KNOWING AND USING RESEARCH IN TEACHING COMPOSITION Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY RICHARD VANDEWEGHE 1977 10239 1 I III; IIIZIIIIILIIII Illlllllllll III! I will; ll ? LIBRARY Michigan State University This is to certify that the thesis entitled KNOWING AND USING RESEARCH IN TEACHING COMPOSITION presented by RICHARD P. VANDEWEGHE has been accepted towards fulfillment of the requirements for Ph.D. degree in English W A / Major professor Date July 28, 1977 0-7 639 ABSTRACT KNOHING AND USING RESEARCH IN TEACHING COMPOSITION By Richard VanDeHeghe In order to establish solid foundations for composition pro- grams, teachers and directors of composition programs should know about the research done in the field of composition and understand how it can be integrated with relevant theory into a comprehensive ration- ale for the design of composition programs. Three central arguments are advanced in this dissertation. First, composition teachers and directors of composition programs should know how to read research reports and should know what research has been done in composition. Second, in designing and carrying out composition programs, teachers and directors should use the insights available to them from research. Third, teachers and directors should integrate relevant theory and research in order to build composition programs soundly informed by the most significant and relevant information available. Chapter I serves as a guide to reading research reports. Three essential elements of research design are examined in it--g§§igg_ validity, test validity, and measurement reliability, These three elements of design are frequently mentioned in the literature on research in composition, but are never defined in it. gut-p. ,: A :4 :i 3M: 4» i: ,: :v'QI JSIN seat: in t 2"". ~ milieu" Richard VanDeHeghe Chapter II is a compilation of research done since l963 in twenty-four areas of needed research prOposed by Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer in Research jn_written Composi- .§jgg_(NCTE: l963). Both published and unpublished research reports are examined, and recommendations for further research are made. Chapter III is an examination of the arguments advanced against using research in composition as well as those advanced in favor of using.it. A theoretical rationale for using the results of research in the design of composition programs is developed. Chapter IV is an investigation of theory which bears directly on the teaching and learning of composition. Insights derived from learning theory, language theory, and composition theory are integrated with insights derived from research in composition to formulate a representative theory of instruction for composition. Chapter V presents two major charges to the English profession. The first is that composition is an intellectually sophisticated field rich in knowledge and presenting numerous empirical, theoretical, and pedagogical challenges for researchers, teachers, and directors. Thus, composition should be recognized as an academic discipline in its own right. The second charge is that composition teachers and directors at all levels should receive Specific training in research, theory, and pedagogy in composition and in related fields. KNOHING AND USING RESEARCH IN TEACHING COMPOSITION By Richard VanDeNeghe A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of English 1977 0‘- t3 -1) IQ"... Copyright by Richard VanDeNeghe 1977 TO Judy, for your interest, suggestions, encouragement, and patience. ii Iuis'r ,::‘a::e and s I alsc .. figsu'arth flflsitior Terr 3 H OERV‘E.’ ACKNOWLEDGMENTS I wish to thank Professor Clinton S. Burhans, Jr., for his guidance and suggestions throughout the preparation of this disserta- tion. I also wish to thank Professors Jay B. Ludwig and Alan M. Hollingsworth for their encouragement and assistance during my study of composition. I am also grateful to Professors Russel B. Nye and Richard Benvenuto for their support throughout my graduate program. ’m-nn 'I I "Ir F J. A IIILIJI. . \g'AH Inter I. CGIICLUS. I‘ll u z: . i’: ”I!" n“ “r“ JJI'N riY o TABLE OF CONTENTS INTRODUCTION . CHAPTER I. DESIGN VALIDITY, TEST VALIDITY, MEASUREMENT RELIABILITY . . II. RESEARCH IN COMPOSITION . III. USING RESEARCH IN COMPOSITION: RATIONALE . IV. A THEORY OF INSTRUCTION FOR COMPOSITION: THEORY AND RESEARCH INTEGRATED . V. CONCLUSION. . . . NOTES BIBLIOGRAPHY iv Page 46 98 116 183 187 207 Ry ti: guns the st "1. Star‘i i3 hih insi (7:. 7r + *3. rese " .CGE'ierS 'I tsearch INTRODUCTION My title--Knowing_and Using Research jg_Teaching,Composition-- governs the structure of this book and the central arguments advanced in it. Starting with the premise that research in composition Offers many rich insights into the teaching and learning of composition at all levels, I propose that composition teachers and directors of com- position programs should know how to read research reports and should know what research has been done in composition. Second, I suggest that teachers and directors should use the insights available to them from research in designing composition programs. Third, I argue that teachers and directors should integrate research findings with rele- vant theory in order to build composition programs which are soundly informed by the most significant and relevant information available. Though much of what I discuss in these chapters may be of interest to research specialists, I am writing primarily for the benefit of research novices, composition teachers, and directors of composition programs, particularly those of the latter two groups who have ignored--for whatever reasons--research in composition. I want to help the reader who does not have the technical expertise of the research specialist to understand research reports and to make informed judgements about the contributions research can make to the design of composition programs. Other s‘zi'ir goals it'ar: Llayd‘ 'y-I-(n'61ian (Jan. DIJ ’ search desii 2's: 3 detail Other writers dealing with research in composition have sought similar goals. The most notable of these writers are Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer, whose Research ig_written Composition,1 published in 1963, provided both an examination of research design and a summary of research throughout this century and also a detailed examination of five research studies. Henry Meckel's chapter in the T963 Handbook gf_Research 1g_Teaching2 also summarized research up to that time. In 1969, J. Stephen Sherwin published 5935. 3 which included Problems in Teaching English: A Critigue gj_Research, a long chapter on investigations into the relationship between writing and traditional gramnar, linguistics, and writing practice. Two other research reports appeared in the sixties and early seven- ties. Richard Braddock's chapter in the l969 Encyclopedia 2f Educational Research4 reviewed research throughout the sixties, and Nathan 5. Blount's chapter in the l973 Second Handbook gf_Research in Teachings examined research in the late sixties and early seventies. Finally, Sara H. Lundsteen's Help for the Teacher of Written Composi- 6 .tigg, published in l976, reviewed research and offered directions for the classroom which derived from the research perspective of the mid-1970's. I intend my book to be read as a companion to these other books and articles on research in composition. Mine differs, however, in four respects. First, in Chapter I, I give considerable attention to essential elements in research design. 0f the books and articles I cited above, only Research ig_Hritten Composition attempts to guide the inexperienced reader by providing some commentary on how to read and evaluate research. I extend that book's guide to research design 12;. tile 0:? 5U \ Pry-fin a: (if SANA-Iii fin-I. Aguirre knob LN. -":r have sig I C. .3 (D ‘1 (D m by closely examining three key elements frequently mentioned in it and in the other summaries but never defined--namely, design validity, test validity, and measurement reliability. Second, in Chapter II, I examine research since 1963 in the twenty-four areas of needed research proposed in Research jg_written _§Omposition. To my knowledge, no one has yet compiled the research which specifically relates to these crucial proposals. In this regard, I examine both published and unpublished research reports, many of which have significant implications for the teaching and learning of composition. I also make recommendations for additional needed research. Third, in Chapter III, I develop a theoretical rationale for using the results of research in the design of composition programs. I examine both the arguments advanced against using research as well as those in favor of it. Finally, in my fourth chapter, I examine theory in fields which bear directly on the design of composition programs. Here, I integrate insights derived from learning theory, language theory, and composition theory with insights derived from research in composition to formulate a theory of instruction for the teaching and learning of composition. My intention here is to present a model of how theory and research can be integrated in order to build the soundest, most informed composi-' tion programs. The reader will notice that I do not direct my remarks to any particular level of instruction--elementary, secondary, or college. I have maintained a comprehensive perspective because the insights I derive from both research and theory apply to many instructional ‘ .EIE‘S. Fifi-Ir .,3.3;;ion ar- The re new?“ I 35% and in viz—:ge and magical ci‘. Itcsition pr tiettaining o 2335 what it ran: state I which n 1'53“. Ieacine rs: to acni eve levels. Furthermore, such a perspective will, I hope, promote greater interaction and exchange among all levels of instruction. The reader will notice, in addition, two themes which form an undercurrent throughout this book. The first is that composition is a serious and intellectually saphisticated field, one that is rich in knowledge and one which presents numerous empirical, theoretical, and pedagogical challenges for researchers, teachers, and directors of composition programs. The second theme follows from the first--that ‘the training of composition teachers at all levels must extend far beyond what it has been in the past and still is today. Given the current state of knowledge about composition, we can no longer ignore the valuable resources that are available to us in research and in theory. Teachers and directors must become familiar with both if they wish to achieve intellectually defensible and more successful programs. ivy: and w iiiilh and t I: j'ewf ”g f CHAPTER I DESIGN VALIDITY, TEST VALIDITY, MEASUREMENT RELIABILITY In 1961, the Executive Committee of the National Council of Teachers of English appointed a committee to investigate the state Of knowledge about composition. Its stated purpose was "'to review what is known and what is not known about the teaching and learning of com- position and the conditions under which it is taught, for the purpose of preparing for publication a special scientifically based report on what is known in this area.”1 Two years later, NCTE published Research jg_Hritten Composition, a monograph prepared by the committee and written by its directors, Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer.* This monograph was the first of a number of research reports published throughout the sixties and early seventies which provided summaries of research in composition. But the Braddock Report was unique in that, along with reporting on research studies, it also examined the critical'tOOls necessary for careful scrutiny of such studies. The Braddock Report stands, then, as an important and valuable document for composition teachers and for researchers in written composition. I find, however, that it has certain *Hereafter referred to as the Braddock Report. ssz'ICT ' "’5 ' {5:795 “W Meir?“ 3;: we REX 4:3“ NSC-J 0'0,’ I .an 1R9 as I." "i ' “5"“;th ?;:‘?I‘ ‘.'.|wde 0 f'ESIItS. r‘ . :g as meaS 32:3, such as I» *i ?:r I- III ti ng «:znns for II In theII .. w ‘Jd -~y sased" I I‘é‘: ' «Sign, EXEC I‘ll discuss 1 cm to IHEI shortcomings which I will address myself to in this chapter. In order to do so, I will briefly review the substance of the Braddock Report. The Report begins with a brief description of its preparation. chapter 11 surveys the methodology and some Of the elements of design in composition research. The first section in this chapter examines vari ables which affect the rating of conpositions--the writer variable, the assignment variable, the rater variable, and the inter-rater vari- abl e- The next section deals with the design of research studies based on frequency counts. In a final section, "General Considera— ti ons ," the authors make suggestions for the critical interpretation and evaluation of research reports by discussing, among other things, the attitude of the investigator, planning of procedures, and reporting 01’ results. Chapter III sunmarizes a considerable amount of research in comos‘ition done in this century. Here the authors list research StUdi es under five general headings: l) environmental factors influ- encing composition; 2) instructional factors influencing composition; 3) rhetorical considerations; 4) objective tests versus actual "pi ti "9 as measures of writing performance; and 5) other considera- fi Ons, such as size of English classes, writing vocabulary, spelling, and handwriting. At the end of this chapter appear twenty-four questions for further research in composition.* In their fourth chapter, the authors select five of the "most 50““d1y based" research studies and examine in considerable detail the design, execution, and results of each study. \_ * . A will discuss these questions, and the research that has been done ‘“ regard to them, in my next chapter. For t :v:;'135, IIIE “'5: ti" 5 am, it P' a.:a"nat‘3n c 23525. Th‘i res: 3.‘ or I; sssitic e: in Iara search that ITSII'J'CIITIQ I rations fc: 'Iv'estigated ‘ I} mgarde Temple, i 133%: the Brad are “classic Ihile . II III-orient For teachers, for researchers, and for directors of composition programs, the Braddock Report is an important document. It is the Ft rst major sunlnary of research to appear in over thirty years.2 Second, it provides for the research novice a useful, though brief, exami nation of the basic tools for a critical scrutiny of research studies. Third, it heightens the professional and pedagogical aware- ness of composition teachers by demonstrating that empirical research in composition exists; by discussing elements of design and measure- ment in layman's terms; and by providing information on composition research that teachers and curriculum designers can implement when constructing writing programs. Fourth, it presents twenty-four recom- mendations for needed research in composition, many of which have been investigated in recent years. Finally, it has become a document I” ghly regarded by subsequent research specialists. Nathan S. Blount, for example, in the Second Handbook 9: Research 13 Teaching (l973), Cal ls the Braddock Report an "indispensable source of information,"3 and a "classic monograph."4 Hhile Blount's praise constitutes continuing recognition of this important document in composition research, it raises an essen- tial question: for whom is the Braddock Report an indispensable sour‘ce of information? In the preceding paragraph I suggest some possibilities. Part of the readership for the Braddock Report con- sists of researchers and teachers with research background and tec“nical training, who no doubt benefit most from the sumnaries Of research and the twenty-four recomnendations in Chapter III. A Se60nd and probably much larger audience consists of research novices a“(I teachers who lack research experience and technical expertise. awe close yrs, I see I hr the bacx; hatch; rese EI‘E', "while ti ssf'yl in‘onra The fi I‘ESECth cha - in cor; Exiting cop; “sign. The a; ‘55 "tymory I Vitend, the re '75" affect re \1 Speci $33:ng by DC “I Wk, EX I’Iii WIS fII‘St The sec 1; n addition to the research sumnaries and recomendations, these readers need the survey of research methods and design in Chapter II and the close examination Of the five studies in Chapter IV. In other no rds , I see the authors attempting to provide this second audience with the background material in research that will help them when exami ning research studies of the sort that appear in the book. ever, How- while this background material does provide much necessary and useful information, I have found it limited in three ways. The first limitation concerns research design. The purpose of the second chapter is to "survey some of the methods and elements Of design in composition research" (p. 6) by pointing out the variables affecting composition rating, frequency counts, and other elements of design- The authors remark that this introduction to research design IS a " cursory review," (p. 26), with which most readers would agree. I ‘i "tend, therefore, to reformulate and elaborate on the variables "hi Ch affect research design by presenting the concept of design W, specifically, internal validity and external validity as deVe‘ oped by Donald T. Campbell and Julian C. Stanley in their monu— mental work, Experimental and Quasi-Experimental Designs for Research, which was first published the same year as the Braddock Report.5 The second limitation concerns testing, or more particularly, $1; Validity. This concept appears often in the Braddock Report, but it is never defined. In Chapter II, for example, we find such “Eferences to test validity as the following: ‘Too few investigators conduct pilot experiments and validate their instruments before undertaking an investigation. (p. 5) he Wt" rel 0th (;. 18) ’Ieafihors E ME W“ are I” it. 333‘ AIS reS~ It": the rear ”.75 I“ r9555 “.5 Ill‘lid:* ‘N 5W; act“ The t y- ' l-Eiiire IE” t-‘I. . ,I 15 fl“ .. m ‘I “Serra .I OR ‘;I EIWIE’ I yiseran I“ or raterf 1+ exafill Cast founc nliahle I SIRTIOI t .r ante of S“ :73," III CR5 wiihg the BUI IL'tIs' 5 MW 35', in analyz :‘I’yE-Ij IIIS nee The worth of such instruments becomes better known, of course, \zhen other investigators attempt to validate the instruments. p. TB The authors again cite the importance of test validity in their review of the Harris study: "The reader's attention is also directed to the manner in which Harris tried out his procedures in a three-month pilot experiment and validated his criteria of measurement before he under- took his regular experiment" (p. 70). wi th For the reader who is unfamiliar the meaning of test validity, such references to its essential r01 e in research are vague and confusing. I will, therefore, examine test validity later in this chapter and detail its importance in assuri ng accurate measurement. The third limitation I wish to consider in the Braddock Report IS Easurement reliability. Like test validity, measurement reli- W is frequently brought up in the Report as an essential co"ISIideration, but it is never defined or clarified. In Chapter II, for example, the authors remark: Ni seman has frequently reported reliabilities in the lower .90's for raters using the general impression method for the English 1 1+ examination. Cast found the general impression and analytic methods more reliable than the other two and the analytic method slightly SUperior to the general impression method. (p. 13) In spite of such references, however, no mention is made of what W1 means. Similarly, in the "Explanation of Statistical Terms“. in Chapter IV, reliability is not explained, though in dis- CI‘SSing the Buxton study, the authors give considerable attention to Buxton's method of obtaining reliability in rating the essays. More- °V9T. in analyzing the Harris study, they point out that Harris refined his measuring instrument so that "all criteria except a and g eseartn my @122 3f the ":iiér of res "5571‘ whetr 31.375" (p. In :3 neip W as thi 1?: cannot b, 3'? 2mm, “'3 ih‘Iest‘ IO reached a good or high level of reliability" (p. 75). I find this fai l ure to define reliability a shortcoming in the Report, and so I intend to examine this concept, which for now may be defined as Egg- si stengy i_I1 measurement. I wish to stress that I am writing for the benefit of the research novice or the teacher who does not have the technical know- ledge of the research specialist. I want to assist the inexperienced reader of research reports and summaries of research "to determine for himsel f whether or not to be convinced by the conclusions Of the studies" (p. 55). By thus extending the Braddock Report, moreover, I wish to help the reader make critical sense of such statements in the Rfiport as this: "if raters are not evaluating for the same qualities, they cannot be expected to rate with validity or reliability" (p. 12). Such comnents appear throughout the Report, as well as in other pub- ”Shed investigations of research. Hith this projected audience in mind, then, I will examine Tl rst, desigg validity; second, test validity; and third, measurement W. In order to illustrate these three concepts, I will draw 0“ the "most soundly based" research studies found in Chapter IV of the Braddock Report. I will rely on these extended summaries rather than on primary materials for three reasons. First, I intend this book to be read as a comparison to other books and articles which deal with research in composition: the Braddock Report; Henry C. Meckel's chapter in the 1963 Handbook gt Research 9_I_i_ Teachingés ‘1- Stephen Sherwin's Four Problems ill Teachflg English: A Critigue 91 Research7 Emma Research ; Richard Braddock‘s chapter in the l969 Encyclgpedia fl 8; Nathan S. Blount's chapter in the l973 Second “earth stac" ”any, I HIS Iats'ials wit? ":SEE'L'". saeci Illifsnm E teal researc 33% tale a '53 asks, "I "trig Specif iSIS' "I0 What 7533 r:— d buent Va ”PHI . s ‘l Va \ Q {in}, ““95 (Var Rt; am EV: “Its 0f the 52:“ Ihich “doth ll Handbook 93: Research _i_n_ Teachingg; Sara III. Lundsteen's Help for the Teacher 21: Written Composition"); and other sources Of a less compre- hensi ve scope. Second, I want to avoid extended description of entire research studies, such description being beyond my purposes here. Final ly, I wish to provide examples from easily accessible materials, materials with which readers of this book may already be familiar. Design Validity ge_s_i_g_n_ validity has received its finest exposition in Donald T. Campbell's and Julian C. Stanley's Eyperimental and Qpasi-Experimental D_esigns for Research, a book which has been highly regarded by research specialists as a "classic exposition Of experimentation in Education“” and a "monumental treatise on experimentation in educa- ti ona‘l research."12 In discussing design validity, Campbell and Stanley make a fundamental distinction between internal validity, “hi Ch asks, "Did in fact the experimental treatments make a difference in this specific experimental instance?"; and external validity, which asks’ "To what populations, settings, treatment variables, and measurement variables can this effect be generalized?" (p. 5). We] validity, in other words, concerns the control of extraneous Var‘i ables (variables other than the treatment variable) in an experi- me“"-: and external validity concerns the generalizability of the res“ ts of the experiment to other situations. I find this an impor- tam; and useful distinction because it focuses attention on the fa(Itors which jeopardize both kinds of validity. These factors, or threats \, to the validity of research design are presented in the campbell and Stanley book in a lucid schema, and I will draw heavily In]; :AIS mate yy‘mn Glen :‘Enerinents :rianIIy ide Iie'standing :Itre one has, “asserts repor 'ESEE'CI'. inves Carats t'esyhith, " :siice effect I h. 5). 3:3 I€SIgh by 13656.58 of "“553 0f the I" he of thee iIIIISIbIe," , I"i ~f . Jeoefil‘dli 12 from this material in my discussion here. In addition, I will also draw from Glenn H. Bracht's and Gene V. Glass' "The External Validity of Experiments,"]3 a 1968 article appearing in the American Educational Research Journal which elaborates on the threats to external validity originally identified by Campbell and Stanley. I believe that by understanding the threats to a valid design--the extraneous variables on the one hand, and the generalizability on the other--the reader of research reports will become better equipped to assess the results of research investigations and their applications. Campbell and Stanley present eight classes of extraneous vari- ables which, "if not controlled in the experimental design, might Produce effects confounded with the effect of the experimental stimu- lus" (p. 5). These variables pose threats to the internal validity Of a design by postulating a plausible rival hypothesis (or hypotheses, in the case Of a number of uncontrolled variables) to account for the "3501 ts of the experiment. "Where an experimental design 'controls' for One of these factors, it merely renders this rival hypothesis imp‘l ausible," write Campbell and Stanley (p. 36). The eight factors "hi Ch jeopardize the internal validity of a design are identified by the authors as follows: 1) .Hi story: This term refers to "the Specific events occurring bet“Wt-Zen the first and second measurement in addition to the experi- mental variable" (p. 5). The key word here is "events": what haPTKened between measurements which was not intended (controlled), but Which nonetheless affected the experimental outcome? That is, I“ general, what the Braddock Report means when the authors state iI "the o o- in , I». “It'd: ti on #I‘ "I “4. "I'UI ICU “3:4”‘13 dbl l Mn»- 4": 'IUU U! (D i”:.o‘4ri‘ar+ "'Ob"i5il. a. . . ' Q I \nA i :xrerh.:.’ 1° . . ES‘IM- U y‘. 13 that if "the investigation entails the comparison of one method of -i nstruction to another, all variables other than the method should be controlled" (p. 25). 2) flaturation: This refers to "all of those biological or psychologi- cal processes which systematically vary with the passage of time, independent of specific external events" (pp. 7-8). For example, if an experiment tests for development of syntactic fluency, we might consider whether during the time lapse between measurements students have naturally grown more syntactically sophisticated, thus calling into question the full effect of the experimental variable intended to produce this result. 3) Testing: This threat pertains to "the effects Of taking a test Upon the scores of a second testing" (p. 5). For example,the occa- Si on of the pretest may generate anxiety, which may in turn affect the StUdent's performance on the test. 0n the second testing, however, the student may have become adjusted to the testing situation, thus r‘eduozzing anxiety and improving performance. Hence the pretest- p0Sttest gain may be partly accounted for by the reduction of anxiety Whig}, was generated by the pretest. As Campbell and Stanley point out, it is well known that students taking achievement and intelligence teSts for the second time usually do better than those taking the test for the first time (p. 9). 0‘) Lnstrumentation: This refers to "autonomous changes in the measur- i“9 instrument" (p. 9) which might explain a change from the first measurement to the second. The section in the Braddock Report dealing With rater variables lists a number of instrumentation threats to ie-zrai aw ‘ .I‘ - [:. .OI 94-60ap‘ rate-:55 Ir! C. I“. : Jay: J6 rise; 3e Orni Q'+ [ Units; ‘9 If. 'F'I 1. mm PIC 55! .I'J' -\ .y l i 0 I PM .u "in“ s 1 ' w? I ' I '4': I :* ., I \l. .I' :. K P‘ u"; T‘Ni ‘ff‘\v I ‘II ' s 14 internal validity, personal feelings and rater fatigue being among the most comnon. Thus, it is desirable in an experiment using pretest and posttest essays that compositions from both experimental and control groups be shuffled and sequenced indiscriminately throughout the rati ng period. Another source of instrumentation contamination brought out in the Braddock Report deals with statistical procedures: an investigator, state the authors, "should be reasonably consistent in his use of [statistical] procedures. He should not, for example, swi tch criterion measures in different parts of an experiment when there seems to be little basis for the change, as did the investigator who used the subordination index as a measure Of language development from grade four to grade eight but then depended upon sentence length as his measure through grades nine to twelve" (p. 24). 5) Statistical Regression: This phenomenon occurs when "persons whose initial scores were toward either extreme (very low or very high Scores) tend to score nearer to the mean on retest than they did on 14 the original test.“ That is, if students have been selected for exDer'imentation on the basis of their extreme scores--say, students who Score extremely low on achievement tests--careful consideration s‘hOUld be given to the possibility Of a regression effect in the data. 6) §El ection: This term refers to biases "resulting from differential v‘:IQIE<:1:ion of respondents for the comparison groups" (p. 5). Experi- me“till and control groups should be equivalent to begin with. If they are not equivalent, if, for example, the experimental group is S“Perior to the control group in some way, then a gain specific to the eXperimental group may be attributed to this initial .gn’p Injflp" o": >t'tvu — "fluor‘w: ‘P- ' Op '. '33U‘3r‘... I p ' - I '1 :rnu; 0": R - :i'scns I i ‘9‘. . n ”In": & . u'. ‘o J,_"‘;¢;A ‘v!Jr l 6;. V '1 V..I NW". t '.'I 15 non- equivalence. Thus, the effects of the experimental treatment might be confounded with the selection difference. The control for this selection threat is brought up in the Braddock Report when the authors cite the need to control such essential selection variables as " the mental ages, writing proficiency, and socioeconomic and inte‘l lectual home backgrounds of the students" (p. 25). Randomization is the best assurance of group equality because it rules out the pass “i bility that the groups would have differed anyway even without th e e xperimental variable . 7 ) Experimental Mortality: This concerns "the selective dropping out 01“ pe rsons from one of the groups" (p. 12). Campbell and Stanley Prov i de a good illustration of this threat: Typically, experiments on teaching methods are spread out over days, weeks, or months. If the pretests and posttests are given ‘i n the classrooms from which experimental and control groups are drawn, and if the experimental condition requires attendance at certain sessions, while the control condition does not, then the differential attendance on the three occasions (pretest, treat- Iiieent, and posttest) produces "mortality" which can introduce S ubtle sample biases. If, of those initially designated as experimental group participants, one eliminates those who fail to show up for experimental sessions, then one selectively Shrinks the experimental group in a way not comparably done in ' the control group, biasing the experimental group in the direc- tion of the conscientious and healthy. (pp. 15-16). E K x eV‘imental mortality may be a particular threat where an experiment e xtehds over a considerable length of time. B ) filection-Maturation Interaction, etc.: This concerns the possi- b ‘ - j 1 ‘I ty that extraneous variables such as history, testing, and In atuY‘ation interact with whatever specific selection differences dis- t ‘ 1 "Quish the experimental and control groups to begin with. If, for example, persons in the experimental group are chosen because of low ....' 'r £153:- IC 3C» éliiiTlC 6C? 22.32: as '3: selecti: titration m“ 32.?“ .u EV )6 academic achievement while those in the control group are of higher academic achievement, a gain specific to the experimental group may be caused as much by the fact of increased attention to their achievement in the experiment itself as by the treatment variable. In this case, the selection difference (low achievement) could interact with the matu ration variable (motivation) to produce a gain that might have occ u r‘red even without the treatment. So far I have discussed only the variables which pose threats to the internal validity of a research design. In the next section I w-i ‘l 1 consider threats to external validity. Though my focus will be c n external validity, the reader should keep in mind that both Elite rnal validity and external validity can be endangered by the same VaP‘i ables. In the closing paragraphs of Chapter II, the Braddock Report urges a number of imperatives in the reporting of results in a research study. One of these imperatives reads: "the nature of the 51:“clients must be described in enough detail to permit the reader to date “mine for which kinds of students the results are applicable, and the ‘i nvestigator must be careful not to generalize his conclusions beyond the limitations of the type of population he sampled from" ( p ~ 27). Here, the authors are referring to external validity. In Eh e1 r earlier comments on variable control, they were also dealing 1 mp1 icitly with questions of external validitx; but I believe these 8 cattered comments can be reformulated and elaborated on more pre- Q‘i 3e13,, as I have tried to do in my discussion of threats to internal v - My. I will draw on an excellent extension of the Campbell and l7 Stan‘l ey book--Bracht's and Glass' 1968 article, in which the authors treat external validity far more comprehensively than did Campbell and Stan 'l ey. Bracht and Glass define external validity_as "the extent and manner in which the result of an experiment can be generalized to di f’f-‘erent subjects, settings, experimenters, and, possibly, tests." The authors go on to divide the threats to external validity into two broad classes: 1) mpulation validity, or "those dealing with gene ralizations to populations of persons (What population of subjects can be expected to behave in the same way as did the sample experi- men tal subjects?); and 2) ecological validity, or "those dealing with the ' environment' of the experiment (Under what conditions, i.e., Sett‘i ngs, treatments, experimenters, dependent variables, etc., can the s ame results be expected?)" (p. 438). External validity, then, COT-Ce ms the applicability of the results of an experiment to a group of De rsons not included in the original experiment and most likely not trea ted under the controlled conditions of the experiment. when we app‘ly the results of a research study to our classroom practice, we are basing that application on judgements of external validity. Egpulation validity draws attention to the relationship between those persons directly involved as subjects in an experiment and those De ”Sons to whom the experimental results are projected. As Bracht and 8" aSs point out, "One of the purposes of a research study is to learn omething about a large group of pe0ple by making observations on a "31 atively much smaller group of subjects" (p. 440). Population M, then, calls for a close identification between these two ofi I<.‘.eeds smoothly from one idea to the next. Terms and criteria Sho I41 (1 be defined carefully, preferably in an operational manner, pemi tting others to use the terms and criteria with the same b esu'l ts" (p. 23). 2 ) mtiple-Treatment Interference: This occurs in instances in which “ two or more treatments are administered consecutively to the same FTe "8 ons" (p. 456). When only 9_r_1_e_ treatment is used, the response to ‘ t Can easily be measured. But when a number of treatments are used, “es houses to subsequent treatments often depend on the earlier treat- Vents. For example, if an experimental writing group's essays are as “9 thoroughly marked, graded, and revised, it may be difficult to mine W“ 1:13} resul' 3 '-an"."'2"19 Madge th: wse to 1 2113‘. be ac: ire Eradiock I o A i 1 ‘ ‘0" 5h: 54-5th \ I-iéutncme Ef 3'3; when a " . ii filth the fit: of ent iii‘flarly, if ‘f‘zfai apph' 2} Occur 11m“ twmnter 3:415» «NJ Repo ““35 come; :(;€r:m ..ls€ntd) f chi-'1 '1 Would n( ‘o r i was»! "i ~S writ “Wt did no 21 dete mine which of these three treatments contribute to the improve- ment of writing performance (if such improvement should be the experi- mental result). 3) Hawthorne Effect: This concerns the possibility that a "subject's know ‘ledge that he is participating in an experiment may alter his res ponse to the treatment. In such cases the experimental results cannot be accounted for entirely by the treatment effect" (p. 457). Th e Braddock Report notes the effect of increased motivation generated by the Hawthorne Effect: "This experiment illustrates . . . the ' Hawthorne Effect,‘ the added stimulation received by an experimental 9P0 up when a new method is being compared to an old method" (p. 26). 4) Novelty and Disruption Effects: This threat appears in instances ‘i n which the treatment, simply because it is new, creates a certain amount of enthusiasm, which may account for the treatment's effects. sim‘i ‘l arly, if the experimenteris unfamiliar with the treatment in the i "‘i t ‘3 al application because of its newness t9__h_i1n_, a disruption effect may occur which will produce effects that may not appear once the e"IDerimenter has become more familiar with the treatment. The B"‘addock Report notes this threat to external validity when the ant-ho rs comnent on the introduction of some irregular element in the exDer-imental situation: "If a procedure or instrument is being used Wh i Ch would not be employed in a regular teaching situation (such as a k~)’Itlograph, recording on a moving drum the starts and stops of a St” dent's writing), steps should be taken to insure that the atypical Q1 e"lent did not affect the outcome of the experiment" (p. 26). The s - “b3 ect's receptivity to newness is thus another element of design re: the exP€ 7'15 eifect C e, 1e quite rages have 5' Ex:=e."""er.‘. ‘s’ances or star's behe iveaTed expe axeararteus a! :19 subjec ‘35 variables the men the WM» .I.:g39, 9X swims ( W 1 Pete “it! or res; "he rESLMS 0] 9h .1: eff 'I'C‘CtS 0' ;.‘-,- 1 u . r. A [a] ' Ch the (“3’ if there ' the results ‘ be thi‘eaite A? 22 that the experimenter must consider, as Bracht and Glass point out: "The effect of some new program in a setting where change is comon ma y be quite different from the effect in a setting where very few changes have been experienced" (p. 439). 5) Experimenter Effect: This threat has to do with unintentional in F1 uences on the behavior of the subjects generated by the experi- menter's behavior, such as encouragement, annoying mannerisms, revealed expectations, etc. It also includes the experimenter's appearance--sex, age, and race--which may operate as another influence on the subject's behavior. These unintentional influences are among the variables to be controlled that the authors of the Braddock Report Ci te when they mention the possible bias produced by the personality, know‘l edge, experience, and attitudes of the teacher in classroom expe riments (p. 25) - 6) pretest Sensitization: This refers to instances in which pre- tes‘lZed subjects become sensitized to the experimental variable through the experience of the pretest. Campbell and Stanley also note this threat but call it the reactive o_r interaction effect 31: testing, "in “h ‘3 ch a pretest might increase or decrease the respondent's sensi- t ‘ - . . 1 V‘Ity or responsiveness to the exper1mental var1able and thus make the results obtained for a pretested population unrepresentative of the effects of the experimental variable for the unpretested universe from which the experimental respondents were selected" (pp. 5-6). Thus. if there is evidence of a pretest effect, the generalizability o F the results to occasions where a pretest will not be administered may be threatened. 71—9—51 '83 s‘t’ez‘. 53338" 3:21:63t “Se. mfifitotm 233%“ and S peer. the 8} MT] this 8“ 1 me. that Emmet” z::‘ (p. 436: a. “=’ ’3‘ : ,..,ral 396“," . = '1 523% exar rm and sum 1'23 presents m; we dilly be ax hi 51' .class repc 7 z‘ . =1. fut the '_Io" I , ‘ ‘ Jy to Si TP‘Q 1; *Lt'actior an}: ““0“ at Et‘ "4'13 0f the :fgnc‘ . On other “‘ ”‘9 here try. wns of 23 7) flosttest Sensitization: This threat occurs when the experimental effect appears because of the posttest. That is, the occasion of the posttest itself may, like the occasion of the pretest, sensitize the subj ect to the expectations generated by the experimental variable. Carrpbell and Stanley stress that the "more obvious the connection between the experimental treatment and the posttest content, the more like 13 this effect becomes" (p. 2l). Bracht and Glass point out, for example, that the wording of posttest questions or illustrations may ”provide a crucial opportunity for the student to acquire the con- ce pt" (p. 436). Both Campbell and Stanley and Bracht and Glass suggest us i ng natural settings to counteract the threat of pretest or posttest Lens ‘3 tization. According to Campbell and Stanley, "Through regular 61 as 5 room examinations or through tests presented as regular examina- tions and similar in content, and through alternative teaching proce- dures presented without announcement or apology in the regular teaCh ‘i ng process, these two sources of reactive arrangements can prODably be avoided in most instances" (p. 22). Similarly, Bracht and G1 ass report that in experiments "where post-test sensitization may effect the measurement of the treatment effect, the experimenter s houj (1 try to employ valid unobtrusive measures" (p. 464). 8 . ) W91 flistory and Treatment Effects: This concerns local condi tions at the time of the experiment which "may affect the res u] ts of the treatment in such a way that the effect would not be Found on other occasions" (Bracht and Glass, p. 464). The threat occurring here is that the effect was indigenous to the historical co - “(11 tions of the experiment, thus making generalizability to other ztzftians d1 1 "teral va‘. it E: hasxeeen‘ [ixlfiflffhté] :‘Tssiy‘ what we this to 0:? talent var‘ I“: which is construc ~‘. V I1 .‘;h~ q' . llrl <4 ar :25: an experi ‘32 lexhdent is being me 3:":3 out the ‘iird upon cri Me. The The should b '1?” d ‘ and that r .., flier-a H w name that ' 24 conditions difficult. Hence, historical conditions may threaten both internal validity and external validity. 9) Measurement gj_the erendent Variable: The dependent variable (experimental outcome) must be clearly defined if we are to know pre- cisely what we are talking about when we generalize the experimental results to other settings. In addition, precise measurement of the dependent variable depends upon the "selection of a measuring instru- ment which is assumed to measure both reliably and validly the under- lying construct" (Bracht and Glass, p. 465). I will discuss reliability and test validity_later; but, for now, suffice it to say that an experimental design is threatened if it is not certain that the dependent variable is in fact being measured (validity) or that it is being measured consistently (reliability). The Braddock Report points out that "Statistical analyses in composition research are based upon criterion measures about which certain assumptions must be made. The nature of these assumptions should be made clear, and there should be fairly adequate evidence that the assumptions are valid and that the criterion measures can be applied reliably" (p. 24). l0) Interaction Qf_Time‘gf_Measurement and Treatment Effects: It is possible that a treatment effect "which is observed immediately after the treatment period may not be maintained at some later time, e.g., a unnth or six months after the treatment period. Most experimenters fail to take the time element into account and thus risk invalid generalization effects to other points in time" (p. 466). This is a crucial consideration fOr research design because it points to the necessity for seeing how much, if any, of the treatment has any iast'fn; hold :2 exserinen 36"., then it 39:: point is an}: be tai :"a:, to see $325, when :‘ter. such is In th “telret‘tjo use”) desii Rims prESei inside, each ’Ejfgjty 9mm,» E‘iI-Een deSlgy 14‘. mm of <1 096m its, 0r fai] 4. {i 'Vgn‘ This Mr . e r 190 25 lasting hold on the experimental subject. If, months or years after the experiment, the subject shows little or no effects from the treat- ment, then the validity of the original experiment is threatened. The same point is made in the Braddock Report: “Often a follow-up measure should be taken, months or even a year after a new method has been tried, to see how learning stands up for experimental and control groups, when instruction and practice lie in the past" (p. 26). Too often such follow-up testing is not included in a research design. In this section I have summarized the threats to the two classes of external validity_in research design--pgpulation validityy and ecolOgjcal validityr-as developed by Campbell and Stanley and by Bracht and Glass. My discussion of these threats as well as the threats to internal validity is intended to aid the reader in the interpretation of experimental results by careful scrutiny of a research design. I will now turn to an examination of three research designs presented by Campbell and Stanley in their book and briefly consider each in light of the threats to internal and external validity enumerated above. Altogether, Campbell and Stanley present sixteen designs along with some variations on them. I will focus on just three of these designs to illustrate how the threats to gggign validity operate on commonly used designs and how each design con- trols, or fails to control, for these threats. I will rely on Camp- bell's and Stanley's graphic notation to describe the features of each design. This simple notation is as follows: An X represents the exposure of a group to an experimental variable or event, the effects of which are to be measured. AI: 0 rep H An R rep?! S'UJQS' Mashed ‘ waved b.’ The left ‘ The 1‘ £6971," Hhic :‘.,s:rat6$ a in, we see i a: :ne outset, haired as a treaten the l :EEIZ' Sin: ii incur to 1 hszner varial =-.-“er.is may I .:i. Athi: mai masm 36' “W J3: ml :6 Anoti when the c - extreme 26 An 0 represents some process of observation or measurement. An R represents random assignment to separate treatment groups. A dashed line-- _____ --represents comparison groups not equated by random assignment. Each horizontal row indicates a group. The left to right dimension indicates temporal order (p. 6). The first design to examine is the "One-Group Pretest-Posttest Design," which the authors call a "pre-experimental" design because it illustrates a number of extraneous variables that can jeopardize internal validity. A scheme of the design is: Thus, we see that one group is used in the experiment; it is measured at the outset, subjected to one experimental treatment, and then measured as a follow-up. But a number of uncontrolled variables may threaten the hypothesis that X caused the 0]____92 difference. One is history. Since only one group is involved, events in addition to X nay occur to students in the group which may cause the change. Another variable is maturation: between the two measurements, students may have grown older, more verbally sophisticated, more tired. A third variable is the effect of testing. That is, the initial measurement may make the experimental group more self- conscious, more alert to the experimental variable, more motivated to change. Another variable is statistical regression, which will threaten the design if the experimental group has been chosen because of its extreme score on 0]. Because of this design's lack of control evertnese (an: shad exahpi e' A secoi high,” which 1! virtue of i ‘ moi which i threaten the "( Q 1‘5 cont iii-3 pmdUCEd E H)\2 Chi w is c :11? Eltreim DC “I? regrESSeS 331 early, 5618 \ 0f the ation V experimental 3:: . only One kji'i‘it. \iflj Dis; \ "t - slf the 27 over these (and even more) variables, Campbell and Stanley cite it as a "bad example" (p. 7) of a research design. A second design is the "Pretest-Posttest Control Group Design," which Campbell and Stanley call a "true experimental design" by virtue of its randomization and the addition of a control group, both of which provide for control of many of the variables that threaten the "One Group Pretest-Posttest Design." The scheme of this second design is: R 0 X 0 History is controlled "insofar as general historical events that might have produced an 01 02 difference would also produce an 03 04 difference" (p. l3). Similarly, if maturation and testing influence an 01 02 change, this should also appear in the control group. Regression is controlled if both groups are randomly assigned from the same extreme pool. "In such a case," write the authors, "the control group regresses as much as does the experimental group" (p. l5). Similarly, selection is removed as a threat through randomization. 