WIWwwwmIIImIIIJ'I III II II .IIII. III I. I'II'IW‘IIIII‘I .542. I'J't'.‘ "y I-M . . A IIIII’I $4.. =\:.Ilt-"3‘ ‘ . II. IIIIIII: I If}? 'I " .III. III. I I" II III IIIII‘IIIIIIII III :I I} "IHIIIIMI II III III“ }} III III" IIIIIIII I If‘II‘I‘I: . " II .IIII " ,l ”IIIWI.WIMIIIMIIIIIII, ‘I “I I}} H IWII III" I" '.IIiI!§3IIiifI"IIL;¥I ”#4:? I III II I III. II. III I JprIIkEIII‘Ifl? @qu I'III -I;IIII III I II'III , IIISIII ' I} IIII'IIIII'I IIIIIIIIIIIII'II” IIIIII'IIII {I} IIIIIII‘I HI I‘III III III’IIIII II '7 .IIII'III II IIIIIIH 1%,? ”3:5“:1? II III “III IIII III“ I . III III III 3: I I III: I “II“ II IIII'I I l ' fififigfijfifizé'; I IIIIII'I .IIIIIIII III III . III;x IIIIIIIQII IIIIIII: I IIIIIIEI I “I I I I "IIIIIII , II‘I‘III I II IIIIIII I IMMIIII I I'I‘IIIII III III IIIIII II “III (II III‘IIIII II'IIIISIIIIIIIIIII 2’3. $3.7? a... II I. I- IIIIII}IWIIII IIIIIII . -'I III-III . :‘II IIIIIIIII III. I II III II IIIIIIIIIII I I I IIIIIIIIIIIIIII. IIIIIE IIIIIIIIIIIIIIII... . III . . I I III. I II _____ I: ‘III)' III. ., I.-,' LE1". [$1 9 '2‘” .zIIII III III I. . “III III III III III III IIIIII...I.II.I.II..I..I III {IIIIIW II" 1I IIIII‘ III I I IIII'IIII‘II II ‘IIIII EI' iL I§I1n1\jj__§1§3\-g; ‘12-? 1",, z: I III I" “I II III; lIIIIIIl I III I “III II” IIIIWI’" III“ III] III III‘II IIIIIIII IIIIIIIWIQII 'I “III IH I IIIII }}}I}II'}IIIIIF1'"§,I “III III III III 41*”?51'a'wéflM‘SIII’E: 3‘. I .IIII iIII. I IIIIIIII. III III II. II III .I II III III IIssIdx xIHIIWI II m ”I II? I I'IIJIIII?%£§IIIIIIY $5315; I «'IIIII IIIIIIII III III IIII III II III II III: II' IIC'II- "III ’IIIIIIi " I II I'Pin’filngn",I’IIg§ IISIIIIIE '1‘: “III I IIIIII 1'III I IIIIIIII ”7 " II III [,7 (an... I"‘:’ "II #33154, I335 II?! E... ISI'ISH.““Q§ 32"} “I I I‘III "I” IIIII WIIIIIII‘I IIIILH'I I 'IIII IIIkI'IMq w :11} :In'f‘? ‘fifi‘fifi"..5fi:§ I III II} III III .II II III, III .‘II II“ II- I.‘ I; I I; IgflI ,IIIIQIIIILQIIIIEYIIfi'JIII HIII} . II I" I“ LIIIIE “I; III fl-ITIIII‘IIi‘ I II" ‘3‘ I *Iffipfiffam’ I" E I I III N IIIUI . “$ng :iIIIIIIIIIIIIII. III III III III'IIM .II III "'IIII"III’II‘II’I‘I'I‘II.Lafiimww'5 II.II_.IIJ ~ I I " . . . jg; ' ‘ III III“ I ‘I “WWI-II... JIIII {I157 'I IIzi‘I :31 3193131, «QM * III III I. III I III III IIIIIIIIII III-II III IIIIIIIIIIIII I III , III III. III... .I III III" III II II. III , III II ‘IIIII I :- .IwIIIIIIIII I 1 ML III’ I I III ‘IIIIIIII III IIIIIII II IIIIHI III: ’II IIIIIIiI. IIIIlII BI III‘IIIII} <71 #1 kg??? IE5}“‘$II .IyIi‘IIIIIgIIIIIg {III} IIHHI : IIIII W W III III III I I} .II‘ }"} I‘IIIII II‘IIIIIISIIIIIIIII } III 5III‘IIiI’I I” “Iii IIIIIEII‘I’Pés-‘g I III III” II’I‘IIIIIIII1 ' III} IIIW III‘ III II II II} III IIIdIIII‘I. IMIIII. IIII'IIIIIM III” I"; I III 31;. I. . II’I “III '1‘} 'I' IIHI‘ IIIIII'II IIIIII III II, II III III 'II‘I I'M.” III I'I’IL IIIII} I1 I 13.;IU'IIII III” mI I3: I III III III. "I. .II L III I ‘III II ...I' ' III III III III. I II. II III... mfg! “III.” IIIIIIIIIII III. II II IIIIII IIII‘III I‘III‘ IIIIII III II II III IIIIIIIII IIIII ”III IIIIIII’ III II'III IIIIIII ”I IIIIIIIIIIIbIIIIIII‘IgIII uaImII JIII'I III I 'II I I I ' .‘I II I I 'II I I, I: IIIIII III II IIII IIIIIII I' IIII II III? I} III” IIII IIIIII 1IIIIIII I’I IIIIIII IIIIII, I"; ““““ WIN“ JIIII II .t I II‘IIII IIIIIII I.“ III I)’. II I I IIIIIIII IIIII IIIII “1.3: IFI IIIII IIII IIIIII III II LI IIIIII l(IIIII {IIIIIIII I‘.‘ II IIIIIIII II I. IIIIIII I"i I'.IIII} IgoIilrlIlI k}, :IIIIII “HI WI IIIIIIIIIIIIIK III“ I,.IIIIIII II’II‘III I“ - III. II .IWI III I I“ IIII III I. III IIIIIIIIII III} III!‘ III» ‘III'III'. I ‘ I‘I‘QMII: III II I II III ." IIIIIIIIIIII}III III'II'I'IIIII‘IIIIIIIIIIIIIIIIIIIIIIIIII I'III IIII‘IIIII ll/I/l/I/I/l/II /l/I/l/l/I/I/III/l/I/l/If/I/l//I/////l/I////l mm 3 1293 10394 5423 This is to certify that the thesis entitled A STUDY OF THE ABILITY OF THIRD AND FIFTH GRADE PUPILS AT HIGH AND LOW READING LEVELS TO RECOGNIZE AND PRODUCE A SUMMARY OF NARRATIVE SELECTIONS presented by l Kathryn Koster Blok has been accepted towards fulfillment of the requirements for Ph.D. Education degree in Majo professor Date JUIJ 13, 1981 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book ntum to remve ichargc from circulation records A STUDY OF THE ABILITY OF THIRD AND FIFTH GRADE PUPILS AT HIGH AND LOW READING LEVELS TO RECOGNIZE AND PRODUCE A SUMMARY OF NARRATIVE SELECTIONS By Kathryn Koster Blok A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1981 ABSTRACT A STUDY OF THE ABILITY OF THIRD AND FIFTH GRADE PUPILS AT HIGH AND LOW READING LEVELS TO RECOGNIZE AND PRODUCE A SUMMARY OF NARRATIVE SELECTIONS By Kathryn Koster Blok This study explored the developmental nature of children's ability to learn from text by comparing performance of students in high and low level reading groups on measures of recognizing and producing summaries. The following questions were asked: 1. Is there a difference between the ability of children to recognize a summary and their ability to produce a summary? 2. Is there a difference between children of different age and reading levels in their ability to recognize and produce summaries? 3. Hill the amount of improvement after instruction be different for children of different age and reading levels on measures of ability to recognize and produce summaries? Procedure The subjects were third and fifth grade pupils in a large (l8,000+). midwest school system. Ability to recognize a summary was measured with a lO-item objective test which required the learner to select the best statement of main idea of a paragraph; to produce a summary, by having the learner write a paragraph summarizing a brief narrative selection. A 17-point scale was constructed for scoring summaries. Kathryn Koster Blok Treatment consisted of eight lessons designed to teach the subjects to write a summary. Post tests were administered to measure improvement on the tasks of recognizing and producing a summary. A difference score was constructed by subtracting the score on the objective test from the score on the production test. Design To analyze the data collected in this study, a 2x 2x 2 ANOVA was performed with third vs. fifth graders, and high vs. low level readers constituting the between-subjects variable, and pre-test vs. post test constituting the within-subjects variable on each of three dependent variables: scores on the objective test, scores on the production test, and the difference score. Appropriate a_pgigri_comparison tests were performed on the data to test the directional hypotheses that, on the production measure, third grade high pupils would improve more than third grade low pupils; fifth grade low pupils would improve more than fifth grade high pupils; and fifth grade low pupils would improve more than third grade low pupils. Em A significant main effect was found for the pre-test/post test factor on the production test but not for the same factor on the objective test. The findings indicate that instruction in producing a summary did not improve performance in recognizing a summary, and that the growth in ability to produce a summary was not measured by the objective test. The results raise the question of the amount of Kathryn Koster Blok confidence that can be placed in objective test scores, and of the kind of decisions that can be made on the basis of scores on objective tests. The significant main effect found for the grade level factor on the production test was attributed to the differences in the larger samples whose scores were used to standardize the scores for the smaller third and fifth grade samples. Planned comparison tests made between pre-test and post test standardized production scores demonstrated a significant improvement for high achieving pupils in both third and fifth grade but not for low achieving students. These findings were interpreted as not supportive of the hypothesis that differences between performance of third and fifth grade students and high and low achievers could be attributed to differences in stages of cognitive development. An analysis of factors contributing to form differences found in the production test suggests that research in schema theory may provide some explanations for the differences between pupils of high and low reading levels. ii © Copyright by KATHRYN KOSTER BLOK 198'! To my panama GM and Geld/rude Koatm (in memo/balm) ACKNOWLEDGMENTS No man is an island. The truth of this became very evident as I progressed through this study. I welcome this opportunity to express my deep appreciation to all those who helped bring this project to fruition. To the chairperson of my guidance committee, Dr. Lois A. Bader, who gave direction at the start of the project, listened patiently and often during the project, and provided valuable comments and enormous encouragement. To the members of my guidance committee, Dr. Ben A. Bohnhorst, Dr. Eugene Pernell, Jr., and Dr. Roy Hesselman for their wise counsel and constructive comments. To my friend and colleague, Dr. Bette Bosma, who understood my frustrations, rejoiced in my accomplishments, and provided support and encouragement when I needed it most. To Dr. John Dow, Superintendent; Mr. Richard Bandy, Director of Elementary Education; Mrs. Elisabeth A. Rowlands, Director of Reading Services; and the teachers of the Grand Rapids Public Schools, who made this study possible. To my sisters, Trudy Koster and Abbie Kenyon, for all the reasons they know best. Finally, the most important acknowledgment is reserved for my husband, James Blok, in whose eyes I have always found the kindest mirror. His confidence in my ability has stimulated me to strive for professional growth, his prayers have undergirded my every effort, his deep love has enriched my life. iv TABLE OF CONTENTS Page LIST OF TABLES .......................... vii LIST OF FIGURES ......................... viii Chapter I. THE PROBLEM ....................... 1 Background of the Study ................ l Statement of the Problem ................ 4 Importance of the Study ................ 6 Purpose ........................ 9 Research Questions ................... 9 Generalizability .................... ll Assumptions ...................... 12 Limitations ...................... 13 Definition of Terms ............... ‘. . . l3 Organization of Subsequent Chapters .......... l4 II. RELATED LITERATURE .................... l6 Introduction ...................... l6 Factors Influencing Reading Comprehension ....... 16 The Complexity of the Reading Process ....... 17 Specific Factors Influencing Comprehension ..... 19 Differences in Reading Behaviors of Good and Poor Readers ....................... 20 Knowledge of Word Meaning ............. 2l Decoding Skills .................. 2l Decoding Rate ................... 2l Specific Reading Disabilities ........... 22 Reasoning Strategies ................ 22 Ability to Organize Text .............. 23 Background Knowledge and Experience ........ 25 Stage of Cognitive Development ........... 26 Methods of Teaching Reading Comprehension ....... 30 Summary ........................ 34 Chapter III. IV. V. Appendix A. B. C. METHODOLOGY ...................... Population and Selection of Sample .......... Procedures ...................... Design ........................ Materials ...................... Instructional Materials ............. Data Gathering Instruments ............ Scoring ....................... Analysis of Data ................... Null Hypotheses ................... Directional Hypotheses ................ Summary ....................... PRESENTATION, ANALYSIS, AND DISCUSSION OF DATA ..... Introduction ..................... Construction of Standardized Scores ......... Research Questions and Statistical Analysis ..... Dependent Variable One: Objective Score ..... Dependent Variable Two: Production Score Dependent Variable Three: Difference Score Directional Hypotheses and Statistical Analysis . . . Summary ....................... CONCLUSIONS AND RECOMMENDATIONS ............ Introduction ..................... Summary of Study ................... Major Results and Discussion ............. Null Hypotheses ................. Limitations ................... Implications for Classroom Instruction ........ Recommendations for Further Research ......... SUMMARY CHECKLIST AND SCORING KEY ........... BIBLIOGRAPHY .......................... vi 101 115 125 127 Table OUT-DUO 10. 11. 12. 13. 14. 15. LIST OF TABLES Distribution of Pupils in the Larger Sample by Grade and Reading Group .................... Distribution of the Sample by Grade and Basal Reading Group .......................... Sequence In Which Tests Were Distributed ........ Mean Raw Scores on Objective Test for Larger Sample . . Mean Raw Scores on Production Test for Larger Sample Analysis of Variance Summary Table for Objective Test for Third Grade Pupils in the Larger Sample ....... Analysis of Variance Summary Table for Objective Test for Fifth Grade Pupils in the Larger Sample ....... Analysis of Variance Summary Table for Production Test for Third Grade Pupils in the Larger Sample ....... Analysis of Variance Summary Table for Production Test for Fifth Grade Pupils in the Larger Sample ....... Analysis of Variance Summary Table for Standardized Objective Score ..................... Analysis of Variance Summary Table for Standardized Production Score .................... Comparison of the Interaction Effect of Reading Level and Pre-test/Post Test Factors on the Mean Standardized Scores on Objective and Production Tests ........ Analysis of Variance Summary Table for Standardized Difference Score .................... The Difference Between Mean Pre—test and Post Test Standardized Scores on the Objective Measure ...... The Difference Between Mean Pre-test and Post'Test Standardized Scores on the Production Measure ...... vii Page 36 37 4O 55 56 57 57 58 58 63 69 73 74 75 76 Figure 10. 11. 12. A.1 LIST OF FIGURES Distribution of third grade pupils in basal reading groups ......................... Form order group by pre-test/post test interaction for third grade .................... Form order group by pre-test/post test interaction for fifth grade .................... Major details in Form A and Form B of the production test .......................... Standardized scores on objective test ......... Standardized scores on production test ......... Mean difference between standardized scores on objective and production measures ........... Standardized production scores for third grade high and low students .................... Standardized production scores for fifth grade high and low students .................... Standardized production scores for third grade low and fifth grade low students .............. Summary of research findings for null hypotheses . . . . Summary of research findings for directional hypotheses ....................... Visual aids to accompany lessons 1 and 4 ........ viii Page 59 6O 61 66 71 78 8O 82 83 88 89 102 CHAPTER I THE PROBLEM Background of the Study The ability to summarize has long been recognized as a component of reading comprehension. Edward Thorndike (1917) concluded his classic study of mistakes in paragraph reading with the observation that "it appears likely that exercises in silent reading to find the answers to given questions, or to give a summary of the matter read . should in large measure replace oral reading" (p. 332). 'Additional evidence of the importance attached to the ability to summarize is found in the number of items testing ability to recognize the main idea of a passage which appear in tests of reading comprehension at all levels. However, a survey of materials used to teach reading revealed that teachers are given little guidance in how to teach the skill of summarizing. Frequently the materials focus on the mechanics rather than on the thought process involved. The strategy suggested for the teacher is often that of simply telling and assigning. An example of this type of instruction is found in the Be a Better Reader series by Nila Banton Smith (l977). The learner is told that the main idea is the most important one, or one that seems to sum up everything else that has been said. He is told that the main idea may be found in different places and that he will have to learn to recognize it wherever he finds it. The learner is then given a series of exercises which require him to select the sentence or foil which gives the main idea, but is not led through the thought process involved nor given a strategy for locating the main idea. Recognizing the importance of summarizing as a component of reading comprehension and the limitations of many materials used to teach this skill, the investigator devised a strategy and prepared materials which, used in conjunction with basal selections, would lead the pupils through a thought process assumed to be involved in summarizing. The instructional strategy and materials were used with a group of six third grade pupils who were identified by the classroom teacher as members of the highest level basal reading group in that class. After an oral pre—test revealed that none of the pupils could explain what a summary was, direct instruction was given by the investigator in four, one-hour sessions over a two week period. At the end of this period of instruction, all of these pupils wrote a summary of a basal selection which met the criteria: answers to "reporter questions" (who, where, when, what, why, how, with what result), written in paragraph form. The strategy and materials were then used by a third grade teacher with an intact class of 24 pupils who were placed in basal reading groups on the basis of scores on a basal reader placement test (see Figure l). Number of Pupils Basal Reader in Basal Group Placement 1 5th 3 4th ll 3-2 5 3-l 4 2-2 _ Figure 1. Distribution of third grade pupils in basal reading groups. After several weeks of instruction, using the materials prepared by the investigator, the classroom teacher reported the following: 1. The pupils in the three highest level reading groups each wrote summaries of basal selections which met the criteria of (a) answers to reporter questions, and (b) written as a paragraph. 2. The pupils in the next to the lowest level reading group each wrote summaries in paragraph form which answered some of the reporter questions. However, they failed to relate the ideas in a way which reflected the main idea of the selection. 3. The pupils in the lowest level reading group wrote answers to some of the reporter questions but did not write in paragraph form. They wrote separate sentences answering specific questions and numbered each of the sentences. Two questions arose from this trial teaching experience: 1. If the strategy and materials had been used with a fifth grade class, would a larger percent of pupils have profited from instruction? 2. Would the nine third grade pupils who were unable to produce a summary have recognized a summary statement if it had been presented in an objective test item? A common theme runs through each of these questions: Do develop- mental differences contribute to children's ability to learn from text? Statement of the Problem The Michigan State Board of Education has set the improvement of instruction in Michigan Schools as a priority goal. As one way of working toward this goal, the State Board of Education initiated the Michigan Educational Assessment Program (MEAP), a statewide testing program in Reading and Mathematics. The Assessment Program uses an objective referenced test system with test items keyed to Michigan minimal performance objectives. The criterion for objective attainment was set by the Department of Education (MEAP Handbook). The MEAP tests are administered every fall to all fourth, seventh, and tenth grade pupils in the public schools in Michigan. Test results are published in local newspapers, with school districts and individual schools identified and ranked from high to low in terms of pupil achievement. While publication of this information is intended to keep open the lines of communication between school and home and to generate public interest in educational programs, there seems to be a tendency on the part of the public to use this information to compare school districts, schools, and teachers. This accountability model, although intended to improve the quality of instruction, frequently puts pressure on both teachers and pupils. Teachers tend to feel pressure to bring all students to mastery level in basic skills even though some pupils may not be ready for instruction in specific skills. Elkind (1979) expressed a concern that the back to the basics movement, with emphasis on specified behavioral objectivesiand teacher accountability, places children under pressure to become compe- tent at an early age and may have negative consequences in terms of emotional problems and poor self-image. A question which arises from a survey of the comprehension items on the Reading section of the MEAP is whether performance on tests which require the reader to recognize the correct response is related to per- formance on a task which requires the reader to produce a correct response. In the real world, when the reader is placed in a situation in which he must extract meaning from the printed page, the reader is not given choices. Instead, he must select relevant features, filter out others, and note the relationships among the features he has selected (Gibson & Levin, 1975). This study will compare the rela— tionship between performance on a test which requires the reader to recognize a correct response with performance on a test which requires him to produce a response. A second question which arises from a survey of the skill areas assessed by the Reading section of the MEAP is whether all third grade pupils are ready for instruction in each of these skills. Of special interest to the writer are the items which measure comprehension, particularly those which measure ability to recognize statements of main idea and summary statements. These skills are among the comprehension skills tested at the beginning of fourth grade so it may be assumed that these skills must be taught in third grade. But may it be assumed that all third grade pupils can profit from this instruction? Although there is no model of comprehension on which all authorities can agree, many models include thinking or reasoning as one component of comprehension. If reasoning and thinking skills are required, then the possibility exists that comprehension may be influ- enced by the child's stage of cognitive develOpment. This study will explore whether all third grade pupils, particularly those assigned to low level basal reading groups, will profit from instruction in producing a summary statement. By comparing performance of third grade pupils with that of fifth grade pupils, the study will explore the possibility that performance is related to cognitive maturity. In summary, this study is concerned with two problems faced by classroom teachers when curricular and instructional decisions are influenced by the content of tests used to measure reading achievement. 