132 137 lllllllllll|||||I||||llllll|||IIIIIIIHIIIIIIIHIIIImlullul “w“ 3 1293 10394 5402 ., ‘Lw: ' .115 ILA R Y Bifiichigm 5mm 0mm? ~' .THEE‘S This is to certify that the thesis entitled A Developmental Study of the Visual Processing of Syntactic Structures in Unrelated and Related Dis- course: A Canparison of Competent Middle School Readers and Competent Adult Readers presented by Barbara Ann Zynda has been accepted towards fulfillment of the requirements for Ph.D. Education degree in l 7 V Major professor Date July 22, 1980 0-7639 ‘ I l—L vmfll’bllu was: 25¢ per day per ttee RETUMIKS LIBRARY MTERIALS: Place in bookre “emote charge fro- circulation records A DEVELOPMENTAL STUDY OF THE VISUAL PROCESSING OF SYNTACTIC STRUCTURES IN UNRELATED AND RELATED DISCOURSE: A COMPARISON OF COMPETENT MIDDLE SCHOOL READERS AND COMPETENT ADULT READERS BY Barbara Ann Zynda 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 1980 M thQ ABSTRACT A DEVELOPMENTAL STUDY OF THE VISUAL PROCESSING OF SYNTACTIC STRUCTURES IN UNRELATED AND RELATED DISCOURSE: A COMPARISON OF COMPETENT MIDDLE SCHOOL READERS AND COMPETENT ADULT READERS By Barbara Ann Zynda Purpose of the Study The purpose of the study was to obtain and analyze data concerning the visual processing of syntactic structures in unrelated and related discourse by competent middle school readers and to compare their performance to that of adult able readers. The study focused on the effects of types of discourse in processing syntactic structures and the effects of left- and right-embedded structures and cognitive processing performance. The behaviors measured were dura- tion of gaze, duration of forward fixation, duration of regression, number of forward fixations, number of regressions, and number of total movements. Materials Materials consisted of unrelated and related discourse using active voice and employing no dependent clauses with the exception of the target embedding. The left- embedded structure was a relative clause that followed and Barbara Ann Zynda modified the subject. The right-embedded structure was a relative clause that followed and modified the sentence object. The conditions were presented with the BBL/Biometrics Reading Eye II, an electronic instrument that employs a photoelectric method to record eye movements on heat- sensitive graph paper. Population and Procedures Thirty sixth grade students were randomly selected from a group of competent readers chosen on the basis of teacher judgment, test scores, and parental consent. Thirty university graduate students were selected from a group of volunteers. They were presumed to be competent readers because of their educational status. All partici- pants were native English speakers. The subjects read five selections silently: an EDL paragraph, a left-embedded structure in unrelated discourse, a right-embedded structure in unrelated discourse, a left- embedded structure in related discourse, and a right- embedded structure in related discourse. Their eye movements were recorded with the EDL/Biometrics Reading Eye II. Major Findings The statistical tests supported the following findings: 1. There were no significant differences in the visual processing behaviors of competent middle school Barbara Ann Zynda readers of left- and right-embedded structures either in the unrelated or the related discourse condition. 2. There were no significant differences between competent middle school readers and competent adult readers in the visual processing of left-embedded structures within the comparison between the unrelated and the related discourse condition. 3. There were significant differences in the visual processing of competent middle school readers and competent adult readers in areas of behavior within particular syntactic/contextual conditions. These were: Duration of gaze LE, SS and RE, SS LE, 77’ and RE, 77’ RE, 55 and RE, 7/" Duration of forward fixation LE, 77 and RE, 77 RE, 53 and RE,7]" Duration of regression LE, SS and RE, SS LE, 77 and RE, 77 Regressions LE, SS and RE, SS Implications of the Study The results of the study indicated the following: 1. Competent middle school readers did not Barbara Ann Zynda significantly differentiate their reading to accommodate either the syntactic or the contextual demands of the material. 2. Competent middle school readers were more like competent adult readers in their fixation behaviors and less like competent adult readers in their duration behaviors. 3. A developmental sequence in psycholinguistic processing seems to be evolving. 4. The function and development of short-term memory regarding visual processing efficiency needs to be investigated. 5. Efficient psycholinguistic processing behaviors cannot be assumed simply because a student scores well on a standardized test. © Copyright by BARBARA ANN ZYNDA 1980 ii To Michael, Susinn, and Anne, My Family iii ACKNOWLEDGMENTS The writer is deeply indebted to Dr. Lois Bader, her chairperson, for her knowledge, guidance, support, and friendship. Dr. Bader is generous to an exceptional degree with her time and effort and encourages a similar effort from her students. The writer acknowledges the members of her committee, Drs. George Myers, James Snoddy, and Roy Wesselman, for their encouragement and guidance, and Dr. Keith Anderson for his cooperation and help. The author wishes to thank Gerald Kusler and Dr. Sal DiFranco for their help in conducting the study. Daniel Pearce, David Thompson, and Judy Thomas are to be thanked for their help, friendship, and support. The author is grateful to Kathleen Hengesbach for providing balance and perspective throughout this endeavor. Lastly, the writer wishes to thank her children, Michael, Susinn, and Anne, for their love and understanding. iv TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . vii Chapter I. THE PROBLEM. 1 Introduction . 1 Fur ose. 2 Bac ground . 2 Research Questions 6 Definition of Terms. . 7 Organization of Subsequent Chapters. 8 II. REVIEW OF RELATED LITERATURE . 9 Introduction . . . 9 Relationship between Reading Comprehension and Syntactic Structures . 9 Relationship between Visual Processing Behavior and Reading Comprehension . 13 Relationship between Visual Processing Behavior and Syntactic Structures. 18 Developmental Nature of Syntactic Processing . . . . . . . . . . . . 21 III. METHODOLOGY OF THE STUDY . . . . . . . . . . . 28 Introduction . . . . . . . . . . . . . . . . 28 Population . . . . . . . . . . . . . . . . . 28 Materials. . . . . . . . . . . . . . . . . . 29 Apparatus. . . . . . . . . . . . . . . . . 29 Data Collection. . . . . . . . . . . . . . . 29 Design . . . . . . . . . . . . . . . . . . . 3l Hypotheses . . . . . . . . . . . . . . . . . 31 Data Analysis. . . . . . . . . . . . . . . . 32 Summary. . . . . . . . . . . . . . . . . . . 32 IV. PRESENTATION AND ANALYSIS OF DATA. . . . . . . 34 Introduction . . . . . . . . 34 Hypotheses and Statistical Tests . . . . . . 34 Hypothesis 1 . . . . . . . . . . . . . . . 35 Hypothesis 2 . . . . . . . . . . . . . . . 36 Hypothesis 3 . . . . . . . . . . . . . . . 37 Hypothesis Hypothesis Hypothesis Hypothesis Hypothesis Analysis of Data Pertaining to Hypotheses 5, 6, and 8. Summary . . . . . . . . meU'I-P V. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summary . . Conclusions . . . Theoretical Implications. . . Visual Processing of Competent Middle School Readers. . . Comparison of Visual Processing of Compe- tent Middle School Readers and Competent Adult Readers . . . Implications of the Study . . Recommendations for Future Research . APPENDICES. A. SAMPLE OF LEFT- AND RIGHT-EMBEDDED STRUCTURES IN UNRELATED AND RELATED DISCOURSE. . B. TABLES CONCERNING THE VISUAL PROCESSING OF COMPETENT MIDDLE SCHOOL READERS, COMPETENT ADULT READERS, AND A COMPARISON OF COMPETENT MIDDLE SCHOOL READERS AND COMPETENT ADULT READERS . . . . . . . . . . . . . . BIBLIOGRAPHY. vi Page 38 39 40 41 42 43 45 48 48 49 52 52 53 56 59 59 62 68 Table 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 A-l A-3 LIST OF TABLES t- Test: Left- and Right- Embeddings in Unrelated Discourse. . . t- Test: Left- and Right- Embeddings in Related Discourse. . t- Test. Left- -Embeddings in Unrelated and Related Discourse. . . . . . t- Test: Right- Embeddings in Unrelated and Related Discourse. . . . t-Test: Student and Adult Processing of Left- and Right-Embeddings in Unrelated Discourse. t-Test: Student and Adult Processing of Left- and Right-Embeddings in Related Discourse. t-Test: Student and Adult Processing of Left- Embeddings in Unrelated and Related Discourse t-Test: Student and Adult Processing of Right— Embeddings in Unrelated and Related Discourse t-Test: Visual Processing of Left- and Right- Embeddings in Unrelated and Related Discourse by Competent Adults. t-Test: Visual Processing of Left- and Right- Embedded Structures in Unrelated and