UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The color-coded cloze procedure : a method to assist adult ESL students in searching for clues to fill… Labrum, Howat Alan 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1992_fall_labrum_howat_alan.pdf [ 8.34MB ]
JSON: 831-1.0078110.json
JSON-LD: 831-1.0078110-ld.json
RDF/XML (Pretty): 831-1.0078110-rdf.xml
RDF/JSON: 831-1.0078110-rdf.json
Turtle: 831-1.0078110-turtle.txt
N-Triples: 831-1.0078110-rdf-ntriples.txt
Original Record: 831-1.0078110-source.json
Full Text

Full Text

THE COLOR-CODED CLOZE PROCEDURE: A METHOD TO ASSIST ADULT ESL STUDENTS IN SEARCHING FOR CLUES TO FILL IN CLOZE BLANKS  HOWAT ALAN LABRUM  B.A., The University of British Columbia, 1972 Dip, Ad. Ed., The University of British Columbia, 1973  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Language Education)  We accept this thesis as conforming to the requi standard  THE UNIVERSITY OF BRITISH COLUMBIA October, 1992  0 Howat Alan Labrum,  1992  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of  L  &U11C.  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  OCTo6.I.  )1)  1992  ELI) iCIi r,,v  1:1.  ABSTRACT Research suggested  by  that  Shanahan students  and did  Kamil not  (1982, use  1983,  inter-sentential  information to improve their doze test scores. Oiler  (1977)  had  discovered  in  their  proficient ESL students the contrary. that the  issue  of too-local  1984)  Chihara and  research  with  more  Building on the idea  reading depended on both the  proficiency and motivation of the students along with the availability of beyond—sentence clues, the present research project was designed using color-coded blanks and parts of speech to highlight inter-sentential clues,  especially re  iterative—word clues. The  purpose  of  the  color—coding  differentiating blanks and words,  was  to  see  if,  by  students could focus more  sharply on the necessary information, primarily re—iterative words that occurred beyond the sentence of each doze blank, and use these words to fill the doze blanks.  The students  in the treatment group were given a key to the relationship between  the  color  and  the  parts  of  speech  but  were  explicitly told to look for re-iterative-word clues. expected, however,  not  It was  that each colored blank would stimulate  the students to look through the passage to find a word of the same color to fill the said blank. to  maximize  the  number  of  It was also the goal  inter—sentential  re—iterative  clues and see how c].oze scores would be affected when color—  iii coding was used as opposed to no color—coding  (the major  control). If  the  inter—sentential  clues  were  present,  then  color—  coding should have resulted in the treatment group having better scores.  Randomly—colored words were used as another  control to determine whether any positive effect had come from the coding or just from color as a motivational factor. Lastly, the pre- and post—tests were non—colored in order to see if the color—coding treatment had just a temporary means of help which led to no lasting gains once removed.. A pilot project was done with a class of twenty adult upper—intermediate  ESL  standard doze test.  students  using  four  forms  of  a  Based on the results the colored—blank  form was dropped so the three classes in the main project could each have one form.  To suit the needs of the more  advanced university students,  new stories were chosen and  prepared using a rationalized doze to maximize the number of re-iterative—word clues.  In the main research project  most of the random-color group (the least proficient group) dropped out after the pre—test. doze  group  received  higher  The non—colored rational  scores  than  the  color—coded  treatment group on all of the tests.  When the mean scores  were graphed both these groups made  steady progress  practice  test  appearing to  to  practice  almost catch up.  test,  the  treatment  Improvement was made  from group from  pre- to post—test by both groups but less by the treatment  iv group,  especially  counted.  when  only  Generally speaking,  inter—sentential  blanks  were  t—tests and a very sensitive  statistical program (“One Between and One Repeated Measures Factor ANOVA”)  confirmed this  improvement but  showed that  the color—coded doze treatment group and the non-colored doze  control  group  in  most  cases  did  not  differ  significantly. The  graphical  analyses  of  the  results  were  more  optimistic in favor of the color-coded treatment than the statistical analyses were but the small sample size (N made the statistical findings unclear at times. in  some  cases  may  have  been  because  of  the  =  13)  Improvement declining  readability levels of subsequent passages. The  number  correctly,  of  blanks  filled,  the  number  filled  and the relationship of these two were analyzed  to determine the confidence and productive confidence levels of each  group.  Results  showed the  color—coded treatment  group were less confident in filling blanks and made limited gains in productive confidence over the control group.  