1. Can decisions about a pupil's ability to extract meaning from text be made on the basis of test items which simply require the reader to recognize a statement of main idea or a summary statement? 2. Can a decision about the timing of instruction in a given skill be made on the basis of when it is tested in the reading program? Importance of the Study It is hoped that this study will provide elementary teachers with guidance in (a) making decisions on the basis of test scores, (b) timing instruction in specific skills, and (c) matching students with materials for reading instruction. At a very practical level, the study may be important because it provides teachers with lesson plans and materials for teaching the skill of summarizing. Teachers often make instructional decisions on the basis of pupil performance on multiple choice items on reading tests. However, outside the school setting the reader must usually extract meaning from print rather than select the best statement of meaning. This study will compare pupil performance on tests of recognizing a summary with performance on the task of producing a summary. The results may suggest to teachers how much confidence may be placed on performance on objective tests. Current emphasis on teacher accountability puts much pressure on teachers to bring all pupils to mastery level in the basic reading skills. Given the many curricular demands placed on elementary teachers, it is important that they present instruction in specific skills when this can be done using the least possible amount of time and resources. This study may provide teachers with guidance for scheduling instruction when the learner is best able to profit from such instruction. The study may also suggest another dimension for teachers to consider when matching students to reading materials. It is commonly agreed that students should be provided with instructional materials which match their reading levels. This match is usually determined by applying a readability formula to the text (measuring word and sentence length) and administering an informal reading inventory to the student (measuring knowledge of word attack skills, meaning vocabulary, and ability to comprehend at the literal and inferential levels). Com- parison of the performance of third and fifth grade pupils in low level basal reading groups may provide insights into the necessity of matching the demands of the reading task to the cognitive level of the reader. Pragmatically, for the elementary teacher the major importance of this study may be the materials and lesson plans provided for teaching children how to summarize narrative material. Dolores Durkin (1978-79), in a study reported in the Reading Researchgguarterly, portrayed teachers as mentioners, assigners, checkers, and interrogators. She then raised the question as to why teachers were not teaching. In a follow-up study, presented at the September 26, 1980 conference of the Michigan Secondary Reading Interest Council, she answered that question. After an in-depth study of the manuals of leading basal readers (Allyn & Bacon, Ginn, Harcourt Brace, Houghton Mifflin, and Scott Foresman), she concluded that teachers were following the manuals--and the manuals were not giving strategies for teaching. For the most part, she found that manual suggestions for teaching could be classified as assigning, telling, mentioning, and checking. The materials provided in this study give teachers very specific strategies for teaching pupils how to summarize a narrative selection. Purpose The purpose of the study was to investigate the developmental nature of children's ability to learn from text by comparing the ability of children of different age and reading achievement levels to recognize and to produce summary statements of narrative material. Specifically, the study sought to determine (a) whether decisions about pupil ability to produce a summary can be made on the basis of performance on a test to recognize a summary, (b) whether differences exist between the ability of pupils of different age and reading achievement levels to recognize and produce summary statements, and (c) whether improvement after instruction in writing summaries might be related to pupil age and reading achievement levels. Research Questiqgs It was the purpose of this study to explore the developmental nature of children's ability to learn from text by investigating the ability of children of different age and reading levels to recognize and produce summary statements of narrative passages. To guide the study, the following research questions were formulated: 1.0 Will there be a difference between the ability of children to recognize a summary statement and their ability to produce a summary of a narrative passage? l.l Will there be a difference between the ability of third grade pupils at a low reading level to recognize a summary statement and their ability to produce a summary? 10 1.2 Will there be a difference between the ability of third grade pupils at a high reading level to recognize a summary statement and (their ability to produce a summary? 1.3 Will there be a difference between the ability of fifth grade readers at a low reading level to recognize a summary statement and their ability to produce a summary? 1.4 Will there be a difference between the ability of fifth grade pupils at a high reading level to recognize a summary statement and their ability to produce a summary? 2.0 Will there be a difference between the ability of children of different age and reading levels to recognize and produce a summary of a narrative selection? 2.1 Will there be a difference between third grade pupils at a low reading level and fifth grade pupils at a low reading level in their ability to recognize a summary statement? 2.2 Will there be a difference between third grade pupils at a low reading level and fifth grade pupils at a low reading level in their ability to produce a summary? 2.3 Will there be a difference between third grade pupils at a high reading level and fifth grade pupils at a high reading level in their ability to recognize a summary statement? 2.4 Will there be a difference between third grade pupils at a high reading level and fifth grade pupils at a high reading level in their ability to produce a summary? 11 3.0 Will there be more improvement after instruction in the performance of pupils at a high reading level than in the performance of pupils at a low reading level? 3.1 Will there be greater improvement after instruction in the performance of third grade readers at a high reading level than in the performance of third grade readers at a low reading level on a task which requires them to recognize a summary statement? 3.2 Will there be greater improvement after instruction in the performance of third grade readers at a high level than in the per- formance of third grade readers at a low level on a task which requires them to produce a summary? 3.3 Will there be greater improvement after instruction in the performance of fifth grade readers at a high level than in the per- formance of fifth grade readers at a low level on a task which requires them to recognize a summary? 3.4 Will there be greater improvement after instruction in the performance of fifth grade readers at a high level than in the per- formance of fifth grade readers at a low level on a task which requires them to produce a summary? Generalizability This study was conducted in eight classrooms in five schools of a large metropolitan school system in western Michigan. The study used only the pupils in the highest and lowest reading levels in each of these classrooms. The classroom teachers, using lesson plans and teaching materials prepared by the investigator, provided the 12 instruction during the time regularly scheduled for each of the reading groups. The tests were also administered by the classroom teacher. Because of the limited number of subjects involved, the lack of control, and the absence of sampling techniques, the results cannot be generalized to similar classrooms. However, the completeness of the account of procedures and the inclusion in the appendix of all lesson plans, teaching materials, and sample test items, should make it possible for any elementary teacher or researcher to replicate the study. Assumptions The study was based on the following assumptions: 1. It was assumed that the sex of the subjects and membership in a minority group would have no bearing on their ability to profit from instruction. 2. Since all classes were taught by teachers with a minimum of five years of experience, and since all teachers were given identical lesson plans and teaching materials, it was assumed that instruction did not vary significantly from class to class. 3. Since instruction was given during the time regularly scheduled for the reading period, it was assumed that time of day and day of week would not affect the study. 4. It was assumed that the same types of reasoning-thinking skills were involved in locating main ideas of paragraphs and summarizing longer selections. 13 Limitations There are two major limitations of this study. First, because of the limited number of subjects, the lack of control, and the absence of sampling techniques, the results cannot be generalized to similar classrooms. Second, the study was limited by factors which operate when research is carried on in the natural setting of an elementary classroom: teacher and pupil absenteeism, teaching priorities, and the interruptions which are a normal part of life in any elementary classroom. Definition of Terms The reader may better understand this study if certain terms are clarified. Basal reading group: A basal reading group is a group of pupils assigned to the same level basal reader on the basis of scores on a basal reader placement test. Cbmerfbrd statement of’objective: This model for writing an instruc- tional objective was developed by Dr. Joseph Comerford (Comerford & Fleury, 1972). It has four parts: (1) teacher's responsibility, (2) content or skill to be taught, (3) pupil involvement or participation, and (4) expected or ultimate pupil behavior. Highest ZeveZ basal reading group: For the purpose of this study, this term is defined as a group of pupils within a given classroom who are reading from the highest level basal reader used in that specific classroom. 14 Lowest level basal reading group: For the purpose of this study, this term is defined as a group of pupils within a given classroom who are reading from the lowest level basal reader used in that specific classroom. Objective test: This term refers to a ten-item test which required the student to select the best statement of main idea from the three sentences which followed a narrative passage. Production test: This term refers to a test which required the student to read a short narrative passage and write a paragraph summarizing the selection. Reporter questions: For the purpose of this study, the term "reporter questions" refers to questions beginning with the words, ghg, where, when, why, what, how, and with what result. They are termed "reporter questions? because they are typical of questions reporters answer in their news reports. Teacher-directed lesson: For the purpose of this study a teacher- directed lesson is defined as a lesson in which the teacher explains and models the learning activity and involves pupils actively in the learning activity. Organization of Subsequent Chapters The content of Chapter I has included the background of the study, statement of the problem, importance of the study, purpose of the study, research questions to be explored, generalizability of the findings, assumptions and limitations of the study, and definitions of terms. 15 In Chapter II, pertinent research and literature relating to the subject content of this study will be reviewed. In Chapter III, a description of the design and methodology used in the study will be discussed. In Chapter IV, presentation of the data collected and analyzed will be reported and discussed. In Chapter V, a summary of the study, appropriate conclusions, and recommendations for future research will be presented. CHAPTER II RELATED LITERATURE Introduction The purpose of the study was to explore the developmental nature of children's ability to learn from text by comparing the ability of children of different age and reading achievement levels to recognize and to produce summary statements of narrative materials. The review of the literature is organized under three main headings: (1) factors influencing comprehension, (2) differences in the reading behavior of good and poor readers, and (3) methods of teaching reading comprehension. Factors Influencing Reading Comprehension A continuing challenge facing classroom teachers and researchers is that of finding ways in which children can be helped to read better. For many years the majority of reading research investigated better ways of teaching word recognition skills. However, recent studies suggest that a high level of skill in word identification is not a sufficient condition for comprehension (Cromer & Wiener, 1976; Oakan, Wiener & Comer, 1971). While decoding training significantly increased decoding speed of single words, it did not improve comprehension performance (Fleisher, Jenkins, & Pang, 1979). 16 17 Reading is now commonly described as a complex, cognitive process, and researchers, influenced by the contributions of psycholinguists and cognitive psychologists, are shifting their focus from word recognition to reading comprehension (Pearson & Samuels, 1980). The growing number of studies of reading comprehension is leading to an increased awareness of the complexity of the reading process and the many specific factors which can influence reading comprehension. The Complexity of the Reading Process When reading is defined as a communication process rather than a decoding process, investigations must take into consideration the many facets of language processing and reasoning-thinking which are related to reading. The many language variables which may influence comprehension are discussed by Bormuth (1969) who views comprehension as a response to the language system. The efficient reader must learn at least the following systems by which language signals meaning: (1) the semantic meanings of words, (2) the ways word affixes influence the semantic meanings and syntactic functions of words, (3) the ways phrase and deep structures are assigned to sentences, (4) the ways the surface and deep structures of sentences govern the modification of word and phrase meaning, (5) the identification of antecedents of pronouns, pro-verbs, anaphora, and other prostructures, (6) the ways structures are assigned to paragraphs and larger units of discourse, and (7) those structures used to modify sentence, paragraph, and section meanings. 18 The complexity of reading comprehension was recognized as early as 1980 when Edmund Huey wrote: "And so to completely analyze what we do when we read . . . would be to describe very many of the most intricate workings of the human mind" (p. 6). Variables related to reasoning and thinking must be considered when investigating reading comprehension. Presently, there is no model of comprehension on which all leaders in the field of reading agree (Gibson & Levin, 1975; Bormuth, 1979). Some regard comprehension as a unitary ability; others view it as a complex set of sub-skills (Beery, Barrett, & Powell, 1973). However, there appears to be much agreement among both scholars and laymen that reading is related to thinking and reasoning (Spache, 1969; Pearson, 1978). An example of the manner in which thinking is conceptualized as an integral part of comprehension is found in the model of Harris and Smith (1980, p. 207): Affection.‘\\~\‘. ' * Quality of Background selection experience 7 , ‘ Thinking Reading Reading , - abil itiesl—T EurposesJ—F comprehension I Language ' ‘ ‘ ’ abilities Physical ' ' factors \ // Affection 19 Using tenth graders as subjects, Olshavsky (1977) investigated the types of strategies readers employ to comprehend an author's message. Her study identified ten different strategies and supported the theoretical position that reading is a problem solving process. Specific Factors Influencing Comprehension There is an emerging literature which recognizes many specific factors which influence comprehension when reading is viewed as a language process and a thinking process. Many studies have found that reading comprehension may be related to factors within the text. These factors include vocabulary difficulty and sentence length (Fry, 1968; McLaughlin, 1969), syntactic complexity (Dawkins, 1975), and conceptual density and abstractness (Harned, 1980). Additional factors within the text noted by Peters (1977) are (a) the method used to present concepts, (b) the organization of concepts in the text, (c) the types of questions presented in the text, (d) the effectiveness with which the text uses adjunct aids such as graphs, charts, and illustrations, and (e) the arrangement of structural aids in the text (i.e., introductions, summaries, sub-headings, etc.). Related to factors within the text which can enhance or impede comprehension are the instructional procedures used by teachers. Comprehension can be enhanced when teachers use differentiated assign- ments, study guides, and directed teaching activities to compensate for difficulties within the text (Herber, 1970; Roe, Stoodt, & Burns, 1978; Estes & Vaughan, 1978). Vacca (1981) emphasizes the importance of the structured overview as a means of increasing comprehension. 20 Factors specific to individual students can influence the degree to which the reader will comprehend textual material. Bader (1980) lists the following as essential for growth in comprehension: con- ceptual background, attending behavior, perception of organization, syntactic ability, semantic knowledge, reasoning ability, reading rate, and recall ability. Research in schema theory has emphasized the role of experience background in comprehension and interpretation of textual material. Anderson (1977) states that "comprehension and interpretation involve an interaction of input with existing knowledge" (p. 416). He notes that when the reader has had prior experiences which relate to the message of the text, the reader is likely to modify the text and introduce new elements which incorporate his or her knowledge and belief about the world. In summary, comprehension is believed to be related to language and reasoning and may be influenced by factors within the text, instructional procedures of the classroom teacher, factors specific to individual readers, and the experience background of the reader. Differences in Reading Behaviors of Good and Poor Readers A number of studies have compared the reading behaviors of good and poor readers in an attempt to account for the differences and to find ways in which to help poor readers achieve at a higher level. Some of these studies have examined learner mastery of specific skills thought to be related to reading comprehension. Other studies have 21 related reading comprehension to reading strategies employed by the reader. A variety of explanations have been offered to account for the differences between good and poor readers. Knowledge of Word Meaning, Karlin (1980, p. 290) states: "Studies by different investigators show that knowledge of word meaning is the most important single factor accounting for variability in reading comprehension." This would sug- gest that the poor readers could be helped by developing vocabulary programs to help them expand their meaning vocabularies. Decoding Skills The results of a study by Samuels, Begy, and Chen (1975) indicated an interaction between fluent reading and speed of word recognition. They found the better reader seems to be able to use context cues and letter cues from the target word more efficiently than poor readers. Their findings seem to suggest that poor readers can be helped to read better by providing training in more sophisticated word recognition strategies. Decodinngate Fleisher, Jenkins, and Pany (1979) examined the effects on comprehension of increasing the decoding speed of poor readers. They found such training increased the decoding speed of single words but did not improve comprehension performance. They concluded that poor readers appear to have difficulty in transferring single word skills to context, but did not offer suggestions as to how poor readers might develop this skill. 22 Specific Reading Disabilities Wiener and Cromer (1967) analyzed the various explanations which have been offered to account for reading difficulty and concluded there are four different models of reading difficulty, each implying a different kind of remediation. These models are: l. a defect model. The difficulty is accounted for by some dysfunction (e.g., sensory impairment). Learning to read cannot occur unless the defect is removed or a different modality is used. 2. a deficit model. Some process or factor is absent. Many remediation programs are based on this model, attempting to help the reader learn something he has not yet learned (e.g., phonetic skills). 3. a disruption model. Some factor is present and is interfering with learning. This factor must be removed before learning to read actively can occur (e.g., anxiety, hyperactivity). 