, the reader's knowledge of grammar and meaning should have told him/her to reject those choices and try again. The psycholinguistic reading model described above is sometimes called a whole language model and is often compared with two other types of models, the phonics and skills models. In l977, Harste and Burke explored the three models of reading (phonics, skills, and whole language) and the instructional implications of each. (See figure 2.2 for a graphic illustration of the models.) graphophonics meaning letter/sound relationships Phonics Skills Whole Language Figure 2.2. (From Harste & Burke, l977) 33 Advocates of the phonics model view reading as a transition from print to sound, and beginning reading instruction focuses on the manipulation of the relationships between sounds and symbols. Those who advocate the skills model believe the word is the start- ing point for reading proficiency, and their instruction focuses on the learning of vocabulary. Proponents of the whole language model view reading as one of four ways in which the abstract concept of language is realized. This orientation as- sumes not only that the systems of language are shared, but that they are interdependent and interactive aspects of a process. (Harste & Burke, l977, p. 37) Instruction, in the whole language model, emphasizes meaning, and beginning reading materials are often built on the oral language ability of the students. It is important to note that all three models include a comprehen- sion/meaning component. The phonics and skills models, however, begin at the lower levels of the language processing (letters and words) and assume that proficiency or automaticity on these levels is a prerequisite for higher levels (syntax and semantics). These models are sometimes called "bottom up" models (Rumelhart, l976). By contrast, the wholelanguage/psycholinguistic model assumes that meaning is central to reading and that any level of processing is neces- sarily influenced simultaneously by both lower and higher levels. This model is often called an "interactive“ model and suggests that bottom-up and top-down processing occur at the same time. In addition to Goodman, other advocates of the interactive model are Rumelhart (l976), Adams and Collins (l977), Smith (l97l, l977), and De Beaugrambe (l978, in press). 34 Another way of looking at the difference in theoretical positions of the various views of reading is to examine the contribution of the reader and the text. For those who operate on a skills or phonics orientation, the meaning to be gained is inherent in the text. The reader strings sounds together to produce words, or words together to produce sentences, in order to gain meaning. Meaning is the linear sum of the words in the sentence. Those who hold a whole language perspective view comprehension as a process in which readers relate the information from the text to the information about the world that they already have stored in their heads in order to construct meaning. What readers gain froma a text, then, is only part of comprehension. Readers also integrate this information with the existing knowledge in their heads. Further, because readers are able to use their knowledge of the world in general and language in particular and are constantly remodeling their existing knowledge, they are able to predict. During the l960's, when Kenneth Goodman and his associates were doing research at Wayne State University, they discovered that the ability to predict is greatly enhanced when "the author's and the reader‘s experience are parallel" (Y. Goodman & Burke, l972, p. l2). Readers, especially beginning readers, must expect print to be mean- ingful. If they do not have this expectation, they cannot make use of all the cue systems (especially the semantic) to make predictions. One way to make print meaningful is to give children stories that are based on experiences similar to their own. This phenomenom is true of adults as well. Teachers of reading generally expect texts 35 on reading to make sense, primarily because they have the background experience to understand them. Unless they are specially trained, they do not expect a text on computor technology to be nearly as mean- ingful because they lack the relevant and necessary experiences to en- able them to predict. Readers, especially beginning readers or those with problems, can be helped to learn to use the cue systems and their knowledge of the world to predict if they are given texts that build on their own ex- periences and are, therefore, predictable. Such texts, however, are difficult to secure because children come to school with differing backgrounds and a wide variety of experiences. A reasonable solution, then, is to allow children to author their own books. In other words, the whole language model of reading not only lends credence to the language experience approach but gives that approach a theoretical basis as well. Assumption Three: Readers Make Use of Their Knowledge of the World Readers make use of their knowledge of the world as well as their perception of reading in particular, which is, in turn, influenced by their past eXperiences with print, the instruction they have received, the theoretical models embodied in that instruction, and the structure of the text itself. Just as knowledge of the world and knowledge of language influence the reader's strategies, so does the reader's perception of reading. What the reader believes about reading is part of that reader's lan- guage specific world knowledge. Some of the ways in which the reader 36 acquires this knowledge is through previous encounters with print and through prior instruction. In the latter case, the teacher's view of reading significantly influences the way that teacher instructs. Fur- ther, the characteristics of the text itself can limit or enhance the reader's utilization of strategies. In l977, Harste and Burke explored the notion that "both the teach- ing and learning of reading are theoretically based" (Harste & Burke, l977). Using instruments such as an interview with open ended ques— tions about reading (Burke, l978) and examples of readers from the three theoretical models (see figure 2.2), Harste and Burke attempted to explore the theoretical orientation of both teacher and student. Readers were asked such questions as: l. When you are reading and you come to something you don't know, what do you do? 2. Who's the best reader you know? What does he (or she) do that makes him (or her) such a good reader? 3. If you were going to help someone with his reading, how would you help that person? (Burke, l978) Based on their responses, readers were classified as having a phonics, skills, or whole language orientation toward reading. In the same manner, the teachers' orientations were also classified. If the teacher chose as the best reader a reader whose miscues were closely related to the graphic display and not corrected even if they were nonsense words, that teacher was classified as having a phonics model of reading. If the teacher chose a student who used phoneme/grapheme correspondences to predict a possible choice and the choice was a whole word (not a nonsense word) but it didn't make sense 37 in the sentence as a whole, that teacher was thought to hold a skills model of reading. Finally, those teachers who chose the student whose miscues were whole words and made sense in the sentence even though they were not very graphically similar to the text, were selected as teachers with a whole language model of reading. After the teachers and students were classified, they were ob- served in the classroom and "subsequent reading performance and class- room behavior was found consistent with the model from which the per— son was operating" (Harste & Burke, l977, p. 32). People's theories of reading, whether of teacher or student, are part of their knowledge of the world and, as such, determine how they attempt to make sense of the world. What they can come to know is directed by what they already know. In terms of reading, this idea is graphically illustrated by the following model (figure 2.3). In this model, the reader's perception of reading-~based on that reader's theoretical orientation—~directs the strategies that reader will use. For example, a reader who believes that reading is decoding to sound will adapt decoding to sound strategies while reading. The type of strategy will, in turn, determine how the reader samples the text. In this case, the reader will attend closely to the letters and attempt to produce sounds that closely resemble the graphic display. The reader's theoretical orientation is largely a result of the instruction that reader has received. Another aspect of the process, however, also has an influence. Just as readers and teachers have theoretical orientations, so do authors. Therefore, the text itself can influence the reader's utilization of strategies and, in turn, 38 Print Setting Text Structure-- Based on author's theoretical model 'of reading instruction Reader's Perception of Reading-- . Based on reader's Directs > Strategy theoretical orientation Selection Figure 2.3. A reader encounters a text. (From Rhodes, l979a as adapted from Harste, in press) 39 modify the reader's perception of reading. In other words, our knowl- edge of the world determines what we are able to know and, at the same time, what we come to know modifies the theory of the world in our heads. The same is true of reading. "What the reader learns from en- countering a given print setting becomes the raw data from which to reformulate anticipatory schemata" (Harste, l978, p. l3). In terms of texts, the editors of books designed for beginning reading instruction have different theoretical orientations and their beliefs are apparent in the texts themselves. A reading program based on a phonics model will include texts, especially at the beginning levels, that focus on sound/symbol relationships and words that are related by the same pattern such as cat, fat, bat, mat. In a skills model, the focus will be on words and sentence patterns. Texts will be composed of stories that utilize only those words previously taught and syntax will be controlled. A whole language program will include a variety of texts produced by professional authors as well as by the readers themselves. The choice of appropriate materials will not be made in terms of controlled vocabulary, syntax, or letter/sound cor— respondences. Rather, materials will be chosen on the basis of whether or not they are predictable. Predictable stories are those stories on familiar topics that are written in natural and familiar language with difficult words and complex structures included, if they are a natural part of a good story and enough content is supplied to support the efforts of the readers to use effective strategies (Rhodes, l979b, p. 22). Because the influence of the text on the reader is of major 4o importance to this study, several relevant studies will be explored in greater detail in the next section. The section will begin, however, with a current survey of the general trends in miscue analysis re- search. Miscue Analysis In the early l960's, Kenneth Goodman began to investigate the com- plexity of the reading process. Working from the simple task of ob- serving children's oral miscues as they read a text they hadn't en- countered before, he began to see that children's miscues were not the result of mere confusion. Rather, the miscues "definitely had to do with the structure of the language. They were attempts to get at mean- ing" (K. Goodman, l976, p. l). From this beginning, a group of re— searchers at Wayne State University began to develop a system for classifying the miscues that readers make. Between l965 and l974 intensive studies of children's miscues were conducted by these researchers. One result of this research was The Goodman Taxonomy of Reading Miscues which evolved concurrently with the studies (Allen and Watson, l976). In more recent years, re- search using the taxonomy and other forms of miscue analysis has con- tinued to increase. Recently, much of that research was brought to- gether in a review of the trends in miscue analysis research (Wixson, l979). Using Wixson's article as a framework, the research she cites as well as other relevant work is discussed in the next section. Finally, several studies more closely related to this research are discussed in detail. 41 Strategy75election Most readers, regardless of age or proficiency, produce a greater number of syntactically and semantically acceptable miscues than grapho- phonically similar miscues (Allen, l969; Biemiller, l970; Burke & K. Goodman, l970; Clay, l968; Kolers, l970; and Weber, l970a). Further, the majority of readers' miscues are syntactically acceptable (Clay, l968; Coomber, l972; Y. Goodman, l976; Menosky, l97l; and Weber, l970b). Finally, readers produce a larger percentage of syntactically acceptable miscues than semantically acceptable miscues (Allen, l969; Burke, l969; l976b; Burke and Goodman, l970; and Y. Goodman, l976). Young Readers Young beginning readers' miscues tend to include a large number of real word substitutions and ”no response" omissions. Word substitu- tions were judged simply as real words or as words that were probably previously learned (Biemiller, l970; Y. Goodman, l976, and Weber, l970a). Omissions were judged to be simply omissions by some re- searchers and to be "no response'l errors to words that readers probably perceived as unfamiliar by other researchers (Biemiller, l970; K. Goodman, l965; Y. Goodman, l976; and Weber, l970a). Mature Readers As readers mature, the percentage of omissions tends to decrease and to be replaced with substitutions of non-words and/or real words the readers have never seen before in print. Initially, for at least the first three grades, the substitutions often represent an increase in graphophonic similarity (Biemiller, l970; K. Goodman, 1976; Y. 42 Goodman, l976; Levine, l976; Schlieper, l977; and Weber, l970a). Less Proficient Readers Less proficient readers tend to produce a relatively larger per- centage of graphophonically similar miscues than more proficient readers. Further, less proficient readers make fewer attempts to cor- rect their miscues, and when they do correct they often correct accept- able and non-acceptable miscues at almost an equal rate (Allen, l976; Au, 1977; Brody, 1973; Clay, l968; and Weber, l970b). Proficient Readers As readers become more proficient, the percentage of graphophoni- cally similar miscues tends to stabilize while the proportion of syn- tactically and semantically acceptable miscues increases (Au, l977; Biemiller, l970; K. Goodman, l976, K. Goodman & Y. Goodman, l977; Greene, l974; and Jensen, l972). For some readers, this phenomenon occurs as early as first grade. Further, proficient readers omit known words that are not essential for understanding rather than un- known words, correct more of their miscues, and tend to correct un- acceptable miscues at a higher rate than acceptable miscues (K. Goodman, l976, Y. Goodman, l976; Greene, l974; Jensen, l972; and Weber, l970b). At this point, it might be helpful to return again to the model presented at the end of the previous section (see figure 2.3). The research discussed thus far represents the various types of strategies readers might select depending on maturity and proficiency, However, as noted in the model, strategy selection is not dependent on maturity and proficiency alone. The reader's perception of reading plays a 43 role as well. The reader's perception is based on his/her theoretical orientation, which is the result of previous encounters with text, previous instruction, and the structure of the text itself. For the purposes of this investigation, the discussion of the reader's perception of reading based on his/her theoretical orienta- tion is divided into two categories: the influence of instruction, which includes the assumption that any instruction is based on a theoretical orientation; and, the influence of the text, which assumes any text the reader has encountered previously was based on the author's theoretical orientation. Influence of Instruction "Instructional method appears to be a major determinant of miscue patterns, at least for readers in the first four grades" (Wixson, l979, p. l68). Cohen (l974) analyzed the oral reading errors of fifty first grade children who were being taught reading by a phonics approach. Data was collected at the beginning of each month for a nine month period. For the first four months, no response was the most common error, but by the second half of the study, the number of no response errors was equalled by that of nonsense errors and exceeded by that of word substitution errors. Cohen concluded that the high incidence of no response errors early in the study was directly related to instruction. Since the students had been trained in the systematic use of letter sounds and were not made aware of other options, their opportunities to select other strategies were limited. Thus, having been directed to focus 44 on letters, but often encountering letter-sound relationships that they had not yet been taught, the readers elected not to respond. In- terestingly, a few readers did attempt to use context but this type of error appeared only at the outset of instruction and disappeared thereafter. Nonsense errors were initially low for both groups. The good readers, however, quickly showed a dramatic increase in nonsense errors during the middle part of the year and a declining trend by the final months. By contrast, poor readers were much slower to begin to pro- duce nonsense errors and by the end of the study, the number of these errors had leveled off but had not started to decline. Since studies of other children being taught by a basal approach rather than a phonics approach made no mention of nonsense responses, Cohen concluded that this type of error was directly tied to the type of instruction received. Using word identification responses of thirty-two first graders Barr (l974) attempted to determine whether or not it was possible to determine a given reader's strategy and to what extent that strategy was indicative of the method of instruction he/she had received. The two types of instruction were phonics and sight word. A phonics strat- egy was inferred if a reader produced non words and few substitutions that could be identified as previously taught words. A sight word strategy was inferred if a reader produced only real words, 75 percent or more from the list of words previously taught. Barr concluded that it was possible to determine a reader's strategy, and that, fur- ther, strategies are influenced significantly by classroom instruction. 45 Even those students who initially form a strategy different from the classroom instructional emphasis change that strategy to conform to instruction by the end of first grade. De Lawter (l975) studied the oral miscues of l69 second graders who had been instructed for two years by means of either a decoding emphasis approach or a meaning emphasis approach. Most substitutions made by children taught by the decoding approach were non words that were graphophonically similar to the expected response. By contrast, the meaning emphasis group produced substitutions that were real words that showed little graphophonic similarity to the expected response. Using the Gray Standardized Paragraphs, Elder (l97l) investigated the responses of American and Scottish children after two and one half years of instruction. The Scottish children received instruction with a strong phonics emphasis and showed more non word substitutions than the American group, which received instruction with a sight-word mean- ing emphasis. Norton (l976) analyzed the miscues of two groups of third graders. One group received phonics instruction; the other received instruction via an analytic eclectic approach. Miscues of readers in the phonics groups were characterized by high graphophonic similarity, numerous non word substitutions, few self-corrections, and few semantically acceptable errors. Miscues of readers in the eclectic group were char- acterized by few non word substitutions, self-correction of approxi- mately half of all errors, and errors that produced semantically ac- ceptable sentences. 46 Taken collectively, these studies clearly establish the influence of instruction on the reader's perception of reading and, in turn, on the reader's selection of strategies. Specifically, all readers in these studies who received phonics instruction tended to produce non- sense or non words more than readers receiving other types of instruc- tion. Further, these non words were usually highly similar, grapho- phonically, to the expected responses. While the other approaches reported in these studies varied some- what, all had at least some emphasis on meaning. For the readers in these programs, the outstanding general trend was the production of real word miscues with little graphophonic similarity to the expected response. Further, some researchers reported that these substitutions were more often semantically acceptable, or more often resulted in semantically acceptable sentences. Finally, one researcher, Norton (l976), reported that the readers in the meaning emphasis group tended to self correct more often. Reader/Text Interaction The influence of texts, the primary focus of this study, is diffi- cult to ascertain. When a reader encounters a text, many factors must be taken into consideration. As Wixson (l979) suggests, "There is also evidence indicating that readers' miscue patterns vary as a func- tion of the interaction among the readers' skills and background, the nature of the written material, and the conditions surrounding its presentation" (Wixson, l979, p. l69). Certainly, the reader's background and skills will influence that particular reader's ability to cope with a given text. Further, as 47 stated above, the reader's background and skills are significantly in- fluenced by instruction, and instruction and materials often comple- ment each other. For example, students receiving phonics instruction for beginning reading usually encounter only those texts that exemplify the same theoretical orientation as the instruction. In other words, the texts used with a phonics approach often contain vocabulary that is restricted to previously taught sound/symbol correspondences and letter patterns. On the other hand, the texts used with a skills ap- proach often contain only those words that have been previously intro- duced. A reader, then, who encounters a steady diet of the same type of text comes to have certain expectations for print materials and to have a particular perception of reading based on prior instruction and complemented by prior experiences with texts. Texts can influence a reader's strategies, as the studies to be reviewed in this section will document. When a reader encounters a text, an interaction occurs. Both reader and text make some contribu- tion to that interaction and both aspects of the encounter must be taken into consideration. In some instances, the reader's contribution proves to be the most important influence. In others, the structure of the text provides the primary influence. Reader contributions. Hocevar and Hocevar (l978) hypothesized that if reading materials are meaningfully related to the beginning reader's concrete experiences, the reader will be able to use semantic cues in addition to syntactic and graphic cues and, consequently, read with greater facility. Students in grades one through three were given two passages to read. Each student had recently had experience, 48 through a Uhands on" science program, with only one of the subjects discussed in the two passages. On the average, the students produced 25 percent fewer miscues while reading the passage related to their recent cognitive experiences. Thomas (l975) matched comprehension of material with reader per- formance and found that the syntactic and semantic acceptability of the reader's miscues is dependent upon the degree to which he/she com- prehends the material. Rousch (l972) observed average readers in grade four and concluded that those students who lacked prior conceptual knowledge of the topic covered in the material produced more graphically similar miscues and were less aware of the need to correct their unacceptable miscues than were readers with a higher level of conceptual awareness. All of these studies support the importance of the knowledge the reader brings to a text based on his/her prior experience and knowl- edge of the language and of the content of the material being read (Smith, l97l). Further, Hocevar and Hocevar (1978) suggest that since it is possible to facilitate the beginning reader's use of semantic cues by basing reading material on recent experiences, a language ex- perience approach appears to be the best choice for instruction at that level. Text contributions. Several studies explore the effect of the text on the reader/text interaction but, because of the type of text used, are of only minimal importance to this study. They are briefly summarized here. 49 Brazee (l976) found that eighth graders produced different types of miscues depending upon whether the material was expository or nar- rative. Carlson (l970) reported that grade four readers produced fewer semantically acceptable miscues while reading content area ma- terial than they did while reading basal selections. Thornton (l973) investigated reader performance on fiction and nonfiction materials and suggested that the miscues were the result of differences arising from the author's writing style. Christie (l977) asked seven and eight year old readers to read two texts. One contained syntactic structures that appear late in language development and one contained structures that appear early. A higher percentage of detrimental mis- cues was observed on the late syntactic structures passage. All of these studies serve to further document the influence of the text. Similar Studies Five studies were found to be more directly related to this study. Dank (l976) analyzed the oral reading miscues of twenty second graders taught by either an integrated reading-language experience approach or an approach emphasizing letter-sound correspondences. Those taught by the decoding approach generated miscues that emphasized letter sound correspondences, produced more non words and fewer omissions, and showed a tendency to produce miscues with high graphic and sOund similarity. Pupils taught by the integrated reading-language experi- ence approach generated more miscues that were semantically acceptable. A major problem in this study in relation to the present study was Dank's definition of language experience. The integrated reading— language experience approach was actually a basal text (Ginn Reading 50 3_5_). Martellock (l97l) asked middle school students to read a story, give an oral retelling, write the story in their own words, and read their composition orally. Among other measures, she compared the oral reading of the students' own compositions with their oral reading of the story selection. This study closely resembles the present study in that a compari- son was made of the oral reading of a text story and a reader produced story. Martellock reported a higher incidence of miscues per hundred words in the reading of their own manuscripts than in the reading of the stimulus story and concluded that the higher incidence of miscues on the students' own compositions was due to the readers' attempts to edit their writing. Martellock's study and this study differ in several ways. First, Martellock's students actually wrote the compositions themselves while the students in this study dictated their stories to the researcher. Second, the compositions in the Martellock study were summaries of the text story, not an attempt to produce a totally new story. Per- haps the physical act of writing the composition and the fact that the composition was a summary both contributed to the reader's desire to edit during the oral reading. Further, Martellock's students were much older than the students in this study. According to Graves' (l979),editing in young children's writing is not a concern in early attempts to produce written discourse. Finally, Martellock reported only miscue and T-unit findings in her study. She did not consider the reader's theoretical orientation or the relative complexity of the 51 texts produced on any other measure except T-units. As stated earlier, when striving to determine the effect of a single text, the interaction between the reader and the text must be considered paramount, and this necessitates giving attention to a number of other factors. Y. Goodman (l967) studied six beginning readers. While the influence of the text was not a primary factor in her study, she did notice that strategy utilization varied across stories and suggested that material written in familiar language, with syntactic and semantic support, would be helpful to children in the beginning stages of reading. Andrews (l976) studied six children from a first grade classroom over a seven month period. She found that while children used some of the strategies included in the instruction, they also developed their own strategies. She concluded that the materials being read had a major influence on the readers' strategy selections, and stated, Perhaps the most significant finding of this entire study is that the material being read has more effect on how a reader reads than the instruction received, the innate ability of the reader, or any other factor looked at (Andrews, l976, p. 253). Because both Goodman and Andrews noted that "strategy effective- ness was determined in part by the familiarity of language, story line, and story content" (Rhodes, l979b, p.33) Rhodes conducted a study with beginning readers to determine the effect of more or less predict- able books. She defined ”predictable" as stories on familiar topics that are written in natural and familiar language with difficult words and complex structures included, if they are a natural part of a good story and enough content is supplied to support the efforts of the readers to use effective strategies (Rhodes, l979b, p. 22). 