W {0.29 ‘« -.J¢X-;.:Z . 5“": v u El: T 3’55? f . I S V‘ '.'.‘.o‘ .1 \uJ. . . a yang-sag» “A... .. *3}. ”75:4 'v x. ‘ . 3;: {33" a... It?" . ”53““ 5);} a? ’5? 3. u 1“ CV .. » I». ‘~ “:9“ , ’2 i . ’2» ' 3 ‘; '- 1' ‘ .v . ; “‘7, Z *x:'4‘;g-.A?‘:'r‘~, 3. x w w “ 1.?" 5* ‘éfi-‘- h. 75 ‘xf‘ge WI, “21.3%.; ‘J;-_- a ‘s. . . , .a «u 9-“ .‘ :‘thwr‘T-zyc. pummhug, 7.1»: a.- -w . -. y... .4». a w 6““ flag:- . w Lg: anusmm ‘I '11 A? h Kg." .. ”wits , ...~. - ~ 4-.. .‘I'.‘-'.f>r.. «a, - #1.» - a Q 'fiE‘ZJ-‘C' a. .‘l Hangar. ‘ ‘W: a: * I x . . ?VV ‘ ,2 . ,. ‘ 1 q: I" g:-.;~a;‘d£;u- . c V. 1- - 3;, ~3r _ . “ . 11.1. V? 3.1;; ,_ xv} §'~Q, 5 J. ., 5%?! «fwd Evlfii'u 3»; . «3.5:;- ‘t. V “'1? ~ “‘73. u: -‘ ‘ (who "a fig». '3?th a. r + .. “5"“ - 1". 1 ' RARIES “it: l u leAN STATE ""NEB W ‘IJ\ \\ H 4 J \\\\'\i§l\‘l\\1‘l\\2\\l§l\§l‘l\l\}‘ lljl'll \ y This is to certify that the dissertation entitled A Study of the Effect of Peer Responding on the Responder as Nriter-Reviser presented by Celeste Anne Resh has been accepted towards fulfillment of the requirements for PhoDo degree in En9115h M or pr ssor Date August 1, 1994 MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771 LIBRARY Mlchlgan State Unlverslty PLACE II REI’URN BOXto roman thb chock“ from your record. TO AVOID FINES return on or before data duo. DATE DUE DATE DUE DATE DUE fil | I MSU In An Affirmative Action/Equal Opportunity Inflation mm! 7—? A STUDY OF THE EFFECT OF PEER RESPONDING ON THE RESPONDER AS WRITER-REVISER BY Celeste Anne Resh A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of English 1994 ABSTRACT A STUDY OF THE EFFECT OF PEER RESPONDING ON THE RESPONDER AS WRITER-REVISER BY Celeste Anne Resh Because studies of peer response have found only minimal support for the effect of peer response when assessing end of term writing samples, the goal of this study was to find a way of identifying the effect of peer response within the process of responding and revising over the period of a semester. If this study could Show that responding has a beneficial effect on the responder as writer, it could help to more firmly support the practice of peer response in the composition classroom. In attempting to discover whether responders learn things about writing that they come to apply in their own writing activities, this study developed a method of analysis that applied measurement categories based on the writing hierarchy to the response and revision data of three first year college composition students of varying abilities. For the three case studies, all of their written responses to peers’ first drafts and all revisions made by the three students were coded into the measurement categories. The categories identified the quality levels of the responses and revisions, the levels at which the changes were suggested or made, the form of the responses that were given, the kinds of revision operations that were undertaken, and the source or motivator for the revisions that were made. Analysis of the data indicated that the measurement categories did allow identification of movement within the writing hierarchy. Further, this method of analysis appears to show that response activity does, at the least, predict future revision activity and possibly even influence it. Analysis of a larger sample of responders, and comparisons with a non—response class, are needed to confirm this preliminary study. Copyright by CELESTE ANNE RESH 1994 To my husband, Jim Resh, my mother, Helen Hering Meuwissen, and my father, Arthur J. Meuwissen ACKNOWLEDGEMENTS I am particularly grateful to my committee chairperson, Dr. Marilyn Wilson, for her unwavering support and invaluable assistance. I am also grateful to Dr. Sharon Thomas for her informed and thoughtful response to drafts, and to Dr. Marcellette Williams and Dr. Larry Landrum for their interest and efforts. Special thanks go to my friends and family for their encouragement and understanding, especially my husband, Jim, and daughter, Jane. vi TABLE OF CONTENTS LIST OF TABLES INTRODUCTION Chapter 1. A REVIEW OF RESEARCH ON THE ROLE OF COLLABORATIVE PEER RESPONSE IN THE COMPOSITION CLASSROOM Peer Response Theory and Practice The Conceptual Basis of Collaborative Peer Response Experimental Studies in Peer Response The Effect of Peer Response on Papers The Effect of Peer Response on Writers The Effect of Peer Response on the Responder as Commenter Peer Response Effect on Responders as Writers Some of the Problems Affecting Successful Peer Response 2. ELEMENTS OF THE STUDY Introduction Classroom Procedures Teacher Explanation Teacher Modeling Student Practice Class Arrangement Class Activities Responding Writer's Notes Other Class Activity Research Participants Means of Procuring the Data Responses to Peers Revisions Made by Responder/Writer Writer’s Notes Means of Classifying the Data Qualities of Writing Levels of Change Form of Response Revision Operations Source of the Revision vii 13 15 l7 19 21 24 33 Method of Analysis MOVEMENT WITHIN THE WRITING HIERARCHY IN THE RESPONSE AND REVISION ACTS OF THREE FIRST YEAR COMPOSITION STUDENTS First Case Study: Sandy Analysis of Written Responses to Peers Analysis of Revisions by Responder/Writer Second Case Study: Bert Analysis of Written Response to Peers Analysis of Revisions By Responder/Writer Third Case Study: Mickey Analysis of Written Responses to Peers Analysis of Revisions By Responder/Writer Conclusions MEASURING THE EFFECT OF RESPONDING ON THE RESPONDER AS WRITER/REVISER First Case Study: Sandy Response in the First Half of the Semester Revisions in the First Half of the Semester (n) Revisions in the First Half of the Semester First Half of Semester Overview Response in the Second Half of the Semester Revision in the Second Half of the Semester (n) Revisions in the Second Half of the Semester Second Half of the Semester Overview Second Case Study: Bert Response in the First Half of the Semester Revision in the First Half of the Semester (n) Revision in the First Half of the Semester First Half of Semester Overview Response in the Second Half of the Semester (n) Revisions in the Second Half of the Semester Second Half of Semester Overview Third Case Study: Mickey Response in the First Half of the Semester Revision in the First Half of the Semester (n) Revisions in the First Half of the Semester First Half of the Semester Overview Response in the Second Half of the Semester Revision in the Second Half of the Semester (n) Revision in the Second Half of the Semester Second Half of Semester Overview Conclusion CONCLUSIONS, IMPLICATIONS, AND QUESTIONS FOR FUTURE RESEARCH Conclusions and Implications The Delayed Effect of Peer Response The Importance of Classroom Method and Setting in viii 55 121 123 123 125 125 126 126 127 128 128 129 129 130 132 133 133 135 136 137 137 137 137 139 139 140 141 141 142 146 147 147 the Study of Peer Response Further Pedagogical Implications Questions for Future Research APPENDIX A: UCRIHS LETTER OF APPROVAL APPENDIX B: REVISION GUIDE FOR READERS/WRITERS RESPONDING TO A IST DRAFT PAPER APPENDIX C: BASIS FOR RESPONSE CATEGORIES APPENDIX D: SANDY’S DATA APPENDIX E: BERT’S DATA APPENDIX F: MICKEY’S DATA BIBLIOGRAPHY ix 152 158 159 162 163 165 168 181 191 206 LIST OF TABLES Table 1. Bert’s Responses to Peers 2. Bert's Revisions by Responder/Writer 3. Sandy’s Responses to Peers 4. Sandy’s Revisions by Responder/Writer 5. Bert’s Responses to Peers 6. Bert’s Revisions by Responder/Writer 7. Mickey's Responses to Peers 8. Mickey’s Revisions by Responder/Writer 9. Sandy's First and Second Half of Semester Performances 10, Bert’s First and Second Half of Semester Performances 11. Mickey’s First and Second Half of Semester Performances Page 58 6O 68 77 88 96 104 112 124 131 138 INTRODUCTION This study focuses on peer response in terms of the value that accrues to the responder as writer. That is, it attempts to discover whether responders, in actively responding to other writers, learn things about writing that they come to apply in their own writing activities. More specifically, if Jane’s responses to other students’ writing moved from the typical beginning writer focus on wording to addressing the higher level concern of idea development, would her revisions of her own writing indicate she was applying the knowledge she demonstrated in her responses? If not, then would this developing sophistication appear in her revision activities at some later date within the semester? In sum, could active reading and responding to peer writing have an effect on the writing knowledge of the responder that could be identified within the process of responding and revising over the period of a semester? Much of the research on peer response has studied its effect on the writing and the writers receiving the response, but very little has considered the effect that could come from the active producing of response. Research results, generally, have not shown a great deal of support 2 for the effect of peer response, however, and that may be at least partly due to where and how those researchers looked for its effects. Shortly after I began to review peer response research, I realized that I needed to look beyond the products and into the processes themselves if I hoped to discover any effects that responding might have in the brief learning period of a semester. Of the little research on peer response directed toward the effect of responding on the responder, I found none that had specifically analyzed both the response and the subsequent revision activity in an effort to identify growth. To accomplish this kind of analysis, I needed to look at a semester's worth of response and revision acts, and I needed to find a way of looking at this data that could provide a more detailed view of the effects of peer response. In preparation for this study, then, I gathered all of the work done in a semester college freshman composition class. The papers provided data regarding responses made to peers as well as revisions made from first to second drafts. In order to identify effects, I compiled a measuring system that would enable me to look closely into the fabric of responding and revising. Sets of writing hierarchy measurement categories were applied to the response and revision data of three students in an attempt to identify any increase of their knowledge of the writing hierarchy that might occur over the period of the semester. If my research could identify that responding has a 3 beneficial effect on the responder, it would provide a distinct kind of support for peer response that other researchers have not been able to ascertain, primarily because of their focus on product rather than process. Most importantly, if my study could show that responding has an effect on the responder, it would justify the use of peer response even more than those studies which show that receiving response is of help to the writer. For, if the only value of response is the receiving of it, then what would be the use of having students respond when the teacher is certainly more capable? Certainly, some composition instructors have found value in peer responding activities because they have allowed the assignment of more writing than otherwise might have occurred due to large class enrollments. The question of large classes sizes aside, however, leaves the larger query, of what value to the student is the act of responding to others? To be able to show that students learn through actively responding would firmly support the practice of peer response in the composition classroom. Further, the method of analysis I have compiled could help to open the way for more detailed analyses of the peer response process. Chapter 1 of this study contains a review of theory, practice, and research related to the role of collaborative peer response in the classroom. The review of the experi- mental research is categorized according to where the 4 researchers looked for peer response’s effect-—on the single paper being analyzed, on the writer, on the responder as commenter, or on the responder as writer. Chapter 2 describes the elements of the study. It includes both a description of the classroom setting within which the study took place, as well as an explanation of the particular measurement categories employed in the study. Chapter 3 provides a detailed analysis of the response and revision acts produced over the semester by the three case study subjects. The object of the analysis is to identify movement within the writing hierachy that appears in the work of the three students. In Chapter 4, I use the writing hierarchy measurement categories to measure the effect of responding on the responder as writer-reviser, the main focus of this case study. The data is presented through comparisons of first and second half of semester performances. In the concluding chapter, I address the implications of the delayed effect of peer response and the importance of method and setting in the study of peer response, as well as teaching implications and questions for further research. CHAPTER 1 A REVIEW OF RESEARCH ON THE ROLE OF COLLABORATIVE PEER RESPONSE IN THE COMPOSITION CLASSROOM Although the history of collaborative writing groups in the composition classroom goes back at least to the 1800's (Gere, 1987), current interest in their usefulness is generally agreed to be a direct result of the 1966 Anglo— American Conference on the Teaching of English at Dartmouth. Within two years of their exposure to the British model of student-centered education that endorsed process over product in writing, American educators joined in what has come to be known as a “paradigm shift" in the teaching of composition (Hairston, 1982). As Thomas Kuhn, who introduced the term, explained in The Structure of Scientific Revolutions (1962), "Paradigms shift when a shift in vision occurs" (119). After Dartmouth, the slow shift toward viewing writing as a process became gradually more visible, and one of the signs was that more and more composition teachers began to address the possibilities of writing groups in the composition classroom. PEER RESPONSE THEORY AND PRACTICE Among the earliest and most prominent spokespersons for peer response was James Moffet who endorsed the practice in Teaching the Universe of Discourse (1968). He argued that learning to use language required “the particular feedback of human response, because it is to other people that we direct speech. The fact that one writes by oneself does not at all diminish the need for response, since one writes for others" (191). More support for the value of writing groups to the process of writing came from Donald Murray’s A Writer Teaches Writing (1968). In it he described how professional writers profit from feedback, and he suggested that students could receive similar benefits by employing writing groups in the classroom. Another call for group work came from Ken Macrorie whose Writing to Be Read (1968) urged that students read their work aloud to one another and receive peer rather than teacher response. The publishing of these three books in 1968 brought a good deal of attention to writing as a process as well as to the practice of group/peer response within that process. Peer review and evaluation, which had been receiving some enthusiastic recommendations prior to Dartmouth (e.g. Dusel, 1957; Mersand, 1961; Johnson, 1962), continued to be appreciated by Peter Elbow (1968) who advised that copies of papers be distributed in a class so students could "judge 7 their effectiveness" (117). Stephen Judy (1970) looked beyond the evaluative aspect and recommended that students share writing with one another because they "probably constitute their own best audience" (216). In 1971 Elbow also noted the importance of audience to his students who had begun to write with the awareness that peers would be reading their work. Two years later, in Writing4Without Teachers (1973), Elbow described a "teacherless" writing class. His students read each other’s writing with the goal being "for the writer to come as close as possible to being able to see and experience his own words through seven or more people" (77) thus providing the writer with a sense of how the reader experienced the writing. Writing teachers and researchers experimenting with the student vs. teacher centered approach to audience found a variety of positive effects. According to Bruffee (1973), peer response contributed more learning as well as better writing. Graber (1974) found that student writing improved in the “teacherless” writing class. Judy (1975) advocated peer editing because “one doesn't need to be an expert in composition and rhetoric to make useful suggestions about the clarity and effectiveness of writing“ (112). He also urged teachers to "encourage group and collaborative projects" in order for students to be able to share their skills and knowledge with one another (113). Such writing group situations, according to Beach (1976), result in revision based on peer evaluation that contributes to 8 intellectual growth. Peer evaluation also provides "a kind of motivation" that Beaven (1977) did not see in other approaches to evaluation. The peer response benefits of motivation, audience awareness, revision skills, and improved writing ability were joined in 1982 by another, the writers’ ability to maintain authority over their own texts. Brannon and Knoblauch (1982) effectively represented the argument that the authority which teachers frequently usurp is returned to the authors when the writing group attends to the writers’ concerns and intentions. The authority and value of writing groups received two boosts in 1984, the first from the NCTE Commission on Composition which encouraged the practice of students commenting on each others’ writing; peer response was officially accepted. The second came from Kenneth Bruffee’s (1984) discussion of the history and ideas behind collaborative learning. Of particular import was his assertion that "no student is wholly ignorant and inexperienced," rather, each is "already a member of several knowledge communities... (and) pooling the resources that a group of peers