4. a difference model. There is a mismatch between the indi- vidual's mode of responding and that required by the material (e.g., the child's verbal language patterns and the language patterns of the material. The learner will not be able to read adequately until a change is made in the material to be read, or in the child's language patterns). Reasoning Strategies Kavale and Schreiner (1979) found a difference in the reasoning strategies used by good and poor readers. Both groups were found to use similar strategies but the above average readers applied them significantly more often than average readers. From their results 23 they concluded that above average readers are more active in attempts to achieve understanding. Included in their recommendations is the suggestion that teachers should teach less skilled readers to apply reasoning strategies. Ability to Organize Text Several research studies find a difference between good and poor readers in the way the readers use the organizational features of the text. These studies seem to support the hypothesis that comprehension is enhanced when the reader is able to detect and use the author's organization of the text. A study by Wiener and Cromer (1966) suggests that a high level of word identification skills may not be a sufficient condition for the occurrence of adequate reading comprehension. In a related study (Oaken, Wiener, & Cromer, 1971), a relationship was found between comprehension and the way in which readers organize input. Golinkoff (1976), after reviewing selected studies, found the results did indicate mastery of decoding skills is one primary factor in the difference between good and poor comprehenders. However, she also found another major factor was the text organization processes used by good readers (obtaining meaning from phrases, sentence, and paragraph units). Cromer (1970), using college students as subjects, compared difference and deficit readers with each other and with good readers. Materials were presented under four conditions: regular sentences, single words, phrases, and fragmented groupings. He found the 24 difference group, but not the deficit group, comprehended as well as good readers when material was presented in preorganized phrases. He concluded this study supported the notion that one source of compre- hension difficulty may be attributed to a difference in the way some readers organize input. Remediation would consist of either teaching beginning readers to group words, or to present materials in meaningful groupings. The relation of organizational aspects to reading comprehension was also investigated by Oakan, Wiener, and Cromer (1971) in a study of poor and good readers at fifth grade level. Before the beginning of the experiment, the poor readers were trained to recognize all the words which would be used in the story. This training did not result in any apparent increase in comprehension. They concluded that an appreciable amount of the poor reader's comprehension difficulty may be attributable to the manner in which he organizes his input. They recommended that instruction as to the means of attending to organiza- tional cues should be included as an integral part of reading instruction. Several studies have examined the effect of advance organizers upon comprehension of good and poor readers. (Advance organizers used in these studies were oral or written statements which introduced the reader to the general idea of the material and to the organizational strategy of the author.) The use of experimenter-provided text orga- nization was found to have positive effects on reading comprehension of below-average and poor readers but did not significantly influence 25 comprehension of good readers (Ausubel & Fitzgerald, 1962; Smith & Hesse, 1969). Some interpret these findings to mean that good readers have little need of advance organizers since they have already devel- oped some sort of strategy for organizing content and that poor readers tend not to organize material independently and, therefore, can profit from their use (Van Blaricom, 1979). Background Knowledge and Experience Research in schemata theory seems to indicate that one of the sources of individual differences in reading comprehension may be differences in background knowledge the students possess about the content of the text material. In a study reported by Steffensen, Joag-Dev, and Anderson (1979), Indians (natives of India) and Americans were asked to read and recall two letters, one that described an Indian wedding and one that described an American wedding. The Indian subjects recalled more idea units from the Indian passage than from the American passage, and the reverse was true of the Americans. The results were interpreted as showing the pervasive influence on comprehension of schemata embodying knowledge of the content of a discourse. The finding of that study suggests that comprehension is impeded when there is a mismatch between the subculture and the majority culture whose viewpoint predominates in the materials children are given to read. Implied is the suggestion that some poor readers may comprehend more if teachers attempt to match materials to the experience background of the pupils, or provide experiences to compensate for the differences in experience backgrounds. 26 Stage of Cognitive Development Current literature in the field of reading and recent studies suggest another factor which may, in part, account for the differences between good and poor readers: the learner's stage of cognitive devel- Opment. Many of these studies are based on Piaget's theory of intellectual development. Much still remains to be done before solid claims can be made for building instructional programs in reading on Piaget's theories. Smith and Robinson (1980) caution that all of the evidence is not in, and that it may be that Piaget's stages cannot be related on a one-to-one basis to reading instruction. Waller (1977) observes that Piaget has written nothing himself about reading. In addition, he states that Piaget considers language and reading as poor media for conveying information, especially before adolescence. He also notes that, because of the complexity of the reading process and the large number of competencies involved in Piaget's theory of thinking, it is very difficult to specify a relation between the two. Nevertheless, Waller believes that within Piaget's theory there are some major implications for reading. He hypothesizes that thinking is a necessary, but not sufficient, condition for reading at any level, for any of its sub-parts, and for any purpose. Waller sees operational thought and perceptual decentration as necessary for successful reading acquisition, and formal intelligence as necessary for comprehension of adult level material. Cleland (1980) comments that Piaget, in his writing, makes almost no direct reference to reading, and that he is more concerned with how 27 the mind works than with what it does. However, he believes Piagetian theory can contribute to existing and future models of the reading process. Cleland relates four aspects of Piagetian theory to reading: 1. The organizational framework of the reader. Reading requires cognitive activity. Piaget's theory of schemata relates to individual abilities to determine meaning from a selection, and to the changes in schemata that a reading selection may require of the reader. 2. The linkage of development and reading. Reading, viewed as a thought process, is linked to development. The characteristics of thinking, described by Piaget at each of his four postulated stages of intellectual development, may be helpful in describing the differences in reading behavior as the child develops. 3. The interrelatedness of language processes. In Piagetian theory, all language processes share a common conceptual base of experiences and the ways in which these experiences have been cog- nitively organized. In reading, this conceptual base influences the meaning the reader will derive from the text. 4. The twin influences of maturation and experience. Maturation is necessary for neurological growth. Experiences are necessary for development of schemata. Cleland holds that a reading model must recognize that a certain neurological development is desirable for reading, but that a number of schemata, formed by experience, are also necessary (Cleland, 180, pp. 10-15). Charles (1974) expresses a belief that, in future years, Piaget's work will attract much more attention in the field of education because Piaget's findings indicate what teachers can, and cannot, expect of 28 students at various stages of intellectual development. Among his list of "Key Ideas" are the following, which are important to this study: 1. Children's mental development progresses through definite stages. 2. Although the stages of mental development occur in a fixed order, different children move from one stage to another at different ages. Further, a child may function in one stage for some matters, while he functions in a different stage for other matters. 3. Children's mental development imposes definite limitations on - what they can learn and how (the conditions under which) they learn. 4. Thoughts grow from actions, not words. 5. Knowledge cannot be given to children. It must be discovered and constructed through the learner's activities (Charles, 1980, pp. 1-5). The literature of the past several years reflects an increasing interest in Piagetian theory as it relates to the reading process and reading instruction. This same interest is reflected in an increasing number of research studies. Among the studies that seek a relation between Piagetian theories and reading instruction, more can be found that are related to reading readiness and beginning reading than to comprehension. A common theme running through many of these studies (e.g., Cox, 1976; Roberts, 1976; Kirkland, 1978) is a concern that children may be pushed into reading activities for which they are not intellectually ready. The researchers emphasize the importance of matching the demands of the reading task to the cognitive developmental level of the learner. 29 Similar concerns are found in the work of Kretschmer (1972, 1975), who envisions a truly developmental model of reading which would someday enable educators to "endow the term 'developmental reading' with more meaning than simply 'that reading instruction which is not corrective or remedial'" (1975, p. 187). Kretschmer notes that, while most behavioral characteristics of preoperational thought have disappeared in the normal child after the age of eight, they tend to reappear when the child of this age deals with content in the abstract (1975, p. 183). After analyzing the responses made by children in grades two to six on the Gate MacGinitie Reading Tests (Primary A, Primary B, and Primary C), Kretschmer found evidence that many of the errors were related to centration. That is, the responses indicated a lack of ability to consider the selection as a whole idea composed of two distinct ideas. The results of this study led to the conclusion that Piaget's model of intellectual development might be profitably applied to thinking in reading. In summary, a number of explanations have been offered to account for the differences between good and poor readers. These include, but are not limited to, mastery of a meaning vocabulary, mastery of specific skills, specific problems of poor readers, reasoning strategies, and strategies the reader uses to organize the text. A growing body of literature is exploring the possible relationship between Piaget's theory of intellectual development and the reasoning- thinking process involved in reading. Some of the characteristics of thought described by Piaget seem applicable to some of the demands of 30 the reading task. The possibility exists that one of the many factors which contribute to the child's ability to learn from text may be the learner's stage of cognitive development. Methods of Teaching Reading Comprehension Although much has been written on the topic of comprehension, comparatively few studies have explored methods of teaching comprehension. In an ERIC search of the literature in which they used compre- hension as the descriptor, Smith and Mason (1980) located 3,539 entries. When they added the descriptors research and teaching methods, they found the number reduced to 193. Applying the criterion "methods which most teachers can adapt and apply to their repertoire of instructional strategies" (p. 1), they found the number of entries reduced to about 50. The studies of methods of teaching reading comprehension apparently have not yet made a major impact on teaching comprehension in the school setting. Durkin (1978-79) presents a substantial amount of evidence that teachers spend little time teaching students how to comprehend written discourse. Pearson and Samuels (1980) state: There is a growing uneasiness among practitioners that something is amiss in our ability to deliver instruc- tion in comprehension. The source of the uneasiness is captured in the perception that while reading test scores in many districts seem to compare favorably to national norms in the primary grades, they begin a skid in the intermediate grades that continues into the secondary grades. There is an invitation, if not a compulsion, to conclude that we do a good job of teaching decoding skills in the primary grades but an inferior job of teaching comprehension thereafter. (p. 182) An elementary principal (Viti, 1980) writes: 31 I have doubts about the quality of current comprehension instruction in our schools, particularly for the bright chil- dren. Too many of them . . . are the victims of a push, push, push environment that accelerates them through books without sufficient regard for in—depth comprehension instruction. (p. 566) Pearson (1980) expresses concern for the confusion that exists concerning reading comprehension. Part of the confusion, he believes is due to the controversy as to whether comprehension is teachable. An extreme position holds that only word identification skills are teachable and that comprehension is dependent on native intelligence and experience. A second position argues that, although comprehension may not be teachable, instructional conditions can be arranged to increase the likelihood that children will understand what they read. On the other extreme is the position that comprehension can be taught directly (p. 4). This controversy is also evident in the literature. Cunningham and Foster (1978) report the success of a strategy for teaching story grammar to sixth graders as a means of increasing reading comprehension. Dreher and Singer (1980) challenge the study, reporting their work with fifth graders which led them to conclude that it is not beneficial to teach story grammar to intermediate students. In a later study, Whaley (1981) reviews the research on story grammar, describes some instruc- tional procedures for enhancing pupil awareness of story structure, and suggests teachers can help children become better readers by helping develop their knowledge of story elements. While the controversy continues as to whether reading compre- hension is teachable, some studies are creating options for teaching 32 comprehension which can be explored by teachers who hold that comprehension can be taught. Pearson (1980) believes comprehension can be taught both directly and indirectly. To teach comprehension directly, he suggests that teachers (a) model the comprehension process, (b) provide cues to help children understand what they are reading, (c) ask pointed, directional questions, (d) offer feed-back, both informational and reinforcing, at appropriate times, and (e) generate useful independent practice activ— ities. To teach comprehension indirectly, he suggests that teachers (a) expand and clarify the child's meaning vocabulary, (b) teach the child how to handle the visuals in the text, and (c) offer guidance regarding how the reader can organize his reading and study of a text (p. 4). The New York State Education Department has prepared a booklet of reading activities (1978) which reflect an awareness of reading as a complex mental and linguistic process rather than an hierarchy of mechanically applied sub-skills. The booklet provides the classroom teacher with a wide variety of options for teaching comprehension: (a) using teaching strategies which employ open-ended and student generated questions to sharpen the reader's focus; (b) using key words as semantic, syntactical, and relational cues; (c) using the cloze procedure in predicting and confirming relationships among cues; (d) categorizing as a process for organizing information; (e) modeling concept attainment; and (f) summarizing and paraphrasing. Smith and Mason (1980), in their review of the literature, found a variety of teaching procedures explored by the researchers. These 33 included practice, questioning, study, cloze, analysis of sentence structure, and the integration of skills involved in other language arts. Smith and Mason conclude their review of the literature with the following observations. 1. There does not appear to be one best approach for teaching comprehension. It is possible students may benefit most from an eclectic approach which combines practice, linguistic principles, increased student involvement, teacher modeling, and systematic application. 2. There is a need for more research to be done in the classrooms at varying grade levels including the primary grades. 3. Most of the studies were done in the seventies. Perhaps research in reading comprehension will come of age in the eighties (p. 7). In summary, numerous articles have been written on the topic of comprehension but few studies have investigated methods of teaching reading comprehension. Although a controversy continues regarding whether comprehension is teachable, the few studies that do exist suggest a variety of options are available to teachers who believe comprehension can be taught. These options include teacher modeling, using questioning techniques, increasing student involvement, and creating pupil awareness of linguistic structures. 34 Summary The review of the literature in Chapter II has explored factors influencing comprehension, differences between good and poor readers, and methods of teaching reading comprehension. Chapter III will present the methodology employed in the study. The population will be discussed, the larger sample and sample described, procedures for collecting and analyzing data will be presented, and the method of reporting data will be explained. CHAPTER III METHODOLOGY In this section the methodology used in conducting the study is described. The population and sample are identified and defined, and the procedure for selecting the sample is explained. The data gathering procedures are outlined, the design of the study is presented, and the treatment and testing materials are described. Finally, the statistical treatment of the data is explained. Population and Selection of Sample The children in this study were third and fifth grade pupils from 13 schools in a large (18,000+) school system in western Michigan whose teachers had volunteered to participate in the study. The schools rep- resented a cross section of the schools in the district and included schools with a high percentage of low socioeconomic, minority, and bilingual children. In all of these schools basal readers constituted the heart of the reading program. Only those pupils placed in the highest and lowest basal reading groups in each classroom were selected to participate in the research project. Pupil placement in basal reading groups was determined by the classroom teacher on the basis of pupil performance on the California Achievement Tests, scores on the district's own Objective Referenced Reading Tests, and teacher observation of pupil response to instruction. 35 36 This larger sample was composed of 264 pupils from 15 third grade classes and 7 fifth grade classes. The breakdown of this group by grade and reading group is presented in Table l. The pre-test scores of the pupils in the larger sample were used to compute the mean and standard deviation of the tests prepared by the investigator for use in this study. Table 1 Distribution of Pupils in the Larger Sample by Grade and Reading Group Number in the Number in the highest reading lowest reading Grade group group 3rd 89 65 5th 69 41 From this larger sample, a sample was selected for this study which consisted of 57 third grade pupils and 65 fifth grade pupils from eight classes in five schools. Selection of the sample was made after teaching materials and parent permission forms were distributed to the teachers but before the pre-test materials were provided. Selection of the sample was influenced by the following factors: 1. Since the project was to extend over a six-week period (four weeks for instruction and two weeks for pre- and post testing). schools 37 with a high rate of pupil mobility or with a high record of absentees were eliminated from consideration. 2. Since materials for instruction and testing would be prepared on a readability level appropriate for the lowest reading level in each class, a decision was made to eliminate any third grade class in which the lowest reading level was below 2-1 level, and any fifth grade class in which the lowest reading level was below third grade level. 3. To obtain a sample that was as representative as possible of an entire class, classes were eliminated if more than one pupil did not receive parental permission for participation. Using these criteria, a sample of 122 pupils was selected from four third grade classes and four fifth grade classes from five schools. The reading range of the third grade pupils in the sample was from beginning second grade reading level to beginning third grade level; of the fifth grade pupils, from beginning fourth grade reading level to beginning fifth grade level. Distribution of the sample by grade and basal reading group is presented in Table 2. Table 2 Distribution of the Sample by Grade and Basal Reading Group High Low Pupils basal group basal group Third grade 26 31 Fifth grade 40 25 38 The five schools in this sample represented a narrower range than the schools in the larger sample. All were located in urban, middleclass neighborhoods. However, since the school district uses busing in an attempt to achieve a degree of racial and academic balance, low socioeconomic groups and minority groups are represented in each of these schools. The schools from which the sample was selected were classified by the administration (Dow, 1981) as schools in which pupil reading skill needs were low or moderate, and in which pupil reading scores over a one-year period were stable or improving. This classification was made by comparing scores of fourth grade pupils on the reading sub- tests of the Michigan Educational Assessment Program test administered in the fall of 1980 with the scores of fourth grade pupils on the reading sub-tests administered in the fall of 1979. Procedures Permission to conduct this study was granted by the superintendent of schools, contingent upon the approval of the director of elementary education and the director of reading services. The research proposal was also evaluated and approved by the director of research. Entrance into the individual schools was gained through a letter to building principals which was sent over the signatures of the directors of elementary education and of reading services, and which encouraged participation in the study. Purpose and procedures of the study were explained in a brief meeting with the reading consultants of the district, who then met 39 with building principals and individual teachers to answer questions and urge participation. Purpose and procedures were again explained at a meeting with the teachers who expressed interest in the study. Assurance was given that no comparisons would be made between teachers or schools, and that all names of children, teachers, administrators, and schools would remain anonymous. Building principals sent letters to parents explaining the purpose of the study and requesting permission for their child to participate. Classroom teachers explained the study to their pupils, requested pupil permission to give the tests to the investigator to be scored, and assured the pupils that no grade would be given and that no decision regarding their reading achievement would be made on the basis of these test scores. After the data were collected and analyzed, all teachers partici- pating in the study were invited to a meeting at which the investigator provided computer print-outs of individual scores and discussed the findings of the study. Design The study was conducted in eight elementary classrooms, four at the third grade level and four at the fifth grade level, using the existing grouping within the classroom. Only those pupils assigned by the teacher to the highest and lowest basal reading groups were included in the study. Two pre-tests and two post tests were administered, one testing ability to recognize a summary statement and the other testing ability 40 to produce a summary of a narrative selection. Two forms of each test were used with each group, with one form administered to half of the high level group and half of the low level group, and the second form given to the other half of each group. See Table 3 for the sequence in which the tests were distributed in each class. Table 3 Sequence In Which Tests Were Distributed Pre-test Post Test Recognize Produce Recognize Produce A 1 B 2 B 2 A l A 2 B l B l A 2 Treatment consisted of 8 teacher-directed lessons, 2 each week for a period of 4 weeks. Instruction was given by the classroom teacher in the time period designated for the basal reading lesson. The teacher followed a series of lesson plans written by the investigator which prepared the pupils to write a paragraph summarizing a narrative passage. At the end of the treatment period, alternative forms of the two tests were administered. All tests were administered by the classroom teacher. 