52 Rhodes found that not only did readers' strategies vary across texts but that they also varied within a given text. As a story be- came more predictable, students began to use more proficient strate- gies and this phenomenon was more prevalent for the more predictable stories than for the less predictable stories. The findings of this analysis of stories by story quarters is consistent with an earlier study by Menosky (l97l) in which she found that reader's miscues change qualitatively as they progress through a story, provided the passage is of sufficient length for the readers to gain contextual support. The research described here is an outgrowth of the study by Rhodes. Based on the psycholinguistic notion of predictability and on the language experience assertion that materials that represent the reader's own experiences facilitate the reading process and are, therefore, predictable, this study examines whether or not reader authored materials facilitate the reading process both because they are more predictable and because they provide more support due to familiar language structure and meaning. Summary In the first section of this chapter, four reviews of language experience research were summarized. Research in that area has under- gone several phases. Early studies investigated only the language experience approach and no attempt was made to compare methods of in- struction. Later studies focused almost exclusively on comparison of methods, and language experience was generally found to be a viable method of reading instruction. Recent research has begun to focus on 53 other factors but no study that attempted to describe the influence of language experience materials on the reading process was located. In the next section, reading was described as a socio- psycholinguistic process in which readers predict, confirm, and compre- hend. Learning to read was characterized as one specific example of how learning in general occurs. Both the reader's knowledge of the world and of language were deemed important to the reading process. Particular knowledge of language included the reader's implicit knowl- edge of the three cue systems of language and the ability to use these systems proficiently. General world knowledge that is specific to the reading process included the reader's perception of reading based on prior-encOUnters with print as well as prior instruction. A socio-psycholinguistic model of reading was presented. Finally, recent miscue analysis research was summarized in an effort to describe strategy selection for mature and less mature as well as proficient and less proficient readers. Since strategy selec- tion is not dependent on maturity or proficiency alone, other research concerning reader/text interaction was also described. Those studies most relevant to this study were described in detail. CHAPTER THREE: METHODOLOGY The purpose of the present study is to compare, using a miscue analysis procedure, the performances exhibited by beginning readers as they encounter a text story and their own dictated story. Procedures The impetus for this study was Hall's suggestion that language experience needed to be re-examined from an in-process perspective. Under perfect conditions, such an examination would include a sample of beginning readers who were asked to read both a basal text story as an example of their usual encounters with reading and some kind of dictated story as an example of the language experience approach. The sample of beginning readers should have had the same instruction, used the same texts, been exposed to the same theoretical model, and acquired approximately the same level of proficiency in reading. In other words, the constraints should be kept as constant as possible. In this study, every effort was made to accomplish this objective; however, some factors were not within the control of the researcher or were inherent in the instruments used. When these constraints appear to have some bearing on the study, they are noted. While the research conditions were not always the best, an effort was made to keep them consistent. m The design of this study has undergone several revisions over the past two years. As a result of two pilot studies, changes were made 54 55 in the method of eliciting the dictated story, the choice of the word- less picture book, the process used for selecting the sample, and the coding procedures. Elicitation method. Initially, this research was undertaken in an effort to clarify procedures for eliciting a story to accompany a wordless picture book. Past experiences with both pre-service and in- service teachers had shown that, while they are interested in using the language experience approach, they are uncomfortable with taking dicta— tion at the beginning levels of instruction. Specifically, many teach- ers feel a need to be in “control" of the interchange with the students, but don't know how to accomplish this. Most attempts to control result- in control of the product (changing the young author's choice of words, syntax, or overall plot of the story) rather than control of the pro- cess (the situational constraints or the interchange between author and scribe). Many teachers are concerned by the lack of "correctness" in their students' endeavors to produce texts. Certainly, the language of most six year old children does not fit the adult norm. However, at this stage of development, children need opportunities to explore communi- cating, not correctness. Six year olds already speak differently from the way they did when they were four or five, and they will continue to develop their linguistic abilities if they are given opportunities to do so and not stifled by the teacher's requirements of "correct- ness." Only after children have learned the “whys" and "whens" of communication, does refinement become significant. "Until then, child- ren must manage to say practically everything they ever have a need to by learning to handle with ease the language that is their own (Stauffer, 56 l970, p. 56). Control of the interchange, then, must focus on the process, not the product. "The effectiveness of any language experience story is directly based on the quality of the interchange between the author and the scribe. The scribe, as the more experienced of the two, has the responsibility for developing and maintaining this interchange" (Burke, l976, p. 2). The most critical part of the language experience interaction is discussion of the experience prior to dictation. Just as adult writers often need the opportunity to talk about and discuss their ideas before committing them to paper, so do children. The initial discussion should be general in nature and, at the same time, allow children the oppor— tunity to organize their ideas through the familiar communication mode of speech. ”Following the general discussion, the teacher should ask specific questions designed to (l) bring out significant points that were missed, (2) suggest alternatives, and (3) provide sequence" (Burke, l976, p. 3). Using the above suggestions as a guide, in the fall of l978, a method for eliciting stories with wordless picture books was developed. (See Appendix A). In this method, particular importance is paid to giv- ing the children opportunities to discuss the total story before begin- ing to dictate. Further, if the initial, general discussion lacks coherence or detail, a series of open-ended questions, designed to probe for more information, follow. In this study, a scribe procedure was chosen in order to ensure elicitation of a complete story that could be compared to the story in the basal text. It was assumed that the added burden of producing the 57 text themselves might prevent the students from attending to the story production. While this procedure was deemed necessary for the present study, it did over simplify the process and may have hindered some students. Actually handling the process of composing, in all its complexity, opens more avenues for exploration of the process and may give readers more support. Harste, Burke, and Woodward (in progress) have found that pre-school children are more willing to read composi- tions they produce themselves than those elicited through dictation even though their own compositions are often not in conventional, representational symbols. Thus, while the scribe procedure was appro- priate for this study, over simplification of the composing process for beginning readers is probably not necessary and may even be detrimental. In instructional settings, emergent readers should have many opportunities to handle the composing process themselves as well as opportunities to dictate stories. Selection of texts. While the elicitation method was being de- veloped, efforts were also made to locate an appropriate wordless picture book. Of the fifty or more available in the local library, most were rejected as poorly illustrated, too abstract, or lacking a story line. Many wordless picture books are designed to teach color, letter, and number concepts and, as such, are not stories. Several Mercer Mayer books were tried, but the stories produced from them by the children were often full of detailed description rather than a story rendition. Finally, an acceptable book, The Bear and The Fly (Winter, l976), was located (Appendix K). The use of this book in two pilot studies as well as the present study has proved quite success- ful. Over twenty children have now produced stories to accompany this 58 book and, in every instance, a fairly well-formed story was produced. A word of caution is in order here. The choice of a wordless pic- ture book to elicit a dictated story was a purposeful choice made by the researcher. Because the children's reading of the dictated story was to be compared with their reading of a basal text story, a rather well-formed dictated story consisting of a protagonist and a series of goal oriented attempts was needed. If the children had been asked to produce a story about a picture, or a group of objects, or a past ex- perience, the results might be very different. For example, in their research with pre-school children, Harste, Burke, and Woodward (in pro- gress) are finding that the children produce more well-formed stories when given the freedom to choose their own subject than when given a wordless picture book (Harste, personal communication, May 30, l980). For the selection of the basal text story, other criteria were used. In order to ensure that the text story closely reflected the model of reading held by the teacher and the students, a selection was made from the basal series (Houghton Mifflin, l97l) in use in the classroom. At the time of testing, the students had just completed the level two pre-primer, Lions, The story, ”Red is Nice," was chosen from the next level pre-primer, Dinosaurs, because it was the most complete story available in that text. None of the children in the study had read the story before. The most important criteria for the selection of both the wordless picture book and the basal reader story was predictability. One aspect of a predictable book is semantic and syntactic redundancy (Rhodes, l976, p. l98). If both the content (semantic) and the language (syn- tactic) are predictable, the story gives the reader more support. In 59 the story, The Bear and the Fly, each time the fly lands on someone, the father bear tries to hit the fly, misses the fly, hits the person, and knocks out that person. In the basal story, "Red is Nice,” the two boys are painting. Each time they finish painting something, they discover they have spilled paint on something else. As a result, they have to paint everthing they spilled paint on until they have painted the tree- house, the doghouse, and the fence. In both stories, the redundancy of the plot enables the reader to predict what will occur in the next epi- sode. Semantically, both stories were judged highly predictable. Syntactic redundancy, which also gives support to the reader, can be accomplished through repetition. In the story, "Red is Nice," simi- lar sentences occur throughout the text: Sentence 8: Red is a good color for a treehouse. Sentence l2: And red's a good color for a doghouse. Sentence 22: Red is a good color for a fence. Sentence 45: Red is a good color for a house. Sentence l3: We'll paint the doghouse red. Sentence 2l: We'll have to paint the fence. Sentence 23: We'll paint the fence. Sentence 44: We'll have to paint the house. Because each repetition is tied to a particular episode, the read- er can expect to find a similar sentence in the next episode. This kind of syntactic redundancy enables a reader to make predictions about the language he/she will encounter and further adds to the support given by the text. Syntactically, "Red is Nice" was judged to be highly pre- dictable. Since The Bear and the Fly has no text, it was not judged on a syntactic measure. 60 Another aspect of a predictable story is the close interrelation- ship“between the text and the illustrations; that is, the story should be as easily recovered from the text as from the illustrations. There- fore, illustrations were one of the prime considerations in the choice of both the wordless picture book and the basal text story. The Bear and the Fly was chosen because the illustrations seemed to clearly carry the story. The stories produced by students in the pilot studies con- firmed this assertion. All the stories were fairly well formed, and quite similar, indicating that the residual story grammar inherent in the pictures was recognized by the students and used in their attempts to generate the text. In "Red is Nice," the text had already been generated by an adult author. While this story does exhibit both a syntactic and a semantic redundancy, the match between the text and the illustrations could have been better. (See results under story grammar in chapter four.) Choice of the text story was limited to the pre primer Dinosaurs (Houghton Mifflin, l97l) because it was the next level of the series used in the classroom. Even though some elements of the story line were left out of the text and could only be gleaned from the pictures, when all three aspects of predictability were considered, "Red is Nice" was judged the most predictable story in the text. Pilot studies. A pilot study was conducted in the fall of l978 and another one in the spring of l979. The first study was informal and included only four children-~two first graders and two second grad- ers. The major results of this study were the development of the eli- citation method and the choice of the wordless picture book. In the spring of l979, a more formal study was conducted. Ten 61 children from a first grade classroom in a surburban school located in a middle-sized, midwestern town participated in the study. As in the previous study, the children were first asked to read a story from their basal reader. Several stories, one from each reading level (primer through beginning third grade) were selected. Some children read more than one selection. The story that elicited enough miscues to study reading performance, but not so many as to render the reading meaning- less, was elected for coding. Each of the ten children in the pilot study also produced a story to accompany the wordless picture book. The major problem with the second study was the difficulty of com- paring results. Performance on one dictated story was easily compared to performance on other dictated stories; however, because several dif- ferent text stories were used in this study, comparisons could not be made across that data or across the two types of stories. Therefore, a selection process had to be developed that would ensure the use of the same text story for all readers. The process developed is described further under the heading "sample.“ Sample Eight first grade students, two female and six male, from a racial- ly integrated, suburban school in a mid-sized, midwestern city partici- pated in this project. The school was chosen because the researcher had worked, previously, with the principal. The teacher was selected by the principal because she expressed a willingness to participate in the study. All students were granted permission, by their parents, to participate in the study in accordance with the guidelines set by the University Committee for Research Involving Human Subjects. (See Ap- pendix B for a copy of the consent form.) 62 Six of the students were six years old and two were seven years old. One student was repeating first grade and one student had been moved up from kindergarten during the second week of the school year. These eight students were chosen from an initial sample of fifteen (the top two reading groups) selected by the classroom teacher as pos- sibly proficient enough to read the text story that had been chosen for the study. In the pilot studies, children had been asked to read several text stories until they reached one that produced enough miscues to evaluate their reading performance. This procedure resulted in a number of dif« ferent text stories and did not allow for a comparison of performance. In the final study, one text story was selected. All fifteen students read the story. Those who made too few miscues were eliminated. Sim- ilarly, those who made so many miscues as to render the text meaningless were also eliminated. Eight children met this criteria. Data Collection In early February, l980, the researcher began working in the class- room. The first two weeks were spent observing, assisting the teacher, and becoming acquainted with the children. Data was collected during the last two weeks of February and the first two weeks of March. An effort was made to collect data at similar points in instruction. Therefore, the top reading group was tested in February, after they had completed the second pre-primer. In March, the second group was tested after they, too, had completed the second pre-primer. In actuality, only a week to ten days separated the dates of data collection. By the time the first group had been tested, the second group was ready for testing. Because miscue research dictates that the story read must be 63 one the children have not seen before and of sufficient difficulty to elicit miscues, the text story was selected from the third pre- primer of the series in use in the classroom. The collection of data required three meetings with the researcher. In the first meeting, each child was taken from the classroom to another location where he/she was asked to read the text story, do a retelling, and answer questions about reading from The Reading Interview (Burke, l978). In the second meeting, each child was asked to dictate a story to accompany the wordless picture book. In the third meeting (the following day) the child's story was returned and he/she was asked to read it. After the reading of both the text story and the dictated story, the child was asked to retell the story. Audio tapes were made of all three meetings. In an effort to make the tasks similar, in meeting one and two, the following procedure was used. In both cases, the child was asked to look through the text and predict from the pictures what the story was going to be about. In the case of the text story, the child was then asked to read the text. In the case of the dictated story, the child was then asked to dictate his/her story. The following day, the dictated stories were returned for reading. Again, a word of caution is advisable. Several constraints were operating in the data collection procedures. First, the location often changed. Some settings were more private and included fewer distractions than other settings. Also, Harste, Burke, and Woodward (in process) have discovered that children sometimes include information from the environment in their dictations. That is, some 64 of their subjects elected to include objects from the physical setting other than the provided stimulus in their stories. (Personal communication, Harste, May 30, l980). While this phenomenon does not appear to be a factor in the present study, certainly the variety of locations for conducting the research might have an effect; however, researcher control of this particular variable was not possible. Second, all data was collected by the same researcher. Generally such consistency is a positive attribute because it generates trust on the part of the subjects. However, in miscue analysis research, subjects are told to read the selection out loud and, if they come to something they don't know, do whatever they usually do when reading alone. They are also told that the researcher cannot give any help during the reading. Some students, especially those who have received instruction based primarily on a phonics and/or skill model, have never been asked to perform this kind of task. Essentially, the instructions ask the reader to be a risk taker and to reply solely on his/her own strategies. While no student in this study refused to follow the directions, several did ask for help a number of times before acquiescing to the instructions. Certainly, a familiarity and subsequent trust of the researcher must be of some importance in this situation. In the case of the retelling, the same researcher variable may be a detriment. Very often the request for a retelling from the same person who heard the oral reading results in a less full retelling than the same request from a person who did not hear the oral reading. In other words, students seem to perceive the request for a retelling of the story as somewhat redundant since the researcher has just 65 heard the oral rendition of the same text. While these particular constraints could not be controlled by the researcher, they were held constant across both texts. Finally, in the collection of the dictated story, students were not allowed, at the end of the dictating session, to read the completed version for the purpose of editing. As suggested below: The fully generated story should be read through without interruption before the close of the session. This reading allows the students to edit any material they find awkward, inappropriate, or inaccurate within the context of the whole. It allows the teacher to catch any inadvertent alterations that may have been made in recording the learner's language. But most important to the learning process, this reading acts to complete the cycle from uninterrupted whole. . . , through consideration to and analysis of the components. . . , back to uninterrupted whole. . . . (Y. Goodman & Burke, l980, p. l9l) Due to concerns over students either using immediate recall if they read their dictated story before the close of the session, or students having a practice opportunity if they read the selection before the end of the session and again a day later for the purpose of coding, the end-of—the-session reading was eliminated. This particular constraint was unfortunate in that it violates a necessary component of the language experience approach. However, the elimination of this particular step was allowed in order to reduce any undue support for the dictated story. Analysis The focus of the two pilot studies was the influence of text on the performance of readers. To that end, analysis was directed at describing the reading behavior on the two tasks. The Evaluation Form of the Reading Miscue Inventory (Y. Goodman & Burke, l976) was 66 used in both cases. While the differences observed proved interesting, the mere description of the differences did not explain why_they occurred. Certainly the two texts did make a difference in strategy selection, but other factors needed to be considered as well. Reading is, after all, a transaction between the reader and the text. Strategies selected are only the outcome of that transaction. What was needed was a fuller understanding of both the reader and the text. In the cyclical process of reading as described in chapter two, readers contribute their perception of reading based on their theoretical orientation. This theoretical orientation is, in turn, derived from past encounters with texts. Texts, too, are based on theoretical orientations that result in a variety of kinds of texts. Thus, both the readers' perceptions of reading and the structures of the texts needed investigation as well. Procedures for obtaining this information are given in detail under the "coding procedure" heading. Coding Procedures As stated above, the primary focus of this study is the influence of text on the performance of readers. To that end, strategy selection as measured by the Evaluation Form of the RMI (Y. Goodman and Burke, l976) was coded first. In addition, an attempt was made to describe the reader's perception of reading as well as to describe the structure of the text. Strategy Selection. On a sentence level, the reader's performance was coded according to the syntactic and semantic acceptability of the sentence produced as well as the degree to which the finally produced sentence deviated from the author's intended meaning. On 67 a word level, miscues per hundred words, were coded as well as correction attempts, sound and graphic similarity, and similarity of grammatical function. In a miscue analysis