V  TABLE OF CONTENTS  ii  2.BS’]?R..ACT TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES ACKNOWLEDGEMENTS AN OVERVIEW OF THE RESEARCH PROJECT CHAPTER ONE: Goal of the Present Research Using the Color Coded doze Procedure Significance of Studying Color-Coded doze Procedure Scope of the Study Definitions of Terms A REVIEW OF THE LITERATURE CHAPTER TWO: Introduction to the Review of the Literature General Setting What is the scope of the research’ What is traditional doze procedure’ What are some of the controversial areas of the doze procedure What are some considerations about students’ motivation and proficiency’ What are the recent education does color—coding as a trend fit How into the present research’ Key Area of Investigation Underlying This Research Paper How do discourse and the nature of doze overlap’ Is doze procedure essentially global or  V  ix Xiii XV 1 1 6 8 12  20 22 24 26 27 32 41 48 55 55 56  What is the chronology of the global\discrete controversy Halliday and Hasan (1971) Caroll (1972), Chihara and Holler (1977) Yamada (1979) Thomas (1980) Shanahan and Karnil (1982) Bridge and Winograd (1982) Leys (1983) Shanahan and Kamil (1983) Shanaharz and Kainil (1984) Henk (1985) Backman (1985)  60 60 60 62 64 68 69 72 73 75 77 79  vi What is the nature of non—native speakers . . . . . of Eng1ish What is the possible role of color in the improvement of non—natives’ doze? . reading scores?..... . . . . . . . . . . What are the contributions of the key areas of investigations to the present research?..... . . . . . . . . . Results of the Review of the Literature Introduction . . . . . . . . . . . . Overview of the method proposed in relation to the 1 iterature . . . goals of the present doze  .83  . . .  . .  .85 .88 89 89  . . . .  89  89 research . . . . . . . . . . . . . . . . . . . . . . ..........09 useofpreviousresearch Implications of review of literature for the present research paper. .. . . .. .. . . . .92 research questions. 95 95 hypotheses . . . . . . . . . . . . . . . . . 99 .99 . . . . . . . . . . . . . . . . . .102 Subjects.... •....... . . . 103 Treatment and Control Groups......... Treatment and Control Passage Forms.............106 106 Nature of the forms. . . . . . . . . . . . . . Method of coloring the passage...... . . . ...... . .108 ..108 Stories Used for the doze Passages..... Introduction to the selection ofthepilotpassages.... 108 Rationale for the type of text analysis (pilot project)........ 110 Presentation of the text analysis form (pilot project) . . . . . . . . . . . . . . . . . . . . . . . . . . .111 Rationale for the type of text analysis (main project).............. 112 Presentation of the text analysis (main project) . . . . . . . . . . . . . . . . . . . . . . 113 Administration of the doze Passages. 115 Overview of the distribution of the text and practice doze passages (pilot project) . . . . . . . .115 Length of time required (pilot project).. .116 Overview of the distribution of the text and practice doze passages (main proj ect) . . . . . . . . . . . . . . 116 Length of time required (main project) 117 Roles of the researcher, teachers, and the students. . . . . . . . . . . . . 118 Scoring of the doze Tests and Data Analysis..... .. . . . . . .. . . .. . . . 121  CHAPTER THREE:  RESEARCH DESIGN AND METHODOLOGY.  . ..  Introduction...... •......... . . . . . . . . . . . . . . . . . . . . . . . . The  The The The  The  The  vii ANALYSISANDDISCUSSION........  CHAPTERFOUR:  Introduction.  .  .  . .  .  . .  .  .  .  .  .  .  . . .  .  .  ...130 . . 130  .  .  .  . . . . .  .  .  .  .  .  .  .  .  .  Changes Made to the doze Procedure after the Pilot  Project......  .  . •.........  .  . . .  . . .  Problems and Solutions in Pilot and Main Projects . . . . . . The Research Data The Pilot Project . . . . . . . . . . . . The Main Project . . . . . . . . . . An overview’ Asumxnaryofthedata Anexplanationofthedata A closer look . . . . . . . . . A statistical look at the data Thehypothesesandresults The problems with the present research design...... . . . . . . . . . . . . . Hope given by the questionnaire The conclusions drawn from the analysis of the productive confidence scores........ . .. . . .. . . IMPLICATIONS OF THE RESEARCH.. CHAPTER FIVE: Introduction. . . . . Future Research Designs BIBLIOGPHY.  . .  . . . .  . . .  .  .  .  .  .  .  .  . .  .  .  .204  •.......  220  APPENDIX C: DISTRIBUTION OF RESEARCH TASKS Role of the Researcher. . . Instructions for Teachers Instructions for Students. . .  .  .200 201  208 208 210  .. .  .  .  .  . . .  .  . .  .  .  .  .  ..  .  . . . . .  ..  .  PRODUCTION OF PILOT CLOZE PASSAGE FORMS..  SPREADSHEET CHART...... .. APPENDIX D: For Collection of Answers and Determination of Scores......  .133 .135 135 139 139 140 175 175 180 198  .  APPENDIX A: PILOT PASSAGES . .. . The Original Texts.... . . . . . . . . . The Color—coded Fornis . . . . . . . . . . . . . A Sample Answer Key...... . . . . . The Distribution of Deleted Words An Analysis of Re—iterative Clues...... APPENDIX B:  .131  .  .  . .  .  .  .  ....  .  ...  .  . .  .  .. .  .  .  .230 .230 243 254 .261 251 .263 .269 .269 272 277 .281 281  viii APPENDICES FOR MAIN PROJECT MAINPROJECTPASSAGES..... APPENDIXE: . . . . The Original Texts.  288  The Color—coded Forms . . . . . . . . . . . . . A Sample Answer Key Sample of distribution of re-iterative words  . . . . .  . . . .  . . • .  .288 . 299 309 311  . . . . . . . . . . . . . . .  APPENDIX F:  PRODUCTION OF THE CLOZE PASSAGES  APPENDIX G:  COMPARISON OF EXACT- AND ACCEPTABLEWORD I4EA}I SCORES . . . . . . . . . . . . . . . . . . .  QUESTIONNAIRES........ ... .. APPENDIX H: FormsforGroupsl,2,and3......  313  .... . .  . . . . . . .  . ... . . .  .. . . ..  316  .320 320  ix LIST OF TABLES 34  TABLE 1  Overall Reading Comprehension Strategic  TABLE 2  Quotes about Recent Education Trends  TABLE3  VariousKindsofCohesion..  .98  TABLE 4  Proficiency Classification of Students  103  TABLE 5  Configurations of the Standard doze Procedure and Variations.....  TABLE 6  107  . ..  Number of Blanks/50 Deletions whose Fillers do not Appear at Least Once Somewhere in the Passage; and the Number of Words in the Passage  . ...  . .... . ....  .111  TABLE 7  An Analysis and Ordering of the doze Passages Based on a) the Number of Non—re—iterative Clues for the Blank Fillers and b) Readability 112 Fou1as. . . . . . . . . . . . . . . . .  TABLE 8  Number of Blanks/50 Deletions whose Fillers do not Appear at Least Once Somewhere in the Passage; and the Number of Words in the Passage.  113  .. ... . .. ... .. ..  TABLE 9  An Analysis and Ordering of the Cloze Passages Based on a) the Number of Non-re-iterative Clues for the Blank Fillers and b) Readability Formulas....................114  TABLE 10  Number of Re—iterative Blanks..................1l4  TABLE 11  Distribution of the Treatment, Variations and Control Cloze Tests, with Variations Applied to Standard (every nth word) Cloze Procedure Format . . . . . . . .  115  TABLE 12  Distribution of the Treatment, Variations and Control Cloze Tests, with Variations Applied to Rational Cloze Procedure Format in which Deletions were Chosen to Maximize Number of Intra- and Inter—sentential Clues....117  TABLE 13  Answer Collection and Blank Score Determination Chart. . . . .  .  124  TABLE 14  Criteria for Development of Weighted Scores....127  TABLE 15  Calculation of Cloze Passage Scores............127  x TABLE 16 TABLE 17  Answer Collection and Blank Score . . Determination Chart . . . . .  129  . . .  131  changes to Cloze Procedure after Pilot Project. . . . . . . . . . . . .  TABLE 18  . .  . . .  Problems and Solutions in Pilot and . . Main Projects.  . . .  .133 143  TABLE19  TestData:Exact—WordScores  TABLE 20  Mean Scores for All 50 Cloze Blanks............147  TABLE 2].  Productive Confidence Scores for All 50 doze Blanks  TABLE 22  . . .  . . . . .  .149  Mean Scores for Intra- and Inter—sentential Blanks  TABLE 23  . . . . . . . . .  ... .. .  *  .154  . . .. .  Productive Confidence Scores for Intra- and . ... . . . .. Inter—sentential Blanks  .  ..l56  TABLE 24  Mean Scores for Intra-sentential Blanks........161  TABLE 25  Productive Confidence Scores For Intra—Sentential Blanks... .... .  ...l63  TABLE 26  Mean Scores for Inter—sentential Blanks  ...168  TABLE 27  Productive Confidence Scores for Inter—sentential Blanks...... . . ..  170  . .... .  TABLE 28  T—tests for the Main Project..............  . .  .183  TABLE 29  Significant Differences......  . .  .186  TABLE 30  Statistical Program: Trend Analysis  TABLE 31  Data Collection Table for ANOVA  TABLE 32  Summary Table: ANOVA.(All 50 Blanks)  TABLE 33  Summary Table: ANOVA (Intra- and Intra sentential blanks) . . . . . .  ... . . . . .. .. .. .  ..J.87 ...188 .189  . . . . . . .  189  Summary Table: ANOVA (Inter-sentential blanks)..... .. . ... . •  190  TABLE 35A Student-Newman-Keuls Test for Variance Score (A) (All5oBlanks)  ......19l  TABLE 34  •1•11•••••••  xi TABLE 35B Student-Newman-Keuls Test for Variance Score (B) (All5OBlanks)  191  TABLE 36A Student-Newman-Keuls Test for Variance Score (A) (Intra- and Inter-sentential Blanks)  . . . .  192  . . . . . . . . . . . .  192  . . . . . . . . .  . . . . . . .  TABLE 36B Student-Newman-Keuls Test for Variance Score (B) (Intra- and Inter-sentential Blanks)  . . . . . . . . . . . .  .  TABLE 37A Student-Newman-Keuls Test for Variance Score (A) (Inter—sentential Blanks) .  . .. .  .193  TABLE 37B Student-Newman-Keuls Test for Variance Score (B) (Inter—sentential Blanks)..  TABLE 38  General Linear Models Procedure  193  (All 50 194  Blanks) . . . . . . . . . . . . . . . . . . . . . . . . . . .  TABLE 39  General Linear Models Procedure (Intra— and Inter—sentential Blanks)...... .... . . .. .. . ...... 194  TABLE 40  General Linear Models Procedure (Intersentential Blanks) . . . . . . . . . .  194  TABLE 41  Post—session Questionnaire.........  . . .... .. . . .  .202  TABLE 42  Treatment Group’s Confidence Level.  . . .. .. .. . . .  .206  TABLE 43  Suggestions for Possible Future Research 213  Designs. . . . . . . . . . . . . . . . . . . .  