41 Materials The instructional materials (see Appendix A) and the testing materials (see Appendix B) were constructed by the investigator for use in this study. The instructional materials prepared pupils to write a summary of a narrative passage. The testing materials con- sisted of two types of tests: an objective test measuring ability to recognize a summary, and a production test measuring ability to write a summary. Instructional Materials The instructional materials included a one-page introduction to the project, a series of eight lesson plans which prepared pupils to write a summary of a narrative passage, and two file folders which contained teaching aids. The first folder contained a set of question starters labeled "reporter questions" and a newspaper article to be used in introducing the concept of a summary. The second folder introduced pupils to the response requirements of cognitive memory and convergent questions. Attached to the instructional materials was a form for recording pupil attendance at each of the eight instructional sessions. The lesson plans were designed for use with any narrative material at any elementary reading level, and were focused on a thought process rather than on the specific content of a given story. Each lesson plan began with a Comerford statement of objective (Comerford and Fleury, 1972) which specified pupil learning activities as well as desired outcomes. For teachers who desired more structure, the plan 42 was continued in script form with set induction, instructional procedure, and suggestion for closure. The plans were designed to stimulate active comprehension, which Singer (1980) defines as "the continuous process of formulating and searching for answers to questions before, during, and after reading" (p. 227). The plans emphasized asking questions a reporter might ask when writing a report. This emphasis on "reporter questions" was intended to increase pupil awareness of the underlying structure of a narrative selection and to lead pupils to the basic recall units required for writing a summary. In the first lesson the concept of a summary was developed inductively by having the learners pose questions a reporter might ask when writing an account of a newsworthy event. The pupils were helped to analyze the opening paragraphs of a news article and were led to the generalization that a summary answers reporter questions, reports only the most important facts, is written in paragraph form, and uses few sentences. These criteria were written on a chart and became the focal point of succeeding lessons. Subsequent lessons were designed to follow the phase out, phase in strategy described by Singer (1980). The plans instructed the teacher to demonstrate the question asking strategy, to verbalize a thought process, and to model the process of writing a summary. The lesson sequence gradually phased out the teacher as a model and phased in the work of the student. 43 In addition to the lessons which taught the pupils to formulate questions, two lessons were planned to teach the pupils the response requirements of cognitive memory and convergent questions. Using the analogy of a target and a puzzle, the pupils were led to observe that one type of question requires a response which can be found in one word or phrase while the other requires a response which combines information from various parts of the narrative. When formulating their own questions, the pupils were encouraged to use both types of questions. In phasing out the leadership of the teacher and phasing in the activity of the learner, the pupils were led through the sequence of observing the teacher summarize a selection, participating with the teacher and the peer group in summarizing a selection, and producing a summary independently. The summaries written by individual pupils were analyzed by the group to determine if they met the criteria for a summary. The final lesson reviewed both the process and product of summarizing a selection. Data Gathering Instruments Two types of measures were constructed for this study: an objective test of ability to recognize a summary and a test of ability to produce a summary. Two forms were prepared for each of these tests. Each form was written at two reading levels, one appropriate for the pupils in the third grade sample and the other for pupils in the fifth grade sample. 44 The objective test of ability to recognize a summary statement consisted of 10 narrative paragraphs, each followed by three statements. All foils made true statements about the passage but two of the foils did not combine the details into a statement of main idea. The learner was instructed to select the best statement of main idea from the three sentences following the paragraph. Each form of the instrument constructed to measure the learner's ability to produce a summary consisted of a one-page story of 25 lines, with 325 to 350 words in each selection. The narrative selections followed the same format: setting, problem, and solution, with eight major details in each passage. The learner was instructed to read the selection and write a summary that tells the story in a short way. Reading ease was adjusted for third and fifth grade levels by controlling the selection of vocabulary items, the sentence length, and sentence complexity. Since mean and standard deviation were to be computed from test scores of pupils in the larger sample, reading ease was adjusted to meet the range in this group rather than to meet the range in the sample. The lowest reading level of third grade pupils in the larger sample was first grade level; of the fifth grade pupils in the larger sample, third grade level. Reading ease was adjusted to these levels. Vocabulary used in the third grade instruments was checked against the vocabulary lists of beginning second grade basal readers; of the fifth grade instruments, against the vocabulary lists of begin- ning third grade basal readers. Sentence length and structure for the 45 third grade instrument was patterned after the sentence length and structure in beginning second grade readers; for fifth grade against sentence length and structure in beginning third grade readers. The following sentences from testing materials used to measure ability to produce a summary illustrate the difference in reading ease. Third grade instrument: "There had been no rain for a long, long time. The water in the lake went down and down." Fifth grade instrument: "There had been no rain for a long, long time, and the water in the lake was getting lower and lower. A teacher's manual was prepared for each of these tests. The directions for administration of the objective test included instruc- tions to the pupils on the procedure for marking the computer answer sheet. Each teacher was provided with a test record form (Appendix B) to insure that half of the pupils in each group, high and low, received one form of the test while the other half received the alternate form. In addition, the investigator labeled each pupil copy with an identi— fying letter and number (e.g., H-l for the first pupil in the high group; L-l for the first pupil in the low group). Scoring In scoring the 10 item objective test, the investigator gave one point for each correct response. To score the test measuring ability to produce a summary, the investigator developed a 17 point scale. 46 The scoring key for the measure of ability to produce a summary (see Appendix C) was developed by the investigator through an analysis of summaries written by pupils in the larger sample who were not members of the pre-selected sample. While reading these summaries, the investi- gator subjectively rated them as "very good," "satisfactory," or "unsatisfactory." The rating was based on the degree to which the summaries met the criteria: written as a paragraph, answers "reporter questions," gives only the most important facts, and uses few sentences. Those rated as "very good" were analyzed and compared with those rated as "unsatisfactory" in an attempt to quantify the differences. The scoring procedure was revised four times, each revision making finer discriminations. For example, in the first attempt to devise a scale, the investigator used terms such as "includes most of the major details and only a few minor details. In the final revision, the number of major and minor details was specified. After each revision of the scale, the investigator applied the scoring procedure to at least 30 sets of summaries written by pupils in the larger sample to determine if all the variables had been accounted for. In the final revision, limiting the number of sentences was not considered in scoring even though it had been included in the criteria given the pupils. It was observed that sentence length and sentence structure in the summaries written by the pupils tended to reflect the sentence patterns in the stOries written by the investigator. The third grade passage was written in simple sentences following 47 the subject-verb pattern. The fifth grade passage used compound and complex sentences. Because of this variation in the stimulus stories, the number of sentences in the pupil summaries was not scored. The final revision produced a 17-point scale which accounted for the variables or format (list or paragraph), structure (setting, problem, solution, in sequence), inclusion of all major details, and exclusion of minor details. A maximum of two points was given for writing the summary as a paragraph; five points for including setting, problem, solution, and proper sequence; five points for including all major details; and five points for excluding minor details. A structured checklist (Appendix C) was prepared for ease in scoring pupil summaries. To check consistency in scoring, the investigator repeated the scoring of one third grade class and one fifth grade class three days after the original scoring. A compar- ison of the two checklists for each summary showed no discrepancies. After scoring all summaries, the investigator recorded the individual pupil scores on the computer record forms. In summary, ability to recognize a summary was measured with a lO-item objective test, with one point given for each correct response. Ability to produce a summary was measured by analyzing the summaries written by the pupils. These summaries were scored using a 17-point scale developed by the investigator. 48 Analysis of Data An analysis of the data collected in this study attempted to answer the questions: 1. Is there a difference between the ability of children to recognize a summary and their ability to produce a summary? 2. Is there a difference between children of different age and reading achievement levels in their ability to recognize a summary and their ability to produce a summary? 3. Will children of different reading achievement levels and different grade levels differ in terms of improvement after instruction in producing a summary? Two sets of scores were collected for each of the subjects in both the larger sample and the sample selected for this study: 1. the score on a lO-item objective test measuring ability to recognize a summary. 2. the score on a test of ability to produce a summary. Two forms of each of the two measures were used, with half of the pupils in the high reading group in each class and half of the pupils in the low reading group in each class given Form A as a pre- test and Form B as a post test. The order of the tests was reversed for the other half of the subjects. A 2)<2 ANOVA was performed to determine whether any differences in scores might be attributed to differences in the two forms of each test. Because a form difference was found on the production test, it was decided to convert the objective scores and the production scores 49 to 2 scores, with a mean of 100 and a standard deviation of 10. A new score was then constructed by subtracting the standard score on the test of recognizing a summary from the standard score on the test of producing a summary to give a difference score. For each of the tests the mean and standard deviation were computed from the set of scores of the larger sample. Mean and standard deviation were computed from the pre-test scores only, and were computed separately for each level, third and fifth. To analyze the data collected in this study, a 2x 2x 2 ANOVA (analysis of variance) was performed with third vs. fifth graders and high level readers vs. low level readers constituting the between-subjects variables, and pre-test vs. post test constituting the within-subjects variable on each of the three dependent variables: ability to recognize a summary, ability to produce a summary, and the difference score. Appropriate g.pripri_comparison tests were performed on the data to test the directional hypotheses of this study. Null Hypothesg§_ To guide the work of the investigator, the following null hypotheses were constructed for each of the three dependent variables: ability to recognize a summary, ability to produce a summary, and the difference score. 1. There will be no difference between the performance of third and fifth grade readers on each of the three dependent variables. 2. There will be no difference between the performance of high level and low level readers on each of the three dependent variables. 50 3. There will be no difference between the pre-test and post test scores on each of the three dependent variables. 4. There will be no interaction effect between the grade level factor and the high level-low level reader factor on each of the three dependent variables. 5. There will be no interaction effect between the grade level factor and the before-after instruction factor on each of the three dependent variables. 6. There will be no interaction effect between the high level- low level reader factor and the before-after instruction factor on each of the three dependent variables. 7. There will be no interaction effect among the grade level factor, the high level-low level reader factor, and the before-after instruction factor on each of the three dependent variables. Directional Hypotheses Based on the review of the literature, the knowledge of results of the teaching experience which initiated this study, and under- standing of the nature of reading, it was possible to make predictions on some of the more theoretically interesting hypotheses. According to Piaget's theory (Charles, 1974) all children go through the same stages of intellectual develOpment but different children move from one stage to another at different ages. On the average, the child at about seven years of age begins to move from the stage of intuitive thought into the stage of concrete operations. Note again that seven is an average age. Some children begin the transition earlier; others begin it later. And no child makes the transition as a clean break. . In general, we can say that the transition occurs first 51 for what we would consider simpler operations. . . . Capacities for performing concrete operations develop one by one, rather than all at once. We must realize that these new understandings are occurring at the expense of former beliefs that are not all that easily shaken off. We can expect that the transitional child, when under stress, will tend to revert to earlier ideas and behaviors. (pp. 13-14) Theoretically, it seems possible that third grade pupils in the low reading group may be in a transitional stage between preoperational and concrete operations, while third grade pupils in the high reading group may be well into the concrete operational stage. Fifth grade pupils in the low reading group may be in the concrete operational stage while fifth grade pupils in the high reading group may be moving rapidly through that stage to the stage of formal operations. Piaget describes the thinking of the preoperational child as transductive, moving from particular to particular, rather than deductive or inductive. He also describes the thinking of the pre- operational child as characterized by centration--focusing on only one aspect at a time. Since summarizing requires combining particulars into a general- ization and focusing on the whole and parts simultaneously, then it may be possible that a third grade pupil at a low reading level might not profit from instruction in summarizing as much as a third grade pupil at a high reading level. In addition, a fifth grade pupil at a low reading level might not have been ready to profit from instruction in finding a main idea when it was presented at a lower grade level, but might be ready to profit from instruction at fifth grade level because of moving to 52 a higher stage of cognitive development. It is also possible that a fifth grade pupil reading at a high level might have acquired this skill at an earlier age and would not profit much from instruction given at fifth grade level. Therefore, the following directional hypotheses were constructed: 1. After instruction in producing a summary, there will be more improvement in the performance of third grade pupils in the high reading group than in the performance of third grade pupils in the low reading group on the measure of producing a summary. 2. After instruction in producing a summary, there will be more improvement in the performance of fifth grade pupils in the low reading group than in the performance of fifth grade pupils in the high reading group on the measure of producing a summary. 3. After instruction in producing a summary, there will be more improvement in the performance of fifth grade pupils in the low reading group than in the performance of third grade readers in the low reading group on the measure of producing a summary. Summary A description of the methodology involved in conducting the study was presented in this chapter. The population consisted of third and fifth grade pupils in a school system in western Michigan whose teachers had volunteered to participate in the study. 53 Scores of the larger sample of 264 pupils in the highest and lowest reading groups in these classes were used to compute the mean and standard deviation for the tests prepared for this project. The sample consisted of 57 third grade pupils and 65 fifth grade pupils in the highest and lowest basal reading groups in eight classes in five schools. The study compared the performance of pupils in the high and low reading groups on measures of recognizing and producing summaries which were administered after they had received instruction in writing summaries of narrative materials. The procedure for obtaining the data was explained. The instruc- tional materials and testing materials were described. The method of reporting the results in the form of an analysis of variance was delineated. Finally, the research hypotheses were stated. In Chapter IV the data will be organized, presented, and discussed. CHAPTER IV PRESENTATION, ANALYSIS, AND DISCUSSION OF DATA Introduction The purpose of this study was to collect, analyze, and compare data concerning the developmental nature of children's ability to learn from text. In this chapter the statistical analysis of data related to the major research questions will be presented and discussed. The study attempted to answer the following questions: 1. Is there a relationship between the ability of children to recognize a summary statement and their ability to produce a summary statement of narrative material? 2. Do children of different ages and different reading levels differ in their ability to recognize and produce a summary of narrative material? 3. Will there be more improvement after instruction in the performance of pupils at a high reading level than in the performance of pupils at a low reading level on measures of ability to recognize and produce summary statements of narrative material? In Chapter III, null and directional hypotheses were presented for each of the dependent variables: (a) ability to recognize a ‘summary, (b) ability to produce a summary, and (c) a difference score obtained by subtracting the score on the objective test from the score 54 55 on the production test. These hypotheses will be restated and the findings will be reported and discussed. Construction of Standardized Scores Two tests were constructed by the investigator for use in this study. The tests were administered to the subjects in the larger sample of third and fifth grade pupils. Means and standard deviations of the raw scores on each of the tests used with the larger sample are presented in Tables 4 and 5. Table 4 Mean Raw Scores on Objective Test for Larger Sample Pre—test Post Test Form Grade order 'p Mean s.d. Mean s.d. Third A-B 76 6.8947 1.880 6.9342 1.982 Third B-A 78 6.5385 1.945 7.2564 2.022 Fifth A-B 55 6.4545 1.989 6.5636 1.793 Fifth B-A 55 6.7273 1.758 6.6909 2.008 Note. Maximum score = 10. Two forms, A and B, had been constructed for each test (objective and production), at each grade (third and fifth). Half of the subjects at each grade and reading level had been given Form A as a pre-test and Form B as a post test with the other half of each group given the same 56 Table 5 Mean Raw Score on Production Test for Larger Sample Pre-Test Post Test Form Grade order p_ Mean s.d. Mean s.d. Third A-B 82 8.1585 2.755 12.2561 3.292 Third B-A 72 11.2222 3.087 9.5278 2.793 Fifth A-B 60 11.1167 2.738 13.1667 2.556 Fifth B-A 50 12.1600 2.590 11.4000 2.241 Note. Maximum score = 17. tests in form order B-A. A comparison of the mean raw scores on Form A of the production test with the mean raw scores on Form B of the pro- duction test seemed to indicate that a form difference might have contributed to the difference in scores. To determine whether a form difference contributed to the difference in mean raw scores on the production test, a 21(2 ANOVA was computed for each test at each grade level (third grade objective, fifth grade objective, third grade production, and fifth grade production) (see Tables 6, 7, 8, and 9). Analysis of Variance Summary Table for Objective Test for Third Grade Pupils in the Larger Sample 57 Table 6 Source g: 115 5 Form group 1 0.0234 0.9453 Subjects X form 152 5.2614 Pre-test/post test 1 11.0429 4.5818* Pre-test/post test X form 1 8.8574 3.6750 Subjects X form X pre-test/ post test 152 2.4101 fp_< .05. Table 7 Analysis of Variaiice Summary Table for Objective Test for Fifth Grade Pupils in the Larger Sample Source g:_ M§_ .