procedure, the first step is to listen to the audio tape and mark the miscues on a previously prepared copy of the story as illustrated below: Qatar, OlOl l. Do you want to go to the library, Ken? Ol02 2. I can't go with you now. (3) (anur’ OlO3 3. I have to paint this tree-red,(. . owl -’3. T gaf C) 4:. @691ng m _ OlO4 I'm(good at painting tree- 0105 5. I can help youSpaint it. C OlO6 6. Then we can go to the library. In this example, miscues numberfl @, ®, and®are all substitutions. In addition, miscue number l becomes a repeated miscue (AM) in line OlO6. Repeated miscues are only counted the first time they occur. Miscues number(:)and(:)are omissions and miscue number (Dis an insertion. This reader also made some attempts to correct. In line OlO6, the reader's attempt was successful (CD). In line Ol04, the reader made two attempts, but both were unsuccessful «:D). In addition, the second attempt resulted in a nonsense word (f). In line OlOS, the reader first said, "I can help you paint it." Then, he/she regressed, abandoned the correct reSponse «:D), and said, "plant it." 68 The next step in coding is to record the information on the coding sheet. (See Appendix C for a sample coding sheet.) Each miscue is listed on the coding sheet as illustrated below: READER TEXT CORRECTION l letter library N 2 ------ house N 3 now ------ N 4 going good N 5 trees tree N 6 ------ houses N 7 plant paint P 8 letter@ . library Y Miscues are coded for graphic similarity, sound similarity, and grammatical function. In each case, the degree of relationship is noted: High (Y), some (P), or none (N). For example, since not much sound or graphic similarity exists between letter and library, the first miscue would be coded (P) for some relationship. However, since letter and library are both functioning as nouns, the miscue would be coded (Y) on grammatical relationship because both words can fill the same grammatical function. Omissions and substitutions such as miscues @, Q, and @as well as repeated miscues, @, are not coded on this measure. Finally, miscues are coded for correction, (Y), no correction (N), and partially correct (P). On a sentence level, the whole sentence, as finally produced, is coded for syntactic acceptability, semantic acceptability, and meaning change. While the reader's first attempt is used for word level coding; at the sentence level, the reader's final attempt is coded. The following examples are given to explain the coding procedures: CZZHZEL, OlOl l. Do you want to go to the library, Ken? In this example, the reader's sentence, as finally produced, is 69 "Do you want to go to the letter, Ken?" While this sentence is Syntactically'acceptable(the reader replaced a noun with a noun) it is not semantically acceptable because it doesn't make sense. Further, it results in a meaning change because the sentence, as produced by the reader, differs from t e author's intended meaning. @féanct Ol05 5. I can help you paint it. In this example, the sentence, as produced by the reader, is both syntactically and semantically acceptable. However, the author's intended meaning is significantly changed by the miscue "plant" for "paint." OlO6 6. Then we can go to theilibrary. In this case, the sentence, including the correction, results in a syntactically and semantically acceptable sentence that coincides with the author's intended meaning. For the purpose of comparison, a mean score was computed for each of the measures described above and for each of the texts. That is, a group average score was used to compare performance on both "Red is Nice“ and The Bear and the Fly for the following measures: syntactic acceptability, semantic acceptability, meaning change, miscues per hundred words, correction, graphic similarity, sound similarity, and grammatical function. In addition, several other comparisions were made. I Because the three least proficient readers in the study made more omissions than substitutions on the text story, this strategy was investigated further. As noted in chapter two, as readers mature, the percentage of omissions tends to decrease and be replaced with substitutuions of non-words and/or real words. The relationship 70 between omissions and substitutions on the two texts was investigated further to determine whether or not the dictated story could make these three students look like more mature readers. A major constraint in the use of RMI coding procedures for determining the reader's selection of strategies is the enormous responsibility placed on the researcher for decision making. Regard- less of the amount of experience the researcher has had with the procedure, new, not-previously-encountered examples of readers' inter- actions with texts often evolve, and consistency is difficult to maintain. However, no other coding procedure, consistent with psycholinguistic theory, is available. The researcher has had con- siderable experience with the Evaluation Form of the RMI. She spent the l978/79 school year at Indiana University where she worked with Carolyn Burke, one of the authors of the Evaluation Form, and Jerome Harste on a project involving the use of the procedure. Further, Margaret Siegel, a graduate student at Indiana University, who also is proficient with the procedure, coded a story judged to be the most representative of the types of miscues generated by the students in the study. The two raters agreed on syntactic and semantic acceptability of sentences 94% of the time. On meaning change, they agreed 88% of the time. Of the total possible miscues (73), one rater coded 69 (94.5%) and one coded 68 (93.2%). Agreement on correction strategies was 95%; graphic and sound similarity, 91%; and grammatical function, 95.7%. While other measures of strategy selection could have been used (reduction of non-words on the dictated story, for example), the ones presented here were chosen on the basis of the researcher's familiarity 71 with the data in conjunction with suggestions made by Jerome Harste and Carolyn Burke. Reader's perception of reading. This category was measured by a description of prior instruction, stated perceptions of reading as measured by The Reading Interview (Burke, l978; Appendix D), and verbalizations made while reading the selection. Prior instruction was measured by the number of days the students had received both phonics and basal instruction, and the number of basal stories read as opposed to opportunities to engage in language experience activities. This information was provided by the teacher. Reading Interviews (Burke, l978) were conducted with both the students and the teacher. As reported in chapter two, Burke and Harste (l977) found that both the teachers' and the students' perceptions of reading as measured by the interview were consistent with classroom behavior. In the present study, the interviews were used to document the models of reading held by both the teacher and the readers as a means of explaining the readers' performances on the texts. Verbalizations (the asides readers make while engaged in the reading process) were used to determine the reader's focus. Rhodes found that children's verbalizations focused more on meaning, while reading the predictable stories, and more on words (or the situation surrounding the reading or the research task) while reading the less predictable stories (Rhodes, l979b, p. lOO). Again, other procedures could have been used; however, prior instruction, stated perceptions of reading, and verbalizations were chosen either because they were consistent with theory or because the data seemed to dictate their choice. 72 Structure of the text. A variety of methods for determining the structure of the text are available. In this study, the following were used: Readability (Spache, 1974), number of untaught words, T-units (Hunt, l966), and story grammar (Stein 8 Glenn, l977b). In addition, the retellings given by the students were scored as a measure of their compre- hension of the text. Readability measures and counts of previously untaught words are similar in that they both focus on word level measures. The Spache for- mula also takes into consideration sentences, but only as a function of length, not syntactic complexity. The Spache formula (l974) was selected for this study because it claims to give the most accurate information for primary level materials. Most primary formulas have a probable error of estimate of six months to a year. In contrast, the standard error of estimate of our new formula is 2 months, i.e. in 68% of the samples, the true reading level will be within plus or minus two months of the estimate found. (Spache, l974, p. l98) Computation of readability using the Spache formula involves the following steps: number of words divided by number of sentences to give an average sentence length figure; multiplication of this figure by .l2l; a count of the number of "unfamiliar“ words, i.e. those words not found on The Revised Word List (Spache, l974); multiplication of this figure by .082; addition of the two figures computed above and the con- stant, .659. (See Appendix E.) For whole texts or very long stories, the above process is done three to five times on samples of 100 words. Since all the stories in this study were relatively short, the entire contents were analyzed. Further, each story was analyzed twice by the researcher and a third time by another person. 73 A recurring problem in the basal story was the use of the ”'5" form of contractions such as "red's" and "there's." According to directions for the application of the formula, contractions such as ”didn't" are considered unfamiliar words unless they appear on the list in exactly the same form. In other words, even if "did" and ”not" appeared on the list, "didn't" would be considered unfamiliar. On the other hand, the "'5" form of possessives is considered familiar. Since the "'5" form of con- tractions seemed most like the ”'5" form of possessives, if the base word was on the list, such words were not counted as unfamiliar. This decision was later verified by Spache (personal communication, June, l980). In order to obtain a word level measure more particular to the sub- jects in the study, the number of previously untaught (not presented yet) words were counted for both the basal story and the dictated stories. At the time of testing, all students had completed levels A, B, and C of the Sullivan Readiness Reader (Behavorial Research Laboratories, l969) and the two preprimers, Iiger§_and L19n§_(Houghton Mifflin, l97l). A list was complied of all the words presented in these five books. Any word not on the list that occurred in a story was considered "untaught.” Since the purpose of this study was to determine the effect of the text on the readers' strategies, two other measurements were made at the word level--the number of unfamiliar (not on Spache list) and the number of untaught (not previously presented) words correctly identified. For example, ”Red is Nice" contained four unfamiliar words according to the Spache word list while The Bear and The Fly contained from four to sixteen unfamiliar words. For each child, both the number of unfamiliar words and the number of these words correctly identified was calculated. 74 On the untaught words measure, ”Red is Nice" contained twenty-three untaught words while The Bear and The Fly dictated stories contained from thirty-one to eighty-three. The same calculations were made on this measure. Because the readability measure and the untaught words measure were both at the word level, an attempt was also made to determine the difficulty of the two texts on a syntactic level. The most commonly used syntactic measure is the T-unit. A T-unit is defined as a "minimal terminable unit" and "consists of exactly one main clause plus whatever subordinate clauses happen to be attached or embedded within it" (Hunt, l966, p. 737). The following example illustrates the coding procedure: l independent clause; l T—unit I drove downtown l independent clause, with I drove downtown and compound predicate; l bought some art sup- T—unit plies 2 clauses, l independent I drove downtown and and l subordinate; l bought art supplies T-unit after I ate my breakfast 2 independent clauses; 2 I drove downtown and I T-units bought some art sup- plies (from Malmstrom 8 Weaver, l976, p. 346) Although the relationships between words, clauses, T-units, and sentences can be measured in,a variety of ways, Hunt's study (l965) suggested that the length of T-unit (words per T-unit) was the best measure of syntactic maturity; therefore, length of T-unit was used to measure syntactic complexity of both the basal story and the dictated stories. One of the problems with the T-unit measure resulted from the 75 researcher acting as scribe. While every effort was made to terminate sentences in print in the same place as the children terminated them in their oral rendition, errors are possible. As the children dictated, each sentence was repeated by the researcher once for approval from the student and a second time, word by word, as the sentence was being written. Further, each session was audio taped and these tapes were used in conjunction with the written version of the story to produce the typed version. Finally, the semantic measure of text structure caused the most difficulty. Various approaches were tried (Kintsch & van Dijk, l978; Mandler & Johnson, l977; Halliday & Hasan, l976). Text analysis is a relatively new field and the coding procedures are not well developed. Generally, the major problem was the basal story. This story, like all the stories in the pre-primers of the Houghton Mifflin series (l97l), is written totally in dialogue. The reason for this particular form is not known by the researcher, but a statement from the Intro— duction to the teacher's edition of the series helps to shed some light on the problem: To read, a pupil must be able to convert printed language into the oral language for which it stands, whether he actually speaks the words and sentences or only thinks how they would sound if he heard them. (from the Teacher's Edition of Tigers, l97l, Houghton Mifflin, p. 7) Perhaps the authors believe that dialogue represents the most familiar form of oral language. Whatever their reason, the story was not easily coded. Of all measures tried, the story grammar approach developed by Stein and Glenn (l977b) proved the most successful. In addition, 76 of all the text analysis research available, Stein & Glenn have done the most work with early elementary children. Thus, their approach was deemed the most appropriate for this study. Story grammars are an attempt to map the abstract cognitive structures that account for how we organize information in long term memory. Most research in this area has centered on recall. However, Applebee (l978), Brown (l977), and Stein and Glenn (l977b) have used story grammars to document the development of story conventions in children. Despite the discrepancies in methodology and purpose, these studies, taken together, have a common thread: Whether or not it is made explicit, all are based on the assumption that individuals tell, retell or recall stories on the basis of an internalization of the story structure-- a story schema--that has been acquired and guides the production or reproduction of a story. (King & Rentel, l979, p. 5) In present study, students were asked to tell (dictate) a story to accompany a wordless picture book. The kinds of stories produced are assumed to be dependent on the acquired story shcema of the authors. Therefore, the stories are likely to differ in complexity. In the Stein and Glenn story grammar, certain elements have been shown to be required for well formed stories. If those elements are missing, the stories are assumed to be less well formed and, therefore, less easily recalled (Mandler & Johnson, l977, p. l38). According to Stein and Glenn, a story consists of a setting and an episode system. Each episode can be further subdivided into the following categories: initiating event, internal response, internal plan, attempt, direct consequence, and reaction. The setting contains both major and minor information, introduces the protagonist, and also gives information about the story context. The initiating 77 event contains an event or action that causes the protagonist to form- ulate a goal. Goal formulation comes about through both the internal response, which may be affective states or cognitions that serve to motivate the protagonist; and the internal plan, which may be subgoals that exist as prerequisites for attainment of the major goal. The attempt includes statements that refer to character's overt, goal— directed behavior resulting in a consequence. A consequence marks the attainment or non-attainment of the goal. Finally, the reaction defines how a character responds to the attainment or non-attainment of the goal (Stein & Glenn, l977b). A sample of a simple episode follows: Melvin, The Skinny Mouse Setting l. Once upon a time, there was a skinny mouse named Melvin . who lived in a big red barn. 3. One day, Melvin found a box of rice Initiating crispies underneath a stack of hay. Event 4. Then he saw a small hole in the side of the box. Internal 5. Melvin knew how good the cereal tasted Response 6. and wanted to eat just a little bit of cereal. Internal [?; He decided to get some sugar first. Plan 8. so that he could sweeten his cereal. 9. Then Melvin slipped through the hole Attempt X; in the box 0. and quickly filled his cereal bowl. ll. Soon Melvin had eaten every bit of Direct the rice crispies Consequence 2. and he had become very fat. . 3. Melvin knew he had eaten too much Reaction [}4. and felt very sad. (Stein & Glenn, l977b, p. 6l) Although all categories are important to a truly well-formed story, some occur more frequently than others. Internal reSponses and reactions, for example, are often ommitted from both folktales 78 and fables (Stein, l978, p. l0) and from children's recalls of stories even if the categories were included in the stories heard (Stein, l978, p. 26). The most frequently recalled categories are the initiating event and the consequence. Further, in a study in which each category was systematically removed, the deletion of the initiating event caused recall of fewer remaining statements for both first and fifth grade students. For the first graders, deletion of the consequence also caused problems (Stein & Glenn, l977c, p. 6). Based on these findings, Stein and Glenn report that the initiating event and the consequence are the most salient characteristics of the episode system (Stein & Glenn, l977c, p. 7 and Stein, l978, p. 26). Further, the major setting statements (not a part of the episode system) are the most frequently recalled statements of all (Stein & Glenn, l977b, p. 9l and Stein, l978, p. 26). Saliency of attempts fall somewhere in the middle. Thus, on a scale of descending frequency, the categor- ies can be listed as follows: Setting Statements Most frequent Initiating Event Consequence Attempt Minor Settings Least frequent Internal Response Internal Plan Reactions Based on their research, Stein and Glenn define a complete episode as containing at least the following: an initiating event (or an internal plan if it includes the goal statement); an attempt; and a direct consequence. Any episode without at least these three categor— ies is considered incomplete. In addition to the delineation of categories and the determination 79 of those categories necessary for a complete episode, Stein and Glenn also specify the causal links between episodes of a story. Three relations between episodes are possible: AND, THEN, and CAUSE. THEN and CAUSE relations are the most common. A THEN relation denotes a temporal juxtaposition and may indicate that one episode is a pre- requisite for the following episode. A CAUSE relationship implies a direct causal connection between two episodes. That is, succeeding episode can only occur because of the conditions set in the prior episode. An AND relationship implies that the two episodes occur simultaneously. This kind of relationship does not occur often. Further, a fourth kind of episodic relationship occurs when one episode is embedded in another. In this kind of relationship, the consequence or reaction of one episode may also function as the initiating event for the next episode (Stein & Glenn, l977b, pp. 67-70). The descriptions given here are designed to illustrate the story grammar as it applies to a very simple and extremely well formed story developed by Stein and Glenn. In real stories, either established folk tales and fables or stories produced by children, the episodes and the categories within them are sometimes not easily explained by the grammar. Many incomplete episodes occur as well as disrupted orders of categories. Further, except for one study Stein and Glenn (l977a) appear to be more interested in the recall of stories than in the production of them. In the l977a study, they did collect stories produced by kindergartners, third, and fifth graders. The stories were elicited by giving the students setting information followed by a request to tell an appropriate story. This study was conducted on the belief that "the type and sequence of categories generated in 80 spontaneous stories should be similar to the pr0posed internal repre- sentation" (Stein & Glenn, l977b, p. ll8). However, critical differences between story comprehension and spontaneous story production surfaced in this study. Most of the more well developed stories did conform to the higher order episodic structure of the grammar, i.e. most had at least initiating event and/or response, and attempt, and a con- sequence; however, the logical structure did not fit all stories: In particular, many stories were logically much simpler and causal relationships between statements were either non-existent, or poorly defined and elaborated. In order to describe these stories, it was necessary to define simpler logical structures such as the Descriptive Sequence and the Reactive Episode, and to describe new informational categories, such as the Activity which indicated non-directed behavior. (Stein & Glenn, l977a, p. lO) The application of this grammar, or any story grammar, to stories produced by children is a highly tentative procedure. As was stated above, the field of text analysis, of which story grannmrs are one small part, is relatively new and the coding procedures are not well developed. As Stein and Glenn have noted, "It is possible...that the category definitions will have to be changed or modified as data are collected” (l977b, p. 58). Adaptations appear to be particularly needed when the grammar is applied to stories generated by young children. In the present study, the story grammar developed by Stein and Glenn (l977b) was used to further explore the structures of both the basal story and the stories produced by the children. Specifically, the use of setting statements, the structure of the episodes, the completeness of the episodes, and the relationships between episodes were investigated. As in other measures used in this study, the 81 coding procedures were checked several times by the researcher. In addition, Karen Feathers, a graduate student at Indiana University who has worked with a variety of text analysis coding procedures, also checked the application of the story grammar to the stories in this study. In addition to word level measures, a syntactic measure, and the story grammar investigation, retellings were also collected. At the conclusion of the reading of both stories, each child was asked to tell what he/she remembered about the story. The retellings were audio taped and later transcribed. The transcribed retellings were then compared with a previously developed retelling guide (Appendix F) to determine the reader's recall of characters, development of characters, events, plot, and theme. The scoring of a retelling used in this study is an adaptation of the procedure developed by Y. Goodman and Burke (l972) and includes a possible lOO points. Retelling scores were compared with each other as well as with the story grammar results in order to determine whether or not the structure of the story affected the reader's ability to produce a retelling. Expected Outcomes A list of expected outcomes was stated in chapter one. These are repeated here with inclusion of the specific instruments and procedures used to measure the results. Measures of Strategy Selection More sentences, as finally produced by the reader, will be judged both syntactically and semantically acceptable on the dictated story than on the text story. Sentences, as finally produced by the reader, will less often result in a meaning change on the dictated story than 82 on the text story. Sentences, as finally produced by the reader, will less often result in a meaning change for the dictated story than for the text story. Readers will produce fewer miscues per hundred words (MPHW) on the dictated story than on the text story. Of the miscues produced, readers will more often correct those miscues on the dictated story than on the text story. Miscues produced on the dictated story will be less similar on a phoneme-grapheme measure than those produced on the text story. Miscues produced on the dictated story will be less similar on a phoneme-grapheme measure than those produced on the text story. Miscues produced on the dictated story will more often be of the same grammatical function than those produced on the text story. Procedures developed by Y. Goodman and Burke for both the Reading Miscue Inventory (l972) and the Evaluation Form of the Reading Miscue Inventory (l976) were selected to measure strategy utilization. In addition, since omissions and substitutions are powerful indicators of reader maturity, the decision was made to measure these two types of miscues separately. Measures of Reader's Perception of Reading The reader's perception of reading will help to explain performance (strategy selection) on the two types of texts. Prior instruction, theoretical model (both students and teacher), and verbalizations made during the reading process were selected to measure the reader's perception of reading. Information about the number of days the students had received both phonics and basal instruction and the number of basal stories read as opposed to opportunities to engage in language experience activities was collected from the teacher 83 in order to measure prior instruction. In order to determine theoretical models, The Reading Interview (Burke, l978) in which both the teacher and the students were asked to answer a series of questions about reading was selected. Finally, verbalizations made by the students while reading were chosen as a means of exploring whether the student's focus was on meaning, words, or the situation surrounding the reading or the research task. Measures of Text Structure The structure of the text will also help to explain the reader's performance (strategy selection) on the two types of texts. Initially, the two texts were chosen based on predictability. A close relation- ship between the text and the pictures as well as syntactic and semantic predictability were considered. The text story was deemed highly predictable on the syntactic and semantic measures, but less predictable on the relationship between text and picture measure. Because the wordless picture book has no text, only the semantic measure, based on the pictures, could be used. However, use of the book in earlier studies showed that children did produce well formed stories. For investigation of the structure of the texts on a word level, a readability formula (Spache, l974) was selected and a count of the number of previously untaught words occurring in the stories was developed. Further, since the correct identification of either unfamiliar (readability formula) or untaught words was considered important, a procedure was developed to measure the number of these words correctly identified. On a syntactic level, the structure of the texts as measured by 84 T-units (Hunt, 1966) was selected as the most appropriate measure. Although the relationships between words, clauses, T-units, and sentences can be measured in a variety of ways, since Hunt's (l965) study suggested that the number of words per T-unit gives the best measure of syntactic maturity, only that measure was used. On the semantic level, the story grammar developed by Stein and Glenn (l977b) was chosen in order to explore the use of setting state- ments, structure of episodes, completeness of episodes, and relation- ships between episodes. Finally, retelling's were also collected as a recall measure of comprehension. Summary In this chapter, the procedures used in the study were presented, including the design, sample, data collection, analysis, and coding pro- cedures. The expected outcomes were restated to include more specific information. Two previous pilot studies caused the design of this study to undergo considerable revision over the past two years. The elicitation method was developed in the first study and is designed to enable teachers to control the process of the interchange between student and teacher rather than the product. The texts were selected on the basis of predictability. The recurring nature of the episodes in both texts was used to judge both texts as semantically predictable. In the case of the basal text story, recurring patterns in the text contributed to predictability on the syntactic level. On the final measure of predictability, a close relationship between text and illustrations, the basal story was judged not as predictable as it might have been. The wordless picture book could not be judged on 85 the last two measures; however, it was chosen from a dozen or more such books used in previous studies as the one that elicited the most well-formed stories. The sample in this study was composed of eight first grade students from a racially integrated, suburban school in a mid-sized, midwestern city. Inclusion in the sample was dependent on student performance on the text story. In order to be designated a "beginning reader" students had to make enough miscues to provide data for the study, but not so many miscues as to render the text meaningless. Data was collected in February and again in March and required three meetings with each child. In the first meeting, The Reading Interview (Burke, l978) was administered and the children were asked to read the basal story. In'Umasecond meeting, the children produced a story to accompany the wordless picture book. In the final meeting, the students read their dictated story. Audio tapes were made of all meetings. Analyses to be performed were selected on the basis of the theoretical model given in chapter two. Since reading is viewed here as a cyclical process in which both the reader's perception of reading and the structure of the text have some influence on the strategies selected by the reader for any particular text, all three areas were deemed important to the investigation and coding procedures were selected or developed for each. Procedures developed by Y. Goodman and Burke for both the Reading Miscue Inventory (l97l) and the Evaluation form of the Reading Miscue Inventory (l976) were selected to measure strategy selection. 