0f the factors which jeopardize external validity, threats to_pgpulation validity can be controlled through random selection from an experimentally accessible population with characteristics shared by the target population. Multiple-Treatment Interference is ruled out because only one treatment is used. The Hawthorne Effect and the Novelty and Disrgption Effect can threaten the generalizability, however, if the student is aware of his participation in the experiment jai I... i n (f 3 (I, e: ‘:r :53. of 1 :‘ass'cae. an: 226‘. pm __ or 2 :etes‘. sens f' is. the 32;.Tetion i a...‘ 3: berar‘ 3:22 a three I. 11., '; ‘ 533:: 6‘5: 6 .3 .6', L‘.\ h-IE :r‘;L‘, i ‘5. o: 0‘ E - l bl 28 and if the experimental setting is too artificial; the best control for both of these factors would be to use a natural setting such as a classroom and embed X5 and Us in the normal classroom routine. Such control procedures would also limit any contamination resulting from pretest or posttest sensitization. I should note, however, that a pretest sensitization threat may still exist if one wishes to general- ize from the pretested experimental group to the unpretested target population in any application of the experiment's results. Finally, the interaction _o_f_ time 9_f_ measurement and treatment effects may also pose a threat unless post-experimental measurement is carried out. This would alter the design scheme by adding additional Os beyond 02 and 04, but it would strengthen the ecological validity of the results. A third design is the "Nonequivalent Control Group Design," which Campbell and Stanley call a "quasi-experimental design" because, while it lacks full experimental control, it is nonetheless advocated in "those settings where better experimental designs are not feasible.* The scheme of this design is: In contrast to the "One Group Pretest-Posttest Design," this design has a control group; but, unlike the ”Pretest-Posttest Control Group Design," this one does not assign subjects randomly from a common *In fact, one of the main themes of the Campbell and Stanley book j§_the usefulness of such designs, which in some instances are preferable to "true experimental designs." One such instance might be a classroom-situated experiment, as I explain above. I}. Q.’ ":J" c'. gpaonnp . d 5'6 LIV CI) 1 —.l— he. er 'r.‘ 3? 6 Med. 5 1 a“ 2' 'i il't'h-e :7"; r I. '3 29 population. This design is most often used in settings where random assignment is not possible or feasible. Hence, it is a widely used design in educational research where intact groups such as classrooms, which are naturally assembled rather than randomly assigned, are used. Though the experimental and control groups do not have pre-experimental equivalence (which would be attained through randomization), there j§_ some equivalence established through pretesting. If similarity between groups is confirmed by the pretest, then certain threats to internal validity can be controlled: "we can regard the design as controlling the main effects of history, maturation, testing, and instrumentation, in that the difference for the experimental group between pretest and posttest (if greater than that for the control group) cannot be explained by main effects of these variables such as would be found affecting both the experimental and the control group" (p. 48). The major threat to internal validity_with this design is with selection- maturation interaction, etc., since the experimental and control groups are not equivalent to begin with. For example, there is the possibility that in spite of similarities derived from pretest data, one group may have a higher maturation rate than another. Such a selection-maturation interaction could thus be confounded with the effect of X. The threats to external validity_are the same as those I enumerated in the "Pretest-Posttest Control Group Design.“ The theoretical designs presented above are but three of the sixteen discussed by Campbell and Stanley. I have chosen these three as sufficient to illustrate the effects on research design of threats to internal and external validity. I will now consider selected design aspects of three of the research projects sumnarized in the Brazecei Repel” “ye studies i :2 :crmwate i‘ni must” resented 650" 33*” the a: research Stud) n‘httntion i seer in cons I find 21:61 is flaw \ threat. gmps' treetfllf assays were W gene's essays er'sed. Thus, "tensive mark) éi‘ect of each Seasick Report 6:: is of each the scrutiny ~::‘:ent revea ww- 1"“ (r, ~ raddcck su 30 Braddock Report as "most soundly based." Though the Report considers five studies in detail, three will be sufficient for my purpose here-- to corroborate the Report's notations on design_validity_and to make a final illustrative application of the design validity_elements I have presented above. One caution, however, is in order. I am well aware, as were the authors of the Braddock Report, that their five selected research studies were not "perfect in all respects" (p. 55). Hence, my intention is not to attack these studies but rather to help the reader in considering the validity of their designs. I find, for example, that the external validity of the Buxton study is flawed by failure to control the multiple-treatment interfer- gggg_threat. The difference between the "Writing" and "Revision" groups' treatment in that experiment is that the "Revision" group's essays were thoroughly marked, graded, and revised while the "Writing" group's essays were commented on only briefly and not graded or revised. Thus, the experimenter claimed three treatment variables-- intensive marking, grading, and revising. Failure to measure the effect of each of these variables is noted by the authors of the Braddock Report: "It is not clear, however, what the relative influ- ence is of each of these three factors" (p. 70). Furthermore, a closer scrutiny of the events occurring to the "Revision“ group during treatment reveals even more variables at work. I will quote from the Braddock summary of this group's treatment: The assignments and papers in the Revision group were treated with considerably more direction. Although the students could develop the assigned tOpic "in their own way," they were expected to write on the same t0pic and to include some critical thinking, a central idea, and material that was organized and (developed. They were encouraged to organize preliminary ideas 31 into an outline before beginning the theme itself, to choose their words and illustrations carefully, and to develOp good unity within paragraphs and transitions between them. They were also warned against using unqualified and fallacious statements. These qualities--as well as errors in spelling, punctuation, and sentence structure-~were marked on the papers, and a few sentences of general evaluation were written at the end, including mention of commendable qualities as well as suggestions for improvement. Each paper was given two grades, one for content and organization, another for general correctness and accuracy. The papers of the students in the Revision group were returned at the beginning of a class period. The general strengths and weaknesses of the essays were pointed out at that time, and excerpts exemplifying certain good features were read to the class to elicit comments on how the effective- ness was achieved. Then the students were required to correct the errors indicated on their papers while the reader went from student to student, giving assistance where it was needed. (p. 61) A number of variables here may well have interacted with the three treatment variables to produce the experimental results. The first is motivation: the students were given more direction and encouragement to write well. The second is mild intimidation: the students were "warned against using unqualified and fallacious state- ments." The third variable is pger interaction, which is a powerful motivational and correctional strategy: excerpts from students' papers were read to the class and discussed. A fourth variable is individual attention: while the students revised their papers, the reader gave assistance to individual students. All of these uncon- trolled variables can be subsumed under the multiple-treatment interference threat to external validity. That is to say, along with the three independent variables the experimenter claims in his con- clusions--thorough marking, grading, and revising--we must also add motivation, intimidation, peer interaction, and individual attention. Since any one (or all) of these additional variables may also account I'C TEVTSE $1.6er . ' "v "a ma ge'e'ai 5:: W59 thEl x"; that r;-'- ,6 -=6« aha. 32 for Buxton's results, the experimenter's third conclusion--that "College freshmen whose writing is graded and thoroughly marked and who revise their papers in light of these matters can improve their writing more than college freshmen whose writing receives a few general suggestions but no grades or intensive marking and who do not revise their papers"--would seem to misrepresent the actual number of variables in the experimental treatment. Similarly, the Smith study reflects a threat to external validity or, more specifically, to pgpulation validity, This is the point that the Braddock Report makes in saying that "a question was raised about how representative the University High School ninth graders would be of school pupils in general" (p. 96). That the students in this experiment were of high intelligence and high socio- economic background limits the applicability of the experiment's result to a rather small prOportion of the larger American school p0pulation, as the Report stresses in the beginning of the summary of this study. The Kincaid study, on the other hand, is cited as being one with a "superior design." This praise derives from Kincaid's careful control of threats to both internal validity and external validity, Factors jeOpardizing internal validity were Controlled in many ways. For example, Kincaid tried to control the history threat by cautious Spacing of the students' writing occasions. He did not wish to space the writing occasions "so far apart that new learning experiences would cause more differences in quality of writing than variations in effi- ciency would cause" (p. 86). Hence, he spaced the two days of writing one week apart. Clearly, his intention was to prevent the intrusion of unintended influences on the experimental outcome. n (3': M r vd-g .. n- r‘ i ’ it ‘6')! I“ u,_ “ f '6' e -. HEI- :" are k‘~_;~ ,‘ '~. ”n! "r. 5. I t 3,3: '2 33 The fatigue factor in the maturation threat was controlled when "every other student wrote on the first t0pic during the first hour, took a ten-minute break, and then wrote on the second topic. The other students first wrote on the second topic and then, after a ten-minute break, on the first tapic" (p. 85). Similarly, Kincaid sought to control the rater variable of the instrumentation threat by having all three graders judge all four papers from each student during the same rating period and by staggering the themes of indivi- dual students in their order of evaluation. Kincaid was equally careful in his control for selection bias. To insure that his subjects were "representative of the freshman population at Michigan State, the investigator relied on the usual registration procedure for the course," thereby obtaining a random sample from the university papulation. Furthermore, as a check on the equivalence between the four groups in the experiment, Kincaid used the first theme written by the students to determine equality in their ability to write. He was thus confident that "each of the four groups was representative of the total group of 80 students" (p. 87). Kincaid also instituted controls for external validity. As I noted earlier in my discussion of ecological validity, for example, a "natural" setting such as a classroom is desirable for the generaliza- bility of an experiment because of its approximation to the conditions of everyday life. It is significant, then, that Kincaid conducted his experiment in regular writing classrooms with regular writing instruc- tors. He also contributed to this natural setting by choosing topics that were “similar in nature to writing assignments used previously in the term" (p. 84). 'r: .i-~ n.- .U a Li. Z a t. u a .1- .. . «L... . my. a ”J 3.1.x. Lula. - I: LHU 34 In conclusion, I wish to add that of the five studies sum- marized in detail in the Braddock Report, I find the Kincaid study the soundest in desigg_validity, I will have further comments to make on this and on the other studies in the Braddock Report in the sections that follow here on test validity_and measurement reliability, Test Validity In the introductory part of this chapter I stated that in my examination of test validity I would point out its importance in assuring that an experimental test "measures what we want it to measure." This is a basic definition of test validity, (There are a number of variations on this definition, but most adhere to essen- tially this idea.) Put another way, the concept of test validity answers the question, "What does the test measure?" For example, if a test is supposed to measure students' skill in assessing grammatical correctness, then we must be sure that the test does in fact measure that skill. If it does not measure that skill adequately, then it nay be said to have low validity as a test of skill in assessing grammatical correctness. Thus we see that this definition of tggt validity relates to l) what is actually being measured by the test; and 2) the gggggg_to which the test measures what it supposedly measures, which is another way of saying that test validity is a matter of degree, that tests are not just considered valid or invalid. A good example of degrees of validity appears in Clinton 5. Chase's book, Measurement for Educational Evaluation, where the author writes: "an intelligence test is valid to the extent that it tells us the truth about an individual's capacity to perform intelligent acts . . . 35 suppose we have two intelligence tests--A and B. We discover that Test A has often provided scores that relate more closely to achieve- ment in complex learning tasks than has Test B. Shall we say that Test A is valid and Test B is invalid? No, both have degrees of validity, but Test A is more valid than Test B."15 There are four types of test validity commonly used in research testing: content validity, construct validity, and two forms of criterion-related validityf-predictive and concurrent. I will explain each here and provide examples wherever possible. Content validity refers to the relationship between the actual content of the test and the subject-matter or performance the test is supposed to measure. Thus, content validity is defined in the litera- ture on educational and psychological measurement as “the extent to which a test measures a representative sample of the subject-matter content and the behavioral changes under consideration.“16 Content validity is a central concern in achievement testing, because such tests are intended to represent instructional objectives as well as the content of instruction. As Richard H. Lindeman writes in Eggga; tional Measurement, "An achievement test has content validity if it represents faithfully the objectives of a given instructional sequence and reflects the emphasis accorded these objectives as the instruction was carried out." Thus, Lindeman continues, "A test in modern algebra would have low content validity for measuring achieve- ment in American history. A test in long division would have low content validity if administered to second-grade pupils. Nhen students criticize a test as not fairly representing the actual 36 content of the course, they are in reality remarking about the test's “17 content validity. Construct validity refers to the interpretation of test scores in terms of some general psychological quality (e.g., some ability, trait, or attitude) known as a construct. In the educational and psychological literature, this kind of validity is defined as "the extent to which test performance can be interpreted in terms of cer- tain PSYChological constructs."18 Examples of constructs are anxiety, intelligence, reading readiness, critical thinking, mechanical interest, study skills, and verbal ability. The value of making inferences based on construct validity_is well put by Gronlund: "There is an obvious advantage in being able to interpret test performance in terms of such psychological constructs. Each construct has an under- lying theory which can be brought to bear in describing and predicting a person's behavior. If we say a person is highly intelligent, for example, we know what behaviors might be expected of him in various specific situations."19 In addition, a test with high construct validity may provide infbrmation to help teachers understand students' academic performance. A diagnostic reading test, for example, may tell us a lot about a person's ability to conceptualize abstract features of language, such as inflections or other phonological properties. A Criterion-related validity is a comparison of a person's test scores with an actual performance (the criterion). The test is valid to the extent that the scores corroborate the actual performance. Of the two forms of criterion-related validities, predictive validity is involved when test scores are used to forecast future performance. 37 Concurrent validity, on the other hand, is involved when test scores are used to estimate an individual's present performance. The differ- ence between these criterigngrelateg,validities_is well illustrated by Gronlund: For example, reading readiness test scores might be used to predict pupils' future achievement in reading, or a test of dictionary skills might be used to estimate pupils' current skill in the actual use of the dictionary (as determined by observation). In the first example, we are interested in prediction and thus in the relationship between the two measures over an extended period of time. This type of validity is called redictive validity. In the second exam- ple, we are interested in gttimatipg_present_status and thus in the relationship between the two measures obtained con- currently. A high relationship in this case would show that the test of dictionary skills is a good indicator of actual skill in use of the dictionary. This procedure for deter- mining validity is called concurrent validity . . . The major difference between the two kinds of validity resides in the time between the two obtained measures. Criterion-related validities are usually reported statisti- cally in the form of correlation coefficients. This statistical procedure expresses the degree of agreement between the original test score and the criterion measure being used as a basis for comparison. Gronlund's explanation of correlation coefficient is concise and lucid. Basically, a coefficient of correlation expresses the degree of relationship between two sets of scores by numbers ranging from +l.OO to -l.OO. A perfect positive correlation is indi- cated by a coefficient of +l.OO and a perfect negative corre- lation by a coefficient of -l.OO. A correlation coefficient of .OO lies midway between these extremes and indicates no relationship between the two sets of scores. Obviously, the larger the coefficient (positive or negative), the higher the degree of relationship expressed. He may use Gronlund's example of a reading readiness test (see above) to illustrate positive correlation and negative correlation. If a person scores very high on the reading readiness test and snseq; csrrela achieve nee-ii he negati V IN". rEii‘ilZp E‘iSure a”Steel 38 subsequently attains superior success in reading, then a positive correlation may be said to exist between the test scores and later achievement. But if a person scores extremely low on the reading readiness test yet still attains superior success in reading, then a negative correlation exists. Perfect positive correlations seldom occur. What, then, is a satisfactory correlation coefficient for the test interpreter to rely on? Most correlation coefficients found in test reports run between .35 and .80; and, though most experts recom- mend coefficients in the neighborhood of .70 or better, they also agree that our evaluation depends on our purpose for using the test. That is, our ideal is a high correlation in a positive direction, but if we do not have any other information upon which to base judgements which must be made, then we may well be satisfied with a validity coefficient of, for example, .38. In doing so, however, we must realize that the .38 correspondence between test scores and criterion measure indicates a small amount of agreement and thus should be inter- preted cautiously. I have presented the four types of test validity independent of one another. It should be noted, however, that the four overlap considerably. This interdependence among the various test validities was stressed most recently in the American Psychological Association's Standards for Educational and Psychological Tests: These aspects of validity can be discussed independently, but only fbr convenience. They are interrelated Operationally and logically: only rarely is one of them alone important in a particular situation.2 39 One example of such interdependence might occur in a diagnostic reading test which could have high construct validity as a test of a person's reading readiness as well as high predictive validity_as a measure of future reading success. Another example appears in the Braddock Report's summary of the Harris study. I remarked earlier in this chapter that the authors of the Braddock Report direct their readers' attention to the manner in which Harris validated his cri- teria of measurement. I wish, therefore, to examine Harris' valida- tion procedure as a way of applying the information I have provided here on test validity. Harris' problem was finding a frequency count test that would validly measure "'the rate of growth of a child's maturing style'" (p. 74). Evidence that a test would in fact measure this construct would be the results indicating that a "satisfactory difference" exists between the test scores of two groups of differently aged children. Earlier measures tried by the investigator did not result in any significant difference between different age groups (i.e., did not demonstrate construct validity). But when the investigator analyzed compositions written by one group of ten year-olds and those written by another group of fifteen yearbolds, he developed a set of eleven criteria "'which occurred sufficiently often to give a clear measurement.'" After computing the difference between the groups, Harris found that the test did indicate a significant difference on all eleven criteria, thus providing the experimenter with a valid test 01’ maturing style. Consequently, Harris used this test in his formal 9*Periment, but not until he determined the reliability_of his measuring instrument. 40 Measurement Reliability Reliability refers to the consistency with which a measuring instrument (e.g., essay, objective test, essay rater) measures from one occasion to another. If an instrument does not give consistent measurements, then it may be said to have low reliability_for whatever it measures. Chase provides a simple and clear illustration by using the yardstick as the measuring instrument: For example, if I measure the length of a room with a yard- stick, I should get about the same result today as I did yester- day. My measuring procedure is reliable. But suppose I have an elastic measuring tape a yard long. Some days I tend to pull it too taut and get more than three feet in each unit. On those days I underestimate the length of the room. Other days, I do not pull it taut enough, and I overestimate the length of the room. My measuring device is inconsistent, or unreliable.23 There are two main ways for determining the reliability of a measuring instrument. The first is through methods of calculating correlation coefficients and the second is through calculating the standard error pf_measurement. Correlation coefficients for determining reliability are calculated and interpreted in the same way as the procedure I described in the section on determining validity coefficients. When determining reliability coefficients, however, agreement is based on the correlation between two sets of the same or very similar measure- ments. This difference becomes clear in considering the methods of calculating reliability coefficients. One method is called "test-retest," in which the same test is given to the same group twice with a certain amount of time inter- vening between the two testing occasions. Scores from the two tests are correlated and the correlation coefficient indicates how 4] consistent the test results are over the time between tests. The advantage of the test-retest method is that it indicates how consis- tent are scores over a period of time. One caution, however, must be observed. As pointed out in Standards for Educational and Psychologi- £21.I£§£§: "retesting is not ordinarily a desirable method of estima- ting reliability because the examinee may remember his or her responses to items from one testing to the next. Hence, memory becomes a systematic source of variance."24 A second method is through using "parallel forms" of the test. This involves making up two different but equivalent test forms (e.g., both sample from the same content, level of difficulty, etc.). One form of the test is administered; and, after some time, the other form is administered. The memory variable which threatened the test- retest method is controlled because the two forms of the test contain different items. Scores from both test administrations are correlated and this correlation coefficient tells us the degree to which the test is consistent between the two forms. A third way of estimating reliability is by means of the "split-half" method. This involves a single administration of a single form of a test. After the test is given, it is divided into two equivalent parts (odd and even numbered items, for example) and the correlation coefficient between these two parts is calculated. Thus, the split-half method indicates the degree to which the two halves are consistent with one another in their measurements. This method is also known as a measure of internal consisteney_in that items on the test are correlated against one another. 42 I pointed out above that there are two main ways of determining the reliability of a measuring instrument. One way, as I have shown, is through calculating the correlation coefficients. The other way is through calculating the standard error pj_measurement. Standard gppppngt_measurement indicates how much we would expect a person's score to vary if he were to be measured a number of times with the same test. That is to say, since there will be minor fluctuations in test scores from one testing to another, any single test score should, be seen as a range of scores rather than an absolute score. He should think of a person's test score as comprised of two components: a tppg_ gpppg_component, which is the score a person would obtain if there were no error in measurement; and a measurement error component. Errors in measurement operate randomly, sometimes increasing a score, and sometimes decreasing it. Standard error pf measurement is thus a statistical procedure which estimates the amount of variation we can expect in test scores due to random errors in measurement. A useful illustration appears in Gronlund's book: For example, let us assume that we have just administered an intelligence test to a class and the results indicate that Mary Smith has an IQ of 97. We note in the test manual that the standard error of measurement is 5. What does this 5 mean with regard to Mary Smith's IQ? In general, it indicates the amount of error that must be taken into consideration in interpreting Mary Smith's IQ score. More specifically, it provides the limits within which we can reasonably expect to find Mary Smith's "true" 10 score. . . . Thus, a range of scores from 92 to l02 would typically be used to describe Mary Smith's per- formance. . . . The standard error of measurement makes it clear that a test score should be interpreted as a "band of scores" rather than as a specific score. With a large standard error the band of scores is large and we have less confidence in our obtained score. If the standard error is small the band of scores is small and we have greater confi- dence that our obtained score is a dependable measure of the characteristic.25 43 Standard error pf_measurement thus gives us a way of expressing a test's reliability in terms of score units: it tells us that a person's or group's obtained scores are not necessarily the "true" scores, but rather scores which fall within an estimated range. It is an estimation of the amount of variation to be expected in test scores due to unavoidable random errors in measurement. Thus, as Gronlund stresses, "The amount of variation in . . . test scores would be directly related to the reliability of the testing procedures. Low reliability would be indicated by large variations in the pupil's test scores. High reliability would be indicated by little variation from one testing to another."26 For anyone reading research reports, an understanding of measurement reliability_is important in interpreting and assessing the merits of the research. This is pointed up in Standards for Educa- tional and Psychologigal Tests, in which the following principle is deemed "Essential": "The test manual or research report should pre- sent evidence of reliability, including estimates of the standard error of measurement, that permits the reader to judge whether scores 27 A5 are sufficiently dependable for the intended uses of the test." a way of applying some of the information I have provided here on reliability, I will examine briefly some of the instances in the Braddock Report where evidence of reliability is presented. In Chapter II of the Report, for example, the authors comment on a study by Stalnaker which emphasized practice rating sessions for composition raters (p. 14). After the first reading of the composi- tions by a number of raters, the correlation between raters ranged between .30 and .75; but, after training, the correlation ranged r9 ~i': lust iii: 4‘" m L (I! It! 4).. 44 between .73 and .98, with an average of .88, thus indicating greater measurement reliability with practice in rating. Rater reliability coefficients are also reported in the Buxton study, where only two raters were used and the coefficient pf_correla- tion was calculated through the Pearson product-moment formula (a very common method for computing the coefficient gj_correlatiop). The authors report a reliability coefficient of .9l for the pretest themes, thus "indicating a high degree of consistency between the two raters" (p. 66). Finally, reliability_calculations were used in the Harris study to insure that his test--which measured "the rate of growth of a child's maturing style"--proved consistent in its measurement of the eleven criteria of maturing style. (See the above section where I discuss Harris' validation procedures used with this test, since insuring both the validity and the reliability of his instrument was one of the merits of the study). In his validation procedure, Harris demonstrated that by relying on his eleven criteria of maturing style, his test did measure differences between the ten year-olds and the fifteen yearbolds. Thus, it proved to be ygiig instrument. To determine the reliability of the instrument, Harris gave the same test to a group of thirteen year-olds. If the test again indicated differ- ences in this median age group, then he could feel confident that it measured the eleven criteria consistently. The results were that "all criteria except a and 9 reached a good or high level of reliability“ (p. 75). Both the correlation coefficient and the standard error pj_ measurement are reported in Table 4 of the summary. The correspon- dence between the low coefficient correlation figures and the high 45 standard error pf_measurement figures in the criteria with lowest reliabilities illustrate the point I made earlier--that the higher the standard error pf measurement obtained, the lower the reliability, In this chapter, then, I have tried to provide information which can assist readers in their understanding and assessment of research reports. I have defined and examined variables which threaten research projects by jeopardizing the internal and external validity of research design. I have also examined and illustrated two other essential facets of research studies--test validity and measurement reliabilitye-by defining each concept and showing how each figures in the reporting of research experiments. In summary, then, a research design should provide control for extraneous variables and should have generalizable results; a research test should measure what it is intended to measure; and, finally, a research measure should measure consistently. CHAPTER II RESEARCH IN COMPOSITION* In "The Crisis in Knowing about Learning to Write," an article which appeared in the September, l975, ADE Bulletin, Robert E. Shafer stated that “It would seem that those of us who are in a position to do so should make every effort to infuse programs of student writing with the 'best that has been thought and said about writing, and that those of us who are not accustomed or experienced in evaluating experi- mental studies of writing should make every effort to familiarize ourselves with their results“ [sic] (p. 56). Shafer is quite right in suggesting that the design of composition programs should take into account the experimental research which has been done in the field. But I believe that part of his statement--"those of us who are not accustomed or experienced in evaluating experimental studies should make every effort to familiarize ourselves with their results" --needs modification. If we consider only the results of research studies, we can easily be led to accept blindly the studies' conclu- sions. He must look beyond the results to questions of design and *In this chapter, I will not use the standard format of foot- noting by numbers. Because I will cite so many research studies, I will use a more convenient system of simply dating the studies in my text. I will provide full reference information for the studies in my notes. 46 47 procedure. To make such information readily accessible was my purpose in the previous chapter-~to help those who are not research specialiSts make critical sense of the results of research by careful scrutiny of some of its essential components: desigp, test validity, and measurement reliability, This is knowing research in one sense-- knowing how to read research reports by understanding research funda- mentals. But we must also know research in a second sense--we must know what specific research has been done in composition. In this chapter, then, I will try to help the reader know research in this second sense. As I noted in the last chapter, there have been major over- views of research in composition in the past few years (Braddock, Lloyd-Jones, Schoer, l963; Meckel, l963; Braddock, 1969; Sherwin, 1969; Blount, l973; and Lundsteen,l976). Each overview summarizes research according to some frame of reference. The Braddock Report, for example, presents topics relating to composition (e.g., environ- mental factors influencipg composition, instructional factors influencing composition, rhetorical considerations, et al.) and a number of studies which explore these t0pics. Similarly, Blount compiles research in the areas of curriculum, the interrelationships ptngammar and writing, vocabulaty, student characteristics, measure- ment, and teacher preparation. In this chapter, I will present an overview of research studies in composition since l963 by reference to the twenty-four suggestions for needed research which were posed in the Braddock Report. These suggestions, which appeared in the form of questions, were as follows: Hhat l of sci Hhat I when ' teach Hhat How d their Hhat his ; At w? intrt drum . Hhat long ency . At w intr . What VEdC ' Hha1 SENS . At l 3961 SUC Str . At' Nri Sch . Hha Sel f0! Hha 10. 11. 12. 13. 14. 15. 16. 48 What kinds of situations and assignments at various levels of schooling stimulate a desire to write well? What do different kinds of students prefer to write about when relieved of the expectations and requirements of teachers and others? What are the sources of fear and resentment of writing? How do the kinds of writing which adults compose vary with their occupations and other factors. What is the effect on writing of having the student compose his paper for different kinds of readers? At which levels of maturation does it seem apprOpriate to introduce the various modes of discourse--narration, poetry, drama. exposition, argument, and criticism? What is the relative effectiveness of writing shorter and longer papers at various levels of maturity and profici- ency? At which levels of maturation does it seem appropriate to introduce the various rhetorical elements of writing? What are the effects of various kinds and amounts of reading on the quality and kinds of writing a person does? What are the direct and indirect effects of particular sensory experiences and guided observation upon writing? At what stages of maturity do students spontaneously seek specific help in improving particular aspects of writing, such as specificity of details, transitions, parallel structure, and metaphor? At which levels of maturation ggg_particular aspects of writing most efficiently be learned? Does the oral reading of rough drafts help the elementary school strengthen "sentence sense"? How does it? What techniques of composition most effectively help build self-discipline and pride in clarity, originality, and good fOrm? What procedures of teaching and learning composition are most effective for pupils of low socioeconomic patterns? What procedures of teaching and learning composition are most effective for pupils learning to write English as a second language? 11. Can 5 18. Can 1 write 19. How 1 or pa 20. Hhat types a vr‘ 21. Hhat 22. How 1 must 23. HOW 1 24. 0f v1 As t1 11th seem ‘ antic“ app 1.52). 1. “We the Br re3.00m to ”"‘te1atir ifiaummk At 1 “Hiram” .333»;- ' dInth ”W d1ffic Elsi-TN e‘. "H SA‘ ”113ch :15 the re 49 17. Can study of the newer types of linguistics help writers? 18. Can formal study of rhetorical theory or of logic help writers? 19. How is writing affected by extensive study and imitation or parody of models? 20. What forms of discourse have the greatest effect on other types of writing? For example, does writing poetry help a writer of reports? 2l. What is involved in the act of writing? 22. How does a person go about starting a paper? What questions must he answer for himself? 23. How does a writer generate sentences? 24. Of what does skill in writing really consist? As the authors of the Braddock Report noted, these questions "which seem fundamental in the teaching and learning of written com- position apparently have gone almost untouched by careful research" (p. 52). I will report, therefore, on what research has been done since the Braddock Report, research which may or may not be in direct. response to these questions, but which nonetheless provides informa- tion relating to these crucial areas. To my knowledge, no one has yet assembled the research studies which pertain to these questions. At the outset, I should state three matters relating to the preparation of my overview. In the first place, some of the questions posed in the Braddock Report deal with psychological matters which are very difficult for researchers to gauge. Question Fourteen, for example--"What techniques of composition most effectively help build SElf-discipline and pride in clarity, originality, and good form?"-- asks the researcher to establish a relationship between technique and two psychological constructs--self-discipline and pride. While many research stud“ and mmveme” 11m: fell inve 7101111195 (or. ‘echnAque)' I 11111 to 5111115 ties (Such as their conc1usi iata derl'Ved S sealest 19”“ iterate f” r11 331111511191 1 re1 mm the 1111‘1111’17-Mi “the“ 9° a‘ 151339111?“ fies-ether a T1 11911. It see stud med 1'11 A as that by inst/vctfon 5:111 grade 1‘5 1tie/f but 15 '93-; A. ,. .1 MW 50 research studies do investigate the relationships between techniques and improvement in such writing skills as clarity, originality, and fbrm, few investigate what impact, if any, a technique has on mental faculties (or, for that matter, what impact mental faculties has on a technique). I will report on both types of studies--those which relate only to skills or achievement, and those which relate to mental facul- ties (such as attitudes toward writing). The former studies base their conclusions on measures of achievement, which limits them to data derived strictly from demonstrable behavior. Still, they do suggest relationships between techniques and achievement which may generate further research into the psychological effects of these techniques. A related matter pertains to the overlap of some questions. Question Twenty-One, for example--"What is involved in the act of writing?"--is closely related to question Twenty-Two--"How does a person go about starting a paper? What questions must he answer for himself?" In such instances, I will group overlapping questions together and report on research most closely related to both. The third matter concerns the selective nature of this over- view. It is selective in two senses. First, I have not included 1 some studies simply because they do not relate to any of the questions posed in the Braddock Report. Thus, I have not reported such studies as that by McElwee (1974), who examined the effects of systematic instruction in proofreading on the spelling_accuragy_of fourth and sixth graders. Another example of a study which is interesting in itself but is not related to the Braddock Report's questions is the tune by Norwood (l974), who conducted an experiment in teaching ehado‘aog this oven t1ose1y tc telling 11" have inve: the eh m £91119. 51 methodology to determine achievement a§_related tg_ethnic origin. In this overview, then, I have reported only on those studies which relate closely to the questions posed in the Braddock Report. Research dealing with such matters as spelling, ethnic origin, vocabulary, teacher preparation, and the like have not been included. This overview is also selective in a second sense. While I have investigated both published and unpublished research studies, I have eliminated some studies because of major flaws in design or pro- cedure. Such was the case with one study which investigated whether students enrolled in a freshman composition course using an experi- mental method improved their writing after fifteen weeks of treatment. The data on the posttests indicated that there was improvement in writing ability, but since no control group was used, the experimenter could not be sure if the improvement was a result of the experimental treatment or not. In another study, experimental and control groups were pre and posttested to determine the effects of teacher-corrected versus peer-corrected writing. But on the writing samples used as the posttest, rater reliability was so low that the writing samples had to be discarded. Scores on the objective posttest were not statistically significant, so the investigator "concluded" that there is no signifi- cant difference between the two methods of correcting student writing. Studies such as these are relatively insignificant or so egregiously flawed that they are hardly worth reporting. In reporting on better designed and more significant studies, Iivill consider both procedures and results. Readers who wish to look 'fin~ther into these studies can easily obtain them either as published documents, or through Universitngicrofilms, or ERIC Document 52 Reproduction Service. Finally, I will, wherever possible, point out areas related to the Braddock Report questions which need further research. 1. What kinds of situations and assignments at various levels of schooling stimulate a desire to write well? l4. What techniques of composition most effectively help build self-discipline and pride in clarity, origi- nality, and good form? These two questions deal both with attitudes toward writing and also with achievement. Because there is little distinction between "kinds of situations and assignments" and "techniques of composition" and because a "desire to write well" necessarily involves "pride in clarity, originality, and good form," I will consider these questions together. Hall, Moretz, and Storm (l976) studied home environments of children who were early writers in school in an attempt to identify what builds positive attitudes toward writing. From infbrmation collected through interviews with parents, the researchers found that most parents of early writers were college graduates who served as models for the children because they wrote in the home themselves; writing materials were easily accessible to the children, as were books, magazines, and newspapers; and parents often engaged in reading and often read to their children. The researchers identified three patterns in this early interest in writing: l) desire to Commmnicate to others through letters; 2) introduction to the names 0f letters and often direct instruction in writing; and 3) help given at the children's requests. 