brings with them to the task may make accessible the normal discourse of the new community they together hope to enter" (644). He pointed out that the success of collaborative conversation depends upon the teacher judiciously designing an assignment that indirectly structures the desired conversation (644). Bruffee’s 9 insistence on the importance of effectively making collaboration a "genuine part of students’ educational development" (652), was capped by his conclusion that to do so "requires new and perhaps more thorough analyses of the elements of our field than we have yet attempted" (652). Gradually, perhaps partly as a result of Bruffee’s call, more research began to focus on the claims of the theorists and practitioners. THE CONCEPTUAL BASIS OF COLLABORATIVE PEER RESPONSE In a recent discussion of collaboration, Harris (1992) characterized it as a "powerful learning tool--to promote interaction between reader and writer, to promote dialogue and negotiation, and to heighten writers’ sense of audience" (369). She added that the tradition of the student in a position of passive acceptance of knowledge from an authority has been replaced by that of the student’s active involvement in the process of collaboration (369). Contributers to this shift from the tradition of passive learning to the recognition of learning as active laid the groundwork for the acceptance of peer collaboration as a sound educational practice through their recognition of knowledge as a social construct. One of the most prominent contributers to this shift was Thomas Kuhn in his book, The Structure of Sgientific 10 Revolutions (1962). By persuasively arguing that a good deal of scientific knowledge is a social construct produced by the scientific community, Kuhn helped to popularize the growing rejection of the fixed view of knowledge. In her comprehensive book on writing groups, Anne Ruggles Gere (1987) traces the history of the fixed concept of knowledge. Gere observes that St. Augustine’s view of knowledge arising from God and disseminated by the priests who studied scriptures began to change in the sixteenth century as science became the source of knowledge. Then Descartes’ assumption that the mind could observe but not interact with the universe continued to support this fixed view of knowledge. The development of quantum physics as well as developments in philosophy gradually led to a move away from knowledge as something to be discovered and toward scientific knowledge as something that is, to some degree, socially constructed. Gere points out that the philospher Richard Rorty extended Kuhn’s argument to include all knowledge and developed his argument with reference to the work of Heidigger, Wittgenstein, and Dewey. Knowledge as social construct appears in fields such as reading and critical theory, as well. Stanley Fish argues that linguistic and interpretive norms are not embedded in the language but in the interpretive community. John Frow talks about criticism that would reach a "meta-interpretive level where our concern is not with the rightness or wrongness of a particular reading but with the formal and social 11 conditions and preconditions of interpretation" (186). Post-structuralist thought recognizes the impossibility of any definitive explanation of texts. Clearly, the non- Cartesian view of knowledge as a social construct has gained significant recognition. This non—Cartesian or social constructivist understanding of the creation of knowledge has also had its influence on education and certainly on writing groups. As Gere explains, "The social View of knowledge, which assumes an indeterminate text in writing groups, supports an enriched concept of writing. When writing constitutes the task of collaboration, the process of working together enables writers to use language as a means of becoming competent in the discourse of a given community. Learning, when conceived in collaborative terms, assumes a socially derived view of knowledge and opposes a fixed and hierarchical one" (75). Support for this social View of knowledge can also be found in the experiments of Lev vygotsky accomplished in the 1920’s and 1930’s. vygotsky’s studies of children’s development of thought and language drew on and surpassed Piaget’s. Reversing Piaget’s view of thought as preceding socialized speech, vygotsky (1981) proposed instead that “the primary function of speech, in both children and adults is communication, social contact. The earliest speech of the child is therefore essentially social... (19) Vygotsky further observed that children pass into a stage in 12 which egocentric speech emerges. When children are moved to stop and think, they tend to think aloud, and this type of speech serves “mental orientation, conscious understanding... it is speech for oneself, intimately and usefully connected with the child’s thinking“ (1981, 133) Eventually egocentric speech, which has its roots in the external, becomes inner speech. As vygotsky continued to study the process of language development, he developed an explanatory concept, the “zone of proximal development,“ a term that describes "the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers“ (1978, 86). Wertsch (1985) affirms that what vygotsky is describing is translatable into small groups of individuals “engaged in concrete social interaction“ (60). Consistent with his description of the initial stages of language development as social in origin, the "zone“ theory not only reaffirms the social element of learning but also acknowledges peer collaboration as a useful learning tool. Forman and Cazden (Wertsch, 1985) find support for peer interaction’s importance in the school based on the Vygotsian theory that social interaction activates intellectual processes which gradually become internalized. Referring to a student of vygotsky who cited his belief that speech creates intellectual capacity only through use of its 13 instrumental capacity, Forman and Cazden reason that peer tutoring should work as well as adult-child interaction. Their study of peer collaboration, where neither peer is explicitly more knowledgable, showed that assuming complementary roles allows collaborators to "solve problems together before they are capable of solving the same problems alone" (341). In their conclusion, Forman and Cazden note that the valuable parent-child interactions that seem to stop when school begins can be continued through peer interaction: "The only context in which children can reverse interactional roles with the same intellectual content, giving directions as well as following them, and asking questions as well as answering them, is with their peers“ (344). Peer collaboration, seen in the light of both knowledge as a social construct and social interaction as the foundation of thought and language, appears to be a potentially useful component of an effective educational system. EXPERIMENTAL STUDIES IN PEER RESPONSE Prior to the Dartmouth conference, and spurred by the post World War II boost in college attendance and the expected deluge of baby boomers, educators had begun to look at the efficacy of peer response in the college classroom. Claims arose that peer response--described variously as peer editing, peer critiquing, peer review, peer evaluation, and l4 peer interaction—-could improve student writing as well as achieve reduction of the teacher’s paper load. A typical example, Charlton Laird’s “Oregon Plan," described in "Freshman English During the Flood," (1956) was the subject of an experiment at the University of Oregon. Laird arranged small groups and asked the students to grade each other’s papers. He reported that his job changed from one of criticizing to one of showing the students how to criticize. As a result, the students received more instruction in writing than in the previous teaching arrangement. In a review of the Oregon experiment, John Sherwood (1958) reported that improvement in writing was basically the same in both the control and the experimental groups. Similar results have been reported in dissertations as recently as the 1980’s. Despite the unproductive-appearing results of the early studies, researchers continued to search for the potential contributions of peer response. As their focus turned away from the simple question of whether teacher or peer response would improve a particular piece of writing and toward the question of what makes better writers, they began to find more to value in peer response. Following is a survey of studies from the past two decades categorized according to the element which seems to be the object of primary interest in each study of peer response: the paper, the writer, or the responder. The first method of gauging the effectiveness of peer response involved comparing the 15 quality of drafts of papers for which the author did or did not receive oral or written peer response. Other studies looked to see what effect peer response has had on the writer who received it. Finally, some studies focus on identifying the value of peer response to the persons actively doing the responding; this act of responding can be assessed in two ways——in terms of its effect on ability to provide helpful response, and in terms of its effect on the writing ability of the responder. While studies concerned with the quality of the effect of peer response on the writer and on the responder appear to have garnered greater researcher attention in the last two decades, interest in whether particular papers will demonstrate the salutary effects of peer response has continued to be under study. The Effect of Peer Response on Papers Researchers searching for statistically significant evidence of the superiority of peer response over other methods of improving papers have yet to gather a great deal of evidence, yet no analysis of papers has proven it ineffective. In her analysis of the papers of high school seniors which had received peer feedback based on checksheets, Jones (1977) found that the students accepted more peer criticisms than they refused. Further, seventy- two percent of the criticisms received and acted upon produced improved papers; the improvements, however, were primarily in the area of diction. Using pre- and post-test 16 essays, Jenson (1979) investigated the effects of peer feedback on writing performance and revision behavior of junior high school students. She found that the experimental groups, who had produced written peer feedback twice for each paper throughout the term, achieved statisical significance of .05 or more for variables which included quality of writing and amount of revision undertaken. In a study that looked at papers of college freshmen over the period of a semester, Ziv (1983) found that, as the students gained more information from her revision strategy sessions and applied that knowledge during oral peer group sessions, the papers improved as writers gradually came to incorporate their peers' suggestions. An analysis of fourth grade students’ first drafts and their revised products by Hittleman (1983), found that peer conference techniques and the content-focused written comments of the teacher appeared equally effective. Bielecki (1987) compared two types of peer response, reader-based and criterion-based editing along with a traditional teacher response to writing. Significant improvement occurred in all three groups of twelfth graders with no statistical difference between them, but it was Bielecki’s observation that the papers receiving response from peer editors seemed to have improved more than the others. DiMento (1988) compared the effect of self— evaluation and peer response on the writing of twelfth— graders and found that although the quality of papers 17 appeared to improve, the data did not support the superiority of either one. In her study of university students, Barcelow-Hill (1988) took a new direction, intending to investigate how critiquing affects the responder/writer, but her evaluative focus was limited to analyzing a writing sample at the conclusion of the study period. While she observed that the experimental group had become relatively adept at peer critiquing, Barcelow-Hill found no difference between the control and the experimental groups in their performance when she analyzed a final writing sample. Although she intended to head in a new direction in terms of looking at the responder/writer, Barcelow-Hill used the traditional analysis of a final writing sample to determine the effect of peer response. Her results reflected those of the Oregon experiment undertaken over thirty years earlier, which could suggest that either there is little or nothing of value in peer response, or there is little or nothing contributed to the understanding of the value of peer response by final product analysis. The Effect of Peer Response on Writers In a review of studies where the assessment focuses on the effect of peers’ responses on the writer, it appears that the attempt to look beyond the analysis of final essays allows researchers to see more of value in the process. In addition to the studies of student writing, analyses of 18 taped oral discussions and student responses to questionnaires have produced insights for some researchers. Danis (1980) found that college sophomores who experienced peer response workshops developed an audience awareness that resulted in an increased understanding of the need for elaboration and explicitness as well as the need to attend to clarity of structure. Writers in the study also cited other advantages including mutual encouragement, the exchange of ideas, and the discovery of insights into their own writing, including the ability to identify their own strengths and weaknesses. Audience awareness was the subject of Reynolds’ (1987) study in which he looked at both the pre- and post-tests as well as the writing exercises and composing aloud protocols of college freshmen. Both the statistical and the descriptive data indicated a significant shift toward audience-consciousness in the collaborative group while the individual treatment group remained oblivious of it. Further, he found that the development of audience- consciousness resulted in a greater ease in producing text for the writers. Similarly, Bender (1989) cited college student responses to questionnaires that indicated group response contributed to the writers’ sense of audience as well as their sense of purpose and style. Bender concluded, based on her study of peer comments and the student writer, that, in general peer comments are "stimulating, motivational, 19 challenging and supportive" (174) to the writer. Gibbs (1990), like Bender and Reynolds, found that the fifth grade writers in her study were strongly influenced by comments received from peers. Similar to the subjects in Reynold’s study, Gibbs’ writers produced a high percentage of reader-based writing as a result of group members’ responses. Additionally, Gibbs concluded that writers developed new revision strategies as they internalized each others’ strategies. She also noticed both improved writing and increased length in texts produced-—accomplishments similar to the greater ease in producing text that Reynolds noted in his writers. The use of research methods beyond examining a final piece of writing apparently allowed these researchers to observe more of value in peer response than occurred in research focused primarily on comparing control and experimental groups’ final writing products. This broader view appears to accompany the shift from the focus on paper quality to a focus on the writers themselves and how they are affected by receiving peer response. The Effect of Peer Response on the Responder as Commenter Many of the studies of peer groups found, among other things, that participating in peer response groups did affect the responder’s ability to respond to or comment upon others’ writing. From explanations of how students become good critics to claims for the internalization of language, researchers found, through observation of group work and 20 through analysis of oral tapes and student responses to questionnaires, that peer response activities affected the responder as well as the receiver of the response. In her study of peer group writing evaluation, Ziv (1983) discovered that while the college freshmen writers had trouble accepting criticisms, the responders’ skills improved over time. She found that while peer comments on surface level matters remained the same throughout the semester, there was a progression in the nature of comments at the conceptual and structural levels. Students eventually moved beyond simply criticizing a writer to being able to offer valid revision suggestions. Ziv credited not only student interaction but also teacher sessions focusing on potential revision strategies for the increase in ability to respond helpfully to conceptual and structural problems. In her study of a college basic writing class, Coleman (1984) used a revision taxonomy to analyze the oral contributions of response group members and found them developing an awareness of the necessity for making changes that would advance the meaning of their texts. She discovered that while they were able to demonstrate this in their group talk about papers, their written work did not yet reveal the increased knowledge. In a similar study of fifth, eighth, and senior high students, Gere and Abbot (1985) found that students participating in peer writing groups grew more adept at discussing writing. The researchers described students as frequently internalizing 21 the language they heard as they participated in group sessions. The effect of the workshop on peer critiquing and subsequently on the participant’s ability to write was the focus of Barcelow—Hill’s dissertation and, as noted earlier, she found no difference between control and experimental groups on a final writing sample. However, she did see both more diagnostic and more sophisticated comments in the experimental group by the end of the term. Wauters (1988) instituted a system in which pairs of college freshmen critiqued another student’s paper. Critiquing