APPENDIX TABLE A  An Analysis of Re-iterative-Word Clues.........261  TABLE B  Method of Color-Coding and Printing Passages...263  TABLE B.l Meanings of the Symbols Used for Color Coding  . . . .  . . . . . . . . . . .  265  TABLEC  RoleofResearcher..............  TABLED  InstructionsforTeachers......................272  TABLEE  InstructionsforStudents  TABLE F  ..........269  ..............277  Special Instructions for Students in the Treatment  Group . . . . . . . . . . . . . . . . . . . . . . . . . .  280  xii TABLE G  Answer Collection and Blank ScoreDetermination Chart (Pilot Project)  281  . . .. .  TABLEH  SampleofPassageAnalyses.....................3ll  TABLE I  Answer Collection and Blank ScoreDetermination Chart (Main Project) ..  TABLE J  Exact- and Acceptable-word Scores Blanks  TABLE K TABLE L  .315  (All  . . . . . . . . . . . . .  . . . .  Exact- and Acceptable-word Scores . . Inter—sentential Blanks  (Intra- and  Exact- and Acceptable-word Scores  (Intra  sentential  TABLE M  .. . . . . . . ..  317  Blanks . . . . . . . . . . . . . . . . . . . . . . . . .  Exact- and Acceptable-word Scores sentential Blanks. . . . .  316  318  (Inter . . . . . . . . . . . .  .319  xiii LIST OF FIGURES Pilot Project GRAPH 1  All 50 Blanks: Four Forms, EXACT-word scores (in %) .  .137  GRAPH 2  All 50 Blanks: Standard vs CCC, EXACT-word . . . scores (in %)  .138  . . .  Main Project (in %)..  148  GRAPH 3  All 50 Blanks: EXACT—word Scores  GRAPH 4  All 50 Blanks: Total no. of Blanks Filled......150  GRAPH 5  All 50 Blanks: Percent of Blanks Filled .. . . .. .... . to Possible  GRAPH 6  All 50 Blanks: Total No. of Blanks Filled  ..... . ..  Correctly GRAPH 7  All 50 Blanks: Correct Blanks/No.  (in %) GRAPH 8 GRAPH 9  .  .  151  . . . . . . 152  of Filled  . . . .....  Intra- and Inter-sentential Blanks: EXACT. .. . . word Scores (in %)  153  . .155  Intra- and Inter-sentential Blanks:Total  no • of Blanks Filled . . . . . . . . . . . • . . . . • • . . . . • . . . . 157  GRAPH 10  Intra- and Inter-sentential Blanks: Percent of Blanks Filled to Possible...................158  GRAPH 11  Intra- and Inter-sentential Blanks: Total No. of Blanks Filled Correctly....... . . .... . . . .159  GRAPH 12  Intra- and Inter-sentential Blanks: Correct Blanks/No.ofFilled(in%)  GRAPH 13  Intra-sentential Blanks: EXACT-word Scores  (in %) .  .  . . . . . . . . .162  GRAPH 14  Intra-sentential Blanks: Total no. of Blanks • Filled...... . . . . . . . . . .  GRAPH 15  Intra-sentential Blanks: Percent of Blanks . . . . . . 165 .. . . . F ii led to Pc s S ib 1 e  .164  xiv GRAPH 16 GRAPH 17  GRAPH 18  Intra-sentential Blanks: Total No. of Blanks . . . . . Filled Correctly...... Intra-sentential Blanks: Correct Blanks/No. . of Filled (in %)  %)  . .. . . •....... ..... ... ... ..... . . . .  . . .... . . . ... . .. .... ... .. •..  .171  to  Possible . . . . . . . . . . . . . . . . .  . . . .  17 2  Inter-sentential Blanks: Total No. of Blanks Filled  GRAPH 22  .169  Inter-sentential Blanks: Percent of Blanks Filled  GRAPH 21  167  Inter-sentential Blanks: Total no. of Blanks Filled....  GRAPH 20  .166  Inter-sentential Blanks: EXACT-word Scores (in  GRAPH 19  .  Correctly. . . . .  . . . . . . . . . . . . . . . . . . .  173  Inter-sentential Blanks: Correct Blanks/No. of  Filled  (in  %)  . . . . . . . . . . . .  ........ . . . . . . .  .174  xv ACKNOWLEDGEMENTS I would  like to thank my advisor,  Dr.  Kenneth Slade,  for his kind and patient guidance in my efforts to complete this  I wish to thank Dr.  Also,  thesis.  statistics advisor,  for all his  To Dr.  in  my  Lee Gunderson,  work on and insights  procedure with ESL students. participated  my  for sharing his special knowledge in a  very warm and enjoyable way. you goes  Walter Boldt,  into  a thank  the  doze  The teachers and students who  color—coded  doze  research  deserve  a  medal for their time and effort. As  is traditional  in writing thesis acknowledgements,  and now more deeply felt and appreciated, parents, wife, many ways. Arabia,  thanks go to my  and children for supporting my efforts in so  To my  family  and my  friend,  Omar,  in  Saudia  I owe a debt a gratitude for making me keenly aware  of how important it is to persist in the opportunities one is given.  1  CHAPTER ONE AN OVERVIEW OF THE RESEARCH PROJECT  THE GOAL OF THE PRESENT RESEARCH USING THE COLOR-CODED CLOZE  PROCEDURE One  goal  of  teaching  reading  comprehension  learners of English as a Second Language them  process  a  passage  in  as  Richards maintained this view  large  (ESL)  chunks  (1980)  to  adult  is to help  as  possible.  and pointed out that  “learners need to develop strategies for interacting with a native meaning  speaker from  so  that  discourse  propositional  are  rather  content  Halliday and Hasan  they  of  (1976),  able  than  single Long  to  extract  being  limited  sentences”  (1981),  total to  64).  (p.  and. Mohan  the  (1986),  supported the same position that coomunication goes beyond the sentence level and involves interaction with a cohesive text. To  students  processing  (as  isolation)  in  the  proposed  opposed  to  advantage  processing  understanding  of  a  of  each  passage  discourse  sentence is  in  two-fold.  