53 Many studies have been done on the effect of teacher criticism of student writing. In an overview of this research, Groff (1975) concludes that, contrary to papular belief, either positive or nega- tive criticism of children's writing gets the same results. That is, the quality of children's writing is not affected by either positive or negative criticism. However, the effect of such criticism on attitudes toward writing is another matter. Gee (1970), for example, worked with 139 eleventh grade students in an investigation of the effects of written comment on expository composition. Students were assigned to three treatment groups: one group received positive com- ments, one group received negative comments, and one group received no comments at all. All of the students wrote four compositions. Before each writing, the previous composition was returned, with appropriate comnents, or no conments at all. Measurement comparisons between the first and fourth compositions were based on the number of T-units and on quality as determined by a rating scale. While Gee found no significant differences in the quality of student writing, he did find that comnents of praise were more effective than negative comments or no comments at all in promoting positive attitudes toward writing. Stevens (1973), working with ninety-one low-performing urban male high school students for ten weeks, investigated what effect positive or negative evaluation has on the quality of writing and on the students' attitudes toward composition. Positive and negative evaluation groups were set up and students wrote five compositions during the study. Stevens found no difference in the quality of the compositions due to the effect of positive or negative evaluation, 54 but he did find that positive evaluation creates positive attitudes while negative evaluation creates negative attitudes. Two other studies deal with related kinds of feedback. Stiff (1967) investigated the effect of three correction methods on the writing of seventy-seven college freshman composition students. The three methods were: 1) marginal comments only; 2) terminal comments only; and 3) combined marginal and terminal comments. The results indicated improvement in all of the students' writing: there was no significant difference deriving from the method of correction. Stiff points out that this result would seem to indicate that the completely corrected paper and the amount of time invested in it may be no more productive than other procedures of correction which are less time consuming. However, Stiff also found that the students in the com- bined marginal/terminal correction group were more pleased with that method than the students in the other groups. The author thus sug- gests that, in the long run, this third method may have a positive effect on student morale and perhaps on performance. Sweet (1966) examined other forms of teacher feedback in his six-week study involving 225 ninth graders. The three methods he employed were: 1) gg_comment, only a numerical score and letter grade; 2) free comment (whatever comment the teacher felt like making; and 3) specified comment (stock reSponses designated in advance for each letter grade, such as A = "Excellent! Keep it up," or C = "Per- haps try to do better"). The three feedback methods were applied to students' objective tests, rather than actual writing samples which are usually used in experiments of this sort. But since Sweet's concern was with measuring the effects of feedback on performance 55 generally, his results are nonetheless suggestive and applicable to actual writing. Students in all three groups demonstrated little short-term effects on test performance due to treatment. However, students in the free comment group did show a signficant effect on scholastic performance over a longer period of time. In addition, only the students in the free comment group showed a positive change in attitude toward English. The researcher called for replication of his study, but for a longer time interval to test the long-term effects of feedback on attitudes. Two researchers comparing different approaches to composition included in their data information on attitudes toward writing. Adams (1971) compared the effectiveness of two methods used in an elective pre-college course. Method A was a highly structured approach which used professional essays as models, limited topics for writing, prescribed forms of discourse and length, mechanical and structural errors marked by a grader, brief comments directed at errors on themes, and required revisions. Method B was described as flexible: models derived from students' writings; no restrictions on t0pics, form, or length; small-group work where students read one another's writing before turning in revisions for the teacher to read; themes evaluated by reSponding to students' thoughts and ideas while mechani- cal and structural errors went unmarked; and long and affirmative comments on papers. The results derived from the STEP (Sequential Test of Educational Progress) test of writing skills, as well as from evaluation of writing samples showed that no significant differ- ences in writing skills existed between students from either group. There were significant differences, however, in attitudes toward n"" 56 the methods: both teachers and students in method B were more enthu-T siastic toward the end of the semester. This study represents two extremes in methodologies; and, though many uncontrolled variables in each method weaken the design, the attitudinal results do suggest that certain elements in method B may increase motivation in students at this level. In a similar experiment, Wahlberg (1970) explored a method of structuring the freshman composition classroom to affect student attitude and improve the learning climate. The control group followed a teacherbcentered lecture format; the experimental group followed a peer-interaction format with a college counselor intervening to show students ways to help one another. While the results showed mixed improvement for both groups, the students in the intervention group felt that more learning took place and that the instructor "cared" for them. A number of methodological experiments in pre-writing_tech- niques have been done. While I intend to report the bulk of these experiments under Question Eighteen (below), one study must be reported here because along with testing the effectiveness of pre- writing techniques, it also gathered data on attitudes. Rohman and Wlecke (1964) worked with students in a college-level sophomore eXpository writing course for one semester. The procedure followed in the experimental group was a six-week unit with the focus on con- cept fOrmation in the prewriting process (stress on the need for experience and thought before the actual writing). The control group fifllowed a traditional basic composition course format: formal study of grammar, logic, and rhetoric; analysis of model essays; and weekly 57 essays with revisions. Post-treatment themes were rated significantly higher in favor of the experimental group. Equally important, however, was the data gathered in the area of attitudes. Where the control group by the end of the course saw writing as an extrinsic problem (concern with mechanics, spelling, rules of good writing behavior), the experimental group saw writing as an intrinsic problem (concern with the value of ideas, thinking, and conceptualization). The authors concluded that as a result of their experiences in this course, the experimental students saw writing as a more worthwhile, more desirable activity. Other comparative methodological studies were concerned less with attitudes and more with achievement. Troyka (1973) investigated the effect of simulation-gaming (role-playing) on the expository prose competence of community college remedial composition students. Using the non-equivalent control group design (see pp. 28-29), the researcher set up an experimental group which was given simulation- based writing experiences focusing on basic rhetorical skills (development by facts, by reason, by incidents, and by comparison/ contrast). The control group was given similar assignments but did not experience simulation-gaming. The experimental group achieved significantly higher scores than the control group on two measures of‘ writing ability-~the STEP section on English Expression, and a ra‘ting scale used to evaluate the themes. Troyka concludes that the experimental treatment not only improved expository writing cuimetence but also proved to be an effective motivational strategy. Some studies compared writing labs/workshOps with more tradi- t1'OI'Ial approaches. Haas, Childers, Babbit, and Dylla (1972) used 58 142 freshman composition students for one semester to investigate the effectiveness of an experimental workshOp method which made use of intensive in-class guidance of daily writing assignments, peer-group problem solving of writing tasks, and condensed descriptions of rhetorical techniques. The control groups followed a format based on lectures on rhetorical strategies and discussions of readings from an anthology. In their results, the researchers found that the experi- mental workshop groups showed superior achievement over the control groups on writing samples rated for rhetorical technique, structure, mechanics, and content. In a similar experiment, Sutton and Arnold (1974) worked with 244 freshmen who scored on the lowest decile on the English scale of the ACT. The purpose of this study was to compare the long-term effects of a writing lab with those of a regular remedial English course upon the achievement and the attrition rate of the students. Students in the writing lab experienced much intensive tutoring while those in the regular course followed a lecture-discussion format. Both groups used programmed texts in spelling, diction, and writing. The researchers found that the writing lab students fared better in their other courses and that the individualized instruction of the writing lab methodology had a significant effect on the future writing grades of these students. Two other experiments with writing labs arrived at less Significant results. Turner (1970) worked with three sections of Jo nior college English to determine whether or not the substitution Of a writing lab for a regular class would improve student writing. We: control groups and one experimental group were set up with 59 evaluation based on judges' ranking of final compositions in the course. The results showed that the experimental writing lab group performed slightly better but not at a level of statistical signifi- cance. Dow (1973) had similar results with another group of college students. One hundred and forty-six students were divided into experimental and control groups. The experimental groups were assigned to a writing lab which was characterized by an informal atmosphere, attractive environment, non-compulsory assignments, non- mandatory attendance, ungraded writing, and extensive student-teacher conferences. The control groups followed a more structured procedure: a fOrmal classroom setting, reading and writing assignments, graded writing, grammar study, research papers, and examinations. Evaluation of both groups consisted of a writing skills test, a test of exposi- tion, and a writing sample evaluated by four raters using an evalua- tion sheet. Dow found that the students in both groups wrote equally as well. Closely related to experiments with writing labs are those which examine the effects of class size and/or individualized instruc- tion. Smith (1974) worked with high school juniors to investigate the hypothesis that the teaching of writing can be improved through individualized and small-group instruction. The researcher used twelve classes. Six were large-class control groups which received instruction directed to each group as a whole. Among the other six Classes were groups which also received instruction directed to the en tire group but with smaller class size, and groups which received in dividualized instruction. Smith found that the students in smaller cl.asses made greater gains in knowledge of writing skills and in 60 writing performance than those in the larger classes and that students of low and average achievement improved more than did students of high achievement. She also found that students in the individualized instruction groups made even greater improvement than those in small classes. An important part of the Smith study was a check on reten- tion of skills six weeks after the experiment: post-experimental testing showed no retention in knowledge of writing skills or in writing performance for students in large classes. Students in small classes showed retention in knowledge of writing skills but no reten- tion in writing performance. Students in the individualized instruc- tion groups showed retention both in knowledge of writing skills and also in writing performance six weeks after the experiment. Lagana (1972) examined an instructional method which employed individualization (diagnostic tests and teacher-student conferences) and peer grouping (students in each peer group chose writing topics, set objectives, and evaluated their writing). The control group operated on a whole-class basis, with the objectives set by the teacher, who also evaluated all writing. The subjects were sixty tenth grade students. Evaluation was based on the STEP writing test and the STEP essay test. Lagana found that "peer evaluation of compositions was at least as effective as teacher corrections and greatly reduced the need for out-of—class teacher time expended in evaluation. Peer evaluation also enabled sutdents to complete more COmpositions while receiving more immediate feedback on each Wr‘iting . . . students were able to progress at their own rate in ac-quiring certain composition skills without repetition of previous \e-arning" (p. 4063A). 61 In a related study, Ford (1973) investigated the effects of peer-editing and grading of themes on the grammar-usage and theme- writing ability of freshman English students. He found that having students edit and grade each others' themes can cause significantly greater gains in their gramnar-usage as well as in their theme- composition ability than having just the course instructor edit and grade the students' writing. Similarly, in an experiment using sixth graders, Sager (1973) investigated whether children who were taught to use a rating scale (composed of four sections on vocabulary, elaboration, organization, and structure) to rate their own compositions and those of their peers would improve the quality of their writing more than students who studied the four criteria of the rating scale but did not use it in evaluating their work. The researcher found that the students using the scale to rate their own work as well as that of their peers did improve the quality of their writing more than did the students who did not use the scale. Another kind of methodological experiment involved programmed instruction. Slay (1968) compared the effectiveness of programmed, formal, and informal approaches to the teaching of grammar in remedial college English. The programmed group used a programmed granular text; the formal group used formal gramnar instruction with a traditional handbook; and the informal group replaced formal grammar instruction with teacher-led discussions of students' writing, along with samples of student writing presented on an overhead projector. The researcher found no significant difference in writing skills among the three groups. Harris (1972) examined the learning 62 effectiveness and cost-time efficiency of programmed instruction for teaching expository writing to college freshmen and high school seniors. Programmed instruction included integrated instructional sequence, behavioral objectives and student knowledge of objectives, cybernetic feedback, and self-instruction. Harris found programmed instruction effective for teaching some high-level cognitive pro- cesses (analyzing informative discourse) and as effective as conven- tional methods for teaching the analysis and production of the kind of discourse which emphasizes logical proof. He also found programmed instruction efficient in terms of cost and time. In an extension of the Rohman and Wlecke study (1964, see above), Burhans (1968) added to the focus on prewriting a stress on writing techniques and structura1_methods. Three approaches to a college level sophomore composition course were compared. The "pre- writing" group emphasized prewriting and rewriting; the "comprehen- sive" group emphasized prewriting, writing techniques (e.g., abstract and concrete language, figurative language, analogy, and exemplifica- tion), and structural methods (development of paragraphs and essays), and rewriting; the "traditional" group emphasized logic, rhetoric, and mechanics. In addition, the "prewriting" and "comprehensive" groups were student-centered and developmental (i.e., from prewriting and writing stages to full essays) while the "traditional" group was material-centered and static (i.e., begin with full essays). Burhans found that students in the "prewriting" and "comprehensive" groups produced writing superior to that produced by students in the "tradi- tional" group. While students in the "prewriting" and "comprehensive" groups showed measurable gains in the areas of wording, flavor, ideas, 63 and organization, none of the three groups proved superior in the improvement of mechanics. Two experiments were concerned strictly with revision. Hansen (1971) investigated whether university students who do teacher-guided revision and rewriting of an essay achieve greater skill in composi- tion performance than students who correct mechanical and grammatical errors with only the aid of a handbook and who do not revise or rewrite. For the self-guided students, then, revision was strictly a matter of proof-reading. In addition, students in the self-guided group wrote more themes without revision, while the students in the teacher-guided group wrote fewer themes but revised each into new themes. The results led Hansen to conclude that there is no assurance that "a student who writes four themes and revises and rewrites each into a new theme will improve his composition skills any more than one who writes eight themes on eight different topics and makes a correc- tion sheet for each" (p. 1473A). The researcher also concluded that editing skills are evidently learned in some way other than through revising and rewriting. This study suggests that if students once understand just what needs to be revised, the actual physical act of revision may be unnecessary. In another study of revision, Effros (1973) worked with ten college freshman composition sections. The experimental groups' procedure was designed to motivate students to revise and rewrite by delaying grades until revisions were completed. The control groups, on the other hand, used minor revision with immediate grades. Results based on the English Expression Test showed that the control group 64 was significantly better, though there was no significant difference between the two groups on the essay test. In an experiment intended to examine creativity in the writing of tenth grade students, Jenks (1965) compared two methods. The first was the "DemOpraxis Journal Method," which consisted of regular journal keeping that included five components: 1) an ideas list; 2) daily writing with three weekly essays focussed on a single tOpic, mood, or opinion; 3) a personal manual with corrections of errors noted by the writer or by members of a peer group; 4) a spelling list; and 5) extra-credit manuscripts. The second method was a regular course of study where students wrote one assigned theme per week and did not keep a journal. Experimental data derived from the Imagina- tive Stories Tasks of the Minnesota Tests gf_Creative Thinking showed that the "Demopraxis Journal Method" contributed significantly to creative development. Since many of the studies I have reported here were conducted in rather short periods of time, evidence seldom indicates that any method being compared with another has any lasting effect. Many researchers report that significant differences might have become apparent had the treatment been carried out over a longer time. Snfith's (1964, see above) six-week post-experimental check is thus an exception worthy of replication. To cite another example, Burrus (1970) conducted a three year experiment with primary childen com- Paring two methods of teaching the mechanics of writing. The "traditional" method placed emphasis on a language textbook and models 0f correctness while the "functional" method emphasized the child's Own language (i.e., stress on mechanics as determined by voice 1 I 65 inflections) and emphasized writing as purposeful communication. Burrus found the functional approach statistically more significant in improving punctuation, capitalization, and spelling. Improvement in usage and sentence structure was slightly higher for the functional group, though not statistically significant. What is more important in this study, however, is the suggestion that attitudes formed in the functional group toward writing as a purposeful, real communica- tion act directed toward peers may have had an impact on the improve- ment of certain writing skills. The Burrus study has too many uncontrolled variables for this causal relationship to be drawn with any real certainty; still, it does suggest a relationship between attitudes and achievement acquired in time which other experimental methodological studies need to investigate. It is clear from many of the studies I have reported here that attitudes students form toward writing are not always considered in methodological comparisons. Indeed, investigations into the relationships between techniques of composition teaching and learning and attitudes which foster improvement in composition pose many ques- tions of a psycholinguistic nature which call for much needed research. What, for example, is the long-term effect on student Inotivation and performance of teacher-sponsored versus self-sponsored writing? Or how does having a genuine purpose for writing influence the development of writing abilities? And what is the impact of feedback from [£11 audiences as Opposed to teacher-only audiences on Student desire to communicate and communicate well. Finally, a research proposafl made by Rohman and Wlecke in their study seems most aPDrOpriate here: researchers should seek "to refine the precise 66 relation of the journal-~that is, the habit of private articulation--to the improvement of a student's attitude and performance as a writer" (Rohman and Wlecke, 1964, p. 108). These kinds of questions view writing as a meaning-centered language process where motivation and attitudes enter into the writing process in as essential a role as do the writing "skills" most studies focus their attention on. The reciprocal nature of attitudes and performance is thus in need of more research which recognizes that the development of attitudes and abilities in writing takes time and that composition methods and approaches are truly "effective" only when their impact on achievement and attitudes is apparent long after treatment. 2. What do different kinds of students prefer to write about when relieved of the expectations and requirements of teachers and others? Varon (1971) examined the content of unsolicited compositions written by fifth and sixth grade students in the years 1963-1968. She found that the major thematic category students preferred was abstract concepts such as 1911; and ha_t_e_. Other categories observed ( in rank order) were: nature, activities, material gggds, and M. Varon also found the greatest use of human referrents in the children's Writing was that of persons generally, followed by self, extra- Mal, world, familial, fanciful, and pg persons at all. Jobe ( 1974) found that when given freedom of choice in selecting topics, students chose fantgy, animals, and personality, in that order. 00be also found that the major influence on choice of topic was 67 internal (students' own ideas), followed by topics derived from personal experiences, and lastly books. Bell (1971) examined 1,502 compositions designed to encourage free expression of (the writing interests of high school students. He found that the interests most frequently expressed by the students (in rank order) were: education, m society, 'life, sports, home, egg world, people, experiences, and friendship. Bell also found that the students were more interested in writing about matters that they perceived as affecting their own lives and that they showed minimal interest in writing about such topics as animals, music, hobbies, travel , space, and literature. Standish's (1970) informal investigation of high school student writing preferences was reported in the April, 1970, Arizona En lish Bulletin: Iqterested in discovering what kinds of composition assignments h19h school students preferred, Patricia Standish (Alhambra HS, Phoenix) asked 256 students to complete a brief questionnaire. e response to item 1, "If you were going to be assigned a composition, which instructions would you prefer to follow?" revealed a preference for an unstructured assignment (an assign- "K-‘mt which left the student free in choice of topic, audience, approach, style, or length) by more than 40% of the students. About 29% preferred the loosely structured assignment, while less than 13% favored highly structured assignments. Item 2 asked students, "If you were going to be assigned a composition, which type of topic would you prefer?" and students indicated preference for topics based on current problems (50%), as Opposed to tapics based on literature (20%), experience (12%), Or the composition book (2%). Item 3 asked students, "If you Were going to be assigned a composition, what type of writing would you prefer?" Students responded to item 3 by indicating a preference for expository writing (36%) over narrative (22%) or descriptive (15%) writing. That these 256 students pre- ferred unstructured assignments is a little surprising. Many texts on writing note that structured topics give the young writer a sense of purpose and direction. (p. 51) P‘L A'zl‘qllwu; -' "I ~ 68 Future researchers might direct their attention toward the writing preferences of college students; content analyses and ques- tionnaire/inquiry techniques may provide interesting data for deter- mining the nature of preferred writing assignments. It may well be, for example, that part of the problem facing beginning writers stems from the conflict between preferred writing tapics and imposed topics. Furthermore, the wide spread of preference reported in such studies as Standish's suggests a need for research which investigates whether providi ng a variety of topics and modes on a writing assignment affects the writing performance of students. 3. What are the sources of fear and resentment of writing? I have considered this question separately from Questions One and Fourteen because it pertains to identifying sources pi: attitudes w writing generally, whereas those questions pertained more to W attitudes formed a; 3 result 9_f_ epecific techniques. Two eCENt studies investigated languaje epprehenswn as a p0551ble source of 1:ear and resentment. In a case study which examined the develop- ‘ne'lt of linguistic security and written fluency, Koch (1975) sought to facilitate linguistic security in college students through small- Show interaction. Pre and posttest comparisons revealed that students involved with small group interaction valued their competence "1 th language more, increased their written fluency, demonstrated 9"eizlter cohesion in their writing, and had greater confidence in thEir ability to speak and write effectively. Brazil (1975) found s‘imilar results; and, unlike the Koch study, he used a control group. Brazil hypothesized that the doctrine of linguistic correctness causes 69 linguistic insecurity. Working at the conmunity college level, Brazil evaluated the overall effectiveness of two contrasting approaches to teaching freshman composition: 1) a dialect-acceptance, student- centered approach; and 2) a language-standardization, teacher-centered students Results favored the dialect-acceptance approach: approach. made greater gains in fluency and overall writing effectiveness. A project by Daly and Miller (1975) came closer to identifying the sources of fear and resentment than did either the Koch or Brazil Daly and Miller reported on the initial development of an studies. The instrument for identifying apprehensive student writers. researchers developed a twenty-six item Lickert-type scale which was designed as an attitudes survey and was tested for validity and Peliabi lity. Students answering the survey are instructed to indicate degrees of agreement or disagreement to such statements as the 1“0110\«11‘ng: --I avoid writing. --I am afraid of writing essays when I know they will be evaluated. --My mind seems to 90 blank when I start to work on a composition. --I expect to do poorly in composition classes even before I enter them. --I have a terrible time organizing my ideas in a composition COUY‘SE. Further research in student writing apprehensions might expand the pioneering work of Daly and Miller to a more precise identifica- tion of fears and resentments. Longitudinal and case-study procedures rt'1th provide information regarding the natural history of reluctant 70 writers. In addition, inquiries into previous school experiences with writing could provide much valuable insight. Related research issues were suggested by Rohman and Wlecke in their study (1964); they encouraged researchers to: Seek to throw more light on the relationship of a person's self- image to his behavior as a writer. What might the validity of a self-image be as a predictor of successful behavior as writer? Seek to uncover to what degree our national "neurosis" about "correctness" has inhibited or encouraged better performance among student writers. What kinds of attitudes, especially in the elementary grades, would provoke better writing among young people? (p. 107) Final ‘ly, I would add that in recent years, an enormous amount of research in dialect variation has become available which can be of much value in attempts to discover the sources of fear and resentment 0f wr'i ting. Thus, research into the relationship of dialect variation and apprehension in writing shows much promise. 4. How do the kinds of writing which adults compose vary with their occupations and other factors? Very little has been done in this area. One study by VanFleet ( 1969) sought to develop guidelines for the content of a college FQDOrt-writing course by gathering data through the analysis of r‘eDOrt-writing textbooks and by interviewing report writers and report r‘eaders at two conpanies. The investigator found that the potential "epOrt writer should be able to do the following: 1) write correctly, Q0"cisely, clearly, and precisely; 2) define and analyze report problems; 3) outline, organize, and write reports of various lengths a"d degrees of complexity and formality; 4) develop particular report Se(:tions, such as introduction and sumnaries; 5) present data 71 graphically and verbally, and distinguish between relevant and irrele- vant details; and 6) follow instructions pertaining to report make-up, physical presentation, and graphic construction. Further research into other areas of specialized writing may be very useful to teachers and curriculun designers of specialized or technical writing programs. 5. What is the effect on writing of having the student compose his paper for different kinds of readers? Most of the research relating to this question deals with peer- audience as eye of the elements in a total methodological approach (see. for example, Lagana, 1972; and, Ford, 1973). A study by McClatchey and McClatchey (1970), however, seems to be in direct response to this question in that peer-audience was the variable in the project. After conducting a pilot study with twenty-nine students, the actual study was begun with fifty-nine freshman composition stu- dents at a university. Students in four classes each wrote four themes. Two of the themes were handed in to the teacher to be graded and cemented on. The third theme was dittoed and distributed in c13:35 groups, taken home and commented on by peers, discussed in class, and then returned to the writers. The fourth theme, exchanged t""“~een pairs of students rather than in groups, was commented on, diSCussed, and returned to the writers. Next, all of the themes were eVa‘luated by a group of outside raters on the basis of interest, egg- SEJEIIEELigp, and organizatipp. A letter grade was assigned for each theEllie. The results indicated that average and above-average students d‘i d better when writing for the teacher. Below-average students, ho‘Mever, did better when writing for their peers. In their 72 conclusions, the investigators suggested that below-average writers make low grades in writing partly because of tension over grades and inability to "psych out" the teacher's wishes. "In any case," they write, "it is apparent from the results of this research that most below-average students, and some who are average or above-average, would profit from writing, at least occasionally, themes that are directed towards their fellow students" (McClatchey and McClatchey, 1970, p. 23). McClatchey and McClatchey call for replication of their inves- tigati on into the impact of audience on student writing, and it would appear from the paucity of studies done in this area that there is still much need for such studies. In addition, researchers might consider investigating what effect audiences other than teachers and peers might have on student writing. _ For example, how is student Writing affected when the audience is the school or the comnunity at 1 3'98 or professional individuals and groups? While it is not uncommn for students at all levels to be engaged in writing for and'Aences of this sort, there has been virtually no research to esti- mate the effectiveness of such practice. 6. At which levels of maturation does it seem appropriate to introduce the various modes of discourse--narration, poetry, drama, exposition, argument, and criticism? Problems inhere in a proper interpretation of this question. For example, are we to take "appropriate" to mean a value judgement, as in "Do children at a certain level peed to be introduced to the e‘lements of drama, or to methods of exposition?" Or are we to ‘i r”lerpret it in a developmental sense, as in "_(_Z_a_i_i_ children at a 73 certain level learn techniques of cri ti cism?" ApprOpriateness, in either sense, has not been dealt with in the research. Another problem lies with the phrase "to introduce." Does this mean _f_g_r_‘_n_i_a_l_ introduction to the modes of discourse? If so, then the reader can look to the methodological studies which involve introducing one of the modes of discourse that I summarized earlier in this chapter. But if "to introduce" means informal introduction, then the question is even more problematic, since children and adolescents of all ages are exposed to and use in their daily language encounters all of the modes of discourse--they narrate, dramatize, argue, criticize, and so forth- The question, then, is too ambiguous for any precise grouping 01’ research studies under it. It calls for a kind of broad develop- menta‘l research which has not been done. Researchers may be better off Pursuing the kind of related issue posed by Sara W. Lundsteen in $.12 for _tfle Teacher 9: Written Composition when she suggests that we " 1 00k to what children 933193 before we talk about what teachers should Q2" (Lundsteen. 1976. p. 17)- 7. What is the relative effectiveness of writing shorter and longer papers at various levels of maturity and proficiency? Researchers have not investigated the effectiveness of compo- 3713? on length either at levels of maturity or at levels of proficiency. A r‘elated matter, however--writing frequenexuhas been looked into. In " Recent Studies of Writing Frequency," which appeared in the first "Huber of Research _i_r_i_1;_11e Teaching pi English (Spring, 1967), Robert H'~"Iting reviews five studies which consider whether increased ””1 ting practice improves writing. He concludes that mere frequency l]~..l. ~o.‘l“o‘u-" .. I' ~ - 74 of writing without accompanying instruction or motivation will not improve writing. Hunting calls for more research in this area, particularly investigations into the relationships between improvement and functional writing assignments (i.e., writing that is meaningful and challenging, as opposed to writing that is merely practice). Sherwin (1969) has a more comprehensive sunmary of research into the benefits of writing practice. From his overview of this research, he concludes that "merely increasing the nunber of assign- ments will not improve the quality of writing" (p. 157). The implica- tion drawn from research on writing freguency seems to suggest that increasing the lqutl of compositions will not improve the quality, though at present we have no research to support or disprove this asserti on. 8- At which levels of maturation does it seem appropriate to introduce the various rhetorical elements of writing? This question poses the same problems as those I discussed under Question Six: what does "appropriate" mean? Does "to intro- due-e" mean formal introduction; or, in this case of rhetorical e‘Ieflients, does it mean raising to consciousness that which people do natural 1y, such as coordination, subordination, transition, etc.? As I Said in my discussion of Question Six, the question is too ambiguous for any precise grouping of research studiesunder it. However, the reader can look to some of the studies cited under other questions here which deal with the introduction of rhetorical elements to students, such as Troyka (1973) and Burhans (1968) under Question One, and Fichtenau (1968) and Gozenba (1975) under Question Eighteen. 75 9. What are the effects of various kinds and amounts of reading on the quality and kinds of writing a person does? There have been many investigations into the relationships between reading and writing. Lacampagne (1969), in his examination 01’ approaches and attitudes toward writing, surveyed over 1,000 twel Fth graders who had been rated either superior or average in writi :19 performance. Among his findings were some correlations betNeen extensive reading background and superior writing performance. Slnri‘liarly, Donelson (1967), in his investigation of 251 tenth graders, ’73llrici that effective writers were widely read and owned more books 't"iif\ ineffective writers. Maloney (1967) tried to identify superior and poor ninth grade writers of expository prose and the qualities that were characteristic of the superior writers. The researcher f0 und that superior writers came from homes where parents bought books r‘egularly and that the students read often and scored high on reading 1:33 ts. Barbig (1968), in a similar study with ninth and twelfth 9"‘aciers, found that the poor writer watched more television and read Fewer books than did more successful writers. Nakamura (1970) inves- ‘i Qated the relationship between the amount of reading and the qua‘l ity of writing done by thirty tenth grade boys. As might be prected, he found that the writers who read more wrote better. St Udents considered good or fair writers owned more books, read a 35""‘Eeaater percentage of the books owned, and were assigned and completed fibre outside reading in school than did the poorer writers. In at'Cl'ition, Nakamura found a close relationship between availability Q T: magazines and newspapers in the home, and the students' ability I" .n’ widen-’45.- w I 4 ,. | 76 to write well. Schneider's study (1970) was an attempt to locate Specific correlations between reading and writing skills. She inves- tigated whether emphasis on reading skills leads to improved writing in a college remedial writing course. Both experimental and control groups followed the same conventional classroom method, except that the experimental group was taught developmental reading in addition to the writing activities. Results were mixed, but in favor of the students in the experimental group: they gained on three posttest meas ures of writing and reading abilities, though only two measures were statistically significant. The author concluded that emphasis on reading skills can lead to improved writing. It would appear from the research that a close connection between reading and writing does exist. None of the studies cited here . however, attempt to articulate the causes of this relationship. A] 1 we can be sure of at this point is that extensive reading con- th‘i butes to success in writing. Why this is so is a rich area for F“ "ther research. 1 0. What are the direct and indirect effects of particular sensory experiences and guided observation upon writing? Ewing (1967) investigated the effect of various stimuli on the Writing produced by third graders. Four sensory stimuli were used: 1) auditory (listening to a musical selection); 2) _v_i_s_u_a_l_ (vi Ewing a film without words); 3) _mgtgp (drawing a picture); and 4) wimal stimulus (being asked to write a story). The students wrote a composition after each of the four stimuli. Five judges ranked the compositions according to overall quality. The 7 77 compositions judged highest in quality were those written with minimal stimulus, followed by those written under auditory, ”visual, and _n_i_9_t_o_r; stimuli. King (1973) sought to determine whether increasing the number of types of sensory stimulation prior to a writing experience would help fourth, sixth, and eighth grade students to write more creatively and to write longer stories. The four stimuli were: 1) aural; 2) _a_1_1_r_'_a_l_ and y_i_s_uayl_; 3) aural, visual, and tactile; and 4) aural, visual, tactile, olfactory, and gustatory. The results were inconclusive. Two related studies focussed on sensory stimuli. Golub and FY‘edrick (1970) compared the differences in the writing of 160 fourth and sixth graders when they were exposed to two variables: 1) _c_o_n_- m visual stimuli (black and white versus colored pictures; and, W versus concrete pictures); and 2) varying instructions gt h°\\~l j_:_9_write (epecific versus general). The researchers found that '31 ack and white pictures seem to be slightly superior to colored pi Ctures for generating more complex and more diverse linguistic st"‘uctures. They also found that abstract pictures were more diffi- CHT t to write about than concrete pictures, which produced more mo(lifying clauses. However, there was no statistical significance for either of these findings; and the variation in instructions had no effect on the students' writing. Donlan (1976) worked with Q] eventh and twelfth graders to determine the effect of four types 01: music on the students' spontaneous writing. He found that wamiliar vocal music interfered more with the quantity and quality 91’" student writing than did familiar vocal music. 78 Finally, Kafka (1971) investigated the effectiveness of three sensory stimuli in helping intermediate students express themselves in writing narrative compositions. He found that students exposed to the- three stimuli--visuel, aural, and tactile--before writing, did not demonstrate better quality in their writing than a control group whi ch wrote without being exposed to the stimuli. In fact, the con- tro‘l group wrote better compositions. Kafka's study, like the others ci ted here, point to the general inconclusiveness among studies done on sensory stimuli. Too much depends on other variables, such as topic, environment, mode of discourse, and, most importantly, indi- V‘i dual student sensitivity to stimuli. Kafka's suggestion that per- haps children write more effectively from internal stimuli than from \external stimuli supports the notion that sensitivity to stimuli of "ha tever sort is far too individualistic for researchers to come to any firm agreement about the relative effectiveness of one stimuli Over another. 11. At what stages of maturity do students spontaneously seek specific help in improving particular aspects of writing, such as specificity of details, transitions, parallel structure, and metaphor? This question presupposes that there is 3 stage or stages of maturity at which students develop such highly motivated self- cl‘irection that they spontaneously grapple with these problem-solving QSpects of writing, when in fact students at a_l_l_ levels can acquire Such self-direction. The essential question, then, has less to do With identifying some level of maturity and more to do with yi_hy and M a self-generated search for help develops. The research studies 79 I cited under Question 0ne--those dealing with attitudes--are the closest we can come to answering this question through research. Hence, there is still need for research in response to this question, though I believe it would be more accurate to rephrase it as follows: "Under what circumstances, environments, approaches, motivational stimuli, etc., gg students at different levels of maturity develop a A sel f-directed problem-solving orientation toward writing?" 1 2. At which levels of maturity ggp_particular aspects of writing most efficiently be learned? This question is similar to Questions Six and Eight in that the many ways of interpreting "aspects" prevent any specific grouping 01: research studies here. Perhaps the best answer research could pro- Vide in response to this question would be that some students, at Various levels of maturation and under various learning conditions, C\aQ learn some aspects of writing. That is to say, the question poses a broad develOpmental issue that so far has been approached only in studies of isolated aspects of writing; and the results of these studies do not lend themselves to the sweeping conclusions this ques- t‘i on seeks. Nonetheless, valuable insight 593151 derive from research which examines developmental aspects of children's and adolescents' whiting. As Lundsteen (1976) points out, there has been some theoretical work concerning "characteristics of children's composi- t‘i on according to increases in age. For example, there appear to be progressions in plot construction, characterization, choice of revealing detail, sequencing, support of main ideas, ability to make choices in forming and arranging sentences, coordination, 80 subordination, and use of transitions (Burrows, 1960; Hunt, 1965). The conpositional thought of children moves from memory of direct, sensory experience to pictured images of concrete objects held in inner speech thought (Vygotsky, 1962). The child's written thought moves from a few words to whole incidents and finally to the complex ordering of experience through various forms of literature, such as the folktale, fable, myth, and fantasy (Nebraska Curriculum Develop- ment Center, 1966)" (Lundsteen, 1976, p. 24). Experimental support for hypotheses such as these may provide some answers to when students can learn what. 13. Does the oral reading of rough drafts help the elementary child strengthen "sentence sense"? How does it? Mills (1970) compared the effects of oral proofreading and Si 1 ent proofreading of rough drafts of compositions by twenty-six 3‘5 xth graders. Students wrote pairs of narratives; one was read Orally for correcting errors in capitalization and terminal punctua- t‘i on, and the other was read silently for the same corrections. Mills 1=0und a significant difference in correcting capitalization errors which favored the oral proofreading method, but found no significant d‘i fference between the two methods in the correcting of terminal punctuation. The researcher concluded that children of this age can t>enefit from both types of proofreading. Further research in this alr‘ea should replicate the Mills study at other levels of schooling. 15. What procedures of teaching and learning composition are most effective for pupils of low socioeconomic patterns? 81 There has been little research in composition specifically directed towards students of low socioeconomic patterns. 'An exception is Fry (1971), who investigated the effects of two variables upon the wri ti ng of 160 ninth grade students of low socioeconomic backgrounds. The two variables were: 1) traditional versus transformational grammar; and 2) direct versus indirect methods of teaching. (The m method was concerned with the correction of specific errors While the indirect method was concerned with generalizing the gramnati- cal el ements without emphasizing Specific errors). Fry found that "Either the proportion of well-formed sentences nor the average - Still‘ll'tltural complexity of sentences was affected by either grammatical appr‘oach, or by either method of teaching. Unlike Fry, however, most resea r‘chers are more concerned with their subjects' achievement level than with their socioeconomic level. That being the case, useful r“fizearch could be done first, on what causal relationships, if any, exi 31: between socioeconomic background and achievement; and, second. 0" strategies for the teaching and learning of composition which are ‘i "fohned by such research in causal relationships. 