Related Discourse by Competent Middle School Readers t-Test: Visual Processing of Left- and Right- Embedded Structures in Unrelated and Related Discourse by Competent Adult Readers . . t-Test: Comparison of Visual Processing of Right- and Left-Embedded Structures in Unrelated and Related Discourse by Competent Middle School Readers and Competent Adult Readers. . . . . . . . . vii Page 35 36 37 38 39 4O 41 42 44 63 65 67 CHAPTER I INTRODUCTION Gibson and Levin (1975) have stated that readers, even though they may not have a specific purpose guiding their reading, retain a general set for understanding, an urge to "make sense." Certain linguistic structures used in written materials may not always facilitate the reader's need to understand. Studies have demonstrated that readers attend longer to those linguistic structures which seem to deter comprehension. These studies have attempted to examine the relationship between linguistic structures and visual processing as they affect reading efficiency. Many of these studies have used eye movement analysis to determine that relationship. We know that the eyes move continuously and that regressions occur. This knowledge leads us to believe as Kolers (Huey, 1968) states, that "what the reader understands from.what he has read is the result of a construction he makes and not the result of a simple transmission of the graphic symbols to his mind” (p. xvii). A reader's eye movements are a picture of his unique behavior in processing printed material. However, little has been done to study the develop- mental aspects of cognitive processing involving the rela- tionship between linguistic structures and visual processing. Cohen (1978) has stated a need for investigating both the possibility of a hierarchy of information-processing stages and how that hierarchy applies developmentally to different stages of reading maturity. This study will attempt to extend our knowledge of the developmental nature of cognitive processing. Purpose The purpose of this study is to examine the cognitive processing performance of competent middle school readers and to compare their performance to that of adult able readers. Specifically, the areas of examination are the effect of types of discourse in processing syntactic structures and the effect of left-and right-embeddings on cognitive processing performance. Background Psycholinguistic Theory Psycholinguists attempt to describe the mental processes involved in the production, comprehension, and acquisition of language. Syntax, the way words combine to form sentences, is one of the components of language com- prehension. Clark and Clark (1977) offer the following outline to describe the individual's strategy for process- ing syntax in order to arrive at meaning: 1. Take in raw speech sounds and retain in "working memory." 2. Organize the speech sounds into constituents. 3 3. Organize the constituents into propositions. 4. Retain the propositions in working memory and purge the memory of the original raw speech. In so doing, exact wording is forgotten while overall meaning is retained. Clark and Clark maintain that these steps probably occur simultaneously and regard this description of linguistic processing as more of a global process rather than a consciously sequential strategy. Psycholinguists also tell us that certain syntactic structures are more difficult to understand than others because of processing restraints which inhibit compre- hension and impede memory. Kimball (1973) proposed the "principle of fixed structure," in which the individual is erroneously committed to a particular "set for parsing a sentence. The principle attempts to explain why the following sentence is so difficult to process: The horse raced past the barn fell. He also formulated the "principle of two sentences” which demonstrates the difficulty in processing a sentence which is really more than one sentence and places a tre- mendous load on memory capacity: That that John departed bothered Mary surprised Max. In discussing processing impediments, psycholinguists have theorized that given information should precede new information in a sentence. As given information, the order 4 of preference seems to be: subject, objects, indirect objects, and noun phrases. Foss and Lynch (1969) demon- strated that left-embedded sentences make more demands on processing capacity than right-embedded sentences. Visual Processing and Syntactic Structures Researchers have used several procedures to demonstrate visual processing behaviors. Eye movement research has been a universally accepted procedure used to describe the visual processing of syntactic structures. Huey (1908) credits Javal with making initial discoveries in the field of eye movements. Tinker made significant contributions in this field by describing visual processing in terms of such visual activities as regressions, pause durations, and flexibility. Tinker (1965) reported that eye movement behavior stabilizes by fourth grade with only small changes occurring up to the high school level when only a few minor improvements may be noted. An earlier study by Gilbert (1953) had reached similar, if not such specific, conclusions. The eye-voice span, first reported by Quantz in 1897, is another procedure used to measure visual processing. Studies using both eye-voice span and eye movements have substantiated the psycholinguists' claim.that certain syntactic structures are mOre difficult to process than others. Levin and Kaplan (1970) and Levin, Grossman, 5 Kaplan, and Yang (1970) used eye-voice span techniques to show the relationship between visual processing and textual demands. Buswell (1920) noted that the eye-voice span was affected positively by age and reading ability and nega- tively by the difficulty of the reading material. Tinker (1965) found that eye movement behavior is considerably affected by the content and requirements of the prose. Bader, Pearce, and Thompson (1980) noted dif- ferences in eye movement processing in related and unrelated discourse. Developmental Aspects of Visual Processing of Syntactic Structures Most of the research concerning the visual processing of syntactic structures has been done using adult subjects, thus ignoring, for the most part, developmental implica- tions in this field of study. However, there are several researchers who have added a developmental component to their investigations. Levin and Turner (1968) involved sub- jects from second, fourth, sixth, eighth, and tenth grades and college undergraduates in order to demonstrate the effect of grammatical structure on eye-voice span. Tatham (1970) compared the performance of second and fourth grade children when reading sentences written on two levels of syntactic complexity. Using third, fourth, and fifth graders, Rode (1974) indicated that the type of syntactic unit (noun or verb phrase) influenced the eye- voice span of younger readers. Barnitz (1980) sought to determine the effect of pronoun-referent structures on the reading comprehension of children in grades two, four, and six. Comprehensive examination of research concerning visual processing of syntactic structures seems to indi- cate that fourth grade is a key maturation point at which time readers become able to take advantage of grammatical structures. Research Questions The following research questions were formulated to guide the investigation: 1. Is there a difference in the visual processing of left- and right-embedded structures in unrelated discourse by competent middle school readers? 2. Is there a difference in the visual processing of left- and right-embedded structures in related discourse by competent middle school readers? 3. Is there a difference between the visual process- ing of left- and right-embedded structures in unrelated discourse and the visual processing of left- and right-embedded structures in related discourse by competent middle school readers? 4. Is there a difference between the visual process- ing behavior of competent middle school readers and the visual processing behavior of competent adult readers? These research questions were expanded and restated in null hypothesis form for statistical testing. Definition of Terms Terms in the study were used according to the follow- ing definitions: 1. Competent middle school readers Competent middle school readers are defined as sixth-grade students selected on the basis of teacher judgment and the attainment of a score falling between the 50th and 75th percentile on the reading comprehension section of the Comprehensive Test of Basic Skills (Level 2, Form S). Competent adult readers Competent adult readers refers to graduate students in the Elementary and Special Education Department of a large Midwestern university.