First, having a broad purview could possibly allow them to take advantage  of the built-in redundancy  in the text to  fill  in  unknown  in  gaps  information  caused  by  especially when a dictionary is not at hand,  words,  or if handy,  would take the student away from the text for a long period of time. reader:  Ashby-Davis  (1984)  pointed out that the normal  2  must check overall understanding of a text or at least inter- and intra sentence meanings to make educated guesses concerning the meanings of unknown words. (p.87) Second,  a  clearer  understanding  wide  purview  could of  lead  the  adult  needed  students  to  a  words.  Thus  a  proficient adult ESL reader can be defined as  one who  is  able to find the most suitable words to fit the meaning of the unknown words in a written passage. It is the goal of this research paper to suggest and test a method of helping adult ESL learners to become more proficient readers. the  review  of  the  The method, to be explained further in literature  in  research design in Chapter Three,  Chapter  Two  and  in  the  is a modified version of  the doze procedure. Henceforward it is referred to as the “color—coded  doze”  (CCC).  rationalization of the  Also  to  be  explained  is  CCC in which as many deletions  a as  possible are made to words which appear more than once in the text. The CCC follows the standard doze procedure in that:  a) every “nth” word, commonly the fifth, is deleted in a passage and replaced with a standard length blank; b) the first and last sentence of a selection are left intact to give the students an idea as to the general direction of the meaning of the text; and C) the passage is about 300 words in length allowing for 50 blanks.  3  The CCC is different from the standard doze procedure in that every word and blank in the CCC passage was given a particular color to designate its part of speech.  Also the  CCC procedure in this research counted proper nouns in the deletion process. such  as  passage found,  This is because some of the proper nouns,  people’s  names,  and  were  thus  especially  when  were used more open  to  the  color—coding  than  once  possibility was  in the  of  applied  being  to  the  draw  the  passage. The students’  purpose  of  the  color—coding  was  to  attention beyond the sentence of the doze blank  to look for clues which would help the students to determine the suitable word to fill in the doze blank. was  in response to researchers such as  This purpose  Cohen  (1980),  who  pointed out that, indeed, there is a tendency for non—native speakers to look locally (within the sentence of the blank) for such clues. color—coded  The purpose of usina tarts of speech to be  was  to  take  advantage  of  the  information  inherent in those parts of speech. including the rammatica1 re1ationshis.  Cohen stated that “doze can also be used to  check for awareness of grammatical relationships” He  illustrated  grammatical  relationships  agreement with the elements in the passage number,  person,  or  whatever)”  (p.  96)  as  (p.  96).  “grammatical  (tense, as  gender,  shown  by  inflections on the parts of speech, resulting in cohesion in the  discourse.  In  making  the  above  statement  and  4  clarification, Cohen gave his support to the importance of grammatical information for doing the doze tests. Notwithstanding the advantages of the parts of speech and their grammatical relationships for  doing doze tests,  it was not the purpose of this research project to teach grammar as an end in itself. color-coded  doze  color—coding.  to  make  Instead,  Nor was it the purpose of the the  students  dependent  on  the  it was the purpose to remove the  color at some future point and to let the students complete the  standard  doze  their awareness  of  exercises, the  while  parts  of  still  speech  making  clues  use  beyond  of the  sentence to help them fill in the blanks. Given  examples  a)  in  the  research  literature on the  doze procedure of ESL students looking inter-sententially for clues to fill doze blanks and b) business ideas,  people  it  students  was to  sentential)  and  students  thought  look level  for to  that clues find  use  colors  the  CCC  beyond  to highlight key  method the  information  sentential level seemed promising. needed to be tested.  observations of how  of  local at  the  teaching (intra inter  However, the CCC method  As will be seen in Chapter Two, there  were a number of hypotheses that should be tested,  but at  this point the prime null hypothesis to be tested was stated as follows:  5  Students trained with the color-coded doze will show no  sianificant  doze  yost—test  difference after  in  scores  treatment  on  when  a  standard  comvared  to  students trained on standard doze vrocedures alone.  In Chapter Four if higher mean scores on the post—test are reported for the treatment group the inference will be made that CCC students were able to make more use of grammatical (syntactic) and lexical (semantic) clues than the students who received lower mean scores.  