1 6. What procedures of teaching and learning composition are most effective for pupils learning to write English as a second language? Much of the research in ESL in recent years has been concerned with practices based on comparative learning theories, such as those described by John Carroll as the audiolingual habit theory, and the %&i tive code-learnipg theoLy (Carroll, 1971). However, very little of this research consists of empirical studies. Morrisroe and Morrisroe (1972), in their survey of 239 articles published between 82 1961 and 1968 which deal with ESL generally, found only seventeen that could be described as empirical research. "Many articles" they note, "deal t with problems in second language teaching, but few dealt with proven ways to solve them" (p. 50). The situation is even worse for research in .ESL directly concerned with the teaching and learning of composition. For example, Dykstra and Paulston (1967) reported on a Programmed method of improving composition skills of foreign students which involved structured language manipulations of model passages. An experiment is included in the report, but it is not described in any detail, and no statistical results accompany it. Much research could thus be done in the area of composition for Students learning English as a second language. Studies such as the One conducted by Friend (1970) could be replicated. Friend exam‘i ned relevant theories and research in linguistics, psychology, and composition theory as they relate to the construction of writing prog rams for students of English as a second language at the inter- med‘i ate and advanced levels. She then presented a writing program based on such information. While Friend's is not an empirical study, it ‘i s a sound example of the kind of investigation that could be hep] icated on an empirical basis. Readers interested in this area or research in ESL should consult Friend's bibliography as well as the bibliographies appearing in Carroll (1966), Croft (1970), the 79\68_yflex__tp_ ERIC Documents _i_n_ Linguistics and the Uncommofly Taught Weep Selected Bibliographies 91 Related Titles (1969), A m Bibliography (1971), and studies indexed in Language and L\anguage Behavior Abstracts. lg‘ .Kl’h £11“ 71. I. 83 17. Can study of the newer types of linguistics help writers? Research into the relationship of modern linguistics and wri ti ng have taken tWo directions. On the one hand, some studies examine whether instruction in linguistics improves writing. Such studies, thoroughly sunmarized in Sherwin (1969) and Blount (1973), do not, as Sherwin concludes "encourage the belief that a linguistic i approach or linguistic knowledge is more effective than a granmatical approach or gramatical knowledge" (p. 156). Sherwin's statement echoes that made six years earlier in the Braddock Report on research in the teaching of traditional granmar: "the teaching of formal granmar has a negligible or, because it usually displaces some instruc- tion and practice in actual composition, even a harmful effect on the 1mpt‘ovement of wri ti ng" (1963, pp. 37-38). The need for any further Ir'eSiearch in this area is unlikely. However, a second direction has emerged in investigations into the relationship of modern linguistics and writing improvement, "aNEl y, studies in sentence combinipg. The major documents here 10¢] ude the Bateman and Zidonis study (1964), which found that the study of generative granmar can improve the effective formation of sentences and increase the complexity of sentences written by ninth and tenth graders. Miller and Ney (1968) worked with fourth graders 170" an entire year, using oral and written drills in sentence °°"bi ning, along with choral readings. Posttest results favored the exDerimental oral/written drill group. These students wrote more words per writing assignment, used the sentence structure practiced 1“ the treatment more, and used a greater proportion of complex 84 sentences than did the control group. Mellon (1969) found that the syntactic fluency of ninth graders can be enhanced through the study of transformational-generative gramnar along with the exercises in sentence con'bining, though it was not clear which affected student writing--the granmar instruction or the sentence conbining exercises. O'Hare (1971) replicated the Mellon study but did not include any instruction in transformational-generative granmar. His work with 300 seventh graders for an entire school year focussed on intensive Practice in sentence conbining. At the end of the treatment he found . that the students in the experimental groups were writing sentences more syntactically mature than the sentences produced by students in the control groups and that the overall quality of the experimental group's compositions was also superior to the control group's composi- tions. Other studies which replicate the sentence combining experi- ments (Oberchain, 1971; Fisher, 1973; Ofsa, 1974; Bivens and Edwards, 1974; and Coirbs, 1975) come to similar conclusions--that sentence “Whining practice improves the syntactic maturity of students in the experiments. Some researchers do not agree with these conclusions (Green, 1972), while still others take the bulk of these findings "1 th considerable skepticism (Marzano, 1976). Nonetheless, enough 9"7 dence has been gathered from enough research to lend much support f0“ the effectiveness of sentence conbining activities. Stotsky (1975), for example, in her comprehensive overview of experiments in santence combining, concludes that these activities will promote silillliactic maturity and will improve the overall quality of student writi ng. 85 18. Can formal study of rhetorical theory or of logic help writers? Fichtenau (1968) examined the growth in written composition of academically above-average children in grades three through six who were taught the concepts of invention, arrangement, and _s_i_:y_1__e__. The researcher found that the only significant difference in writing g, skills at posttest occurred at the third grade level. He concluded that there is little relationship between teaching these selected rhetorical concepts and the improvements of written composition at . these grade levels. Gozemba (1975) asked whether rhetorical training through y_i_s_u_a_l_ media (slide-tape programs, films, and photographs) would be more effective than rhetorical training through yeppel means a10ne in improving four writing skills of college freshmen: 1) ability to clearly state a thesis; 2) ability to carefully state an argument; 3) ability to deliberately substantiate the argument with examples; and 4) ability to skillfully express ideas with varied sentelnce structure. The researcher found that training through y_i_s_ge_l_l_ "'Qd‘ia was extremely effective: the gains of the experimental group i" all four writing skills were nearly double those of the control 9"°UD. Some researchers have focussed strictly on m. Rohman and Nlecke (1964), as I reported under Questions One and Fourteen f°UDd that college saphomores in a treatment group which emphasized °°"|<:ept formation in the prewriting stage wrote significantly better themes than did students in the control group which did not focus on Praflriting concept formation activities. Odell (1974) used freshman composition students for one semester in an experiment emphasizing 86 mic discovery procedures developed by Kenneth Pike. No control group was used; rather, the researcher predicted the changes that would appear in students' writing from pretest to posttest and deter- mined how likely it was that these changes could be attributed to chance. Because of the nature of this design, the results must be taken as tentative; posttest analysis of essays revealed that students were in fact using at least some of the operations that they were taught in the course. I should also note that Hoyer (.1974) has com- Pl1ed a useful annotated bibliography on the invention process in i couposition and on the act of creativity. She presents four sections on the following: 1) general works on invention; 2) taxonomic heuristics; 3) discovery through persona; and 4) multi-observational aPPV‘Oaches. Each section is divided into subsections on theory. Praetice, and research. Three studies considered the effectiveness of generative rhetoric in improving writing. Hardaway (1969) investigated whether generative rhetoric is more effective than traditional rhetoric in 1mlilt‘oving the writing skills of college freshmen. The experimental group received instruction in generative rhetoric of the sentence and paragraph, read from models, and did exercises. The control group analyzed sentences and paragraphs by focussing on loose, b“ anced, and periodic styles; types of sentences (simple, compound, Etc~ ); and topic sentence, unity, coherence, and emphasis in the pa"agraph. Hardaway found no significant differences between the “'0 groups, though mean scores for the experimental group were slightly higher in the areas of focus and structure, content, sentence construction, fluency, and general impression. Miller (1972). 87 in his experiment involving college students, investigated what effects the Christensen Rhetoric Program has upon student attitudes toward conposition and upon the use of free modifiers in their writing after a lapse of time from instruction. He found that the program did not affect attitudes toward composition, but he did find the program superior to traditional methods in helping students to expand ideas 1" in sentences and paragraphs and to continue to do so after leaving instruction. Similarly, Hazen (1972) compared the effectiveness of . the Qristensen Rhetoric Program with a traditional write-revise aPpr‘oach at the community college level. Ten writing skills were the h.“ criteria for improvement: organization, ideas, development, tone, Style, reasoning, sentence structure, usage, punctuation, and SW” ing. Hazen found positive results and concluded that the Christensen Rhetoric Program will promote writing skills at this level superior to the skills of students taught by the write-revise aPPPoach. In a related study, Sanders (1973), working with junior c3°1'lege freshman composition students, compared James Kinneavy's "aims" approach (which stresses expressive, literary, persuasive, exD'l oratory, scientific, and informative aims which govern the “01 ces writers make in the process of writing) with a traditional "modes" approach (which stresses techniques relevant to the various “”des of exposition). Though both groups improved their writing. sanClers found no significant differences between them resulting from (fit-her approach. Finally, Klein and Grover (1970) investigated whether instruc- tion in symbolic logic would effect improvement in composition and 88 logical sentence analysis for students in grades nine through twelve. The researcher found that instruction in logic has a significant effect on sentence logic analysis but does not contribute to improve- ment in students' essay writing skills. A valuable suggestion for further research in this area was proposed by Braddock when he stated that "It would be interesting .-. for someone to do a critical synthesis, 'What Research in Reading A Suggests to Writers,‘ which may get at the effect of rhetorical con- siderations on various eypes g}: readers, not merely on composition teachers or raters--the usual yardstick for this kind of research" (Braddock, 1969, p. 451). 19. How is writing affected by extensive study and imitation or parody of models? - Pinkham (1968) emphasized the characteristics of "good wl‘iting" in an experiment involving 180 fifth grade students from u"‘ban and suburban areas for a fourteen week period. Students in the expe rimental group followed procedures based on stressing the charac- tel"‘istics of "good writing" found in selections from children's literature, along with actual writing and revision. The control group also wrote and revised, but did not receive emphasis on the mdel characteristics. Pinkham's results indicated a significant ‘11 1’ference in favor of the experimental group on the STEP writing tes 12. On the STEP essay test, however, Pinkham found no significant 9'1 1"ference between the groups, though there was a positive gain for StUdents in the urban area. Calhoun (1971) investigated the effect 01’ analysis of essays on reading and writing abilities of college 89 ccumuaosition students. Sixty-four students in the experimental groups analyzed essays through a series of ten lessons geared toward arti- lating the rhetorical techniques used in the readings. Fifty-eight students in the control groups had no such systematic instruction in araas‘lysis, though all other elements of instruction were the same for both groups. The researcher found that systematic analysis of ‘rileeitorical techniques contributes to an increased awareness of those techniques when they are encountered in reading. But no evidence was found to indicate any transfer of this awareness to writing; that is, there were no significant gains for either group on the compositions rated as posttests. An interesting study related to essay analysis was conducted by Stewart (1966). From a group of 77 anthologies used in freshman connF><3sition courses across the country, he analyzed the underlying rat-i onales of the anthologies and the rationales of directors of fres hman composition programs who use these readers. He then cata- 1Milled the rationales and compiled the following list of those most 1:“eqldently underlying the texts and their uses: --texts offer advice for the beginning writer; --texts offer the study of language as the proper content of a composition course. «texts stimulate interest in tapics for writing; «texts offer prose models; I --texts offer critical reading and thinking which lead to better writing. Of the five rationales, the last two were the most popular. Stewart challenged the imitation _o_f_ models rationale by citing the 90 difference between analysis (the picking-apart of a reading selection) and synthesis (the putting-together act of writing) and then asked what goes on in the student's head that allows him to make the trans- ference from analysis to synthesis. He saw, finally, three functions the texts serve: 1) they add to a liberal education; 2) they expose students to good writing; and 3) they are a useful introduction to methods of literary criticism. Since the use of models in the teaching of writing is such a wideSpread practice, it is surprising that more basic research has not been done in this area. Stewart's challenge to the imitation of made‘l s rationale raises fundamental issues which researchers ought to look into: l) what is the process whereby the analysis of reading 581ections influences the production of writing?; and 2) to what extent is the imitation theory compatible with research into language r'01::essin ? 20. What forms of discourse have the greatest effect on other types of writing? For example, does writing poetry help a writer of reports? Only one study examined the transfer potential between forms 0f discourse. Shapiro and Shapiro (l97l) investigated the suggestion that student improvement in writing poetry would result in improve- ment in writing prose and in improvement in student attitude toward H terature generally. The researchers used 82 fourth graders in metropolitan schools for six weeks. Procedures followed in the exPerimental group consisted of activities related to the study of Poetry through studying poems, listening to poems. and writing poems. students in the control group used the Roberts Series along with the 91 same number of writing Opportunities as the students in the experi- mental group. Evaluation of post-treatment writing samples was based on a rating scale which assessed: 1) unity of thought; 2) organiza- ti on and fluency; 3) opening and closing sentences; 4) originality and imagination; and 5) emotional appeal. Results favored the experimental group on poetry writing, prose writing, and attitudes toward literature. Among their conclusions, the researchers felt that the results favored the experimental students because of the freedom and scape poetry provides for linguistic expression and because of the provision of an alternative mode for self-expression through Tan 9 uage. The Shapiros' conclusion echoes one of the implications Burh ans drew from his experiment on the college level: "Interestingly enough, students who do well even in the conventional composition ngram have quite often had high school courses in 'creative' "P1 ting. Our study indicates that they do well not because training and experiences in writing fiction and poetry help them specifically "i th the problems of exposition but because in 'creative' writing COUrses they have become personally and subjectively involved in and “minitted to the processes of writing itself, and this involvement Carries over into every other kind of writing they do" (Burhans, 1968, p. 37). Burhans' statement suggests at least two possibilities f0!“ further research in this regard: l) how does personal involvement affect the written product?; and 2) what elements of the process of W"'iting in one form of discourse transfer to writing in another form? 2l. What is involved in the act of writing? 92 22. How does a person go about starting a paper? What questions must he answer for himself? A number of researchers have attempted to characterize ele- Emig (l97l) used a case study method Students ments in the composing process. tc» examine the composing processes of eight twelfth graders. composed themes aloud and provided autobiographies of their writing experiences. From her observations and from the data collected from true: writers, Emig constructed an outline of the composing process a] ong with a narrative account of the steps in that process. She found that the students engaged in two modes of composing: first, tire: reflexive, characterized by l) focus on the writer's thoughts arlci 'feelings; 2) sense of a self-directed audience; 3) affective exp1 oration; and 4) a personal approach. Second, the extensive, marked by l) focus on an other-directed conlnunicable message; 2) cog- ni t‘i ye exploration; and 3) an impersonal, reportorial approach. Emig fk’lllwd that the composing process for these two modes is further C“ a racte ri zed by processes of different lengths with different clustering of components. For the twelfth graders in this sample extensive writing occurs chiefly as a school-sponsored activity. Reflex- ive writing is a longer process with more elements and compon- ents than writing in the extensive mode. Reflexive writing has a far longer prewriting period; starting, stapping, and contemplating the product are more discernible moments; and reformulation occurs more frequently. Reflexive writing occurs often as poetry; the engagement with the field of discourse is at once committed and exploratory. The self is the chief audience--or, occasionally, a trusted peer. Extensive writing occurs chiefly as prose; the attitude toward the field of discourse is often detached and repor- torial. Adult others, notably teachers, are the chief audience for extensive writing. (p. 9l) 93 In his profile of the composing process of a twelfth grader, Mischel (1974) also collected data through observations of the student composing and through interviews with the student about his writing and about the composing process. The writing was essentially narra- ti ve and was done in forty-five minute sessions, though no specific Mischel found that the student started by think- time limit was set. Then he just began ing out what he wanted to say before saying it. writing and proceeded in a linear manner. There was very little prewriting activity, nor was there any planning on paper--all was mental. During the physical act of composing, the observer noticed little more than the student verbalizing his thought, then writing it The student down, while occasionally hesitating over a word or phrase. pai d little attention to correcting mechanical errors; his focus was 0" meaning and plain expression, though he did do some rereading and rev-i sing later. Graves (l975) examined the writing processes of seven year-old chi 1 dren. He too used a case study method based on analysis of Chi 1 dren's writing, interviews with the children on their views of the ‘i r own writing, as well as interviews with other children on their concepts of a good writer. Other sources of information included testing, interviews with parents, and observations of the children in several environments. Graves' findings in regard to learning envi ronments present significant implications for classroom practice: l) Informal environments give greater choice to children. When children are given choice as to whether they write or not as to what to write [sic], they write more and in greater length than when specific writing assignments are given. nulvhg !. .1! LV' ‘fi. 94 2) Results of writing done in the informal environments demon— strate that children do not need motivation or supervision in order to write. 3) The formal environments seem to be more favorable to girls in that they write more, and to greater length, than do boys whether the writing is assigned or unassigned. 4) The informal environments seem to favor boys in that they write more than girls in assigned or unassigned work. 5) In either environment, formal or informal, unassigned .. ’ writing is longer than assigned writing. (5) An environment that requires large amounts of assigned writing inhibits the range, content, and amount of writing done by children. 7') The writing developmental level of the child is the best predictor of writing process behaviors and therefore transcends the importance of environment, materials and methodologies in influence on children's writing (p. 235). Sawkins (197l) investigated the procedures 60 fifth graders used when writing narrative themes. The students were interviewed after writing two conpositions. Among her conclusions, Sawkins found that better writers are more concerned with the content of their wri‘t:1’ng (ideas, organization) than poorer writers, who are more con- ce‘”fieed with the mechanics of writing (spelling, punctuation, capitali- zat-i o"). In related research into the composing process, Ney (l974) deve] oped a model of the sentence combining operation in an effort t0 EXplain its effectiveness. Basically, Ney sees the mental opera- t1‘3'153 of the sentence combining activity as one of raising to a 5“ f‘--<:onscious level of control "linguistic resources which are "mate to the students" (p. 168). Once these resources are on a c°“S<:ious level, the student can use them in his written performance. 95 Finally, Cooper and Odell (1976) investigated whether profes- sional writers attend to the sound of their writing during the com- posing process. Eight subjects were used in this study--two university teachers and scholars, two columnists, two news writers, and two technical writers. The researchers found that the M of these . subj ects' writing does not play a very significant role in their composing processes. Their main concerns (in rank order) were: l) enabling their readers to understand with ease; 2) clear expression 0f their ideas; 3) appropriate style; and 4) the _s_9_u_n_d_ effect they imag'i ned their writing would have on their audience. Conventional matters of correctness mattered little. Examination of the composing processes of writers at all leve1 s is a rich area for further research. Graves (1975) remarks that "future research in writing should continue to explore the feasi bility of the case study method." He notes, in addition, that "Further studies are needed to investigate the developmental histories 0f (11' fferent types of children in relation to writing and the writing pmcess"(p. 24l). Researchers might also use a model of the writing Pf‘Ocess such as Emig's (197l), which locates specific components in the process, in order to measure how variations in those components (e. 9 - . assigned versus unassigned topics) affect the written product. 23. How does a writer generate sentences? The vast amount of work done by linguists over the past two de"Tades precludes any attempton my part to identify the multitude o f studies relating to this question. The interested reader need 0 My consult the enormous body of research by 96 transfonnational-generative linguists on the concepts of competence and performance, surface and deep structure, and child language acquisition, to find a wealth of theoretical and empirical responses to this question. 24. Of what does skill in writing really consist? The studies of the composing process which I presented above are attempts to make the kind of discoveries needed to provide some answers to this question. In addition to these studies are two by researchers who have attempted to identify specific "skills" employed by s uccessful writers in the act of composing. Stallard (l972) exami ned the writing behavior of good student writers from a high schoo “I senior class. His data were based on observations made on Students writing an expository essay under laboratory conditions. He. 1EOLJnd that good writers write slowly, take time to read segments 0f th eir work at intervals during the writing process--and read the fina‘l paper and revise it. They do not consider identifying a parti- cular audience for their writing, nor do they demonstrate concern for D1 anning the structure of paragraphs or the structure of the enti he essay. He must keep in mind that Stallard's experiment was under laboratory conditions; hence the processes demonstrated may not be characteristic of less artificial situations. In a related study, Hooks (1972) sought to identify what elements of writing are con- sideY‘ed most essential by professional writers. She collected data fm'“ written documents of Hemingway, Faulkner, Fitzgerald, and Thomas Wolfe and from criteria professional book reviewers use in t . . he‘ V‘ evaluation procedures. She found that the elements of effective 97 writing include: l) the view of composition is that of a total process; 2) the origin of ideas lies in the writer's background and personal experience; 3) the purpose of writing is to conmunicate an idea to an audience; 4) the notion of audience determines language and style; 5) reading others' works and constant writing will develop style; and 6) revision is necessary for succinct presentation of ideas. Further research into the "skills" or elements involved in wri ting might consider the question proposed by Lundsteen: "Would a child who has insight into the writing process do better in the 1orig run? Would a longitudinal study show that ability to discuss the writing process is reflected in the quality of the writing, after all? Would the kind of writing involved make a difference in the "e‘ationship between quality of product and ability to discuss the Process?" (Lundsteen, l976, p. 57). It should be clear by now that a wealth of research in compo- 51111 on is available to composition teachers, to directors of composi- ‘11 On programs, as well as to other researchers. It should also be C] Ear that there is much to be examined in the teaching and learning of Composition at all levels, that opportunities for much needed r'QSQarch are plentiful. In this chapter, I have tried to point out a number of areas where research has been done, as well as areas where much valuable research can be done. In the next two chapters, I Will consider usigg research by focussing on how the findings of r“Search in composition can be implemented in the design of composi- ti°n programs. “" in III—l. ll CHAPTER III USING RESEARCH IN COMPOSITION: RATIONALE In the Arizona English Bulletin of April, 1970, Kenneth L. Donel son precisely located an area of professional ignorance among Eng‘l ish teachers at all levels when he asked, "why is it that an Eng'l i sh teacher will modestly brag about his willingness to spend time on a study of literary criticism (whether it be Frye or Brooks and Warren or Krieger or Booth or others) or rhetoric (whether that Of Aristotle or Campbell or Christensen or what have you) and see no value in learning anything about educational research (and usually demeam it in the process of any discussion)" [sic].l While Donelson's remark pertains to English teachers in general, I find it particularly aPl3'| i cable to composition teachers and directors of composition pro- grams who, in spite of nearly a century of research in composition, are too often unaware of this research or ignore it even when they know it exists. In a l973 study, for example, Lucille Shandloff s“"Veyed twenty-seven junior college curriculum designs in composition to See whether the findings of research} in written composition were be‘i '19 used. She found little indication that designers implemented. or QVen Mm the findings of empirical research. Among her c°nclusions, Shandloff cited l) a need for research findings to be dis$emi nated widely among composition curriculum designers; 2) a need 98 99 to narrow the, gap between discovery of knowledge and its implementa- ti on into curricula; and 3) a need to reeducate both planners of curricula and classroom teachers. Shandloff's findings support those of Willard D. Memering who, in his l97l study of teaching practices at the secondary and college levels, found little evidence of theo- reti cal or research support for approaches used in the teaching of conpositi on . 3 Since the teaching of composition, along with the teaching of literature, form the core of English curricula, it is puzzling that so many curriculum designers and teachers are unfamilar with relevant research. Donelson suggests one reason when he notes that teachers' "rel uctance to read research (and the consequent lack of application of research supported findings to the classroom) stems from fear or lack of understanding or lack of training. That word, 'research,’ often brings stultifying memories of jargon (we do not use jargon in 5W] ‘3 sh teaching, after all), mathematics (and aren't English teachers s“PIDC>$ed to be afraid of numbers?), tests (which do not fit into the huméln‘istic tradition), and Greek letters (why can't they say what they mean without all that mumbo-jumbo), all apparently pointless ”“1 impenetrable to English teachers" (p. 1)- There is much truth in Donelson's suggestion that this reluc- tance comes from a basic unfamiliarity with research. Numbers and jargon can dissuade even the most dedicated English professional from discovering and applying relevant research. Yet, as Donelson further points out, this reason smacks of a weak rationalization for igno'F‘ance: "Any attempt at comnunicating ideas or information or r esea rch is based on some sort of language, and of course the reader 100 trying to get the idea or information or research must know the language. But reading Moliere in the original demands ability to read French, reading Frye's ANATOMY OF CRITICISM assumes a background in literary criticism and some awareness of other critics, reading Shakespeare with any intelligence demands awareness of Elizabethan Eng] ish, and reading anything in modern rhetoric assumes some back- ground in ancient and renaissance rhetoric. And English teachers seem able and willing to pick up some other language, if the results seem to be worth the time and effort" (p. 4). I contend that the time and effort are worth it. Research in composition has provided far too much information for curriculum designers and teachers to ignore. As Donelson points out, "Perhaps one reason the professions of law and medicine are respected is that the 9 cod lawyer or doctor keeps informed on what is going on in resea Y‘ch. Teachers who do not know what is being done and who is doing it in research simply have lost touch with basic information they must know" (p. 4). Curriculum designers and composition teachers, then, must ove Y‘come their ignorance of relevant research--ignorance of the 9’“ stence of research and ignorance of an understanding of research-- if they are to benefit at all from the insights it can provide into the teaching and learning of conposition. Since, as Dwight L. Burton poi rited out in l973, "Research activity in the field has represented a V'i V‘tual explosion,"4 it is sad that so few English educators are awa Y‘e of what research has been done in composition; and it 1'5 even sadder that those who are aware of it can be described as they were b y R01 and Harris in l968: "I do not suppose it is entirely unfair 10] to guess that when the average interested teacher of English is pre- sented with an account of a piece of research he turns to the general description of the problem, then to the summary and conclusions, brushing with slight horror enroute against a few clotted tables of correlations."5 Thus far in this book, I have tried to remedy this state of affairs by providing readers with background for reading research intelligently. In Chapter I, I attempted to provide insight into some of the essentials of research methodology; in Chapter II, I sought to compile significant research in crucial areas to inform the reader of what research has been done and what more needs to be done. In this chapter, I will discuss how research in composition can be used in educational settings. First I will consider some of the key arguments advanced against using research in composition, and then I will consider a number of arguments which support its value. A frequent criticism of research in composition concerns expectations. That is, for all the time, energy, and money invested in research, we should expect it to provide more conclusive answers to the issues and problems it examines. Instead, however, we find statements such as the following being made by research reviewers themselves: I Today's research in composition, taken as a whole, may be com- pared to chemical research as it emerged from the period of alchemy: some terms are being defined usefully, a number of procedures are being refined, but the field as a whole is laced with dreams, prejudices, and makeshift Operations.6 Anyone who has read a considerable portion of the research in ‘the teaching and learning of English composition knows how Inuch it leaves to be desired. quay. ' 102 Most reviewers of research shake their heads over the quality of much of what they review, and listings of common faults that make research inconclusive or invalid are strikingly similar over the past twenty-five years. (Burton, p. l6l) Statements such as these could easily lead us to simply shrug off all research as inconclusive and trivial; yet, given a multi-faceted pro- cess like composition, it seems no small wonder that research projects are laced with difficulties, errors, and questionable conclusions. In any event, research is not the panacea for writing problems; rather, it is the probe which examines important aspects of such problems. It does not provide the magic solutions we so often seek; rather, it provides directions which point toward those solutions. In short, we must not hold inflated expectations for research in composition. As Burton remarks, "Perhaps we have expected too much of an effort that is relatively young and we may have underestimated the difficulty of probing, for our particular reasons, the mysteries of human behavior, its development and change" (p. 161). A century of educational research has shown that to look for simple solutions to complex problems is to look simplistically, and in vain. A second argument leveled against the value of research in composition is that it is frequently incompatible with traditional practices. "I don't need research to tell me how to teach," says the angry teacher, "I've been teaching this way for a long time and I get results; experience tells me I'm right." And, indeed, the teacher often may well be right. Campbell and Stanley remark in Qgerimental and Quasi-Experimental Designs for Research that when "across the centuries many different approaches have been tried, if so'ne approaches have worked better than others, and if those which [VIVA .-‘ u 103 worked better have therefore, to some extent, been more persistently practiced by their originators, or imitated by others, or taught to apprentices, then the customs which have emerged may represent a "8 valuable and tested subset of all possible practices. Seeking insight and direction from both traditional practice and experimental research, the alert educator must see the values and limitations of each. E The relationship between traditional practices and research can be considered from three perspectives: l) the two may be incom- patible because traditional practices are right and research wrong; .2) they may be incompatible because traditional practices are wrong and research right; and 3) they may be compatible because both are right. From the first perspective, for example, traditional practices and years of classroom experience tell us that practice in revision improves writing skills. Yet this belief clashes with the results 01’ a study conducted by Barbara L. Hansen, who found that the act of "evi sing alone may not lead to improvement in writing performance.9 Hansen's findings conflict not only with traditional practices and experience but also with other research investigations into the e""“f’i cacy of revision (see, for instance, the studies cited in the B r‘acldock Report, pp. 35-36). Until researchers replicate the Hansen study and come to conclusions similar to hers, the results must be ta ken as tenuous. An example of the second perspective--that traditional prac— ti Qes are wrong and research right--can be seen in the long-held 5% j ‘i ef that formal instruction in granmar leads to improvement in 104 writing. A half-century of reserach involving all kinds of students at all levels has consistently concluded that, as the Braddock Report unequivocally states, "the teaching of formal grammar has a negligible or, because it usually displaces some instruction and practice in actual composition, even a harmful effect on the improvement of writing" (pp. 37-38). Yet many writing teachers and composition in , prOgram directors continue to accept unquestionably the principle and practice of improving writing through formal instruction in abstract gramnar. From the third perspective, both traditional practices and research often agree, as in the relationship between reading and vvriting. Educators have long held that linguistic production (i.e., writing and speaking) is facilitated by linguistic reception (i.e., reading and listening). As I pointed outin the previous chapter, much research supports this integrative position: that is, that all the language functions reinforce one another; and, while we may not as .Yet know precisely why, we are sure that linguistic production is a1°‘l’ected by linguistic reception. From this overview of the compatibility of traditional prac- tiCe and research, it should be apparent that the relationship be"i‘rleen the two is one of reciprocity. Since the goals of both are the same-~that is, better understanding of learning and better ways or teachinguboth are vital to the educative process. We need tr‘a‘flitional practice insights based on the cunulative experience of Yea "s of classroom teaching as much as we need the refining insights 0? research. Campbell and Stanley, in their discussion of “tradi- ti Q“ a1 wisdom," express this dual need succinctly when they state 105 that "Experimentation thus is not in itself viewed as a source of ideas necessarily contradictory to traditional wisdom. It is rather a refining process superimposed upon the probably valuable cumulations of wise practice. Advocacy of an experimental science of education thus does not imply adopting a position incompatible with traditional wisdom" (p. 4). A third criticism of research concerns its truth value. "What,“ critics argue "does research 'prove'?“ If we look to research to "prove" a theory or hypothesis, however, we operate under a funda- That is, mental misconception of the nature of experimentation. experimental hypotheses are never "proved," but rather exposed to disconfirmation and either disconfirmed or not disconfirmed. Once- again, Canpbell and Stanley sum the matter: "experimental results never 'confirm' or 'prove' a theory-~rather, the successful theory is tested and escapes being disconfirmed. . . . The results of an eXperiment 'probe' but do not 'prove' a theory. An adequate hypo- thesis is one that has repeatedly survived such probing-~but it may 37"!st be displaced by a new probe" (p. 35). Thus, the truth value of research is not that of definitive "proof" or confirmation of hypotheses but rather that of disconfirmation of rival hypotheses. Th ‘5 s is why control of variables (rival hypotheses) is so important in experimental research. "Varying degrees of 'confirmation' are co... ‘Ferred upon a theory through the number of glausible rival W available to account for the data. The fewer such .9] alasible rival hypotheses remaining, the greater the degree of Q9" firmation'" (Campbell and Stanley, p. 36). In other words, the ”Q h e one experiment controls for rival hypotheses, the greater its 106 truth value. Because of the persistent problem of control of vari- ables in research, therefore, we must recognize the crucial role of multiple experimentation and the need for replication--that most often research studies reveal only partial "truths" and that the balance of "truth" is found in multiple examinations of the same or similar hypotheses. Closely related to the truth value argument is a fourth argument which focusses on flaws 13 research. As I tried to point out earlier in this chapter in my discussion of expectations, educa- tional research and research in composition in particular (with its many variables) is not without its flaws. Burton remarks that "Control of variables is, of course, the eternal ghost haunting experi- mental studies" (p. 166). Flaws in research are most crucial, how- ever, if we consult only isolated studies. Here again, the value of experimental research lies in repeated experimentation. Flaws in Separate studies tend to cancel one another out when the combined experimental outcomes are the same. James R. Squire, in his recent aPticle entitled "Research Can Make a Difference," remarks that "No one of these studies, by itself, may seem overly impressive; not i "frequently their controls and manipulation of data seem suspect. Yet taken together they tend to yield similar findings . . . the "lost apparent influence results over the long run from the continued i ""'Dact of a large number of related studies that seem to point in the s ame direction."10 If we seek "conclusiveness,“ therefore, we must consider 1: h e long-range results of research. For example, the flaws that mar 'Tl an) of the investigations into the relationship between the formal ‘ 107 teaching of grammar and writing improvement matter little when the bulk of these studies point to the conclusion that writing is not improved through formal grammar study. Similarly, though many inves- tigations in the last decade into the effects of teaching modern lin- guistics on improving writing may be suspect in one way or another, most conclude that the teaching of modern linguistics may improve F“ one's knowledge of abstract grammar but not one's writing. Other examples which are less "conclusive" but still bear the weight of replicated findings are those studies which support the efficacy of small peer-groups and/or writing labs and those studies which find a relationship between audience-awareness and writing improvement. While these studies are more recent and thus do not have the influ- ence of nunbers that the gramnar studies have, they nonetheless cancel out each others' flaws in their conmon findings. It is in multiple studies such as these that we find, as Squire states, the "long-range effects that shape and influence the direction of change in curriculum and instruction."” Of the arguments in favor of using research in composition, the first is that research findings often challenge our assmnptions about how conposition is learned. It is a healthy challenge because i t forces us to critically examine our beliefs and, at times, to "bdify them. Consider, for example, a traditional assumption about the composing process. The traditional composition-handbook view of the composing process (which still informs much contemporary tgaching practice) is that of a simplified, mechanical think-outline- wh ‘i te-revise procedure. But recent research into the nature of the ‘ 108 composing process shows that it is a complex, multi-faceted phenomenon. Janet Emig's study of the composing process of twelfth graders iden- tifies many components of this process, along with many factors which affect it (and, ultimately, the written product).12 In addition, Donald H. Graves' study of seventh graders makes paramount the impact of the learning environment on the composing processes of students.13 a In research indirectly related to the composing process, we see the key functions of prewriting activities in the improvement of composi- 14 tion skills. Studies such as those by Taylor and by Rohman and Wlecke15 suggest that far more emphasis be placed on the prewriting stage in the process of composing. Finally, studies which investigate ‘6 and the causes of linguistic insecurity, such as those by Brazil Koch17 suggest that much of what teachers view as "writing problems" ame actually matters of general language apprehension. Research also challenges our assumptions about how composi- trion is taught. Consider, for example, two common practices which are part of almost every composition classroom. The first is the '“Clle of audience in the writing situation. While few pe0ple would d‘5sagree that in any human comnunicative act an audience is a _s_i_n_e_ -SJJ!E§HQQQ, how often is it that the writing which goes on in a composi- ti on class is directed toward a real audience? Moreover, how often do teachers consider the impact audiences other than themselves have 18 upon their students' writing? Hooks' study clearly shows that a udience-awareness is crucial to the professional writer; while at 19 Ford,20 and the same time, studies by McClatchey and McClatchey, L figanafl indicate that the presence of peer audience improves the q uality of student writing. The crucial pedagogical matter at issue ‘ 109 here is whether or not provision is made in the classroom for other than teacher-directed writing audiences. To assume that audience does not matter is to ignore the persuasive findings of a number of studies. A related pedagogical matter concerns the nature of evaluation of student writing. As I pointed out in the previous chapter, research .p indicates that either positive or negative criticism will produce the ‘ 22 same results in short term achievement. Many studies, however, conclude that positive and negative criticism have definite and oppo- site impact on attitudes toward composition; and, as I further remarked in the last chapter, these studies suggest that in the long run, it is 'through the reciprocal, integrative nature of positive attitudes and achievement that genuinely self-directed learning and improvement in composition occurs. Thus, teachers must examine their traditional evaluation procedures in light of both short-term and long-term goals 1'1 achievement and attitude-formation. On a larger scale, research challenges our assumptions about &riculum design. If we assume, for example, that students learn Dr‘i marily by being told about something, then the classroom model ‘tifliit commonly evolves is the lecture or the lecture/discussion model. Jr'f’.. on the other hand, we assume that students learn not only'by MMM something but also by d_oigg_ and by observing, then the model that emerges is likely to be the workshop. Thus, tradi- t i Onal writing classroom models are seriously challenged by compara- t‘i Ve methodological studies such as those by Haas, Childers, Babbit, a“ cl 1mm}3 Sutton and Arnold,24 and Smith,25 all of which find the 110 workshop/l ab model more effective than the lecture/discussion model in improving composition performance. Another aspect of curriculum design that research challenges Most curricula are set up along the lines of is its time structure. students take "courses" during "terms"; certain time constraints: and they study "units" which are organized into "blocks," all of which make for order and supposed efficiency. We can articulate precise objectives for these time units; and, after a specified lapse of time, determine whether or not the objectives have been fulfilled. But, when our objectives are to develop competencies in composition, the time constraints of traditional curriculum design are not consis- tent with the findings of much research which concludes that the development of language competencies takes much more time than is provided in a six-week "unit" or a ten-week "term." As Burton observes, "studies which have featured experimental treatment over a period of only a few weeks or months have been predestined to conclu- 8 ‘ions of 'no significant differences,‘ since one thing that is known, at least, is that improvement in general aspects of writing ability 'i S a slow, gradual process." This is not to say that certain writing "Skills" cannot develOp in short periods of time, for as Burton goes on to point out, "it may be possible to identify those aspects of Wf‘i ting ability that can be expected to show growth in short periods. Se ntence embedding may be one example" (pp. 177-178). If, however, We expect general writing improvement in a ten-week "term" or a six- Week "unit," then it may well be that our expectations are too high. In fact, a frequent theme which emerges in many studies in conposition i s that if more time were available, the researchers could come to 111 more conclusive results. It would appear that some researchers, like sane teachers, expect more thanis possible in a given length of time. The development of composition competencies, as with all forms of language conpetence, takes time. That being the case, we are better off asking "What 23p be learned?" rather than "What shguld be learned?" within the time constraints of curriculum design. Furthermore, serious consideration should be given to extending the time constraints of curriculum design so that it is more in accor- dance with the actual exigencies of developing competence in composition. One of the most important functions research serves is that of an informed vehicle to implement educational change. There is cause for regret, as Sara W. Lundsteen points out in Help £93133 Teacher o_f Written Composition, over the "lag of twenty-f1 ve to thirty-five years between the discovery of new knowledge and putting that knowledge into action."26 Were teachers and administrators to give greater consideration to the findings of research, much evidence Could be marshalled to support the changes both groups so often seek to bring about. Again, Lundsteen remarks that "Often changes in s Chool practice are stalled because sufficient evidence to support Change is not available or because diverse opinions or confusion °Ver contradicting theories exist."27 In the field of composition, n'l11<:h evidence is, in fact, available; but, through professional 1 Quarance or lack 'of training, it is seldom brought to bear on the Cle‘3'Fsionumaking process. Where methodological disputes exist, for e"al'rnple, research findings can, if not settle such disputes, at least Se we to refine methodologies. Furthermore, where educational 112 problems hamper educational progress, research can provide much valuable insight into solutions to these problems. This is not to say that research can solve all the ills in the teaching of English, but only that it can provide a wealth of information vital to the search for solutions to these ills. Nathan S. Blount supports this conviction in stating a few years ago that "In 1970 it seems clearly possible to solve significant educational problems in the teaching of English by research.