[ Form group 1 2.1992 0.446 Subjects X form 108 4.9276 Pre-test/post test 1 0.0722 0.0325 Pre-test/post test X form 1 0.2910 0.1311 Subjects X form X pre-test/ post test 108 2.2188 58 Table 8 Analysis of Variance Summary Table for Production Test for Third Grade Pupils in the Larger Sample Source g1, MS_ .5 Form group 1 2.1601 0.1840 Subjects X form 152 11.7348 Pre-test/post test 1 110.6992 l7.9145*** Pre-test/post test X form group 1 643.0625 104.0676*** Subjects X form group X pre-test/post test 152 6.1792 *** p_< .001. Table 9 Analysis of Variance Summary Table for Production Test for Fifth Grade Pupils in the Larger Sample Source 9:, M§_ f_ Form group 1 7.1367 0.7532 Subjects X form group 108 9.4745 Pre-test/post test 1 22.6953 6.4166* Pre-test/post test X form 1 107.6679 30.4410*** Subjects X form X pre-test/ post test 108 3.5369 * p_< .05. *** p_< .001. 59 Tables 6, 7, 8, and 9 show the form order group main effect, the pre-test/post test main effect, and the form order group by pre- test/post test interaction effect. This interaction effect shown for the production test is the most important finding of the four analysis of variance tests computed to discover form differences. For all subjects in both grades, production test Form A raw scores were lower than production test Form B raw scores. The subjects in form order group A-B, who received Form A of the production test as a pre-test and Form B as a post test, showed considerable gain when the post test was administered. The subjects in form order group B-A, who received Form B as a pre-test and Form A as a post test, showed a loss rather than a gain when the post test was administered (see Figures 2 and 3). $15— |.— — c —- .2 _ 4.: g —- :3 1()- ;_ _ O. | I.— I a) .— S. o — s 5_ 5 : A-B——— g — B-A ---‘- (D 1— 2 0 I l Pre-test Post Test Figure 2. Form order group by pre-test/post test interaction for third grade. 6O 15 +2 r U) a, F" l- c. 2 OS — 4; F— 810- U b- O S. 0'- f‘ I 1"" i f L. §5: 3 _ A-B -——————- CO 3: r B-A - - - - 5 t . z: 0 J Pre-test Post Test Figure 3. Form order group by pre-test/post test interaction for fifth grade. The form difference in the production test came as no surprise to the investigator because of a phenomenon observed while scoring the summary paragraphs written by both third and fifth grade subjects. The production test required the pupils to read a narrative selec- tion and write a paragraph summarizing the story. Each form of the production test provided the pupils with a one-page story containing eight major details (see Figure 4). While scoring the summary paragraphs the investigator observed that many pupils, from both third and fifth grade, perceived the problem in Story A as the storm, and the solution that the turtle stayed deep down in the water until the storm was over. Since Michigan is in a tornado belt and the schools were preparing the children for the spring tornado drills, the experience background of the pupils might have 61 Details Story 1 (A) Story 2 (B) turtle mice dock winter--needed home water low barn not get on dock mean cat storm chased mice tree down dog came near dock chased cat get on dock mice safe Figure 4. Major details in Form A and Form B of the production test. contributed to this modification of the text which resulted in lower scores for the summaries written for this passage than for those written for Story B. ' Because of the form difference that was found in the production test, the raw scores on all the tests were converted to standardized scores, using only the pre-test scores of the larger sample. The standardized scores were computed separately for the objective tests and the production tests, Form A and Form B, and for third grade subjects and fifth grade subjects. They were constructed such that the standardized score had a mean of 100 and a standard deviation of 10 (a Z score). 62 Research Questions and Statistical Analysis The subjects in the study were given pre-tests to measure their ability to recognize and produce summary statements. Treatment con— sisted of eight lessons designed to teach the subjects to produce a paragraph summarizing a narrative selection. After the treatment, post tests were administered to measure improvement on the tasks of recognizing a summary and producing a summary. A difference score was computed to compare improvement after instruction on the measures of recognizing and producing a summary. To facilitate the presentation of the results and the discussion, each dependent variable (objective score, production score, and dif- ference score) will be presented separately and each of the hypotheses discussed in relation to the dependent variable. Qgpendent Variable One: Objective Score An analysis of variance was computed to examine each null hypothesis in relation to the first dependent variable: scores on the objective measure of ability to recognize a summary statement. The results of this ANOVA are summarized in Table 10. Each of the null hypotheses presented in Chapter III will be restated for the first dependent variable: the score on the objective measure of ability to recognize a summary statement. The findings will be presented and discussed. 63 Table 10 Analysis of Variance Summary Table for Standardized Objective Score Source 2 ‘g: M§_ .[ Grade 1 1,260.5 10.5946** High/low l 1,542.0 12.9607*** Grade X high/low 1 87.0 0.7312 Subjects X high/low X grade 118 118.97 Pre-test/post test 1 45.5 0.695 Pre-test/post test X grade 1 15.0 0.2296 Pre-test/post test X high low 1 13.5 0.2066 Pre-test/post test X grade X high/low 1 3.0 0.0459 Subjects X grade X high/low X pre-test/post test 118 7,708.5 **p_'<.01. ***p_ < .001. 64 1. There will be no difference between the performance of third and fifth grade pupils on an objective measure of ability to recognize a summary statement. This hypothesis was rejected. The grade factor main effect was significant, E_(1, 118) = 10.595, p_< .01 (see Table 10). 2. There will be no difference between the performance of high and low level pupils on an objective measure of ability to recognize a summary statement. This hypothesis was also rejected. The ANOVA revealed a signif- icant main effect for the high/low level factor, E_(l, 118) = 12.96, p_< .001 (see Table 10). 3. There will be no difference between the pre-test and post test scores on the objective measure of ability to recognize a summary statement. This hypothesis was not rejected. No significant difference was found between the mean standardized scores of the pre-test and post test, §_(l, 118) = 0.696, p_> .05. 4. There will be no interaction effect between the grade level factor and the high/low level factor on the objective measure of ability to recognize a summary statement. This hypothesis was not rejected. No grade level, high/low interaction main effect was found, [_(1, 118) = 0.731, p.> .05. 5. There will be no interaction effect between the pre-test/ post test factor and the grade level factor on the objective measure of ability to recognize a summary. 65 This hypothesis was not rejected, f_(1, 118) = 0.229, p_> .05. 6. There will be no interaction effect between the pre-test/ post test factor and the high/low level factor on the objective measure of ability to recognize a summary statement. The hypothesis was not rejected. No interaction effect was found between the pre-test/post-test factor and the high/low level factor on the objective measure, f (1, 118) = 0.206, p_> .05. 7. There will be no interaction effect among the pre-test/post test factor, the grade factor, and the high/low level factor on the objective measure of ability to recognize a summary statement. The hypothesis was not rejected. No interaction effect was found among these factors, E_(l, 118) = 0.045, p_> .05. The failure to reject hypotheses 5, 6, and 7, which predict no interaction effect, is illustrated in Figure 5. Summary and discussion of findings. Analysis of data on scores on objective tests revealed that five of the seven null hypotheses could not be rejected. There was no significant main effect for the pre-test/post test factor; there were no significant interaction effects for the high/low, grade factors; the pre-test/post test, grade factors; the pre-test/post test, high/low factors; and the pre-test/post test, grade, and high/low factors. A significant main effect was found for the grade factor. Mean scores, as presented in Figure 5, seem to indicate that the third grade pupils had higher scores than the fifth grade pupils. However, it must be remembered that the standardized scores of third and fifth grade Mean Standardized Score 66 110 r- 105-— .P________.——————"""TT-"—’.——‘Tfl—flfi.. C .—._o_o—O-—O—°—'—°—°—°—'—. L— t —————————————————— O 100 — (- .OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. 95-— 1. 90 — : Third grade high ———-——— __ Third grade low -—._.- 85 C. Fifth grade high -'—-~— J Fifth grade low ------- or-A I l Pre-test Post Test Figure 5. Standardized scores on objective test. 67 subjects were computed separately, using the means and standard deviations of the pre-test scores of the larger sample of each grade. As a result the standardized scores of third and fifth grade pupils cannot be used as a basis for comparison. The difference in the test scores of third and fifth grade pupils very likely is due to the fact that the third grade sample was superior to the larger third grade sample on whose scores the standardized scores were constructed, while the fifth grade sample was more representative of the larger fifth grade sample. Almost half of the 15 third grade classes in the larger sample were from inner city schools while only one of the seven fifth grade classes in the larger fifth grade sample was from an inner city school. The high number of low achieving pupils in the larger third grade sample had an influence on the construction of the standardized scores for third grade. A similar influence was not made by the few low achieving pupils in the larger fifth grade sample. Thus, the difference in the larger samples must be considered when comparing the standardized scores of third and fifth grade pupils. There was also a significant main effect of the high/low factor. This was not an unusual finding since the teachers had placed the pupils in the high and low level reading groups on the basis of test scores and observation of response to instruction. It could be expected that the scores on high level pupils would exceed those of low level pupils. 0f more interest is the lack of a significant main effect for the pre-test/post test factor. It had been assumed that the same type of 68 reasoning-thinking skills were required for both recognizing a summary statement and producing a summary. Thus, instruction in producing a summary could improve performance in recognizing a summary. However, the results seem to indicate that instruction in producing a summary did not transfer to recognizing a summary. Dependent Variable Two: Production Score Table 11 summarizes the data obtained from an analysis of variance of the production scores. Each null hypothesis will be restated in relation to this dependent variable and the findings will be presented and discussed. 1. There will be no difference between the performance of third and fifth grade pupils on the measure of ability to produce a summary. This hypothesis was rejected. A significant grade factor main effect was found, f_(l, 118) = 17.908, p_< .001. 2. There will be no difference between the performance of high and low level pupils on the measure of ability to produce a summary. This hypothesis was rejected. For both third and fifth grade levels, the mean score of the high level pupils exceeded that of the low level pupils. A high/low factor was found to be significant, f_(l, 118) = 22.934, p_< .001. 3. There will be no difference between the pre-test and post test scores on the measure of ability to produce a summary. This hypothesis was also rejected. Post test scores were significantly higher than pre-test scores, E (l, 118) = 28.171, p_< .001. 69 Table 11 Analysis of Variance Summary Table for Standardized Production Score Source fl M_S_ _F_ Grade 1 1,555.5 l7.9089*** High/low 1 1,992.0 22.9345*** Grade X high/low 1 23.0 0.2648 Subjects X high/low X grade 118 86.8559 Pre-test/post test 1 1,376.0 28.1717*** Pre-test/post test X grade 1 28.0 0.5732 Pre-test/post test X high/low 1 228.5 4.6782* Pre-test/post test X grade X high/low 1 24.5 0.5016 Subjects X grade X high/low X pre-test/post test 118 48.8432 *p_< .05. ***p < .001. 7O 4. There will be no interaction effect between the grade level factor and the high/low level factor on the measure of ability to produce a summary. The hypothesis was not rejected, [_(1, 118) = 0.264, p_> .05. 5. There will be no interaction effect between the pre-test/post test factor and the grade factor on the measure of ability to produce a summary. The hypothesis was not rejected, E_(1, 118) = 0.573, p_> .05. No interaction effect was found. 6. There will be no interaction effect between the pre-test/ post test factor and the high/low level factor on the measure of ability to produce a summary. A significant interaction effect was found between the pre-test/ post test factor and the high/low level factor, [_(1, 118) = 4.698, p_< .05. The hypothesis was rejected. 7. There will be no interaction effect among the pre-test/post test factor, the grade factor, and the high/low factor on the measure of ability to produce a summary. No significant interaction effect was found among these factors. The hypothesis was not rejected, §_(1, 118) = 0.501, p_> .05. Summary and discussion of findings. Results of the ANOVA applied to the dependent variable, scores on the production test, showed that three of the seven null hypotheses could not be rejected. There was no significant interaction effect of the grade, high/low level factors; the pre-test/post test, grade factors; and the pre-test/post test, high/low level, and grade factors. Mean Standardized Score _ 110 s E L. 105 - 100 — 95 — 9o - p as); Orr-l Figure 6. 71 Third grade high Third grade low - -— —— Fifth grade high -:-—--- Fifth grade low -------- I Post Test Standardized scores on production test. 72 There was a significant main effect for the grade factor. Again, third grade scores appear to be better than fifth. However, as with the objective measure, this might be due to the differences in the third and fifth grade samples on which the scores were standardized. The significant main effect of the high/low level factor could be expected on the basis of the criteria used by the teachers to assign pupils to high and low level reading groups. The significant main effect of the pre-test/post test factor (see Figure 6, page 71) is of interest when compared with the same factor on the objective test (see Figure 5, page 66). While instruction in producing a summary did not appear to improve ability to recognize a summary, it did seem to influence ability to produce a summary. Of particular interest is the significant interaction effect of the pre-test/post test factor with the high/low factor when this interaction is compared with the interaction of the same factors on the objective test. While there was little improvement of either group on the post test of the objective measure, the high level pupils made significantly more improvement on the production post test than did the low level pupils (see Table 12). Dependent Variable Three: Difference Score Table 13 summarizes the data obtained from an analysis of vari- ance of the difference score which was constructed by subtracting the standardized objective score from the standardized production score. Each null hypothesis will be restated in relation to this dependent variable and the findings will be presented and discussed. 73 Table 12 Comparison of the Interaction Effect of Reading Level and Pre-test/Post test Factors on the Mean Standardized Scores on Objective and Production Tests Reading level Pre-test Post Test Difference Objective tests: High 103.1231 104.4807 1.3576 Low 98.5004 98.8757 0.3753 Production tests: High 102.5536 109.3555 6.8019 Low 98.7130 101.5712 2.8582 1. There will be no difference between the performance of third and fifth grade students on the measure of the difference between scores on the production test and objective test. The hypothesis was not rejected. The mean difference score for third grade subjects was 2.0608; for fifth grade subjects, 1.5460. The grade factor main effect was not significant, E_(l, 118) = 0.0964, p_> .05. 2. There will be no difference between the performance of high and low level students on the measure of the difference between scores on the production test and objective test. This hypothesis was not rejected. The mean difference score for high level students was 2.1527 compared with 1.4541 for the low level students. No significant high/low level main effect was found, _F_ (l, 118) = 0.1776, 11> .05. Table 13 Analysis of Variance Summary Table for Standardized Difference Score Source 511 M_S_ f. Grade 1 15.6195 0.0964 High/low 1 28.7716 0.1776 Grade X high/low 1 20.8160 0.1285 Subjects X high/low X grade 118 161.9790 Pre-test/post test 1 926.0653 7.3455** Pre-test/post test X grade 1 1.6136 0.0127 Pre-test/post test X high/low 1 129.2418 1.0251 Pre-test/post test X grade X high/low 1 11.9692 0.0949 Subjects X high/low X grade X pre-test/post test 118 126.0719 **p_< .01. 75 3. There will be no difference between the pre-test and post test scores on the measure of difference between scores on the production test and objective test. This hypothesis was rejected. The mean difference score on the pre-test was -0.1785; on the post test, 3.7852. There was a signifi- cant pre-test/post test factor main effect, {_(1, 118), 7.3455, Ef3.01. This difference is the result of an increase in the post test score of the production test. While there was little change between the mean pre-test and post test scores on the objective measure, there was a large increase in the mean post test score on the production measure (see Tables 14 and 15). Table 14 The Difference Between Mean Pre-test and Post Test Standardized Scores on the Objective Measure Mean standardized score Grade Pre-test Post Test Difference Third 102.8598 104.2521 1.3923 Fifth 98.7637 99.1042 0.3405 76 Table 15 The Difference Between Mean Pre-test and Post Test Standardized Scores on the Production Measure Mean standardized score Grade Pre-test Post Test Difference Third 102.8559 108.3774 5.5215 Fifth 98.4106 102.5494 4.1388 4. There will be no interaction effect between the grade level factor and the high/low factor on the measure of difference between scores on the production test and the objective test. -This hypothesis was not rejected. There was no significant grade level, high/low interaction effect, {_(1, 118) = 0.1285, p_> .05. 5. There will be no interaction effect between the pre-test/post test factor and the grade factor on the measure of difference between scores on the production test and the objective test. The results did not yield a significant interaction effect between the pre-test/post test and grade factors, [_(1, 118) = 0.573, p_> .05. The hypothesis was not rejected. 6. There will be no interaction effect between the pre-test/post test factor and the high/low level factor on the measure of the differ- ence between scores on the production test and scores on the objective test. 77 The hypothesis that there would be no significant interaction effect of pre-test/post test, high/low factors was not rejected, {_(1, 118) = 1.0251, p_> .05. 7. There will be no interaction effect among the pre-test/ post test, grade, and high/low level factors on the measure of difference between scores on the production test and the objective test. This hypothesis was not rejected. Results did not yield a significant interaction effect of the pre-test/post test, grade, and high/low level factors, [_(1, 118) = 0.0949, p.> .05. While no group made a significant gain on the objective measure, both third and fifth grade pupils, and both high and low level pupils, made substantial gains on the production test which resulted in a large difference score for all subjects. (The difference score was obtained by subtracting the standardized score on the objective test from the standardized score on the production test). The lack of pre-test/post test, grade, high/low interaction is demonstrated in Figure 7. Summary and discussion of findings. The hypotheses related to the last dependent variable, the difference score, were formulated in an attempt to explore the relationship between performance on the objective measure with the performance on the production measure. Since the objective and production tests were two different types of measures, the mean scores on the two tests cannot be compared. The relationship between performance on these two measures may best be examined by comparing the difference between pre-test and post test Difference Score 78 Third grade high -2 — Third grade low -—--— - Fifth grade high —----- Fifth grade low --°°°°° l I Pre-test Post Test Figure 7. Mean difference between standardized scores on objective and production measures. 79 scores on each of these measures as presented in Figures 5 and 6. The lines connecting pre-test and post test scores on the objective measure are almost flat. The lines connecting pre-test and post test scores on the production test show a sharp rise. This phenomenon is reflected in the significant main effect of pre-test/post test for the difference score. This was the only significant main effect or interaction effect found using the difference score. Directional Hypotheses and Statistical Analysis Based on a review of the literature, knowledge of the results of a teaching experience which initiated this study, and an understanding of the nature of the reading process, the investigator formulated directional hypotheses as alternatives to three of the null hypotheses relating to the dependent variable: scores on the production test. Appropriate planned comparison tests were performed on the data to test these directional hypotheses. Each of the directional hypotheses from Chapter III will be restated and the findings presented and discussed. 