On the other hand, it can  be argued that the clues might not have come from beyond the sentence but from within the sentence of the blank, albeit with the help of the color-coding.  Though this problem and  the resulting hypotheses are better left to Chapter Two after a review of the literature on doze procedure has been done. Before proceeding, however, it should be noted here that, because of the necessary limited scope of this research project, there is an explicit focus on the following question.  Having trained using the CCC, do  students become more aware of inter—sentential same—word re iterative word fillers that match the deleted words, than students who have trained with non—color—coded parts of speech, color—coded blanks only, or randomly—colored blanks and words?  Clearly, re-iterative words of the kind ‘ust  mentioned above are the easiest type of cohesive words to quantify for purposes of identification in preparation of  6  the doze passages.  This and other types of cohesive words  are given and illustrated in Table 3 at the end of Chapter Two.  Their interaction in and effect on the CCC research  project can be analyzed and possibly determined by an analysis of the answers given by students on the pre—tests, practice tests, and post—tests.  SIGNIFICANCE OF STUDYING COLOR-CODED CLOZE PROCEDURE  As will become evident in the review of a portion of the abundant literature about the doze procedure, the doze procedure curriculum, example, doze  has  been  as well Schoenfeld  procedure  seen as  to  have  in other  (1980)  a  place  the  language curricula.  and Valmont  could make use  in  (1983)  students’  of  ESL For  suggested  knowledge  of  syntax, phonics, and word meanings to improve the students’ prediction and confirmation strategies. Other uses of doze procedure include: a) teaching (Soudek & Soudek, 1983 b) testing  —  1984),  ( Foley, 1983),  c) readability (Rye, 1982), ...but  the  doze  procedure  has  deficiencies,  attempted  solutions to which have generated many of the studies to replace  it  with  modified  versions.  through a series of experiments  Shanahan  and  Kamil  (to be outlined in Chapter  7  Two) have even tried to disparage the doze procedure as a teaching device.  This present study intends to build on the  strengths of the doze procedure by trying to overcome one of its possible weaknesses  its questioned ability to help  —  students look beyond the sentence of the blank to find clues to a filler for the blank (Shanahan & Kamil, 1982, 1983, and 1984; Leys, 1983). The use of color—coding as a so.ution to this supposed inability of students to look beyond the sentence needed to be  tested,  not  dust  as  an  intellectual  exercise,  but  determine whether there was sufficient benefit to  to  justify  the extra time and expense it takes to prepare color-coded passages. society  Certainly is  attracting  playing and  it a  can  be  greater  categorizing,  said  and and  that  greater  the  color role  means  to  in  our  both  in  producing  color—coded products are becoming more efficient and cheaper all the time.  Thus as the benefit-to—cost ratio continues  to increase, the use of color—coded c].oze procedure becomes more feasible.  This is all the more reason why the color—  coded doze can and should be tested in a controlled way. Because  color—coding  has  become  so  prevalent  in  our  everyday lives, even in a limited way in education, it seems ever more speech  and more plausible to use color—coded parts  in the teaching of ESL.  It  should be noted  of  that  Gattegno was using color—coding to teach phonics and to a lesser degree, grammar in 1957, and in lesser ways language textbook writers predate even this work  (Dale,  1926,  Hay &  8  Wingo,  The problem with color-coding has been and  1954).  may continue to be with the necessary standardization of the color-code, which in itself is difficult to attain, and the constraints this would put on teachers and textbooks if they did not want to confuse the students.  Scope of the Study The  color—coded  doze  experiment  paper was divided into two sections, different location urban setting  described  this  each carried out in a  in Greater Vancouver,  (population  in  a  1.5 million).  large  The  Canadian  first was  a  pilot Drolect in which 18 young adult ESL students, largely from the Pacific Rim and studying at a major private college in Burnaby, B.C., took part. their  college’s  They were defined according to  categorization  intermediate level.  system  as  nineteen— learners  to from  in  (chosen  from  twenty—two—year Japan,  program  Vancouver, a  total  year  (some  in  old  studying  in  at  five days a week in  of  university three  one  The  hundred  level  classes  the University  participated.  anthropology and political second  upper-  The second was the main prolect in which 60  Ritsumeikan English Columbia  the  The college students were in a program  in which they studied five hours a day, 14—week terms.  