- Syntactic Structures Syntactic structures refer to relative clauses following and modifying either the sentence subject or the sentence object. Right-embedded sentence A right-embedded sentence is a sentence contain— ing a relative clause that follows and modifies the sentence object. EX. The policeman stopped the car that was speeding. Left-embedded sentence A left-embedded sentence is a sentence containing a relative clause that follows and modifies the sentence subject. EX. The girls who sat in front fell asleep. Unrelated discourse Unrelated discourse refers to three sentences of differing subject matter. Related discourse Related discourse refers to sentences cohesive with regard to semantic content. Organization of Subsequent Chapters Chapter II will contain a review of pertinent literature in the following areas: 1. The relationship between reading comprehension and syntactic structures. The relationship between visual processing behavior and reading comprehension. The relationship between visual processing behavior and syntactic structures. The developmental nature of syntactic processing. Chapter III will present a description of the materials and procedures employed in this study. The design of the study will be presented. Chapter IV will report the results of the data collected, compared, and analyzed for this study. Chapter V will include a summary of the investigation, appropriate conclusions, implications, and recommendations for further research. CHAPTER II REVIEW OF RELATED LITERATURE Introduction The review of related literature is organized under four major headings. These are: l. The relationship between reading comprehension and syntactic structures. 2. The relationship between visual processing behavior and reading comprehension. 3. The relationship between visual processing behavior and syntactic structures. 4. The developmental nature of syntactic processing. The findings of the research in these areas serve as the basis for developing the hypotheses of this study. The Relationship Between Reading Comprehension and Syntactic Structures Thorndike (1917) stated in his landmark study that: Reading is a very elaborate procedure, involving a weighing of each of many elements in a sentence, their organization in the proper relations one to another, the selection of certain of their connota- tions and the rejection of others, and the cooperation of many forces to determine final response. (p. 323) He noted the difficulty students encountered in responding to questions Which tested their literal under- standing of long, involved sentences. Thorndike attempted 10 to analyze the responses and hypothesize on the reasoning processes that produced those responses. He theorized that reading comprehension necessitated an active, involved participation on the part of the reader. Almost fifty years later, Chomsky (1965) reinforced Thorndike's notion concerning the complicated interaction involved in language processing when he developed his theory of transformational-generative grammar. This theory posits that every sentence can be represented on a surface level and a deep structure level. A reader must not only be able to decode words and understand their individual meanings, but must also be aware of the grammatical inter- relationships in order to retrieve the underlying meaning of the sentence. Smith (1971) emphasized the importance of grammar as the link between the surface and deep structure levels of language, although, as he says, ”There is no simple correspondence between surface structure of language and meaning." (p. 29) Marcus (1971) noted that a reader may find print material ambiguous if there is a discrepancy in the relationship involving decoding words, knowing the mean- ings of the words as dictated by context, and understanding the mechanics of grammatical structures. Ike attempted to discover the relationship between syntactic com- plexity and comprehension facility by developing a diagnostic tool to measure comprehension of syntactic structures for intermediate grade students. He reported 11 that prepositional phrase modifiers and sentences which contained relative clauses in the subject-verb-object pattern of independent clauses were among the most difficult to comprehend. A hierarchy of difficulty was established concerning specific prepositions and semantic groups of prepositions by Foust (1973). Those prepositions denoting temporal and abstract relationships were identified as the most difficult, followed by those prepositions denoting directional and positional relationships. Adding to the information concerning levels of syntactic difficulty, Fagan (1971) investigated the effect that the number and types of transformations had on reading compre— hension. He found that appositives, participles, gerunds, and genitive embeddings were the most difficult. He also reported that complex transformations rather than numerous transformations adversely affected reading comprehension. More supportive data were obtained by Stoodt (1970), who found significant correlations between reading comprehension and comprehension of various connectives. Richek (1974) studied children's comprehension of three anaphoric forms (noun, pronoun, and null) and the effects of contextual variations on these forms. She found signifi- cant differences among the three forms with comprehension dropping to 60% in the null form. Consistent with Fagan's conclusions was the suggestion that the complexity of a sentence affects comprehension of the structures in that 12 sentence, but Richek added that efficient methods of measuring complexity had not yet been defined. However, Fodor, Garrett, and Bever (1968) identified the main verb as a determiner of complexity in a sentence, since verbs limit the syntactic choices with which they can occur. Kolers (1970) and Miller and Coleman (1967) reported on the relationship between grammar and context. Kolers found that readers make fewer errors over the final three- fifths of sentences, and Miller and Coleman found more correct answers at the ends of sentences using a cloze procedure. They concluded in each case that the greater amount of syntactic data facilitated the readers' quest for understanding. Using a technique that presented a sentence blurred in varying degrees, Sawyer (1971) reported that readers were able to recognize the existence of right-embeddings under poorer blur conditions than left-embeddings. Also found easier to recognize under more blurred conditions were "by + agent" phrases in passive sentences compared to simi- lar prepositional phrases in active sentences. "By" phrases introducing locatives were more easily recognized than "by" phrases introducing agents in simple passive sentences. Difficulties with reading comprehension can be attri- buted to the inability to organize textual material into meaningful units. In order to study this phenomenon, Cromer (1970) chose two groups of junior college readers 13 who were poor in reading comprehension, either because of vocabulary deficits or because of word—by-word reading. These groups with a group of good readers were presented stories in four different modes: sentences, single words, phrases, and fragmented groupings. The word-by-word readers comprehended better in the phrase mode than they did in any other mode, performing just as well as the good readers in that mode. Remediation based on chunking information into constituents may prove to benefit the reading comprehension of word-by-word readers. In summary, the reader must integrate the phonological, lexical and syntactic information in the efficient pursuit of comprehending the deep structure of print material. How» ever, certain syntactic structures appear to be more diffi- cult to process than others. Therefore, it seems reasonable to continue to investigate syntactic structures and their specific effect on language processing in order to facili- tate the reader's role as comprehender. The Relationship Between Visual Processing Behavior andIReading,Comprehension In his book, Huey (1908) cites Javal's studies as providing the critical discoveries that undergird eye movement research. Javal noted that eye movements were discontinuous, and that reading, or the specific seeing of words and letters, occurred during the pauses of the eye movements. He also concluded that there was a pause every ten letters and that this was about the amount of input 14 that could be seen clearly in one fixation. He observed that the upper half of the line was more important and con- cluded that the fixation point of the eye moves along between the middle and top of the small letters. More recently, Tinker (1958) completed a compre- hensive review of the literature on eye movements, as well as conducting research in the field. In one study, Tinker (1965) compared prose of varying levels of difficulty and reported that pause duration in eye movements was directly related to