86 In addition, the decision was made to measure omissions and substitu- tions separately. For the reader's perception of reading, three areas were selected for investigation: prior instruction, theoretical models, and verbalizations made by the readers during their interaction with the texts. In order to explore the structure of the texts, two word level measures, a syntactic measure, and a semantic measure were either developed or selected from existing procedures. Finally, the expected outcomes from chapter one were restated. In this instance, additional information specific to the methods chosen for evaluation were added. CHAPTER FOUR: RESULTS In chapter two, a socio-psycholinguistic model of reading was presented. In this model (figure 4.l), reading is viewed as a cyclical process composed of three parts: reader's perception of reading, strategy selection, and text structure. The reader's perception of reading directs strategy selection; which, in turn, determines how that reader samples a text; which, in turn, modifies the reader's perception of reading. In each encounter with print, all of the reader's prior experiences with print exert an influence on that reader's performance. For example, if the reader has primarily had experience with phonics instruction and texts that embody that instruction, the reader will employ strategies that reflect the phonics/skills orientation. Further, each encounter with print serves to remodel the reader's perception for the next text he/she encounters. In other words, both perception of reading and text structure have a powerful influence on the way a reader samples print. This study is concerned with the influence of the text and seeks to determine whether or not a dictated story can cause a student to change strategy selection. Therefore, while the primary focus is on a comparison of strategy utilization on the dictated story as opposed to a story selected from the instructional materials used in the class- room, the perception of reading and the structure of the text also need exploration. 87 88 Print Setting Text Structure-- ased on author's heoretical model 'of reading instruction Reading-- . Strategy Based on reader's DlCECtS ,. Selection theoretical orientation Figure 4.1. A reader encounters a text. (From Rhodes, l979a as adapted from Harste, in press) 89 Measures of Strategy Selection In chapter one, a set of expected outcomes were stated for measures of strategy selection. These outcomes were stated in terms of the in- formation available for the administration of the Evaluation Form of the Reading Miscue Inventory (Y. Goodman & Burke, l976). They can be divided into three categories: sentence level measures; miscues per hundred words and correction strategies; and word level measures. Each category will be discussed separately. Sentence Level Measures For the sentence level measures, the following expected outcomes were stated: More sentences, as finally produced by the reader, will be judged both syntactically and semantically acceptable on the dic- tated story than on the text story. Sentences, as finally produced by the reader,'will less often result in a meaning change for the dic- tated story than for the text story. (See chapter three for examples.) The results of the sentence level measures are given in table 4.l. Based on overall performance on the text story, the students are listed in this table in order of proficiency. In other words, Phil was judged the most proficient reader and Jeremy was judged the least proficient reader. If the sentences produced on the dictated story were more often syntactically and semantically acceptable (see chapter two for examples of acceptability standards), the dictated story score should be higher than the text story score. Similarly, if the sentences less often resulted in a meaning change for the dictated story, that score 90 Table 4.l Reading Miscue Inventory Results Sentence Level Individual Scores Syntactic Semantic Meaning Acceptability Acceptability Change Phil text story dictated story Tim text story dictated story Jack text story dictated story David text story dictated story Karen text story dictated story Cindy text story dictated story Julian text story dictated story Jeremy text story dictated story 94.0% 94.0% 94.0% 100.0% 100.0% 100.0% 90.0% 90.0% 84.0% 96.0% 93.0% 93 0% 84.0% 73.0% 67.0% 71.0% 60.0% 55.0% 82.0% 71.0% 65. % 79.0% 73. % 73.0% 73.0% 67.0% 56.0% 93.0% 93.0% 81.5% 43.0% 41.0% 35.0% 65.0% 60.0% 55.0% 42. % 35.0% 23.0% 33.0% 33.0% 33.0% 25.0% 17.0% 6. % 71.0% 65.0% 65.0% 91 should also be higher than the score for the text story. The results indicate that five of the eight readers (63%) produced more syntacti- cally acceptable sentences; six of the eight readers (75%) produced more semantically acceptable sentences, and seven of the eight readers (88%) produced more sentences that closely approximated the author's intended meaning. Of these three scores, the semantically acceptable score is the most important (Y. Goodman & Burke, l976, p. l3). This score is designated the "comprehending score" because it reflects the degree to which the reader is able to construct meaning while reading. In other words, it indicates the degree to which the reader is able to use predicting, confirming, and comprehending strategies. The results cited in table 4.l indicate that 75% of the readers in this study were able to use these strategies more proficiently on the dictated story than on the text story. MPHW and Correction Measures For the miscues per hundred words and correction strategies, the following expected outcomes were stated: Readers will produce fewer miscues per hundred words (MPHW) on the dictated story than on the text story. Of the miscues produced, readers will more often correct those miscues on the dictated story than on the text story. (See chapter three for examples.) As indicated in table 4.2, six of the eight students produced fewer miscues per hundred words on the dictated story than on the text story and five of the eight students corrected more often on the dic- tated story than on the text story. In addition, Julian's correction score for the dictated story is only one percent lower than his score 92 Table 4.2 Reading Miscue Inventory Results Miscues per Hundred Words and Correction Score Individual Scores Miscues per Hundred Words Correction Phil text story 7.4 62.0% dictated story 2.2 100.0% Tim text story 9.0 31.0% dictated story 6.7 57.0% Jack text story 11.8 38.0% dictated story 16.2 19.0% David text story 13.0 16.0% dictated story 12.6 32.0% Karen text story 16.7 6.0% dictated story 21.1 18. % Cindy text story 24.2 18.0% dictated story 18.3 12.5% Julian text story 32.7 8.0% dictated story 27.9 7.0% Jeremy text story 41.6 5.0% dictated story 33.0 21.0% for the text story, indicating that the dictated story certainly was no more difficult for Julian than the text story. In addition to the semantic acceptability score, the correction score has also proven to be a good indicator of reader performance (Y. Goodman & Burke, 1976). Again, this score indicates the degree to which the reader is able to use predicting, confirming, and compre- hending strategies to construct meaning. Generally, a reader who is not reading for meaning attends more to letter/sound correspondence or prior knowledge of the word than meaning, and tends not to correct as often. (See chapter two for a review of miscue analysis research.) The results in table 4.2 indicated that 63% of the readers in this study corrected miscues more often on the dictated story than on the text story. Word Level Measures At the word level, the following expected outcomes were stated in chapter one: Miscues produced on the dictated story will be less similar on a phoneme-grapheme measure than those produced on the text story. Miscues produced on the dictated story will more often be of the same grammatical function than those produced on the text story. (See chapter three for examples of grammatical function standards.) The results (table 4.3) show that seven of the eight students in the study produced miscues that were less similar on both graphic and sound similarity for the dictated story than for the text story. Readers who use strategies based on a phonics model of reading usually predict and confirm at the letter level. That is, they produce mis- cues that are graphOphonically similar to the text words but are often 94 Table 4.3 Reading Miscue Inventory Results Word Level Substitutions in Context Individual Scores Graphic Sound Grammatical Similarity Similarity Function Phil text story 82.5% 82.0% 18.0% dictated story 75.0% 75.0% 0.0% Tim . text story 88.0% 88.0% 35.0% dictated story 83.0% 83.0% 83.0% Jack text story 92.0% 92.0% 36.0% dictated story 89.0% 86.0% 67. % David text story 94.0% 87. % 48.0% dictated story 72.0% 68.5% 53.0% Karen text story 72.0% 72.0% 53.0% dictated story 49.0% 45.0% 48.0% Cindy text story 88.0% 84.0% 24.0% dictated story 65.0% 47.0% 53.0% Julian text story 75.0% 78.0% 44.0% dictated story 45. % 45.0% 35.0% Jeremy text story 80. % 78.0% 51.0% dictated story 91.0% 87.0% 57.0% 95 nonsense words or words that do not make sense in the total context. Readers who use strategies based on a skills model of reading usually predict and confirm at the word level and produce the expected response if they know the word, and an omission if they do not know the word. ( See chapter two.) The fact that 88% of the students in this study produced miscues that were graphophonically less similar on the dictated story than on the text story points to a powerful influence of the text. Generally, "proficient readers tend to have lower high similarity scores than in- effective readers" (Y. Goodman & Burke, 1976, p. 21). Thus, the ma- jority of readers in this study used more proficient strategies on the dictated story than on the text story. Results reported in table 4.3 also indicate that five of the eight readers in the study produced miscues that were more often of the same grammatical function on the dictated story than on the text story. Further, although none of Phil's miscues were of the same grammatical function as the expected response, all of his miscues were corrected, indicating that he recognized the lack of grammatical agreement and was able to use that information to correct. Of the remaining five readers, all, or 63%, were able to use more proficient strategies on the dic- tated story than the text story as indicated by more miscues of the same grammatical function. While graphophonic similarity and grammatical function do not give as clear an indication of comprehending as semantic similarity and successful correction, they do give an indication of the reader's per— ception of reading. Almost all of the graphic and sound similarity scores are very high, indicating that these readers pay particular 96 attention to letter/sound relationships at the expense of other strate- gies that might help to achieve meaning. This indicates that these readers perceive reading as a decoding-to-sound task. In order to assess over-all performance, an average score on eaCh of the measures discussed so far was computed for both stories. The results are presented in table 4.4. Syntactic acceptability, semantic acceptability, meaning change, grammatical function, and successful correction scores are all higher for the dictated story than for the text story as predicted. The scores for graphic similarity, sound simi- larity, and miscues per hundred words are all lower for the dictated story than for the text story as predicted. In other words, all scores reported in table 4.4 confirm the predicted results. The readers in this study produced more syntactically and semantically acceptable sen- tences, more sentences that approximated the author's intended meaning, fewer miscues per hundred words, more successful correction attempts, miscues that were less similar on a graphophonic measure, and more sentences that were similar on a grammatical function measure. Thus the readers in this study were able to make more proficient use of pre- dicting, confirming, and comprehending strategies on the dictated story than on the text story. Omission and Substitution Measures In this study, the three least proficient readers overwhelmingly chose to omit any word they perceived as unknown. Since mature readers produce more substitutions than omissions (Biemiller, 1970; K. Goodman, 1976; Y. Goodman, 1976; Levine, 1976; Schlieper, 1977, and Weber, 1970a), the decision was made to pursue this phenomenon further. The 97 Table 4.4 Reading Miscue Inventory Results Mean Percentage Scores Text Story Dictated Story Syntactic Acceptability 66.6% 76.0% Semantic Acceptability 61.0% 72.1% Meaning Change 53.7%(P+N)a 69.4%(P+N) Miscues per Hundred Words I 19.5% 17.3% Correction 23.0% 33.3% Graphic Similarity 83.9%(Y+P)b 71.0%(Y+P) Sound Similarity 82.6%(Y+P) 67.1%(Y+P) Grammatical Function 38.6% 49.5% aPartial meaning change (P) and no meaning change (N) were combined. bVery similar (Y) and partially similar (P) were combined. 98 results, given in table 4.5, show that omissions on the dictated story did go up for five of the readers, indicating that on this measure these readers found their dictated story somewhat more difficult. (This result is consistent with measures of the structure of the text to be discussed later in this chapter.) Nevertheless, while omissions did increase for most of the proficient readers, this increase did little to change their ability to use effective strategies. A compari- son of figures 4.2 and 4.3 confirm this statement. In figure 4.2, the text story, the five more proficient readers were clearly more willing to take risks and produce some kind of re- sponse even for the words they were unsure of. In all five cases, the percentage of substitutions was far greater than the percentage of omissions. Further, two readers, Jack and Phil, produced a response for every word and never omitted. On the other hand, the three less proficient readers were not willing to take risks and produced omissions much more often than they produced responses. Figure 4.3 gives the same information for the dictated stories. Two trends are apparent here. Omissions went up for most of the readers and, at the same time, the overall trend was towards moderation. For the three most proficient readers, Phil, Tim, and Jack, omissions in- creased on the dictated story. Omissions also increased for one of the two middle range readers and for one of the three less proficient readers. For the proficient readers, the increase in omissions indicates that they were able to take even more risks on the dictated story than they took on the text story. The model of reading held by these students (see Reader's Perception of Reading section) indicates that they value a no omissions strategy. Yet, truly proficient readers should be able Table 4.5 Percentage of Omissions and Substitutions Individual Scores Text Story Dictated Story Omissions Substitutions Omissions Substitutions Phil 0.0% 80.9% 20.0% 80.0% Tim 3.8% 65.3% 14.3% 85.7% Jack 0.0% 73.5% 4.6% 83.7% David 5.4% 83.8% 11.9% 86.4% Karen 12.8% 76.6% 5. % 81.6% Cindy 59.7% 37.3% 43.7% 53.1% Julian 60.0% 40.0% 68.7% 29.6% Jeremy 61.2% 38.8% 41.0% 59.0% aPercentage of total number of miscues. 100 .zLOHm uxmu mg» mcwuwmg m_w;3 mums m:o_w:pwpmn:m new mcommmeo mo cemwcwasou < .mwgwmczmwd saweee eee.=e see_o emcee eeseo some ewe _e;e % w W m m w u u w 9 u 0 w 8 new 0 mnzm mso mnzm mac maam mso mazm mso mazm mac maam mac mnzm mac mnsm mso l J 1 l I ll 1 l l I 1 l 1 L l l l l Acoum uxmh o— mp om mm om mm. oq me om mm om mm on mn ow mm om mm 101 .zcoum vmumuu_u ecu mcwummc mpwsz meme m:o_p:u_umn:m use mcowmmmso eo cemwgmqeoo < .mwv mesa?“ xsmcmc :mw_:w xvcwu :mcmx uw>mo xomw see cha 9 .V Z 9 9 .V 8 00 ..L 00 00 IL 00 Z 6 1L 6 8 E 00 IL C4 9 ..I. Co .V C.. .V O 0 0 O 9 /. IL IL 9 Cu .V 6 I. 9 I. .00 O 0 % o/6 o/6 6/6 % o/6 o/o % % o/o % ~76 % o/o % % mnzm mso mnam mso maam mso mnzm mac mnsm m=5 mnsm mso mnzm mac mnzm mac steam eeeeeeeo llllllllllllllllllll op mp om mm om .mm ow m¢ om mm om mo 0m mm ow mm om mm cop 102 to omit, at times, and trust to subsequent text to provide further information. Thus, faced with a more difficult text, the increase in omissions is a positive indication that these readers were able to con- tinue using the most efficient strategy available to them. For two of the least proficient readers, who chose to omit more often than substitute on the text story, the reverse in this trend on the dictated story also indicates a positive trend. The middle range readers simply changed positions. David produced more omissions on the dictated story and Karen produced more omissions on the text story. Yet, both were in the moderate range in both cases. The trend toward moderation, then, is evident for all readers except Julian. On both texts, Julian produced more omissions than substitutions. The remaining seven readers all moved toward a more efficient utilization of strategies. The less proficient readers were able to take more risks on the dictated stories by increasing substitutions; the more proficient readers were able to take more risks by increasing omissions; and the middle range readers maintained their moderate approach. Summary of Measures of Strategy Selection The expected outcomes for measures of strategy selection were divided into three categories: sentence level measures; miscues per hundred words and correction strategies; and word level measures. On a sentence level, more students produced syntactically and semantically acceptable sentences that closely approximated the author's intended meaning on the dictated story than on the text story. In addition, more students made fewer miscues per hundred words and more often cor- 103 rected those miscues on the dictated story than on the text story. And, finally, more students produced miscues with less graphic and sound similarity and more grammatical function similarity on the dic- tated story than on the text story. When individual scores were aver- aged in order to compare group performance on the two stories, the re— sults showed that, on all measures, the students (as a group) were able to make more proficient use of predicting, confirming, and comprehend- ing strategies on the dictated story than on the text story. In addition to the information available from the Evaluation Form of the RMI (Y. Goodman & Burke, l976), omissions and substitutions were compared. On the dictated story, the five more proficient readers were able to maintain their strategy of producing more substitutions than omissions. Furthermore, two of the three less proficient readers were able to reduce their over reliance on omissions and also produced more substitutions than omissions. Overall, for all but one reader, the trend was towards moderation. In conclusion, the dictated story did appear to exert an influence on the strategy selection of the readers. On every measure investi- gated, students exhibited more effective reading strategies while read- ing the dictated story than while reading the text story. In order to explain reader performance, the discussion will now turn to the reader's perception of reading and the structure of the text. Measures of Reader's Perception of Reading In chapter one, the expected outcome for this section was stated as follows: The reader's perception of reading will help to explain performance (strategy selection) on the two types of texts. Prior in- 104 struction, theoretical model (both students and teacher), and verbali- zations made during the reading process were selected to measure the reader's perception of reading. Prior Instruction All the students in this study were from the same classroom and had received approximately the same amount of instruction from the same teacher. Further, they had all been exposed to the same texts; and, at the time of testing, had acquired approximately the same level of proficiency in reading. Because of a strike, the school did not open until October. Early in that month, the teacher began instruction in phonics. After approximately two weeks of instruction with teacher-produced, phonics materials, the students began instruction in the Sullivan Readiness Reader (1969). All students had to finish book A of the Sullivan se- ries before they could begin the first pre-primer of the Houghton Mifflin series (1971). At the time of testing, all students had had approximately seventy days of instruction in the Sullivan series and approximately fifty days of instruction in the Houghton Mifflin series. They had finished levels A, B, C, and D in the sullivan series and they had read nine stories from the basal reader (Houghton Mifflin). Instruction in the Sullivan series focuses on letter/sound rela- tionships. Students are given lists of minimal pair words and asked to discriminate between them. The following list provides an example: pin, pan, man; nip, nap; pit, pat; tip, tap; and so on. Each pair, or trio, consists of words that are different from the preceding word by only one letter. This kind of instruction is illustrative of the 105 phonics model discussed in chapter two (figure 2.2). The type of instructional procedures used in conjunction with the Houghton Mifflin series can be characterized as based on a skills model (figure 2.2). In this case, the focus is on the word. We do not want to teach pupils to ”sound out" or "puzzle out” the pronunciations of the words in their reading matter. They already know perfectly well how to pronounce those words if the vocabulary in their early reading matter consists entirely, as it should, of words they have already heard and even spoken time and time again. (Iiggyg, teacher edition, 1974, p. 18) Not only is the focus on the word, but it is on the "known” words. Therefore, each succeeding story includes only those words that have been used in previous stories plus the few "new" words introduced for the assigned story. As the authors state, "It becomes imperative that pupils recognize instantly most if not all of the words in a sentence Except the one to be decoded" (Tigers, teacher edition, 1974, p. 21). Given the materials chosen by the teacher in this study, one could predict that her instruction would focus on letter/sound relationships and knowledge of vocabulary. In this classroom, the morning reading period (one to one and a half hours) was devoted to phonics instruction and children were drilled on letter/sound relationships. In the after- noon, the basal readers were used for approximately the same period of time. As new vocabulary was introduced for each story, the children were given cards with the words printed on them. The children were expected to keep the cards in their desks and use them to practice the words. During reading instruction, if the child experienced difficulty with a word, the teacher either tried to help the child sound it out or told him/her the word. 106 Students in this study also received a limited amount of language experience instruction. Once a week, they had an opportunity to do creative writing for about 30 minutes in the afternoon. The first few weeks were devoted to group dictation of a story written by the teacher on the blackboard. The children then copied from the black- board. Later they wrote their own stories about pictures, current topics, the weather, holidays, and so on. If they didn't know how to spell a word, they were expected to ask the teacher. In addition, consistent with her model, the teacher did manipulate syntax and vocabu— 1ary as she felt appropriate so that these experience did not directly lead to the dictation experiencein this study. This type of instruc- tion had occurred approximately seventeen times, or 8% hours, before the children were tested. Because of the texts used and the instruction offered, the children in this classroom primarily developed letter and word level strategies. In addition, they generally expected the teacher to confirm correctness. Theoretical Model In order to determine the theoretical model of both the teacher and the children, The Reading Interview (Burke, 1978) was administered. In this interview, a series of open-ended questions about reading are asked. (See Appendix D for a sample interview form.) The teacher in this study is in her late twenties and has a B.A. in education plus twenty hours of graduate work. In her interview, she stated that she learned to read by a sight word approach and didn't learn anything about phonics until she took a reading methods course in college. She said that she teaches reading through both sight words 107 and phonics, although she uses context and phonics herself. She thought that her sister was a better reader than she was because her sister reads “phonetically." The teacher's statements made in response to the interview questions help to explain the kind of instruction she offers in her classroom. Clearly, this teacher believes that a combination of phonics and drill on sight words will enable students to become profic- ient readers. Her instruction centers on phonics and vocabulary. When the students were given the same interview, their responses reflected both the teacher's theoretical model and the instruction they received in this particular classroom. When they came to something they didn't know, all eight readers stated that they would "sound it out." Three also suggested that they might ask someone else, probably the teacher. One said he would stop and think about it. One suggested skipping it. One would read more and one would think up a more simple word. The readers selected a good reader, for the most part, because that person was smarter (in a higher group) or practiced a lot (read a lot of stories). Two also mentioned that that person listened. Four of the eight readers thought that their good reader never came to something he/she didn't know. Three thought "maybe" and only one definitely stated "yes." All eight thought that their good reader would sound it out if he/she g1g_come to something he/she didn't know. In helping other students, all eight readers suggested sounding out as a means of giving help. In addition, two suggested that they would tell the person the word or find someone who could tell them the word. One suggested skipping it and one suggested thinking about it or looking at it and spelling it. 108 When asked what their teacher would do to help readers in trouble, five students said she would help them sound it out or sound it out for them. The other three thought she would tell them the word. When asked to describe how they learned to read, six of the eight students named a teacher as the person who helped them learn to read. The other two named a parent. In addition, two of the children who named a teacher claimed that they also had received help from a parent. Interestingly, four of the five more proficient readers indicated that they had received help at home while none of the three less proficient reader indicated that they received help at home. When asked what those people did to help them learn to read, five of the eight readers said that their instructor helped them to sound out or spell out the words. Two said that the person told them the word or told them if the word was wrong and one simply stated -that he practiced a lot with his dad. When asked what they would like to do better as a reader, all but one responded on a word level. They either wanted to learn more words, or sound out words better, or read all the words, or not mix up the words. In addition, one also suggested that she'd like to be better at skipping words. One child refused to answer the question, stating several times that he would rather do something else. The results of the interviews (both teacher and children) are con- sistent with the kind of texts used and the kind of instruction offered and received in the classroom. On a continuum of theoretical models from phonics through skills to whole language, both teacher and students in this study clearly fall somewhere between phonics and skills. 