at  of  sixty  in  ESL the  British students  students)  were  science majors mostly in their  their  third  or  fourth  year)  at  Ritsumeikan University in Japan, visiting Canada for a nine-  9  month  English  Social  course.  Science  Engineering,  and  Law,  administration, such topics.  In  Japan  they  International Literature,  had  been  Relations, Economics  studying  Science and  and  Business  so the ESL course content revolved around  According to their TOEFL scores they were not  quite up to the required standard for university entrance. More specifics on their English proficiency levels will be given in Table 4 in Chapter Three. The language skill area being studied in the present research was  reading  and  learning  to  use  context  clues,  particularly parts of speech with a special emphasis on re iterative words  to help  find clues that would enable the  students to restore deleted words from passages of between approximately 295 and 330 words each. taken  from two books.  personalities,  The book,  written  Patricia Raymond  for  (1987),  ESL  These passages were  CANread, students  was used  for the  about and  Canadian  edited  college  by  class.  Canadian Society, A Macro Analysis, a college text by Harry H.  Hifler  Seven  (1991)  passages  was were  used  for  selected  the (one  stories) for the college students.  Ritsuiueikan each  from  students. separate  Six passages were chosen  for the Ritsumeikan students. With regards to the reading level of the pilot project stories,  Little  (1988)  in his review of Raymond’s CANread  stories and two ESL schools known to this researcher all assessed students.  her  stories  at  the  intermediate  level  According to a Plesch readability (1948)  for  ESL  rating  10  of the parts of the stories used in the pilot study, most of the stories were mostly in the grade seven to eight range, although  one  readability  was  (1977)  at  the  sixth  grade  A  rating gave somewhat the same  (For details please refer to Table 7 page 102.)  level.  Fry  results.  in Chapter Three on  Ad hoc trials of two of the stories with ESL  students had suggested that the doze forms of the CANRead stories  would  have  to  be  given  to  upper-intermediate  students or even higher if strategically possible. The passages for the Ritsumeikan students ranged from a readability level 12 to 15, with most being at level 15, the most difficult. sub—program  in  The assessment was done using a readability a  shareware  computer  Count vl.2 by Chris B. Sakkas.  program  called  Word  An analysis of the passages  will be given in Table 4 in Chapter Three. The CCC study revolved around the questions of how to get students to look beyond the immediate sentence of the blank to find the suitable (preferably the deleted) word and to show whether or not the students were, beyond the  sentence.  in fact,  The methodology was  to  looking  divide  the  students into groups and give each group the same standard doze passage with the groups kinds  were of  given  color  same deletions,  passages  clues  as  which is  had  but the various  explained  treatment  amounts  forthwith.  or The  passages of the main treatment group had all the words and blanks  color—coded  according  to  parts  of  speech.  The  passages of the secondary control group had dust the blanks  11  color—coded.  (In order to allow for more students to be  involved in the more important treatment cells, this blankonly colored treatment was not used in the main project.) The passages of the tertiary control group had all the words and blanks colored randomly.  The passages of the primary  (most important) control group had no color at all. In Chapter Two, will  focus  primarily  a selective review of the literature on  research which  has  a  bearing  on  finding an answer to the questions of how to help students look for inter—sentential clues and to show whether or not the students are doing so.  The review will also show the  basis of the research design.  Chapter Three will present  the resulting design of this two month study.  (The study  was carried out over two periods of about a month each, one period for the pilot and one for the main research project.) The results will be reported and commented upon in Chapter Four.  In Chapter Five conclusions will be made about the  effectiveness of the doze procedure and what influence, any,  color—coded parts  with adult ESL students.  of  speech had  on  it,  if  particularly  In addition, recommendations will  be given to improve both the color—coded doze procedure and the method of analyzing the results.  12  DEFINITION OF TERMS  doze  (color—coded)  the same as the standard doze with  —  the exception that in the color—coded doze all the parts of  speech  are  present  color—coded.  research  paper,  Also,  for  proper  the  nouns  are  purposes  of  the  counted  in  the  deletion process.  doze  (fixed-ratio)  in a fixed manner,  —  a passage in which words are deleted  e.g.  every 5th word is deleted.  It is  also called the “random doze”.  