difficulty of the text; that is, both the content and the requirements of the prose. Difficulty of the text, attributed to the readers' particular vocabulary deficiencies and comprehension difficulties, affected the number of fixations as reported by Buswell, 1920; Judd and Buswell, 1922; and Woodworth, 1938. Purpose for reading can also affect visual processing. A developmental study by Levin and Cohn (1968) used students from second, fourth, ninth, and eleventh grades to determine if the eye-voice span changed according to each of three stated conditions--normal reading for pleasure, careful reading in preparation for questioning, and fast reading to search into the main idea. As has been proven to be the case in other studies, the older children had longer eye-voice spans. The eye-voice span was shortest for careful reading, and longest for quick skimming, with normal reading falling in between the two. The study demonstrates that students as early as second grade can 15 understand that there are different purposes for reading and can make accommodations according to those purposes. Again measuring the eye-voice span, Levin and Kaplan (1970) concluded that the eye-voice span was affected by the Structure and contextual predictability of written material by showing that the eye-voice span is longer for a sentence than for a random string of words. The performance of good readers tends to demonstrate an adaptability to the nature of the material. walker (1938) noted that good readers differed from less-skilled readers in being more flexible in their response to the nature of the material. The pattern of the skilled readers regarding fixations and pauses reflected the comprehension demands of the material read. In contrast, poor readers tended to use the same eye movements regardless of the textual demands. Heiner and Henderson (1974) reported similar findings in a study which examined the eye movement behavior of skilled and less skilled seventh grade readers using material containing cloze deletions and non-treated materials. The good readers increased their fixations and regressions on the cloze materials, while the poor readers made fewer fixations and regressions on the cloze materials. The researchers speculated that the poor readers did not adjust their visual processing strategies to accommodate the demands of the cloze material, but merely intensified their efforts by extending the duration of their fixations. 16 Different findings were reported by Goltz (1975) using skilled and less-skilled readers at the college level. He indicated that the two groups adjusted their reading to accommodate different purposes, but that the nature of the accommodation was different. Skilled readers appeared to be more efficient in that their "regular reading" produced shorter eye fixations and larger right-going saccades. Their slower reading indicated longer duration of fixations rather than increased fixations. Skilled readers used regression to augment what they had read at a slower speed, while less-skilled readers used regression to compensate for what they had missed at an accelerated pace. Mackworth (1972) also compared poor and skilled readers in an experiment calledthe "missing word task.” The children, from second, fourth, and sixth grades, were asked to choose a word from a list of nouns and a list of verbs to fill in the blank in a sentence. The eye movements of the subjects were recorded, and the duration of time spent looking at the wrong list was computed. Good readers spent half as much time looking at the wrong category as the poor readers. Both groups made significant improvement between second and fourth grade, and while the good readers continued to improve, the poor readers made very little change between fourth and sixth grades. Hochberg, Levin, and Frail, in an unpublished manu- script cited in Gibson and Levin (1975), probed the effect of peripheral vision on reader performance when interword 17 spaces in the reading text were filled with letterlike, 'meaningless symbols. Fifth graders were slowed in their reading more than second graders. This was interpreted to mean that younger children were reading'word-bydword, while the older children were accustomed to using interword spaces, picked up peripherally, to move their eyes across units larger than words. This study indirectly tested Hochberg's (1970) earlier findings that peripheral vision is an important aid in deciding where fixations will occur. Rayner (1974) investigated the areas of the periphery that provide the reader with different kinds of useful information. His ingenious computer technique monitored the fixation patterns of adult readers in situations in which a critical word was manipulated in the text. Rayner's data indicated that information about word shapes and about initial and terminal letters is picked up begin- ning about seven to twelve character spaces in the periphery, but that recognizing the meaning of a word begins only about one to six character spaces to the right of the fixation point. Rayner's research, plus previous research by McConkie and Rayner (1973), demonstrated the complex processing that occurs during eye movements. Other findings gleaned from McConkie and Rayner's research were that word length affects reading farther into the periphery than does word shape or specific letter identification, and that blank spaces and beginnings of sentences are rarely fixated. On the basis 18 of their research, one can conclude that what is "seen" includes both the foveal region and the periphery working together to provide the reader with input to decide on the location of the next fixation. In the area of regressive eye movements, Stern (1978) reported that a fixation occurring just before a regression is shorter than normal. He reasoned that as the reader becomes aware that he is not making sense out of the material, he quickly stops processing at that point. In summary, eye movement research indicates that skilled readers' eye movements are flexible and adaptive, accommodating to the nature of the material and to the nature of the task. Peripheral vision also seems to be a factor in efficient reading, although more needs to be known about its specific function. The Relationship Between Visual Processing Behavior and Syntactic Structures Kolers (1970) and Weber (1970), working with mature readers and beginning readers, respectively, have observed that the reader makes predictions about what will follow as he reads. These predictions are based on what the reader has internalized about how his language is structured. Are there linguistic structures that are more predictable and thus easier to comprehend and to process than others? Levin, Grossman, Kaplan, and Yang (1972) found that the left-embedded sentence type significantly limited the amount of information the reader could process as measured 19 by the eye-voice span. Wanat (1971) substantiated this finding in his study of the eye movement patterns of mature readers when reading isolated sentences containing left- and right-embeddings. Although the number of forward fixations were the same for both types of sentences, left- embedded forms required more total time to process. Bader, Pearce, and Thompson (1980) reported similar findings concerning forward fixations in the unrelated sentence condition. However, their study also compared left- and right-embeddings in both unrelated sentences and connected discourse. While they found that sentences were processed more slowly than paragraphs, the left- embedded structure was processed more slowly than the right-embedded structure in both contextual conditions. They did not find a significant difference in forward fixations in the unrelated sentence condition, but they reported a difference in number and duration of regressions in the unrelated left-embedded sentence condition and the left-embedded, connected discourse condition. Carpenter and Just (1977) examined pronoun referents to determine if regressive eye movements are selective and indicative of integrative processing. Sentences were constructed to provide the reader with strong linguistic cueing to the pronoun referrant; other sentences did not. Results indicated that this linguistic cueing facilitated identification of the pronoun referrant as evidenced by 20 regressive eye fixations. The researchers also posited that regressive movements may not necessarily indicate inefficient reading habits, but may reflect interpretive processing behavior. Taylor (1960) observed similar eye movement behavior in mature readers, concluding that these readers engage in an active search for understanding when they are puzzled by different or obscure syntax and vocabulary. This search is characterized by regressive eye movements and subsequent sweeps. Mehler, Bever, and Carey (1967) observed the visual