1. After instruction in producing a summary, there will be more improvement in the performance of third grade pupils in the high reading group than in the performance of third grade pupils in the low reading group on the measure of producing a summary. A planned comparison (see Figure 8) was made between the pre-test standardized production scores and the post test standardized production scores for third grade high students. The comparison demonstrated a significant improvement, f_(l, 118) = 20.0268, p_< .05. A similar 80 110 —l O 01 IIITIIIITTI 100 lllrlll to 01 1111 Standardized Production Scores £0 0 TFITTTTI Third grade high Third grade low — ---— 1 1 Pre-test Post Test in Figure 8. Standardized production scores for third grade high and low students. 81 planned comparison was made for third grade low students. The comparison was not significant, [_(1, 118) = 2.5296, p_> .05. Thus, support was found for the directional hypothesis that third grade high students would demonstrate more improvement on their production scores than third grade low students. 2. After instruction in producing a summary, there will be more improvement in the performance of fifth grade pupils in the low reading group than in the performance of fifth grade pupils in the high reading group on the measure of producing a summary. The planned comparison (see Figure 9) made between the pre-test standardized production score and the post test standardized production score for fifth grade high students demonstrated a significant improve- ment, {_(1, 118) = 8.9834, p_< .05. A similar planned comparison was made for the fifth grade low students. This comparison was not sig- nificant, [_(l, 118) = 2.4008, p.> .05. Support was not found for the directional hypotheses that more improvement on the production measure would be made by fifth grade low than by fifth grade high students. On the contrary, exactly the opposite was found. The improvement for fifth grade high students was significant, for fifth grade low students it was not. 3. After instruction in producing a summary, there will be more improvement in the performance of fifth grade pupils in the low reading group than in the performance of third grade pupils in the low reading group on the measure of producing a summary. The planned comparisons (see Figure 10) made between the pre-test and post test standardized production scores for the low third grade Standardized Production Scores 82 l. 110 — r— L _ 0,..- 105 2. Km" r .,-" : 0'"! 100 — L L“ 000...... r- .Oooooo'°..... 95 — l... t 90 F : Fifth grade high —--—--—- — Fifth grade low -~----- 85 — j, 0e . . Pre-test Post Test Figure 9. Standardized production scores for fifth grade high and low students. Standardized Production Scores 83 1 1 l 110 11111 105 100 ITTIITI q 95 111111 90 Third grade low ———- «— Fifth grade 10w 000000000 85 I VTTTIFI I 1 Pre-test Post Test °—1—\ Figure 10. Standardized production scores for third grade low and fifth grade low students. 84 pupils and for the low fifth grade pupils did not demonstrate a significant improvement for either group [third grade low, E_(1, 118) = 2.5296, p_> .05; for the fifth grade low, {_(1, 118) = 2.4008, p_> .05]. Thus, no support was found for the directional hypothesis that improvement of fifth grade low level pupils on the production measure would exceed that of the low level third grade pupils. Summary Chapter IV described the standardized scores, presented the research questions along with the statistical analysis, and summarized the findings for the directional hypotheses. In Chapter V the major results of the study, implications, and recommendations for further research will be presented. CHAPTER V CONCLUSIONS AND RECOMMENDATIONS Introduction This study eXplored the developmental nature of children's ability to learn from text by comparing the performance of students in high and low level reading groups in third and fifth grade on measures of recognizing and producing summary statements of narrative material. Answers were sought to the following questions: 1. Is there a difference between the ability of children to recognize a summary statement and their ability to produce a summary? 2. Is there a difference between children of different age and reading levels in their ability to recognize and produce summary statements? 3. Will the amount of improvement after instruction be different for children of different age and reading levels on measures of ability to recognize and produce summary statements? These research questions were related to practical problems faced by elementary teachers: 1. Can decisions be made about children's ability to extract meaning from text on the basis of performance on objective tests which require students to select a summary statement? 85 86 2. Can a decision about the timing of instruction in a given reading skill be made on the basis of when the skill is tested in the reading program? In this chapter the study will be summarized, the major findings related to the null and directional hypotheses will be presented, implications for classroom instruction will be discussed, and recommendations for further research will be made. Summary of Study The population consisted of third and fifth grade pupils from a large (18,000+) metropolitan school system. Tests prepared by the investigator were administered to a sample of 264 subjects. Ability to recognize a summary was measured with a lO-item objective test; ability to produce a summary was measured by having the students write a paragraph summarizing a one-page narrative passage. After a form difference was found in the two forms of the produc- tion test (writing a summary), the means and standard deviations of the scores of pupils in the larger sample of 264 subjects were used to standardize the scores of the smaller sample used in this study. Null hypotheses were constructed for each of the three dependent variables: ability to recognize a summary, ability to produce a summary, and a difference score (computed by subtracting the score on the objec- tive measure from the score on the production measure). Directional hypotheses were constructed for three of the hypotheses relating to the production variable. 87 To analyze the data collected in this study, a 2x 2x 2 ANOVA was performed with third vs. fifth graders and high vs. low level readers constituting the between-subjects variable, and pre-test vs. post test constituting the within-subjects variable on each of the three dependent variables. Appropriate g_pyjp§j_comparison tests were performed on the data to test the directional hypotheses of this study. Significant main effects were found for the grade factor on the objective test (Ef3.01) and production test (p;=.001); for the high/low factor on the objective test (Ef3.001) and production test (Ef<.001); and for the pre-test/post test factor on the production test (Ef1.001) and difference score (Ef3.01). The only significant interaction effect found was for the pre-test/post test by high/low factors on the production measure (Ef1.05). The planned comparison tests on the directional hypotheses found support only for the hypotheses that third grade high students would demonstrate more improvement on their production scores than third grade low students. All findings are summarized in Figures 11 and 12. 88 Dependent variable Objective Production Difference Null hypotheses test test score . There will be no difference Rejected Rejected Not between third and fifth p_ < .01 p_<:.001 rejected . There will be no difference Rejected Rejected Not between high and low p <:.001 p_< .001 rejected . There will be no difference between pre-test and post Not Rejected Rejected test rejected p_< .001 p_< .01 . There will be no interaction effect between grade and Not Not Not high/low factors rejected rejected rejected . There will be no interaction effect between pre-test/ Not Not Not post test and grade factors rejected rejected rejected . There will be no interaction effect between pre-test/ Not Rejected Not post test and high/low rejected p_< .05 rejected . There will be no interaction effect among grade, high/low . Not Not Not and pre-test/post test rejected rejected rejected Figure 11. Summary of research findings for null hypotheses. 89 Directional Production hypotheses test 1. Third grade high pupils will improve more than third grade low pupils Supported 2. Fifth grade low pupils will improve more than fifth grade high pupils Not supported 3. Fifth grade low pupils will improve more than third grade low pupils Not supported Figure 12. Summary of research findings for directional hypotheses. Major Results and Discussion The results for the null and directional hypotheses will be presented and discussed separately. Null Hypotheses Grade level. The grade level main effect was significant for both the objective and production tests. The mean standardized objective score for the third grade sample was 103.5559; for the fifth grade sample, 98.9339. The mean standardized production test score for the third grade sample was 105.7167; for fifth grade, 100.4799. However, it cannot be concluded that there is a difference in the ability of third and fifth grade students to recognize and produce a summary. The difference can be accounted for in terms of the difference in the larger samples whose scores were used to standardize scores of the smaller sample. The third grade sample was superior to its larger sample, which contained many low achieving students from inner city 90 schools. The fifth grade sample was more representative of its larger sample, which contained only four low achieving students from one inner city school. Additional evidence that there was no true difference between third and fifth grade subjects is found in an analysis of the difference score, for which no significant main effect was found. For the third grade sample, the difference between the mean standardized scores for the objective and production tests was 2.0608; for the fifth grade sample, 1.5460. Highllow. A significant main effect was found for the high/low factor on both the objective and production tests. The mean standard- ized objective score for the high students was 103.8018; for the low students, 98.6880. 0n the production test, the mean standardized score for the high students was 105.9545; for the low students, 100.4779. This was not an unusual finding since the teachers had assigned pupils to high and low reading groups on the basis of test scores and response to instruction. Again, no significant main effect was found for the high/low factor on the difference score. The difference score for high level pupils was 2.1527; for low level, 1.4541. Pre-testlpost test. The findings on the pre-test/post test factor are of particular interest. None of the groups (third or fifth, high or low) made significant gains on the post test of the objective test (see Figure 5, p. 66). All of the groups made significant improvement on the production post test (see Figure 6, p. 80). 91 It had been assumed that the same reasoning-thinking skills were involved in both recognizing a summary and producing a summary. Thus, instruction in producing a summary could have resulted in an increase in the post test scores on the test of recognizing a summary. However, in this study, the instruction given in producing a summary did not bring about a change in the objective test score. A significant main effect was found for the pre-test/post test factor on the difference score. The mean difference score between the production and objective pre-tests was -0.1785; between the post tests, 3.7852. This difference is the result of the increase in the post test scores of the production test. Interaction effects. The only significant interaction effect found was for the pre-test/post test by high/low factor on the pro- duction test. The high level pupils made significantly more improve- ment on the production posttest than did the low level pupils (see Table 12, page 73). This factor will be discussed in more detail when the directional hypotheses are presented and discussed. Directional hypotheses. The three directional hypotheses, formulated for the production test, were stimulated by a report from a third grade teacher who, in a pilot study, had used the lesson plans and teaching materials prepared by the investigator. He reported that the pupils in the highest reading groups had profited from instruc- tion in writing a summary but those in the lowest reading groups had not developed a concept of a summary paragraph. 92 This led to speculation that ability to summarize might be related to the pupil's stage of cognitive development. Summarizing requires the reader to focus simultaneously on both the main idea and details of a passage. It might be possible that less skilled readers in third grade would be in transition from the preoperational-intuitive thought stage to the concrete Operational stage. As such, they might still be at a stage of cognitive development in which they would find it diffi- cult to decentrate--to focus on the whole and the parts simultaneously. 0n the other hand, the highly skilled readers at third grade level might be well into the concrete Operational stage and moving toward the stage of formal operations. As such, they should be able to profit from instruction in summarizing. It was hypothesized that the highly skilled fifth grade pupils would be well into the concrete operational stage and would already be skilled in producing summaries. Therefore, their test scores would show little improvement. The less skilled fifth graders might now be moving into the stage of concrete operations and would be ready to benefit from instruction in summarizing. As a result, their test scores could show much improvement. This reasoning led to the construction of three directional hypotheses as a substitute for those presented for the dependent variable, the production test: (1) third grade high level pupils would show more improvement on production test scores than third grade low level pupils; (2) fifth grade low level pupils would show more improvement than fifth grade high level pupils; and (3) fifth grade 93 low level pupils would show more improvement than third grade low level pupils. The planned comparison tests rejected two of these three hypoth- eses. The fifth grade high level readers showed significantly more improvement on the production test scores than did the fifth grade low level readers. The comparison between the standardized production scores on the pre-test and post test for the fifth grade high level students demonstrated a significant improvement, {_(l, ll8) = 8.9834, p_< .05. For the fifth grade low level students, the comparison was not significant, [_(l, ll8) = 2.4008, p_> .05. Thus, the hypothesis that fifth grade low level pupils would show more improvement on the production test than the fifth grade high level pupils was rejected. A similar planned comparison was made for low level third grade pupils. The comparison was not significant, {_(l, ll8) = 2.5296, p_> .05. Thus, the hypothesis that fifth grade low level pupils would show more improvement than third grade low level pupils was rejected since neither demonstrated a significant improvement. The directional hypothesis that third grade high pupils would show more improvement than third grade low level pupils was not rejected. The planned comparison made between the pre-test standard- ized production scores and the post test scores for third grade high level pupils demonstrated a significant improvement, {_(l, ll8) = 20.0267, p_< .05. However, in the context of the findings for all three directional hypotheses, this finding is not important. 94 While the greater improvement of the third grade high students over third grade low students on the standardized production test scores supports the theory that differences in stages of cognitive development might contribute to differences in the achievement of high and low level pupils, the other results of the planned comparisons do not support this theory. This does not necessarily mean that differences in stages of cognitive development do not contribute to differences in achievement of pupils reading at high and low levels. A variety of factors may have influenced these findings. Perhaps the range between low level third and fifth grade pupils was not wide enough to permit differences to emerge. Differences might have been noted if comparisons had been made between low level third grade pupils and those at sixth or seventh grade. It is possible that the significant improvement of fifth grade high level pupils over the low level pupils may have been related to instructional procedures. It was assumed that, since summarizing is tested at the fourth grade level on the MEAP test, fifth graders had received instruction in this skill. However, as Bader (l979) noted when discussing pupil performance on tasks relating to main idea, some pupils are tested, retested, and drilled on skills which appear on tests, but have never had the concept explained to them. It is possible that fifth graders had not previously been given direct instruction in summarizing so that fifth grade high level students had not mastered this skill in an earlier grade. It is also possible, 95 since MEAP test items are of the objective type, prior instruction prepared pupils to respond to objective test items but had not prepared them to produce a summary. However, if instruction in producing a summary had not been given in earlier grades, then the improvement in the scores of the fifth grade high level students would be expected. Limitations Conclusions cannot be made without considering the many limitations of the study. When the investigator decided to do this study in a naturalistic setting, the decision was made with a full awareness of all the difficulties which might be encountered: teacher and pupil absenteeism, teaching priorities, and the many interruptions which are a normal part of life in any elementary classroom. All of these fears were realized--and a few problems were encountered which had not been anticipated. In some schools the teaching schedule had to be adjusted to fit special school assemblies. In all schools the research schedule was interrupted by a snow day--which extended the schedule into the week set aside for parent-teacher conferences. In one school, a teacher's bout with the flu delayed the project. In many schools the schedule was delayed because of high pupil absentee- ism. In one class eight pupils were sent home with the flu on the day originally scheduled for a pre-test. Subjects were lost when teaching priorities did not permit rescheduling a test for pupils who had been absent. Other subjects 96 were lost when teachers rearranged the class list to delete pupils who transferred out or to add pupils who transferred in, with the result that several pupils were given the same form as pre-test and post test. In one class all the pre-test data on producing a summary was lost when the little girl who was "paper passer" for the day picked up the summary paragraphs and distributed them with the other papers to be taken home. Consideration should also be given to the limitations which might result from the use of two very different types of tests. It may be that the production test was more sensitive to pupil improvement since the scoring procedure recognized even slight changes in pupil perfor- mance while the scoring for the objective test recognized only right and wrong answers. As indicated in Chapter I, there are also limitations due to the limited number of subjects, the lack of control, and the absence of sampling techniques. Implications for Classroom Instruction One question raised by the results of this study is that of the degree of confidence that can be placed in the scores on objective tests, or what kinds of decisions can be made on the basis of scores on objective tests. In this study, if only the objective test had been administered at the end of the eight instructional sessions, the conclu- sion would have been reached that no learning had taken place (see Figure 5, page 60). In reality, as presented in Figure 6 (page 7l), 97 all groups had profited from instruction. Decisions based on objective test scores should be made only after a careful study of what the score tells about the performance of the pupil. Related to this is the question of how teachers evaluate pupil learning. If one sees only the spread between the high and low level student (see Figure 6, page 7l), there is the likelihood of developing the mental set of viewing these pupils as good and poor readers. How- ever, if one views the curve between the pre-test and post test scores of the low level students (Figure 6, page 7l), the mental set changes. The low level pupils are learning, but at a slower rate. Adjusting the instruction to the rate of learning of the low level pupil, and making this pupil aware of the growth he or she is making, could help both teachers and pupils to eliminate the labels of "good" and "poor." The results of the study also suggest that children learn what they are taught. If this is the case, teachers should examine the reading curriculum and reading materials to determine the degree to which the reading program is preparing pupils for experiences outside the school setting. Many instructional materials limit pupils to selecting correct answers, matching statements, and writing brief answers to literal level questions. Outside the school setting, when the reader must extract meaning from print, he is not given choices of responses. He must note relationships, sort out the relevant and irrel- evant details, and organize his statement of meaning. Teachers should consider putting less emphasis on fill-in-the-blank materials and more emphasis on helping children become book-thinkers who organize their own responses. 98 The study also suggests that all pupils are not necessarily ready for instruction in a given skill at the time that it will be tested in the reading program. Since the MEAP tests include items on sum- marizing at the fourth grade level, it may be assumed that such instruction should be given to all pupils at the third grade level. However, this study found that many third grade pupils at a low reading level did not benefit from such instruction. Teachers should not feel compelled to include all pupils in specific skill instruction if they have reason to believe that a few of the pupils are not ready to benefit from such instruction. Nor should teachers hesitate to take such a decision-making role. Although it was not the intent of the study to evaluate the effectiveness of teachers as decision makers, the results of the study suggest that teachers do well in this role as evidenced by the decisions they made when placing pupils in reading groups. While there were exceptions, for the most part those pupils placed in high groups did achieve at a high level; those in low groups, at a lower level. Recommendations for Further Research The findings and the limitations in the findings of this study suggest the need for further research in several areas. A question posed in this study was whether reading achievement might be related to stages of cognitive development. Results of the study did not permit any conclusions to be reached concerning this relationship. Refinement of the study using a wider range of grade levels or changing the response from a written to an oral form might provide some answers. 