pairs changed each time the class met to critique papers, and Wauters observed that they benefited from this exposure to different standards which subsequently caused them to examine the validity of their own. Bruffee’s call for more thorough analyses in the field of composition appears to be being met by those researchers whose focus has turned toward the act or process of collaboration and away from a sole concentration on the product of it. Peer Response Effect on Responders as Writers As researchers look for more than the information gained from evaluating a final piece of writing, they seem to be seeing more deeply into the fabric of the process of developing as a writer. When Danis (1980), who found no actual improvement in final writing samples, listened to what her college sophomores were saying about their peer 22 group experiences, she found evidence of internalization of writing knowledge. She cited students who commented on their newly developed tendency to keep in mind previous criticism as they composed new papers: "The group put a question mark by ’it,’“ said one; "now, when I type ’it,“ I stop and think... I’m explaining myself better now" (141). Although analysis of final writing may not have discovered it, the interviews make it clear that peer response affected these writers as they approached fresh writing assignments. In students' learning logs and in interviews with students, Coleman (1984) found evidence of the college basic writers’ growth in ability to self-monitor. From almost no self-monitoring on their first papers during group discussion, the students in her study made more comments about their own papers than others did by the end of the semester. She also noted that students were ”taking their audience into consideration much more as they monitored their own writing" (27), an improvement she treated as a separate issue and credited to the teacher; however, it would seem more likely that interaction with a peer audience of responders is more directly related to audience awareness than any admonition a teacher could produce. Similarly, Freedman (1987), whose data included audio and videotapes, ethnographic note taking and all writing materials, found the ninth grade peer response groups in her study capable of aiding students in responding to their own writing. Recently, research by Baum Brunner (1990) has offered a 23 broader vision of the contributions of response to the responder as writer. Her research on twelfth graders’ writing and revision involved analyzing both audio—taped sessions and the drafts and revisions produced in a twelfth grade writing class. In analyzing the data, she offered a vocabulary for describing the kinds of contributions that peer response offers to the writing student. The traditional form of response centered around evaluating strengths and weaknesses of a text is called Intra-Textual talk. Baum Brunner suggests that two other forms of talk are also involved, Inter-Textual, a type of talk that connects different types of texts, and Extra-Textual Response which is talk that helps students learn how to talk about texts. Growth in ability to talk about text included the ability to be direct and directive in collaboration with others while focusing on a few topics. As a result of her study, Baum Brunner concluded that the strength of the relationship between the various styles of talk and writing itself may require creating changes in one of the modes in order to make changes in the other one. In looking closely at four individuals, she found that three of the four made changes in both their response talk as well as in their writing, and the fourth made few changes at all. Her conclusion goes beyond the question of whether commentary is of value or not and asserts that commentary is a bellwether of changes in writing ability, an assertion that is the focus of my own research for this dissertation. 24 As research has moved beyond the product itself and begun to look along the process of writing and responding, we seem to be finding more to value in terms of what peer response has to offer in the composition classroom. Relevant to composition research is the comment of the forest biologist who has developed an expertise in studying forest canopies, a part of the forest ecosystem that ground- bound scientists had ignored: "Where you put your traps is going to have a huge effect on what you get" (Pennisi, 409). Some of the Problems Affecting Successful Peer Response Research on the effects of peer response can be complicated not only by the choices made about what aspects to focus upon, but also by the dynamics of the collaborative interaction within the classroom setting. Bruffee (1984) identified five elements without which the collaborative conversation among peers can break down: students need knowledge of the subject being written about, knowledge of the assignment, sensitivity to others’ needs and feelings, knowledge of the conventions of discourse, and knowledge of the standards of written English (644). Likewise, Karen Spear (1988) identified limitations that included "lack of ability or intellectual readiness to read and revise effectively,” and “basic confusions about the nature of process and product in students and teachers alike" (51—2). This confusion, Spear explains, is manifested in several ways including “miscues from teachers that collapse process into product," a lack of interaction skills that affects 25 both the students’ reading and revision of others’ drafts, and restrictive reading strategies (52). That it is no small task to judiciously design an effective collaborative conversation, as Bruffee puts it, becomes apparent in a review of the literature on peer response. Among the studies reviewed in this survey of the literature, the predominant cause of problems in peer response appears to be a combination of Bruffee’s element of sensitivity and Spears’ identification of a lack of interaction skills among the peers. Danis (1980) noted in her study of college sophomores that a quarter of the papers had major difficulties overlooked by the students in their oral response sessions. In follow—up interviews over half of the students said they had been reluctant to comment negatively on peer papers. Danis primarily blamed both student uncertainty regarding the peer response task as well as limitations related to the social situation due to fear of offending peers (148). Similarly, Ziv (1983) observed that the oral and written responses of college freshmen were often valid criticisms, but the receivers of the criticism resented and thus ignored it. Eventually, however, the students began to make helpful suggestions which were subsequently implemented by their peers. Ziv felt her students "became good critics as a result of their interaction with their peers and because throughout the semester, I interspersed peer group sessions with class sessions about revision strategies they might use 26 when rewriting their papers" (9). It seems likely that her instruction in revision strategies was the primary reason for improvement in group interaction, and apparently the passage of time helped bring sensitivity to the peer responders. The need for instruction in sensitivity and interaction skills appears in a case study of three college freshman writers. Berkenkotter (1984) sees her case study subjects respond to oral peer feedback in three entirely different ways, only one of which seems unharmed by the experience. She concludes by cautioning that students’ sense of authority over their texts may be threatened by peer response that is not appropriate. Once again sensitivity affects peer response effectiveness. George (1984) discovered other failings in her study of oral peer groups in a college freshman English class. One of the most serious failings, she noted, was that even after a particularly good discussion, the writers did not recognize as worthwhile the useful comments that had been made. Second, in their reading of essays, the responders did not look critically at how ideas were presented, but rather tended to pick up an idea and discuss it in isolation. Finally, two of the three groups, the "leaderless" and the "dysfunctional" groups (321), did not achieve productive interaction in terms of giving one another useful advice. George’s proposed solutions include asking students to bring questions about their papers to the 27 group, to talk through their papers before or after reading them and explain where they had difficulty, and taping the sessions or asking students to review their discussion at its conclusion. In a study of oral peer response groups in two ninth grade English classes, Freedman (1987) found that the peer response groups could help students respond to their own writing, and the group members encouraged and questioned one another, but did not engage in talk that evaluated other writing. The students not only avoided discussing writing problems, they also resisted using response sheets provided by the teacher. Solving the sensitivity problem would no doubt improve the interaction skills of these students. More recently, Wauters (1988) and Mabrito (1991) have confronted the sensitivity issue in their peer response groups. Wauters noted the tendency for disillusioned composition instructors to give up on peer response which they find results in "disappointingly brief or trivial" verbal exchanges that produce insignfificant changes in their drafts (157). Citing numerous limitations of the Bay Area Writing Project model of small group oral response, she argues that a code of mutual protection keeps most groups from honest interchange and results in "aimless verbal abandon" (158). Acknowledging that some groups do improve over time, Wauters notes that the conflicts of weaker groups also tend to grow over time. Further problems result when initial groups shrink as the term progresses. Wauters’ 28 solution for her college freshmen is not to add more teacher directed activities in order to control the groups, but an alternative model which she calls "non-confrontational critiquing" (157). In this approach students are paired up to read a paper of another student which will be revised after the writer reads the peer comments. Since the student whose paper is being read is reading another paper somewhere else in the room, there is no time wasted on unfruitful confrontations. Not only does this model avoid many of the group problems, according to Wauters, it also increases participation of students who usually contributed little to group discussions. She allows students to choose their own partners for peer critiquing, but she also asks them to find new partners each time. Their job is to read, defend their opinions to each other, reach a consensus, and write their response to the paper. Wauters spends the first two weeks of the course on peer critiquing training. She emphasizes the importance of convincing the students that they already have the skills necessary to critique but, near the end of this period, she gives them a peer critiquing form to use which she argues establishes clear criteria that allows students to respond more meaningfully. Although Wauters may be propping up her students in the first part of her training and then undercutting them by supplying a critiquing form, it is quite clear that she has avoided the sensitivity and interaction problems that can sabotage small group oral peer response. 29 Sensitivity to small group interaction is the focus of Mabrito’s (1991) study of eight first year college writers, half of whom were "high apprehensive" writers and the other half of whom were "low apprehensive" (509). Observing that it takes a composition teacher time to establish a community of trust with any peer groups, Mabrito points out that it is even more difficult when dealing with the high apprehensive’s stress at being placed in a situation where not only is the writing made public, but public verbalization in response to other writing is also demanded. Using electronic mail as the medium through which the students gave and received peer response, Mabrito found that the group of high apprehensives participated on the same level as the low apprehensives in terms of giving response, a considerable improvement over their face—to-face performance. Further, they responded more to e-mail suggestions for revision than they did to face-to-face recommendations. While e—mail response is not available to the average writing class, the effect is much like that achieved by Wauters in her non-confrontational critiquing pairs arrangment. In an article about things that go wrong with student- centered teaching, Mary Rose O’Reilly (1989) noted how students who have been through progressive education "put their chairs in a circle, like well-modified rats, as dutifully as students of 1967 faced the front," thus appearing obliging on the surface while remaining resistant 30 underneath (144). We find similarities within the dynamics of collaborative interaction in some peer response group classrooms. Students will obligingly respond with innocuous comments because they fear both offending others as well as receiving retaliatory remarks. Since sensitivity to confrontational interaction occurs as a problem in many studies of peer response, this problem needs to be acknowledged when we look at peer response research that does not report the kind of success that its theoreticians and proponents believe is possible. Sensitivity, while a difficult problem, is not the only one as Kenneth Bruffee makes clear in his effective discussion of collaborative learning (1984): Organizing collaborative learning effectively requires doing more than throwing students together with their peers with little or no guidance or preparation. To do that is merely to perpetuate, perhaps even aggravate, the many possible negative efforts of peer group influence: conformity, anti—intellectualism, intimidation, and leveling—down of quality. To avoid these pitfalls and to marshall the powerful educational resource of peer group influence requires us to create and maintain a demanding academic environment that makes collaboration-—social engagement in intellectual pursuits--a genuine part of students’ educational development (652). As important as those elements critical to the successful practice of collaborative peer response are to the classroom, so, too, is the acknowledgement by researchers of the presence or absence of those elements in the peer response situations they are studying. Without a comprehensive review of the critical elements and potential 31 pitfalls present in the classroom being studied, researchers risk being in the position of finding only that the particular teacher has inadequately organized a peer response situation rather than whether or not the practice itself is of educational value. Over the past few decades, the practice of peer response has steadily gained adherents among composition specialists, if not solutions to problems that have existed since its inception. Conceptual support for viewing peer response as a learning tool continues to grow as the social view of knowledge gathers a wider audience. However, until recently, research has not found support of any appreciable degree for the practice of peer response in the composition classroom. Unimpressive results seem to occur most often when the focus of the study is on the final product of composition. As research moves beyond looking simply at the end product, some researchers are finding that peer response activities apparently increase audience awareness, and others are seeing an increase in ease in talking about and producing text. These results suggest that peer response studies are likely to find more support for the practice as they move beyond the final product of writing and look more closely into the fabric of the processes that the writers go through as they write, respond, and revise. An additional reason for the absence of stronger support in research results may be due to studies that focused on classrooms 32 where peer response situations were under—developed. As should teachers in the implementation of collaborative activities, so should researchers attempt to identify the existence or absence of elements critical to successful peer response, thereby providing an informative context within which their results can be assessed. Attending to this important consideration, as well as to the processes involved in peer response, should help researchers to provide a clearer view of the effect and value of peer response in the composition classroom. CHAPTER 2 ELEMENTS OF THE STUDY INTRODUCTION Increasingly, peer response has become the subject of much study, but little has focused on the internalization of writing knowledge as a result of active response to peer writing. One of the few to look into this area, Baum Brunner (1990), concluded that the talk she heard as twelfth graders commented on each others’ papers was a bellweather of changes in their writing ability. She suggested that the strength of the relationship between the various styles of talk and writing itself may require creating changes in one of these modes in order to make changes in the other. (She seems to be suggesting that, for example, students who focus only on mechanical concerns in writing as well as peer response might move beyond that in response, say, if they are taught to move beyond that level of concern in their writing.) Her conclusions seem well—founded and may help to explain those studies that found peer response to have little if any effect on writing. It seems logical to assume that ineffective response techniques would affect the outcome of response, yet studies have continued to reach 33 34 weak conclusions about the effect of response itself without considering that the effect may depend upon the techniques or methods of responding. As this researcher noted in the previous chapter, some of these weak findings may be due to the focus of the research, but it also seems likely that they may result from the particular approach to response used by the groups being