processing behavior of college students using three kinds of ambiguous statements. They formulated the rule that the reader fixates on the first half of each constituent. How- ever, serious criticism concerning the validity of the study was raised, although some researchers have credited the study with pointing out the need for more investigation in the area of ambiguity. Wanat and Levin (1968) were also concerned with structures that may result in ambiguous processing. They measured the eye-voice span of undergraduate students to study their processing behavior of passive sentences which contained either an agent-present relationship or an agent- deleted relationship. The eye-voice span was longer for the agent-included sentences. It was hypothesized that the agent-included sentence was easier to process because there was a direct relationship between the surface structure 21 and the deep structure, thus eliminating ambiguous interpretation. In summary, studies have analyzed the effects of different syntactic structures on the cognitive processing of written materials as measured by eye-movement behavior. While more studies of this type are needed, certain factors should be considered in conducting this kind of research. It is interesting that one of wanat's (1970) criticisms of the previously mentioned Mehler, Bever, and Carey study (1967) concerned their use of ambiguous sentences, arguing that these forms are atypical of a normal reading situatiOn. This position could be carried a step further, arguing that reading unrelated sentences is also atypical. However, few studies in this area of linguistic structures examine read- ing behavior using connected discourse. There appears to be a need for studies to approximate natural reading behavior by using connected discourse as the vehicle for probing cognitive processing of syntactic structures. The Developmental Nature of Syntactic Processing Huey (1908) agrees that, depending on our reading purpose, we all at some time or another read by letters, word, or phrases, using whatever method will facilitate the extraction of meaning. However, in order not to mis- lead us, he adds that, "we see, too, that the reader's acquirement of ease and power in reading comes through increasing ability to read in larger units." (p. 116) 22 Just what do we know about how the reader acquires the ability to process larger units of syntactic structure 'more efficiently? Chomsky's study (1969) revealed that children were still engaged in active syntactic acquisition up to age 9 and perhaps even beyond. It was also noted that while children vary in their rate of acquisiton, they do not vary in the order of acquisition. Developmental studies done by Levin and Turner (1966) indicated that effective use of syntactic structures increases with age and skill, which is interpreted to mean that the skilled reader is sensitive to the syntactic regularities of language and uses higher order units to acquire meaning. Gibson and Guinet (1970) found that children read inflectional endings with fewer errors than non-inflectional endings on words of equivalent length. This tendency increased from third to fifth grade. Levin and Turner (1968) hypothesized that the phrase was the unit of decoding and attempted to test it develop- mentally by measuring the eye-voice span. They asked college undergraduates and students in second, fourth, sixth, eighth, and tenth grades to read phrase-embedded sentences in both active and passive voice, as well as unstructured word lists. Findings revealed that the eye- voice span increased with age and was significantly longer for sentences than for the unstructured word lists for all ages and sentence types. An important finding of the study 23 was that all of the age groups, except second graders, ended their eye-voice span at phrase boundaries. Readers as early as fourth grade chunked their oral reading into phrase units, but this tendency did not appear to grow stronger between fourth grade and adulthood. However, Resnick (1970) found this tendency to be stronger in adults, although also present in children. Levin and Turner's (1968) findings are consistent with Tinker's (1965) research which described eye movement patterns in reading as becom- ing adultlike at the fourth grade level. Steiner, Weiner, and Cromer (1971) further corrobo— rated the previous developmental findings when they examined the effect of a type of comprehension training on good and poor readers in fifth grade. When given a word-by-word presentation of a story, good readers made anticipation errors that were syntactically and semantically sound and often identified whole phrases rather than single words. This would suggest that the fifth graders were capable of structuring their reading into phrase units. This study supports a body of evidence demonstrating that the phrase is the major decoding unit for skilled readers. Rode (1974) studied the effects of different syntactic structures on the eye-voice span, using subjects in the third, fourth, and fifth grades. Consistent with other studies, the eye-voice span increased with age. The data also indicated that the type of phrase, noun, or verb, had a powerful effect on the eye-voice span of the subjects at 24 all age levels with the span expanding in noun phrases and constricting in verb phrases. The eye-voice span was also found to terminate at verb phrases significantly more often than at noun phrases. The last finding raises the possi- bility that the clause is the syntactic unit that readers use for decoding rather than the phrase. Rode also specu- lated that short-term.memory development may be a determin- ing factor in explaining why the third graders in her study were able to chunk words into phrase units but not clause units. The fourth grade subjects began to chunk clause units together, and Rode hypothesized that an increase in short-term memory enabled them to make more efficient use of language ability. Siler (1973) compared the effects of syntactic and semantic violations on the oral reading performance of second and fourth graders. His findings, similar for both grade levels, indicated that the category of syntactic violations and the category of combined syntactic and semantic violations were more disruptive than the category of semantic violations, but no more disruptive than each separately. Syntax appeared to have a greater disruptive effect than semantics in oral reading performance. On the other hand, syntactic complexity seems to be an aid to comprehension and recall, rather than a hindrance, according to a study done by Pearson (1974) using third and fourth grade readers. Students preferred cohesive, more heavily embedded forms to less cohesive, less heavily embedded forms. The finding of the aided recall section 25 of the experiment led to the conclusion that a causal relation cannot help but be stored in a unified, sub- ordinated chunk. Children appeared to store causal rela- tions in long-term memory in a cued form and retrieve it in that form (which may have been more natural and familiar to them), regardless if the initial input was cued or uncued. It may be proposed that the children in Pearson's study responded as they did because the forms they chose were more familiar to them. Hypothesizing that compre- hension is aided by familiar syntactic structures, Tatham (1970) prepared sentences on two levels of syntactic com- plexity and had second and fourth graders read them. Comprehension was better on those selections which coincided syntactically with frequent patterns of oral language. Howe ever, as the reader matures, comprehension must expand to include written styles that are decidedly different from spoken patterns. Barnitz (1980) used children in grades two, four, and six to determine the development in comprehension of selected pronoun-referent structures. Three linguistic comparisons were made along the dimensions of referent type, referent order, and referent distance. Analysis of the data demonstrated that: (l) noun phrase pronomials were easier to comprehend than sentential pronomials; (2) structures with forward reference were easier to come prehend than structures with backward reference; and 26 (3) there was no difference between intra-sentential and inter-sentential structures. A.hierarchy of difficulty for each grade was also constructed based on these findings. Barnitz concluded that except for a few esoteric structures, pronoun-referent structures are generally comprehensible by the time children reach sixth grade. Sheldon (1974) studied the developmental aspects of language processing when she compared the strategies used by adults with those used by four-and five-year-olds when processing relative clauses. The clauses were inserted either after the subject or after the object, and related either to the subject or