109 Verbalizations While reading, the students often produced "asides" that were not part of the text. These were termed verbalizations and coded in a man- ner similar to Rhodes (1979b) (See appendix G for results). If the child questioned the sense of a sentence containing a miscue, commented on a story event, or brought his/her own knowledge or feelings to bear on the story, the verbalization was classified as having a focus on meaning /M/. If the child commented on the difficulty of a word(s) or the text, or made a statement related to a word recognition strategy, the verbalization was classified as having a word focus /W/. Situation verbalizations /S/ included comments on the directions given, the physi- cal appearance of the text, the reader's place in the text, or the length of the story. Rhodes found large differences in the verbalizations made as the children read the more and less predictable stories (Rhodes, 1979b, p. 99). In general, far more verbalizations relating to meaning were made while reading the predictable stories (64%) than while reading the less predictable stories (7%). From these results, it was predicted that the children in this study would make more statements relating to meaning while reading the dictated (more predictable) story than while reading the text story. While a higher percentage of verbalizations relating to meaning were made on the dictated story than on the text story, the differences cited here (table 4.6) are not nearly as large as those cited by Rhodes. In this study, by far the largest percentage of verbalizations on book texts exhibit a focus on letters and words. Nevertheless, this finding clearly confirms the findings presented for 110 Table 4.6 Focus of Verbalizations Text Story Dictated Story Number Percentage Number Percentage Meaning [M] 5 12.2% 5 20.8% Word [W] 33 80.5% 15 62.5% Situation [S] 3 - 7.3% 4 16.7% prior instruction and theoretical models. As evidenced by their focus while reading the two texts, these students reflected both the theoreti- cal model they espoused in the interviews and the instruction they re- ceived in the classroom. Summary of Measures of Reader's Perception The expected outcome for the reader's perception of reading was measured by prior instruction, theoretical model, and verbalizations. Due to the kinds of texts selected and the instruction received, the students in this study exhibited primarily letter and word level strate- gies. The prevelance of these strategies was further documented by information from the interviews: The teacher stated that she taught reading through phonics and sight words; all eight children suggeSted sounding out as their first strategy, both for themselves and for help- ing others; and, finally, by far the largest percentage of verbaliza- tions for both stories focused on the word level. All three measures, then, indicate that the children in this study perceive reading as a decoding-to-sound process. 111 Measures of Text Structure The expected outcome for this section was stated as follows: The structure of the text will also help to explain the reader's performance (strategy selection) on the two types of texts. For investigation of the structure of the texts on a word level, a readability formula (Spache, 1974) was selected and a count of the number of previously un- taught words occurring in the stories was developed. 0n the syntactic level, the structure of the texts as measured by T-units (Hunt, 1966) was selected. On the semantic level, the story grammar developed by Stein and Glenn (1977b) was chosen in order to explore the use of set- ting statements, structure of episodes, completeness of episodes, and relationships between episodes. Finally, retellings were also collected as a recall measure of comprehension. Word Level Measures In order to determine the relative difficulty of the text story and the dictated stories, Spache's readability formula (1974) was used. In figure 4.4, the results are given in a bar graph. The text story "Red is Nice" was measured at a 1.6 (first grade, six months) level. The dictated stories ranged in difficulty from 1.7 to 3.1. Therefore, as compared with the stories dictated by the children, on this measure, "Red is Nice" should have been the easiest story to read. However, the results of the RMI do not support this statement. Rather than being the easiest story, "Red is Nice“ was more difficult than the text story. On the group measure, all of the subjects in this study exhib- ited more efficient strategies on the dictated story than on the text 112 Had—ad—JddadNNNNNNNNNNwW o—amwhmmwooxoo—«wammwooxOOH 1117 1111111111111111 Jeremy Phil Red is nice Jack 11m Cindy Karen Julian David Figure 454.A comparison of readability levels of the text story and the dictated stories. 113 story. Clearly, the text story did not offer the support needed to enable the students to use proficient strategies. Thus, in this study, the readability formula did not predict which story would be the most difficult for the reader. In order to eXplore a word level measure more particular to the students in this study, a measure of untaught words was developed. (See chapter three for an explanation.) Because the authors of the basal text used in this classroom suggested that children need a con- trolled vocabulary in order to be able to "decode” new words, this measure was deemed necessary. In other words, the instruction given the children in this study was based on the assumption that if a story contained too many words not previously taught or not introduced for the particular story to be read, the students would not be able to decode those words. From this assumption one could predict that a story with an inordinate amount of previously untaught words would be incomprehensible to the readers. However, this prediction was not true in the present study. As seen in figure 4.5, the text story con- tained only twenty-three words not previously taught while the dic- tated stories contained thirty—nine to eighty-three words not pre- viously taught. Yet, as a group, the students in this study exhibited more proficient strategies on the dictated stories, despite the in- ordinate number of untaught words, than they did on the text story. Thus, the dictated story offered the support the readers needed in order to perform proficiently. And, again, the focus on word level measures, even a measure designed particularly for the students in the study, did not predict which story would be the most difficult. 114 100 95 9O 85 8O 75 70 65 60 55 50 45 40 35 30 25 20 15 10 23 39 39 45 46 53 58 71 83 Red is nice Jeremy C1ndy Phil Tim Karen Jack Julian DaV1d Figure 4,5, A.comparison of the number of untaught words in the text story and the dictated stories. 115 In addition to a readability measure and an untaught words meas- ure, two other areas were explored. Given the instruction received by the students, their theoretical model, and their penchant for focusing on words while reading, it was predicted that both the unfamiliar words (Spache readability formula) and the untaught words would not be easily identified. Therefore, the number of both unfamiliar words and untaught words correctly identified was computed. Further,.the in- crease in number of such words correctly identified on the dictated story was also computed. For example, in table 4.7, the number of un- familiar words occurring in the dictated stories always exceeds the number of unfamiliar words included in the text story. Thus, every reader took the risk of including words that, according to the Spache list (1974), should not have been familiar. In addition, all readers, except one, were able to correctly identify more unfamiliar words on the dictated story than on the text story. Only Jack correctly iden- tified the same number of unfamiliar words on both stories. There- fore, all the readers were able to process at least as many unfamiliar words on the dictated story as on the text story and seven readers were able to process even more-~up to four times as many. On the untaught words measure (table 4.8), again, all readers took the risk of including words that they had not been taught through prior instruction received in the classroom. In addition, all readers were able to correctly identify more untaught words on the dictated stories than on the text story. In fact, with the exception of Phil, all readers correctly identified at least twice as many untaught words on the dictated story as on the text story. Phil missed 116 Table 4.7 Readabilitya Unfamiliar Words Correctly Identified Individual Scores Unfamiliar Number Words Correct Increase Phil text story 4 4 dictated story 5 5 +1 Tim text story 4 3 dictated story 7 6 +3 Jack text story 4 3 dictated story 6 3 +0 David text story 4 3 dictated story 16 9 +6 Karen text story 4 4 . dictated story 8 6 +2 Cindy text story 4 2 dictated story 4 3 +1 Julian text story 4 3 dictated story 7 4 +1 Jeremy text story 4 l dictated story 7 4 +3 aSpache Readability Formula 117 Table 4.8 Untaught Words Correctly Identified Individual Scores Untaught Number Words Correct Increase Phil text story 23 23 dictated story 45 45 +22 Tim text story 23 22 dictated story 46 46 +24 Jack text story 23 20 dictated story 58 43 +23 David text story 23 20 - dictated story 83 67 +47 Karen text story 23 17 dictated story 53 43 +26 Cindy text story 23 12 dictated story 39 27 +15 Julian text story 23 12 dictated story 71 32 +20 Jeremy text story 23 7 dictated story 39 23 +16 118 doubling his score by only one word and, at the other extreme, David correctly identified over three times as many untaught words on the dictated story as on the text story. In this classroom, the philosophy of the text authors and the teacher indicated that the number of unfamiliar or untaught words en- countered in a given story should be kept to a minimum in order to facilitate decoding of "new“ words. Given the opportunity to produce their own stories, the students were able to overcome this restriction and included many unfamiliar and untaught words in their stories. Further, they were able to process a surprising number of supposedly ”unknown" words. Syntactic Level Measure On the syntactic level, the structure of the texts as measured by T-units (Hunt, 1966) was selected as the most appropriate measure. (See chapter three for a description of the procedure.) Although re- lationships between words, clauses, T-units, and sentences can be meas- ured in a variety of ways, Hunt's (1965) study suggested that the num- ber of words per T-unit gives the best measure of syntactic maturity; therefore, only that measure was used. Figure 4.6 shows a comparison of the results. The text story, "Red is Nice" had an average of 5.36 words per T-unit while the dictated stories ranged from 3.93 to 8.7 words per T-unit. On this measure, the text story was not the least difficult story produced. However, only one child, Jeremy, produced a story that was considered less difficult syntactically. Jeremy was considered the least proficient reader in the study. On the text story, he produced fewer syntactically and semantically acceptable 119 llllllLllllll 1 7.0 "—" LDCDLDOLDOLOOLDOLDOLO 06mm¢¢mmNszF LO 5.36 6.50 6.70 6.92 6.94 27 3.93 4pm puma ”9.11110 4390 uauex 11Hd 1111] 99.111 $1 pea Kweeep -unit in the text story A comparison of words per T and the dictated stories. Figure 4.6. 120 sentences and more often altered the intended meaning than any other student. He produced the most miscues per hundred words and corrected the least often. Further, on both word level measures given earlier, Jeremy's dictated story was rated only slightly more difficult than the text story. Apparently, Jeremy is not particularly adept at deal- ing with print in either a receptive or a productive mode. Neverthe- less, the majority of dictated stories, as measured on a syntactic level, were considered more difficult than the text story. Again, the syntactic measure did not predict the most difficult stories, as all of the students, despite the increased syntactic complexity of the dictated stories, performed more proficiently on their dictated stories than on the text story. Semantic Level Measure On the semantic level, the story grammar developed by Stein and Glenn (1977b) was chosen in order to explore the use of setting state- ments, structure of episodes, completeness of episodes, and relation- ships between episodes. (See chapter three for an explanation of the procedure and appendix H for the results.) Setting statements. According to Stein and Glenn (l977b), a story is composed of a setting and an episode system. The setting statement generally serves to introduce the protagonist and give in- formation about the story context. In their investigations of recall of stories by children, Stein and Glenn (l977b) found that the setting statement is the most frequently recalled information given in a well- formed story. 121 All the students in this study started their story with a setting statement (table 4.9). Three of the statements included the classic folk tale beginning "Once upon a time." The other five statements con- sisted of a description of the characters in the story and their present activity. The text story, on the other hand, contained no set- ting statement, but began, rather abruptly, with the question, "Do you want to go to the library, Ken?‘I In fact, all of the stories in the pre-primer, Dinosaurs (Houghton Mifflin, 1971) began in a rather abrupt fashion. Of the seven stories included, only one begins with any kind of setting statement. In addition to setting statements, four of the children in the study elected to include closure statements as well (table 4.10). While Stein and Glenn do not discuss closure statements, the kinds of endings included by the children in this study are also common in folk- tales. Again, no closure statement was included in the text story. Structure of Episodes. In addition to setting statements, a well-formed story also contains an episode system comprised of an ini- tiating event, an internal response, an internal plan, an attempt, a direct consequence, and a reaction. (For definitions and examples of each, see chapter three.) Further, some categories occur more fre- quently than others. In their research, Stein and Glenn (l977b) found that a well-formed episode must contain at least the following: an initiating event (or an internal plan if it includes the goal state- ment); an attempt; and a direct consequence. Any episode without at least these three categories is considered incomplete. In the present study, the text story and all the dictated stories were coded using the Stein and Glenn story grammar (appendix H) and 122 Table 4.9 Use of Setting Statements R is N ------------------------------------------------------------- B 8 F Once upon a time there was three bears and they were eating breakfast. B 8 F Once upon a time, they were eating dinner. B 8 F The mother and the father and the little girl were eating breakfast. B 8 F They're eating breakfast. B 8 F Once upon a time, there lived three bears. B 8 F The three bears were sitting at the table. B 8 F The three bears were eating breakfast. B 8 F The three bears were eating dinner. Table 4.10 Use of Closure Statements U) 2 The end. That's the end of the fly. WWWWWWWWJU Q°Q°Q°Q°Q°Q°Q°Q°do 'n'n'l'l'fl'fl'l'l'n'f'l the percentage of complete episodes was computed (table 4.11). The most striking aspect of these results is the lack of complete episodes in the text story. For every episode in the "Red is Nice" story, the attempt category was missing, resulting in no complete episodes. Every episode had an initiating event and all but one had an internal plan, but none of the plans were ever carried out. 123 Table 4.11 Percentage of Complete Episodes Number Number Percentage Percentage Complete Incomplete Total Complete Incomplete aR is N 0 6 6 % 100% b8 8 F (1) 5 0 5 100% 0% B 8 F (2) 4 1 . 5 80% 20% C8 8 F (3) 5 2 7 71% 29% a8 8 F (4) 4 2 6 67% 33% B 8F (5 ) 3 2 5 60% 40% B 8 F (6) 3 2 5 60% 40% B 8 F (7) 2 4 6 33% 67% B 8 F (8) 2 4 6 33% 67% aNo episode included an attempt category. Goals were always stated as part of the internal plan. bOne episode did not include an attempt. The goal was stated as part of the internal plan. cOne episode did include an internal plan rather than an initia- ting event but no goal was included in the plan. 124 In contrast, the stories produced by the children ranged from 100% complete episodes to 33% complete episodes. In all but two of the child- ren's stories, over half of the episodes were complete. One other striking aspect of this particular data was the correspondence between reading performance and production of stories. The data in table 4.11 is listed in order of reader proficiency as determined by performance on the text story. That is, the first Bear and Fly story was produced . by the most proficient reader in the study and the last Bear and Fly story was produced by the least proficient reader in the study. Interest- ingly, the percent of complete-episodes-data follows exactly the same pattern. That is, the most proficient reader in the study produced the most complete story, the second most proficient reader produced the second most complete story, and so on. Relationship between episodes. In addition to the delineation of categories and the determination of those categories necessary for a complete episode, Stein and Glenn (1977b) also specify the causal links between episodes of a story. Three relations between episodes are possible: AND, THEN, and CAUSE. THEN and CAUSE relations are the most common. A THEN relation denotes a temporal juxtaposition; a CAUSE relation implies a direct causal connection between episodes; and an AND relationship implies that the two episodes occurred simultaneously. Stein and Glenn have suggested that the ability to retain sequences might be dependent on the relation between episodes, with CAUSE being the strongest relation, THEN second, and AND the weakest (Stein 8 Glenn, l977b, p. 116). In the text story, over half the episodes were CAUSE related while only two dictated stories have even one CAUSE relationship (Table 4.12). 125 The most probable explanation for the lack of CAUSE relationships in the dictated stories is the age of the children. For example, the children in this study often did not handle causation very successfully and frequently combined events in a story with "and then," denoting a temporal relationship. (See Appendix I for examples of the children's dictated stories.) Table 4.12 Relationship Between Episodes THEN CAUSE AND (I) Z Q°Q°Q°Q°Q°Q°Q°Q°-h 'TT'T'I'T‘I'TI‘T‘IW'I'TIW‘I mmbw—bm-b-bm COCA-#0000) OOOOOOOOO Retellings In addition to word level measures, a syntactic measure, and the story grammar investigation, retellings were also collected. At the conclusion of the reading of both stories, each child was asked to tell what he/she remembered about the story. The retellings were compared with a previously developed retelling guide (Appendix F) to determine the reader's recall of characters, development of characters, events, plot, and theme. The scoring of retellings used in this study is an adaption of the procedure developed by Y. Goodman and Burke (1972) and includes a possible 100 points. 126 Table 4.13 Retelling Scores Red Is Nice The Bear and the Fly Phil 59% 76% Tim 62% 67% Jack 50% 63% David 52% 61% Karen 58% 60% Cindy 66% 60% Julian ' 55% 69% Jeremy 37% 67% In general, the retelling scores for the dictated stories were higher than those for the text story (Table 4.13). Only Cindy failed to produce a better retelling for her dictated story than for the text story. In addition, it was hoped that the retelling scores for the dictated stories would reflect the semantic structure of the texts. In other words, a more well formed story should have produced a better re- telling. If that were the case, the retelling scores for The Bear and the Fly should appear in descending order in the table. While each score is slightly lower than the one preceding it for the first six readers, the pattern does not hold for the remaining two readers. Julian and Jeremy produced the two least well-formed stories in terms of complete episodes; yet, their retelling scores rank second and third respectively. One possible explanation may be that the experience of producing and reading their own stories gave more support to these less proficient readers than it did to the other readers in the study. Summary of Measures of Text Structure In this section, the structure of the text was investigated. Two 127 measures (a readability formula and a list of untaught words) were used to measure the difficulty of the texts on a word level. On both measures the dictated stories were found to be more difficult than the text story. Words per T-unit (Hunt, 1966) was used to investigate the syn- tactic complexity of the text and revealed that all but one of the dictated stories was more syntactically complex than the text story. On the semantic level, the story grammar developed by Stein and Glenn (l977b) was used to explore the structure of the text. The dictated stories were found to be more well-formed in that they included setting statements and complete episodes. The text story had no setting statement and no complete episodes. The relationship between episodes was also investigated. Because the CAUSE relationship is the strongest, one would expect better recall of stories with more CAUSE relationships. However, retellings were also collected and, although the dictated story contained proportionately more CAUSE relationships, only one student produced a better retelling for the text story than the dictated story. Summary of Results In this chapter, the results of the three types of measures (strategy selection, perception of reading, and text structure) were given. All of the strategy selection measures indicated that the students, as a group, clearly performed better on the dictated story than on the text story. In the investigation of the reader's perception of reading, the measures indicated that all of the children in this study held a similar theoretical model. On the continuum from phonics through skills to whole language, the students in this study fell somewhere between a phonics model and a skills model. Consequently, when encount- ering difficulty with a text, their strategies consist of either sound- 128 ing out the word or asking someone to tell them the word. They usually do not use knowledge of syntax or semantics to predict or confirm their word level guesses. The measures of text structure indicated that on both a word level and a syntactic level, the stories produced by the students were more complex than the text story. However, on a semantic level, the dictated stories were generally more well-formed than the text story. Finally, the retelling results suggested that the dictated stories were recalled better than the text story. A major criticism of the value of reader-authored stories might be that the readers are able to perform more proficiently on their dictated stories because the stories are their own. In order to read effectively, readers must be able to control all three cue systems-~the graphophonic, the syntactic, and the semantic. When the student functions as both author and reader, control over the semantic cue system is greatly in- creased. The reader has greater opportunity to "know" the meaning of the story because he/she wrote the story. However, as seen in the discussion of the structure of the text, the students produced dictated stories that were more difficult on both a word level and a syntactic level than the text story. Yet, they were able to read these stories proficiently. Obviously, greater control of the semantic system enabled these readers to cope with more difficult vocabulary and syntax. Thus, the question is not whether or not the dictated stories were ”easier" to read. In evaluating reading difficulty, the three cue systems must be kept in mind. The fact that the reader was also the author for the dictated story offered an advantage on the semantic level. Further, the story grammar results confirmed the semantic accessibility of these dictated stories. On the other hand, on a word level and a 129 syntactic level, the dictated stories were judged more difficult. Because the readers in this study appeared to focus mostly on the word level, the increase in unfamiliar and untaught words alone should have rendered proficient performance on the dictated stories nearly impossible. Yet, the readers were able to overcome this difficulty. Further, the increase in syntactic complexity should also have proven to be a problem, but it wasn't. The real question here is what kind of support is the most valuable to the reader? From the results of this study, it appears that semantic support may offer the greatest help to the reader. For the students in this study, greater control of the semantic level enabled them to more proficiently handle difficult vocabulary and syntax. CHAPTER FIVE: CONCLUSIONS Results The purpose of this study was to determine whether or not stories produced by the readers themselves could alter reading performance. Using information from the Evaluation Form of the Reading Miscue Inven- tory to assess proficient reading strategies, it is clear that the eight first grade readers in this study did alter their reading performance on the dictated story; and, further, were able to use more proficient strategies while reading the dictated story than while reading the text story. In addition to strategy selection, the reader's perception of reading and the structure of the text were also investigated. While measures of reader perception did not explain group performance on the dictated story, they did help to explain some individual performances. One of the measures of text structure, the semantic measure, did seem to explain performance on the dictated story. Measures of Strategy Selection Strategy selection as measured by the Evaluation Form of the Reading Miscue Inventory (Y. Goodman 8 Burke, 1976) was used to measure reading performance. Information available from the Evaluation Form includes: syntactic acceptability; semantic acceptability; loss of meaning; miscues per hundred words; correction attempts; graphic simil- arity; sound similarity; and grammatical function. As a group, the readers in this study performed more proficiently on all measures on the dicta- ted story. 