studied. Ideally, a study of the effects of peer response should occur in a classroom where the teacher has had the time to develop a peer response approach that has resolved the kinds of drawbacks noted in the previous chapter, particularly the problems of sensitivity and the lack of interactive skills. The task of judiciously designing an effective collaborative conversation, as Bruffee has recommended, needs to be addressed before a researcher can attempt to assess the impact of response on the writer. My early experiences in designing a peer response classroom led me to modify my approach until I felt confident that it was, as I told my students, as valuable to give as it was to receive response. The problems that led to my modifications of the early response format were related to oral response, critiquing sheets, and small groups of three to five assigned to the same group for the semester. Oral response exacerbated the sensitivity problem that Bruffee and others have acknowledged as limiting the students’ responses to innocuous comments in order to avoid having their own feelings hurt when their turn came around 35 I noticed, too, the same problem that Nystrand (1986) observed, "groups that proceed by listening rather than by reading rarely go beyond sentence level concerns" (189). Critiquing sheets seemed to create resistence, possibly because they usurped the students’ authority at the same time the students were being told they had the authority to respond, or possibly because they distracted the readers from genuine reading and responding. Permanent small groups, while intended to create security, did not necessarily accomplish it due to personality conflicts during oral response; in addition, when some of the members were absent, the one or two left had little beneficial interaction. To solve the small group problems, I instituted a different format. When first drafts were due, half the class came one day, and the other half came the next. The problem of absent students depriving group members of response was alleviated by having the half class group of a dozen or so sit in a single circle with a chair in the center as the exchange point for papers. If some of the students were absent, this did not create any problem because the group was still large enough to ensure a full class period of response activity. Students who wished to quietly discuss a paper were encouraged to do so, but not at the expense of avoiding written response. This reduction by half not only helped to create a more intimate atmosphere, but also allowed me time to be able to confer individually 36 for a few minutes with each student while the others continued their responding. The sensitivity and interaction problems were resolved partly by the institution of written rather than oral response. The other resolution to those problems was achieved by requesting that all response be in the form of either praise or questions rather than direct critical comments. Hillocks (1986) found support for this view in his review of research involving positive and negative comments, "Negative comments have negative effects, and positive comments—~on the average-~have positive effects" (221). The result of these adjustments in my approach to peer response over the years has been a generally confident class that responds with pertinent praise and genuine questions that lead to improved second drafts. Since I had assured my students that response would not only lead to improved papers but to their overall improvement as writers, I became very interested in seeing what their response comments and their revisions would show over the period of a semester. After reviewing research on the subject, I concluded that if the researcher’s focus is not limited to comparisons of final products, and if the responding situation has the earmarks of success (substantive responses and revisions and satisfied students), then a study of the process of response and revision should yield some useful information about the relationship between the two. Following is a description of 37 the writing class from which my data was gathered-—a class much in the form of Hillock’s (1986) "environmental mode“ where the teacher provides only brief introductory lectures, then attempts to structure activities which will potentially engage students in focusing on the principle of the writing task and also help them to provide appropriate feedback for their peers (122). CLASSROOM PROCEDURES Preparation of the students for written peer response was accomplished in less than two class meetings. An explanation of the desired focus and form of response was followed by the teacher’s modeling appropriate response. Student practice was reviewed and received appropriate response. The students wrote five papers over the semester with a varying combination of expressive, exploratory, informative, and persuasive features. Papers were introduced through a variety of prewriting activities including brainstorming, visualizing, listing, freewriting, small group discussions and activities, demonstrations, and examples. The students recorded in their Writer’s Notes the information they would need when they wrote their papers. Each paper assignment was intended to create the potential for the development as well as the logical organization of ideas. Following are brief descriptions of the assignments as they were presented to the students: 38 Paper I——A narrative description of an accident that you have experienced—~an actual one, or an embarrassing moment, or a happy coincidence. Conclude with a reflection on the experience. Paper II——A description of an argument/disagreement in which you were involved where neither of you was absolutely right or wrong. Each side is told from that person’s point of view, and the conclusion contains analytical or reflective commentary. Paper III-~A consumer investigation of a product you use. After developing a list of questions, test the product and/or survey peers in relation to your questions. Consider how best to organize your information as you write up your survey and its results. Paper IV--Interview someone from an older generation. If your subject talks about a number of events, you’ll need to consider logical methods of organization and transition as you compose your paper. If the person tells about one event in simple chronological order, you will want to add your own comparative or reflective comments throughout the paper. Paper V--From the journal entries you’ve been doing in response to editorials and opinion columns, select a topic on which you would be interested in writing either an opinion essay or a letter to the editor. After referring to the section on logic in the course handbook, try to find three different types of support for your opinion, and then 39 decide the most effective way of organizing your essay or letter. In preparation for these paper assignments, one to three classes were devoted to pre-writing activities. Each of the papers was revised once, and the second drafts were subjected to editing sessions primarily concerned with mechanical and spelling concerns. Twice in the semester they selected one of the second drafts and rewrote it into a final polished draft, but these third drafts and their editing corrections were not a part of my study. Following is a more specific description of the class preparation, arrangement, and activities. Teacher Explanation In explaining the focus of response, I discussed the four categories of writing qualities and emphasized the order of importance usually assigned them, e.g. Diederich and his group (1974): Ideas, Organization, Wording, and Flavor. I also explained that while spelling and mechanical considerations were of importance in writing, they would not be addressed until the second drafts of students’ papers were written and responded to in a second group editing session. In my explanation of the form of response, I pointed out that direct critical comments seem to raise anxiety more than they lead to writing improvement. I strongly recommended that they avoid direct criticism and instead ask 40 questions when they were concerned about something in a peer’s paper. In addition, I explained, they should also look for things to praise, because we usually learn more when we’re told what it is we are doing right than when the focus is primarily on our errors. I told them that I would keep them focused on producing praise and questions by reminding them to "Mind your P’s and Q’s" each time they looked at first drafts. The moving of response away from negative comments and toward both questions and positive comments helps to create an environment for the kind of extra-textual talk that Baum Brunner (1990) has identified as helping students learn to talk about texts. Teacher Modeling To model appropriate response, I used a sample essay, reading through it once and then again. The second time I included responses to the paper in the form of positive statements about parts that were particularly well written and pertinent questions about parts that were not so well done. Student Practice The students were given a short essay with a variety of problems as well as some good parts, and they were asked to practice giving written responses to it. I collected these and reviewed each one prior to the next class. Continuing 41 to model, I praised their successful efforts to respond, and asked questions about responses that were not appropriate in either form or focus. (e.g. "Could you ask a question instead of being directly critical of his opening sentence?") When the practice papers were returned, the students were also given a "Revision Guide for Readers/Writers Responding to a First Draft Paper" to further emphasize the type of response they were being asked to do (see Appendix B). Class Arrangement When first drafts were due, half the class came to one class meeting, and the other half came to the next. In addition to providing a more intimate atmosphere for peer response, a primary reason for this arrangement was to allow time for the teacher to confer individually with each student. The writers’ task in conference was to tell the teacher first, what they knew they had done well in the paper, and second, what they thought they might work on when they revised. The teacher rarely offered more than agreement since a quick read through their papers usually showed that the peers were addressing both successes and problems in their written responses. Upon arriving on a first draft day, students were asked to place their desks in a large circle, leaving enough room between their chairs so they could get in and out easily. 42 An empty desk chair was left in the center of the circle to hold papers that were placed there when students finished responding to them. They then picked up another paper to work on, and they continued this pattern until near the end of class time. Class Activities After asking the students to be sure their names were on their papers, the teacher went around the circle picking up the papers and then passing them to other students for reading and response. In the infrequent event that students arrived without a paper, they were asked to read over another student’s shoulder rather than deprive someone else of having a paper to read-~at least they would have the advantage of seeing what others had written. When students finished reading and responding to a paper, they placed the papers on the center desk and picked up another that they had not yet seen. Depending on the assignment and on the length of time any individual took to respond to a paper, each paper received between four and eleven sets of response over the class period of approximately 90 minutes. Respondinq Prior to the response sessions, the teacher regularly reminded the students to respond to each other’s writing with P’s and 0’5 and asked the students to keep in mind what they knew about the writing assignment and what they knew 43 about writing in terms of content and development of ideas, organization of ideas, language, and voice. As they responded to peer writing, the students kept a record of the papers they had read. Writer’s Notes In addition to writing responses on the papers of peer writers, students were also asked to keep a record of the authors of the papers they read as part of their class Writer’s Notes. Beside each name they were asked to write one or a combination of the following: what was best about that writer’s paper, or what needed the most revision work, or what their best response was to that person’s paper. The purpose of this activity was to insure their continued active participation in the response sessions. At the end of the class period, they were asked to write another Note which they addressed to themselves and in which they predicted what they would work on when they revised. Often these notes reflected what they had identified in conference, but they also included improvements suggested by classmates. They would refer to these notes when they began their revision work on their papers at home. Writer’s Notes were used for a variety of activities including recording paper assignments, pre-writing activities, and occasional questions related to response and revision activities. 44 Other Class Activity Students were asked to keep a journal based on their summaries and responses to newspaper op-ed articles. They did three a week until it was time to begin the fifth paper of the course. The fifth paper assignment asked students to write either a letter to the editor or an opinion piece. RESEARCH PARTICIPANTS My case study centered on a freshman English class at a mid-sized Michigan university. Of the twenty-four students in the class, I selected three for an in-depth analysis. From the half of the class that had the better record of attendance and participation, I looked for students who had completed all of the first and second draft of all five paper assignments because the revisions they made were part of the basic data to be studied. Equally important was their attendance and participation in each revision session since their responses were also part of the basic data to be analyzed. Finally, selection was based on differences in the overall quality of writing. Of those in the half of the class who had met the main criteria, two females were considerably better than average writers, two females were above average, and one male was a writer of average ability. In the belief that a variety might produce more information, I selected one from each of the two higher categories as well as the average writer to be the focus of my study. The three, Mickey, Sandy, and Bert, were all 18 year-old 45 freshmen who had grown up and been educated in Michigan. Approval for the study was granted June 18, 1992, by the Michigan State University Committee on Research Involving Human Subjects (see Appendix A). The students in the class signed forms giving permission for their work to be analyzed when the semester was completed. MEANS OF PROCURING THE DATA For the in-depth analysis of the three students’ response and revision processes, data was gathered from two main sources: Responses written on peer papers throughout the semester, and revisions made from first to second drafts on all five papers assigned. Additionally, information was gathered from the Writer’s Notes, a form of journal or writer’s notebook which contained, along with assignments and pre-writing exercises, the students responses to various questions relating to their writing, responding, and revising activities. Responses to Peers Response data was gathered from first drafts where peer responses had been written in different colored pencils selected by each responder. The colored pencils allowed a fairly efficient method of discerning the written responses of each student. All responses were recorded except for the occasional editing correction of a spelling or mechanical N... -—O r. {v- 2. .pk v. :4 C e F. 46 error . Revisions Made by Responder/Writer Revision data, the second main source of information, was taken from comparisons between first and second drafts of all five papers written during the semester. All revisions were recorded except for the occasional editing correction of a mechanical or spelling error. Writer’s Notes Separate from the primary sources of data is the Writer’s Notes information that came from the students’ writing notebooks which were used for various in-class activities. Two Writer’s Notes entries provided information regarding the students’ working and conceptual views of revision. A Writer’s Note assigned a week before the end of the semester requested that they ”Please recall what you know or have learned about writing in this class that will be of value to have thought about/or to refer to as you read and comment on your classmate’s first drafts." The second part of this Writer’s Note asked "Did preparing this list help? Explain." A week later, for the final Writer’s Note of the class, I asked students to give I'your current perception of what is involved in revision. Compare it with your sense of revision prior to taking this class. Give your opinion regarding the peer group work in which you participated." Although not subject to the analysis that "F’ ‘ _..