“28 ; Along with implementing and supporting change, research serves to educate (or re-educate) teachers and adninistrators. As I .. t. pointed out earlier, research forces us to re-examine our assumptions about learning, about teaching, and about curriculum design. In addition, it forces us to supply rationales for our practices which are based on more solid ground than our feelings, our experience, or our subjective evaluations. Our approaches to teaching and learning Ought to be based, as Donelson emphatically states, "on something better than 'Students do seem to learn from it,’ or 'I'm excited about it and the kids seem to be,‘ or 'It seems to work better than (or as well as) what we have used,’ or 'Kids will thank me someday for all I'm doing for them,‘ or 'We've always done it this way,I or . I t's good for students,‘ or 'You can't get into college without it,‘ 0" 'This is part of our cultural heritage,‘ or 'You can't get a job Without it.‘ All these answers are invalid and intellectually i n defensible without some sort of supporting evidence, yet each like the voice of the turtle is heard in our land" (p. 3). Thus, research serves both as the vehicle for educating those who institute change as well as the vehicle for informing the kinds 113 of changes that are made. In this regard, I quote once more from Donelson, who synthesizes both of these functions of research when he concludes that [teachers] had better have demonstrable evidence to support their inevitable contentions that the new should (or should not) replace the old. When the implementation of the new calls for no funds, then it is simply academically and intellectually indefensible not to have some proof, and that is the teacher's problem, no one else's. But when implemen- tation of materials or techniques costs money, then the already budget-pressed school administrator has every right to ask, "What makes you think it will work?" or "How will you know it does anything better than the way you're teaching now?" (p. 7) I would amend Donelson's remark to add that administrators as well as ‘teachers are responsible for marshalling evidence for whatever curri- culum changes they promote. In conjunction with implementing and supporting change in c:urriculum practice, research should support whatever theoretical .Iirameworks inform curriculum design. This is particularly true in the field of composition, where in the last two decades an explosion Of knowledge has occurred in relevant theoretical areas. With new ‘3 nsights and research into the nature and functions of language and ‘i fItO the psychological, sociological, economic, environmental, and physiological influences on learning, the demand for broader founda- ‘iZ‘i ons for curricula becomes much greater.7 We must know and apply What what is being done in such fields as learning theory, modern 1 inguistics, psycholinguistics, sociolinguistics, language acquisi- t ‘5 On, reading theory, and composition theory. For too long, teachers and administrators (particularly those i n English) have ignored (for whatever reasons) the findings of both theory and research. Memering's investigation of theories underlying 114 practices in composition on the secondary and college levels, which 29 found that no true theory underlies any of the current practices, is sadly echoed by Burton when he notes that "The composition teacher, as everybody knows, can show no respectable theory; His discipline boasts no scholarship but is planned by dolts, manned by drudges, and avoided if possible by everyone“ (p. 174). Memering's and Burton's observations, coupled with Shandloff's findings that there is little indication that composition program designers imple- mented, or even knew about research in composition,30 clearly indicate just how weak are the designs of most composition programs. We can no longer afford to ignore the wealth of information that is available to us both in theory and in research. As Mina P. Shaughnessy points out‘in her article "Basic Writing," "The teacher must know more about 1 anguage and learning than English teachers have had to know in the Pastumore about how people acquire languages, how language functions 7 n different social settings, what writing is as a product and a pro- cess, and what theories of learning might inform his pedagogy. The teacher faces, in short, the formidable task of extrapolating from a "lawyer of complex, even turbulent, fields whatever insights and infor- Ination will serve him in his work with his students."31 Where Composition programs and practices are informed by theoretical frame- ‘VO rks deriving from such fields, research enters as the cutting edge, the means of supplying solid support for the rationales and conditions Uhder which composition can be taught and learned. In this chapter I have tried to present a comprehensive di 3 Cussion of the value and use of research in composition. I have 115 considered arguments both Opposed to and in favor of using research. Toward the end of this chapter, I have argued for the design of com- position programs which are based on relevant theory and supported by the findings of research. In the next chapter I will demonstrate how research in composition can be used in conjunction with relevant theoretical frameworks to arrive at a theory of instruction for the teaching and learning of composition. CHAPTER IV A THEORY OF INSTRUCTION FOR COMPOSITICN: THEORY AND RESEARCH INTEGRATED So far I have argued that composition teachers and directors of composition programs should know research--know how to read research reports, and know what research has been done in composition. I have also argued that teachers and directors should 952. research in the design of composition programs. My third argument is that teachers and directors should also know about and use relevant theory in designing composition programs. If we seek composition programs ‘5 nformed by significant theoretical and research findings, then theory and research must be integrated. In this chapter, therefore, I will examine theory pertinent to the development of a rationale for the teaching and learning of composition. I will then present a number of directions from research which contribute to a theory-based '3 "structional rationale. Finally, I will integrate both theory and r‘esearch in order to formulate a theory of instruction for the teaching and learning of composition. In "Beyond Literacy," an essay which appeared in the March, I 973, ADE Bulletin, Alan M. Hollingsworth argues that English must bec=0l11e an interdisciplinary field: "I believe that many of us in Eng] ish must learn about, much more about, other fields of study, 116 117 other subjects, other arts, other sets of learning activities. I believe that English must become interdisciplinary, but with caution and no illusions. In the 1970's English must become interdisciplinary, multi-disciplinary, crossdisciplinary." Hollingsworth puts forth four reasons for his argument. The first is to insure our self-preserva- tion--that is, we must know far more about our field than we ever "We must," Hollingsworth 2 have before in order to answer our critics. writes, “be tough, saphisticated, knowledgeable." The second reason is to avoid the curriculum waste which derives from uninformed thinking and lack of historical awareness. The third reason is to reaffirm the dynamic nature of literary criticism in the curriculum tzy recognizing its expanding frontiers. Finally, English must become 1 nterdisciplinary in order to support and extend curriculum reform. Hollingsworth's thesis is even more pertinent today than it was in l973. As we move closer to the l980's, it becomes more and "'Or‘e apparent that a virtual explosion of knowledge has occurred in numerous fields, many of which, just a few years ago, were generally <=<>t1sidered to be remote from the concerns of "English" as we knew it. And, in the field of composition, in particular, we now know that these fields bear directly on our work. Robert E. Shafer puts the Case very well in his l975 ADE Bulletin article entitled "The Crisis 1 "I Knowing about Learning to Write" when he says: If we look only briefly at some of the ongoing research on writing and learning to write, we will see that we perhaps ‘do not have a "crisis" at all, but rather an explosion in understanding how learning to write and to read actually 'take place. Although some of the research is being done in English education programs and in departments of linguistics, as well as occasionally in departments of English, the «explosion in knowledge about human communication processes 'is derived from such seemingly diverse fields as linguistics, symb educ Alth bedf disc than larl rese sense to understa consider Cong051t sional 8. than men Hhen NECGS exam; on a is tr 0f V( Phenc the ( Socic Write and 1 dEfiy "rite acti< EXDGI Semic a wri EHce audie a wri IISQ reSee duce SEIeC Phenc as a ASSOC dl3tr dEman be h' 118 symbolic logic, cognitive psychology, information theory, educational theory, and rhetorical theory and research. Although, as Shakespeare put it, "misery makes strange bedfellows," it may well be that a new look at this inter- disciplinary research will be more valuable at this time than increasing the number of our "speculations," particu- larly when these are ill-grounded and based upon no research at all. Clearly, Shafer is using the term "research“ in its broadest sense to include both theoretical and empirical findings. We can understand how comprehensive this interdisciplinary research is by considering the following quotation from Carl Klaus' recent College Composition and Communication article, "Public Opinion and Profes- sional Belief," in which he discusses writing as a process rather than merely as a product: when writing is understood as a process, the study of it necessarily demands an interdisciplinary approach. For example, a writer in the act of using language is drawing on a unique set of verbal possibilities (idiolect) which is the product of the writer's interaction with shared sets of verbal possibilities (dialects). Understanding these phenomena and their impact on the process of writing requires the expertise of such disciplines as psycholinguistics, sociolinguistics, and linguistic anthropology. Likewise, a writer in the act of using language is inescapably discovering and fbrming experience, for the words which a writer uses to define experience not only communicate but also shape that writer's perception of experience. Understanding the inter- action of language, cognition, and perception requires the expertise of such disciplines as cognitive psychology, semiotics, and transformational linguistics. Furthermore, a writer in the act of using language is communicating experi- ence for one kind of purpose or another to one kind of audience or another. Understanding the interaction between a writer's social intention and language requires the exper- tise of such disciplines as rhetorical theory and communication research. Finally, of course, a writer uses language to pro- duce a piece of writing which embodies in its particular selection and arrangement of words the interaction of all the Inhenomena I've described thus far, and countless others, such uas a writer's aesthetic intentions, or mental and emotional iissociations that transpire during the process of writing, or (distractions that interrupt the process, or the technical (demands of a particular subject matter--too many in fact to the listed here. Understanding the relationship of that sel bro sty of ana one on the muc the app experti “By thi become ence it that ma for.‘ 8 With ma believe and abo Shaughn 119 selection and arrangement of words to the phenomena that brought it into being requires at last the expertise of stylistic analysis, which in turn depends upon a variety of disciplines, such as literary criticism, rhetorical analysis, psychoanalysis, and statistics. Clearly, enough, one discipline or another can provide only one perspective on only one element or stage in the process of writing. If the process is to be wholly understood, if we are to know as much about writing as we possibly can, then we must bring to the study of it as many disciplines as are possible and appropriate. The number of interdisciplinary fields in which Klaus demands expertise is overwhelming in its scope and depth, as he points out: “By this point, no doubt, many of you are wondering how you could ever become familiar with research in all those disciplines and what differ- ence it would make to your teaching even if you did."5 It is unlikely that many of us can conceivably gain the ideal expertise Klaus calls f’or.~ But certainly we can become far more familiar than we are now with many of these fields. Since our work is in composition, I believe we must know as much as we can about learning, about language, and about pe0ple learning language, particularly composition. Mina P. Shaughnessy is quite right when she states recently that the "teacher must know more about language and learning than English teachers have had to know in the past--more about how people acquire languages, how language functions in different social settings, what writing is as a product and a process, and what theories of learning might inform his pedagogy. The teacher faces, in short, the formidable task of extra- polating from a number of complex, even turbulent, fields whatever insights and infbrmation will serve him in his work with his students."6 In the sections which follow, then, I will examine theoretical Principles and implications of learning theory, language theory, and the0' cons nagn know comp.- how ‘ how, inst of v as a task Prev 120 compgsition theory, My intention is to devel0p a rationale for a theory of instruction in composition. Thus, I will not directly consider every related interdisciplinary field; aside from the sheer magnitude of such an enterprise, I am not qualified to do so and I know of no one who is. Rather, I hOpe to demonstrate how theory in composition and related fields can be used to develop a rationale: how this rationale can be supported by the findings of research; and how, together, they can produce a sound and practical theory of instruction for the teaching and learning of composition. An appropriate view of "learning" must derive from a theory of what learning is. It must account for the complexity of a person as a thinking creature, and it must account for the diversity of tasks faced by a person when he learns. Today two learning theories prevail in education. The behaviorist theory views learning as habit formation: a response to a stimulus becomes habitual through rein- fbrcement. Accordingly, the learner is viewed as a creature of habit; and his learning tasks are accounted for by observing the demonstrable behavior during which the habits are formed. While some aspects of human behavior may be accounted for by habits (e.g., some actions and reactions), this view of learning seems inappropriate when we consider the powerful non—observable dimension of learning, namely, thinking. A second learning theory and one increasingly prevalent today--the cognitive--is centrally concerned with the way people make sense of their world--that is, with the way they interpret events in ‘their world and organize these events in order to make sense of their 121 world. It is a more powerful theory of learning because it views the learner as a thinking creature, and because it accounts for far more learning tasks than the purely demonstrable. He may contrast these two learning theories pedagogically. Typical of the behaviorist pedagogy is some form of rote learning, or, what Carl R. Rogers calls the "mug and jug" approach: "Most traditionally-taught courses attempt to 'elicit' responses from students by stimulating them, coaxing them, lecturing to them, etc. Courses structured along these lines are analogous to the 'mug and jug' pedagogical procedure 'where the teacher represents the fount of knowledge (the Jug) and the student represents the yawning recep- In contrast, learning typical of the cognitivist pedagogy is what Rogers calls "significant learning": ”By significant learning I mean learning which is more than an accumulation of facts. It is learning which makes a difference--in the individual's behavior, in the course of action he chooses in the future, in his attitudes and in his personality."8 While it is true that rote learning would also make a difference in the individual's behavior and course of action he chooses in the future, what is most crucial here is that "signifi- cant learning" does not depend on an external stimulus to initiate or sustain it. Rather, this kind of learning is self-motivated--the learner perceives a problem which is important to him and which he really wants or needs to solve. It is thus far more intricately tied to his attitudes and personality than is any form of rote learning. Le_a_[_r lean ' inpor lean a cog grate as Sr exan' in cc that lati< Cate< by V; 1 fl! Stru. unive test: 8559, test- IEStl StTuC teste Comic 0r wh 122 Frank Smith makes a similar distinction in Comprehension and Learning when he distinguishes "meaningful learning" from frote learning": IThe aspect of learning that I regard as by far the most ‘ important is that which can and often must be self-initiated by the learner. I shall call such learning 'meaningful' because it involves a cognitive change that makes sense, because it is intimately inte- grated with everything else the learner knows about the world." If, as Smith points out, "what matters is meaningfulness,"9 then we must examine Ameaningful learning" in light of its theoretical orientation in cognitive learning theory. Psychologist Ulric Neisser, in Cognition and Reality, writes that "Cognition is the activity of knowing: the acquisition, organi- zation, and use of knowledge.“10 This activity of knowing is predi- cated upon the learner's cognitive structure, a concept which goes by various, but similar, definitions. I will rely on two here which I find most lucid. The first is what George A. Kelly calls a "con- struct": "we consider a construct to be a representation of the universe, a representation erected by a living creature and then tested against the reality of that universe. Since the universe is essentially a course of events, the testing of a construct is a testing against subsequent events. In other words, a construct is tested in terms of its predictive efficiency."n Kelly's definition presents the key components of cognitive structure: it is a representation, with predictive value, which is tested against the‘world. Frank Smith elaborates on these key components in his discussion of the elements of cognitive structure, (Jr what he calls a "theory of the world“: "The theory is a summary of experier will try tc reaning the 'constitutet for learnil learning: of an 'expe light of ti See our intera< 0f the worl ihen we teg either cont It is in tr because "16 tive struct learning Si inadequote we WW c predlCtable our theOry Thus W4 Sml'th: 123 of experience; it is memory. It is the evidence upon which children will try to interpret new events--the only basis for any sense or meaning that they can impose upon the world. And finally the theory ‘constitutes their expectations for the future; it is the foundation for learning.f Smith's metaphor for the learner is that of the scientist, which is the most-used metaphor in the cognitive view of learning: "Scientists construct hypotheses, which become the basis of an 'experiment,‘ and they confirm or modify their theories in the light of the experimental results” (p. l2). Seen in this way, our cognitive structures form the basis for our interactions with our world. We operate according to our theories of the world, theories which are made in light of past experience. When we test our theories of the world (i.e., form hypotheses), we either confirm them or encounter something which challenges them. It is in the challenge that the opportunity for learning arises, because "learning is the process of elaborating and modifying cogni- tive structure when it fails to make sense of the world. . . . A learning situation arises whenever our cognitive structure proves inadequate for making sense of the world . . . And we learn--that is, we modify cognitive structure--so that our experience will become more predictable, so that in the future we will have a better match between our theory of the world in the head and our experience" (Smith, pp. ll8-l19). Thus, "learning,“ from the cognitive perspective, is hypothesis-testing, As such, it involves the four steps outlined by Smith: en SO C6 C05 mal tic hat Stu 0P0! Whit Stuc mean 18" 124 (a) the eneration of a_hypothesis, or a tentative modification or elaboration of any of the . . . components of cognitive structure; (b) a test of the hypothesis involving some direct ' interaction with the environment in order to obtain feedback; (c) feedback, which provides new information against which the predicted consequence of the original hypothesis can be compared, (d) acceptance or r ection of the hypothesis. If the feedback is positive--1f tfie consequence ofithe tested hypothesis is compatible with the predicted consequence--the hypothesized change in cognitive structure is confirmed. If the feedback.is negative--if the result of the test is con- trary to the rediction--the hypothesis is rejected or modi- fied. (p. 223) From this model of the learning process, many principles derive which have significant implications for learning in general and language learning (including compoSition) in particular. The first principle concerns the problem-solving nature of the hypothesis- testing model. Meaningful learning j§_problem-solving, as Smith emphasizes: "Learning, in other words, is a process of problem- solving. We learn because we do not understand, cannot relate, cannot predict. Everything we know, then--the current state of our cognitive structure--is a consequence of all our previous attempts to make sense of the world" (p. ll9). If, therefore, we seek to create meaningful learning situa- tions in composition, we must focus attention on the problem-solving nature of learning. Carl R. Rogers suggests that we "permit the student, at any level, to be in real contact with the relevant problems of his existence, so that he perceives problems and issues which he wishes to resolve."12 This suggests that composition students be given the Opportunity to grapple with real, personally meaningful issues and problems in the content of their writing. This is not to argue far an exclusive focus on experience-based writing but simply to urge teachers to consider the real problems students UT St ho ta sk‘ the the Not lea t Eur-\wlt " 125 face in their lives and build on them in the composition classroom, problems which are frequently as intellectual as those which teachers so often impose on students. Problem-solving also applies to the "skills” of writing. Since the most prevalent rationale for develOping writing skills is to improve competence in written communication, the trial-error- feedback methodology of hypothesis-testing is directly applicable. Smith, for example, points out the problem-solving inherent in the development of writing skills: "This skill involves predicting the uncertainty of a listener (or reader), and organizing the surface structure so that just that uncertainty is reduced. . . . Estimating how much prior knowledge on the part of a listener or reader can be taken for granted constitutes a major part of a speaker or writer's skill” (p. ll2). Reduction of the reader's uncertainty, then, becomes the writer's goal; the success or frustration of this goal comprises the learning situation. In Teaching Egg.Univer e gf_Discourse, James Moffett succinctly sums up the potential of trial-error-feedback learning in the composition classroom: If, as I believe, writing is learned in the same basic way other activities are learned--by doing and by heeding what happens--then it is possible to describe ideal teaching prac- tices in this way and compare them with some current practices. Ideally, a student would write because he was intent on saying something for real reasons of his own and because he wanted to get certain effects on a definite audience. He would write only authentic kinds of discourse such as exist outside of school. A maximum amount of feedback would be provided him in the form of audience response. That is, his writing would be read and discussed by this audience, who would also be the coaches. This response would be candid and specific. Adjustments in language, form, and content would come as the writer's response to his audience's response. Thus instruction would always be individual, relevant, and timely. These are precisely the virtues of feedback learning that account for its great success.13 ext noi bet the no can str tun Errc lear thei BeCdl offer there Ilsks OCH 126 The second principle to consider is inherently part of the trial-error-feedback methodology. I refer to the risk-taking which must go on in any hypothesis-testipg_situation. That is, whenever meaningful learning occurs, the learner takes certain chances when he puts his cognitive structure to the test. He literally experi- ments, saying, in effect, "This is what I believe," or "This is what I think should happen." The amount of uncertainty present in his experiment determines the degree of risk involved. George A. Kelly notes that "there are times when a person hesitates to experiment because he dreads the outcome. He may fear that the conclusion of the experiment will place him in an ambiguous position where he will no longer be able to predict and control. He.does not want to be caught with his constructs down. He may even keep his constructs strictly to himself lest he be trapped into testing them prema- turely."14 All learning, then, involves risks and the possibility of error. Learners who are reluctant to take risks trade an enormous learning potential for a degree of safety. Yet, as Smith points out, "Low risk takers are unlikely to learn very efficiently because of their reluctance to take the chance of being wrong" (p. l97). Because the possibility of error always exists in risk-taking, effective learning frequently exacts a price. Smith remarks that there are economic issues that the learner considers when taking risks: the initial investment of time, interest and effort, the ultimate value of the learning achievement; the probability of being successful; the rewards or disadvantages of alter- native outcomes. A child might be regarded as making a by C0 pr CD! for so: cle Edi. ide I‘ea Pre sty CES b00 enCC 127 cost-benefit analysis before he enters into any learning trans- action, with his present and predicted emotional states as variables that are taken into account. If the costs of a par- ticular learning task outweigh the estimated benefits, a child is unlikely to accept such a pointless bargain. (p. 226) Thus, mistakes are an essential part of learning. We learn by being wrong as much as we learn by being right, provided, of course, we have the freedom to learn by being wrong (i.e., if the price exacted is not too great). This is particularly true in the composition classroom, where the necessity for risk-taking is para- mount. For example, seldom are students encouraged to explore half- formed ideas, or conceptual ambiguities in their writing, even though such exploration often leads to greater originality and improved clarity in writing. Smith remarks that "One of the tragedies of our educational system is that it seems to result in reluctance to bring ideas to the surface and expose them to criticism" (p. ll4). The reason for such reluctance may be that the demands for intellectual precision weigh too heavily against conceptual risk-taking. In another sense, reluctance to take risks with the rhetorical] stylistic/mechanical aspects of composition* affects the entire pro- cess of composing, as Mina P. Shaughnessy illustrates in her recent book, Errors and_§xpectations: If a writer is not worried about being wrong, if he sees a chance fOr repairing and perfecting his copy at a later point before anyone sees it, he will be free to think about what he means and not worry so much about the way he is saying things, a worry that almost inevitably cuts him off from his best grammatical intuitions. Furthermore, by withholding closure on his sentences, he is more likely to work on them *For the sake of brevity, I will hereafter use "stylistic" to encompass matters of style as well as of the rhetorical and mechanical (e.g., punctuation) dimensions of composing. a: ct at co an uh tht 128 and, in the process, begin to be aware of his power to make choices (semantic and organizational) that brings him closer and closer to his intended meaning.1 Thus we see that if learners are to exploit the risk-taking dimension of learning, a healthy attitude toward error is essential. Error must be considered an avenue toward meaningful learning, not something to be avoided at all costs. In the trial-error-feedback methodology, errors are of the utmost significance. A third principle from cognitive learning theory--decentrali- zgtiggy-corresponds to the way Piaget used this term in his work on the developmental stages of children's thought. Piaget examined the talk of children as a reflection of their thought processes. He found two classes of talk: the eggrcentric and the socialized. In the first, the child “does not bother to know to whom he is speaking nor whether he is being listened to. He talks either for himself or for the pleasure of associating anyone who happens to be there with the activity of the moment. This talk is ego-centric, partly because the child speaks only about himself, but chiefly because he does not attempt to place himself at the point of view of his hearer."16 The communicative use of language in the e o-centric stage is subjective and private. Piaget writes that the child speaks "in a language which disregards the precise shade of meaning in things and ignores the particular angle from which they are viewed, and which above all is always making assertions, even in argument, instead of justifying them. . . . In a word, the child hardly ever asks himself whether he has been understood. . . . he does not think about others when he talks" (p. 40). 129 In socialized talk, on the other hand, the child does place himself at the point of view of his bearer; he realizes that others do not always see things as he does, and adjusts his talk accordingly. The communicative use of language is more public in that the child is more concerned with making himself understood by others. The transition from eggycentric thought to socialized thought is a process of decentralization-~of seeing things from another's point of view. But this process is not confined to the child's develOpment. It is an inherent part of the functioning of everyone's thought and language, as Piaget points out: We shall quickly realize the full importance of ego-centrism if we consider a certain familiar experience of daily life. We are looking, say, for the solution of some problem, when suddenly everything seems quite clear; we have understood, and we experience that sui generis feeling of intellectual satisfaction. But as soon as we try to explain to others what it is we have understood, difficulties come thick and fast. These difficulties do not arise merely because of the effort of attention needed to hold in a single grasp the links in the chain of argument; they are attributable also to our judging faculty itself. Conclusions which we deemed positive no longer seem so; between certain propositions whole series of intermediate links are now seen to be lack- ing in order to fill the gaps of which we were previously not even conscious; arguments which seemed convincing because they were connected with some schema of visual imagery or based on some sort of analogy, lose all their potency from the moment we feel the need to appeal to these schemas, and find that they are incommunicable; doubt is cast on propositions connected with judgements of value, as soon as we realize the personal nature of such judgements. (pp. 45-46). An interesting parallel to Piaget's investigation of intellec- tual development in childhood is William G. Perry, Jr.'s examination of thought processes in late adolescence. In conducting research for Forms 9! Intellectual and Ethical Development ig_the College Years, Perry observed the talk of college-level students; and, from his 130 data, he devised a schema which chronicles the evolutionary stages of cognitive develOpment. Basically, Perry's schema begins with rela- tively simple dualistic thinking (absolute right-wrong, good-bad, etc.); moves through a number of transitional stages marked by increasing modification of outlook; and develops to final, more mature, stages of relativistic thinking and affirmation of the self in a pluralistic world. It is a movement, in Perry's words, "away from a naive egocentrism to a differentiated awareness of the environ- l7 ment." Both Piaget's and Perry's investigations of decentralization are vital to our understanding of the learning process. If meaningful learning amounts to the modification of cognitive structure, then decentralization (i.e., the ability to broaden one's perspective) is at the heart of learning. In this regard, James Moffett is right when he says that "Learning is a matter of 'decentering,' of breaking through our egocentricity to new points of view not determined solely by our physical vantage point in time and space or by our emotional preferences. We achieve decentering by adapting ourselves to things and peOple outside ourselves and by adopting points of view initially foreign to us. . . . This amounts to expanding one's perspective" (p. l48). The implications of decentralization encompass both the con- ceptual and the stylistic dimensions of composition. In the concep- tual domain, it suggests that barriers to effective communication of one's ideas are ultimately bound up in the process of decentralized thinking. As Moffett points out, "Probably the majority of communica- tion problems are caused by egocentricity, the writer's assumption that ence vrii awai prot tes lea the the the eit Nor int 11:0 C0” Obj the C09 131 that the reader thinks and feels as he does, has had the same experi- ence, and hears in his head, when he is reading, the same voice the writer does when he is writing. It is not so much knowledge as awareness that he needs" (p. 195). Moffett argues further that problems diagnosed as stylistic are also problems of decentralization: The fact is, I believe, that writing mistakes are not made in ignorance of common-sense requirements; they are made for other reasons that advice cannot prevent. Usually, the student thinks he has made a logical transition or a narrative point, which means, again, he is deceived by his egocentricity. What he needs is not rules but awareness. Or if he omits stylistic variation, metaphor, and detail, he does so for a variety of reasons the teacher has to understand before he can be of use. . . . Particular instances of failing to do what one thinks one is doing, and of failing to use the full resources of language, should be brought to light, the consequences revealed, the reasons explored, the need for remedies felt, and the possi- bilities of solution discovered. (pp. 202-203) The three principles that I have discussed so far--hypothesis- testing, risk-taking, and decentralization--all lead to a view of the learner as an active participant in the learning process. This is the fourth principle that must be established--the gct1!§_nature of the learning process. In the modification of cognitive structure, the learner engages in a thinking process in which new information is either assimilated with stored infbrmation, or rejected. In other words, decisions are made as to the perception, selection, and integration of "what's out there.“ Ulric Neisser calls this a "constructive process," in which "The perceiver is active. To a considerable extent he chooses what he will see, selecting some objects for attention and perceiving some of their properties rather than others. . . . By constructing an anticipatory schema [i.e., cognitive structure], the perceiver engages in an act that involves 132 information from the environment as well as his own cognitive mechanisms. He is changed by the information he picks up.”8 A mistaken view of the learner, then, is one that sees him as passive, for it does not consider the deliberate and intentional aspects of the learning process. As Smith emphasizes, "The process of generating and testing hypotheses about the world is all that any child has or needs in order to make sense of progressively more and more of the world around him. But in order to learn in this way, the child must interact with the world. Such learning is active; it involves deliberately seeking information that will facilitate the process of constructing a theory of the world" (p. l27). A view of the learner as active, then, focuses our attention on his own participation in cognitive development. To "educate" students is, as Neil Postman and Charles Weingartner state in Teaching a§_g Subversive Activity, "to elicit from students the meanings that they have already stored up so that they may subject those meanings to a testing and verifying, reordering and reclassifying, modifying and extending process. In this process, the student is not a passive 'recipient'; he becomes an active producer of knowledge."19 - The final principle to consider is the process-orientation of cognitive learning theory--that is, how people learn. There are basically three modes of learning: 1) by doing; 2) by observing; and 3) by being told about something. David R. Olson, in his l973 §ghggl_ Beyj ! article entitled "What Is Worth Knowing and What Can Be Taught,“ provides a useful illustration of each: "One may learn that a stove is hot by (l) touching it, that is, through direct contingent experience . . .; or (2) by seeing someone recoil from touching it, 133 that is, through modeling or observational learning; or (3) by hearing the sentence 'The stove is hot.'"20 These modes of learning complement one another. No one alone constitutes the_mode of learning; they all contribute to the process. As Smith points out, "A teacher's responsibility must be to understand the advantages and limitations of the various modes of learning, and to relate these factors to the objectives of their instruction and the prior knowledge of the children involved" (p. 222). And, in the composition classroom, where the objectives of instruction center around improving written competence, the advantages and limitations of the three modes are clear. What students "learn" when they are "taught" writing is the "skill" of effective expression and communi- cation through writing. How, then, does one learn this skill vis a vis the three modes of learning? Olson contends that while "demonstrations and descriptions have some effect on the acquisition of skills," they "affect action indirectly by indicating the features of the effects or consequences of a performance against which the actual performa- tory attempts can be evaluated" (p. 4l). He argues, in addition, that ”Skills are acquired, not through being told, but primarily through practice, through performatory action coupled with feedback," and that writing skills in particular "must be taught largely on a practice-and-correction or tutorial basis“ (pp. 39-40). Olson's arguments about the limitations of learning through observation and description are echoed by Smith's attitude toward the acquisition of skills: "A skill cannot be summed up in words, though we may be able to provide a learner with helpful hints about what to 134 concentrate on, or about how to perform a particular sequence of operations. . . . skill is the way in which you use what you know or believe, and cannot be directly communicated either through language or through demonstration" (p. 2l8). Students learn to develop writing skills primarily by using them, though they may also learn through observation and description as supplementary strategies. Given the fact that writing is not just g_skill, but many skills, it becomes even more essential that the most effective means of achievement competence in these skills is to attend to them while they are in the process of developing--because students do not develop one skill at one time, but many skills simul- taneously. This emphasis on ggigg_strongly supports process-orientation in the classroom, so that writing is treated in process-~while it is actually being done, or during the working through of completed drafts. And, pppcess-orientatipp complements all that I have said so far in regard to the principles of learning theory. First, it reaffirms the natural model of the learning process--the problem- solving nature of_hypothesis-testing. The central pedagogical impli- cation here is the utilization of the classroom workshop, where, with its trial-error—feedback methodology, writing is treated in process. Second, process-orientation provides opportunities for the rigky tgkigg which is necessary for meaningful learning: conceptual and stylistic errors are thus essential learning devices to be handled g§_thgynggggr_in the writing process. Third, process-orientation promotes decentralization in that it focuses on both conceptual and stylistic conflicts which arise during the process of saying what one 135 wants to say. Fourth, it actively engages the learner in the learning process through constant and guided attention to what he is doing while he is doing it. Process-orientation in the composition classroom attends directly to both the conceptual and stylistic aspects of writing. Piaget's remark that “It may well be through quarreling that children first come to feel the need for making themselves understood" (p. 65) clearly points to the conceptual refinement that occurs in the trigg; . error-feedbapk_methodology. Such refinement can come through feedback offered by peers, as Moffett contends--"Group reactions establish a consensus about some objective aspects of the writing and identify, through disagreement, those aspects that involve individual value judgments" (p. l94). Or, conceptual refinement can result from feed- back offered by teachers, as Perry suggests--"The good teacher becomes one who supports in his students a more sustained grOping, explora- tion, and synthesis."21 Acquisition of writing skill§_is also greatly facilitated by process-orientation, as we see in the model of skill learning which appears in Stephen N. Judy's book, Qgplorations in the Teaching 9_f_ SecondarypEnglish: I will experiment (and get feedback). x / g. . I have I need these these skills. new skills. I can do I want to .//’/;7 this with ~e_~__________,;;.do something them. new. 136 The cycle helps to show how language skills--from basics to rather complex thinking skills--grow and develop. The speaker or writer, whatever his age, has an intuitive knowledge of his bank of skills--that is, what he can do with language. "I have these skills," he says to himselTT'"therefore I can do these kinds of things.“ But no member of the community of language is static; each person is growing, so the speaker or'writer comes to feel, "I want to do new things to express and share my growing experiences." Thus he concludes, "I will need new skills to accomplish this task." As he attacks a new task, ha reinforces his present bank of skills while adding to it. 2 It should be clear by now that the principle of process- orientation is the central tenet of cognitive learning theory and that the other principles I have cited here are corollary dimensions of this tenet. The implications fbr the composition classroom that I have drawn from all of these principles will become more familiar as I consider other relevant theoretical fields. A second essential theoretical area is language theory. I believe that the teaching and learning of language processes-- speaking, listening, reading, and writing--should be based on an adequate understanding of the nature of language. Accordingly, I will, in this section, discuss three key aspects of contemporary language theory--gpamnar, language variation, and language acquisition. My discussion of each will be brief because my concern is with central principles and their implications for composition, rather than with any kind of detailed, comprehensive exposition of the particulars of language theory. The teaching and learning of composition is invariably infbrmed by a theory of grammar. For example, if we subscribe to a prescrip- tjye_grammar, such as the traditional grammar derived largely from Latin, then our view of grammar is that it is bound by rules of 137 correctness--rules which determine whether language forms are "right" or "wrong." Yet such a view of grammar is inadequate in that 1) it is static--it is not concerned with the way language changes or with the way language is actually spoken: and 2) it does not address itself to meaning but instead fbcuses on the surface features of language. A more powerful grammar, therefore, would be one which is both descriptive and also concerned with meaning. Structural grammars, while more adequate because they are descriptive rather than prescriptive, are also limited in that their focus is on the purely fbrmal properties of language--that is, on the observable surface of language. Hence, meaning plays a small part in a structural grammar. Transformational-generative grammar, on the other hand, is the most powerful theory of grammar we have today because it not only describes the surface features of language but it also attends to the underlying meaning components and focuses on the way language is generated. Three main principles of transformational-generative grammar are: l) the distinction between surface and deep structure; 2) the distinction between competence and performance; and 3) the generative nature of grammar. The first principle distinguishes between two levels of language. Surface structure refers to sensory data, the observable linguistic utterance (speech sounds, or marks on paper), while deep structure refers to meaning--the underlying thought structure. By "underlying thought structure" I mean the syntactic and semantic structure which is not necessarily apparent in surface features. 