130 131 Measures of Reader's Perception of Reading The reader's perception of reading was measured by prior instruc— tion, theoretical model, and asides made during the process of reading the texts. The results showed that both the teacher and the students viewed reading as a decoding-to-sound process. On all measures, the emphasis was on letters or words, with little attention paid to context. Measures of Text Structure Two Word level measures, one syntactic measure, and one semantic measure were used to describe the structure of the texts. On both the word level and syntactic level measures, the overwhelming majority of dictated stories (all but one) were found to be more difficult. All of the dictated stories contained more unfamiliar and untaught words than the text story, and all of the dictated stories (except one) contained more syntactically complex sentences than the text story. An assess- ment of the structure of the texts on the semantic level showed that, generally, the dictated stories were more well-formed. They included setting statements and complete episodes while the text story did not. The text story appeared more well formed only on the relationships between episodes. Discussion The results of this study did support the contention that ' language experience activities, in this case a story dictated to ac- company a wordless picture book, can enable young readers to process print more proficiently. Given the perception of reading held by the students in this study and the increased complexity of their dictated stories, performance on the dictated story was surprising. Because of 132 an over reliance on sounding out or prior knowledge of words, these students should not have performed as well as they did when faced with a task that appeared to be more difficult. The penchant to sound out should have resulted in an increase in the percentage of miscues that were very similar on the graphic and sound similarity measures. As a group, this result did not occur. Rather, sound/symbol correspondences decreased on the dictated stories. Further, the perception that reading is a list of words—-either known or unknown--resu1ted in more omissions (words perceived as unknown) than substitutions (words perceived as known) for the three less proficient readers in their encounter with the text story. Yet, on the dictated story, two of these readers were able to reverse that strategy and produced more substitutions than omissions. On the group measures, then, the reader's perception of reading did not explain performance. However, it did help to explain some individual performances. One reader, for example, was not able to per- form more effectively on any measure of strategy selection on the dictated story. Interestingly, this reader was the more over-dependent on the ”sounding out” strategy than any of the other readers in the study. Faced with an unknown word, Jack almost always chose to use multiple attempts at ”sounding out," even though in his interview he suggested that "sounding out" didn't always work. Jack often produced as many as eight to ten attempts, usually resulting in non—words. However, while his strategy was consistent across quarters for the text story, the number of non-words decreased sharply in the third and fourth quarters of the dictated story. Since word and syntactic level measures showed that the dictated 133 stories were more difficult than the text story, this measure did not explain performance either. Again, given the increased difficulty of the children's stories, their performance was surprisingly proficient. 0n the semantic level, the dictated stories were found to be more well—formed (contained setting statements and complete episodes) than the text story. Only the semantic level measure appeared to explain reader performance. Perhaps the completeness of episodes in the child- ren's stories gave them the kind of semantic knowledge necessary to go beyond their usual letter and word level strategies and pay more attention to meaning. Given that the children's stories were more similar to the narratives generally found in well-written children's literature, the readers may have been able to use their general know— ledge of stories to gain access to better strategies. Traditionally, the level of difficulty of a text has been measured on word and sentence levels. Uncommon words and syntactically complex sentences have been assumed to present the reader with a more difficult task. The results of this study do not support this idea. Rather, it seems that uncontrolled vocabulary and complicated syntax can be handled efficiently if students are allowed to produce their own texts. Reader-authored texts may be more accessible precisely because the readers were the authors. Certainly, interest in reading what one has produced must be a factor. Beyond general interest, the fact that the reader-author already "knows" the meaning of the story must enable him/her to use the semantic cue system more effectively. The increase in semantically acceptable sentences and correction of miscues as well as the drop in sentences that changed the intended meaning of the author, all point to a more effective use of the semantic cue system of 134 language when the readers encountered their own stories. Finally, not only were the dictated stories more well-formed than the text story, but well-formedness was directly related to reading proficiency for individual readers. The most proficient readers produced the most com- plete stories while the least proficient readers produced the least complete stories, indicating that knowledge of the categories necessary for a well—formed story appears to enable a reader to both produce and process stories more proficiently. The results of this study indicate that the text ggg_influence strategy selection. However, traditional readability measures do not appear to predict more accessible texts as readily as semantic measures. Further, even reader-production of a more predictable text does not ensure more proficient reading strategies for all readers. Some readers may choose to monitor instruction rather than the text. However, for most of the readers in this study, the text did have a positive' influence. Implications Because this study is limited to only eight children, the results cannot be generalized to all readers. The sample was too small and not selected randomly; the students had only one opportunity to engage in a language experience activity; and only one type of language ex- perience activity, dictation to accompany a wordless picture book, was used. The implications of this preliminary investigation, however, do merit discussion. Teacher Preparation The assertion that both the teaching and learning of reading are 135 theoretical (Harste 8 Burke, 1977) was demonstrated by the teacher and students in this study. The teacher claimed that she taught reading through phonics and sight words. Both her choice of materials and her use of instruction time demonstrated that her teaching reflected her beliefs. ‘The students, too, in their interviews and verbalizations, reflected the teacher's theoretical model. (Even though the effect of a more predictable text (the dictated story) enabled the students to utilize reading strategies more efficiently, their general reading patterns did not change much.‘ Students, it seems, also have theoretical models of reading and, while some texts may enable them to use that model more efficiently, their basic approach to print is maintained. Teachers like the one in this study who have been exposed to only. one model of reading need to be made more aware of the different theo- retical models. Where teacher training focuses solely on one model of reading, an introduction to all approaches: phonics, skills, and whole language would constitute a more functional preparation of teachers. Reading and Composition In this study, the relationship between reading and oral and written composition was highlighted. A comparison of reader profic- iency and percentage of complete episodes in dictated stories revealed that the more proficient readers produced the most well-formed stories. Conversely, the less proficient readers produced the least well-formed stories. Knowledge of story structure appears to influence both pro- ductive and receptive modes. Initially, a child need to be read to in order to develop a story framework. Once the child begins to develop a schema for stories, the related processes of reading and oral and 136 written story composition should complement each other. Each attempt to produce a story should further the student's general knowledge about stories so that better predicting, confirming, and comprehending stra- tegies can be used on the next attempt to read a story. The interrela- tionship between reading and composing and the fact that knowledge from one mode can influence performance in another are strong arguments for language experience activities. The fact that the stories produced by the children in this study were ranked by percentage of complete episodes in exactly the same order as the students were ranked by proficient use of reading strategies, points to the fact that knowledge available for receiving language, reading, is also knowledge available for producing language, composition. Instructional procedures need to be built on this interrelationship. Not only do children need opportunities to author their own stories for reading; they need opportunities to engage in a variety of writing activities in order to explore the myriad relationships between the two modes. Of particular importance is the opportunity to handle the writ- ing process themselves. Attempts to simplify or control the production of writing strips away support. Rather, instruction (and research) should be designed for exploration of the complexity of the writing process in an attempt to discover how such complexity generates support. Less. In this study, restrictions of vocabulary and sentence complexity did not seem to ensure more proficient reading strategy utilization. Measures of difficulty on both the word and sentence level indicated that the dictated stories should have been more difficult to read. 137 Yet, the children read their dictated stories with greater ease even though the vocabulary and sentence structure were not those they had previously encountered in classroom instruction. Attempts to simplify texts based on word level and syntactic level measures appeared to reduce needed semantic support. Perhaps the over simplification of texts for beginning readers impedes rather than supports children's attempts to make sense of print. Both teachers and publishers need to be made more aware of the lack of support given by some types of texts. If the texts currently used in public schools generally render learning to read more difficult than necessary, changes need to be implemented. With increased know- ledge, authors of reading materials could produce more predictable texts and teachers could make better choices when selecting materials, especially for beginning readers. Future Research In this study, the attempt to document and explain reader perform- ance proved to be a complex task. Many factors contribute to the process called reading and attempts to isolate various sub-skills have not re— sulted in adequate descriptions of reading. Newer models that emphasize the interaction between reader and text hold more promise. Substantial progress has been made in recent years, but the effort needs to be continued. In this study, several aspects of both the contribution of the reader and the contribution of the text were examined, but other factors need to be investigated as well. Readers, for example, interact with print in a variety of settings outside the classroom. What effect do these interactions have on their personal model of reading? Texts, 138 too, reflect a model of reading and have been shown to elicit different kinds of strategies depending on the model of instruction held by the authors of the texts (Rhodes, l979b), but more texts need to be studied. Finally, the interaction between reader and text does not take place in a vacuum. The particular environment in which reading occurs can also influence strategy selection. The students in this study were removed from the classroom for the purpose of testing. If the data had been collected in the classroom or during a special writing clinic or at home, would the results have been different? As psycholinguistic models of reading move to soci-psycholinguistic models of reading, an adequate description of the reading process becomes increasingly complicated. The acquisition of print occurs in many places other than the classroom and under many different types of constraints. Future research should continue to investigate not only language experience activities but all reading in light of these newer models. Attention should also be given, in future research, to newly devel- oped text analysis procedures. Attempts to map the cognitive structures that control the production and recall of stories are very new and not well developed; but, in time, may result in better methods of assessing readability as well as the production of more readable texts for instruc- tion. In this study, the stories produced by the readers were found to be much more well-formed than the story taken from the basal reader. Attempts to describe other basal reader stories could be made using the story grammar employed in this study (Stein 8 Glenn, l977b) or any of several other text analysis procedures currently being developed. Only one story from one pre-primer was described in this study. Other stories 139 from the same pre-primer could be compared or stories from a variety of basal reading series could be compared. Traditional readability measures, too, need to be reassessed in light of current research. In this study, the semantic level measure appeared to predict reader success better than any other measure. More research is needed in this area. As researchers expand their knowledge of language, the study of all language encounters becomes increasingly difficult. This investiga- tion centered on one instance of one type of language experience activity. Still, the attempt to document and explain that one activity revealed a maze of interrelationships. Perhaps, in order to better understand language use, and reading as one instance of language use, future research will necessarily be limited to the study of only a few subjects at a time. A truly close examination of language, in all its complexity, cannot be achieved in any other way. The description of even a single reader's interaction with a single piece of print is not a simple task. As J. Harste has suggested, ”Any instance of written language use...is best viewed as the orchestration of a complex social event" (Harste, 1980, p. 2). Traditionally, reading educators, being cognizant of this fact, have interpreted it to mean that simplification of the process engenders accessibility. This study questions that interpretation. Conceptually and instructionally, the findings of this study mandate the development of more viable models of written language which recognize that written language, in its natural complexity, support control. BIBLIOGRAPHY BIBLIOGRAPHY Adams, M. J., 8 Collins, A. A schema-theoretic view of readipg_(Tech. Rep. 32). Urbana, IL: University of Illinois, Center for the Study of Reading, April 1977. Allen, P. D. A_psycholinguistic ana1ysis of the substitution miscues of selected oral readers in grades two, four, and six and the relation- ships of these miscues to-the reading process: a descriptive study. Unpublished doctoral dissertation, Wayne State University, 1969. Allen, P. 0. Some general implications concerning specific taxonomy categories. In P. 0. Allen 8 D. J. Watson (Eds.), Findings of research in miscue analysis: classroom implications. Urbana, IL: National Council of Teachers of English, 1976. Allen, P. 0., 8 Watson 0. J. (Eds.), Findings of research in miscue analysis: classroom implications. Urbana, IL: National Council of Teachers of English, 1976. Allen, R. A. More ways than one. Childhood Education, 1961, 38, 108-111. Anderson, R. C., Reynolds, R. E., Schallert, D. L., 8 Goetz, E. T. Frame- works for comprehending discourse. American Educational Research Journal, 1977, 14(4), 367-381. Andrews, N. C. Six case studies in learning to read. Unpublished doctoral dissertation, Indiana University, 1976. Applebee, A. N. The child's concept of stopy: ages two to seventeen. Chicago: University Press, 1978. Ashton-Warner, S. Teacher. New York: Simon 8 Schuster, 1963. Au, K. H. Analyzing oral reading errors to improve instruction. 1be_ Readinngeacher, 1977, 31, 46-49. Barr, R. The effect of instruction on pupil reading strategies. Reading Research Quarterly, 1974-75, 10(4), 555—582. Bennett, S. W. The key_yocabulary in organic reading: an evaluation of some of Ashton-Warner's assumptions of beginning reading. Unpub- lished doctoral dissertation, University of Michigan, 1971. 140 141 Betts, E. A. Factors in readiness for reading. Educational Administra- tion and Supervision, April 1943, 29, 199-230. Biemiller, A. J. The development of the use of graphic and contextual information as children learn to read. Reading Research Quarterly, 1970, 6, 75-96. Bond, G., 8 Dykstra, R. The cooperative program in first grade reading instruction. In R. Stauffer (Ed.), The first grade reading studies. Newark, DE: International Reading Association, 1967. Brazee, P. E. A qualitative and quantitative description of eighth grade students' oral reading in both narrative and expository materials. Unpublished doctoral dissertation, University of Northern Colorado, 1976.. Brazziel, W. F., 8 Terrell, M. For first graders: a good start in school, Elementary School Journal, 1962, 62, 352-355. Brody, D. P. A psycholinguistic comparison of oral reading behavior of proficient and remedial readers. Unpublished master's thesis, Rutgers University, 1973. ‘ Brown, G. Children's sense of story in relation to reading and writing, Unpublished master's thesis, The Ohio State University, 1977. Burke, C. L. A language experience. Unpublished manuscript, Blooming- ton, IN: Reading Department, Indiana University, 1976. (a) Burke, C. L. A psycholinguistic description of grammatical restructur- ings in the oral reading of a selected group of middle school children. Unpublished doctoral dissertation, Wayne State Univer- sity, 1969. Burke, C. L. The readipg interview. Bloomington, IN: Reading Depart- ment, Indiana University, 1978. Burke, C. L. The three cue systems: #2--syntactic. In P. 0. Allen 8 D. J. Watson (Eds.), Findings of research in miscue ana1ysis: classroom implications. Urbana, IL: National Council of Teachers of English, 1976. (b) Burke, C. L., 8 Goodman, K. S. When a child reads: a psycholinguistic analysis. Elementarnynglish, 1970, 41, 121-129. Carlson, K. L. A psycholinguistic description of selected fourth grade children reading a variety of contextual material. Unpublished doctoral dissertation, Wayne State University, 1970. Carstensen, L. M. Language experiences found in teacher guides for first grade reading textbooks. Unpublished doctoral dissertation, Arizona State University, 1973. 142 Chomsky, C. Write first, read later. Childhood Education, 1971, 41(6), 296-299. Christensen, K. E. Language facility of kindergarten children. Elemen- tary English, 1972, 42, 1107-1111, 1119. Christie, J. F. The effect of later appearing syntactic structures on seven- and eight:year-old children's oral reading errors. Unpub- lished doctoral dissertation, Claremont Graduate School, 1977. Clay, M. M. A syntactic analysis of reading errors. Journal of Verbal Learning and Verbal Behavior, 1968, 2, 434-438. Clay, M. M. What did I write? Auckland, Australia: Heinemann Educa- tional Books, 1975. Cohen, A. Oral reading errors of first grade children taught by a code emphasis approach. Reading Research Quarterly, 1974, 19(4), 616- 650. Cohen, S., 8 Kornfield, G. Oral vocabulary and beginning reading in disadvantaged children. The Reading Teacher, 1970, 24, 33-38. Coomber, J. E. A psycholinguistic analysis of oral reading errors made by good, average,.and poor readers. Unpublished doctoral disser- tation. The University of Wisconsin, 1972. Cooper, C. R., 8 Petroskey, A. R. A psycholinguistic view of the fluent reading process. Journal of Reading, December 1976, 184—207. Cox, V. E. Reciprocal oralcy/literacy recogpition skills in the lang- uage production of language erperience students. Unpublished doctoral dissertation, University of Arizona, 1971. Dank, M. E. A study of the relationship of miscues to the mode of formal reading instruction received by selected second graders. Unpublished doctoral dissertation, University of Massachusetts, 1976. DeLawter, J. A. The relationship of beginning reading instruction and miscue patterns. In W. D. Page (Ed.), Help for the reading teacher: new directions in research. Urbana, IL: National Conference on Research in English, ERIC Clearinghouse on Reading and Communication Skills, National Institute of Education, 1975. Elder, R. 0. Oral reading achievement of Scottish and American child- ren. Elementary School Journal, January l97l,.11, 216-230. Farr, R. Reading: what can be measured. Monograph from the ERICZCRIER Reading Review Series. Newark, DE: International Reading Associa— tion Research Fund, 1969. 143 Giles, D. E. The effect of two approaches to reading instruction upon the oral language development of first grade pupils. Unpublished doctoral dissertation, North Texas State University, 1966. Goodman, K. S. A linguistic study of cues and miscues in reading. Elementary_English, 1965,.4g, 639-643. Goodman, K. S. Reading: a conversation with Kenneth Goodman. Glen- view, IL: Scott, Foresman, 1976. Goodman, K. 8. Reading: a psycholinguistic guessing game. Journal of the Reading Specialist, May 1967, 4, 126-135. Goodman, K. S. What we know about reading. In P. 0. Allen 8 D. J. Watson (Eds.), Findings of research in miscue analysis: classroom implications. Urbana, IL: National Council of Teachers of English, 1976. Goodman, K. S., 8 Goodman, Y. M. Learning about psycholinguistic processes by analyzing oral reading. Harvard Educational Review, 1977, 41, 317-333. Goodman, Y. M. Developing reading proficiency. In P. 0. Allen 8 D. J. (Eds.), Findings of research in miscue analysis: classroom implications. Urbana, IL: National Council of Teachers of English, 1976. Goodman, Y. M. A psycholinguistic description—of observed oral reading phenomena in selected beginning readers. Unpublished doctoral dissertation, Wayne State University, 1967. Goodman, Y. M., 8 Burke, C. L. Reading miscue inventory: evaluation form. Indiana University, School of Education, Reading Department, 1976. Goodman, Y. M., 8 Burke, C. L. Reading miscue inventory: rprocedure for diagnosis and evaluation. New York: Macmillan, 1972. Goodman, Y. M., 8 Burke, C. L. Reading strategies: focus on compre— hension. Holt, Rinehart, and Winston, 1980. Graves, D. What children show us about revision. Language Arts, March 1979, 56(3), 312-319. Greene, B. H. L. A psycholinguistic analysis of the multiple attempts produced in oral reading by selected readers upon the single appearance of an item within a text and upon items recurring throughout a text. Unpublished doctoral dissertation, Wayne State University, 1974. Hahn, H. Three approaches to beginning reading instruction--ITA, language experience, and basic readers. The Reading Teacher, 1966, 12; 590-594. 144 Hall, M. A. The-development and evaluation of a language experience a — proach to reading with firstagrade cultura11y disadvantaged child- rgp, Unpublished doctoral dissertation, University of Maryland, 1965. ‘ Hall, M. A. The language experience approach for teaching reading: a research prospective. Newark, DE: The International Reading As- sociation, 1977. Halliday, M., 8 Hasan, R. Cohesion in English. London: Longman, 1976. Harris, A. J., 8 Serwer, B. L. Comparing reading approaches in first grade teaching with disadvantaged children. The Reading Teacher, l966,_12, 631-635. Harris, G. R. The effects of communication skills through authorship: a language experience suppjement on the development of reading skills with primary students. Unpublished doctoral dissertation, University of Idaho, 1972. Harste, J. C. Language as a social event. Paper presented at the an- nual meeting of the American Educational Research Association, Boston, April 11, 1980. Harste, J. C. Reading and reality. In J. Robinson (Ed.), Whole lan- guage for whole children. Columbia, MO: Reading Department, University of Missouri, in press. Harste, J. C., 8 Burke, C. L. A new hypothesis for reading teacher re- search: both teaching and learning of reading are theoretically based. In P. 0. Pearson (Ed.), Reading; theory, research, and practice. Twenty-Sixth Yearbook of the National Reading Confer- ence: Mason, 1977. Harste, J. C., 8 Burke, C. L. Toward a socio-psycholinguistic model of reacing comprehension. Viewpoints in Teaching and Learning, 1978, 4p 3). Harste, J. C., Burke, C. L., 8 Woodward, V. A. Children, their language, and world: initial encounters with_print. NIE Research Grant NIE-G-79-Ol32, l979. Harste, J. C., 8 Carey, R. Comprehension as setting. Monograph in Language and Reading Studies, 1979, p, 4-22. Henderson, E., Estes, T., 8 Stonecash, S. An exploratory study of word acquisition among first-graders at midyear in a language experi- ence approach. Journal of Reading Behavior, 1972, 4, 21-31. Hildreth, G. Experience related reading for school beginners. Elemen- tary English, 1965, 42, 280-97. 