- a 47 the responses and revisions were, these two sets of notes provided an informative look at the difference between the students’ concept of revision and their actual practicing knowledge. MEANS OF CLASSIFYING THE DATA Five means of classifying the data were employed. Three were used in categorizing Responses to Peers (Qualities of Writing, Levels of Change, and Forms of Response), and four were used in classifying the Revisions by Responder/Writer (Qualities of Writing, Levels of Change, Revision Operations, and Sources of Revisions). Responses to peer writing were coded according to the following three categories: QUALITIES OF WRITING * Ideas * Organization * Wording * Flavor LEVELS OF CHANGE * Theme * Sentence * Phrase * Word 48 FORMS OF RESPONSE * Question * Positive Comment/Praise * Critical/Negative Comment Revisions made between first and second drafts were coded according to the following four categories (note that the first two are the same as the first two analysis categories used to code responses): QUALITIES OF WRITING * Ideas * Organization * Wording * Flavor LEVELS OF CHANGE * Theme * Sentence * Phrase * Word ~REVISION OPERATIONS * Reordering * Addition * Substitution * Deletion W: 49 SOURCE OF REVI S ION * (rr) Due to response received * (n) Not due to response received An explanation of each of these categories and their individual sub-categories follows. Qualities of Writing *Ideas *Organization *Wording *Flavor I classified both revision and response data according to the categories designated in Diederich’s Measuring Growth in English (1974) in which he described his work with teachers in developing a consistent approach to rating writing. Through surveys, a factor analysis, and workshops with teachers, Diederich’s study evolved the following list of factors or qualities of writing: ideas, organization, wording, and flavor. (See Appendix C for a detailed description of Diederich’s categories.) I referred to Diederich’s descriptions of these four categories for guidance as I classified my data. Examples of responses that would be placed in the Idea category included "Did you want her to jump?" or "Nice .r~ 'v‘ a. v‘ F‘- ‘A 50 background info." Responses typical of those found in the category of Organization included "Could you add a transition to make it easier to see one side from the other?“ "Good organization of tests.“ Responses typical of the Wording category included "Maybe a different word could be used?" and "Good choice of words." The Flavor category had few entries and they were difficult to categorize since the comments might easily have been designated as related to Wording. One comment identified in this category, "Nice job!" appeared to refer to the flavor of the author’s commentary. Levels of Change *Theme *Sentence *Phrase *Word Both the response and the revision data were classified according to the four “levels of change" categories that Nancy Sommers (1980) has used in describing revision: word, phrase, sentence, and theme-—referring to the extended statement of an idea. The similarity of the "levels of change“I to the "qualities of writing“ invites the question, "why use both?". In fact, each offers information that the other cannot. Sommer’s category indicates the level of the change, while Diederich’s "qualities“ give information about ~: .\~ Ibp L¢ Ax» s «\s \,\x it 51 the quality of the changes at any level. For example, looking at Sandy’s responses to peers on the first paper, one can see that four of 24 comments were word related which suggests that a sixth of her responses were the typical "thesaurus" type comment seen in freshman writers. Yet, a look at the "qualities" categories shows that three of the four responses were rated as Idea—related comments with only one being a lower level "thesaurus" kind of comment in the Wording category, thus indicating that three quarters of Sandy’s word level responses were of a higher quality than might have been expected or discerned. Determining the level of change intended in a response given was more difficult than identifying the level in an actual revision. In coding “levels of change" in responses given, I took into consideration the context within which the response was made. An example of a response in the word category is "Maybe used too often?" (Responder was referring to a particular word). A response referring to a convoluted phrase asked, "Could you make this clear?" At the sentence level were comments such as "Good setting" (referring to a sentence that provided setting details), and "Could this sentence be made clear?" Responses coded at the theme level included those which appeared to encompass more than a sentence either in area addressed or in what was requested. Examples: "Very good intro." and "What about the other two tests?" Coding “levels of change" was less complex when the a‘. If -5 .A -2 t; 52 source of data were actual revisions by the Responder/Writer. Examples are as follows: A word level change might be the substitution of one word for another. A phrase level change could involve deleting duplicate information. A sentence level change could be the reordering of a formerly confusing statement. A change at the level of theme could involve the addition of more information on a particular topic. Form of Response *Question *Positive Comment/Praise *Critical/Negative Comment Responses to peer writing were also sorted according to the form of the response produced on peer papers. In this category, responses fell into one of three areas: direct critical comments, direct praise, and questions or indirect suggestions. Although students were asked to avoid direct critical comments, their past exposure to this form of response assured that the critical comment could not escape being one of the forms of response. Typically, direct criticism tended to center on word and sentence level concerns. Among the types of comments produced were "Confusing" or "These last sentences seem choppy“ or “This sentence could be made 2." Direct praise only occasionally attended to word level 53 concerns. Compliments tended to come in comments such as, "Very nice, I can §§§_what happened" or "Good use of quotes" or "Good informative intro." As with direct praise, the questions rarely addressed word level concerns. Characteristic questions included ones like "What do you feel about it and what it means to you?" and "Could you insert where you are talking about so the reader doesn’t get confused?" or "Could these paragraphs be combined because the ideas seem to fit together?" Occasionally, a response would be composed of two sentences-~the first a direct comment, followed by a helpful question. In these cases, the entire response was coded as a question. When a response was phrased like a question, but with no end punctuation, it was assumed that the responder intended to question, and the response was recorded as a question. In cases where a question mark was appended to a sentence that was not specifically phrased like a question, the question mark was taken as a signal of the responder’s intent to question, and the response was coded as a question. Although not a category that would supply direct information about the processes of responding and revising, "Forms of respone" data allowed me to speculate about correlations between the form (critical comment, praise, or question) and the level of concerns within the "qualities of writing" and "levels of change" categories. That is, I was interested in the possibility of identifying a connection . .7— 54 between the way a response was shaped and the Responder/ Writer’s attention to lower or higher order concerns. Revision Operations *Reordering *Addition *Substitution *Deletion Revisions by the responder/writer were analyzed not only by the "qualities of writing" and the "levels of change" categories but also according to the four "revision operations“—-deletion, substitution, addition, and reordering--that Sommers (1980) used. In her research, she noted that student writers most frequently used the operations of deletion and substitution. By classifying responses in these categories, I would be able to see if my student writers followed the pattern Sommers identified in her own study. In recording the revision data, I found these four categories adequate for describing all of the revision acts that occurred. Typical descriptions of revision might be "Deleted unnecessary verb," "Substituted a similar phrase," Added information,“ or "Reordered a sentence for clarity." Source of the Revision *(rr) Due to the response received *(n) Not due to the response received ' 1 '1 l 55 Finally, I identified the source of the revision, whether or not it was due to "response received" on the first draft or not. If a revision occurred, and a response on the first draft appeared to have contributed to the revision in any way, the revision was identified as due to response received (rr). If the change could not be credited to any written response, the coding indicator was (n). The purpose of this category was to try to identify possible correlations between movement up toward high level concerns and increased competency at undertaking revisions on one’s own initiative. METHOD OF ANALYSIS Data from the responses to peers as well as the revisions by the Responder/Writers were coded into the appropriate categories and subcategories as identified above. (See Appendices D, E, and F.) Upon completion of the coding, I prepared a table for each of the Responder/Writer’s response data and one for each of the Responder/Writer’s revision data. Because of the multiple categories of analysis, separate tables were necessary. (Completed tables for both the response and the revision data for one student are included on the following page to help clarify the analysis process being described. All three sets of data are included in Chapter 3 where their implications are more fully explained.) The numbers in the Response—related tables represent 56 the amount of responses in each sub—category within the categories of "qualities of writing," “levels of change," and "forms of response." (Note that the "qualities of writing" category totals are on the horizontal lines that begin with the bolder letters of I (for Ideas), 0 (for Organization), W (for Wording), and F (for Flavor). The numbers in the Revision—related tables represent the amount of responses that occurred in the sub-categories of the categories of "qualities of writing," “levels of change," "revision operations," and "sources of revisions." The tables were designed so that they would display the data in a hierarchical fashion; that is, if a student moves up the writing hierarchy in terms of kinds of responses or revisions over the semester, this movement would appear as a movement from lower left toward upper right. A student who begins, as does Bert, with heavy word emphasis, but who gradually moves toward a greater focus on organization and theme, will have a table where the clusters of the numbers entered in each box gradually move from the bottom left area of word concerns toward the upper right area of organization and thematic concerns over the course of the semester. 57 Table l Bert’s Responses to Peers 58 50:0 topcaot o. 290E map a. 88. 2.096 0. .20. .0: 0p 802:8th 85.9.6 050... .05 5.00200 .0553 .0 32.030. 05 35:50 .90: 6.5903 50:05:00.0 .309 53.0.. $5000.00 @8008. .0 ”Eco... 05 059:8 EoEEoO .0250 .EoEEoO 0228.. €05,030 “0 .a .0 5.00900 .0965 .0 ”.05.. 0:. 9.568 0605 60:28.... 6.0.2.. 05>) ”56030.03 . sax III-mum n. __ 8 8 am. __ em 9 3 3__ 2 2 n. a. e _ m __ e e m N N = N N N _ _ =0. N N. N mm o __ o = o = o a o _e 1 __ i __ 5. r __ = __ = ‘10 I o __ o __ o _?_ _ __w_ N N .5 = = =e. _ __s_ N N o __ = = = a = __ __ .0 on N N = o =e. _ _ _KN N _ _ Arm m n o __ __ =8 m. m o = =e. _ _ = a on N N = .= _3N N _ _ =. .0 SN _ _ =8; e n =8... e __3e N N 480 N N =— =NN N _ _ = = o R _ _ __NN N _ ._ __..N N N 8 N _ _ S e N e __a BN _ _ =8 e e _ =3 e e EN N _ _ =2 n m . 0 I. I. h. I. I. w m. 5 en. en 24—... m. 5 an. ea 63%... m. 5 5... ea 63 a. m. 5 :9. ea 63:... m. 5 en. en aka—nu > 500.. >_ 500.. 5.0 On. __ 500a. 2000a .3: Eve.— 0. 89.2.8: . e.gea 59 Table 2 Bert’s Revisions by Responder/Writer 2 Tel 1') I 1". R:- vicious- by 1(0 «ponder/5 Vrlh-r 6O 6:0 rouge. o. 20.06 020 a 88 _ £098 0. .22 .0: 8 390.0028 3.6.9.“. 080» .9: 2000.8 .92.; .o 8530. 05 39:8 .26: .0520; 5053.090 .88. ”$50.. .698 205058 .8332. 05 3658 8:28 .cofisefim .8562 8:863. ”932. 002002 8009.2 0. 02p .0: n c236. 800E E. l .5332 .0 0059. ".5 .5828 .866 .o £26.. 2.. 33:8 285 682:8 68>: .662. teenager...) 3. essence. 3 55.5.? N w Hana. 58.52.02.039: b. nae—35¢ 61 A detailed discussion of the performances displayed in the tables is given in Chapter 3 where each student’s data is presented separately in order to look for evidence of movement within the writing hierarchy in both responses and revisions undertaken. My analysis of the response performances occurred paper by paper in chronological order. As I analyzed the various categories in the second and subsequent papers, comparisons were made to the student’s performances in the previous papers. The discussion of the student’s response performance over the semester was then followed by a paper by paper analysis of the student’s revision performance over the semester. Again, as I looked at the data from each paper, I made comparisons with the performances in the preceding papers. This method of analysis not only provided an elucidation of each student’s performance as seen in the categories used to analyze the processes of response and revision, but also offered a sense of the student’s movement within the writing hierarchy over time. In Chapter 4, where the focus is on determining if responding has any correlation with revising behaviors, the data was analyzed in terms of a comparison between first and second half of semester performances. (Because the first half of the semester had two paper assignments and the second half had three, the results were averaged.) More specifically, I compared the "qualities of writing“ and the "levels of change" data in the responses to those in the r, . -v' C' r;".'. ~,,n .}—> .— ANNA... ,- "<~r "’~,_ .4" he.“ ‘..-p_. y- l ~.. T" ( T 62 revisions. With each student, I looked for correlations between the categories; e.g. will their revisions reflect what they say when they respond to peers? If not, will it happen over time, and will time also show movement from lower to higher level concerns? An additional question I hoped to answer involved the number of revisions on a Responder/Writer’s paper which were due to response received (rr). I compared the (rr) revisions with other data on the graph, both within an assignment and across the semester, for the purpose of looking for any correlation between movement up toward higher level concerns and increased competency at undertaking revisions on one’s own initiative. The fourth chapter concludes with a comparison of observations based on the three in-depth studies. In making these comparisons, I attempted to determine whether internalization of writing knowledge appeared to be similar in the cases of all three students. In the final chapter, I use the information gathered from the Writer’s Notes entries written in response to my questions regarding the students’ concepts of revision and their actual practicing knowledge of it. While not a part of the main study, these notes provided an illuminating view of certain aspects of the study. CHAPTER 3 MOVEMENT WITHIN THE WRITING HIERARCHY IN THE RESPONSE AND REVISION ACTS OF THREE FIRST YEAR COMPOSITION STUDENTS In order to accomplish the primary purpose of my research, measuring the effect of responding on the responder as writer, I needed to find a more sensitive method of measuring the changes that occur in a student’s writing knowledge than is usually seen in the typical end of the semester study of final writing products. The purpose of this chapter is to demonstrate that it is possible to identify changes in writing knowledge within the time period of a semester. To achieve a closer look at the fabric of response and revision, I selected the five ways of measuring the acts associated with responding and revising which were described in the previous chapter. Briefly, response acts were viewed through the lens of Diedrich’s "qualities of writing" (Ideas, Organization, Wording, and Flavor), and Sommers’ ”levels of change" (Theme, Sentence, Phrase, and Word). They were also sorted according to the "forms of response" (Question, Praise, or Criticism) employed by the responders. Revision acts were subjected to four levels of analysis, two 63 64 of which were the "qualities of writing" and "levels of change" used in the response analysis just mentioned. In addition I used Sommers’ "revision operations" (Reordering, Addition, Substitution, and Deletion). I also analyzed the revision acts in relation to the "source of revisions" (whether the act occurred due to response received or not). In analyzing the response and revision activities of the three students in this case study, I looked for interrelationships among the categories to provide the best picture possible of the hierarchical levels the writer was concerned with during each paper assignment. I was able to trace evidence of growth from lower to higher levels of the writing hierarchy as the responder—writer attended to writing concerns in both response and revisions. In connection with movement to higher levels, I also looked for a developing ownership of revisions, because higher level revisions that occur due to the writer's own initiative rather than to response received from others would indicate more concretely the student’s developing experience as a writer. These combinations of measurements provided a great deal of information as well as mutual support for findings, a support similar to that offered by the ethnographic technique of triangulation. The measurement totals are displayed in tables that accompany each case study discussion of the response and revision acts produced by Sandy, Bert, and Mickey. In addition to the information ~n~ FYI. .— .n- v- v‘ .unyq 0..-»! a-..‘.~ v-54-» '73 ( (I) (I) I) “A v ‘7‘. "r; ._ ‘ ~. a. \. r- ‘n Ll-yk; \ r- ‘\ Ina, t.