138 While much controversy exists among linguists today over the nature of deep structure, there is, nonetheless, agreement that the principle meaning components of language do not reside in surface structure. Most crucial here, then, is the location of meaning in deep structure, and not in the surface features of language. The second principle is the distinction between linguistic ggmpetence and linguistic performance. Competence refers to the unlimited potential every native user of a language has gained from experience in his linguistic environment-~an internalized set of rules about language which provide the user with an unconscious sense of what is "grammatical" in his language and what is not. Performance refers to what a user does with language--the sentences he actually produces. What is crucial here is the distinction between what a language user gag gg_and what a user actually ggggj fOr we know from linguistic research that linguistic competence far exceeds linguistic _performance. Because every native speaker has a built-in, unconscious sense of what is grammatical in his language and what is not, we cannot judge his knowledge of grammar purely on the basis of his linguistic performanpg. He knows more about his grammar than his performance may indicate. The third principle concerns the generative nature of language. Transformational:generatiye_grammar is a process grammar. It posits that each linguistic utterance has a deep structure which, through a series of transformations, forms a surface structure. Surface structure is governed by the application of transformational “rules" which indicate whether an utterance is grammatical or ungrammatical. It is thus generative in that by fOllowing these 139 rules, we could produce all of the possible sentences in the language. In other'words, transtrmational:generative grammar starts with the parts of an utterance in deep structure and moves to the whole utter- ance in surface structure. The major implication to be drawn from these three principles of language theory is the_ppoce§§:0rientation of transformational- generative grammar. Since meaning is located in deep structure and only to a small degree in surface structure (e.g., a passive trans- fOrmation may slightly alter the deep structure), then the writing process, like the speaking process, is one of transforming deep struc- ture into an adequate surface structure. Thus, in the composition classroom, the most important Operations in the writing process occur at the deep structure level, where meaning is located. Secondary operations on surface structure, then, take the form of stylistic revisions--adjustments in clarity and precision which are necessary to approximate intended meaning in deep structure. Failure to recog- nize this essential distinction too often results in a distorted view Of the process Of composing--that is, the confusion of the writing process with the written product. When Frank Smith comments that many peOple hold a "perverse idea that good writing should spring fully formed out of a writer's head" (p. 192), he is reaffirming the view that the primary operation is the generation of meaning and that the secondary Operation is a matter of revising, or editing, which comes at the end Of the writing process, not during it. What this amounts to finally is a distinction between "writing" and "editing." Students who confuse the two erect unnecessary blocks in their own writing process (e.g., hypercorrection or fear of language); they fail to 140 realize that there is nothing wrong with saying something unclear or "incorrect" (according to the conventions of edited English) while engaged in the process of saying it, and that clarity and "correct- 'ness" come at a much later point. Another dimension of grammar as process-orientation lies in the competenceaperformance principle. If teachers recognize that all students have a basic competence in their native language, then the teacher's job becomes one of helping students have better access to their competence grammar--that is, of activating competence to the level of performance through constant use of language. This is the theoretical rationale behind sentence-combining exercises--that the activity of sentence-combining is one in.which students' unconscious knowledge of the abstract rules of grammar is made conscious through the process of manipulating actual sentence parts. Thus, students are made aware of what they already "know" at the level of competence. In summary, then, the value of increased familiarity with transformational:generative grammar is that it provides insight into the nature Of language and how it is produced and, by implication, how language is produced in the generation of written discourse. The separation of the meaning dimension of language production--geep_ structure--from the communicative dimension--surface structure-- facilitates a more informed view of language processing, particularly the writing process. Another important aspect of language theory is language variation. We live in a linguistically pluralistic culture, one which is marked by inter-dialectal varieties of language use-- regional, ethnic, and social dialects--as well as by intra-dialectal 141 varieties--social language "styles" such as spoken and written, formal and infbrmal, and "consultative" and "intimate."23 1 Given, then, this wide variety of language uses, we would be mistaken to assume, as so many prescriptive grammar texts do, that there is only one "correct" level of usage, only one "standard" way of speaking or writing. And, just as there are many dialects, each more or less appropriate depending on the context in which it is used, so there are many language "styles," each also appropriate, depending on its contexts. As Jenefer M. Giannasi says in her recent essay, "Dialects and Composition," the "teacher of written composition must be aware of the scope, influence, and uses of the various dialects (more accurately, varieties) of the language. *Questions about dialect status, code switching, mutual intelligibility, and social attitudes may be answered only if the researcher, teacher, and composer can differentiate varieties by their situational and contextual cate- gories."24 Another implication of language variation concerns the cultural and personal dimensions. That is, a person's dialect is closely tied to his values; it is an intimate part of his cultural and personal identity and must be so respected. Failure to recognize this simple fact Often results in imposed language standards in the schools and in the problems which go along with this. Mina P. Shaughnessy, in Errors and Expectations, makes clear the problems which can arise when the cultural aspects of a person's language are ignored: When we remember the ways in which the majority society has impinged upon the lives of most [basic writing] students and when we recall the student's distrust of teachers and their language, engendered over years of schooling, it is difficult 142 to see how the desire to identify with the majority culture, and therefore its public language, could possibly have sur- vived into young adulthood. At best we might expect deeply ambivalent feelings about "making it" in a course that teaches what is perceived as an alien dialect. Even the instrumental motive [i.e., a practical use for the language] is likely to be weak among students whg are not yet in the habit of seeing themselves in careers. 5 Nor can the personal aspects of language use be ignored when language standards are imposed. As Lou Kelly points out in a 1974 College Composition and Communication essay, "Teachers who reject a person's language reject the person. Teachers who cling to their Obsession with grammar are not serving the student or the educational system; they are preserving the notion that, though all men are created equal, the language you learn in the home and community where you are created stamps you inferior if it is not 'correct.'"26 Teachers can build on the competence students possess in their own dialects by providing opportunities for students to express themselves in their own dialects or idiolects without undermining confidence in their ability to write. This is the essence of the writing section of the l974 Conference on College Composition and Communication resolution on "Students' Right to Their Own Language": If we name the essential functions Of writing as expressing oneself, communicating information and attitudes, and dis- covering meaning through both logic and metaphor, then we view variety Of dialect as an advantage. In self-expression, not only one's dialect but one's idiolect is basic. In communica- tion one may choose roles which imply certain dialects, but the decision is a social one, for the dialect itself does not limit the information which can be carried, and the attitudes may be most clearly conveyed in the dialect the writer finds most congenial. Dialects are all equally serviceable in logic and metaphor. Perhaps the most serious difficulty facing "non-standard" dialect speakers in developing writing ability derives from their exaggerated concern for the least serious aspects of writing. If we can convince our stuaents that spelling, 143 punctuation, and usage are less important than content, we have removegya major obstacle in their developing the ability to write. This is not to say that "anything goes" at all times, but that in the process of writing, there is as much place for the features of a divergent dialect as there is for the features of Edited American English dialect. When peOple write, they rely on the language forms which most facilitate the expression of their ideas and attitudes; and, when what they have written is intended to be communicated to others, then the accepted conventions of written language can be implemented to bring about the most effective communi- cation. Clearly, I am distinguishing here between writing and revising/editing. As Lou Kelly remarks in her College Cpmposition and Communication article, "We must let them speak their own language on paper, with their own voices, without worrying about the social conventions. Then, to help them get the responses they hope for from the people they want to reach with their ideas, we must help them learn to become competent copyreaders."28 An expanded awareness of language variation can help composi- tion teachers deal more effectively with the divergent dialects students bring to the classroom. Teachers can build on the strength of divergent dialects by defining for themselves and for their students the place of these and "standard" dialects in the writing process. That there is such a place is unmistakable: that most students are aware of it is less certain and perhaps the cause of many writing problems. The third aspect of language theory to consider is language acquisition--the process by which children learn their native 144 language. This is an essential part of a theoretical framework for a language learning situation such as composition because many valuable insights derive from an awareness of the powerful intellectual achievement of children in learning their first language. In this section, therefore, I will attend to the process of language acquisi- tion by examining some of the principles involved in it and their implications for later language learning. All languages operate according to systems of pplg§_which underlie the phonological, syntactic, and semantic components Of language. The central principle in the process of language acquisi- tion, then, is the child's unconscious construction of the pgleg that govern his understanding of the structure of language. The method employed by the child in acquiring these rules is the second principle to consider: he tests hypothesegf-conducts linguistic experiments to discover what regularities occur in the language he is surrounded by. The psychological processing involved in the hypothesis-testing is outlined by Jerry A. Fodor in Smith's and Miller's The Genesis pf Language: In the first place, there is a body of observations that the child must be assumed to make, a body of data aBOut his language provided by the child's exposure to the verbaliza- tions of adults, siblings, and so on. Second, there are whatever learnin principles the child employs to organize and extrapo ate these observations. Third, there is the application of the principles to the data, the body of knowledge about the structure of his language that the child-cum-fluent-speaker will employ in speaking and under- standing the language.29 Another principle of language acquisition is the central role Of_epyironment. While most psycholinguists contend that such environmental factors as imitation, practice, and reinforcement play 145 little, if any, role in language develOpment at early stages, there is some agreement that verbal interchanges between child and adult such as expanding, prompting, and echoing facilitate language development.* What is most important is that the adult facilitates the child's use of language by manipulating and encouraging it. Furthermore, since, in the course of language development, the child is constantly experimenting, he is unavoidably taking linguistic pjgkg, An environment which encourages linguistic 51§k§_ (e.g., an adult who does not constantly correct the grammaticality of utterances) is one in which fluency develops more readily. The fourth principle of language acquisition to consider con- cerns motivation. From a psycholinguistic point of view, the child is primarily motivated to acquire language in order to make sense of his world by controlling it through language (e.g., the past can only be *In expanding, an adult modifies a child's syntactically immature utterance to one of greater maturity, as in: Child: "Give Mommy." Parent: "Give it to Mommy?" Child: "Give it Mommy." In prompting, an adult reformulates a statement which the child may not have understood because of its syntactic structure, as in: Parent: "Where is your cup?" Child: (Silence). Parent: "Your cup is where?" Child: "My cup on table." In echoing, an adult imitates a child's partially unintelli- gible sentence, but substitutes for the unintelligible part something which hopefully corresponds to the child's intended meaning, as in: Child: "I want oom." Parent: "You want your spoon?" Child: "I want spoon." 146 summoned and communicated through language). Whatever other purposes the child may have in using language--practical, emotional, imagina- tive--the primary motive arises from a personal need to use language to regulate and control his world. That the child achieves active mastery over most of the structure of language in just a few short years is testimony to his highly motivated self-direction. Since, as Frank Smith notes, "the way language is learned tells us a good deal about learning in general" (pp. 3-4), the proceSS' of language acquisition provides a model of natural language learning which can inform the teaching and learning of reading and writing. While most writers focus on the relationship between the principles of language acquisition and the development of reading ability, few concentrate on what insights language acquisition holds for the development of writing ability. But since one's native language is learned without any formal instruction (hence, "natural“), the language acquisition model provides implications for the composition classroom. The first implication concerns linguistic competence and the teaching of grammar. The distinction between linguistic competence and linguistic performance is central here: since the linguistic competence of school-aged children enables them to understand and produce grammatically correct sentences, it is questionable whether teaching grammatical principles is necessary--certainly not in the early or middle grades and probably not in later grades either, with the exception of classes where the nature of language is the subject of study. This is not to say that children and adolescents have reached adult levels of sophistication in vocabulary and syntax, 147 but only to suggest that they will reach these levels eventually as their language development proceeds at its natural pace. Furthermore, as I pointed out in Chapter II, research has repeatedly demonstrated that there is no correlation between the teaching of formal grammar and improvement in writing ability. We should, instead, assimilate the natural language learning strategies that the child employs regularly without the help of instruction; he learns his grammar through exposure to language and through hypothesis-testipg(trial-error-feedback). There should, therefore, be plenty of exposure in the composition classroom to new language features (sentence structure, vocabulary) and plenty of opportunity to experiment with them. One essential way of facilitating exposure is through reading. In Teaching English tg Speakers pf English, Bradford Arthur writes that "an almost entirely passive exposure to formal, literary English can lead to the ability to produce this style in 'composed' prose." The child will not merely imitate what has been read; rather, as in the case Of Arthur's third grader, “her passive assimilation Of literary prose enabled her to produce an approximation of literary prose herself, with little difficulty and no special training or encouragement."30 James Britton also notes the value of reading for learning writing when he says that the effect of reading upon writing is that "As in their speech, children 'absorb' a knowledge of the grammatical system of the English language without ever having made that language explicit, so they take in the same kind of awareness 31 of the way the written language works." Along with reading, 148 exposure can be increased through talking, a language activity which may introduce new language features, help initiate latent features (in student competence), and thereby facilitate linguistic pegfgp: gapgg, Clearly, then, the development of writing ability can be fOstered naturally through a close integration of all the other language functions--reading, talking (and, by extension, listening)-- in a holistic framework. As James Britton remarks, "I believe that the continuous use of language by speaking or writing or listening or reading, the use of continuous language, is the really productive factor in all language work" (p. 7l). A second implication concerns the role of meaning in language acquisition. Since children learn language through hypothesis- testipg in a meaningful context, our emphasis in the composition classroom should fall primarily on the meaning of utterances and only secondarily, if at all (depending on the purpose for the writing), on correctness. Furthermore, in keeping with the idea of_bypothesis- testipg, it is only natural and necessary that mistakes will be made, in the areas Of both grammar and semantics. Mistakes are particularly useful in the composition classroom, as James Moffett points out: "The learner simply plunges into the assignment, uses all his resources, makes errors where he must, and heeds the feedback. In this action-response learning, errors are valuable; they are the essential learning instrument" (p. 199). Another implication which follows from the previous one is the view of the child as an aptly; agent in language development. The child is not a mere passive mechanism while learning language, as imitation theories hold; rather, he is constantly engaged in 149 linguistic experimentation. This role of the child as active agent should be maintained in the composition classroom. Teaching the ' grammatical and rhetorical rules of language use puts the student in a passive role, encourages what Moffett calls “error-avoidance," and takes valuable time away from actually using language. Frank Smith notes that "the skills of saying something efficiently, and of having something to say in the first place, are best developed by being put to use" (p. l9l). Accordingly, students should be encouraged to write a lot (free writings, journals, compositions), in many forms (expository prose, fiction, poetry, etc.) and for real audiences (classmates, teachers, others). A fourth implication of language acquisition for teaching writing is the role of environment. The child takes many linguistic risks when learning language, makes many "mistakes," and learns by these "mistakes," all of which go on in an environment encouraging linguistic experimentation. If we want students to take the risks which are necessary for meaningful learning, then we must encourage an environment which is free from fear of making mistakes. As Frank Smith points out, "effective learning has little chance of taking place without the possibility of error on the part of the learner' and error frequently demands a price" (p. 226). This view of an environment which encourages risks has two applications. First, we must ask ourselves if language development is being encouraged when too much attention is given to surface features at the expense of the content of utterances. Second, we must examine what we see as the intellectual functions of writing. On the one hand, the intellectual function of writing is the effective 150 communication of ideas; here, the focus in on intellectual clarity in the written product. On the other hand, the intellectual function Of writing is also to serve as a means toward achieving clarity--that is, the writing process itself functions as a way of promoting intellectual development. Here, the focus is on intellectual clari- fying in the writing process through risk-takipg. As James Britton says,in "the essential process of sorting experience which goes on in the writing . . . we give [the student] the Opportunity to write above- all because this encourages him to cope with something that is an immediate concern, an immediate problem to him; he is dealing with a part of the pgy_for himself and we put the stress upon that, at the same time admitting that he will learn to write by writing" (pp. 28- 29). If we encourage students to take intellectual (as well as linguistic) risks, then we must shift our focus from the written product to the writing process. And, if our students are to see the process of writing both as a means of shaping experience and as a means of learning from it, then it follows that we should stress the process of writing as a way of facilitating intellectual development. That is particularly true on the secondary and college levels where demands for intellectual sophistication are greatest. As I pointed out in my earlier discussion of William G. Perry, Jr.'s research on the intellectual and ethical development of college-age students (see pp. 129-l30), an environment which supports intellectual exploration greatly facilitates the growth and refinement of intellect. Finally, by encouraging intellectual risk-taking in the process of writing, we may well generate in our students the kind of vigorous, 151 genuine writing that usually accompanies strongly motivated, personally meaningful utterances. The final implication to be drawn from language acquisition again concerns motivation. The efficacy of strong, personally moti- vated self-direction gives us a target for the composition classroom. Frank Whitehead, in The Disappearing Dais, points such direction for writing teachers: In writing as in speaking, what matters more than anything else is that one should have something to say that one really cares about saying. It is this powerful impetus from within that is essential in order to carry the prentice writer past the obsta- cles and difficulties of his task; yet all too Often this is entirely missing from the attitude with which the secondary school child approaches his weekly composition homework. The first aim of the teacher must be to manoeuvre the child into a position where he feels this impetus within himself; somehow he has to engender in his pupils an urgently felt impulse towards communication, moreover, which demands for i s fulfill- ment the permanence Of the written form of language. In this section, I have focused on the process of natural language learning because I feel that an awareness of this process provides useful insights for the teaching and learning of writing. The principles and implications I have drawn from the principles of language acquisition are operationally real: I have seen them applied on the elementary, secondary, and college levels. When students have the Opportunity to use their natural language learning strategies, they learn, and learn well. Composition theory is a profoundly ill-defined field which encompasses a number of disciplines ranging from traditional fields such as philosophy, history, and rhetoric to more recent and rela- tively unexplored ones such as speech act theory, tagmemics, and psycholinguistics. While I haven't the space here to consider how 152 all such areas contribute to composition theory, I will attend to three key aspects of composition theory which relate to formulating a theory of instruction for composition. These are: l) structure of composition; 2) seguence in composition; and 3) the composipg_process. In Tradition and Reform j_r_1_ the Teachim g_f_ English: A History, Arthur N. Applebee argues that the specification of goals for instruc- tion in English must derive from an answer to what he calls the “perennial question“--"'What is English?'": To answer it is to specify implicitly which goals are central and which of lesser importance. If, for example, English is defined as a set of mechanical skills in language use, a goal such as "good spelling" may emerge near the top of the hier- archy. It becomes important in itself and instruction can be focussed directly upon it. This has in fact sometimes happened because spelling has been defined as a mark of a good educa- tion; students have been tested and drilled in spelling for its own sake. If, however, English is defined as a way to order and understand the world through language, then spelling becomes a secondary goal. The focus of instruction will be on using language in a significant exploration of the world, with spelling simply a skill which is useful but not central in that process. Though spelling may still be taught directly, such teaching will have to be assessed in terms of its effect on the larger goal rather than simply in terms of improvement in spelling ability.33 Once we can define "English," then, we can determine its structure. John Dixon suggests some possible definitions in Gypytp Throggh English, his report on the Dartmouth Seminar, the l966 Anglo- American Conference on the teaching of English. Dixon rejects overreliance on either writing éhill§.°F knowledge of literary and cultural heritage as the ultimate definition of English because each isolates the other. He then proposes that English is to be defined not as some thigg, but "by process, a description of the activities 34 We engage in through language." For Dixon, then, English is neither exclusively skills nor knowledge, but rather the process of 153 acquiring both. And, by extension, composition is not a product like a skill or a packet of knowledge but a process, as Dixon remarks: "The question 'What is English?‘ invites a different form of answer from, say, 'What at our best are we doing in English classes?‘ If we wish to describe a process, composition for example, the first question will tend to suggest the finished product (the marks on the page even) rather than the activity of bringing together and composing the disorder of our experience. 'What . . . doing' will suggest nominal forms of verbs (bringing, composing) and thus help to keep activities in mind."35 In a similar fashion, James Moffett, in Teaching the Universe .pf Discourse, argues that a definition of English as either skills or content creates an inadequate dualism which fails to recognize the integrative nature of the two. Instead, Moffett reintegrates skills and content in what he considers to be the definition of English--"all discourse in our native language" (p. 9) regardless of its content (e.g., literature, history, biology, etc.). Furthermore, all discourse consists of three interdependent elements which com- prise its structure: 7 The elements of discourse are a first person, a second person, and a third person: a speaker, listener, and subject; informer, informed, and information; narrator, auditor, and story; transmitter, receiver, and message. The structure of discourse, and therefore the super-structure of English, is this set of relations among the three persons. But in order to exploit this venerable trinity, we must get beyond its innocent look. Within the relation of the speaker to his listener lie all the issues by which we have recently enlarged the meaning of "rhetoric"--what A wishes to do by speaking of such and such a subject to B. Within the relation of the speaker to his subject lie all the issues of the abstractive process-- how the speaker has symbolically processed certain raw 154 phenomena. But of course these two relations are in turn related: what and what for are factors of each other. As with all trinities, the FETations of persons is a unity-- somebody-talking-to-somebody-about-something. And, lastly, within the relation of the listener to the subject lie all the issues which we call comprehension and interpretation. In proposing this structure, I am thinking that the student would learn the skills of operating our symbol system by role-playing first and second persons in all the possible relations that might exist between the student and a subject, and between him and a speaker or listener. (p. l0) For Moffett, then, English is all discourse; and, a student learns English through the process of shifting roles among the elements of discourse. Like Dixon, Moffett defines English by pro- cess, or activity, where language is used in real, authentic discourse (somebody-talking-to-somebody-else-about-something). I see Moffett's definition of English as a more precise articulation of what Dixon means when he defines English as a process. For both writers, the focus is on language in Operation, language used for real purposes and directed toward real audiences. Shifts among the three "persons" are the very stuff Of composition, for such shifts inform every aspect of composing--style, point of view, structure, order, and so on. The structure of composition, then, is the nature of discourse --the shifts among the three persons. I find Wayne Booth's definition of the "rhetorical stance" most appropriate in this regard because it succinctly places the three elements of the nature of discourse squarely within the field of composition: The common ingredient that I find in all of the writing I admire . . . is something that I shall reluctantly call the rhetorical stance, a stance which depends on discovering and maintaining in any writing situation a proper balance among the three elements that are at work in any communica- tive effOrt: the available arguments about the subject itself, the interests and peculiarities Of the audience, 155 and the voice, the implied character, of the speaker. I should like to suggest that it is this balance, this rhetori- cal stance, difficult as it is to describe, that is our main goal as teachers of rhetoric. Our ideal graduate will strike this balance automatically in any writing that he considers finished. Though he may never come to the point of finding the balance easily, he will know that it is what makes the differegge between effective communication and mere wasted effort. This passage also points to the structure Of composition as a process-- "discovering and maintaining"--where the success of the written pro- duct depends exclusively on attention to the interaction Of the three elements of discourse in the writing process. To summarize this view of the structure of composition, then, English is defined as process and includes all discourse; and the structure of composition is the nature of discourse--the interaction of the three elements of discourse in the writing process. The second aspect of composition theory concerns seguence. Is there any sequence in the development of composition abilities indigenous to the structure of composition and applicable to a theory of instruction? In order to answer this question, I must extract three principles from the structure of composition as I have defined it. These three principles are: l. That language be used, not studied as an Object in itself. 2. That language be used as it really exists-~in a speaker- audience-subject context. 3. That language be used as it occurs naturally--in a variety of contexts and for a variety of purposes. Many composition theorists and teachers advocate a sequence marked by such variety of authentic discourses, in which at one end we have completely personalized, private discourse (e.g., the 156 self-directed monologue), while at the other end we have imperson- alized, public discourse (e.g., the legal document). Such a sequence would cover the range of discourse situations and purposes within the rhetorical stance. A brief examination of the versions of this sequence proposed by composition theorists will help to make it more concrete. James Britton, for example, proposes a sequence of modes of writing which begin with expressive writing and develop into either referential or poetic writing. In expressive writing, the writer's audience is, for the most part, himself; there is little concern with the amenities of public written discourse: and the modes usually include diaries, journals, and other forms of self-directed writing. Referential writing, on the other hand, is directed toward an audience other than the self; there is much concern with the communicative conventions of public written discourse; and the modes include expository prose and other forms of informative discourse. Egetip_writing is literary; concern with conventions takes the form of a verbal construction--a verbal performance; and the modes include fiction, drama, and poetry. The interesting part of Britton's model is his location of expressive writing as the starting point for either referential or poetic writing: Referential a~£xpressiveH—~ LPoetic "Expressive writing" writes Britton, "is for us, then, the center point--still a kind of matrix, tending to be on the move, either to referential writing on the left, or to formal, poetic writing on the right“ (p. 43). Britton believes that a sequence of writing 157 instruction should begin with the expressive and proceed develop- mentally toward either end. This is particularly true, notes Britton, in the composition classroom, where the central concern is usually with referential writing: You cannot, I believe, teach the referential, the expository, as such: it has to be arrived at, it seems to me, by the shedding Of certain aspects of the expressive. The shedding process is highly intricate: it is as though there existed a delicately adjusted threshold which allows the integrity and individuality of the writer to move through into the writing yet leaves the finished product Objective, referential. By short circuiting the process, I believe we produce the form of expository writing without the vigour, the personality, Of a writer--a linguistic tool which will have very limited uses indeed. (p. 48) Britton's sequence, therefore, begins with expressive modes of writing and moves to public modes in order to sustain the personal sense Of the writer in the written piece. His is essentially a humanistic rationale which is aimed at developing the kind of personal "voice" which gives a distinctly personal flair to public discourse. Janet Emig proposes a similar sequence in The Composing Processes pf Twelfth Graders. She recasts the Britton model as follows: Expressive field of Reflexive < discourse > Extensive For Emig, all writing is primarily expressive--that is, it "expresses the thoughts and feelings of the writer in relation to some field of 37 discourse." Expressive writing can evolve toward the reflexive, which is "the mode that focuses upon the writer's thoughts and 158 feelings concerning his experiences; the chief audience is the writer himself; the domain explored is often the affective; the style is tentative, personal, and exploratory." Or, expressive writing can evolve toward the extensive, which is "the mode that focuses upon the writer's conveying a message or a communication to another; the domain is usually the cognitive; the style is assured, impersonal and often reportorial" (p. 4). Emig argues that American secondary schools give too much attention to extensive writing and too little attention to reflexive writing, which results in a "limited, and limiting, experience" (p. 97). She prOposes, therefore, a range of writing experiences which includes both modes of writing. Emig's proposal for a breadth of writing experiences suggests the kind of sequence I am proposing here, one which includes a range of modes as it moves from the personal to the public. James Moffett presents a finely articulated sequence which is based on a human development model: "Ideally this sequence would correspond both to [the student's] own intellectual and emotional growth and to some significant progression in 'symbolic transforma- tion,‘ as Suzanne Langer has called the human processing of the world" (p. 13). Moffett's sequence is founded on different kinds of discourse, where the student moves “in his writing and reading from one kind of actual discourse to the next in a sequence which permits him to learn style, logic, semantics, rhetoric, and literary form contin hoffet m 'U‘UQIr-Pc-f—{l—hc—fmt-O-CL‘U‘D / —O c- O a"!- (D 159 continuously through practice as first or second persons"* (p. 13). Moffett's sequential model becomes a series Of movements: l. From the implicit, embodied idea to the explicitly formulated idea. 2. From addressing the small, known audience like oneself to addressing a distant, unknown, and different audience. 3. From talking about present objects and-actions to talking about things past and potential. 4. From projecting emotion into the there-then to focusing it on the here-now. 5. From stereotyping to originality, from groupism to individuality . . . (p. 57). At the heart of Moffett's sequential model is the process of abstraction which goes on constantly in the "somebody-talking-to- somebody-e1se-about-something" interaction in the nature of dis- course. Moffett also refers to this as the "I-it-you" relation: The referential relation of I-it must be crossed with the rhetorical relation of I-you, in order to produce a whole, authentic discourse. Rhetoric, or the art of acting on someone through words, is an abstractive act. That is, one performs the same activities in pitching a subject to an audience as one does in extracting that subject from raw phenomena: one selects and reorganizes traits of things, digests, codes preferentially. A course in rhetoric teaches how to present material successfully, how to find subjects; how to choose words and sentence structures, how to enchain items in sequence and patterns. Both abstracting from and abstracting for concern the same kinds of choice. The difference is whether the speaker-subject relation or the speaker-listener relation is determining the choice-- the extracting from the source or the anticipation of audience response. Representing reality to oneself and presenting it to others are merely two aspects of the same process, which is abstraction. Once coding is verbal, we are hard put to conceive of it as solely abstracting from. *"Persons," here refer to the elements of discourse. See quote from Moffett, above, pp. l53-l54. 1.1.!!! 160 In fact, I will make the assertion that neither abstracting from nor abstracting for exists apart from the other in the universe of discourse. 'Composition' means handling both dimensions at once; a speaker always stands in some relation to both his subject and his audience. (pp. 31-32) Moffett's sequence Of activities--his "spectrum of discourse" --takes the theoretical form of a hierarchy of levels of abstraction and the practical form of a wide range of writing options which follow the series of movements I outlined on p. 159. Interior Dialogue (egocentric speech) Vocal Dialogue (socialized speech) Correspondence Personal Journal Autobiography Memoir Biography Chronicle History Science Metaphysics Clearly, Moffett's sequence, like those of Britton and Emig, begins with the self and moves outward, in a developmental pattern which parallels the psychology of the learner: The primary dimension of growth seems to be a movement from the center of the self outward. Or perhaps it is more accurate to say that the self enlarges, assimilating the world to itself and accomodating itself to the world, as Piaget puts it. The detailed forms which this movement takes are various and often paradoxical. In moving outward from himself, the child becomes more himself. The teacher's art is to move with this movement, a subtle act possible only if he shifts In 161 his gaze from the subject to the learner, for the subject is in the learner. (p. 59) With Moffett's sequential model, we arrive at the intersection of a structure fOr composition--the nature of discourse--with a sequence for composition--a movement from the personal to the public, from the less abstract to the more abstract, and from the self in relation to itself to the self in relation to the world. Such a sequence may be indigenous to the structure of composition in that it embodies the essential principles of that structure that I stressed on p. 155: that language be used; that language be used as it really exists--in a speaker-audience-subject context; and, that language be used as it occurs naturally--in a variety of contexts and for a variety of purposes. A third aspect of composition theory is the nature of the composipg_p:pcess. Implicitly or explicitly, whatever view teachers hold of the composing process will inform many of their classroom strategies. Similarly, whatever view students hold of the composing process will inform many, if not all, of their writing strategies. Thus, a theory of the composing process is essential both for teachers and for students. The composing process has been described, however, in a variety of models. A 11222! model, for example, conceives of a uni-directional process which can be divided into stages, such as choosing and narrowing a subject, outlining, writing, and revising. Prewriting, writing, and rewriting is another, more familiar way of describing this process. A second model is what might be called cyclical, in which the "stages" of writing are less discrete. The 162 writer continually fluctuates between matters of content, style, and structure in what is essentially a recursive give-and-take between form and content, with formal methods of invention assisting in the process. A third model is the discovery conception of the composing process. Here, the act of writing itself serves as a way of genera- ting content and style. The writer simply writes to find out what he wants to say and how he wants to say it; he is governed by no set procedures other than making discoveries through the act of writing, which is viewed as an organic process. These three models focus on the generative operations of the mind in the conposing process. An alternative to these is the behavioral model, which is less concerned with the mind as the genera- tive source of thought and language and is more concerned with the writer's interaction with the external environment as the stimulus Of thought and language. That is, the writer relies on talk:-with other writers and speakers--before and during the composing process. Talk becomes a way Of both discovering and refining thought and language. Clearly, then, the composing process can be viewed in many ways. Each view has its benefits as well as its limitations. The 11222! model, for example, works well when the writer knows enough about his subject to discern its structural and developmental patterns or knows just how he wants to "say it." But it is an inadequate model when the writer does not possess such knowledge--that is, when he must discover his subject's ordering principles by writing or talking his way to them. Similarly, the eyclical, discovery, and behavioral models are most helpful when the writer is in the position 163 of not knowing what he thinks until he says it, as E. M. Forster put it. In short, all models of the composing process bear on the act of writing in its largest context. They are not mutually exclu- sive but rather complementary. It would appear, then, that the best model of the composing process is what I refer to as the integrative model--one which overcomes the limitations of subscribing to any one model by recognizing the strengths of all and relying on these strengths in various writing situations. The integrative model is more useful for two reasons. First, it is comprehensive--it embraces all models of the composing process. Second, it is flexible enough to become relative to the writing task at hand and adaptive to many writing situations; that is, it helps students to prepare themselves for the universe of writing situations they face in their everyday lives by making them aware of many ways to proceed in diverse writing tasks. In summary, I have proposed that the structure Of composition is the nature of discourse--which is itself an operational definition of the communicative process. I have proposed a sequence in composi- tion which is learner-centered in that it is patterned on stages of psychological growth. Finally, I have proposed an integrative model Of the composing process which is more comprehensive and flexible than is any singular model; adherence to any one model, I have argued, simplifies and constricts the complex Operations of the writing pro- cess while it fails to recognize the immense diversity of real writing situations. lea ove kee thi; mig chal 1 w of - tea: l‘ese "'It Luna IEar rESe 164 In my extrapolation of principles and implications from learning theory, language theory, and composition theory, a number of overlapping themes have emerged. This is as it should be, for in keeping with the "explosion of knowledge" discussed at the start Of this chapter, we now know that many common insights derive from what might appear to be diverse and unrelated fields. Here are the insights common to the three theoretical fields I have examined in this chapter: d . The role of process in learning, language, and composition. The prpblem-solving nature Of learning. The trial-error-feedback model Of_hypothesis-testing. Conceptual and stylistic risk-taking. Conceptual and stylistic gegentralization. The role of environment in meaningful learning. The meaningrcentered context of learning and language use. The active nature of the learner and the learning process. KOCDNOSU'IéUN The role of audience in language use. I will have more to say in regard to these common insights at the end of this chapter when I formulate a theory of instruction for the teaching and learning of composition. But first, I will consider research in composition once again to discern its common ground with my discussion Of theory. In Help fOr the Teacher gj_Written Composition, Sara W. Lundsteen presents twelve recommendations for the teaching and learning of composition. These twelve recommendations derive from research conclusions over the past fifty years as well as from the 165 conclusions formed by leading spokesmen in English and the language arts. Lundsteen proposes that these twelve points "could stand as a platform upon which to build programs for children and young people." Lundsteen's twelve points, which appear below, offer many directions for the composition classroom which are soundly based on research and much critical thinking: 1. Oral lan ua e base. Written composition needs to be tied to oral language. Conversation and "free talk" are the basis for consciously structured reporting, story telling and retelling, both original and from other sources. The confidence and fluency that stem from composing with spoken words and body language is essentially related to composing with written words. Dramatics and pantomine add other dimensions to the composing process and promote vividness of characterization and economy of action, as well as clearer conceptualization of plot and episode. Partnerships in writing Spring almost spontaneously from person- tO-person talk and from dramatization, whether spontaneous or planned. No doubt the most Obvious and helpful link between oral language and written composing involves dictation by a child to a teacher who puts the child's spoken words into visible written form. Whether done on an individual or a group basis, hearing and seeing one's unique combina- tion Of words--whether a question or phrase, picture caption or whole sentence--is an exhilarating experience. This method also introduces children to the concept that written symbols stand for oral symbols. Gradual growth from dictation to shared writing to independent writing seems a natural sequence for a great majority of ready young learners. Environment. A varied environment that stimulates many kinds of creative reSponse adds depth and increases potential for selection of content. Experiences with music, dance, paint, clay, drama, rhythmic expression, and other media foster zest for experiment and inventive- ness in the whole realm of curriculum--including writipg, We need to spend more time on what happensbefore a child writes. Inner motivation. Motivation to communicate comes from within. It is innate. Shared by humans of every culture, and apparently by some lower mammals, this urge is universal. For humans, writing is part of the urge. Oral languages historically advance into written forms; roughly 2000 of the more than 3000 oral languages now 166 in existence have gradually developed to the point that they have written forms. Teachers cannot “motivate" children to write; they can only stimulate them. When actually used to relate to peers and adults, children's writing is a vehicle both for self-expression and for affecting their audience. In almost any setting, children's unassigned writing exceeds their writing on assigned tOpics. Individual selection is not only possible but requisite, even when part of a COOperative authorship of books or letters or reports. Composing in writing is an intensely individual process. The individual writing conference between teacher and child may assist this process. The contribution pf children's literature. Children's Titerature can contribute greatly to the written and oral composing of children. Awareness in children of what a story is grows from early exposure to stories heard and read. From the earliest here-and-now accounts of Objects and events shared with infants and toddlers, through the "safer" folk tales and (at somewhat later ages) tales made of sterner stuff, through imaginative and realistic literature, the oral and written heritage is a vast resource for children to draw upon in their own composing. Using known characters for new exploits can help many children to be truly original; for example, a popular dilemma from literature children Often use is that of being the youngest or weakest, yet being able to achieve or overcome. Audience. Various audiences help to shape the style and content of writing. Stories written for younger friends or classes have characteristics quite different from those written for older students or adults. This adapta- tion of writing style for different audiences holds whether the writing is factual, imaginative, speculative, or persuasive. And, since the product belongs to the producer, a story, poem, or any other writing should not be given to an audience until the young author releases it to that audience. Positive response. Enjoyment of stories and reports, appreciation for a bit of original phrasing or a unique character or event, is the response to be encouraged. This reaction is positive and yet selective. Rating and grading have no part in unifying writer and audience. Appreciation of factual accuracy and questions reflecting a new curiosity awakened by a report show that an author has respected and affected his or her audience. Such positive reaction from teacher and peers is evaluative and is the kind of stimulus that builds motivation for further writing.' Teachers help children by looking for strengths and pointing them out. ' 10. 