to the object of the sentence. She found that children and adults use the same processing strategies but differ in the emphasis placed on the parti- cular strategies, which explained the particular types of sentences on which both groups made the most errors. Adults seemed to over-rely on the Adjacency Strategy, which caused them to make more errors in those sentences in which the relative clause directly followed the subject, and in those sentences in which the relative clause fol- lowed the object making the object of the sentence the object of the clause. Children, however, relied on the Parallel Function Strategy, and so made more errors in those sentences in which the noun phrases and the relative clauses did not have similar functions. The children also over-relied on 27 the Extraposition Strategy when attempting to find the antecedent to the relative pronoun in object relatives. In summary, readers as early as third grade have demonstrated an ability to use syntactic structures to process reading material effectively. This processing ability increases with age and skill to include more demanding syntactic understanding. Reading material is processed into meaningful constituents, evidence having been presented for both the phrase and the clause as the main unit of decoding. Younger readers also seem to respond better to familiar syntactic structures in the early stages of reading. We also have a large body of information con- cerning the processing of anaphoric forms. However, comparatively little can be said concerning significant developmental stages in processing syntactic structures. There have been relatively few studies that have attempted such an investigation. More studies need to consider adding a developmental component to the design. As it is, it is difficult to synthesize these individual studies, each dealing with a particular aspect of syntactic processing and a particular age group, and to weave a coherent develoPmental thread through them. CHAPTER III. METHODOLOGY OF THE STUDY Introduction The chapter describes the methodology employed in conducting the study. The data collection procedures were similar to those used in the Bader, Pearce, and Thompson (1980) study concerning the effect of related and unrelated discourse on the processing of syntactic structures. Their results, obtained with adult competent readers, will be compared to the results obtained with competent middle school readers. Population The population was composed of thirty middle school students in the sixth grade who were randomly selected from a group of competent readers. The competent readers were identified on the basis of teacher judgment, the attainment of a score falling between the 50th and 75th percentile on the reading comprehension section of the comprehensive Test of Basic Skills, Level 2, Form S, and parental consent. The test scores were based on results obtained during spring of the 1978-79 school year. All participants were native English speakers. 28 29 Materials Materials developed by Bader, Pearce, and Thompson (1980) were used. These consisted of sentences using active voice and employing no dependent clauses with the exception of the target embedding. These were either left-embedded sentences that contained a relative clause that followed and modified the subject or right-embedded sentences that contained a relative clause that followed and modified the sentence object. The unrelated sentence condition consisted of three sentences of equal length (11 words) dealing with differing subject matter, while the related sentence condition consisted of paragraphs extend- ing over SeVen.lines that were cohesive with regard to semantic content. The target sentence in the unrelated discourse condition was the second sentence, while the target sentence in the related discourse condition appeared on the sixth line. Apparatus The conditions were typed on IBM pica, single spaced on 3%" x 5" cards and presented with the EDL/Biometrics Reading Eye II, an electronic instrument that employs a photoelectric method to record eye movements on heat- sensitive graph paper. Data Collection The subjects were tested individually during the school day in a small room designated for this purpose. 30 Initially students were to be tested after school, but because of the short time (approximately five to ten minutes) needed to complete the procedure, a change was ‘made. The principal agreed to allow the subjects to be tested during the school day, provided that they were taken from class during times that were not crucial, such as study periods or free reading times. The subjects were given a prepared set of instructions informing them.of the general operation of the Reading Eye II. They were told that they would be asked to read several different selections during the recording process. They were instructed to read naturally and to pay attention to the material on the card so that they might be able to paraphrase the selection afterwards. The subjects were not asked to paraphrase; the instructions were used to focus the attention of the subjects on the task. After the subjects' eyes were aligned properly, they were instructed to close their eyes between selections and to keep their heads motionless throughout the experiment. The experiment included five reading selections: an EDL paragraph, a left-embedded sentence in isolation, a right- embedded sentence in isolation, a left-embedded sentence in related discourse and a right-embedded sentence in related discourse. At the conclusion of the experiment, subjects were asked if they encountered any reading difficulties or if they reread any portion of the five selections. Their responses were recorded. 31 Design A 2 x 2 (context & structure) design was used to test the effect of left and right embeddings in conditions of unrelated sentences and connected discourse. Within each of the context conditions, subjects were assigned on a rotating basis two of eight constructions containing left- embeddings and two of eight constructions containing right- embeddings. No subject received left-embedded and right- embedded sentences from the same pair. Hypotheses 1. There will be no difference in the visual processing of left- and right-embedded structures in unrelated discourse by competent middle school readers. 2. There will be no difference in the visual processing of left- and right-embedded structures in related discourse by competent middle school readers. 3. There will be no difference in the visual processing of left-embedded structures in unrelated discourse and related discourse by competent middle school readers. 4. There will be no difference in the visual processing of right-embedded structures in unrelated discourse and related discourse by competent middle school readers. 5. There will be no difference in the visual processing of left- and right-embedded structures between competent middle school readers and competent adult readers within the unrelated discourse condition. 6. There will be no difference in the visual process- ing of left- and right-embedded structures between competent middle school readers and competent adult readers within the related discourse condition. 7. There will be no difference in the visual processing of left-embedded structures between competent middle school readers and competent adult readers within the comparison between the unrelated and the related discourse condition. 32 8. There will be no difference in the visual processing of right-embedded structures between competent middle school readers and competent adult readers within the comparison between the unrelated and the related discourse condition. Data Analysis A matched pairs t-test was used to investigate the interaction between the visual processing of syntactic structures in unrelated and related discourse. A t-test for independent variables was used to compare the visual processing behaviors of competent middle school readers and competent adult readers. The independent variables are the unrelated and related discourse conditions, the left-embedded and right- embedded syntactic structures, and the maturity levels of the subjects. The dependent variables are the visual processing behaviors which include: gaze duration, forward fixation duration, regression durations, number of forward fixations, number of regressions and total eye movements. Summary The chapter described the methods and procedures used in the study. The eye movements of competent sixth-grade readers, who were good comprehenders, were recorded, while the subjects read sentences with left- and right-embeddings in unrelated and related sentence conditions. Their performance was to be compared to that of adult competent readers. 