~v ‘s Q‘ \g N ~“.. . “‘h ‘ III \ ‘t “V; s a ;>,‘ “‘ . Ny- ~§ 65 provided by the totals in each column, the tables are hierarchically arranged so that if the student moves from word focused concerns to higher level concerns over the course of the semester, this movement will be apparent in the movement of clusters of numbers from the lower left area of word concerns toward the upper right areas of organization and thematic concerns. That is, without attending to the specific totals on the data sheets, it is possible to visually apprehend the student’s progress by looking for movement of the number clusters from the lower left toward the upper right. As the discussion of each case study indicates, movement does occur in terms of the students’ knowledge about writing within both the process of response and of revision. Thus, the detailed discussions of the student’s movement within the writing hierarchy indicates that this measuring system does offer the potential for measuring the effect of responding on the writer, which is the subject of Chapter 4. First Case Study: Sandy Of the three students, Sandy’s writing ability fell midway between the other two students; her writing was generally in the mid— to high B range, and her effort and participation were usually among the highest in the class. Her efforts with both her Writer’s Notes and her journal entries were above average. In total number of responses given to peers, Sandy was also in the middle; however, she w. .3 ~.~ C. A: v. Q. 3.. r N. A; A h. N: CA p» a: L. 3. r. rk. ‘V .1 CC r. A. 66 made more revisions than either Bert or Mickey. Analysis of Written Responses to Peers Sandy produced a total of 83 written responses on her peer’s first drafts over the course of the semester, ten less than Mickey, and 39 more than Bert. Possibly because the response to the first paper followed so closely on the heels of the hour of practice on how to respond, the first paper contained her highest number of responses, 27. The number of responses for subsequent papers were 17, 16, 11, and 15 respectively. Sandy's responses fell into every sub- category within the three main categories ("qualities of writing," "levels of change," "forms of response”) except one. None of her responses ever appeared to deal with the "flavor" or "voice" of writing. The likely reason for this may be that this quality tends to be a concern of more accomplished writers. 67 Table 3 Sandy’s Responses to Peers Sand 3/ ’11::- I" _,.. .650 topcno. 2 200:. 030 a. 88. 2800.0 2 .22 .0: 00 80030000 00.6020 060. .65 2000.60 .0553 .0 ”0.2030. 0:. 000.50 .0>0_n. 050.03 00:00:00.0 .0000. ”0.2.0; 600000 0800.0. .0 ”E2. 0:. 02.05.00 E08800 00.50 ..:0EEoO 02:80 60.0030 .0 .0 .0 5.00200 .00:0:0 .0 :99. 0:. 03.0500 060.: .00:0.:0m .08.:0 .0.0>> 20020.03 68 ; >00. N N allfl. e. __ no .N.. o o.__ on Ell-m; 8 N. NLUN ..o=:e N o o e.o.N . m__ N.o.n o NFeNe... o LM 0 o __ o o __ o __ __ __ __ __ __ __ _F __ __ 0 Jr.” 0 . . m fi. N . . __.. . = __N. N . . _T . . __ __ __ __.. . = :0. m . N03. . . "IL—7 __NN N m __ o __NN e v a N . . __ __ __ e N . . __N. . . __ __N. N m =2 N N __ =0 . . n. . . 2 e e N =% .. o. . =8 .. o. . S .N m. N . . __ = N. e N . we BeeN __€Ne. New. see. 0 e . . =2 N N =0N e v =2 m N . .N n e . L I. l. I. I. we :0. :0 03—7. we :. :0. :0 03=$ m. 5 :0. :0 “vi—.0.v we 5 :0. :0 03 .x. m. 2 :0... :0 03 II III } I I r l- =. .0000 __ .0000 _ .0000 m 02.09 0.00.. 2 8000000“ 69 Responses to Paper I: In the category of "qualities of writing," 88% of Sandy’s responses were primarily in the Ideas sub-category with 8% in Organization, and 4% in Wording. As with the large number of responses she produced on these first papers, it seems likely her intensive focus on ideas was at least partly due to the recent lesson on how to respond. Comparatively, in the “levels of change" category, she addressed the Theme level 66% of the time with Sentence and Word at 17% apiece and no Phrase level responses. Sandy’s emphasis on higher level concerns as she responded is reflected in the "qualities" category and supported by the "levels" category results. As she successfully addressed higher order concerns for the most part, Sandy used the Praise and Question "forms of response" far more than the Critical Comment. Sandy’s Comments were 25% of her responses, with Praise and Questions at 75%, or 54% and 21% respectively. Although a problem with critical comments is that they can frequently tend to focus on word level concerns, Sandy's comments only did so once. One reason why her comments reached a higher level may be because they were not confined to the typical direct critical comment but came in an indirect form instead, e.g., “Maybe explain your father’s reaction" and "Maybe describe what you saw in the car." Examples of her primary form of response--Praise at the Theme level-—are "Very good intro," "Good description. Can see the picture,“ 70 and "Good conclusion." Although her Praise comments nearly always treated concerns beyond the word level, they tended toward the general. However, Questions rated at the Theme level were considerably more specific; for example, "What was the after effect?" and "What was people’s reaction in car and yours?" For a beginning composition student, Sandy, in this first paper, did remarkably well in avoiding that tendency of the inexperienced writer—reviser to attend primarily to wording. An important question in regard to Sandy’s responses is, will she maintain this admirable focus over the semester, and, if she does, will improvement be seen in terms of more specific responses at the higher levels if she continues to respond with such a high percentage of Praise. Responses to Paper II: Sandy’s responses in the Ideas category for the second paper assignment totaled 65%, less than we saw in the first paper, but the loss shows up primarily in the Organization category with 23%. Wording contains only 12% of her responses. The movement toward Organization responses is probably related more to the Paper II assignment format than anything else, since it called for two points of View. Just as she remained focused on the higher order "qualities," Sandy also maintained the emphasis on Theme in the ”levels of change" category. However, unlike the first paper where Sentence and Word level were equal, she 71 addressed the Sentence level twice as often as the Word level. Hierarchically, this may indicate that she was beginning to move away from a concern with word level changes. At the same time, Sandy’s "forms of response" seem to correlate with the kinds of responses she produced. Comments dropped to 6%, while Praise increased to 64% and Questions increased by a nearly a third to 30%. Her positive comments became more specific; nearly half gave reasons for the compliment, e.g., "Good conclusion. Was weighted out equal on both viewpoints," and "Good intro! Good background info." Sandy also asked more well developed questions, e.g., "What do you think about now since the incident has occurred?“ Her comment on wording was not even a criticism but an actual suggestion for an improvement: “both of you" is substituted for "you two." As this overview shows, Sandy maintained her admirable beginning and even showed a bit of movement up the category hierarchies. Responses to Paper III: In this third paper, Sandy’s Ideas responses increased to 69%, probably because she had no responses in the Organization category. Since this paper was a type of survey, it tended to organize itself during the planning stage and those Sandy read may have presented no cause for concern in regard to organization, at least at the Theme level. Instead, 31% of Sandy's responses were found in Wording. Again, a correlation can be seen between the ||||||. ~u. ,r. ;. l. 4.. . . .. g a. . \ . ‘ .. r. .: .. u S E .3 .. E C . r. C a. . C . C 2. P. . .w r . .1 .. H 0. . .H a. 5 Ti T. .7 .3 2,: ‘1. r C r. .v r ‘ ‘ _... .r. .M a. . .. 3. A. 3 S . . N. r. c. N. S a. . .3. N. C E Q. .3 A. V.“ v . .C 1 C. C. .3 0. C. S v. . ,. .. . r . F. A. C. : .3 r U .J .. 0 . .. .3 I. Z. r v . 2. c. :u L. H. N: v... «C A. A... 7. ma. : . -. .1. r“ .r. a. 5. Lu ;. L. .. .. u s... .. a. .. A. Q. .0 .2 .n u n. «0 r... I .. 0 n." a. .1 a. v. v. a? .u r. 7.. C. .3 T. .L n: .t «a. v. . . A: . . pr. 72 “qualities of writing" percentages and those in the "levels of change" category with Theme responses totaling 63%-— within 6% of the Idea related responses. The slight increase in Wording responses was accompanied by increases in both the Word and Phrase sub—categories of "levels of change" and a fifty percent reduction in Sentence oriented responses. The improved quality of the Word and Phrase responses is worth comment because over half have been phrased as helpful questions such as "Do you want this word here?“ In regard to the use of Questions as a "form of response," the third paper had the highest percentage of questions up to that point—-44%; the same percentage was shared by the Praise responses. The Comments doubled, but at 12% (or a total of two out of 17) this is not a concern. It is notable that Sandy was asking even more questions, and there were substantive ones: "Is this an important issue to you--long-lasting taste?" and "Great background to your testing categories, are there any other tests being conducted? Information ones?" Further, her Praise responses maintained, if not improved upon, the specificity that was noted in the second paper, e.g., "Good conclusion. Good recommendations" and "I like how you added this part in to relate to thirst quench." Overall, even with the heavier focus on word level concerns in this third paper, a gradual movement up the hierarchy is evident. 73 Responses to Paper IV: Responses to the fourth paper indicated a reversal of the Organization and Wording distribution in the third paper, as well as an increase in Idea level concerns. Perhaps because this paper involved gathering information that did not, like the product survey, predetermine the format, Organization level responses reached a high of 27% of Sandy’s total responses to peers. With no responses in Wording or Flavor, the lion’s share of 73% were found in the Idea sub-category. A similar upper level scoring occurred in the "levels of change" sub—category of Theme where 82% of the responses were focused. The remaining 18% were aimed at the sentence level. Remarkably, Sandy paid no attention to word level concerns in either the "qualities of writing" or the "levels of change" categories. Likewise, there were no Critical Comments in the "forms of response" category. Responses of Praise increased to the level seen in the second paper, and Questions decreased 8% in comparison to the previous paper. While the Praise responses were not generally as specific as was seen in more recent papers, the questions continued to be pertinent and helpful. Praise responses varied between "Very good!" (a response to a well- developed idea) and "Good explanations." Pertinent questions, for example, involved organizational matters, "Do you think these could all be related?" and development of ideas, "Could you explain this? More examples?" While responses of praise may not have become any more specific in hr“ _,.- v, I... III . , , l : . . . . . . v u , . r u r . «\V Y s w u .r . r“ 4. 2. .H w. .3 J. h. ,. A. .. . L. ~. .. .Q h. A... . v . . a . :0. 3L «D WA. Wu. Q. .Q by . \ AW. . .. .2 S. r. .. L. a. m. T . up. .2. n 2. a: u. t. L. a. .3 . . .. L. up. :0 . . .C .. . :0 2. yr“ -nu LI. .3 .3 .... ..;.. C. C. .L . r K . . .J. .. s. u. A. a. n. C. vs T. ,.« ”A s - A p.“ v L. a... _“ 74 the fourth paper, and while the percentage of Question responses was slightly less than in the third, the long view of Sandy’s responses shows both an increase in Praise and Questions as well as more specific concerns at the higher levels across the semester. Responses to Paper V: The increase in Idea level concerns that appeared in the past two papers seemed to have reached a pinnacle with the fifth paper where all of Sandy’s responses fit into the the sub-category of Ideas in the "qualities of writing" category. Correspondingly, responses for "levels of change" were only at 7% each for the Phrase and Sentence sub— categories, with the bulk, 86%, in the sub-category of Theme. The focus on development of ideas showed just as much in the "levels" sub-category of Phrase, "Could you explain this word(s) more?" as it does in a Sentence level response of "Does this sound confusing to you?“ and in Theme level responses such as "What do you feel about it and what it means to you?" This final paper, an opinion piece, had as its foundation regular journal entries based on reading opinion pieces throughout the semester. Possibly there is a connection between her journal work and Sandy's complete focus on ideas in this paper. It is most likely due to Sandy’s total focus on ideas that not only were there no Wording comments, but, in terms of the "forms of response,” 75 there was just one in the sub-category of Critical Comment. And even this comment was an indirect and very specific suggestion: "In intro, maybe state your opinion of it and what caused you to have this opinion." Specificity occurred as well in the Praise responses for the most part, e.g., “Good paper! Showed reason and explained why!" There is certainly the possibility that this particular writing assignment did much to keep Sandy focused on the higher levels of concern, yet such a result has not been unexpected considering the gradual movement Sandy's responses have made up the various category hierarchies. Analysis of Revisions by Responder/Writer Sandy made a total of 80 revisions over the course of the semester, compared to 55 for Bert and 63 for Mickey. As with her responses to peers, Sandy’s greatest effort occurred on the first paper, followed by quantities of 13, 9, 26, and 5 respectively. Revision analysis involved four separate categories, "qualities of writing" and "levels of change," which were also used in the previous response analysis, and two others, "revision operations“'and “sources of revision." Sandy’s revisions fit into every sub- category. Even "Flavor," which had not been addressed in responses to peers, was a focus of revision later in the semester . 76 .31.. .1 _ _ =__ .._ L L L i, .l L L L. .l L r; .. L- l c L L .. L > .31.... >_ if: .L S F! r i = 3:231 ‘LGQCQ 5-2-87. v m.- H 2 5% L.~=L>>\La~=~=q~v..u- K: itaétusaz Table 4 Sandy’s Revisions by Responder/Writer 77 avcam 00:0 tonczo. 0. 20.08 030 m. .88. 2.096 0. .90. .0: 00 A“000500.00 00.6.9.0 069 .05 5.00050 .95....) .0 8.5030. 05 3.0800 .96.“. 650.02. 50.53090 .800. 55.0.. 20028 22.0.30 6833.. 05 35.68 8:200 £22.33 .5502 05.00.02. ”0.05. 002000. 08090. 0. 0:0 .0: a. c0330. 3008 E. -- .c0v..>0. .0 0050.... “.5 2000.00 .0965 .0 £05.. 05 00.0800 0602.. 089:3 .0825 0.0.5 "£60206? u >8. 3. mm mm Np a w 5 c8 .3 .5 w 0 H308 geisha—0:008: .3 gauge“ II'I .. P... .A‘ ~v- \ y- u... Q" 'r .- ..\~ - » ~_‘. «50 v a L N. 78 Revisions on Paper I: The ”qualities of writing" sub—category, Ideas, contained 40% of Sandy’s revising efforts on this first paper; 30% of her focus was on Organization, and the other 30% was on Wording. Without a review of the "levels of change" category, these results would seem very impressive for a freshman writer. However, a survey of the "levels" sub-category of Theme indicates that none of the revisions were at this level--a result more in keeping with what might be expected of a freshman writer. Yet, Sandy’s revisions were not limited to the Word and Phrase levels, since her Sentence level revisions reached 44% of the total changes made. Although this focus on Sentence level or Word and Phrase looks good for a first paper, the question is, will it bear up under the scrutiny available through the analysis category of "revision operations" which identifies the sophistication level of the changes. Considering that substitution and deletion have been identified as less sophisticated operations, and Sandy used them only a third of the time in her first paper, she seems to have done fairly well. This perception is supported by a review of the operations involved in her Sentence level revisions—-six of twenty-seven were the subject of Reordering, five were Additions, and only one was a Substitution. Since beginning writers are known to favor substitution and deletion operations, Sandy's performance indicated she was a bit more sophisticated than might have (I) «“9 ' V (I) r. t.‘ c.._. ’1! w A s 79 been expected. Finally, it is important to consider the "source of the responses" which led to Sandy's revisions-~whether changes were primarily due to response received from others (rr) or due to her own initiative (n). It would seem likely that a first paper would have more revisions due to response received, and then less as the semester progresses, but Sandy has almost 60% self-initiated revisions. This is not too surprising, however, considering her performance in the other categories. It will be interesting to see whether Sandy will have maintained the standard she set in this first paper or even moved farther up the category hierarchies, and whether doing so will be accompanied by an increase of self-initiated revisions overall. Revisions on Paper II: Sandy’s rather surprising number of Idea related revisions in the first paper did not turn out to be a fluke. With her second paper, 62% of her revisions were at this highest level of concern within the "qualities of writing" category. This increase profited from a reduction in Wording level as well as Organization level revisions, with Flavor still not a subject of revision. Accompanying the slight increase in revisions related to Idea development was the appearance of 15% in the "levels of change" sub—category of Theme. This change reduced the Sentence score slightly while leaving Phrase and Wording about the same. As such, .’, I. I ll Y... If“ V“ Q. r_- . ‘ o \ . - 4 n A v. .. .. u. . :1: 2.. r“ .. .gu. .. r. .. :. L. .f. 1 .. p: A: .3 A. a: .n. .. . L. C . . . . ¢ . . i a... a: a ‘ «C .c I. . . v. r . .a . u. n. C. . . . ”l. «u. .I. ‘Q 3. w,” L. r... 2. L. 2,. a. s3 .5 ,u n. v. Q. .1 Q. .r.. A . r? w. 4. r... v. 2. .nn r“ w. an r a. n. v. d . . .. . . . . . . . n . . . . .3 I. a). 