145 Hocevar, S. P. 8 Hocevar, D. Facilitating beginning reading with the use of meaningful context. Journal of Reading Behavior, 1978, 1g (3). Hoover, 1. Historical and theoretical development of a language exper- ience approach to teaching reading in selected teacher education in- stitutions. Unpublished doctoral dissertation. University of Ari- zona, 1971. Hunt, K. W. Grammatical structures written at three grade levels (NCTE Report No. 3). Champaign, IL: NCTE, 1965. Hunt, K. W. Recent measures in syntactic development. Elementary Eng- lish, l966,.43, 732-739. Jensen, L. J. A psycholinguiStic analysis of the oral reading behavior of selected proficient, average, and weak readers reading the same material. Unpublished doctoral dissertation, Michigan State Uni- versity, 1972. Kelly, A. M. Sight vocabularies and experience stores. Elementary Eng- lish, 1975, pg, 327-328. King, M. L., 8 Rentel, V. Toward a theory of early writing development. Research in the Teaching_of Engjish, Fall, 1979. Kintsch, W., 8 van Dijk, T. Toward a model of text comprehension and production. Unpublished paper, Department of Psychology, Univer— sity of Colorado, 1976. Kolers, P. A. Three stages of reading. In H. Levin 8 J. P. Williams (Eds.), Basic studies on reading, New York: Basic Books, 1970. Kopel, 0. Reading readiness: its determination and use. ‘Teachers College Journal, January, l942,_1§, 64-70. Lane, K. B. A description, analysis and evaluation of three approaches to the teaching of reading. Unpublished doctoral dissertation, North Texas State University, 1963. Lee, D. M., 8 Van Allen, R. Learning to Read Through Experience. Englewood Cliffs, NJ: Prentice-Hall, 1963. Levine, E. F. The relationship between threepatterns of oral reading errors and specific reading strategies. Unpublished doctoral dis- sertation, Northwestern University, 1976. Malmstrom, J., 8 Weaver, C. "Syntactic maturity” and syntactic appro- priateness in teaching writing. In N.A. Johnson (Ed.), Current topics in language. Cambridge, MA: Winthrop, 1976. 146 Mandler, J. M. 8 Johnson, N. S. Remembrance of things parsed: Story structure and recall. Cognitive Psychology, 1977, 2, 111-151. Martellock, H. A. A psycholinguistic description of the oral and writ- ten lapguage of a selected group_of middle school children. Un- published doctoral dissertation, Wayne State University, 1971. Mc Canne, R. Approaches to first grade English reading instruction for children from Spanish speaking homes. The Reading Teacher, 1966, _13, 70-75. McC. Gallager, J. A. Vocabulary retention of lower-class students in language experience and in basal approaches to the teaching of reading, Unpublished doctoral dissertation, University of South Carolina, 1975. McKenzie, M. Learning to read through reading. In E. Hunter-Grundin 8 H. U. Grundin (Eds.), Reading: implementingythe Bullock Report. London: United Kingdom Reading Association, 1978. Menosky, D. M. A_psycholinguistic description of oral reading miscues generated during the reading of varying portions of text by selec- ted readers from grades two, four, six and eight. Unpublished doctoral dissertation, Wayne State University, 1971. Minsky, M., A framework for representing knowledge. In P. Winston (EdL), The psychology of computer vision. New York: McGraw Hill, 1975. Neisser, V. Cognition and reality. San Francisco: W. H. Freeman, 1976. Norton, 0. E. A comparison of the oral reading errors of high and low ability first and third graders taught by_two approaches--synthetic- phonic and analytic—eclectic. Unpublished doctoral dissertation. The University of Wisconsin, 1976. O'Donnell, C., 8 Raymond, 0. Developing reading readiness in the kinder— garten. Elementary English, 1972, 49, 768-771. Packer, A. Ashton-Warner's key vocabulary for the disadvantaged. LIE: Reading Teacher, 1970, 23, 559-564. Pearson, P. 0. Research in reading comprehension: an interactive_per— spective. Paper presented at International Reading Association, 1978. Peck, L., 8 McGlothin, L. E. Children's information and success in first-grade reading. Journal of Educational ngchology, December 1940, g1, 653-664. Rhodes, L. K. Comprehension and predictability: an analysis of begin- ning reading materials. Monograph in Languagg and Reading Studies, 1979, a, 100-131. (a) 147 Rhodes, L. K. The interaction of begjnning readers' strategies and texts reflecting alternate models of predictability. Unpublished doctoral dissertation, Indiana University, 1979. (b) Rhodes, L. K. Predictable books: an instructional resource for meaning— ful reading and writing. In J. C.llarste 8 M. Atwell (Eds.), Main— streaming, the special child, and the reading_process. Bloomington, IN: Reading Department, Indiana University, 1976. Rousch, P, D. A_psycholinguistic investigation into the relationship between prior conceptual knowledge, oral reading miscues, silent reading, and post-reading performance. Unpublished doctoral dis- sertation, Wayne State University, 1972. Rumelhart, D. E. Notes on a schema for stories. In D. G. Brown 8 A. Collins (Eds.), Representation and understanding: studies in cog- nitive science, New York: Academic Press, 1975, 211-236. Rumelhart, D. E. Toward an interactive model of reading(Tech. Rep. no. 65). La Jolla, CA: University of California--San Diego, Center for Human Information Processing, March 1976. Rumelhart, D. E., Ortony, A., 8 Montague, W. E. The representation of knowledge in memory. In R. C. Anderson 8 R. J. Spiro (Eds.), ' Schooling and the acquisition of knowledge. Hillsdale, NJ: Law- rence Erlbaum, 1977. Schank, R. C. Identification of conceptualizations underlying natural language. In R.C. Schank 8 K. M. Colby (Eds.), Computer models of thought and language. San Francisco: Freeman, 1973. Schlieper, A. Oral reading errors in relation to rade and 1 l O O f Sk111. The Reading Teacher, 1977, 31(3), 2839287. eve O Schomer J. H. A study of the reading achievement ' , . . of first and second grade students 1n a commun1cation skills through authorship pgggram, Unpublished doctoral dissertation, University of Idaho, Smith, F. Comprehension and learning: a conceptual framework for teachers. New York: Holt, Rinehart 8 Winston, 1975. ‘ Smith, F- Eéycholinguistics and readin . New Yo . . smithiggi Understanding reading. New York: Holt, Rinehart 8 Winston, Smithiggé Understanding reading. New York: Holt, Rinehart 8 Winston, 148 Spache, G. D. Good reading for poor readers (rev. ed.). Champaign, IL: Garrard, 1974. Spache, G. D. A new readability formula for primary grade reading mater- ials. Elementary School Journal, 53, 410—413 Stauffer, M. A. Comparative effects of a language arts approach and basal reader approach to first grade reading achievement. Unpub- lished doctoral dissertation, University of Delaware, 1973. Stauffer, R. G. The effectiveness of language arts and basic reading approaches to first grade reading instruction. The Reading Teacher, 1966, go, 18-24. Stauffer, R. G. The language experience approach to the teaching of reading. New York: Harper and Row, 1970. Stauffer, R. G., 8 Hammond, W. Effectiveness of a langooge arts and basic reader approach to first grade reading instruction, co-oper- ative research project No. 2679. U. S. Department of Health, Edu— cation, and Welfare, 1965. Stauffer, R. G., 8 Hammond, W. The effectiveness of language arts and basic reader approaches to first grade reading instruction--extended to the third grade. Reading Research Quarterly, 1969, 4, 468-499. Stein, N. L. How children understand stories: a developmental analysis (Tech. Rep. No. 69). Champaign, IL: Center for the Study of Reading, March 1978. Stein, N. L., 8 Glenn, C. G. A developmental study of children's con- struction of stories. Paper presented at the Society for Research in Child Development, New Orleans, March 1977. (a) Stein, N. L., 8 Glenn, C. G. An analysis of story comprehension in elementary school children. In R. Freedle (Ed.), Multidisciplin— ary approaches to discourse comprehension. Hillsdale, NJ: Ablex, 1977, 53-120. (b) Stein, N. L., 8 Glenn, C. G. The role of structural variation in child- ren's recall of simple stories. Paper presented at the Society for Research in Child Development, New Orleans, March 1977. (c) Thomas, K. J. Syptactic/semantic acceptability and semantic similarity of oral reading errors as functions of variation in attained comp- rehension. The University of Arizona, 1975. Thorndyke, P. W. Cognitive structures in comprehension and memory of narrative discourse. Cognitive Psychology, 1977, p, 77—110. 149 Thornton, M. F. A_peycholinguistic description of_purposive oral read- iog and its effect on comprehension for supjects with different reading backgrounds. Unpublished doctoral dissertation, Wayne State University, 1973. Vilscek, E. C. What research has shown about the language-experience program. A decade of innovations: approaches to beginningoreading. Newark, DE: International Reading Association, 1968, 9-25. Vilscek, E. C., 8 Cleland, 0. Comparison of the basal and the co-ordi- nated language experience approaches in first grade reading instruc- tion, co-operative research project No. 2729. U. S. Department of Health, Education, and Welfare, 1966. Weber, E. M. An investigation of the effects of two readingpreadiness programs which were administered by parents to theiropost-kinder- garten children on measures of readiness, listening, and beginning reading. Unpublished doctoral dissertation, Michigan State Univer- sity, 1975. Weber, R. M. A linguistic analysis of first grade reading errors. Reading Research Quarterly, 1970, 5, 427-451. (a) Weber, R. M. First graders' use of grammatical context in reading. In H. Levin 8 J. P. Williams (Eds.), Basic Studies on reading, New York: Basic Books, Inc., 1970. (E) Wells, 1. M. An investigation designed to test the feasibility of using visual literacy techniques and the language experience approach to reading to develop the reading abilities of remedial fourth grade readers. Unpublished doctoral dissertation, Michigan State Univer- sity, 1975. Wixson, K. L. Miscue analysis: a critical review. Journal of Reading Behavior, 1979, 11(2), 163-175. Wrightstone, J. Research related to experience records and basal readers. The Reading Teacher, 1951, §(l), 5-6. BIBLIOGRAPHY OF INSTRUCTIONAL MATERIALS Durr, W. K., et al. Houghton Mifflin readers: Dinosaurs, Lions, Tigers. Boston: Houghton Mifflin, 1971. Sullivan, M. W. Readiness reader. Palo Alto, CA: Behavioral Research Laboratories, 1969. Winter, P. The bear and the fly. New York: Crown, 1976. 150 APPENDICES APPENDIX A Elicitation Method 151 ELICITING STORIES WITH WORDLESS PICTURE BOOKS by Sharon Thomas STEP ONE (explain purpose) I want to collect some stories that children have made up to go with picture books. I have a book here that has pictures but no writing to tell the story. I want you to tell me a story to go with the pictures in this book. STEP TWO (looking at the book) Here is the book. First, I want you to look at it and see what the story is about. Then, I want you to tell me about your story. After that, we'll write it down, and when we finish you can read your story to me. (Give the student plenty of time to look at the book.) STEP THREE (discussing the story) (Note: Close the book but leave it in front of the child. Start the discussion with the book closed but if the child wants to look at the book, let him/her do so.) Have you seen this book before? Tell me, in your own words, what your story is going to be about. (If child doesn"tre5pondcw*gives a very short response, WAIT before going on to questions.) Probing questions: Who is in your story? What do they do? Where does this story take place? What happens to make the story interesting? What's the problem in this story? What would be a good ending for this story? STEP FOUR (dictating the story) Now, I want you to tell me your story in your own words and I will write it down for you. When we finish, you can read your story. STEP FIVE (reading the story) Now, I want you to read your story to me. If you come to something you don't know, do whatever you usually do when you're reading alone. When you finish, I'm going to ask you to tell me about your story. APPENDIX B PARENT PERMISSION LETTER Consent Form 152 Dear Parent(s): I am a graduate student at Michigan State University. During winter term, I will conduct a research project at Forest View School. Because I would like your permission to allow your child to partici- pate, I am taking this opportunity to explain my project. For my study, I will ask the children to perform the following tasks: 1) Read a story from their basal reader. 2) Dictate a story to accompany a wordless picture book. 3) Read the story that they dictated. 4) Answer ten questions about how they learned to read. All of the children's responses will be tape recorded and I will use the information to develop a profile of your child's reading strategies. I will share the information with your child's teacher and with you, if you so desire. The information obtained from this project will appear in my dissertation. However, neither the city nor the school will be named and the children will be referred to by first name only. If your child has an unusual first name and your prefer an alias, please let me know. In order for your child to participate, you must sign the accompanying "Informed Consent" form and return it to your child's teacher. If at any time you wish to withdraw your child from the project, you are free to do so. Or, if your child wishes to withdraw, he/she is free to do so. If you have any questions, please call me. Thank you for your time and interest. working with your child. Sincerely, Sharon K. Thomas Office: 353-0834 Home: 153 332-5983 I look forward to 154 INFORMED CONSENT FORM Child's Name As the legal parent/guardian of the above named student I hereby give my permission for his/her participation in a research project being conducted by Sharon Thomas at Forest View School. I understand that my child will be asked to read a story, dictate a story, read the dictated story, and answer questions about his/her reading. I further understand that all responses will be tape recorded and that a description of my child's reading may appear in Ms. Thomas' dissertation. Finally, I understand that I may withdraw my child from the project at any time or my child may withdraw himself/herself at any time. Parent/Guardian Address Social Security Number Date APPENDIX C Reading Miscue Inventory Coding Sheet 155 C u u: u i. 111 . . . . n ..,. u u u u u u u u w w u u L a ._ ~ : .. 2: .. . c 533 «38 «39.3 5 a... .22. at :33: ~33. .53.: ~33. A.» a u z .— .. ._ p n m. z a r z y .. » . u u R H N S .3323. a At: 25:... venues: 1:: pr. a.” 3:... 5:: unsung-3.0 9:30» vasaouo . A: 039:! “yo—=9 .202 o ..n. p u unannou< «Qua-ov4 coco-eon so 3» «a a0 J o J m «9.2.0 1.5.3.: «due-ion 3.93:5» ..— .03.... e 593.. o 3 «a 3 .0 ~23: 8:5,...” .39» 38 annisJ .: =14u3ua¢059m ~O>~J 3.3) baby“ 9:. .201 G.,—Ovoud dCuficoAOBQSU Ono—Om 0.3:: APPENDIX D Reading Interview 156 READING INTERVIEW by Carolyn L. Burke Name Age Date Occupation Education Level Sex Interview Setting 1. When you are reading and you come to something you don't know, what do you do? Do you ever do anything else? 2. Who is a good reader that you know? 3. What makes her/him a good reader? 4. Do you think that she/he ever comes to something she/he doesn't know when she's/he's reading? 157 Yes When she/he does come to something she/he doesn't know, what do you think she/he does about it? No S that she/he does come to something that she/he 599%953- doesn't know., [9 e“ What do you think she/he does about it? If you know that someone was having difficulty reading how would you help them? a What would _________. teacher do to help that person? your How did you learn to read? 158 What did (they/you) do to help you learn? 9. What would you like to do better as a reader? 10. Do you think that you are a good reader? Yes No Additional Notes: APPENDIX E Spache Readability Formula Worksheet 159 Worksheet for Application of the Spache Readability Formula for Grades 1-111 Book Date Author Publisher Page Page Page Page From From From From To _ To To To 1. Total number of words 2. Number of sentences 3. Number of words not on Revised Word list 4. Average sentence length (Divide l by 2) . Multiply 4 by .121 . Multiply 3 by .082 . Add constant .659 .659 .659 ooucnm . Estimated grade level (Add 5, 6 and 7) Average of estimate Analyzed by .659 Date APPENDIX F Retelling Scoring Sheets 160 Reader Red Is Nice Character Analysis Recall (15) Development (15) Ken paints treehouse, doghouse, fence, and wants to paint house Mike paints treehouse, doghouse, fence, and wants to paint house Dad doesn't want his house painted Events (30) Mike wants to go to the library. Ken has to paint his treehouse. Mike offers to help so that they can go to the library. The boys paint the treehouse red. The boys get paint on the doghouse. They paint the doghouse red. The boys get paint on the fence. They paint the fence red. Dad comes. home . 10. The boys discover paint on the house. 11. Dad discovers paint on the boys. 12. The boys want to paint the house. 13. Dad wants boys to paint themselves. 14. The boys don't want to be painted red. 15. Dad doesn't want the house to be painted red. \OCIJNO'IU‘IDWN-d Plot (20) Each time the boys paint something,they'spill paint on something else and have to paint that too. Theme (20) People should think before they act. Retelling Score 161 Reader The Bear and the Fly Character Analysis Recall (15) Father Bear Mother Bear Baby Bear the dog the fly Events (30) The three bears are eating. A fly comes in. Father Bear makes a mess. The fly lands on Mother Bear. Father tries to get the fly. He hits Mother Bear. Mother Bear is knocked out. The fly lands on Baby Bear. A OOCDVmUl-th—J o o o o o o o o o o 11. He hits Baby Bear. 12. Baby Bear is knocked out. 13. The fly lands on the dog. 14. Father Bear tries to get the fly. 15. He hits the dog. 16. The dog is knocked out. Father Bear tries to get the fly. Father Bear tries to get the fly. Development (15) tries to get the fly is knocked out is knocked out is knocked out comes in the house, lands on the table, the mother, the baby, and the dog; gets away 17. Father Bear climbs on a chair on the table. 18. He tries to get the fly. 19. Father Bear falls off. 20. He knocks himself out. 21. The house is a wreck. 22. The fly gets away. Plot (20) Each time Father Bear tries to get the fly, he knocks somebody out. Father Bear makes a mess and the fly gets away. Theme (20) People should think before they act. Retelling Score APPENDIX G Verbalizations 162 VERBALIZATION: Text Story and Dictated Story Text Story Dictated Story Where was I? (Lost place) /S/ Want me to read it out loud? /S/ Is that an "i" or an "f"? /W/ What was that? (Commenting on a Is that an "i" or an "e"? /W/ liquid paper correction.) /5/ Got to be an "i". /W/ Guess what that is. (Mother) /W/ Not "a". /W/ No, let's see. Wouldn't think that There's that word again.(library) was funny. (He misplaced the period /W/ and has the sentence ending with This is "arf, arf, arf."(Written "little.") /M/ in the picture.) /W/ Wait a minute. Let me say that word Same as for the next word. /W/ again. ("Little.") /W/ I don't know what this is. (Know) Again? No. (Says "agin" for "again.") W W Right there.(Comes to "library" Where is it? (Looking for the word again. Goes back to first "again.") /W/ page and points to the first It's a "and." (Corrects miscue) /W/ occurrence of library.) /W/ No, it says--tuh-ay-bul, table. /W/ Two vowels are in here. (Trying No. Can't be that! (Read "The break- to sound out "surprise.") /W/ fast was runned." for "The break- Hmmm. That word's back here. fast was ruined.") /M/ (Fence) /W/ Then he flew near you. Then she looked No, doesn't make any sense. at him. (Commenting on what (Read: "The finished is red." happened next. Not in text.) /M/ for "The fence is red.") /M/ Can't figure this out. Is it "once Fend out? Huh uh (no). (Read: upon a time?" /W/ "We'll fend out." for "we'll Is that "the?" /W/ find out.") /M/ Is that "over?" /W/ Forget about that stuff. That's What if you can't read it? (Before too hard. (commenting on his reading the story.) /W/ attempt to sound out "find".) Is this buzz? (Result of attempt to /W/ sound out "Buzz.") /W/ Easy! Right? (Commenting on attemptOh, I don't know that word. ("Around.") to sound out "boys.“) /W/ /W/ The paint. Look at it. Red paint. Is this "Flew?“ (Father) /W/ /M/ I'm on this page. /S/ Boy! I shouldn't have picked this There's Dad. (Commenting on new kind of words. /W/ character in the story.) /M/ Geez. Did you draw these pictures End of the story. /S/ or were they in here? /S/. I don't know this word. (Nice) /W/ The pictures don't stay very well. This is a long word.(1ibrary)/W/ /s/ Hmmm (Looking at paint. Says, Oh! Geez! (Exclamation. Translated "P1ant?") /W/ as "Oh, look what's happening Wet? No. (Wet for what) /W/ here.") /M/ Let's see. What's this word. Geez! Now the bear's knocked out. More? No. Hmmm. /W/ /M/ Wait! (Says "we'll for ”will." Did Julian have trouble on his words? Then says, “wait" and changes /W/ it to "will.") /W/ Like? wait! (Has omitted "like." 163 Table VERBALIZATIONS: Text Story and Dictated Story Text Story Dictated Story Goes back and produces "like" but isn't sure yet.) /W/ I figured it out what I think it is. I tried to sound it out. ("Nice.") /W/ Well, I wanted to figure it out because when I start reading in it (this book) and I have to read this page I might have to read this word. ("House.") /W/. Is this "goed?" ("Good.") /W/ Wait a minute. (Says "you" for "can" and corrects.) /W/ No. (Says "will" for "we" and corrects.) /W/ No. (Says "will for "we'll" and corrects.) /W/ Everything's red! Hee,hee,hee. (Commenting on the picture.) /M/ No, I don't know that word. ("Ken.") /W/ No. Commenting that he doesn't know the word "now." /W/ I don't know that word. What that was. ("House.“) /W/ I just can't read this book. /W/ Ohh! (groan) No, I don't. (Doesn't know the word "find.") /W/ Ohh! (groan) (Omitted three words in sentence.) /W/ This is kind of hard. /W/ APPENDIX H Story Grammar Results Initiating Event Response Plan Reaction Consequence, Initiating Event Response Plan Reaction Consequence Initiating Event Response Plan Reaction Consequence Reaction Initiating Event * Response Plan (not developed) ------------------------------------------ I1 --1 .......................................... EC-- ............................................ 7 164 Red is Nice 00 you want to go to the library, Ken? I can't go with you now. I have to paint this treehouse red. 'm good at painting tree houses. I can help you paint it. Then we can go to the library. That looks good! Red is a good color for a tree house. We got paint on the doghouse. I know what we can do. Dad got all this red paint for me. And red's a good color for a doghouse. e'll paint the doghouse red. I'm getting sick of painting. We can't stop now. There! The doghouse is painted. We got red paint on the fence! r--- ----CAUSE We'll paint the fence. Then we can go to the library. his will surprise Dad! I'm hot and I'm sick of painting. Can't we go to the library now.? No. We have to get this painted. Get to work. There! The fence is red. Will your dad like it? We'll find out. Here he comes now. Do we have much more paint? Yes. We'll have to paint the fence. ed is a good color for a fence. hat's going on here, boys? I had to paint my tree house. And paint got on the doghouse. Then we had to paint the doghouse. And paint got on the fence. Then we had to paint the fence. .......................................... .——SL---CAUSE (um-THEN We'll have to paint the house. Red's a good color for a house. L---CAUSE ..................................................................... THEN 165 Initiating Event {JLOOK boys! here's red paint on you. Plan e'll have to paint you. Response Oh, no! Red isn't a good color for me. I don't want to be painted red. And I don't want a red house! * These two episodes could be considered reactions to the first three episodes. Setting (major) Setting (minor) Initiating Event Plan Attempt Consequence Attempt Reaction Consequence Initiating Event Attempt Consequence Initiating Event Initiating Event (Predictive) Initiating Event Attempt .Consequence Reaction Initiating Event (Predictive) Initiating Event Attempt Initiating Event Attempt Initiating Event Consequence Attempt Plan Consequence 166 Phil Once upon a time there were three bears and they were eating breakfast. And then a fly came in and the addy Bear saw it. oAnd then he got his fly swatter and he was swatting at it. Then he hit some juice and it spilled. And he hit almost everything on the table. And then he was hollering and the fly went away for a minute. ------------------------------------------------- THEN {And then it landed on the Mother Bear's head and he swatted it. And he hit mother's head and she got knocked out. ------------------------------------------------ THEN And then it was coming near the Baby Bear and it's going to land on her nose and then it landed on her nose and the Daddy Bear swatted it. And then she went knocked out. And then he was mad at the fly and he was coming towards it. ------------------------------------------------- THEN It was going to land on the dog's head And then it landed on the dog's head. And then the Daddy Bear swatted it and it was on the dog's head. And then the Daddy Bear swatted it and it was on the dog's head and the dog went knocked out. ------------------------------------------------- THEN Then he chased the fly. And then the Daddy Bear got a chair. And then he standed on the chair to get the fly. And then he fell down and he got knocked out too. And then the fly went out the window. Setting (major) Initiating Event Attempt Consequence Attempt Setting (minor) Attempt Consequence 167 Tim Once upon a time, they were eating dinner. And a fly came in nd it was bothering all of them he dad got up to hit it. e hit it. and all the stuff came over And she broke her silverware. Then the dad got up to hit it and they were watching it. Then he reached over and hit the mom on the head. nd he got the mom's head and the fly took off. And the mom was knocked out. And one of the things on the table were broke off. Then the mom was knocked out. -------------------------------------------------------------------- THEN Initiating Event Attempt (Predictive) Consequence ' Attempt Consequence Reaction Initiating Event Attempt Consequence, Initiating Event -- --------------------------------—--------‘------------‘ ----‘ {Then the dad was chasing the fly Attempt Setting (minor) Plan Consequence .._-...._.._-..---_..__..----..---............---_-......-....-..--..-.._..-_---....-...\ --—-----------------------------------. nd the fly was coming on her head. And then the dad was going to slap her. Then the fly got away. And he slapped her on the head. Then she was knocked out. - Then the dad was mad. Then the fly was going on the dog's head. Then dad slapped him all around the house. Then the mom had one eye open. Then the dad was climbing up on the chair. nd he was falling down. Then they were all knocked out. The t.v. was cracked. The plant was cracked. And the fly got away. ---------- THEN 1 ..... 