33 The eye movements were photographed with the EDL/ Biometrics Reading Eye II, and the materials used in the machine were developed by Bader, Pearce, and Thompson (1979). The statistical procedures were designed in association with the research consultants at Michigan State University. In Chapter IV, the data are presented, analyzed, and organized. CHAPTER IV PRESENTATION AND ANALYSIS OF DATA Introduction The purpose of the study was to obtain and analyze data concerning the visual processing of syntactic structures in related and unrelated discourse by competent middle school readers and to compare their performance to that of adult able readers. The areas of examination focus on the effect of length of discourse in processing syntactic structures and the effect of left-and right-embeddings on cognitive processing performance. The methodology for the collection and treatment of data was described in the previous chapter. The chapter will present the statistical analysis of the findings as they relate to the hypotheses constructed for the study. Hypotheses and Statistical Tests The data concerning the visual processing behaviors of competent middle school readers were tested using a matched pairs t-test. A t-test for independent variables was used to test the data comparing the visual processing behaviors of competent middle school readers and competent adult readers. 34 35 Hypothesis 1. Ho 1: There will be no difference in the visual processing of left- and right-embedded structures in unrelated discourse by compe- tent middle school readers. The hypothesis was tested with a matched pairs t-test. (See Table 4.1). Table 4.1.--t-Test: Left- and Right-Weddings in Unrelated Discourse Varflflfles bl SD t LE,ES‘mx1RE,EB Duration.of gaze .14 1.02 .73 Duration of ferward fixation -.04 .70 -.32 Duration of regression .07 .60 .62 Forward.fixations -.13 3.39 -.22 Regressions .70 2.55 1.50 Total movements (R and F) .57 4.97 .62 *p < .05 Key-IE lxitqmtedfimg RE Eughtrmbaifing SS Sentence (unrelated discourse) 77- Paragrwh (related discourse) There were no significant differences in the visual processing of left- and right-embedded structures in unre- lated discourse by competent middle school readers. Hypothesis 1 was accepted. 36 Hypothesis 2. Ho 2: There will be no difference in the visual processing of left- and right-embedded structures in related discourse by competent middle school readers. The hypothesis was tested with a matched pairs t-test. (See Table 4.2). Table 4.2.--t -Test: Left- and Right-Fabeddings in Related Discourse vanuflfles 14 SD t LE,WV ariRE Zr Duration.of gaze .28 1.12 1.38 ‘Duration of forward fixations .16 .52 1.59 Duration of regression .03 .65 .23 Fonmufl.fihmuions .00 2AM} .00 Regressions -.07 3.10 -.12 Tonalunwamrms GiandID -.L3 £100 -.15 *p < .05 There were no significant differences in the visual processing of left- and right-embedded structures in related discourse by competent middle school readers. Hypothesis 2 was accepted. 37 Hypothesis 3. Ho 3: There will be no difference in the visual processing of left-embedded structures in unrelated discourse and related discourse by competent middle school readers. The hypothesis was tested with a matched pairs t-test. (See Table 4.3) Table 4. 3.--t-Test: left-weddings in Unrelated and Related Discourse venuufles M SD t IE,E§;amdlE,Zr Duration of gaze .31 1.00 1.69 Duration of forward fixations .01 .60 .12 Duration of regression .07 .66 .59 Forward fixations .50 2.83 .97 Regressions .67 3.65 1.00 Total movarents (R and F) 1.10 5.71 1.06 *p‘<.05 There were no significant differences in the visual processing of left-embedded structures in unrelated and related discourse by competent middle school readers. Hypothesis 3 was accepted. 38 Hypothesis 4. Ho 4: There will be no difference in the visual processing of right-embedded structures in unrelated and related discourse by compe- tent middle school readers. The hypothesis was tested with a matched pairs t-test. (See Table 4.4) Table 4.4.--t-Test: Right Embeddings in'Unrelated,and.Related.Discourse \kmiflfles M SD t RE,E§lam1RE,7T' Duration of gaze .11 .83 .73 Duration of fbrward.fixations .11 .67 .87 Duration of regression .03 .58 .24 Fbrward fixations -.63 2.99 -1.16 Regressions ~ -.03 2.43 - .08 Total movarents (R and F) -.67 4.50 - .81 *p‘(.05 There were no significant differences in the visual processing of right-embedded structures in unrelated and related discourse by competent middle school readers. Hypothesis 4 was accepted. 39 Hypothesis 5. Ho 5: There will be no difference in the visual processing of left- and right-embedded structures between competent middle school readers and competent adult readers within the unrelated discourse condition. The hypothesis was tested with a t-test for independent variables. (See Table 4.5) Table 4.5.--t-Test: Student and Adult Processing of Left- and lfightJhbeifingsinlkuehuedIfismmnse vacufifles t LE,EB,mleE,SS Duration of gaze -2.01 * Duration of forward fixations - .77 Duration of regression -2.10 * Forward fixations - .83 Regressions -2.15 * Total movements (R and F) -l.66 *p < .05 The data indicated that there were significant differ- ences between the middle school readers and the adult readers concerning the visual processing of left- and right- embedded structures in the unrelated discourse condition. These differences occurred in the areas of duration of gaze, t = —2.01, p‘<.05, duration of regressions, 40 t = -2.10, p < .05, and number of regressions, t = -2.15, p OB HMUOH oo.H mo.m no. mGOHmmouwom no. mm.~ om. mcoaumxflm oumsuom mm. we. no. mommmohwou mo coaumusn Na. or. AC. cowumxwm ouwzpom mo Gowumunn mo.H oo.H Hm. oumw wo coaumusn 1R .5 26 mm :5 ma. 1 oo.m mH.u mucoao>oe HmDOH NH. 1 oa.m no.1 . wooemmonwom oo. w¢.N oo. mcowumxwm oumzhom mN. no. mo. conmopwou mo cowumusn mm.H on. ma. sowumxwm ohmshom mo nowumnsa mm.H NH.H mu. onmw mo coaumusn )R .E was 1R :3 No. 3...» 3. a. a 5 35888 H38. om. mm.~ on. mcowmmonwom mm. : mm.m mH.a mcowumxwm ohmsuoh No. on. no. soammoumoh mo cowumuoa mm. 1 on. so. Gowumxwm oumshom mo cowumupn mm. No.H «H. onmw mo cowumuun mm .mm mam mm .mA «u am 2 . moanmwuo> muoomom Hoonom oaoowz usouomaou ho omusoomwn ooumaoa mam noumHoucD cw mounuoohum oooooosm:unwwm nan numoq mo wnwmmoooum Hmdmw> "umoenunu.an< manna 64 mo:. V.m ”umou ooafimuuoBu Mom 1 ouoz weneemaam-unwnm mm waneemaam-ummn mu Goauwocoo nmmuwmumm .kw coauwocoo oocouoomuoopse mm 1 mom am. 1 om.¢ 50.: Am a mv muaoao>oa amuoa mo. 1 m¢.N no.1 macammoumom 0H.Hu mm.~ no.1 maowumxwm mumshoh «N. mm. mo. sowmwouwou mo soaumusa mm. mm. Ha. cowumxwm oumzhom mo nofiumuaa mm. mm. Ha. oumw mo nowumuna Lw.mm pom mm .mm Ru am 2 moanowum> Aemsaauaouo H-< magma 65 EVEN mmnm Ana 3 a 5 35868 H38. ma mm H om maoammouwom 1.. no. ma . N om . H 303me oumgom ma. mm. 00. nowmmouwon mo acaumnso RH¢.N om. mm. Gowumxam oumshom mo coeumusn rom.~ am. He. onmw mo doaumusn Lu .5 28 mm .3 + m¢.m sm.~ w.a Am a my mucoao>oa Hmuoa + mq.m mo.~ mm.H maoflmmouwom om.H mo.H we. moowuoxwm mumsuom # mH.¢ «m. cm. Gowmmouwou mo compouna NN.H am. NH. cowumxwm oumshom mo cowumuaa % qw.¢ on. ma. onw mo Gowumhao )Rsa was karma L.¢¢.m mm.m m.~ Am a mv mucoao>oa Hmuoa L.mH.m Nw.N mo.H mcowmmouwom mo.a mm. mm. maowuoxfim ohmshom H Hm.m am. mm. commmouwon mo nofiumunn mm. as. no. coaumxwm ohmshom mo aowunusn A ma.¢ ow. ow. our» mo cowumunn mm .mm mam mm .mq «u am 2 moanmwnm> muoomom pano< unouomaoo mo omunoomwn noumHom poo ooumaoucs cw monouoauum oopooofimnuzwwm com lumen mo wcwmmoooum Hmsmfl> "umoaauun.~u< canoe 66 .owma .commsonh pom .oonmom .Hoomm Eoum moummo< .80. v R .8. v 3. .mo..V or ”umou moafimuuo3u Mom : ouoz waneeooam-uawnm mm wcnaemnam-uumu mu Gowuwoaoo smuwmumm .kw cowuwocoo mucoucomnoounfi mm 1 mom mm.a om.~ o.H muco8o>oa.aouoa no. qm.N oo.~ maoammouwom +nm.~ mo.~ mH.H maowumxwm oposhom mo.Hn am. no. u Goammouwon mo defluounn Hum.m mm. mm. coaumxflm oumznom mo Gowumuaa smm.m no. om. «New mo coaumusn LR.mM pom mm .mm so nm E moanmwum> Aemsawucoov N-< magma 67 Table.Ar3.