2 . LI... v. ml. », L o u 80 these revisions showed only a small movement up the writing concerns hierarchy. In looking at those changes in the light of "revision operations," it is possible to see a slight hierarchical improvement relative to the percentage of Additions to Ideas. There, a small movement, from 33% in the first paper to 46% in this paper, can be seen. Overall, however, the "revision operations" of Deletion and Substitution have not lessened but have actually gained a small percentage since the first paper. While there has been a kind of balance between gains and losses in the three main categories, there has been a small increase in the percent of revisions Sandy made on her own initiative. Although not much movement can be observed from the first to the second paper, Sandy is at least maintaining the performance established in the first, and her self—initiated revisions have even gone up by 10%, a respectable increase. Revisions on Paper III: With no revisions occurring in the "qualities of writing" sub-categories of Organization or Flavor, both Ideas and Wording increased slightly over previous papers to 67% and 33% respectively. Possibly because the paper topic was a survey, the Theme sub-category percentage was slightly less than the previous score, but the Sentence sub-category 81 received a boost to 44%, resulting in a 9% higher score of revisions in the upper section of "levels of change." The “revision operation" of Addition came to 55%, just under the previous paper’s score. Not too surprisingly, the Wording revisions had 33% located in the lower level sub- category of Substitutions. Nevertheless, with over half of her operations in the upper levels, Sandy is continuing to do fairly well at addressing more global issues. In assessing the results of the second paper compared to the first, the primary change noted was a 10% increase in self—motivated revisions. With this paper, the percentage of self—initiated revisions as well as the other categories generally seem to be remaining about the same. Revisions on Paper IV: The "qualities of writing" category of Ideas, which had held fairly steady since the first paper, dropped this time, but it was due to Organization related revisions which grew to 38%. This made for a 21% percent increase in the upper level of "qualities," while Wording revisions dropped to a remarkably low 8%. One revision is in the Flavor sub- category this time, but its story is not an upwardly mobile one—-the operation performed on it was a deletion, and the source of the change was a response received from a peer. But in other revisions, a surprising upward turn is found. A comparison of the "levels of change" upper and lower categories shows that while Theme has only 8%, if it is n-‘vn v.,.-.a. V‘V'V .— A“... . «7" .4--. . o1 - Q 82 combined with Sentence level changes, the upper level was where 69% of all revisions were made. Word level changes dropped to a low of 8% and with Sentence level changes at 23%, Sandy’s changes at the lower levels showed a healthy decrease from the previous paper’s lower level Word and Phrase changes total of 44%. A corresponding upward movement showed up in "revision operations" with 72% of all revisions due to Reordering or Addition operations. Even the revision operation on the two Wording entries were split between upper and lower level operations. The unusual incidence of Addition operations within the Organization sub-category occurred most likely because this assignment, which was based on interviews, required choosing one format or another for presenting the information. Sandy changed her mind about how she was going to do this while she was writing her second draft which resulted in her adding transition sentences to the paper. The kind of ownership demonstrated by the addition of those transition sentences raised the number of self- initiated revisions in this paper to a new high of 85%. With word level concerns at an all time low, and the upper halves of "levels of change" and "revision operations" at all time highs, the upward movement is striking. However, considered in the context of Sandy having changed her mind about her format while she was in the midst of revising, the increases probably do not signify a dramatic permanent change. Yet, they do indicate that Sandy now has the FA,A,«> y..v5.. . --rv\vr~' .n..l. . y . .- 1 -‘.~ ~ In- \ .0 .. ,. 0...- .- sgnnJ-A. A! "a ”V “v v.. 5 ‘QYW: ‘. \A‘, .v: "E“ . 83 potential for taking considerable responsibility for improving her own writing. Revisions on Paper V: A definite difference exists between the fourth and final papers beginning with the actual number of revisions: 26 on the previous paper and only 5 on the last paper. In terms of the "qualities of writing," the percentages look more like those seen in the first paper than in the previous one, with 60% in the upper levels and 40% in Wording. The "levels of change" Theme sub—category had no revisions, and Sentence had only one, or 20%. Word and Phrase revisions totalled 80%, a surprisingly high figure that these lower level concerns had never seen before this. Altogether, the downward movement shown in the two main categories of ”qualities of writing“ and “levels of change" seems somewhat jarring in comparison with the previous paper. In “revision operations" a similar downward movement is apparent: only 20% are Additions. The others are in the lower levels of either Substitution or Deletion operations. While it is disappointing to see this reversal, it is interesting to notice the definite interrelationship between the main categories of "qualities," "levels of change," and "revision operations." Whatever direction the writer- reviser goes, these three categories of analysis tend to parallel each other. As might be expected, the "sources of revision" 84 information is similar to what might be expected based on what has happened thus far. Not one revision was due to her own inititative. The "sources“ score has paralleled the other main data sources not just in its lack of upward movement but in its reaching a new low. Oddly, this last source of information provides a potentially reassuring possibility: the revisions were all due to others’ responses, thus they may not necessarily represent Sandy’s level of sophistication as a writer and self-initiating reviser. As with the previous paper, there may be some connection between Sandy’s revision performance and the paper assignment. Since the paper assignment was based on a journal assignment that ran through the semester, Sandy had a considerable amount of time to think about her topic for the opinion piece. Possibly, Sandy did what has been observed in more experienced writers, she thought about the topic long enough to have done in her head the kind of revisions she might have done had she not thought about the topic for so long. A similar possibility is that, in having read opinion pieces throughout the semester much like what she wrote, she knew enough to be able to compose a piece that satisfied her personal understanding of what such a piece of writing required. Also, since this paper did not necessarily come from a great deal of personal experience, it may not have had enough potential for ownership. Finally, it should be noted that Sandy received no peer 0"“; -p-v p! ,....A- .‘A (v v-- “A- .Av‘ s, ..a rt: rm. 1...“ 0.. ‘61 s v R\M 85 responses that asked for anything more than word level changes, and one of the comments of praise she received was fairly accurate in its assessment: "Good paper, you illustrated all the points you planned to cover and did that well." Considering the accuracy of the comment, perhaps Sandy’s revision performance on this paper is not such a jarring move downward after all. After taking into account the circumstances of this last paper, and looking at Sandy's revision performance across the semester, it is difficult to conclude that she has made a distinct move up the writing hierarchy overall. Yet, the slight increase in ownership in the first three papers, followed by the dramatic increase in the fourth suggests that, despite the final paper showing, Sandy's taste of greater ownership will not be lost on her. SECOND CASE STUDY: BERT Bert was the lowest achiever of the three students in most respects including his writing quality which averaged in the low to mid-B range. Bert’s participation in class and his efforts with journal and Writer’s Notes assignments were average and sometimes above average. In total number of responses given to peers Bert was by far the lowest, and his revisions were also less than either of the other two students. This is not to say that Bert was an uncooperative student; in fact, he was a likeable person and he took the class as seriously as the average student does. 86 Analysis of Written Response to Peers Bert’s total production of written responses to peers over the semester was 43, considerably less than Sandy’s 83 and Mickey’s 92. As was the case with Sandy, Bert's first effort at responding produced his largest number of responses, followed by 7, 7, 9, and 4 respectively for the rest of the semester. Although he had considerably less responses than Sandy, Bert's fell into all the sub— categories except Flavor, just as Sandy’s responses did. Since he was the least accomplished of the three writers and the most representative of the average student in terms of effort and quality of work, it would have been surprising to see Bert address Flavor as a concern. 87 «LU»- > .1. r :. \/_ .r f r .5. Table 5 Bert's Responses to Peers mu a. 5 3H. ... .r.» m. r. 2 5C.» .5 0C. . \ kryflma .alVl 2L00A~ A: v.02: CQVU- 88 .90 ..00c30. 0.. 20.0... 030 a. $8. 2890 0.. .20. .0: 00 “00250.00 008.000 0E0... .9: 6000.00 .0553 .0 «0.50:0. 05 00.0800 .08.“. 050.03 60:00:00.0 .800. 0.3.0.. 2000.00 0800.0. .0 ”Ear 0c. 0.....0E00 ..c0EE00 .0055 ..cmEEoo 02.80 50.0000 .0 .0 .0 5.00080 .0086 .0 .....0>0.. 05 00.0600 060.: 08028 08.20 .0.0>> “£00300; >00. IlmlN 2 __ 8 8 __.Il._mo_ _ am a. E ... N. ...N m. .... —_ v . m = o o a F. N N N N m . . 2 N N. N m _I o .r o =1 dilll __ o o ... _. __ 0 _. __ __ a r __ _. 0 _ o o o =3 _ . 2 N N a = =2 . . 2 N N 0 = __ n. . __ for .8 N N o ... . . _TN N . . .._.m m. m o .7 a m m o __ ... _ . __ n. __8 N N __..N N . . 0 8N . . 8.0 o m 8.. 8 fit. N N =80 N N . NN N . _ __ 0 0N . . NN N . . ..N N N ...N N . . 8 o N v n. .mN . . 8 m .V . =8 0 v 0N N . . .0. m m 0 m. m. m. . m. m. =0. 0. ... cm. .0 03—7 0. ... cm. :0 “i=0. 0. ... so. :0 03—7. 0. 5 c9. :0 03—7. 0. 5 cm... :0 .03 > .0000 >_ .0000 ... .0000 __ .0000- _ .0000- run m 0309 8000 3 8800mm;— 89 Responses to Paper I: In the "qualities of writing" category, Bert's responses were similar to Sandy’s as far as distribution between upper and lower levels. The Ideas sub—category received 56% of his responses with Organization receiving 31% for a total of 87% at the upper level, leaving Wording with only 13%. Within the ”levels of change" category, Bert’s responses total 75% at Sentence level, and 13% at Theme level for an upper level total of 88% leaving the rest at the Word level. Although his upper level scores in both categories are not quite near the level Sandy reached, it is clear that, in responding, he could attend to more than the word level concerns that Sommers (1980) noted in beginning writers. The "forms of response" Bert used were 19% Questions, 38% Praise, and 44% Critical Comment. The Comments were split between word level concerns at 13% and advice about sentence organization, e.g., “Maybe these sentences could be put together.“ No questions or positive comments were directed at word level considerations. Instead, the act of questioning brought forth responses such as, “What ended up happening?" and “Did anyone get in trouble?" Like Sandy, Bert's positive comments tended toward the general, but did not descend to wording concerns: "Good description" and “Good setting" were typical. Bert’s responses to this first paper are somewhat impressive for a freshman composition student of average VA “A. ~Y .‘v- 9O ability. He probably focused less on word concerns than he might have without the class lesson on forms of response and qualities of writing. However, as his sentence level responses in the organization category indicate, the lesson was not a miracle cure for beginning writers’ difficulty in attending to higher level concerns. Nevertheless, through questioning and positive comments, Bert definitely thought about writing beyond the word level. It will be interesting to see whether he, like Sandy, will continue at this level of thinking about writing throughout the semester with a little movement and a hint of progress, or demonstrate a definite movement up the hierarchy of writing concerns. Response to Paper II: For the second paper, Bert’s "qualities of writing" responses were within three percentage points or less of his responses in the first paper: the Ideas total was 57%, Organization was 29%, and Wording was 14%. Like Sandy, Bert kept his Wording at around the same level as in the first paper. Also like Sandy, Bert’s emphasis in the “levels of change" category remained in upper levels with Theme actually doubling to 29%. In Organization, he limited himself to just one comment about sentence arrangement: “Could this sentence be made 2?" represented only a 14% emphasis on this element of writing. Along with the greater emphasis on Theme, Bert’s percentage of questions asked also increased, to 58%, as he wvl‘ .u a. ‘v‘ uny .0 A (I) .l\ K). 91 asked about Ideas and Organization issues. As did Sandy’s in the second paper, Bert’s critical comments decreased, dropping from 44% to 14%. As would be expected, the comment was word related: “Maybe used too often.“ Unlike Sandy who improved her positive comments, Bert's percentage of Praise given remained about the same as did the somewhat general nature of his comments, eg., "Good“ written next to a section of good detailed description. Bert’s responses demonstrated what was also observed in Sandy’s responses--a correlation between the form of the response and the levels addressed. As would seem likely, questions address higher order concerns, and criticism seeks a lower level. With this set of responses, Bert generally maintained his first paper response performance, and showed a promising gain in percentage of questions asked and a doubling of attention to the theme level of change. In terms of overall progress, Sandy and Bert seem fairly similar with slight movements up some of the category hierarchies. Responses to Paper III; In this third paper, Bert increased his percentage of responses in the Idea category considerably, from 57% in the previous paper to 86%. None of his responses focused on *word level concerns, and in Organization his positive (mamment was at the Theme level: "Good organization of tests." For "levels of change" Bert addressed none but the 'Fheme level which correlated exactly with the "qualities of 92 writing" score of 100% for the combined Idea and Organization categories. Bert has outpaced Sandy’s 30% lower scores in those categories, but in pure numbers, he has produced less than half of the responses she did. Nevertheless, his positive comments and questions, while not large in number, have become increasingly more specific and thus more helpful. He gave no critical comments, instead increasing his Praise score to 42%. His percentage of questions, 57%, was comparable to his performance in the previous paper, with the specificity increasingly more helpful, e.g., ”What about the other 2 tests?" For the second time in a row, Bert has had a higher percentage in the Question category than has Sandy. It seems likely there is a connection between his asking more questions and out— pacing Sandy in his attention to higher levels of concern in the "qualities" and "levels" categories. Responses to Paper IV: Bert's responses for the fourth paper assignment all fell within the ”qualities of writing" category, a performance resembling Sandy’s where all responses were in the upper levels. If, as speculated earlier, Sandy had a quarter of her comments in Organization because of the particular paper assignment, then Bert's lack of attention to this kind of organization response could reflect his status as a less sophisticated writer than Sandy. Within the ”levels of change" category, Bert focused on Theme 93 related responses 66% of the time with the rest of the focus on Sentence level issues thus approximating Sandy’s attention to those same two areas. Bert’s percentage of responses in the form of questions, at 56%, was nearly identical to the preceding two papers, and the questions continued to be helpfully specific, e.g. "Could you make this more clear?” regarding a particular sentence, and "Could you give more details on how the night ended?" at the Theme level. Positive comments dropped by half, however, when Critical Comments went from zero in the previous paper to 22% of this one; yet, the quality of these comments was rather high since, in both cases, he was agreeing with peer responses regarding development and clarity. In responding to this paper, Sandy dropped back in percentage of questions, and her positive comments did not become any more specific although she was addressing issues at the higher side of the "levels of change" category. Comparatively, Bert has been a little more consistent in terms of movement upward. He has maintained his percentage of questions over the last three papers, and he has not addressed word level concerns for the last two. Responses to Paper V: Bert, unlike Sandy, addressed Organization concerns as well as Ideas in the final set of papers. Three quarters of his responses were at the sentence level and the same ones » ‘- A. .1,- , HAV‘“ ydoq- , 'r‘V"-' I 5“. "my “YA :v-v {YV‘ b. l“ w y‘.& w ' .... .. .. A... 0. .. .... A c .3 ... Q. t. a. 2. r.. I. .3 L-.. . .3 v . C. a. a... N 0 . 0 «x. A. a 94 were also in the form of questions. The Theme level response was a rather general, "Nice conclusion." The sentence level responses were specifically about order and confusion, e.g., "Could you maybe reword this?" and even a throwback to earlier times, "Could you maybe make this 2 sentences?" While, like Sandy, Bert maintained his higher level focus for the most part, it was somewhat puzzling that the actual number of his responses was only four until I saw his Writer’s Note regarding this last session of responding. He wrote that the class writing skills had increased, so "most of my comments were just simple things, no major problems." Perhaps that reason coupled with the kind of writing assignment that it was (the opinion piece based on a semester of reading and responding to opinion pieces) is enough of an explanation. Analysis of Revisions By Responder/Writer Bert's semester total of revisions was 55, 25 less than Sandy, and 8 less than Mickey. In order of papers assigned, the number of revisions Bert made were, 15, 15, 7, 16, and 2. These revisions are categorized within the four main categories of "qualities of writing," “levels of change," "revision operations," and "sources of revision." In his revisions, as in his responses, Bert addressed concerns in all areas but that of Flavor, a sub-category unlikely to be used by less sophisticated writers. 