11. 167 Negative criticism should be avoided because it implies rejection. Red-penciled correction and authoritarian com- ment can thwart the confidence needed for further explora- tion. Editing with a child when a paper is to be put into "good" form for display or in a permanent collection can and must be a supportive relationship. The purpose of editing is to help young authors say what they want to say. Even another child serving as editor can adapt a sharing attitude for the purpose of making the writing clear, precise, and easy to understand. If serious reshap- ing of a sequence of ideas is needed, this should be done on a separate paper. Respect for children's own words is thus maintained, and planning or accepting of a better approach is less likely to damage self-esteem. Drafts. Children's first drafts are usually messy. Words are omitted or spelled wrong; handwriting is Often poor. This is often true of many adult authors. One of the truisms Of composing is that ideas forging ahead of one's pencil or typewriter cause many surface errors. Oral display. Not every piece of writing needs to be corrected or copied. Much of a child's writing is best read aloud, if the child permits, and filed in a private folder. An audience might be large or it might be as small as one peer or one teacher. If the story, verse, letter, or report is to be read by persons other than the teacher-intermediary, then editing and writing in appro- priate form are usually necessary. Seeking the author's approval for the finished copy helps to build pride, a strong force for further interest in writing. DevelOpmental irregularity. Development in writing occurs iniirregular spurts. Although learning curves may appear when exact test scores are smoothed into a growth picture, such ratings are neither accurate nor appropriate for composition. Not every story or other piece Of writing is better than the preceding one for a child pg for a professional. Teaching needs to be based on developmental knowledge of children's composition. Observation. Developing powers of observation is essential to the writing process. Welcoming oral comment upon observations strengthens abilities needed in composing-- for example, how the sand looks or feels when it blows, how birds fold their wings when they alight, why people prefer to be in groups rather than alone, or how an author makes an idea clear. Voice. As children mature in supportive environments, they develop an individual "voice." They must be helped to understand who they are (in positive terms) and what values 12. 168 they stand for in order to develop their own style and project their creativity into their products. Creative problem-solving. Creative problem-solving, an important part of composition, can strengthen essential processes of selection. The word creative implies child autonomy, child choice, some areas of the unknown, not just being handed a writing task, topic, or problem. Teachers, of course, can provide frameworks for writing problems ("Why don't you try composing a tall tale, some- thing like those we've been reading and talking about."), but the substance of the composition needs to come from the child's own Observation and imagination.38 To Lundsteen's twelve points I would add the following addi- tional points which derive from the research I have compiled in Chapter III: 13. 14. 15. 16. Fear pf lan ua e. Hostility or indifference to writing OTTEn stems from rejection Of divergent dialects in the classroom. Also, students are Often inhibited in acquiring written fluency because of an over-emphasis on "correctness" in written composition. (See Question Three in Chapter II.) Writipg Labs/Workshpps. Students make greater gains in writing competency when classroom procedures are based on a workshop model. A small class meeting in an infor- mal atmosphere with intensive tutoring available greatly facilitates develOpment of composition skills. Class sessions marked by in-class guidance of writing assign- ments and group problem-solving of writing tasks focuses on writing in process and provides opportunities for problems to be met when they occur. (See Question One in Chapter II.) Collaboratiye_learning. Peer interaction in the composi- tion classroom is an effective pedagogical strategy. Students can collaborate at all stages of the writing process, from prewriting to revising and editing. Colla- borative learning is particularly effective for average and below-average composition students, and a powerful motivational device for students at all ability levels. (See Question Five in Chapter II.) Readin . A close relationship exists between reading bac ground and writing competence, as Lundsteen demon- strates in her comments on the contribution of children's literature. Though researchers cannot say precisely why, 169 findings indicate that effective writers most Often read more than less effective writers. (See Question Nine in Chapter II.) 17. Sentence combining. If elaborate sentence structures are considered positive increments to writing competence, then sentence combining exercises can promote syntactic fluency and can improve the overall quality of student writing. Thus, such activities can be a useful and effective supple- ment to other activities in the composition classroom. (See Question Seventeen in Chapter II.) 18. Writin _preferences. Professional writers indicate that the or1gins of their ideas lie in their backgrounds and personal lives. Inquiries into student preferences indicate some agreement on interest in writing about topics which affect their own lives in some way. Since research has indicated that a relationship exists between the quality Of writing and the writer's interest in his subject, then student writing preferences should be given consideration in the classroom. In addition, since there is a wide spread Of specific writing interests among students at all levels, then equal consideration should be given to providing a variety of writing topics on any writing occasion. (See Question Two in Chapter II.) 19. Pedagogy. As I indicated in my examination of many of the studies under Question One in Chapter II, certain pedagogi- cal procedures affect both writing performance as well as attitudes toward writing. More important, however, is the long-range effect pedagogy has on the reciprocal nature of performance and attitudes, for it is clear that some pedagOgical procedures facilitate the development of writing competence more than do others, and thereby promote a continued and renewable interest in the development of writing proficiency. (See Question One in Chapter II.) These nineteen points, then, illustrate strongly supported directions in composition research. None are based on isolated experimental studies; rather, they derive from a number of research studies which come to similar conclusions. They apply to a range Of levels of schooling and to a range of instructional conditions. Along with the insights which derive from the theoretical fields that I summarized earlier, these nineteen points must be taken into considera- tion in curriculum design at all levels. 170 Since curriculum usually implies specific methods and materials, I prefer instead to develop a "theory of instruction"--a phrase popularized by Jerome S. Bruner. According to Bruner, a theory of instruction "sets forth rules concerning the most effective way of achieving knowledge or skill. . . . A theory of instruction . . . is concerned with how what one wishes to teach can best be learned, with improving rather than describing learning." Thus, a theory Of instruc- tion is concerned with the process of learning:' And, it is intri- cately related to pertinent theory, as Bruner points out: "a theory of instruction must be concerned with both learning and development and must be congruent with those theories of learning and development to which it subscribes."39 Though a theory of instruction is less concerned with specific methods and materials, these matters are still involved, but only to the extent that they relate to a theoretical framework. The focus must be on the most effective means for facili- tating learning. Bruner establishes four components of a theory of instruction: 1. Predispositions. This refers to attitudes which facili- tate the learning process--"learning in general or a particular type of learning. For example, what sorts Of relationships with people and things in the preschool environment will tend to make the child willing and able to learn when he enters school?" (pp. 40-41). 2. Structure. By this, Bruner means the underlying principles Of a subject, or, ways of thinking about the fundamentals of a subject. 3. Se uence. This refers to a developmental way Of presenting whatever is to be learned. 4. Reinforcement. Bruner uses this term in the generally accepted sense--"the nature and pacing of rewards and punishments in the process of learning and teaching" (p. 41). But he does not intend the term to be taken in a behavioral psychology sense; rather, he refers to 171 reinforcement in the cognitive sense Of knowledge of the results of a learning trial which satisfies some inner need for such knowledge: "Learning depends upon knowledge of results at a time when and at a place where the knowledge can be used for correction . . . 'Knowledge of results' is useful or not depending upon when and where the learner receives the corrective information, under what conditions such corrective information can be used, even assuming appropriateness of time and place of receipt, and the form in which the corrective information is received" (p. 50). A theory of instruction, then, because it i§_theoretically oriented, has greater applicability than a specific curriculum or instructional model. That is, it is more concerned with ppy_some- thing is learned than with what is learned. It is thus a more valu- able guide to curriculum design because it is learning-centered rather than materials or methods-centered. As such, a theory of instruction is primarily geared toward the learning process, as Bruner emphasizes: "a theory of instruction seeks to take account Of the fact that a curriculum reflects not only the nature of knowledge itself but also the nature of the knower and of the knowledge-getting process. . . . Knowing is a process, not a product" (p. 72). I turn, finally, to a theory of instruction for the teaching and learning of composition. On the basis of all that I have dis- cussed previously in this chapter--the principles and implications of learning theory, of language theory, and of composition theory and the insights derived from research in composition--I will formulate a representative theory of instruction which integrates both theory and research. As a frame of reference, I will use the four components of a theory of instruction prOposed by Bruner. Though I am mainly concerned with the principles of a theory of instruction for composi- tion, I cannot avoid occasionally referring to methods and materials. 172 What is most essential is that the use of particular methods and materials be informed by a theory of instruction which is solidly based on relevant theory and supported by research. In short, methods and materials abound in curriculum; what is needed is sound reasons for using them. The first component of a theory of instruction for composition concerns predispositions toward learning. The goal here is to develop attitudes which facilitate the learning process in the most effective ways. In this regard, we may recall the powerful motivational forces at work in "meaningful learning" in general and in language learning in particular. In both instances, the learner is confronted with a problem-solving situation which creates its own inner motivation--to find a solution in a trial-error-feedback procedure. Thus, genuine problem-solving becomes a model for fostering learning predisposi- tions. While I must admit that artificiality is difficult to entirely remove from the composition classroom, problem-solving situations such as the following can provide meaningful occasions for writing prob- lems: "Can you write a story about your camping trip for the other kids in class to read?“ or, “Can you write a script for a radio play that you and your classmates can produce?" or, "Can you polish up this story for other readers so we can put it in our class anthology?" or, "Can you make your audience draw appropriate inferences from a paragraph based on purely observable facts?" or, "Can you make this same point, but for the campus newspaper audience, who know less about your subject than your present audience?" Problem-solving situations such as these focus on writing in a meaning-centered context (not an "exercise," but a real discourse); 173 they involve students in the dynamics of writing for real audiences; and, they initiate the cyclical model of skill acquisition that I presented on p. 135. This is precisely the kind of creative problem- solving Lundsteen refers to when she cites research supporting this approach to composition: The selective processes in creative problem-solving parallel some of the composing processes. Some parallels include the gathering and selection of details (observations or facts), the planning and selection of procedures, and the planning and selection of ways to evaluate the results, or consequences ("Did my 'funny' story make the group laugh?") In essence, children can apply what they know about produc- tive problem-solving to composing in writing. They need that same quality of creative autonomy and that same selectivity that they have used before on unknown, undecided, and unmastered areas. . . . (p. 5) Inherent in problem-solving are three principles which I cited throughout my examination Of learning theory, language theory, and composition theory: risk-taking; decentralization; and meaningfulness. Earlier in this chapter I discussed how conceptual risk-taking --readiness to expose one's ideas, however half-formed, to an audience--and how_§tylistic risk-taking--writing without fear of making mistakes in rhetorical, stylistic, or mechanical aspects of composing--can be promoted as a natural part Of learning to write. Research support for conceptual risk-taking_derives from Piaget's work with children and from Perry's work with older students,40 both of whom found that intellectual maturity is closely related to one's willingness to expose one's ideas to the scrutiny of others. Further- more, the research by Walter Loban in 1961 shows a close correspondence between conceptual and etylistic risk-taking. The results of the Loban Project are summarized in the Braddock Report as follows: "'Those subjects who proved to have the greatest power over 174 language . . . were the subjects who most frequent1y_used'language.tg express tentativeness. Supposition, hypothesis, and conditional statements occur much less frequently in the language of those sub- '"41 The Loban Project makes clear jects lacking skill in language. the interrelatedness of both kinds of risk-taking: students who are willing to take_epnceptual risks also explore more writing strategies. The research by Piaget, Perry, and Loban also supports the principle of decentralizatipn: students who are encouraged to explore their ideas through talking and writing recognize both the validity and the limitations of their ideas, as well as the most effective means for expressing them. In regard to meaningfulness. I have repeatedly asserted that meaning is at the heart of all learning--the desire to make sense of the world and of language and, in composition, the desire to transmit meaning to an audience. Once we remove the audience from the context of language use, we remove one of the most essential purposes for using language--to communicate. Furthermore, we know from the research conclusions cited by Lundsteen that audience figures cen- trally in the writing process: "Various audiences help to shape the style and content of writing" (p. 31). Lundsteen's conclusions are corroborated by the numerous studies I cited under Questions One, Fourteen, and Five in Chapter II, studies which show that when students act as audiences for one another's writing, their writing improves more than when no audience is provided. Predispositions toward learning composition, then, can be generated by creating genuine and meaningfulproblem-solving situa- tions in which risk-takipg, decentralization, and meaningful language 175 g§e_are encouraged. I would stress again that the learning model most compatible with_problemrsolving is trial-error-feedback, which pro- vides for maximum operation of the principles of problem-solving. The second component of a theory of instruction for composi- tion concerns structure. Since I have proposed that the structure of composition is the patgge pf discourse (somebody-talking-to-somebody- else-about-something), a theory of instruction should be infOrmed by the features of the ggtgre pf discourse. Of these features, the first is that discourse is a process of shifting roles among the three elements of a discourse occasion (writer-audience-subject). Second, the nature of discourse is both a thought and a language process, which means it integrates all language functions--writing, reading, speaking, and listening--with thinking in order to generate and trans- mit thought. Finally, it is an active process--active both in the generation of meaning and in the communication of meaning. If we combine all of these features and relate them to research on instructional methodologies, we see that the most effec- tive instruction model is the writipg_lab or workshop, where students engage in the shifting relations between writer-audience-subject; where they use language (hopefully) for real purposes--to get effects on an audience; where they integrate all language functions in the total process of writing; and where they are actively involved in the pro- cess of language production. The focus of the workshop is on the process Of composing, be it during the actual writing, or during the discussion and revision of drafts. We see research support for the workshop methodology in Lundsteen's comments on_epyironment, inner motivation, audience, 176 drafts, and oral display. Additional research support derives from the numerous studies of the effectiveness of writing labs/workshops that I cited under Questions One, Fourteen, and Five in Chapter II. . The third component of a theory of instruction is seguence. In my earlier section on composition theory, I advocated a sequence in composition which moves from the personal to the public, from the less abstract to the more abstract, and from the self in relation to itself to the self in relation to the world. This sequence is based not only on the thinking of leading spokesmen in composition theory but also on models of intellectual, ethical, and emotional growth. Such a sequential model suggests an instructional strategy in which students begin writing about personal experiences and, through a series of transitional stages, move gradually to writing about more public concerns--that is, matters of a more experientially generali- zable nature. As I pointed out earlier in this chapter, many sound and con- vincing positions on this sequence are advocated by such writers as James Britton, Janet Emig, and James Moffett. But, since researchers have not investigated the effectiveness of sequential models in com- position, a need for such research exists. Presently, however, the experiences of many writing teachers lends much support to the imple- mentation Of the sequence I am prOposing here, as Mina P. Shaughnessy points out in a recent essay: "It would be difficult to argue against the accumulation of experience in basic writing that suggests auto- biographical content, expressive fOrms, and write-talk or feel-think models Of composing as mOst effective for beginning writers, even where the intent 15 to end up with formal academic writing."42 177 The final component of a theory of instruction concerns reinforcement. From both learning theory and language theory, we know that an environmen§_is most conducive to learning when errors are con- sidered as much a part of learning as are successes. This is parti- cularly true in the composition claSsroom, where students continually learn from the mistakes they make. Most teachers, however, look on mistakes as something to be avoided, primarily because teachers too Often focus exclusively on the written product rather than on the writing process. But, when mistakes are seen as a natural part of the writing process, not something to be avoided at all costs but rather something to learn from, then mistakes are vital--they define, as James Moffett says, “what is gOOd": I think any learning psychologist would agree that avoiding error is an inferior learning strategy to capitalizing on error. The difference is between looking over your shoulder and looking where you are going. Nobody who intends to learn to do something wants to make mistakes. In that sense, avoidance of error is assumed in the motivation itself. But if [the student] is allowed to make mistakes with no other penalty than the failure to achieve his goal, then he knows why they are to be avoided and wants to find out how to correct them. Errors take on a different meaning, they define what is good. (pp. 199-200) One aspect of reinforcement, then, involves reinforcing the value of mistakes in the learning process. Another aspect concerns evaluation. We know from learning theory and from language theory that positive evaluation is the most effective strategy for developing positive attitudes toward learning, particularly if we seek long-range results, such as the development of writing competency. This is not to say that there is no place for negative criticism in the composi- tion classroom, but only to suggest that the power of positive evaluation be Observed and implemented most effectively. Students 178 with a history of writing problems most often have had their weaknesses drilled into them while their strengths have been ignored. Depending upon a student's writing history and his purpose for writing a particu- lar piece, a careful balance of positive and negative criticism may be the most effective formula. It is clear from research in composition that positive evalua- tion is more effective in generating and maintaining positive attitudes toward writing. Also, students learn more from their mistakes when such mistakes are treated positively--as a perfectly normal part of learning to write. As Lundsteen remarks in her sixth point--Positive response--"Teachers help children by looking for strengths and point- ing them out" (p. 3); and, as I pointed out under Questions One and Fourteen in Chapter II, studies done with older students also support the value of positive evaluation of writing. The conclusions reached by Thomas C. Gee in his 1970 study of evaluation procedures used with eleventh graders summarizes well the research basis for my proposal that teachers make greater use of positive evaluation (while not excluding negative evaluation): Students seem to have more patience in working on their composi- tions if they think they will be rewarded for what they do well and if they are encouraged along the way. This study indicates that to assist the building of positive attitude, teachers must give a pat-on-the-back for the improvements that the student makes. To withhold praise until the student has achieved an ideal performance is educationally unsound. It is an easy thing for teachers to mark a set of papers by correcting errors in grammar, usage, spelling, and punctuation and by making suggestions for improving organization, transition, and clarity. It is Often somewhat harder to find several points to commend, but students need encouragement and rightly merit praise for things well done. Their continued improvement apparently comes from recognition of what they do well in addition to what they do not do so well. Certainly their confidence and pride in their efforts, and their enjoyment of writing, are enhanced by 179 a teacher's assurance that they are beginning to master the skills required for good writing. The form of evaluation is another matter which bears on reinforcement. From the research I cited under Questions One and Fourteen in Chapter II, we know that certain forms of feedback on student writing are more effective than others. For example, Stiff (II, p. 54) found that combined marginal and terminal comments were more effective than either marginal or terminal cOmments alone in having long-range positive effects on student attitudes toward writing and possibly on writing performance. Furthermore, Sweet (II, pp. 54-55) found that spontaneous comments made by teachers on stu- dent writing effected a significant improvement in student performance over an extended period Of time. Similarly, Adams (II, pp. 55-56) found that students were more enthusiastic when their themes were evaluated primarily on the basis of thoughts and ideas. These studies suggest that the best form of evaluation is the written commentary which focuses primarily on the content of composi- tions. The efficacy of simply grading compositions finds little support in the research, as Lundsteen points out: "Rating and grading have no part in unifying writer and audience" (p. 3). But since grading is such a permanent fixture in education today, this presents a dilemma: how gg teachers grade student writing? Perhaps the best alternative to grading individual compositions is the pro- cedure of filing papers in a portfolio which Moffett proposes: after students have completed work on a composition, their writing goes "into folders for each student and when the teacher has to evaluate student work for the benefit of administration, he makes a general 180 assessment of the writing to date. No grades are given on individual papers" (p. 198). Though researchers have not investigated the effectiveness of grading on a portfolio basis, there are strong indications from compo- sition testing methodology that such a procedure is necessary when we consider the variability in student writing performance--that is, no one writes well at all times. This variability and its implication for testing are brought out in the Braddock Report when the authors remark that "if an investigator wishes to measure individual students' improvement in writing, he should provide for at least two writing occasions as a pretest, at least two as a post-test, and count the rating only of the better composition on each occasion. If three writing occasions are used for each test, it may be wisest to average the ratings of the two best papers, but more research needs to be done on this possibility."44 Furthermore, a number of studies cited in the Braddock Report's section on the writer variable (e.g., the Kincaid Study) provide evidence of fluctuation in students' composi- tion abilities over a number of writing occasions (see Braddock Report, pp. 6-7). Grading On-a portfolio basis, then, would make allowance for composition fluctuation while it would also implement an evaluative measure which is supported by research methodology. One final point to make with regard to reinforcement concerns the intrinsic reinforcement inherent in language use. That is, when we speak to someone about something, we can gauge the success of our utterance by whether or not our listener has understood, has been moved, has been convinced by it. This is equally true when our utterance is written, which suggests that the mere fact of audience 181 feedback is the most natural means of reinforcement--positive or negative. In the writing lab/workshOp I have proposed here, reig- forcement is built in: a student can make his own evaluation by simply heeding feedback from his audience. If the student's audience consists of peers, then powerful motivational and correctional bene- fits accrue to the learning situation. The effectiveness of peer evaluation is well put by W. H. Auden, in The der's Hand, when he discusses using apprentice poets to comment on each other's work: The apprentices do each other a further mutual service which no Older and sounder critic could do. They read each other's manuscripts. At this age a fellow apprentice has two great virtues as a critic. When he reads your poem, he may grossly overestimate it, but if he does, he really believes What he is saying; he never flatters or praises merely to encourage. Secondly, he reads your poem with that passionate attention which grown-up critics only give to masterpieces and grown-up poets only to themselves. When he finds fault, his criticisms are intended to help you improve. He really wants your poem to be better.45 The instructional model which derives from this theory of instruction would be a writing workshop where students are engaged in solving writing problems which present meaningful challenge and create a need to extend writing skills. In the writing workshop, students actively participate as both writers and audiences; they encounter real purposes for writing--to get effects on an audience--in a genuine give-and-take communicative context; and they integrate all language functions in the total process of writing, from prewriting to writing and editing. Students write in a sequence which starts from an autobiographical base and moves to writing of a more publicly generalizable nature. Within the sequence, students gain experience with many forms of discourse by writing in a variety of contexts and for a variety of purposes. Students learn to write in an environment 182 which develops positive attitudes toward acquiring writing competence by encouraging the exploration of thought and language through reduc- tion of threats to conceptual and linguistic experimentation by building on writing strengths; by recognizing the powerful learning value of errors; and by evaluating student writing on the basis of writing portfolios. I have concluded this chapter with a theory of instruction for the teaching and learning of composition which is based on relevant theory and supported by research. I must emphasize that this is.a theory of instruction, not the theory of instruction. I have sought to present a way of formulating a theory of instruction which is soundly based on insights from learning theory, language theory, and composition theory and research. My firm belief is that the composi- tion curriculum must be based on such a theory of instruction--one which derives from the most recent available knowledge. For too long, teachers and curriculum designers have made decisions based on logic, intuition, and experience alone while they have ignored significant theory and experimentation in many areas which bear directly on their work with students. CHAPTER V CONCLUSION Underlying all that I have written so far are two major charges to the profession of English. The first is that the field of compOsition must be considered a serious intellectual discipline in itself. Writing is a highly complex activity which for too long has been taken for granted as something almost anyone can teach without special study or training. English teachers who claim successful writing programs seldom provide informed rationales for their success; this is not to slight these teachers' success in any way, but merely to cite a need for understanding why some writing programs are success- ful. As Frank Smith notes in meprehension and Learning, "many teachers have great insights into instruction, but are painfully unaware of the theoretical justification for their intuitions."] Those teachers who admit continued frustration and failure in the teaching of composition join their fellows in the search for better methods, better textbooks, better tricks and plays, all the while wondering why their students can't learn to write the way methods/ ‘textbooks/experience/intuition/logic say they should. 4 Yet, as I have tried to point out here, composition can and should be approached systematically as a serious intellectual disci- pline. In the March, 1977, Change, E. Michael Walsh states that 183 184 since we know little about writing, "we live in a world of band-aids and hocus-pocus." According to Walsh, the reason for this is our failure to examine fully the skill of writing: The root cause of this situation is not that writing is too complex an activity to be examined but that few have ever thought of examining it. The English faculty's long-standing antipathy to scientific experimentation (labeled "educa- tionist" and therefore bad), its emphasis on literature, and its de facto denigration of expository writing have all resulted in the assumption that it is not possible to examine writing with the systematic methods applied to most other phenomena. When composition is seen as an intellectual discipline in its own right, then the wealth of insights available from research and theory can be brought to bear on our understanding of what happens when a student learns to write and what should happen when a teacher teaches writing. Today, the call for better results and more system- atic knowledge on the part of writing teachers and directors of writing programs comes from many directions. The media and the general public mount the familiar bandwagon with their outcries over apparent declines in "basic skills." In our professional journals, we see increasing demands for more informed rationales for writing practices and programs. In 1974, for example, David E. Eskey wrote in a College English essay that "We should, as responsible profes- sionals, thus insist that no English teacher be turned loose in the classroom until he has mastered at least the fundamentals of social and regional dialectology. Such a teacher will know better than to try to sell his students a single brand of English as the one and only English fOr all times and places. He will deal with the many dialects of the language, and the natural shifting of styles within dialects, as the typical situation it is."3 Another example appears 185 in the May, 1977, issue of College Composition and Communication, in which Myrna J. Smith insists on composition programs which are solidly based on learning theory: However, in our search for new and better ways of teaching composition, we must go beyond the suggestions of just anyone's successes and ground our teaching in theories of knowing. As the psychologists and psycholinguists refine the theories of epistemology. we teachers must be ready to apply those theories to classroom practice. In our return to basics, let's be sure those basics haxe solid footing in what is most basic of all: how one learns. My second charge to the profession is that composition teachers must be broadly trained to fulfill their professional responsibility and to keep abreast of the significant developments in their field-- that is, they must be specifically trained as writing teachers through preparation which incorporates background in related research, theory, and pedagogy and which is accompanied by experience in teaching writing. In the December, 1976, gellege Composition and Communication, Carl Klaus asks, shouldn't we demand of ourselves and of those who follow us in our calling not only the ability to read and write proficiently but also a form of training, a kind of knowledge, that uniquely fits us to teach writing, as surely as medical school prepares doctors to practice medicine, or law school prepares lawyers to practice law? If we genuinely believe in the dignity of our calling, then we should assure it the dignity that can only come from our being truly professional about it. And I cannot for the life of me imagine any other way of being pro- fessgonal about it than to know as much about writing as we can. Can we, as a profession, continue to justify sending out writing teachers whose background is mostly in literature and whose preparation in composition, if any, consists of having taught writing, with little or no training in the theoretical, research, and pedagogical foundations of the discipline? I believe that we 186 cannot. As Walsh states, "Composition should be as important a subject for research as literature, possibly even more important given the large percentage of an English department's time devoted to teaching it."6 In this regard, I must reaffirm two Of the recommendations to the profession made by the 1974 Carnegie Conference on the State of Undergraduate English: --That despite the centrality of this responsibility [for effective instruction in writing] in today's colleges and universities, teachers of English have received their formal instruction in literature and have ordinarily received inadequate (if any) instruction in the teaching of writing. --That because the profession has not been educated to meet its responsibilities in the area Of writing, it has neither a full understanding of the difficulties students have in mastering writing, nor fully effective methods for helping them become better writers; that it often, in fact, employs methods detrimental to the development of good writing. ‘ Clearly, these two charges to the profession are reciprocal. When composition is recognized as a serious intellectual discipline in its own right and when the training of English teachers includes specific preparation in composition and related fields, then composi- tion programs at all levels can be firmly rooted in the most signifi- cant develOpments in research and theory and be implemented by people who know why they teach as they do and why their students learn as they do. NOTES NOTES INTRODUCTION 1Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer, Research jn_Written Composition (Urbana, 111.: National Council Of Teachers of English, 1963). 2Henry C. Meckel, "Research on Teaching Composition and Literature," in Handbook pf Research pp Teaching, ed. N. L. Gage (Chicago: Rand McNally, 1963). 3J. Stephen Sherwin, Four Problems in Teaching English: A Criti ue pf_Research (Scranton, Pa.: International Textbook Company, I969). 4Richard Braddock, "English Composition," in Enc clo edia pf Educational Research, ed. R. L. Ebel (New York: Macmillan, l869). 5Nathan S. Blount, "Research in Teaching Literature, Language and Composition," in Second Handbook pf Research jn_Teaching, ed. R. M. W. Travers (Chicago: Rand—McNally, 1973). 6Help for the Teacher pf_Written Comppsition, ed. Sara W. Lundsteen (Urbana, 111.: ERIC Clearinghouse on Reading and Communica- tion Skills, 1976). 187 CHAPTER I 1Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer, Research in_Written Composition (Urbana, 111.: National Council of Teachers of English, 1963), p. 1. Further references to this book will appear in my text. 2The most comprehensive summary previous to the Braddock Report was R. L. Lyman, Summary pf Investigations Relating_tg_Grammar, Language, and Composition, University of Chicago Supplementary Educational Monographs, NO. 36 (Chicago: University of Chicago, 1929). 3Nathan S. Blount, "Research on Teaching Literature, Language and Composition," in Second Handbook gf_Research in_Teaching, ed. R. M. W. Travers (Chicago: RanndNally, 1973), p. 1084. 41o1a., p. 1088. 5Donaid T. Campbell and Julian c. Stanley, Experimental and ggasi-Experimental Designs fpr Research (Chicago: Rand McNally, 1966). First publishedin Handbook of Research gg_Teaching, ed. N. L. Gage [Chicagoz Rand McNaii§2"T§BBJTT"FEFTEEF'references to this book will appear in my text. 6Henry C. Meckel, "Research on Teaching Composition and Litera- ture," in Handbook of Research gn_Teaching, ed. N. L. Gage (Chicago: Rand McNaiin'l§E§)T" 7J. Stephen Sherwin, Four Problems _1'_n_Teachin En lish: A Criti ue pf Research (Scranton, Pa.: International Textboo Company, 1969). _ 8Richard Braddock, "English Composition," in Enc clo edia pf Educationel Research, ed. R. L. Ebel (New York: Macmillan, 1969). 9See note 3. 10 Help for the Teacher pf Written Composition, ed. Sara W. Lundsteen (Urbana, Ill.: ERIC Clearinghouse on Reading and Communica- tion Skills, 1976). 188 189 ‘llGlenn H. Bracht and Gene V. Glass, "The External Validity of Experiments," American Educational Research Journal, 5 (1968), 265. 12Richard E. Snow, "Representative and Quasi- Representative Designs for Research on Teaching," Review Lf Educational Research, 44 (1974), 265. 13See note 11. All further references to this article will appear in my text. 14Clinton 1. Chase, Measurement for Educational Evaluation (Reading, Mass.: Addison-Wesley, 1974), pp. 202-203. 15Ibid.. pp. 56-57. 16Norman E. Gronlund, Measurement and Evaluation in Teaching, 3rd ed. (New York: Macmillan, l976),p pp 81- 82. ‘7Richard E. Lindeman, Educational Measurement (Glenview, 111.: Scott, Foresman, 1967), p. 37. 18 Gronlund, p. 93. 22Standards for Educational and Psycho_pgicalw Tests (Washing- ton, D. C: American Psycholog1ca1 Association nc , p. 26. 23Chase, p. 74. 24Standards, p. 48:" 25Gronlund, pp. 114-115. Ibid., p. 114. 27Standards, p. 50. CHAPTER II Adams, Vernon A., "A Study of the Effects of Two Methods of Teaching Composition to Twelfth Graders," Diss. University of Illinois at Urbana-Champaign, 1971. (University Microfilms NO. 72-12, 055 A_TESOL Biblipgraphy: Abstracts pf ERIC Publications and Research Reports, 1969-1970 (Wash1ngton, D.C.: leachers of English to Speakers of Other Languages), 1971. Barbig, Evelyn V., "An Exploration of Growth in Written Composition to Determine the Relationship of Selected Variables to Poor Writing in Grades Nine and Twelve,“ Diss. University of Tennessee, 1968. (University Microfilms NO. 69-1231) Bateman, Donald R. and Zidonis, Frank J., The Effect pf a_Stud .pf Transformational Granmar pp the Writing pf Ninth an enth Grfiders, NatiOnal Council of Teachers of English Researchi Report No. 6 (Urbana, 111.: NCTE, l966). Bell, James B., "A Study of the Written Composition Interests of Senior High School Students," Diss. University of Oregon, 1971. (University Microfilms NO. 7l-23,lO4) Bivens, William P. and Edwards, Allan B., "Transformational Grammar and Writing Improvement," paper presented at the Twenty-Fifth Annual Meeting of the Conference on College Composition and Communication, Anaheim, Calif., April 4-6, 1974. (ERIC Document Reproduction Service, ED 101 361) Blount, Nathan 5., "Research on Teaching Literature, Language and Composition," in Second Handbook gf_Research in Teaching, ed. R. M. W. Travers (Chicago: Rand McNally, l973)? Braddock, Richard. “English Composition," in Encyelppedia gf_Educa- tional Research, ed. R. L. Ebel (New York: Macmillan, 1969). Braddock, Richard, Lloyd-Jones, Richard, and Schoer, Lowell, Research in_Written Composition (Urbana, 111.: National Council 0 Teachers of English,‘1963). Brazil, James M., "On Teaching Composition at the Community College," Diss. University of Michigan, 1975. (University Microfilms No. 75-29.156) 190 191 Burhans, Clinton 5., Jr., Extended Testing gj;a_Unified Experimental Course in pompositi on _i_n a_ Variety _o_i: Materi aTs’ and Formats, USOE Cooperative Research Project NO. 7-1149 (East Lansing, Mich.: Michigan State University, 1968). Burrows, A. T., Children's Writin : Research jn_Composition and Related SEills (Urbana, .: National CouncilTOf Teachers of English, I960). (ERIC Document Reproduction Service, ED 090 546 . Burrus, Dorothy J., "A Three-Year Comparative Study of the Functional Approach to Teaching the Mechanics of Usage," Diss. Oklahoma State University, 1970. (University Microfilms NO. 71-11,110) Calhoun, James L., "The Effect of Analysis of Essays in College Compo- sition Classes on Reading and Writing Skills," Diss. Boston University School of Education, 1971. (University Microfilms Carroll, John B., "Current Issues in Psycholinguistics and Second Language Teaching,“ paper presented at the Fifth Annual TESOL Convention, New Orleans, La., March 5, 1971. (ERIC Document Reproduction Service, ED 052 643) Carroll, John 8., "Research in Foreign Language Teaching: The Last Five Years," in Lan ua e Teachin : Broader Contexts, North- east Conference on the Teaching of Forei n Languages, ed. R. C. MeadI'Jr.'TNeW_VOr : Mhieria 5 Center, 966). Combs, Warren E., "Some Further Effects and Implications of Sentence- Combining Exercises for the Secondary Language Arts Curricu- lum," Diss. University of Minnesota, 1975. (University Microfilms No. 75-21,042) Cooper, Charles R. and Odell, Lee, "Considerations of Sound in the Composing Process of Published Writers," Research in_the, Teaching 9: English, 10 (1976), 103-115. Croft, Kenneth, TESOL 1967-68: .A Survey (Washington, D.C.: Teachers of English to Speakers Of Other Languages, 1970). Daly, John A. and Miller, Michael 0., "The Empirical Development of an Instrument to Measure Writing Apprehension," Research in tpe_Teaching pf English, 9 (1975), 242-249. Donelson, Kenneth L., "Variables Distinguishing Between Effective and Ineffective Writers in the Tenth Grade," Journal 9: Experi- mental Psycholggy, 35 (1967), 37-41. Donlan, Don, "The Effect of Four Types of Music on Spontaneous Writings of High School Students," Research jn_tpe Teaching gf_Eng1ish, 10 (1976), 116-126. 