--t-Test: Comparison of Visual Processing of Left- and RhgukEmuxwedERIMCUErm hathrdkuedemrlRekued InsomuselnrConrflentbfiddh:SdmxflfiRafibrsamd CammtauzAduthemkms Variables t* LE, SS and RE, SS Duration of gaze -2.0l* Duration of forward fixation - .77 Duration of regression -2.10* Forward fixations - .83 Regressions -2.15* Total movements (R & F) -1.66 LE,7T and RE,’/7" Duration of gaze -3.36T’ Duration of forward fixation -2.29* Duration of regression -2.15* Forward fixations - .86 Regressions -l.86 Total movements (R & F) -1.59 LE, 33 and LE,77' Duration of gaze - .47 Duration of forward fixation -l.52 Duration of regression .48 Forward fixations «1.08 Regressions .22 Total movements (R & F) - .49 RE, 83 and RE,77’ Duration of gaze -2.15* Duration of forward fixation -2.77f Duration of regression .35 Forward fixations .75 Regressions .32 Total movements (R & F) - .34 Key - SS Three-sentence condition zr’ Paragraph Condition LE Left-Embedding RE Right-Embedding Note - For two-tailed test: *p <_.05, p 4;.01 BIBLIOGRAPHY BIBLIOGRAPHY Bader, L. A., Pearce, D. L., and Thompson, D. R. Effect of discourse on processing of left- and right-embedded syntactic structures. Perceptual and Motor Skills, 1980, 59, 883-889. Barnitz, John G. Syntactic effects on the reading compre- hension of pronoun-referent structures by children in grades two, four, and six. Reading Research Quarterly, 1980, 15, 268-289. Buswell, G. T. How adults read. Supplementarnyeading Monographs, 1937, (Whole no, 45). Carpenter, P., and Just, MgA. Reading comprehension as eyes see it. In M. A. Just and P. A. Carpenter (Eds.), Cognitive Processes in Comprehension, Hillsdale, N.J.: Lawrence Erlbaum Associates, 1977. Chomsky, Carol. The acquisiton of syntax in children from five to ten. Cambridge: MIT Press, 1969. Chomsky, Noam. Aspects of a theory of syntax. Cambridge: MIT Press, 1965} Clark, Herbert H., and Clark, Eve V. Psychology and language. New York: Harcourt, 1977. Cohen, Karen. Eye activity in the study of the reading process. In Frank Murray (Ed.) Models of efficient readin . Newark, Delaware: International Reading Association, 1978. Cromer, W. The difference model: A new explanation for some reading difficulties. Journal of Educational Psychology, 1970, Q1, 471-483. Fagan, William T. Transformations and Comprehension. The Reading Teacher, November, 1971, 25, 169-172. Foss, D. J., and Lynch, R. H. Decision processes during sentence comprehension: Effects of surface structure on decision times. Perception and Psychophysics, 1969, 5, 145-148. 68 69 Foust, C. The relationship between understanding preposi- tions and reading comprehension (Doctoral dissertation, Ohio State University, 1973). Dissertation Abstracts International, 1974, 34, 4983A. (University Micro- films No. 74-03166) Gibson, E. J., and Guinet, L. The perception of inflections in brief visual presentations of words. Journal of Verbal Learning and Verbal Behavior, 1971, I9, 182-189. Gibson, E. J., and Levin, H. The psychology of reading. Cambridge: MIT Press, 1975. Gilbert, L. C. Functional motor efficiency of the eyes and its relation to reading. Universit of California Publications in Education, 1953, II, I59-232. Goltz, T. H. Comparison of the eye movements of skilled and less-skilled readers (Doctoral dissertation, Washington University, 1975). Dissertation Abstracts International, 1976, 36, 6413B. (University Micro- films No. 87-14058) Heiner, W. H., and Henderson, J. S. Information processing strategies of good and poor readers as evidenced by eye movements. Lewisburg, Pa.: Bucknell University, 1974. (Eric Document Reproduction Service No. ED 098535) Hochberg, J. Component of literacy: Speculations and exploratory research. In H. Levin and J. P. Williams (Eds.), Basic Studies on reading. New York: Basic Books, 1970. Huey, E. B. The s cholo and eda o of readin . Cambridge: r ress, . r1g1na y pu 13 ed, Macmillan, 1908) Judd, C. M., and Buswell, G. F. Silent reading: A study of the various types. Supplementary Educational Monographs, 1922, (Whole No. 23). Kimball, J. P. Seven principles of surface structure parzing in natural language. Cognition, 1973, 2, 5- 7. Kolers, P. A. Three stages of reading. In H. Levin and J. P. Williams (Eds.), Basic studies on reading. New York: Basic Books, 1970. 70 Levin, H., and Cohn, J. A. Studies of oral reading: XII. Effects of instructions on the eye-voice span. In H. Levin and J. P. Williams (Eds.), The Analysis of reading skill, Final report, Project No. 5-1213; Contract No. OE6-10-156, Cornell University and U.S. Office of Education, December, 1968. Levin, H., Grossman, J., Kaplan, E., and Yang, R. Constraints and the eye-voice span in right- and left- ggbggded sentences. Language and Speech, 1972, 15, Levin, H., and Kaplan, E. L. Grammatical structure and reading. In H. Levin and J. P. Williams (Eds.), Basic studies on reading. New York: Basic Books, 1970. Levin, H., and Turner, E. A. Sentence structure and the eyegyoice span. Project Literacy Report, 1966, Z, 7 - . Levin, H., and Turner, A. Sentence structure and the eye-voice span. In H. Levin, E. J. Gibson, and J. J. Gibson (Eds.). The analysis of reading skill. Final report Project No. 511213, Contract No. OE6-10-156, Cornell University and U.S. Office of Education, December, 1968. Mackworth, J. F. Some models of the reading process. Learners and skilled readers. Reading Research Quarterly, 1972, 7, 701-733. Marcus, A. D. The development of a diagnostic test of syntactic clues in reading. In R. Leibert (Ed.), Diagnostic viewpoints in reading. Newark, Delaware: International Reading Association, 1971. McConkie, G. W., and Rayner, K. The span of the effective stimulus during fixations in reading. Paper presented to the American Educational REsearch Association, New Orleans, 1973. Mehler, J., Bever, T. G., and Carey, P. What we look at when we read. Perception and Psychophysics, 1967, 2, 213-218. Miller, G. R., and Coleman, E. B. A set of thirty-six prose passages calibrated for complexity. Journal of Verbal Learning and Verbal Behavior, 1967, 6, 851-854. ‘ 71 Pearson, P. D. The effects of grammatical complexity on children's comprehension, recall, and conception of certain semantic relations. Reading Researach Quarterly, 1974-1975, lg, 155-192. Quantz, J. 0. Problems in the psychology of reading. New York: Macmillan, 1897. Rayner, K. The perceptual span and peripheral cues in reading. Cognitive Psychology, 1975, 2, 65-81. Resnick, L. B. Relations between perceptual and syntactic control in oral reading. Journal of Educational Psychology, 1970, pl, 382-385. Richek, M. A. Readin comprehension of anaphoric forms in varying linguistic contexts. Reading Research Quarterly, 1976-1977, l2, 145-165. Rode, S. S. Development of phrase and clause boundary reading in children. Reading Research Quarterly, 1974-1975, l9, 124-142. Sachs, J. S. Recognition memory for syntactic and semantic aspects of connected discourse. Perception and Psychophysics, 1967, 2, 437-442. Sawyer, D. J. Intra-sentence grammatical constraints in readers' sampling of the visual dis la (Doctoral dissertation, Cornell University, 1 71 . Dissertation Abstracts International, 1861-1972, 32, 1927A. (University Microfilms No. 71-24, 5227 Sheldon, A. On strategies for processing relative clauses: A comparison of children and adults. Journal of Psycholinguistic Research, 1977, 6, 305-317. Siler, E. R. The effects of syntactic and semantic constructs on the oral reading performance of second and fourth graders. Reading Research Quarterly, 1973-1974, 2, 583-602. Smith, F. Understanding reading. New York: Holt, 1971 Steiner, R., Weiner, M., and Cromer, W. Comprehension training and identification for poor and good rggderg. Journal of Educational Psychology, 1971, 62, 5 -51 . Stern, J. A. Eye movements, reading, and cognition. In J. W. Senders, D. F. Fisher, and R. A. Monty (Eds.), Eye movements and the higher psychological functions. Hillsdale, N.J.: *Lawrence Erlbaum Associates, 1978. 72 Stoodt, B. L. D. The relationship between understanding rammatical conjunctions and reading comprehension $Doctoral dissertation, Ohio State University, 1970). Dissertation Abstracts International, 1861-1972, 31, 4635A. (University Microfilms No. 71- 07, 576) Tatham, S. M. Reading comprehension of materials written with select oral language patterns: A study at grades two and four. Reading Research Quarterly, 1970, S, 402-426. Taylor, S. E. Eye movement photography with the Reading E e. New York: Educational Development Laboratories, Inc , 1960. Tinker, M. A. Bases for effective reading. Minneapolis: university of Minnesota Press, 1965. Tinker, M..A. Recent studies of eye movements in reading. Psychological Bulletin, 1958, 55, 215- 231. Thomas, J. Visual processing of competent and less competent adult readers. Paper presented to the International Reading Association, St. Louis, 1980. Thorndike, E. L. Reading as reasoning:- A_study of mistakes in paragraph reading. Journal of Educational Psychology, 1917, 2, 323-332. Walker, R. Y. A. A qualitative study of eye movements of good4rea2§rs. American Journal of Psychology, 1938, 51, 72- l. Wanat, S. F. Relations between language and visual processing. In H. Singer and R. B. Ruddell (Eds.), Theoretical models and processes of reading. Newark, Delaware: International Reading Association, 1976. Wanat, S. F. , and Levin, H. The eye-voice span: Reading efficiency and syntactic predictability. In H. Levin, E. J. Gibson, and J. J. Gibson (Eds.), Eng analysis of reading skill, Final report, Project No. 5-1213, Contract No. OE6-lO-156, Cornell University and U.S. Office of Education, December, 1968. Weber, R. M. A linguistic analysis of first grade reading errors. Reading Research Quarterly, 1970, 5, 427-451. WOodworth, R. S. Experimental Psychology. New York: Holt, 1938. IES 111111111111