4 ----- THEN Setting (major) Setting (minor) Initiating Event Plan Consequence Attempt Consequence Reaction, Initiating Event Attempt Consequence Attempt Consequence Reaction Initiating Event Attempt Consequence Reaction Attempt Consequence Initiating Event Attempt Consequence Reaction Initiating Event 168 Jack Slhe mother and the father and the little girl were eating breakfast. {And the dog was watching. Then, "Zoom!" Out from the window came a fly. And the Papa saw it. Then he got the fly swatter and almost got it. Then whapped again and he almost got it again. And the table was knocked. And all the water came out of the glasses. And so the breakfast was ruined. ............................................................. ”(----- He said, "I'll get you next time." -------------------------------------------- ”AL-THEN Then he came down again and again and again. And he missed him still. He whapped and whapped again and he still missed. "Next time, I'll get you for sure.” ------------------------------------------------------------- >M-----THEN {And the fly went on the mother's head. And Papa hit it nd she was knocked out. {Ber little girl cried. So did the PUPPY- -------------------------------------------------------------------- THEN So he kept on swinging and swinging until he just about got it. ------------------------------------------------------------- 7e-----THEN Then he got on the girl's head. And he whapped it with the swatter. And she got knocked out. Then the dog was sad. .......................................... p- -- And the dad was getting angrier and angrier and angrier. y ‘-- --THEN Attempt Consequence Attempt Plan Consequence 169 Jack Until, he hit the dog's head. Then the dog was knocked out. ------------------------------------------------- THEN {Eben he ran after the fly. He swatted nd swatted and swatted and swatted and swatted. Then he got a chair and put it on the table. Then he limbed up on it. And he fell down. And so they're all knocked out. The fly went around and around and around. Then he went out the window. (”""*\gf—‘*W That's the end of the three bears and the dog. Setting (major) Setting (minor) Initiating Event Setting (minor) Attempt (Predictive) Setting (minor) Initiating Event Plan Setting (minor) Consequence Reaction Consequence Setting (minor) Consequence Initiating Event Attempt Consequence Reaction Minor Setting Consequence Reaction 170 David hey're eating breakfast. The dog's begging. And the Mama Bear is drinking some juice. And the Father Bear is eating his breakfast And the little girl's putting the breakfast ‘n her mouth And then the bee comes flying in. Then the father looks up. nd the mother was cutting up her breakfast. And the little bear's looking at her father. And the napkin's hanging over the table. And the Father Bear's going to whack the bee with the fly swatter. ------------------------------------------------- THEN And then the little bear's biting her claw. And then the Father Bear's looking at the bee. and he is ready to smack it. And the Mother Bear's lifting up her paws like this and she's looking at Father Bear. hen the bee didn't get smacked. And the Mother's juice and the Father's juice went over the table. And the juice was flinging over the bread. And.one piece of bread was flying in the air. Then the Mother Bear looked at the father sort of mean like. And then the father's beans were flying up in the air. And his milk was splattering all over his rms. And then everybody's looking up at the ceiling. nd the juice is spilled on the little girl's plate. And the bread basket was dumped. And two pieces of bread were on the table. And then the father's milk glass was dumped. ------------------------------------------------- THEN And then the bee went up on the mother's head. He smacked the mother's head, nd the bee didn't get smacked. {And the little girl was looking at her mother a little worried. And then the mother's mouth was open a little bit. hen the mother got knocked out. The dog was getting kind of worried that he was going to hit him and the little girl. And the father's getting ready to smack the bee again. ------------------------------------------------- THEN Initiating Event Consequence (Predictive) Reaction Consequence (Predictive) Plan Attempt Consequence (Predictive) Reaction Initiating Event Plan Consequence, Initiating Event Attempt Plan Minor Setting Consequence .............................................. V 171 David And the bee was flying toward the little girl. ow she was going to get knocked out. {And the little girl was pointing at the mother and telling him something. Then he was going to knock her out pretty soon, And the Father Bear's getting ready to smack the bee on her head. And then he smacked her over the head. And then she was about to fall. Then the dog was trying to get the girl awake. Then the father bear was getting kind of mean. ---------------------------------------------- A;---THEN Then the dog standed up. And the bee was getting close to him. nd then the Father Bear was getting ready to smack him over the head with the fly swatter. ---------------------------------------------- q.-}N-- but the bee still didn't get hit. 1 -i--CAUSE And the Father Bear was chasing the bee. Then he was getting a chair and he was utting it on the table to smack him. And the Mother Bear was waking up. Then the Father Bear was starting to fall. And he drops the fly swatter. And now everybody's knocked out. And then the bee went out the window. ................................................................... 9L- Setting (major) Setting (minor) Initiating Event Setting (minor) Plan Attempt Consequence Attempt Consequence Reaction {They were eating dinner A 172 Karen Once upon a time, There lived three bears. IX fly came in and headed towards Father Bear Mother and Baby just ignored it. Father got the fly swatter and tried to kill the fly. He slapped the fly swatter on the table and everything went all over. He didn't get the fly 4K so he tried again. L ............................................. »--4-CAUSE e slapped Mama Bear on the head. Mama Bear screamed. Mama Bear went to sleep. ather Bear still tried to kill the fly. ............................................................... lL---THEN Initiating Event Response Attempt Consequence Reaction Initiating Event Response Attempt t came towards Baby Bear Father said, I'Hold still." And hit Baby Bear on the head. Baby Bear fell down and went to sleep. Father Bear got mad. ------------------------------------------------- THEN The fly came towards the dog. Father Bear said, "Hold still, dog." And hit the dog on the head. --------------------------------------------------------------------- THEN Attempt Plan Reaction Consequence Closure {gather Bear ran around with the fly watter trying to get the fly. He pulled a chair up onto the table. Mother opened one eye. Father Bear fell. He went to sleep. Their house was a mess. The fly went out the window. The end. Setting (major) Setting (minor) Initiating Event Setting (minor) Plan Attempt Consequence Attempt Consequence Reaction {fhey were eating dinner A 172 Karen Once upon a time, There lived three bears. fl\ fly came in and headed towards Father Bear Mother and Baby just ignored it. Father got the fly swatter and tried to kill the fly. e slapped the fly swatter on the table and everything went all over. He didn't get the fly so he tried again. ............................................. mil-CAUSE e slapped Mama Bear on the head. Mama Bear screamed. Mama Bear went to sleep. ather Bear still tried to kill the fly. ............................................................... 1 (----THEN Initiating Event Response Attempt Consequence Reaction Initiating Event Response Attempt t came towards Baby Bear Father said, "Hold still." And hit Baby Bear on the head. Baby Bear fell down and went to sleep. Father Bear got mad. -------------------------------------------------THEN The fly came towards the dog. Father Bear said, "Hold still, dog." And hit the dog on the head. --------------------------------------------------------------------- THEN Attempt Plan Reaction Consequence Closure {gather Bear ran around with the fly watter trying to get the fly. He pulled a chair up onto the table. other opened one eye. Father Bear fell. He went to sleep. Their house was a mess. The fly went out the window. The end. Setting (major) Initiating Event Consequence Initiating Event Attempt Consequence Reaction Initiating Event Attempt Consequence Initiating Event Attempt Consequence 173 Cindy The three bears were sitting at he table. And then the fly flew in. And then the Papa Bear saw the fly oing for the food. And then all the food tipped over because of Papa's fly swatter. And the drinks spilled and everything. ----------------------------------------------- THEN And then Papa Bear saw the fly going on Mama's head. He swatted Mama's head. nd Mama laid down on the table. Baby Bear was pretty sad. And then Baby Bear started yelling at Papa Bear. And then Baby Bear told Papa Bear what he did to her. ----------------------------------------------- THEN And then Papa Bear saw that the fly was going on Baby Bear's head. And then he swatted Baby Bear ’ and Baby Bear fell down. ............................................... THEN And then Papa Bear saw the fly going through the air. And he tried to swat the fly. And then all the bears were laying down except Papa Bear. And then the fly flew out. Setting (major) Initiating Event Plan Attempt Consequence Reaction Consequence Initiating Event Plan Attempt Initiating Event Setting (minor) Initiating Event (Predictive) Consequence Initiating Event Attempt Consequence Reaction Initiating Event Attempt Reaction 174 Julian The three bears were eating breakfast. Then, a fly buzzed inside. Then the Daddy Bear got his fly swatter and tried to hit the fly. The daddy made a mess nd still was chasing after the fly. The fly flew up high so the daddy father couldn' t get him. ---------------------------------------------- THEN {50, the fly decided to land on the mother's head Then the father took the fly swatter and hit the mother on the head. ---------------------------------------------- THEN he fly flew over near the plant. The daddy did not know that he broke a glass. he fly was going to land on the bear's nose. but father still was standing up straight. ---------------------------------------------- THEN {The fly was landing on the girl's head.~ All of a sudden, "Bam!" The girl was falling down on the ground. The dog looked at her. But the bear still had mean eyes. ---------------------------------------------- THEN head. hen the father saw him and the father hit the dog on the head. The father was still chasing the fly. {:he fly started to land on the dog's -----------------------------------------——---THEN 175 Julian The father picked up the chair and still hanged on the fly swatter. Plan The father put the chair on the table. _ And standed on the other chair for a stair. The father bear got up on the chiar Consequence (Predictive) nd the chair was ready to fall. Consequence The father bear fell right down off the table Setting (minor) and the mother bear still was rubbing her feet. Consequence he house looked a wreck. And three people were knocked out. Setting (minor) nd the mother bear was still rubbing , her feet. Closure The end. Setting (major) Initiating Event Plan Attempt Reaction, Initiating Event Attempt Consequence Reaction ..................................................... 1--------------THEN "Buzz." "There's a fly, Daddy. Get him" Initiating Event Attempt Attempt Consequence Reaction Initiating Event Attempt Consequence Attempt Plan Consequence Closure 176 Jeremy 'Buzz." A fly went into their room and was flying around. Father Bear took a fly swatter. "Bam! Boom! Smash!" f:[he bears were eating dinner. I ----------------------------------------------- THEN Then he yelled, "Get that fly." --------’------------------_-- ----------------- "Whack!" Hitted Mother Bear. Mother Bear was knocked out. Papa Bear went for the fly. "Buzz." “Whack!" ----------------------------------------------- THEN Papa Bear went after the fly. While Baby Bear was knocked out. Their dog was worried. ----------------------------------------------- THEN "Buzz." "Whack!" Mother Bear was knocked out. ----------------------------------------------- THEN And Father Bear went to get the fly. He got the chair up on the top of the table. . apa Bear fell He fell down and knocked himself out. Everything was broken. "Buzz." "Buzz." That's the end of the fly. APPENDIX I Dictated Stories 177 Phil N. Once upon a time there was three bears and they were eating breakfast. And then a fly came in and the Daddy Bear saw it. And then he got his fly swatter and he was swatting at it. Then he hit some juice and it spilled And he hit almost everything on the table. And then he was hollering and the fly went away for a minute. And then it landed on the Mother Bear's head and he swatted it. And he hit the mother's head and she got knocked out. And then it was coming near the Baby Bear and it's going to land on her nose. And then it landed on her nose and the Daddy Bear swatted it. And then she went knocked out. And then he was mad at the fly and he was coming towards it. It was going to land on the dog's head. And then it landed on the dog's head. 178 Phil N. 2 And then the Daddy Bear swatted it and it was on the dog's head and the dog went knocked out. Then he chased the fly. And then the Daddy Bear got a chair. And then he standed on the chair to get the fly. And then he fell down and he got knocked out too. And then the fly went out the window. 179 Tim Once upon a time, they were eating dinner. And a fly came in. And it was bothering all of them. The dad got up to hit it. He hit it and all the stuff came over. And she broke her silverware. Then the dad got up to hit it and they were watching it. Then he reached over and hit the mom on the head. And he got the mom's head and the fly took off. And the mom was knocked out. And one of the things on the table were broke off. Then the mom was knocked out. And the fly was coming on her head. And then the dad was going to slap her. Then the fly got away. And he slapped her on the head. Then she was knocked out. Then the dad was mad. 180 Tim - 2 Then the fly was going on the dog's head. Then the dad slapped him and the fly got away. Then the dad was chasing the fly all around the house. Then the mom had one eye Open. Then the dad was climbing up on the chair. And he was falling down. Then they were all knocked out. The t.v. was cracked. The plant was cracked. And the fly got away. 181 Jack The mother and the father and the little girl were eating breakfast. And the dog was watching. Then, "Zoom.” Out from the window came a fly. And the Papa saw it. Then he got the fly swatter and almost got it. Then he whapped again and he almost got it again. And the table was knocked. And all the water came out of the glasses. And so the breakfast was ruined. He said, ''I'll get you next time." Then he came down again and again and again. And he missed him still. He whapped and whapped again and he still missed. "Next time, I'll get you for sure." And the fly went on the mother's head. And the Papa hit it and she was knocked out. Her little girl cried. So did the puppy. So he kept swinging and swinging until he just about got it. Then he got on the girl's head. 182 Jack - 2 And he whapped it with the swatter. And she got knocked out. Then the dog was sad. And the dad was getting angrier and angrier and angrier. Until, he hit the dog's head. Then the dog was knocked out. Then he ran after the fly. He swatted and swatted and swatted and swatted and swatted. Then he got a chair and put it on the table.“ Then he climbed up on it. And he fell down. And so they're all knocked out. The fly went around and around and around. Then he went out the window. That's the end of the three bears and the dog. 183 David They're eating breakfast. The dog's begging. And the Mama Bear is drinking some jdice. And the Father Bear is eating his breakfast. And the little girl's putting the breakfast in her mouth. And then the bee comes flying in. Then the father looks up. And the mother was cutting up her breakfast. And the little bear's looking at her father. And the napkin's hanging over the table. And the Father Bear's going to whack the bee with the fly swatter. And then the Little Bear's biting her claw. And then the Father Bear's looking at the bee and is ready to smack it. And the Mother Bear's lifting up her paws like this and she's looking at the Father Bear. Then the bee didn't get smacked. And the mother's juice and the father's juice went over the table. And the juice was flinging over the bread. And one piece of bread was flying in the air. Then the Mother Bear looked at the father sort of mean like. And then the father's beans were flying up in the air. And his milk was splatteing all over his arm. 184 David - 2 And then everybody's looking up at the cieling. And the juice is spilled on the little girl's plate. And the bread basket was dumped. And two pieces of bread were on the table. And then the father's milk glass was dumped. And when the bee went up on the mother's head, he smacked the mother's head and the bee didn't get smacked. And the little girl was looking at her mother a little worried. And then the mother's mouth was open a little bit. Then the mother got knocked out. The dog was getting kind of worried that he was going to hit him and-the little girl. And the father's getting ready to smack the bee again. And the bee was flying toward the little girl. Now she was going to get knocked out. And the little girl was pointing at the mother and telling him something. Then he was going to knock her out pretty soon. And the Father Bear's getting ready to smack the bee on her head. And then smacked her over the head. And then she was about to fall. 185 Then the dog was trying to get the girl awake. Then the Father Bear was getting kind of mean. Then the dog standed up. And the bee was getting close to him. And then the Father Bear was getting ready to smack him over the head. And then he smacked him over the head with the fly swatter but the bee still didn't get hit. ‘ And the Father Bear was chasing the bee. Then he was getting a chair and he was putting it on the table to smack him. And the Mother Bear was waking up. Then the Father Bear was starting to fall. And he drops the fly swatter. And now everybody's knocked out. And then the bee went out the window. David - 3 186 Karen Once upon a time, there lived three bears. They were eating dinner. A fly came in and headed towards Father Bear. Mother and Baby just ignored it. Father got the fly swatter and tried to kill the fly. He slapped the fly swatter on the table and everything went all over. He didn't get the fly so he tried again. He slapped Mama Bear on the head. Mama Bear screamed. Mama Bear went to sleep. Father Bear still tried to kill the fly. It came towards Baby Bear. Father said, I'Hold still.‘I And hit Baby Bear on the head. Baby Bear fell down and went to sleep. Father Bear got mad. The fly came towards the dog. 187 Karen - 2 Father Bear said, "Hold still, dog." And hit the dog on the head. Father Bear ran around with the fly swatter trying to get the fly. He pulled a chair up onto the table. Mother opened one eye. Father Bear fell. He went to sleep. Their house was a mess. The fly went out the window. The end. 188 Cindy The three bears were sitting at the table. And then the fly flew in. And then Papa Bear saw the fly going for the food. And then all the food tipped over because of Papa's fly swatter. And the drinks spilled and eyerything. And then Papa Bear saw the fly going on Mama's head. He swatted Mama's head. And Mama laid down on the table. Baby Bear was pretty sad. And then Baby Bear started yelling at Papa Bear. And the Baby Bear told Papa Bear what he did to her. And then Papa Bear saw that the fly was going on Baby Bear's head. And then he swatted Baby Bear and Baby Bear fell down. And then Papa Bear saw that the fly was going on the dog's head. So Papa Bear swatted the dog's head too. And then Papa Bear saw the fly going through the air. 189 Cindy - 2 And he tried to swat the fly. And then all the bears were laying down except Papa Bear. And then the fly flew out. The 190 Julian The Three Bears with a Dog three bears were eating breakfast. Then, a fly buzzed inside. Then the Daddy Bear got his fly swatter and tried to hit the fly. The The $0, daddy made a mess and still was chasing after the fly. fly flew up high so the daddy father couldn't get him. the fly decided to land on the mother's head. Then the father took the fly swatter and hit the mother on the head. The The The but The A11 The The But fly flew over near the plant. daddy did not know that he broke a glass. fly was going to land on the bear's nose. father still was standing up straight. fly was landing on the girl's head. of a sudden, "Bam!" girl was falling down on the ground. dog looked at her. the bear still had mean eyes. The 191 Julian - 2 fly started to land on the dog's head. When the father saw him and the father hit the dog on the head. The The and The And The and The father was still chasing the fly. father picked up the chair still hanged on the fly swatter. father put the chair on the table. standed on the other chair for a stair. Father Bear got up on the chair the chair was ready to fall. Father Bear fell right down off the table and the Mother Bear still was rubbing her feet. The And And house looked a wreck. three people were knocked out. the Mother Bear was still rubbing her feet. The end. 192 Jeremy The bears were eating dinner. "Buzz." A fly went into their room and was flying around. Father Bear took a fly swatter. "Bam! Boom! Smash!" Then he yelled, "Get that fly." "Whack!" Hitted Mother Bear. Mother Bear was knocked out. Papa Bear went for the fly. "Buzz." "There's a fly, Daddy. Get him." "Buzz." "Whack!" Papa Bear went after the fly. While Baby Bear was knocked out. Their dog was worried. “Buzz." "Whack!" Mother Bear was knocked out. 193 Jeremy - 2 And Father Bear went to get the fly. He got the chair up on the top of the table. Papa Bear fell. He fell down and knocked himself out. Everything was broken. "Buzz." "Buzz." That's the end of the fly. APPENDIX J Text Story Mike: Ken: Mike: Mike: Ken: Ken: Mike: Ken: Ken: Mike: Ken: Mike: 194 RED IS NICE Do you want to go to the library, Ken? I can't go with you now. I have to paint this tree house red. I'm good at painting tree houses. I can help you paint it. Then we can go to the library. That looks good! Red is a good color for a tree house. We got paint on the doghouse! I know what we can do. Dad got all this red paint for me. And red's a good color for a doghouse. We'll paint the doghouse red. I'm getting sick of painting! We can't stop now. There! The doghouse is painted. We got red paint on the fence! Do we have much more paint? Yes. We'll have to paint the fence. Red is a good color for a fence. We'll paint the fence. Then we can go to the library. Ken: Mike: Ken: Ken: Mike: Ken: Dad: Ken: Mike: Ken: Mike: Ken: Dad: Ken: Mike: Dad: This will surprise Dad! I'm hot, and I'm sick of painting. Can't we go to the library now? No. We have to get this painted. Get to work. There! The fence is red. Will your dad like it? We'll find out. Here he comes now. What's going on here, boys? I had to paint my tree house. And paint got on the doghouse. Then we had to paint the doghouse. And paint got on the fence. Then we had to paint the fence. Oh, Ken! We got paint on the house! We'll have to paint the house. Red's a good color for a house. Look, boys! There's red paint on you! We'll have to paint you. Oh, no! Red isn't a good color for me. I don't want to be painted red. And I don't want a red house! APPENDIX K The Bear and the Fly PLEASE NOTE: Copyrighted materials in this document have not been filmed at the request of the author. They are available for consultation, however, in the author's university library. These consist of pages: 196-212 University. Microfilms International 300 N. ZEEB R0,, ANN ARBOR, Ml 48106 1313) 761-4700 196 GE; Ez 21322.35 :25 :52; 335 .3 33m 5 197 oéowuaAhoé ana— . - 23-2. 5. £25.58 63:. ._ 7.53 225... steam .u 5:381:88. .3 .338.— 38.33% 5.3 a: a :33 8 3+5 :2— < "muafifiam ..n: 05 1:: .33 2E. Janna £853 3am =e...5:s=& E 9.36.3.0 80.5.80 no >353 .328 out: 5 ‘35.:— vcd out: gQ—o “€92 hawk 2: .3 93.2.93 23.83 .3121 :9ch :33 amiss; .7: x33 9:. aces-.53.: 2C. a v a e p a a 3 60.:qu 5:39:60 ems—333m .30on .3 «15:5 5 Engage—2E... 3.3353 329:; me 5.5m .555.— 2: =_ 2339855.: 68.: ..r.Z .48? >32 .c=:..>< 32:.— 0:0 ..2: .E..:....:...:_ FEED S commit—E 3 2:2.» nor—3!: £05233. 2: as ...gcu=.:.... .7529; .:._.:_ .2: 255.3 {air—2:: .3 65.238 $532332... ._5_=E_P.E £18.52... .532: an: .3 no 5...... L9... 5 633.cmcfio .8 .2389. treatauu a E 19:5» €835..va on has. 5.33:3...— 35 .«c ta.— ez datum...» n23.— :< .553 .33 .3 82 e 22:80 k5 a»... ... . ,. ’3 p _:N” w.‘\-V* ‘ - \\/ ‘1‘»..— ( V5; \ \ r \ ..a .. 0...! . . . Onto-n, . ...:.. 5... .- CIC. 0'- .. o. to... u... h . k. 1.... a .13... . .. ..ou‘:...:~:“\\ 200 - é ..\.\\ .lro ., .....o....m. )— . . . “JV \.. ..I4...._ 1 H... .. .\ .. s“ \ .\. ../r E ....»...a.. . ...,.,, . ..,/....IV. Q5 .\.:.... \«\. ......a. 1../.../;.1//,..\ \....././,..,....,W/. . . . . x . v . . s I . . ...1. s ’2 . “1.1;... .4: x .21.... $3.. . . /. .... . .......... .... TA”)... .\..~. ......N. z . . .\\.....11............... ...»...rmxma..1.ws.1.;\\\\\m.(N....w./.~fl///I/nth..... J}... ....... 11... .. ...”..w < b ., 11.11.11 .4 .1\..:..1 ......v...\.._. T... ....I I . . .. . ... . na. 1 ..\. 4 ... or...- .... 1...... "5.x. ... .5. ...... ... .7 .us .. .sM .12...» {fr \\\. ‘1'] ..u\\.\. \x. .... .x\\ .. n13? . .... -.1 ‘- ~‘ cv-‘~ ’51,. K» 41... .. Q ...... 0.... ‘_ .., "1 \ "u“00--c' v \ o . 202 is“... . .... . . . .. «vawfiumj .. NR. , \.......\.\..::s.. . .....- {\éx -: ...“... ...... x... “WEN N.A.; ~ . . . 1. z. . . ‘ 3.”)! / . _ i. 5...,,..\.§, 5. “x Ax \.\\~ . «m . . s . . ...”...w ,L . ._ 3 ,1, . . . M. “N. Rfifidxfin 2 , .... .E x ...» ...u up... 3. . _. :5»... 44335. T h . .., .. .. .p—wnswxwo—mnu. 1%.”: a. . . ...7 RV“. . . A.».1..x\fif§¢. . 4 . . st... w¢ m.‘ .... \. s ‘ Q ~ J», . \\\..:mv.v II . .‘nlo ..."Ncu‘u. . .5. ..K. ~.\\ ..qu as: ‘ .mhs HH‘ \mkhll‘lh fifl VVVJ ..u. . \ ~ .3. x 3. \. .II. . u . I‘AMJH . ..1 n x: m C. (Q: (III-7.3).!" . I ‘ 3 . . ) . / u.~.\.. \‘§.\\_ bazcg§wiwquvx . t. .... , .\...:% . . V .‘ .H . . . m ...—...:Om . .. , . . . 4 . . ‘ .. .- g“. _..w&\:§.\\h . . a fl 203 . ... . ,.<. .. s... ...o . . 4.” w 77$...VA1 p: .. .¥2%aw1 . .5 LL? r. 5.» . ~. .... . 1 I. . \--.'..’.D \ . do ... .\ . . . . . -.-xfl‘x. A/w ..u......1¢..: ..x. ... .x.Elfé4§t§wrafimxf‘ufiaaéyég, a. .\ - \‘..u.\au.lh .m,..\. a «\HIQ I.o 7 Q . » .. . . \v&&;4a..\g.ré.a . .. 3gfirx.-..?. fix. . . \ I\ . Alhv If. \ ' L ......w .. i g§§§€ . $..V%wfimewN&pmeafifikme$% .1. 13$ .. ”CI" fri‘ r-NI: .‘ flu .-L ...... Ta... H.741. . if” 204 .- \\\ .‘Ih. 44N .cfiln“\ .l .- ..4. . ...»w'o‘......l..\\ . .\\.4 ...x \\».\\.§§$N&§+..£ex¢. Vamp. . . A .. . . \: kaxixamkfiru waft...» xx. .... ..4 v: “Wm; .. 4 “Sand .. urn 4 ..I,.., ‘3 . .. «43% 43.4. . .. ...... ..4 . “Ara“ . .\¥.x ... . \. 4 .. x4. Ls: .‘M .4 P .n.‘ - n . A . w . . . ... .. . \ ...; o.....x....4\.\ PEkfi Xx .. . . .... ...\\4 . s , _ . \.\\| ~ ... . H.. .......\ W.....\\..\L.\\\...4 ..r\, . .. H... .5 \\ ...4 k .. . , p. .. .... J.,... ...:(2. . . 4 . . . . . . _ ft . . . . .....-T. . a. if ....t“( u.~.pvp aflf "fr... “$14.7wa . _ - x‘ \ 7.,va w Pcfifi K O \ Q“. ‘i K \ f‘ ‘L' ..r 5.4.