192 Dow, Ronald H., "The Student-Writer's Laboratory: An Approach to Composition," Diss. Boston University, 1973. (University Microfilms No. 73-23,550) Dykstra, Gerald and Paulston, Christing B., "Guided Composition," English Language Teaching, 21 (1967), 136-141. Effros, Charlotte, "An Experimental Study of the Effects of Guided Revision and Delayed Grades on Writing Proficiency of College Freshmen," Diss. New Haven University, 1973. (ERIC Document Reproduction Service ED 079 764) Emig, Janet, The Composing Processes pf Iyelfth Graders, National Council of Teachers of English Research Report No. 13 (Urbana, 111.: NCTE, 1971). Ewing, June 8., "A Study of the Influence of Various Stimuli on the Written Composition of Selected Third Grade Children," Diss. Columbia University, 1967. (University Microfilms NO. 68-5044) Fichtenau, Robert L., "The Effect of Teaching Rhetorical Concepts of Invention, Arrangement and Style on the Written Composition of Selected Elementary School Children in Grades Three Through Six," Diss. Florida State University, 1968. (Univer- sity Microfilms No. 69-16,386) Fisher, Kenneth 0., "An Investigation to Determine if Selected Exer- cises in Sentence-Combining Can Improve Reading and Writing," Diss. Indiana University, 1973. (University Microfilms No. 74-2586) Ford, Bob W., "The Effects Of Peer Editing-Grading on the Grammar- Usage Ability of College Freshmen," Diss. University of Oklahoma, 1973. (University Microfilms No. 73-15,321) Friend, Jewell A., "A Writing Program for Students of English as a Second Language, Based on a Critical Examination of Relevant Research and Theories in Linguistics, Psychology, and Compo- sition," Diss. Southern Illinois University, 1970. (Univer- sity Microfilms NO. 71-11,440) Fry, Danny J., "The Effects Of Transformational Grammar Upon the Writing Performance of Students of Low Socio-Economic Back- grounds," Diss. Memphis State University, 1971. (University Microfilms No. 72-7572) Gee, Thomas C., "The Effects of Written Conment on Expository Composition," Diss. North Texas State University, 1970. (University Microfilms NO. 71-551) Golub, Lester S. and Frederick, Wayne C., “An Analysis of Children's Writing Under Different Stimulus Conditions," Research in_the Teaching pf_English, 4 (1970), 168-180. 193 Gozemba, Patricia A., "The Effect of Rhetorical Training in Visual Literacy on the Writing Skills of College Freshmen," Diss. Boston University School Of Education, 1975. (University Microfilms No. 75-20,950) Graves, Donald H., "An Examination of the Writing Processes of Seven Year Old Children," Research jg_tge Teaching pf English, 9 (1975). 227-241. Green, Everest A., "An Experimental Study of Sentence-Combining to Improve Written Syntactic Fluency in Fifth-Grade Children," Diss. Northern Illinois University, 1972. (University Micro- films NO. 73-4169) Groff, Patrick, "Does Negative Criticism Discourage Children's Composition?" Language Arts, 52 (1975), 1032-1033. Haas, Virginia J., Childers, Perry R., Babbitt, Elizabeth and Dylla, Sandra, "English Composition by Workshop," Journal gf_Experi- mental Educatipp, 40 (1972), 33-37. Hall, MaryAnn, Moretz, Sara A. and Storm, Jodellano, "Writing Before Grade 0ne--a Study of Early Writers,“ Language Arts, 53 (1976), 582-585. Hansen, Barbara L., "The Effect of Teacher-Guided Theme-Revision on Composition Performance of University Freshmen," Diss. Ball State University, 1971. Dissertation Abstracts, 31, (1971), 1437A. (University Microfilms No.771-24,379) Hardaway, John M., "Generative Rhetoric: An Analysis of its Influence on the Writing of College Freshmen," Diss. Arizona State University, 1969. (University Microfilms No. 69-20,768) Harris, Maverick M., "The Effectiveness of Progranmed Instruction for Teaching Expository Writing to College Freshmen and High School Seniors," Diss. University of Texas at Austin, 1972. (University Microfilms No. 73-7562) Hazen, Carl L., "The Relative Effectiveness of Two Methodologies in the Development of Composition Skills in College Freshman English," Diss. North Texas State University, 1972. (Univer- sity Microfilms No. 73-2905) Hooks, Janice, "An Analysis of Writing Skills as Described by Selected Professional Writers," Diss. Indiana University, 1972. (University Microfilms No. 73-6978) Hoyer, Linda E., "The Invention Process in Composition: A Selected, Annotated Bibliography, 1950-1974," Master's thesis, Texas A & M University, 1974. (ERIC Document Reproduction Service ED 103 914) 194 Hunt, K. W, Grammatical Structures Written at Three Grade Levels, National Council of TeaChers of English Research Report No.3 (Urbana, 111.: NCTE,1965). Hunting, Robert, "Recent Studies of Writing Practice," Researchi "LEE. Teaching gf_English, l (1967), 27- 40. 1968 Index to ERIC Documents in Linguistics and the Uncom n1 Taught Languages and Selected Bibliography_ Of’RETEte 1t es (Washington, D. C: ERIC C earinghouse for Lingu1st1cs, 1969). Jenks, Eleanor C, "An Experimental Method to Develop Creativity of Tenth Grade Students," Diss. University of Oregon, 1965. (University Microfilms NO. 65- 12 ,221) Jobe, Ronald A, "Factors That Influence Children' 5 Free Choices of TOpics for Creative Writing," Diss. University of Minnesota, 1974. (University Microfilms NO. 75- 12,087) Kafka, T. Toby, "A Study of the Effectiveness of Four Motivational Stimuli on the Quality of Compositions of Intermediate Stu- dents in One School District," Diss. St. Johns University, 1971. (University Microfilms NO. 71- 30 ,213) King, Richard P., "Sensory Approach to Creative Writing: A Study of the Effect of Increasing the Number of Types of Sensory Stimuli Intended to Motivate Children to Write Creatively," Diss. University of North Dakota, 1973. (University Micro- films No. 74-14,904) Klein, Marvin L. and Grover, Burton L, An Assessment Lf the Effective- ness Lf S mbolicL Lo ic in the Teachi_g Lf Composit1on ymouth,1s.,1970).—(ERIC Document Reproduction Service ED 059 209) Koch, Carl J., "Small Groups in the Composition Class: A Case Study of Developing Linguistic Security and Written Fluency," Diss. University of Michigan, 1975. (University Microfilms NO. 75-29, 160) Lacampagne, R. J., "A National Study of Selected Attitudes and Approaches to Writing of Twelfth-Grade Students With Superior Writing Performance Versus Those With Average Writing Per- fOrmance," Diss. University of Illinois, 1969. (University Microfilms NO. 69-10,757) Lagana, Jean R., "The Development, Implementation, and Evaluation of a Model for Teaching Composition which Utilizes Individualized Learning and Peer Grouping," Diss. University of Pittsburgh, 1972. Dissertation Abstracts, 33 (1973), 4063A. (University Microfilms No. 73-4127) 195 Language and Lan ua Behavior Abstracts (LaJolla Calif.: United States Intem ationalUniversity). ’ Lundsteen, Sara W. ed., Hel for the Teacher Lf Written Composition (Urbana, Ill. :ER earifigfiouse on Reading and Communication Skills, 1976). Maloney, Henry B., "An Identification of Excellence in Expository Composition Performance in a Selected 9A Population With an Analysis of Reasons for Superior Performance," Diss. Columbia University, 1967. (University Microfilms No. 68-2432) Marzano, Robert J., "The Sentence Combining Myth," nglish Journal, 65 (1976), 57- 59. McClatchey, Barbara and McClatchey, Joseph, "The Writer's Audience: What is its Effect on Quality of Student Composition?" Arizona English Bulletin, 12 (April 1970), 20-23. McElwee, Gregory W., "Systematic Instruction in Proofreading for Spelling and Its Effects on Fourth and Sixth Grade Composi- tion," Diss. University of Wisconsin-Madison, 1974. (University Microfilms No. 75-7594) Meckel, Henry C. "Research on Teaching Composition and Literature, " in Handbook of Research Ln Teaching, ed. N. L. Gage (Chicago: Rand ndNally, 1963) Mellon, John C. , Transformational Sentence- Combinin : A Method for Enhancing_the Development Lf S ntactic Fiuenc in En lish Composition, National Council of Teachers of EngTish Research Report No.10 (Urbana, Ill. :NCTE, 1969). Miller, Barbara and Ney, James W., "The Effect of Systematic Oral Exercises on the Writing of Fourth-Grade Students," Research jg_the Teaching g: English, 2 (1968), 44-61. Miller, Tyree J., "A Quantitative Study of the 'Free Modifiers' in Narrative- -Descriptive Compositions Written by Black College Freshmen After Leaving the Influence of The Christensen Rhetoric Program and a Study of Their Attitudes Toward Written Composition," Diss. Ball State University, 1972. (University Microfilms No. 73-1225) Mills, H. W., "A Comparison of the Effects of Oral Proofreading and Silent Proofreading of Rough Drafts of Compositions in Helping Sixth Grade Children Develop Greater Sentence Sensitivity," Diss. University of Kansas, 1970. Mischel, Terry, "A Case Study of a Twelfth- Grade Writer," Research in the Teachiggfo English, 8 (1974), 303-314. 196 Morrisroe, Michael and Morrisroe, Sue, "TESL: A Critical Evaluation of Publications, 1961-1968," Elementary English, 49 (l972), 50-61. Nakamura, Richard, "What is the Relationship Between the Amount of Reading and the Quality of Writing by Tenth Grade Boys?" Arizona English Bulletin, l2 (April 1970), 24-28. (Lincoln: University of Nebraska—Press, l936). Document Reproduction Service ED 013 806-ED 0l3 81]) Nebraska Curriculum DeveIOpment Center, A Curriculum for En lish (Efifif Ney, James W., "Notes Toward a Psycholinguistic Model of the Writing Process," Research jg_the Teaching gj_English, 8 (1974), lS7-l69. Norwood, Mary Catherine D., "Achievement as Related to the Ethnic Origin of the Student in an Experience-Centered Approach to the Teaching of Freshman Composition," Diss. University of Tulsa, 1974. (University Microfilms No. 74-l9,266) Oberchain, Anne, "Effectiveness of the Precise Essay Question in Programming the Sequential Development of Written Composition Skills and the Simultaneous Development of Critical Reading Skills," Master's thesis, George Washington University, 1971. Odell, Lee, "Measuring the Effect of Instruction in Prewriting," Research in the Teaching gf_English, 8 (1974), 228-240. Ofsa, William J., "An Experiment in Using Research in Composition in the Training of Teachers of English," Diss. University of Illinois at Urbana-Champaign, l974. (University Microfilms NO. 75"]19711) O'Hare, Frank, Sentence Combining; Im rovin Student Writin Without Formal Grammar Instruction, National Council of Teacaers of English ResearchTReport No. 15 (Urbana, Ill.: NCTE, 197l). Pinkham, R. G., "The Effect on the Written Expression of Fifth Grade Pupils of a Series of Lessons Emphasizing the Characteristics of Good Writing as Exemplified in Selected Works from the Area of Children's Literature,“ Diss. University of Connecti- cut, 1968. (University Microfilms No. 69-02168) Rohman, D. Gordon and Wlecke, Albert 0., Pre-writing; The Construc- tion and A lication of_Mgdels for Concept Formation in Writing, UggE Cooperative Research Project No. 2|7Z (East Lansing, Mich.: Michigan State University, 1964). Sager, Carol, "Improving the Quality of Written Composition Through Pupil Use of a Rating Scale," Diss. Boston University School of Education, l973. (University Microfilms No. 73-23,605) 197 Sanders, Sara E., "A Comparison of 'Aims' and 'Modes' Approaches to the Teaching of Junior College Freshman Composition Both With and Without an Auxiliary Writing Lab," Diss. University of Texas at Austin, 1973. (University Microfilms No. 74-05321) Sawkins, Margaret W., “The Oral Responses of Selected Fifth Grade Children to Questions Concerning Their Written Expression," Diss. State University of New York at Buffalo, 1971. (University Microfilms No. 7l-l6,463) Schneider, Virginia L., "A Study of the Effectiveness of Emphasizing the Teaching of Reading Skills to Improve Composition Skills in Remedial English Classes at Kansas City Community Junior College," Diss. University of Kansas, l970. (University Microfilms No. 71-13,359) ’ Shafer, Robert E., "The Crisis in Knowing About Learning to Write," ADE Bulletin, No. 6 (Sept. l975), 52-57. Shapiro, Phyllis P. and Shapiro, Bernard J., Poetry Instruction: .LEE Effect pg Attitudes Toward Literature and the Ability tg_Write Prose, 197l. (ERIC Document Reproduction Service ED 058 2705 Sherwin, J. Stephen, Four Problems in Teaching En lish: A Critigue gj_Research (Scranton, Pa.:-Internationa ext ooF Company, 1969,. Slay, Allan L., "A Comparison of the Effectiveness of Programmed, Handbook, and Non-Formalized Grammar Instruction in Remedial College Freshman English Composition," Diss. St. Louis University, 1968. (University Microfilms No. 69-l6,048) Smith, Doris 1., "Effects of Class Size and Individualized Instruction on the Writing of High School Juniors,“ Diss. Florida State University, l974. (University Microfilms No. 74-2546l) Stallard, Charles K., "An Analysis of the Writing Behavior of Good Student Writers," Diss. University of Virginia, l972. (University Microfilms No. 72-33,385) Standish, Patricia, in "Shoptalk," Arizona English Bulletin, l2 (April l970). Sl. Stevens, Alfred E., "The Effects of Positive and Negative Evaluation on the Written Composition of Low Performing High School Students," Diss. Boston University School of Education, 1973. (University Microfilms No. 73-23,617) Stewart, Donald C., "Essays, Analysis, and--Better Writing?" Research jg_the leaching gj_English, 3 (l969), 42-51. 198 Stiff, Robert, "The Effect Upon Student Composition of Particular Cozrectgon Techniques,” Research jg thg Teaching glenglish, l 1967 , 54-75. Stotsky, Sandra L., "Sentence-Combining as a Curricular Activity: Its Effect on Written Language Development and Reading Comprehension," Research jfl.the Teaching gj English, 9 (1975), 30-71. Sutton, Doris G. and Arnold, Daniel S., "The Effects of Two Methods of Compensatory Freshman English," Research jg_the Teachigg_ gj_English, B (1974), 241#249. Sweet, Roger C., Educational Attainment and Attitudes Toward School as g_Function'gf_Feedback jn_thg_Form gf__Teachersl Written Comments, I966. (ERIC Document Reproduction Service ED 015 [635 Troyka, Lynn 0., "A Study of the Effect of Simulation-Gaming on Expository Prose Competence of College Remedial English Composition Students," Diss. New York University, 1973. (University Microfilms No. 73-30.136) Turner, Geraldine C., Experimental Research _i_r1t_h_e_ Use 913 Writing L39, St. Petersburg Junior College, 1970. (ERIC Document Reproduction Service ED 041 576) VanFleet, Ellamaye, "Guidelines for Determining the Content of a Collegiate Report-Writing Course," Diss. University of Tennessee, 1969. (University Microfilms No. 70-17,856) Varon, Ruth L., "Examination of the Content of Unsolicited Composi- tions Written by Upper Elementary School Children for Their Teacher's Eyes Only," Diss. Columbia University, 1971. (University Microfilms No. 71-24,168) Vygotsky, L.)S., Thought and Language (Cambridge, Mass.: MIT Press, 1962 . Wahlberg, William A., "The Effect of Process Intervention on the Attitudes and Learning in a College Freshman Composition Course,” Diss. University of Michigan, 1970. (University Microfilms No. 71-15,334) CHAPTER III 1Kenneth L. Donelson, "Should English Teachers Be Expected to Read and Conduct Research?" Arizona En lish Bulletin, 12 (April 1970), p. 4. Further references to this art c e will appear in my text. 2Lucme Shandloff, "The Relationship of Freshman Composition Curriculum Practices in Florida Public Community Junior Colleges to Research in the Teaching of Written Composition," Diss. Florida State University, 1973. 3Willard D. Memering, "Recent Theories and Practices in the Teaching of Composition," Diss. Florida State University, 1971. 4Dwight L. Burton, "Research in the Teaching of English: The Troubled Dream," Research in thg_Teachin gj_En lish, 7 (1973), 161. Further references to this article will appear n my text. 5Roland Harris, "Some Thoughts on Research and the Teaching of English," Research jg_the Teaching_gj English, 2 (1968), 6. 6Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer, Research in Written Composition (Urbana, 111.: National Council of Teachers 5? English, 1963), p. 5. 7Richard Braddock, "English Composition," in Enc clo edia 91 Educational Research, ed. R. L. Ebel (New York: Macmillan, 1969l, p. 449. 8Donald T. Campbell and Julian C. Stanley, Experimental and Quasi-Experimental Designs for Research (Chicago: Rand McNally, 1966), p. . Further references to this boOk will appear in my text. 9Barbara L. Hansen, "The Effect of Teacher-Guided Theme- Revision on Composition Performance of University Freshmen," Diss. Ball State University, 1971. 10James R. Squire, ”Research Can Make a Difference," Research .19 the Teaching gf_English, 10 (1976), 64. 199 200 nIbi ., p. 53. 12Janet Emig, The Com osin Processes gf_Twe1fth Graders, National Council of Teachers 0 ng isfi Research Report No. I? (Urbana, 111.: NCTE, 1971). 13Donald H. Graves, "An Examination of the Writing Process of feven)Year Old Children," Research jg the Teaching_gj_English, 9 1975 . 14Philip H. Taylor, "The English Composition in the Junior School: Prepared or Unprepared?" Educational Research, 5 (1962). '50. Gordon Rohman and Albert 0. Wlecke, Prewritin : The Construction and_Applicationlgf_Mode1s for Conce t Formation TF- Writin , USOE Cooperative’Research Project No. 74 (East Lansing, Mich.: Michigan State University, 1964). 16James M. Brazil, “On Teaching Composition at the Community College," Diss. University of Michigan, 1975. 17Carl J. Koch, "Small Groups in the Composition Class: A Case Study of Developing Linguistic Security and Written Fluency," Diss. University of Michigan, 1975. '8Janice Hooks, "An Analysis of writing Skills as Described by Selected Professional Writers," Diss. Indiana University, 1972. 19Barbara McClatchey and Joseph McClatchey, "The Writer's Audience: What is its Effect on Quality of Student Composition?" Arizona English Bulletin, 12 (April 1970). 20Bob W. Ford, “The Effects of Peer Editing/Grading on the Grammar-Usage Ability of College Freshmen," Diss. University of Oklahoma, 1973. 2lJean R. Lagana, "The Development, Implementation, and Evaluation of a Model for Teaching Composition which Utilizes Individualized Learning and Peer Grouping," Diss. University of Pittsburgh, 1972. 22Patrick Groff, "Does Negative Criticism Discourage Children's Composition?" Language Arts, 52 (1975). 23Virgina J. Haas, Perry R. Childers, Elizabeth Babbit, and Sandra Dylla, "English Composition by Workshop," Journal gj.Experi- mental Education, 40 (1972). 201 24Doris G. Sutton and Daniel 5. Arnold, "The Effects of Two Methods of Compensatory Freshman English," Research in thg Teaching _o_f English, 8 (1974). 25Doris I. Smith, "Effects of Class Size and Individualized Instruction on the Writing of High School Juniors," Diss. Florida State University, 1974. 26Sara W. Lundsteen, ed., Help for the Teacher gf_Written Composition (Urbana, 111.: ERIC Clearinghouse on Reading and Communi- cation Skills, 1976), p. 1. 271bid. 28Nathan S. Blount, "Research on Teaching Literature, Language and Composition," in Second Handbook of Research in Teaching, ed. R. M. W. Travers (Chicago: RanndNalTy, 1973), p. 109 . 29Memering, see note 3. 30Shandloff, see note 2. 3lMina P. Shaughnessy, "Basic Writing," in Teachin Composi- tion: 10 Bibliographic Essays, ed. Gary Tate (Fort ort : exas Christian UniversityPress,il976), p. 154. CHAPTER IV 1Alan M. Hollingsworth, "Beyond Literacy," AQ§_Bu11etin, No. 36 (March 1973), pp. 6-7. - 2Ibid., p. 7. 3Robert E. Shafer, "The Crisis in Knowing about Learning to Write," ADE Bulletin, No. 6 (Sept. 1975), pp. 53-54. 4Carl Klaus, "Public Opinion and Professional Belief," College Composition and Communication, 27 (1976), 337. 5Ibid. 6Mina P. Shaughnessy, "Basic Writing," in Teachin Composi- tion: lQ_Bibliographic Essa 5, ed. Gary Tate (Fort ort : Texas Christian University Press, 1976), p. 154. 7Carl R. Rogers, Freedom §g_Learn (Columbus, Ohio: Merrill, 1969). p. 35. "“"“ "“" 8Carl R. Rogers, Qg_Becoming 3 Person (Boston: Houghton Mifflin, 1961). p. 280. ‘_""' 9Frank Smith, Comprehension ggg_Learning (New York: Holt, Rinehart and Winston, 1975), p. 121. Further references to this book will appear in my text. loUlric Neisser, Cognition and Reality (San Francisco: W. H. Freeman and Company, 1976), p. 1. 1iGeorge A. Kelly, A Theory gj_Personality(New York: Norton, 1963). p. 12. 12Rogers,OgBecominggPerson, pp. 286-87. 202 203 13James Moffett, Ieaching_the Universe 2: Discourse (Boston: Houghton Mifflin, 1968), p. 193. Further reierences to this book will appear in my text. 14Kelly, p. 14. 15Mina P. Shaughnessy, Erro[§_and Expectations (New York: Oxfbrd University Press, 1977), pp. 79-80. 16 . Jean Piaget, ng_Lan ua e gpg_Thought pf thg_Chlld, trans. from 3rd ed. by Marjohie Gahain (l926: rpt. London: Routledge 8 Kegan Paul, 1959), p. 9. Further references to this edition will appear in my text. 17william 6. Perry, Jr., Forms _o_f_ Intellectual and Ethical Development in the Collegg_Years (New York: Holt, Rinehart and Winston, 19681, p. 204. 18Neisser, p. 57. 19Neil Postman and Charles Weingartner, Teachin pg g_§gp: versive Activity (New York: Delacorte Press, 1969), p. 32. 20David R. Olson, "What Is Worth Knowing and What Can Be Taught," School Review, 82 (1973), 35. Further references to this article will appear in my text. 2‘Perry, p. 211. 22Stephen N. Judy, Ex lorations _ip the Teachipg _o_f_ Secondary English (New York: Dodd, Mead, 1974). pp. 43-44. 23"Consultative" and "intimate" are technical terms used to describe language styles in Martin Joos' book, The Five Clocks (New York: Harcourt, 1961). 24Jenefer M. Giannasi, "Dialects and Composition," in Teachin Com osition: lQ_Biblio ra hic Essa 5, ed. Gary Tate (Fort Worth: iexas C ristian UniverSlty ress, 19 6 , p. 277. 25Shaughnessy, Errors and Expectations, p. 125. 26Lou Kelly, "Is Competent Copyreading a Violation of the Students' Right to Their Own Language?" College Composition and . Communication, 25 (1974), 255. 204 27"Students' Right to Their Own Language," Special Issue, Collgge Composition and Communication, 25 (Fall, 1974), 8. 28Lou Kelly, ppp, p. 253. 29Jerry A. Fodor, "How to Learn to Talk: Some Simple Ways," in The Genesis of Lan ua e, ed. Frank Smith and George A. Miller (Cambridge: MIT-Fress, 1966), p. 106. 30 . . . Bradford Arthur, Teachln Engllsh tg_Speakers.gf_Engllsh (New York: Harcourt, 1973), p. 64. 31James Britton, "Progress in Writin ," in Explorations jp_ Children's Writing, ed. Eldonna L. Everetts Urbana, I11.: National Council of Teachers of English, 1970), p. 38. Further references to this and other articles by Britton in this book will appear in my text. 32Frank Whitehead, The Disappearipg Dais (London: Chatto & Windus, 1966), p. 161. 1 33Arthur N. Applebee, Tradition and Bgform jp_the Teachin pf; En lish: a Histor (Urbana, 111.:‘Nationa1 Council of'TEachers of English, 197 , p. 253. 34John Dixon, Growth through English (New York: Oxford University Press, 1967), p. 7. 351bid., p. 10. 36Wayne C. Booth, “The Rhetorical Stance," in Teachin Freshman Composition, ed. Gary Tate and Edward P. J. Corhett (New YOFk: OxfoFd University Press, 1967), p. 184. 37Janet Emig, Ihg Composing Processes gj_Twelfth Graders, National Council of Teachers of English Research Report No. 13 (Urbana, 111.: NCTE, 1971), p. 36. Further references to this book will appear in my text. 38Hel ‘fgp_the Teacher gf_Written Composition, ed. Sara W. Lundsteen Ur ana, Tiil: ERIC Clearinghouse on Reading and Communica- tion Skills, 1976), pp. 1-4. Further references to this book will appear in my text. 39Jerome S. Bruner, Toward g_Theor ‘2: Instruction (Cambridge, Mass.: Belknap Press, 1966), p. 40. Furt er references to this book will appear in my text. 205 40See notes 16 and 17. 4(Richard Braddock, Richard Lloyd-Jones, and Lowell Schoer, Research in Written Com osition (Urbana, 111.: National Council of Teachers 5? English, 963 , p. 32. 42Shaughnessy, in TeachingComposition: 19_Bib1iographic Essays, p. 151. 43Thomas C. Gee, "Students' Responses to Teacher Comments," Research in the Teachin of English, 6 (1972), 219. This article is a summary of’ihe authoris— 1970 Doctoral dissertation entitled "The Effects of Written Comment on Expository Composition," North Texas State University. 4i§§§gg§gh_jp_Written Composition, p. 7. 45 p. 40. W. H. Auden, The Dyer's Hand (New York: Random House, 1956), CHAPTER V 1Frank Smith, Comprehension and Learnipg_(New York: Holt, Rinehart, 1975), p. 246. 2E. Michael Walsh, in "How Can Colleges Deal With the Increasing Incidence of Student Weakness in Basic Learning Skills?" Change, 9, No. 3 (1977), 50. BDavid E. Eskey, "The Case for the Standard Language," lelggg_English, 35 (April 1974), 774. 4Myrna J. Smith, "Bruner on Writing,“ College Cgmposition and Communication, 38 (1977), 133. 5C3P1 Klaus, "Public Opinion and Professional Belief," Collgge Composition and Communication, 37 (1976), 337-38. 6valsh, p. 50. 7"Recommendations to the Profession," AQ§_Bulletin, No. 6 (Sept. 1975), p. 60. 206 BIBLIOGRAPHY BIBLIOGRAPHY Adams, Vernon A. _A Study of the Effects _o_f_ 1112 Methods _o_f_ Teachin Composition tg_TweTfth Graders. Diss. University of Illinois at urbana-Champaign,i1971. Applebee, Arthur N. Tradition and Reform jg_thg_Teaching 9: English: g_History. Urbana, 111.: National Council of Teachers 0 English, 1974. Arthur, Bradford. Teachipg_English tg_Speakers g: English. New York: Harcourt,i1973. A_TESOL Bibliography: Abstracts of ERIC Publications and Research ngorts, 1969-1970. Washihgton, D.C.: Teachers of English to Speakers of Other Languages, 1971. Auden, W. H. The Dyer's Hand. New York: Random House, 1956. Barbig, Evelyn V. Ap_Ex loration of Growth ip_Written Composition to Determine the Relationship_of Sglectedvariables tg_Poor hhfitin jg_Grades Nine and TweTve.i_Diss. University of Tennessee, Bateman, Donald R. and Zidonis, Frank J. The Effect 913 Study g_f_ Transformational Grammar gg_the Writipgflgf_Ninth and Tenth Graders. Nationa1 Council of Teachers of English Research Report No. 6. Urbana, 111.: NCTE, 1966. Bell, James B. A Study of the Written Comosition Interests 5E Senior High Sc oBT'Students. Diss. University ofOregon, 1971. Bivens, William P. and Edwards, Allan B. "Transformational Grammar and Writing Improvement." Paper presented at the Twenty- Fifth Annual Meeting of the Conference on College Composition and Communication, Anaheim, Calif., April 4-6, 1974. Blount, Nathan S. "Research in Teaching Literature, Language and Composition." In R. M. W. Travers, ed., Second ngdbookIQf Research jp.Teaching. Chicago: Rand McNally, 1973. 207 208 Booth, Wayne C. "The Rhetorical Stance." In Gary Tate and Edward P. J. Corbett, eds., Teachipg_FreShman Composition. New York: Oxford UniversityPress, 1967. Bracht, Glenn H. and Glass, Gene V. “The External Validity of Experi- ments." American Educational Research Journal, 5 (1968), 437-474. Braddock, Richard. "English Composition." In R. L. Ebel, ed., Encyclopedia of Educational Research. New York: Macmillan, 1969. Braddock, Richard, Lloyd-Jones, Richard and Schoer, Lowell. Research jp_Written Composition. Urbana, I11.: Nationa1 Counci 0 Teachers of English,1963. Brazil, James M. Qo_Teachiog_Composition ot_the Community College. Diss. University of Michigan, 1975. Britton, James. "Talking and Writing," "Progress in Writing,“ "Language and Experience," "Student Writing and Evaluation." In Eldonna L. Everetts, ed., Ezplorations jp_Children's Writin . Urbana, 111.: National Council of Teachers of English. 1970. Bruner, Jerome S. Toward o_Theory of Instruction. Cambridge, Mass.: Belknap Press, 1966. Burhans, Clinton S., Jr. Extended Testing of o_Unified Experimental Course jp_Composition jp_o_Variety of_Materials and Formats. USOE Cooperative Research Project No. 7-1149. East Lansing, Mich.: Michigan State University, 1968. Burrows, A. T. Children's Writing: Research jp.Compositiop and Related Skills. Urbana, 111.: National Council of Teachers of English, 1969. Burrus, Dorothy J. IA,Three-Year Comporative Studyof the Functional Approach to Teachin the Mechanics of Usage. Diss. Oklahoma tate University, 970. Burton, Dwight L. "Research in the Teaching of English: The Troubled Dream." Research ip_the Teaching of_Eng1ish, 7 (1973), 160-189. Calhoun, James L. The Effect of_Ana1ysis olessays jp_College Compo- sition Classes op Reading op__ riting Skll s. Diss. Boston University Sohool o E ucation, 97 . Campbell, Donald T. and Stanley, Julian C. Experimental and ansi- Experimental Designs for Research. Chicago: Rand McNally, 1966. ’ 209 Carroll, John 8. "Current Issues in Psycholinguistics and Second Language Teaching." Paper presented at the Fifth Annual TESOL Convention, New Orleans, La., March 5, 1971. "Research in Foreign Language Teaching: The Last Five Years." In R. G. Mead, Jr., ed., Language Teachin : Broader Contexts, Northeast Conference op_the Teachipg of. oreign Languages. New York: MLA Materials Center,‘1966. Chase, Clinton 1. Measurement for Educational Evaluation. Reading, Mass.: Addison—WeSTey, 1974. Combs, Warren E. Some Further Effects and Implications of_Sentence- Combining Exercises fOr the Secondar Language Arts Curriculum. Dlss. University of Minnesota, 1975 Cooper, Charles R. and Odell, Lee. "Considerations of Sound in the Composing Process of Published Writers." Research jp_tpo_ Teaching of English, 10 (1976), 103-115. Croft, Kenneth. TESOL, 1967-68: 'A_Survey. Washington, D.C.: Teachers of English to Speakers of Other Languages, 197D. Daly, John A. and Miller, Michael D. "The Empirical Development of an Instrument to Measure Writing Apprehension." Research jp_tpo_ Teaching of English, 9 (1975), 242-249. Dixon, John. Growth Through English. New York: Oxford University Press, 1967. Donelson, Kenneth L. "Should English Teachers Be Expected to Read and Conduct Research?" Arizona Epglish Bulletin, 12 (April 1970 . 1-7. "Variables Distinguishing Between Effective and Ineffective Writers in the Tenth Grade.” Journal of ExDerimental Psychology, 35 (1967), 37-41. Donlan, Don. "The Effect of Four Types of Music on Spontaneous Writings of High School Students." Research pp the Teachjog_ of_Eng1ish, 10 (1967). 116-126. Dow, Ronald H. The Student-Writer's Lgboratory: ‘Ap_Approach to_ Composition. Diss. Boston University, 1973. Dykstra, Gerald and Paulston, Christina B. "Guided Composition." English Language Teaching, 21 (1967), 136-141. Effros, Charlotte. ‘Ap_Experimenta1 Study of the Effects of Guided Revision and Delayed Grades op_Writing Proficiéncy‘of_Co||ege FreSHmen. iDiss. New Haven'UnlverSlty, 210 Emig, Janet. The Composing_Processes Lf Twelfth Graders. National Council of Teachers of EngTish— Research Report No.13. Urbana, Ill. :NCTE, l97l. Eskey, David E. "The Case for the Standard Language." College English, 35 (April 1974). 769-774. Ewing, June B. A Study_ of the Influence Lf Various Stimuli Ln the Written Composition Lf Selected Third Grade Children. Diss. COTUmbia UnlverSlty, 1967. Fichtenau, Robert L. The Effect of Teaching Rhetorical Concepts Lf Invention, Arrangement and—Style on the written Composition Lf Selected Elementar School Children in GradesiThree Throogh Slx. Dlss. Florlda State University, 1968. Fisher, Kenneth D. An Investi ation to Determine if Selected Exercises in Sentence-Combining gan Improve Reading_and—Writi_g. Diss. IhdianaUniverSlty, 1973— Fodor, Jerry A. "How to Learn to Talk: Some Simple Ways." In Frank Smith and George A. Miller, eds., The Genesis of Language. Cambridge, Mass.: MIT Press, 1966. Ford, Bob W. The Effects of Peer Editi_g/Grading_Ln the Grammar-Usage Abilit Lf College Freshmen. Diss. UniverSlty of Oklahoma, 1973. Friend, Jewell A. A Writing Program for Students of En lish as a Second Languae ,Based Ln a Critical Examination of RETevant ResearCh and‘ eories in Linguistics, Ps ch01o , and Co_posi- tion. Diss. Southern 1T1inois University, 1970. Fry, Danny J. The Effects of Transformational Grammar Upon the Writing Performance_ of Students Lf Low Socio- Etonomic Backgrounds. Diss. Memphis State University, 1971. Gee, Thomas C. "Students' Responses to Teacher Comments." Research 1p the Teaching of English, 6 (1972), 212-219. . The Effects Lf Written Comment Ln Expository Composition. Diss. North Texas State University, —1970. Giannasi, Jenefer M. "Dialects and Composition. " In Gary Tate, ed., Teachin Composition. 10 Biblio ra hic Essays Fort Worth: Texas ChristianUnlverSlty ress, 6. Golub, Lester S. and Fredrick, Wayne C. "An Analysis of Children's Writing Under Different Stimulus Conditions." Research in £pe_Teaching of_Eng1ish, 4 (1970), 168-180. 211 Gozemba, Patricia A. The Effect of Rhetorical Trainin Ln Visual Literac Ln theWritingSEills of Co l1ege_ res men. Diss. Boston Unlversity School of Education, 1975. Graves, Donald H. "An Examination of the Writing Processes of Seven Year Old Children.“ Research jg_the Teachigg_gf English, 9 (1975), 227-241. Green, Everest A. An Experimental Stud Lf Sentence- Combinin to Improve Written Syntactic F uengy_7n Fifth-Grade Cfii laan. Diss. Northern Illinois University, 7197?. Groff, Patrick. "Does Negative Criticism Discourage Children's Composition?“ Language Arts, 52 (1975), 1032-1033. Gronlund, Norman E. Measurement and Evaluation Ln Teaching. 3rd. ed. New York. Macmilian, I976. Haas, Virginia J. et al. "English Composition by Workshop." Journal 9: Experimental Education, 40 (1972), 33-37. Hall, MaryAnn, Moretz, Sara A. and Storm, Jodellano. "Writing Before Grade 0ne--a Study of Early Writers." Language Arts, 53 (1976), 582-585. Hansen, Barbara L. The Effect Lf Teacher-Guided Theme- -Revision 2g Composition Performance— Lf University_Freshmen. Diss. Ball State University, 1971‘. Hardaway, John M. Generative Rhetoric. An Analysis Influence Ln the Writing Lf College Freshmen. Diss. Arizona State University, 1969. Harris, Maverick M. The Effectiveness of Programmed Instruction for Teaching_Exposito_y_Writing_to CollegeiFreshmen and High School Seniors. Diss. University OF'Texas at‘Austin, l972. Harris, Roland. "Some Thoughts on Research and the Teaching of English." Research jg_the Teaching gf_Eng1ish, 2 (1968), 5-13. Hazen, Carl L. The Relative Effectiveness of Two Methodologies Ln the Development LfCompositionSkilTE'Ln Colle e FreShman English. Diss. North Texas State—UnivFrSlty, 972. Hollingsworth, Alan M. "Beyond Literacy." ADE Bulletin, No. 36 (March 1973), 3- 10. Hooks, Janice. An Anal sis Lf Writing Skills as Described by_Selected ProfessiFnal r ters. Diss. Indiana University, i972. 212 Hoyer, Linda E. The Invention Process in Composition. A Selected, Annotated—Biblio ra fi |§56-|97Z'. Master 5 thesis, Texas A& M University, |E7%. Hunt, K. N. Grammatical Structures Written at Three Grade Levels. National Council of Teachers of English Research Report No.3. Urbana, Ill. :NCTE, l965. Hunting, Robert. "Recent Studies of Writing Practice. " Research in the Teaching_Lf English, 1 (l967), 27- 40. 1968 Index to ERIC Documents in Linguistics and the Uncommonly Tau Tt [an ua es and electe iography of Related Titles. as 1ngton, D.C .:'_ERIC Clearinghouse for Linguistics, l969. Jenks, Eleanor C. A_n_ Experimental Method 1:}; Develop Creativitygj Tenth Grade Students. Diss. University of Oregon, 1965. Jobe, Ronald A. Factors That Influence Children's Free Choices 2: To ics for Creativeflriting, Diss. University of Minnesota, l9gl Joos, Martin. The Five Clocks. New York: Harcourt, 196l. Judy, Stephen N. Explorations in the Teaching_Lf Secondary English. New York. Dodd, Mead, l974. Kafka, T. Toby. A Stud of the Effectiveness of Four Motivational Stimuli on t e uETit of Comp051t1ons of Intermediate Student ts 1n'ofie c 00 'DTstrict. *Diss.'_§t. Johns University, __71______ Kelly, George A. A_Theorz.gf_Personality. New York: Norton, 1963. Kelly, Lou. "Is Competent Copyreading a Violation of the Students' Right to Their Own Language?" College Composition and Communication, 25 (l974), 254-258 King, Richard P. Sensory Approach to Creative Nritipg; A Study_ of the Effect of Increas1ng thF Number Lf Types of Sensor Stimul1 IntFnded to Motivate Children to Write _Creativejy_. Diss. Univers1tyo of’North Dakota, 1973. Klaus, Carl. "Public Opinion and Professional Belief." College Composition and Communication, 27 (l976), 335-340. Klein, Marvin L. and Grover, Burton L. An Assessment Lf the Effec- tiveness of S mbolicL Lo ic in the TeaChing of Composition. ymout M‘Wfis. , C‘Document Reproduct1on Service ED 059 209. 2l3 Koch, Carl J. Small Groups in the Composition Class: A Case Stud ‘gf_Develppin Linguistic Securit ppngritten FTuency. 155. University 0%Mifl1igan, 1975. Lacampagne, R. J. _A National Stpgygj Selected Attitudes and Approaches pp_Writing pf lyplfth-Grade Students With Superior Writ1n Performance Versus Those With Average Writing Performance. iDiss. University of Illinois, 1969. Lagana, Jean R. The Development, Implementation, and Evaluation pf a Model for Teachin Composition which Utilizes Individualized Learningfafid eer Groupigg. Diss. University of Pittsburgh, 1972. Language and Languagg_Behavior Abstracts. LaJolla, Calif.: United States International University. Lindeman, Richard E. Educational Measurement. Glenview, Ill.: Scott, Foresman, l967. Lundsteen, Sara W. Ed., Help for the Teacher of Written Composition. Urbana, Ill.: ERIC Clearinghouse on Reading and Communication Skills, 1976. Lyman, R. L. Summary prInvestiggtions Relating_pp_Grammar, Language, and Compos1t1on. UniVersity of Chicago Supplementary E uca- tional Monographs, No. 36. Chicago: University of Chicago, 1929. Maloney, Henry B. Ap_Identification‘gf_Excellence jp_Expository Composition Performance _i_r_1 _a_ Selected 9A Pppulation With _a_n_ Anal sis pf Reasons for Spperior Performance. Diss. Columbia Un1versity, l967. Marzano, Robert J. "The Sentence Combining Myth." English Journal, 65 (l976), 57-59. McClatchey, Barbara and McClatchey, Joseph. "The Writer's Audience: What is its Effect on Quality of Student Composition?" AArizona English Bulletin, l2 (April l970), 20-23. McElwee, Gregory W. Systematic Instruction jp_Proofreading_for Spelling and Its Effects pp_Fourth and Sixth Grade Composi- tion. iss. UniVersity of Wisconsin-Madison, l97l. Meckel, Henry C. "Research on TeaChing Composition and Literature." In N. L. Gage, ed., Handbook 2: Research pg_Teaching. Chicago: Rand McNally, l963. Mellon, John C. Transformational Sentence-Combining: _A_ Method 1931 Enhancing the Development pf_S ntactic Fluenc .ip En lish CompoSition. National Counci o eac ers o Englisfi Research Report No. 10. Urbana, Ill.: NCTE, 1969. 214 Memering, Willard D. Recent_Theorig§ and Practices 1p the Teaching f Composition. Diss. Florida State University, l97l. Miller, Barbara and Ney, James W. "The Effect of Systematic Oral Exercises on the Writing of Fourth-Grade Students." Research 111 the Teachinggf English, 2 (T968), 44-6l. Miller, Tyree J. ‘A Quantitative Study pf the 'Free Modifiers' in Narrative-Descriptive Compositions Written py Black Colldge freshmen After Leavingt e n uence pf_The Christensen Rhetoric Prggram and 3 Study of Their Attitudes Towardi Written Composition. Diss. Bdil State university, 1972. Mills, H. W. ,A Comparison ngthe Effects pf Oral Proofreading and Silent Proofreadin gj_Rou h Drafts of Compositions in Helping Sixth Grade Children Develpp_Greater_§entence SensitiVity. DiSs. University of Kansas, 1 Mischel, Terry. "A Case Study of a Twelfth-Grade Writer." Research lg the Teaching pf_English, 8 (1974), 303-3l4. Moffett, James. Teaching the Universe pf Discourse. Boston: Houghton Miff 1n, T968. Morrisroe, Michael and Morrisroe, Sue. "TESL: A Critical Evaluation of Publications, l96l-l968." Elementary English, 49 (l972), 50-61. Nakamura, Richard. "What is the Relationship Between the Amount of Reading and the Quality of Writing by Tenth Grade Boys?" Arizona English Bulletin, l2 (April l970), 24-28. Nebraska Curriculum Development Center. [A Curriculum for English. Lincoln: University of Nebraska Press, 1966. Neisser, Ulric. Cognition and Reality. San Francisco: W. H. Freeman and Company, 1976. Ney, James W. "Notes Toward a Psycholinguistic MOdel of the Writing Process." Research jp_the Teaching_pf English, 8 (1974), l57-169. Norwood, Mary Catherine D. Achievement pg Related tg the Ethnic Origin of the Student ip_gp_Experience-Centered Approach pg_ the Teadhin pf Freshman ngposition. Diss. Un1vers1ty of Tdea, [974. Oberchain, Anne. Effectiveness pf_the Precise Essa Question in Egg: rammin the'SequentiaTDeveiopment pf, r1tten omposiiion ghills and'the SimUltaneous Developmeptflpf_CriticaT‘Readin Skills. *Master's thesis, GeorgeWashington University, l97l. 215 Odell, Lee. "Measuring the Effect of Instruction in Prewriting." Research in theerachjng_ j English, 8 (1974), 228-240. Ofsa, William J. ‘An_Experiment in Usin Research in Composition in the Training of Teachers of n ish. DissT—University of‘— ITTin01s at Uhhana-Champaign, i974. O'Hare, Frank. Sentence Combining: _lmproving Student Writing Without Formal Grammar Ihstruction. National Council of Teachers of English ResearEhReport No. l5. Urbana, Ill.: NCTE. l97l. Olson, David R. "What Is Worth Knowing and What Can Be Taught." School Review, 82 (Nov. 1973), 27-43. Perry, William 6., Jr. Forms pf Intellectual and Ethical Development }p_the Collegg_Years. New Ydrk: Holt, Rinehart and Winston, 6B. Piaget, Jean. The Lan ua e and Thought of the Child. 3rd ed., 1926; rpt. Lohddn: Routiedge & Kegan Pdd1:_l95§. Pinkham, R. G. Ing_Effect gn_the Writt§n_Expression gf_Fifth Grade Pupils pf g_Series prLessonsEmphasizing the Characteristics of Good Writing as Exemplifiedl1p_Se ected Works from pp; 'AFea p;_C 1 renTE Literature. Diss. University of Connecti- cut, 68. Postman, Neil and Weingartner, Charles. Teaching p§_g_Subversive Activity. New York: Delacorte Press, 9. “Recommendations to the Profession." ADE Bulletin, No. 6 (Sept. l975), 58-60. Rogers, Carl R. Freedom pp Learn. Columbus, Ohio: Merrill, 1969. Qn_Becoming g_Person. Boston: Houghton Mifflin, 1961. Rohman, D. Gordon and Wlecke, Albert 0. Pre-writing; Ihg_Construc- tion gng A lication pj_Models for Concepthormation jn_ Writin . U E Cooperative Research Project No. 2|74. East Lansing, Mich.: Michigan State University, 1964. Sager, Carol. Improving tpg_ ualit ijWritten_Composition Through Pu il pp; p_ g atin ca e. Diss. Boston Ufiiversity School 0 ucation, l9 . Sanders, Sara E. A_Comparison of 'Aims' and 'Modes' A roaches tp_ the Teaching 91‘ Lmior C—ollege Freshman Compos1tion Both hfiih and 1thout an Auxilia Writing Lab. Diss. University of Texas at—AustihT51973. -——' 216 Sawkins, Margaret W. The Oral Responses of Selected Fifth Grade Children pg uestions Concernin ‘Their Written Expression. Diss. State n1vers ty 0 ew ork at Buffalo, 9 Schneider, Virginia L. A Stud of the Effectiveness pf Em hasizin the Teachin of Readin SFiTTE'tp_Improve Composition Shilis ih_Remedial Ehglish Classes at Kansas Cit Community Junior u 'Sdllege. Diss. n1versity od’kansas, 9 . Shafer, Robert E. “The Crisis in Knowing About Learning to Write." ADE Bulletin, No. 6 (Sept. 1975), 52-57. Shandloff, Lucille. The Relationshi .pf Freshman Com osition Curricu- lum Practicesdifii or1 a u ic Communitydunior Colleges pp. Research jp_the Teachin prWritten Composition. D1ss. lorida State Univer§it , l973. Shapiro, Phyllis P. and Shapiro, Bernard J. Poetry Instruction: lp§_ Effects pp Attitudes Toward Literature and'the Abilit pp. Write Prose. 197l. ERIC—Document Reprdduction erv1ce E57658 276 Shaughnessy, Mina P. "Basic Writing." In Gary Tate, ed., Teaching Composition: lg Bibliggraphic Essgys. Fort Worth: Texas Christian UniversityPress, 1976. . Errors and_§gpectations. New York: Oxford University Press, 1977. Sherwin, J. Stephen. Four Problems in Teaching En lish: A Critigue pf_Research. Scranton, Pa.:-Thternat1onal Text ooh Company, 969. Slay, Allan L. A Comparison 211 the Effectiveness p: Progranmed, Hand- book, and Non-Formalized Grammar Instruction in Remedial CoilegE_Freshman English Composition. Diss. SE1 Louis Un1versity, . Smith, Doris I. Effects of Class Size and Individualized Instruction pp png_Writin pjfhfighSchool thiors. Diss.iFlorida State University, 974. Smith, Frank. Com rehension and Learning. New York: Holt, Rinehart and Winston, l975 _ Smith, Myrna J. "Bruner on Writing.“ .Qpllegenggmposition and Communication, 38 (l977), 129-I33. Snow, Richard E. "Representative and Quasi-Representative Designs fOr Research on Teaching." Review pf_Educational Research, 44 (l974), 265-29l. 217 Squire, James R. "Research Can Make a Difference." Research 1_ EEE Teaching pf English, 10 (1976), 63-65. Stallard, Charles K. An Anal sis of the Writin Behavior of Good Student Writers. Diss. Uniiehgity of Virginia, l972. Standards for Educational and Ps cholo ical Tests. Washington, D.C.: AmeFican Psychological soc1ation, Inc., 1974. Standish, Patricia. In "Shoptalk." Arizona English Bulletin, 12 (April 1970), 51. Stevens, Alfred E. The Effects pf Positive and Nggative Evaluation pn_tng_Written Com osition pf Low Performin High School Students. Diss. Boston niverSity SEhool o E ucation, 1973. Stewart, Donald C. "Essays, Analysis, and--Better Writing?" Research inthg Teaching 91: English, 3 (1969), 42-51. Stiff, Robert. "The Effect Upon Student Composition of Particular Correction Techniques." Research ifl.£fl£ Teaching pj_English, l (1967), 54-75. Stotsky, Sandra L. "Sentence-Combining as a Curricu1ar Activity: Its Effect on Written Language Deve10pment and Reading Comprehen- sion." Research 1p_tpg Teaching ijEnglish, 9 (1975), 30-71. "Students' Right to Their Own Language." Special Issue, College Composition and Communication, 25 (Fall, 1974). Sutton, Doris G. and Arnold, Daniel S. "The Effects of Two Methods of Compensatory Freshman English." Research in pnngeaching Lf English, 8 (1974). 241-249. Sweet, Roger C. Educational Attainment and Attitudes Toward School as a Function of Feedbaék in the—Fdrm ofrTeachers‘lWritten panama 537 Mcfieh't—Rrprodddt—F—ion ervic“—e ED 015 Taylor, Philip H. "The English Composition in the Junior School: Prepared or Unprepared?“ ‘Egucational Research, 5 (1962), 57-62. Troyka, Lynn Q. A Study _o_i: EDS. Effect gt Simulation-Gaming 91 Ex ositor Prose Cpmpetence of Colle e Remedial n lish Composition Students. *Diss.—New YorE University, i973. Turner, Geraldine C. Experimental Research in the Use of p_Writing Lab. St. Petersburg thior College,_l97OT'-ERIC—Document ‘Rddroduction Service ED 041 576. 218 VanFleet, Ellamaye. Guidelines for Determining the Content of a Colle iate Re art-Writing'tburse. Diss. lUfiivers1tyo 61' Tennessee, lSES. Varon, Ruth L. Examination Lf the Content p£_Unsolicited Compositions Written W_y. UpperElementary School ildren forFTheir TeaCHér' s 155 . Eyes ny. 1a University,m Vygotsky, L. S. Thought gng Languagg. Cambridge, Mass.: MIT Press, 1962. Wahlberg, William A. The Effect Lf Process Intervention L".£QE Attitudes and Learningin_ a Colle e Freshman Composit1on Course. llDiss. Un1versity of gan, Walsh, E. Michael. "HoW Can Colleges Deal with the Increasing Inci- dence of Student Weakness in Basic Learning Skills?" Change, 9, No.3 (1977), 50-52+ Whitehead, Frank. The Disappearing Dais. London: Chatto B Windus, l966. "IIIIIIIIIII'MIHIIIIT