99 There is no evidence that the improvement in production scores was the result of instruction. A change in design to include a control group might answer this question. The study provides evidence that instruction in producing a summary does not influence performance in recognizing a summary but says nothing about whether instruction in recognizing a summary will influence per- formance in producing a summary. Since concern about MEAP test scores leads many teachers to provide instruction in taking objective tests which require the student only to recognize a summary, it would be of interest to determine whether such instruction is preparing pupils to function outside the school setting where they must produce a summary. Based on these observations, the following recommendations are made for further research. 1. The study should be replicated using a wider range of grade levels, comparing performance of third grade pupils with that of pupils in grade six or seven. 2. The study should be replicated with low level students, providing them with options such as dictation or the use of a tape recorder to negate any constraints that might have been imposed by the requirement of a written response. 3. The study should be replicated using a control group which would not receive instruction to determine whether improvement in production scores was related to the treatment. 4. The study should be replicated substituting instruction in recognizing a summary for instruction in producing a summary to lOO determine whether such instruction would transfer to performance in producing a summary. The findings and the limitations in the findings also suggest that research in other areas is needed to account for the differences between readers who perform at high and low levels. The results of this study do not permit explaining the differences in terms of stages of cognitive development. It is possible that Piaget's comments about reading readiness apply to other stages of reading develOpment as well. Piaget is quoted as stating, The concept of readiness is not bad but I am not sure that it can be applied to reading. Reading aptitude may not be related to mental age. There could easily be a difference of aptitude between children independent of mental age. But I cannot state that as a fact because I have not studied it closely. (Hall, 1970, p. 30) The response of readers to the story in Form A of the production test, in which the pupils modified the text to conform to their expe- rience, seems to support the statement of Pearson and Johnson (l978) that "experience is at the heart of comprehension, especially as comprehension is assessed in today's schools" (p. 44). The findings of this study suggest that research in schema theory may be a more fruitful area to explore in attempting to account for the differences between high and low level readers. APPENDIX A INSTRUCTIONAL MATERIALS OVERVIEW OF INSTRUCTIONAL PROCEDURES Pre-Tests l. 2. Administer an objective test of ability to recognize a summary. Have the pupils write a paragraph summarizing a previously read basal selection. Instructional Episodes 1. 2. 3. 7. 8. Introduce the concept of a summary through analysis and discussion of a news article. Review the criteria for a summary statement and model a summary for the pupils. Provide a group experience in writing a summary by (a) reviewing the criteria for a summary, (b) assisting the pupils in analyzing a previously read basal selection, and (c) guiding the dictation of a summary of that selection. Teach the response requirements of convergent questions and relate convergent questions to the process of summarizing. Review the response requirements of convergent questions, provide individual practice in formulating convergent questions, and guide group evaluation of convergent questions written by individual pupils. Review the criteria for a summary, analyze a previously read basal selection, and provide individual practice in writing a summary of that selection. Guide group evaluation of summaries produced in session 6. Review the process and product of summarizing a basal selection. Post-Tests l. 2. Administer the alternate form of the objective test of ability to recognize a summary. Have the pupils write a summary of a previously read basal selection. INSTRUCTIONAL GUIDELINES Administer objective test Form A to one-half of the pupils in each group, and Form B to the other half of the pupils in each group (high and low instructional level). As a post test, give each pupil the alternate form. This will minimize the possibility that any difference which may be found was a result of the instrument rather than a difference in pupil performance. Use a different basal selection for each of the instructional episodes 2, 3, and 6. During each of the instructional episodes the learners should be led through the criteria for a summary. They should be encouraged to refer to the visual aids. lOl Reporter Questions Who? What? When? Where? Why? How? With what result? Newspaper Article Target Questions ask you to find the exact words that answer the question Puzzle Questions ask you to put together information from several places in the story Figure A.l Visual aids to accompany lessons 1 and 4. Lesson l: A. 103 LESSON PLANS FOR INSTRUCTIONAL EPISODES Introducing the concept of a summary Comerford Statement To introduce the concept of a summary paragraph by having the learner: l. 2. 3. 4 so that state the questions a news reporter might want to answer when he writes a news report listen to the reading of the introductory paragraph(s) of a newspaper article state which questions were answered in the article count the number of sentences in the introductory paragraphs the learner can define a summary as answers to reporter questions, written in paragraph form, using as few sentences as possible, and relating only the most important facts. Procedure I. Set induction "When a reporter takes notes on a fire, accident, or other newsworthy event, he tries to answer questions he thinks the reader or listener might want answered. Suppose you were reading about a fire in our city. What questions would you want the reporter to answer?" (If necessary, the teacher asks probing questions to elicit the words: who, what, where, when, why, how, with what result. As the pupils supply the words, the teacher lists them on the board under the heading "Reporter Questions.") Presentation "As I read the beginning of this news article, listen for the answers to the reporter questions." (Hold up file folder with reporter questions and news article.) "Which questions were answered? What is the answer?" (As pupils supply the answers, list them on the board next to the appropriate question word, limiting response to a word or phrase as you record answers.) "We have answered many of the questions. The answers were found in just these few sentences." "As I read these sentences again, I will pause at the end of each sentence. Count the number of sentences the reporter used to answer all these questions.“ How many? (Write number supplied by pupils.) l04 "See how little of the article I have read." (Point) "A reporter tries to answer these questions in the first part of his article. We call this a summary. The reporter uses this summary to tell you the most important facts in the shortest way possible." Closure "Today we learned what a summary is. Look at the helps on the file folder to answer my questions. What questions does a summary try to answer? How many sentences did the reporter use in his summary? Did he number the answers to the ques- tions or write them in a paragraph? Did he give all the facts or just the most important ones?" "Using your own words, who can tell us what a summary is?" (If necessary, use probing questions to elicit the criteria: answers to reporter questions, written in as few sentences as possible, written in paragraph form, giving only the most important facts.) (Prepare the following chart and keep it visible during each of the succeeding lessons. Encourage the learners to refer to it as often as necessary.) A Summary answers most of the reporter questions uses as few sentences as possible is written like a story tells only the most important facts th-d Lesson 2: A. 105 Modeling a summary Comerford Statement To demonstrate the process of writing a summary by having the learner: l. 2. 3. so that supply the question words used by a reporter supply a word or phrase from a basal selection which answers each question watch and listen as the teacher demonstrates how to combine these words and phrases into a summary the learner can state orally the criteria for a summary, and describe orally the procedure to follow in writing a summary. Procedure I. Set induction , "With your help, I am going to write a summary of the story in our reader, (give name and_page number). If you need to, look at the file folder to answer my questions." (Mount folder with reporter questions and news article. Encourage learners to refer to it as often as necessary.) Presentation "To get ready to write a summary of this story, I have to find answers to the reporter questions. Give me the ques- tion words so I can write them on the board. If you forgot some of them, look at the file folder." (Write "Reporter Questions" on the board as a label. Beneath write the question words as they are supplied by the learners.) "Now help me find the answers. Let's start with the question who, Look through the story in your reader and tell me what to write after who," (Write words or phrases supplied by the learners. Keep the response as brief as possible. Continue until the questions have been answered. In some cases, it will not be possible to answer all the questions.) "Now I am ready to write a summary. How many sentences did the reporter use? Did he number the questions and write answers, or write in paragraph form?" "I will try to answer these questions in just a few sentences. I will begin with whg_and what and see how many facts I can fit into one sentence. l06 (As you write your summary, point out that you are using the words and phrases that they gave you in response to the reporter questions. Check off each of the questions as you incorporate the response in your summary.) Closure "Let's see if I have written a summary. Did I answer most of the reporter questions? Did I use just a few sentences? Did I write it as a paragraph? Did I tell only the most important facts?" "Who can tell what I did to write a summary?" (Use probing questions as needed to elicit the statements: listed reporter questions, found the answers in the story, wrote in paragraph form, used as few sentences as possible, told only important facts.) 107 Lesson 3: Providing a group experience in writing a summary A. Comerford Statement To provide practice (group) in writing a summary by having the learners: 1. state the question starters for teacher to write on board 2. participate in an analysis of a previously read basal selection to locate answers to reporter questions 3. respond to questions about the criteria for a summary so that the learners can dictate to the teacher a summary of the basal selection which incorporates all the answers listed on the board. Procedure 1. Set induction "Yesterday you helped me write a summary. Today I am going to help you write a summary of a different story in our reader. Please turn to (story title andpage);f 2. Presentation "Look at our file folder and tell me what we are going to do first." (If necessary, use probing questions to elicit: listing the reporter questions, locating the answers in the story, listing the answers on the board next to the appropriate question word.) "Now that we have the reporter questions and answers, we are ready to put them together in a summary. In what form should we do it? How many sentences should we use? Which facts should we tell? What can we do to our list on the board to make sure we are using all the answers?" "Talk to each other for a few minutes. Try out some sentences. Now, who will give me a sentence to begin with?" (As sentences are dictated, have pupil check off facts on list on board.) 3. Closure (After group has dictated the summary, have the learners state the criteria for a summary and evaluate the summary on the board against the criteria.) Lesson 4: 108 Teaching response requirements of convergent questions A. Comerford Statement To teach the response requirements of convergent questions by having the learners: 1. look and listen as the teacher presents a chart explaining the different response requirements of cognitive memory (target) and convergent (puzzle) questions under teacher direction, orally formulate convergent questions based on a previously read basal selection state (as the teacher records on the board) the page and line numbers where words or phrases provide class to the answers to these questions count the number of references for each question to insure that clues to the response are found in two or more places in the basal selection so that the learners can describe convergent questions as questions which are answered by combining clues from two or more parts of the reading selection. B. Procedure 1. Set induction "When we wrote a summary, we started with reporter questions. Today we're going to ask two kinds of reporter questions: (show file folder) target questions and puzzle questions." Presentation "First, I'll ask a target question. A target question asks you to look for just one word or phrase that tells the answer. (Point to folder.) See if you can find just one word that answers this question." (Ask a question that has just one correct answer, that can be answered with one word or phrase.) "This time I'll ask a reporter question that is a puzzle question. Look at the pieces at the tap of this folder. t is not a picture now, but if I could take the pieces off and put them together, it would be a picture of ." (Wait for pupils to supply word.) "Puzzle questions do not have an answer that can be found in one word or phrase. The answer is in the story, but you have to find pieces, or clues, in several parts of the story. Here is a reporter question that is a puzzle question. See how many places in the story give you a clue, or a piece of, the answer. When you find a clue, don't tell me the answer. Just give me the page number and the line on which you found the clue." 109 (Write on the board a question which is answered in the story but which requires the learner to put together information from several places in the story. An example might be, "How do you know the bear had a hard time finding something to eat?" As the children find clues, write down the page number and line number. After two or three chil- dren have contributed, have all members of the group look up these references. If time permits, write another question on the board and repeat the procedure.) Closure "There are two kinds of reporter questions. What are the two kinds? What do you do to find the answer to a target question? What do you do to find the answer to a puzzle question?" (Have the file folder chart visible. Encourage the learners to refer to it if necessary.) Lesson 5: 110 Providing individual practice in writing convergent questions A. Comerford Statement To provide individual practice in writing convergent questions by having the learner: 1. 2. respond orally to questions about the characteristics of puzzle questions listen to the teacher presentation of the relationship of puzzle questions to the thought process involved in writing a summary write one or more convergent questions, based on a previously read basal selection, using the question words of "reporter questions" as question starters share these questions with the group and participate in a search of the basal selection to determine whether the clues to the answer are found in more than one place in the basal selection so that the learner is able to describe a convergent question and can use convergent questions to guide the writing of a summary. B. Procedure 1. Set induction "Today we are going to be reporters asking puzzle questions about a story to help us get ready to write a summary. Who can tell us what a puzzle question is?" (Have the file folder visible. Encourage use of it.) "Let's turn to the story on page (This may be any story which all members of the group have read and with which they are still familiar.) "Write a puzzle question on your paper. Underneath the question, write the number of the page and the line on which you found a clue to the answer. Remember, to answer a puzzle question, you must put together information from several parts of the story." (Give help as needed. After each pupil has written at least one puzzle question, collect the papers and call the group together. Without identifying the writer, put one of the questions on the board. Have the pupils find clues and give the page number and line number of the clue. Have the pupils decide whether the question is a uzzle question. Have them explain why it is or is not.) 111 Closure "Sometimes when we get ready to write a summary, we ask ourselves puzzle questions. Who can explain what a puzzle question is?" (Have file folder visible. Encourage use, if necessary.) Lesson 6: A. 112 Providing individual practice in writing a summary Comerford Statement To provide individual practice in writing a summary paragraph by having the learner: l 2: 3. 4 give orally the major features of a summary give orally the "reporter questions" supply answers to reporter questions based on a previously read basal selection watch as the teacher writes these answers on the board so that the learner, working independently, can write a summary which combines the information listed next to the reporter questions on the board. Procedure 1. Set induction "Now that we know what a summary is, we are going to try to write a summary of the story" Presentation "Before we write a summary, who can tell us what a summary is?" (All charts or file folders should be visible: "reporter questions," "target and puzzle questions," and "A summary is. . . ." "If you will tell me the reporter questions, I will write them on the board." "Using your readers, give me any answers you find." (List answers next to question words. The answers should be given in a single word or short phrase.) "Now you are ready to write a summary. Using the answers we have on the board, write the story in a short way. You will probably each have a different way of writing the summary, but you should all have the most important facts, using as few sentences as possible." Closure "If you have finished writing a summary, read it over and see if you have remembered to write it the way our chart tells us to." (Point to chart: "A summary is. . . .") 113 Lesson 7: Guiding group evaluation of summaries A. Comerford Statement To strengthen awareness of the criteria for a summary by having the learner: 1. state the criteria for a summary 2. listen to the teacher read two or more of the summaries written by pupils in lesson 6 3. respond to questions about the content of these summaries so that the learner will be able to tell whether or not a paragraph meets the criteria for a summary, and will be able to suggest additions or deletions for summaries read by the teacher. Procedure 1. Set induction "When I wrote a summary of a story, you helped me decide whether I had remembered all the rules for writing a summary. Today we will help each other with the summaries you wrote." 2. Presentation "Help me finish this: 'A summary is. . . .'" (Write the statements on the board as pupils give them.) "As I read this summary (00 NOT IDENTIFY WRITER), listen to see if the summary follows all these rules." "Which reporter questions were answered? Was it written like a story? Did it use only a few sentences? Did it tell only the important facts? (Keep the discussion positive. If additions or deletions are suggested, involve the learners in a discussion of how this might improve the summary. Do not allow the discussion to become critical. Be sensitive to the self-image needs of the writer.) 3. Closure "When you read a paragraph to see if it is a summary, what do you look for?“ Lesson 8: A. 114' Providing a final review of writing summaries Comerford Statement To review the process and product of summarizing a basal selection by having the learner: l. respond to questions about the kinds of questions reporters use when writing a news report respond to questions about the form and content of a summary respond to questions about the way cognitive memory ("target") and convergent ("puzzle") questions can be helpful in writing a surrma ry so that the learner can explain orally the process and product of summarizing a basal selection. Procedure 1. Set induction "We have learned many things about writing a summary. I have turned all our charts around so that we cannot see them. Let's see if you can help me remember what the charts tell us." (Have the three charts turned so writing faces wall.) "The first chart tells us the questions a reporter thinks of when he gets ready to write a report of a news event. See how many we can remember. I will write them on the board. When we cannot think of any more, I will turn the chart around and we will compare our list with the chart." (Repeat this with each of the charts.) Closure "Using the charts to help you, who can explain what you do when you write a summary of a story?" (Ask probing questions. Refer to the charts as needed.) APPENDIX B OBJECTIVE AND PRODUCTION TEST MATERIALS Number of Pupils Types of Tests Test Forms Listing Pupil Names Assignment of Tests Dates of Testing EXPLANATION OF TEST RECORD FORM My study is comparing the performance of pupils at high and low instructional levels, so I am collecting data only from these two groups in your class. I have used the numbers you gave me on your application form to determine how many tests to order for your class. (If you wish to use these tests with the other pupils in your class, you have my permission to duplicate them.) I have prepared two kinds of tests: 1. a lO-item test of ability to recognize a main idea (or summary sentence) 2. a story to test the ability of the pupils to produce a summary (write a paragraph which gives a summary of the story). There are two forms of each of these tests: 1. Form A and Form B of the objective test 2. Story 1 and Story 2 of the test of ability to write a summary The Record Sheet has room only for the names of the pupils in your high level instructional group and your low level instructional group. Please write these names in alphabetical order. As the test record shows, I have assigned specific tests to specific pupils. Since I did not have the names of your pupils, I have written the corresponding number on the test to be given to each pupil. For example, the first name on the high instructional level test will receive the test labeled "H-l"; the second name on that list will receive the test labeled "H-2," etc. The first name labeled "L-l"; the second one, the test labeled "L-2," etc. This distribution of tests is a critical feature of my statistical analysis. The pre-tests should be given a day or two before you begin the eight special teaching lessons. The post tests should be given a few days after teaching the eight special lessons. If you have any questions, please call me at home: 455-634l. 115 116 Teacher Grade School TEST RECORD Please list names in alphabetical order High Instructional T T Low Instructional Level Group . . . . Level Group Rec. - lO-item objective test of ability to recognize a summary statement. Pro. = Measure of ability to produce a summary. 