95 Table 6 Bert’s Revisions by Responder/Writer 96 to: ..0..0 ..00::0. 0. 20.0... 030 0 $8. 2.00.8 0. .0.0. .0: 00 A300.3900 00.6.9.0 080m .0:. 5000.00 .05...) .0 ”0.5000. 0:. 00.9:00 .0>0.0 050.03 60:00:00.0 ....000. ”0.3.0.. 5000.00 0:30.000 .0:00.>0.. 0:. 00.8.00 :0..0_0o 00.5...QO .:0...00< .0:..00000 no.m.<~. 002000. 3:003. 0. 0:0 .0: o. 5.030. 800.: .5 .. .8030. .0 00.39. ”.5 2000.00 .00:0:0 .0 0.00.. 0:. 00.0800 050.: .00:0.:0m .0820 0.02. ”c..:0».:Q03 >3. 8 a 03 ... .. . m 0 HQ 0.... .0._._><..00=0%0~— .3 20.35% 97 Revisions on Paper I: The majority of Bert’s revisions were at the Wording level, with 13% in Ideas. The emphasis on Wording is not surprising since Bert seems to represent the average freshman composition student. Not surprising, either, is his focus in the "levels of change" category" on Words 60% of the time and Phrase changes at 33% with the remaining 17% at: the Theme level. The lower level "revision operations" of deletion and substitution constitute the bulk of Bert's revisions. With the Substitution score at 67% and Deletion at 13%, the inescapable conclusion is that Bert definitely represents a a beginning writer. A review of the “source of responses" in Bert’s first set of revisions looks encouraging in terms of his ownership of the writing, because over half of the revisions have been on his own initiative rather than because of responses he has received. Yet, when a comparison is drawn between this performance and the levels of the revisions he addressed, it Shows the small number of revisions at the Idea level were due only to response received from peers, while the large number of word-related revisions were his idea. This, then, Serves as one more confirmation of Bert’s status as a beginning writer, particularly in comparison with Sandy’s InOre sophisticated revision performance in this first paper. 98 Revisions on Paper II: Bert’s revisions at the Idea level remained low in this second paper, just as the bulk of his emphasis was in the ”qualities of writing“ sub-category of Wording at 93%. In like fashion, his level of change remained primarily at the Word level with 60% and the Phrase level at 33%. In the category of "revision operations,“ Bert also continues in the same fashion as the first paper, with 86% at the Substitution and Deletion levels. Additionally, his self-initiated revisions and their focus on word-level concerns remain essentially the same. Like Sandy, Bert maintained the performance established in his first attempt at revising a paper, albeit at a considerably lower level on the writing hierarchy. Revisions on Paper III: Something happened to Bert. The percentage of his revisions in the Ideas and Wording categories has reversed to 71% in Ideas and 14% in Wording, with another 14% showing up in the sub-category Organization. In the “levels of change" category, another turn-around is evident: no revisions fell into the formerly heavily used Word category, and just 28% were in Phrase revisions. With 43% and 28% respectively in the Sentence and Theme categories, it is clear Bert has moved up the writing hierarchy in this paper. This move to higher levels is echoed in the “revision operations" category as well. Where in the first two papers 99 Bert relied heavily on Substitution and Deletion, now only 14% of his changes are in Substitution. The majority are in Addition, at 71%, with the other 14% in Reordering. When his “sources of revision" are considered, it might seem reasonable to suspect that this upward move would have been due to responses received from more knowledgeable peers. It is all the more surprising, then, to note that four out of the five revisions at the Idea level were on his own initiative, and the one Phrase change in Wording was due to response received. Compared to Sandy, who virtually showed no movement in this paper, Bert was a steamroller. The third paper, a survey, had the potential for being fairly well developed in pre-writing planning and thus requiring less substantive revisions. With Bert, that may not have been the case, he simply may not have developed his paper as much as he eventually realized was necessary after reading other papers. Revisions on Paper IV: Unlike Sandy, whose growth spurt in the fourth paper was not extended to the fifth, Bert, for the most part, maintained the levels he achieved in the previous paper. His Idea development score dropped to 19%, but that was offset by his 56% score in Organization. The Wording score of 25% did result in a small Word focus where none had existed the previous paper. The "levels of change“ category r1 100 continued to hold more than two-thirds of the changes in the upper levels with Theme rising to 63% over its previous 28% allocation. In "revision operations," Addition accounted for 69% of the total changes in the paper, and, since all of the additions were in the Idea and Organization categories, this was, again, a sign that Bert was maintaining the growth seen in the third paper. However, in the case of "sources of response," Bert lagged a bit behind his previous score, perhaps because in actual numbers his total revisions on this paper were only seven. Where this paper was, for Sandy, a considerable movement upward, for Bert it was a matter of maintaining the leap forward he’d made in the third paper and even moving a bit more up the writing hierarchy in most of the categories. Revisions on Paper V: In this final paper Bert has little in common with Sandy. One exception is that both completed only a small number of revisions, Bert with two to Sandy’s five. In contrast with Sandy’s downward movement, Bert's final revisions were confined to the area of Ideas, thus maintaining, if not moving upward. The "levels of change“ he used were Sentence and Theme, and the “revision operation" of Addition was used for both. The other element that Bert had in common with Sandy was that the "source of revision“ was response received rather than from Bert’s own . v- 101 initiative. The possible reasons for the limited revisions and the lack of revision initiative have been discussed in previous sections. Briefly, as in Sandy’s case, Bert may have thought about the opinion he was expressing for a long enough period so that he had, like an experienced writer, done much of the revision in his head. Additionally, as cited earlier, Bert’s thought that his peers’ papers were really done well probably applied to his own as well because it was among his best work. Despite the limited revisions and lack of self-initiated revisions, Bert has not dropped back into word-level considerations but has maintained the progress upward that began with the third paper. Unlike Sandy, once he made the move up the writing hierarchy, Bert never took any appreciable step backward, rather he maintained and even made gradual adjustments upward. THIRD CASE STUDY: MICKEY Mickey was the best writer among the three students in this case study; her writing was in the high B to A range, and both her journal entries and Writers' Notes were always done well. In responding to peer writing, she produced the most responses of the three. However, in making revisions, Mickey was midway between Sandy and Bert with a total of 63. Considering that more experienced writers have been shown to make less revisions than less experienced writers, it is not surprising that Mickey’s total places her between Sandy and 102 Bert. Analysis of Written Responses to Peers Mickey’s responses to peer writing over the semester reached a total of 92, nine more than Sandy and nearly twice as many as Bert. Like the other two, Mickey’s first set of responses were more than for any other paper. Her subsequent responses to peers totaled 17, 23, 15, and 11, respectively. Mickey’s greater skill as a writer would seem to account for her responses numbering more than either of the others, and it should also explain why she addressed issues in every sub-category including Flavor, an area unaddressed by Sandy and Bert. 103 Table 7 Mickey’s Responses to Peers 104 ..0..0 ..00::0. 0. 20.0E 030 a. $8. 2.000 0. .0.0. .0: 00 «000200.00 000.020 060. .0:. 2000.00 .0:...5. .0 ”0....030. 0:. 0.....0E00 .0>0_0 .0:.0.0>> .:0..0N.:00.O .0000. 0.30.. 2000.00 00:008. .0 250.. 0:. 00.02.00 .:0:.:.00 .00....0 ..:0:.Eoo 02.....00 .:0..m0:0 “O .0 .0 2000.00 .00:0:0 .0 000.. 0:. 0.....0E00 020:. .00:0.:0m .08.:0 0.03 .:..:0¢..:0.03 >00. _. 8 alum-Iona 8 .N 2 ... m. 8 8 .v 8 .m a ......fl __ .. 0 m . .= 2 m .. N .= «N 2 . t o. . ON 3 m N N T... . o ..o . . =0 N . . .3 . . .To .. _. .. ._ 0 ._ o . . ..o N . . ..o . . L* 0 . _. __ .0 7.. . . 2 N N .. o o. . ..2 .. N N...» ..o . . ._ ... N . . 0 .. o . . .. .... . . ..0 . ac . . .. o . . .r. . ..0 ..N m . N .2. . ... ... .__o N . . .W . . .3 v v _0 .. ..o . . ... N N 0 .0. N N =2 N N ..o N . . __ ._ ... .o . . LE .. N . ... .E . . ... N N .0 .3 . m . . ..mm m .. . ..N. o. ... m ..Nm ... o. v ..8 2 ... v .. .3 . . ..o . . _. ..N. N N __.V . . ..0 ..8 c v ..o . . _8 2 o. m =8 .. . v 82 o v .0 .2 N . . EN N N N o v N ..o . . 0W. v v 0 I. I. I. I. I. ... m. ... c0. :0 03—7. W. ... :0. c0 0% m. 5 :0. :0 fies—T. m. 5 cm... :0 03 .... m. c. :0... :0 03 - >.0000 >_.0000- . ... .0000 __ .0000 ..0000 .30.: .. 0.00... 0.000 0. 0025003— 105 Responses to Paper I: Like Sandy and Bert, Mickey’s first set of responses were heaviest at the Ideas level in the “qualities of writing" category. Her responses at that level reached 69% of her total, with 15% in Organization, and the rest in Wording. In the "levels of change" category, Mickey’s attention to Theme level concerns, at 53%, was somewhat less than Sandy’s, but considerably more than Bert’s. A third of her responses were at the Sentence level with the remaining 16% divided between Phrase and Word indicating that her concerns were in the upper half of the hierarchy for the most part. Mickey’s "forms of response“ found her with more in Questions and less in Critical Comments than Sandy and Bert, a result that probably reflects her greater experience as a writer. Her entries in the Critical Comment category tend not to be direct criticism but rather indirect suggestions accompanied by explanations, e.g., “These two words are a little confusing. Maybe you could combine them ("less smart"). The specificity of her Praise comments was clearly superior to those produced by Bert and Sandy. Most positive comments contained reasons, as seen in these two examples: "Very nice, I can §§§_what happened” and "Very suspenseful story." Just as Bert’s overall spread of responses was not surprising for an average writer, neither are Mickey’s response results. In contrast with Sandy’s performance, 106 Mickey differs primarily in the spread of the "forms of response" totals and in her higher level of specific responses. Considering she is the better writer of the three, it would seem logical to expect to see Mickey respond to other writing by focusing on upper level concerns, asking pertinent questions, and giving specific positive as well as critical comments. A question worth pursuing is, will such a writer benefit from doing more of what she already does well? Responses to Paper II: Unlike either of the others, Mickey’s focus on Ideas in the “qualities of writing" category increased appreciably, from 69% to 82%. Organization contained 6% of the responses, as did Wording and Flavor. In the category of ”levels of change," Mickey’s Theme score raised a few points to 59% with the rest in Sentence and none in the lower half of the "change" hierarchy. This upper level score is very similar to that of Bert and Sandy, while the use of “forms of response" tends to differ. Where the others increased their questions and decreased their critical comments, Mickey instead increased her amount of praise by nearly fifteen points to 70%. Since it is such specific praise, the loss in the Question column does not seem detrimental as far as the loss of specificity that could have occurred. Her praise often teaches as much as it pleases, e.g., "Good way to show how you felt" and "Nice way to move into other’s 107 View point." At this point in the analysis of Sandy and Bert’s responses, I was able to observe a correlation between the form of response and the levels addressed. In Mickey’s case, she has just two responses in the sub-category of Critical Comments, but both are in the Theme column of Ideas, and both are phrased as indirect suggestions, e.g., "Maybe add her reactions." It may be that since the more skilled writer is less likely to address word level concerns, the correlation between critical comments and word level is appr0priate only for the less experienced writer. Responses to Paper III: With the third paper, Mickey maintained at 82% the gain in Idea responses that occurred in the second paper. The rest of her comments are distributed equally between Organization and Flavor. Again, as in the previous paper, all of her comments fall into the Sentence and Theme columns, this time with Theme taking 40% more of her responses. Questions have slightly increased to the level of the first paper, and, with no Critical Comments, Praise is at a high of 74%. That so much of her response occurred in the form of praise rather than questions would be more troublesome if the praise weren’t so specific, e.g., "Good explanation of test“ and "Good transition." At this point, Mickey’s results resembled Bert's with a 108 major focus on the upper levels of the "qualities" and "changes" categories, rather than Sandy's where a third of her responses resided in the Wording section. Responses to Paper IV: Mickey, like Sandy, had the majority of her responses in the upper levels of Idea and Organization. Because of the organization possibilities of this assignment, it seems logical that the two more sophisticated writers would address that issue as they responded to peer papers. Mickey also had a 13% Wording response and 6% in Flavor, a category never addressed by Bert or Sandy. That this is the third time in a row Mickey had given thought to the author’s voice is another indicator of her higher level of sophistication as a writer, although it should be noted that she never used the descriptive terms of "flavor“ or "voice" in her response. As would be expected, Mickey’s responses are mostly in the upper half of "levels of change," a score shared by the others. In "forms of response," Mickey had a turn—around; her previous emphasis on Praise has exchanged places with the Questions column. This time, Mickey has over half of her responses in Questions and a third in the Praise column. The responses in Praise vary from her earlier performance in that some are not as detailed about what she appreciates as the earlier ones had been. However, her questions continued to invite necessary improvements, e.g., “Could you add a 109 transition here to connect the two sentences?" In comparison with Sandy and Bert, Mickey doubled her Questions while Sandy dropped back a little, and Bert continued along with over half of his responses in the form of questions. A review of the progress of the three responders through this fourth paper shows Mickey to be performing at about the same level overall while Bert and Sandy have both moved up the writing hierarchy in terms of less word level concerns. All three have improved in percentage of questions asked, an achievement that could help to account for the move up the writing hierarchy in two of the three. Responses to Paper V: In this final paper, Mickey continued to perform in much the same fashion she had since the first paper. Over 90% of her attention was focused in Ideas and Organization, and she dealt with one problem in the Wording category. The word concern was handled with a Comment: "You may not need this word." This gently suggestive remark mirrored word level comments in the first paper set of responses, both in tenor and in low percentages. After the upsurge in questions in the previous paper, Mickey returned to the two— to—one Praise and Question ratio that appeared in all of her other sets of responses. Although Sandy and Bert did not do at all badly from the first time they responded, they were never at the helpful level that Mickey could achieve. However, they did show some movement up the writing hierarchy over the semester whereas Mickey primarily 110 maintained the skill she displayed at the beginning. Analysis of Revisions By Responder/Writer Mickey’s total revisions over the semester reached a total of 63, just eight more than Bert undertook and seventeen less than Sandy handled. The number of revisions per paper, in chronological order, are 11, 4, 12, 25, and 11. The four main categories into which these revisions are sorted are "qualities of writing," "levels of change," "revision operations," and "sources of revision." Despite having made responses to peers that could be placed in the Flavor column, Mickey addressed every other sub-category but Flavor as she revised her papers throughout the semester. 111 Table 8 Mickey’s Revisions by Responder/Writer M3. H “H I'M“. n 0" Alt Mn: 0 [in/u ‘q- in“ , unufh )pnm um! In! )_ “‘H “can tun v ‘v n )Al ’II -[h u; \n 'M '\ I- up nzn n tHH ‘) .J, )1 "1(9) r). ~(‘DI‘FNLI 1,." u ‘ 0' '1-(lll A5; u "1:: « r. l‘ "rm; .ru l c." A): \lL-p‘t v "'D _l‘l "HIM ’1 ) ‘1! 101M.) “ 1:: w ’l )| n. .4».~1 (1‘. ‘1'?! n nut u )\4 In “M 112 ”010000 .... 5000000015180 HmUH m m . £0 00 .0 £0 0 > m 0 ... D m x x3. . £0 .00 000...." .290. 00.90. m00.0000. .0030 0030.00 .00 ..0<0_m 0. 000000. 00.0002. .0.” .8500 0. 03000. I .0. 3003 .000...00. 9.92.0300. 00.0..00 0030.00 .00 ..0<.m.000_ 000.0200». 00.0002. ..O.<<.r .009. 0000.09.00. 5.0.0.00. m_0