117 TEACHER'S GUIDE--FINDING THE MAIN IDEA A. Directions to the teacher 1. Distribute the test materials. Make sure each pupil gets the test assigned to him. See the folder "Record of Assigned Tests. Demonstrate how to mark the answer sheet. Administer the test in two sittings a. First sitting—-items 1 through 5 b. Second sitting--items 6 through l0 c. Allow as much time as needed for each sitting. Tell pupils any single word they do not know, but DO NOT READ ANY PART OF THE TEST TO THEM. B. Directions to the pupils 1. Introduction "During the month of February I will teach several reading lessons on finding the main idea. We will begin by finding out what you know already." Distribution of materials a. Test--"Finding the Main Idea" (Be sure each receives the right form.) b. Answer sheets c. No. 2 pencils (Mrs. Rowlands informed me each school has a supply). Completing the test a. Identifying the papers "Look at the paper that says 'Finding the Main Idea.‘ The first line says 'Pupil.' On that line is a letter and a number. That is a secret code for your name. Write your name next to the secret code. Next, write the date, my name, and the name of our school." (Put these on board-- give help as needed.) "Now look at the small sheet. (Hold up the answer sheet.) Turn the paper sideways. (Demonstrate.) Write your name, my name, and the date." b. Marking the answer sheet (Write on the board--[a], [b], [c], [d], [e]) "Next to your name, on the side of the paper, it shows how to fill in your answer. After you read each story, decide on the answer. Look at the letter in front of the answer. 118 Find the number of the story on the answer sheet. Find the letter in the row next to the number. Take your pencil and completely fill in the box around that letter. For example, if you choose sentence Q, fill in the box around 9," (Demonstrate on board.) "FILL IN ONLY ONE BOX IN EACH ROW. IF YOU CHANGE YOUR ANSWER, COMPLETELY ERASE THE FIRST BOX. Taking the test "Now we are ready. Read each story. Under each story there are three sentences. Find the sentence that tells best what the story is about. Find the sentence that is most like the story. Then go to the answer sheet and fill in the answer." "If there is any word you do not know, raise your hand and I will tell you the word." "Stop when you finish story number 5. We will do the others tomorrow." 119 Sample Objective Test Items Level I (Third Grade) This was the day of the school fair. All the children had balloons. They put notes on cards and tied them to the balloons. Then they let the balloons sail into the sky. a. The children had fun at school. b. The children sent notes by tying cards to balloons. c. The children liked balloons at school Level 11 (Fifth Grade) Bob, Sue, Fred, and Alice wanted to go to the roller skating rink, but they did not have enough money. "We could have a car wash and earn money, said Sue. Alice and Bob got out the hose, pails, and old towels. Sue and Fred made signs that said: "Car Wash--75¢)." Many cars came. The children were very busy. Soon they had enough money to go roller skating. a. The children thought a car wash would be a good way to earn money. b. The children washed cars to earn money to go roller skating. c. All the children worked to get ready for the car wash. 120 TEACHER'S GUIDE-~NRITING A SUMMARY Directions to the teacher 1. 2. 3. 4. Distribute the test materials. Be sure each pupil gets the correct story, Story 1 or Story 2. See the folder "Record of Assigned Tests." Tell pupils any single word they do not know, but DO NOT READ ANY PART OF THE STORY TO THEM. Allow as much time as needed to read the story and write the summary. This test should be administered in one class period. Directions to the pupils 1. Introduction "During the month of February I will teach several lessons on 'Finding the Main Idea.‘ Today I want to find out what you know already about writing the main idea of a story." Distribution of materials a. Test--Story No. l or Story No. 2. Make sure each pupil receives the story assigned to him. b. Writing paper--the kind used for daily writing assignments. Identifying the papers a. "At the top of the story is a line for your name. On that line is a letter and a number. That is a secret code for your name. Write your name next to the secret code. Then write my name, the name of our school, and the date on this paper." (Have these written on the board for them to copy.) b. "00 the same with the writing paper. Write your name, my name, the name of the school, and the date at the top." Completing the test "As you read the story, ask yourself: 'What is the main idea of this story? What is this story about?’ After you finish reading, write a summary of the story. Tell the story in a short way." "If there is any word you do not know when you are reading, raise your hand and I will tell you. When you write the story in a short way, you may go back to the story as often as you like. If you need help in spelling a word, I will help you." 121 Pupil Date Teacher School WRITING A SUMMARY-~LEVEL 1, STORY 1 The sky was blue. The sun was warm. Tim Turtle sat on the old dock looking at the children. He saw the little boy fishing. He saw the little girl playing in the water. Tim Turtle was happy. Every summer Tim Turtle made his home under the old dock. Every morning he went up on the dock and watched the children. But this summer it was hard for Tim Turtle to get up on the dock. There had been no rain for a long, long time. The water in the lake went down and down. Then, one day, the water was too low. Tim Turtle could not get up on the dock. He tried and tried. Then he swam away. Tim Turtle found a rock in the sun. He sat on the rock and looked at the boats on the lake. But he was not happy. He wanted to be on the old dock. He wanted to watch the children play. One day Tim Turtle swam past the dock. The mother was calling the children. "Come in," she called. “A storm is coming." The sky became dark. The wind blew and blew. The rain came down hard. Tim Turtle saw lightning in the sky. He heard the loud thunder. Tim Turtle was afraid. He went deep down in the water. He stayed there until the storm was over. Tim Turtle went for a swim. He swam past the old dock and saw the boy fishing. He saw the girl playing in the water. Then he looked at the old dock. A big tree had come down in the storm. The tree was in the water near the dock. Tim Turtle swam to the tree. He went up on the tree. What a surprise! He could get from the tree to the dock. The sky was blue. The sun was warm. Tim Turtle sat on the dock and watched the children. Tim Turtle was happy again. 122 Pupil Date Teacher School WRITING A SUMMARY--LEVEL I, STORY 2 The snow was deep. The winter was cold. The little field mice looked for a warm home. They ran into the horse barn. “This looks like a good home," they said. "Maybe we can live here." The mice liked the home in the barn. They made a nest in the hay. They ate the corn the farmer had put there to feed the birds. They ran and played all day. At night they talked to the horses. There was one thing they did not like. The farmer had a mean cat. The cat stayed in the house at night. In the morning the cat came to the barn. Then the mice had to look out. The cat hid under the tools in the barn. She jumped out at the mice and made them run. Sometimes the mice ran to the t0p of the barn. The cat could get up there, too. The mice had to run again. One day the farmer came to the barn with a dog. "You look like a fine watch dog," said the farmer. "You can live in this barn. You can take care of my farm. The dog walked all around the barn. He saw the little mice but did not run after them. Just then, the mice saw the cat come in. The mice ran as fast as they could. They hid under the hay. The dog saw the cat, too. He did not want the cat in his barn. He barked and barked. He ran after the cat. The cat ran all the way to the house. The dog watched the cat go into the house. Then he walked slowly back into the horse barn. How the mice loved that dog. Now they did not have to look out for the cat. They could run and play all day. "This horse barn is a good home," they said. "We can live here all winter." 123 Pupil Date Teacher School WRITING A SUMMARY--LEVEL II, STORY 1 It was a warm, sunny day. Tim Turtle sat on the old dock watching the children. The boy was fishing and the little girl was playing in the water. Tim Turtle felt happy. Every sunmer Tim Turtle made his home under the old dock. In the morning he would climb up on the dock to watch the children. But this summer it was not easy for Tim Turtle to get on the dock. There had been no rain for a long, long time and the water in the lake was getting lower and lower. Then, one day, the water was too low. Tim Turtle tried and tried but he could not get on the dock. He swam away sadly. Tim Turtle found a rock in the sun. He climbed up on the rock and sat there watching the boats on the lake. But he was not happy. He wanted to be on the old dock watching the children play. One day, as Tim Turtle was swiming past the old dock, he heard the mother calling the children. "Come in quickly," she called. "A storm is coming." The sky grew very dark, the wind blew hard, and the rain came down fast. The lightning flashed and the thunder rolled. Tim Turtle was afraid. He went deep down in the water and stayed there until the storm was over. Tim Turtle went for a swim past the old dock. He saw the boy fishing and the girl playing in the water. Then he looked at the old dock. A big tree had fallen down in the storm and the tap of the tree was in the water close to the old dock. Tim Turtle swam to the tree and climbed on it. What a surprise! He could get from the tree to the dock. The sky was blue and the sun was warm. Tim Turtle sat on the dock watching the children. Tim Turtle was happy again. 124' Pupil Date Teacher School WRITING A SUMMARY-~LEVEL II, STORY 2 The winter was cold. The snow in the field was very deep. The field mice looked for a better home. When they saw the horse barn they said, "Let's go in here. Perhaps we can live here this winter." The mice were pleased with the home in the barn. They made a nest in the hay. They found a bag of corn that the farmer had put there to feed the birds. This would be food enough for all winter. There was lots of room to run and play all day. At night they talked to the horses. There was just one thing they did not like about the home in the barn. The farmer had a mean cat. The cat stayed in the house all night but, in the morning, the cat always came to the barn. Then the mice had to be very careful. The cat would hide under the tools in the barn and jump out when the mice came by. Some- times the mice climbed up to the top of the barn to get away from the cat. But the cat could climb up there, too. Then the mice would have to start running again. One day the farmer came into the barn with a dog. The dog was a fine watch dog. The dog would live in the barn and look after the farm. The dog walked all around the barn. He saw the little mice but did not run after them. Just then, the mice saw the cat come into the barn. They were so afraid that they hid in the hay. The dog saw the cat, too, and became very angry. This was his barn and he did not want the cat in it. Barking loudly, the dog ran at the cat. The cat ran out of the barn and started for the house. When the cat was in the house, the dog walked slowly back to the barn. How the mice loved that dog. Now they did not have to look out for the cat. They could live in the barn all winter. APPENDIX C SUMMARY CHECKLIST AND SCORING KEY SUMMARY CHECKLIST Pre- Post test Test 1. paragraph _____ _____ 2. setting, problem solution, sequence _____ _____ 3. major details _____ _____ 4. minor details _____ _____ Details Story l Story 2 turtle mice dock winter, needed home water low barn not get on dock mean cat storm chased mice tree down dog came near dock chased cat get on dock mice safe Pre-test Pre—test Post Test Post Test 125 126 SCORING OF SUMMARIES PRODUCED BY PUPILS Criteria Score 1. Is the summary written as a paragraph? a. Written as a paragraph ................. 2 b. Written as a list of answers .............. d 2. Does the summary answer the relevant "reporter" questions? a. Reports setting, problem, solution in sequence ..... 5 b. Reports setting, problem, solution, but not in sequence ........................ 4 c. Reports any three of the above ............. 3 d. Reports any two of the above .............. 2 e. Reports any one of the above .............. l 3. Does the summary contain all the major details? a. Reports all 8 major details .............. 5 b. Reports 7 of the 8 major details ............ 4 c. Reports 5 or 6 of the major details .......... 3 d. Reports 3 or 4 of the major details .......... 2 e. Reports less than 3 of the major details ........ l 4. Does the summary contain any minor details?* a. Summary does not include any minor detail ....... 5 b. Summary includes only 1 minor detail .......... 4 c. Summary includes 2 or 3 minor details ......... 3 d. Summary includes 4 or 5 minor details ......... 2 e. Summary includes more than 5 minor details ....... l *A major detail that is repeated is scored as a minor detail. BIBLIOGRAPHY BIBLIOGRAPHY Anderson, R. The notion of schemata and the educational enterprise. In R. Anderson, J. Spiro, & W. E. Montague (Eds.), Schooling and the acquisition of knowledge. Hillsdale, N.J.: Lawrence Earlbaum, 1977. Bader, L. A. Elements of comprehension. The Michigan Reading Journal, 1979, 13, 67-68. Bader, L. A. Reading diagnosis and remediation in classroom and clinic. New York: Macmillan, 1980. Beery, A., Barrett, T. C., & Powell, W. R. Elementary reading instruc- tion: Selected materials (2nd ed.). Boston: Allyn & Bacon, 1973. Bormuth, J. R. An operational definition of comprehension instruction. In K. S. Goodman & J. T. Fleming (Eds.), Psycholinggistics and the teaching of reading. Newark, Del.: International Reading Association, 1969. Brekke, B. W., Williams, J. D., & Harlow, S. 0. Conservation and read- ing readiness. Journal of Genetic Psychology, 1973, 123, 133-138. Charles, C. M. Teacher's petit Piaget. Belmont, Calif.: Fearon Publishers, 1974. Chester, R. 0. Reading comprehension: A pragmatic approach to assessment and instruction. In R. T. Williams (Ed.), Insights into why and how to read. Newark, Del.: International Reading Association, 1976. Christopherson, S. L. The effects of knowledge of semantic components of discourse on summarizing text. Paper presented at the Annual Meeting of the American Psychological Association, New Orleans, La., September 1974. (ERIC Document Reproduction Service No. ED 188 152) Cleland, C. J. Piagetian implications for reading models. Reading wor1d, 1980, _2_Q,’ 10-15. Comerford, J. F., & Fleury, B. J. The objectives controversy. Journal of Teacher Education, 1972, 23, 27-31. Cox, M. B. The effect of conservation ability on reading competency. ReadinggTeacher, 1976, 39, 251-258. 127 128 Cromer, W. The difference model: A new explanation for some reading difficulties. Journal of Educational Psychology, 1970, 31, 471-478. Cromer, W., & Wiener, M. Idiosyncratic response patterns among good and poor readers. Journal of Consulting Psychology, 1966, 39, 1-10. Cunningham, J. W., & Foster, E. The ivory tower connection: A case study. Reading Teacher, 1978, 31, 365-369. Dawkins, J. Syntax and readability, Newark, Del.: International Reading Association, 1975. Dow, J. Report to the board of education on the fall, 1980, administration of Michigan educational assessment program. Grand Rapids, Mich.: Grand Rapids Public Schools, 1981. Dreher, M. J., & Singer, H. Story grammar instruction unnecessary for intermediate grade students. Reading_Teacher, 1980, 33, 261—268. Durkin, Dolores. What classroom observations reveal about reading comprehension instruction. Reading Research Quarterly, 1978-1979, 14, 481-527. Durkin, Dolores. Basal reader research findings. Paper presented at the Text Comprehension Research Symposium sponsored by the Michigan Secondary Reading Interest Council, Ann Arbor, Michigan, 1980. Elkind, 0. Reading, logic and perception: An approach to reading instruction. In J. F. Hellmuth (Ed.), Educational therapy (Vol. 2). Seattle: Special Child Publications, 1969. Elkind, D. The child and society, New York: Oxford University Press, 1979. Estes, T., & Vaughan, J. Reading and learning in the content classroom. Boston: Allyn & Bacon, 1978. Fry, E. A readability formula that saves time. Journal of Reading, 1968, 11, 513-516. Fleisher, L., Jenkins, J. R., & Pany, 0. Effect on poor readers' comprehension of training in rapid decoding. Reading Research Quarterly, 1979, 13, 30-48. Geiselman, R. E., & Samet, M. G. Summarizing military information: An explanation of schema theory. Human Factors, 1980, 33, 693-705. Gibson, E. J., & Levin, H. The psychology of reading. Cambridge, Mass.: The MIT Press, 1975. 129 Golinkoff, R. A comparison of reading comprehension processes in good and poor comprehenders. Reading Research Quarterly, 1975-1976, 11; 623-659. Guthrie, J. T. Reading comprehension and syntactic responses in good and poor readers. Journal of Educational Psychology, 1973, 33, 294-299. Hall, E. A conversation with Jean Piaget and Barbel Inhelder. Psychology Today, 1970, 3, 25-31, 54-56. Harned, L. A comparison of written discourse within four elementary social studies series with regard to syntactic structure, conceptual density and abstractness, and vocabulary difficulty, Unpublished doctoral dissertation, Michigan State University, 1979. Harris, L. A., & Smith, C. B. Readingyinstruction: Diagnostic teaching in the classroom. New York: Holt, Rinehart, & Winston, 1980. Herber, H. Teaching reading in content areas. Englewood Cliffs, N.J.: Prentice-Hall, 1970. Huey, E. The psychology and pedagogy of reading. New York: Macmillan, 1908. Reprinted Cambridge, Mass.: MIT Press, 1968. Karlin, R. Teaching elementary reading (3rd ed.). New York: Harcourt, Brace, & Jovanovich, 1980. Kavale, K., & Schreiner. The reading processes of above average and average readers: A comparison of the uses of reasoning strategies in responding to standardized comprehension measures. Reading Research_9uarter1y, 1979, 13, 102-128. Kirkland, E. R. A Piagetian interpretation of beginning reading instruction. Reading Teacher, 1978, 31, 497-503. Kretschmer, J. Measuring reading comprehension. Unpublished doctoral dissertation, Ohio State University, 1972. Kretschmer, J. Toward a Piagetian theory of reading comprehension, Reading World, 1975, 14, 180-187. Levin, J. R. Inducing comprehension in poor readers: A test of a recent model. Journal of Educational Psychology, 1973, 33, 19-24. Mandler, J. M., & Johnson, N. S. Remembrance of things parsed: Story structure and recall. Cognitive Psychology, 1977, 3, 111-151. 130 McKenzie, M. Schematic reading and learning. In R. T. Williams (Ed.), Insights into why and how to read. Newark, Del.: International Reading Association, 1976. McLaughlin, G. H. SMOG grading--a new readability formula. Journal of Reading, 1969, 13, 639-646. Michigan State Board of Education. MEAP, Michigan educational assessment handbook. Lansing, Mich.: Michigan State Board of Education, undated. New York State Education Department. Comprehension through active processing: A set of interactive instructional models. Albany, N.Y.: Bureau of Reading, 1978. (ERIC Document Reproduction Service No. ED 163 420) Olshavsky, J. E. Reading as problem solving: An investigation of strategies. Reading_Research Quarter1y, 1976-1977, 1;, 654-674. Pearson, P. 0., & Johnson, D. 0. Teaching reading comprehension. New York: Holt, Rinehart, & Winston, 1978. Pearson, P. 0., & Samuels, S. J. Editorial: Why comprehension? Reading Research Quarter1y, 1980, 13, 181-182. Peters, C. W. Evaluating social studies materials: Readability formulas are not the only answer. Social Studies, 1977, 31, 70-77. Phillips, 0. G. Piagetian perspective on science teaching. Science Teacher, 1976, 53, 30-31. Reder, L. M. The role of elaboration in the comprehension and retention of prose: A critical review. Educational Research, 1980, i9) 5'53. Roberts, K. P. Piaget's theory of conservation and reading readiness. Reading Teacher, 1976, 39, 246-250. Roe, B., Stoodt, B. D., & Burns, P. C. Reading instruction in the secondary school. Chicago: Rand McNally, 1978. Samuels, S. J., Begy, J., & Chen, C. C. Comparison of word recognition speed and strategies of less skilled and highly skilled readers. Reading_Research Quarter1y, 1975-1976, 11, 72-86. Singer, H. Active comprehension: From answering to asking questions. In C. McCullough (Ed.), Inchworm, inchworm: Persistent problems in reading instruction. Newark,—Del.: International Reading Association, 1980. 131 Smith, N. B. Be a better reader, basic skill edition. Englewood Cliffs, N.J.: Prentice-Hall, 1977. Smith, N. B., & Robinson, H. A. Reading instruction for today's children (2nd ed.). Englewood Cliffs, N.J.: Prentice-Hall, 1980. Smith, N. J., & Mason, G. E. Creating options for the teaching of reading comprehension: Research on methodology. Paper presented at the College Reading Association Annual Conference, Baltimore, Maryland, 1980. Spache, G., & Spache, E. B. Reading in the elementary school (2nd ed.). Boston: Allyn & Bacon, 1969. Steffensen, M. S., Joag-Dev, C., & Anderson, R. C. A cross-cultural perspective on reading comprehension. Reading Research Quarterly, 1979, 13, 10-29. Thorndike, E. L. Reading as reasoning: A study of mistakes in paragraph reading. Journal of Educational Psychology, 1917, 3, 323-332. Thorndike, R. L. Reading as reasoning. Reading Research Quarter1y, 1973-1974, 2, 135-147. Vacca, R. T. Content area reading. Boston: Little, Brown, & Co., 1981. Van Blaricom, G. The effect of passage organization on main idea comprehension at three response levels. In C. Pennock (Ed.), Reading comprehension at four linguistic levels. Newark, Del.: International Reading Association, 1977. Viti, R. Letters to the editors. Reading Research Quarter1y, 1980, 15, 524. Waller, T. G. Think first, read later! Piagetian prerqujsites for reading. Newark, Del.: International Reading Association, 1979. Whaley, J. F. Story grammars and reading instruction. BeadingTeacher, 1981, 35, 762-770. Wiener, M., & Cromer, W. Reading and reading difficulty: A conceptual analysis. Harvard Educational Review, 1967, 31, 620-643.