doze (modified)  —  a doze passage which varies in one or  more aspects from the traditional doze passage.  doze  (random)  a  —  passage  according to a set pattern, random  insofar  specific  as  words  the  which  in e.g.  deletor are  which  words  are  deleted  every fifth word. has  no  control  deleted  in  a  fixed  It is  over  the  manner.  (Another name for this is the “fixed—ratio doze”.  doze (rational) deleted, found  e • g.  all  elsewhere  iterative  —  words  a passage in which words are selectively nouns  in were  the  and verbs text  targeted  project of the present research)  at for  only.; least  words once  deletion  that  are  (these in  the  main  13  doze  (standard)  —  a passage  in which every nth word  is  deleted, with the exceptions that proper nouns are skipped over (not deleted or counted) and all words in the first and last sentences are left intact.  doze (traditional)  clozentropy  the  —  (see traditional doze)  the same as the standard doze  process  of  collecting  native  speakers’  responses to doze blanks in order to determine  possible  —  acceptable  responses  for  scoring  non-native  speakers’  answers.  cohesive  the  —  state of words  having a relationship with  other words.  color—coded parts of speech  —  see Parts of Speech:  (color  coded)  confidence  —  themselves  as  the  amount  of  faith  manifested  by  the  the  number  students of  doze  have  in  blanks  filled.  confidence (productive)  —  the number of correct answers as a  proportion of the number of doze blanks filled.  14  conjunctive words  are  -  the state of having a relationship in which  joined with conjunctions  such as  “and”,  “but”,  etc.  constraints (discourse)  the limitations put on the meaning  —  of the passage by each of the words within it. if the passage is describing “dogs”, cannot be “it”.  —  the pronoun reference  (see textual constraints)  constraints (textual)  context  For example,  the same as discourse constraints  —  the sum of the inter-relationships involved in a  reading passage.  cues (grapho—phonemic)  -  clues from the written form of the  sound of word.  cues  (re—iterative)  —  clues that are the same word as or  which contain the root of the deleted word.  cues (semantic)  cues  —  (syntactic)  clues giving lexical meaning  —  clues  giving  syntactical  (functional)  meaning  deletion (fixed-ratio)  -  a pattern of deletion in the doze  procedure where every nth word is removed.  15  deletion (rational)  a selective deletion of words based on  —  a particular teaching need, rather than a random (every nth word) deletion.  discourse  —  the  chunk of passage  larger than a  sentence,  i.e. the whole passage and all the inter-relationships.  Erra  an  —  abbreviation  for  inter-sentential blanks, words are deleted.  filler  -  a  combination  of  intra-  and  the blanks left when re-iterative  (see Intra— and Inter—)  in a doze passage a word which is required to fill  in a blank which has been left when a word has been deleted; ideally a filler would be re-iterative, as has  been deleted;  the  i.e.  the same word  filler may or may not be  found  somewhere in the text.  gap filling exercise  —  any exercise which requires missing  information to be found and inserted.  guessing  strategy  —  a  systematic  necessary solution to a problem, blanks  by  carefully  and  way  -  to  determine  the  such as filling in doze  logically  considering  the  information available to the reader and deciding upon the most suitable word to use as a filler.  16  Inter—  —  a short form for inter-sentential blanks, the space  left when words are deleted from a doze passage.  (see Erra  and Intra-)  Intra—  —  a short form for intra—sentential blanks, the space  left when words are deleted from a doze passage (see Erra and Inter).  intermediate level  -  the level of ESL proficiency in whiöh  the students are familiar with all the English tenses in the active  voice,  and  are  learning  the  passive  voice,  conditionals, modal auxiliaries and adjective clauses.  inter—sentential clues:  words that are located beyond the  sentence of the doze blank.  intra-sentential clues:  words that are  located within the  sentence of the doze blank  parts  of  speech  (color-coded)  condensed to five categories, color a) red:  blue:  verbs  adverbs e)  C)  nouns and green:  black:  conjunctions, pronouns, “not”  —  the  parts of speech are  each designated with its own  (subject and object)  adlectives  and  articles  pronouns b) d)  purple:  remaining aroups including the following  prepositions,  interj ections,  relative  17  parts of speech ideas, test  things,  (traditional) e.g.  a)  —  “student”,  nouns:  people,  “Vancouver”,  places,  “philosophy”.  b) pronouns: representatives of nouns, e.g. “I”, “me”;  relative pronouns—-those pronouns which who”,”  which”  C)  verbs:  actions,  e.g.  of  conjunctions:  joining  prepositions: “in”,  “on”  verbs,  words  h)  e.g.  words,  which  e.g.  adverbs:  “often”  “and”,  location  e.g.”  “think d)  “famous” e)  “quickly”,  relate  interiections:  ideas,  “write”,  adjectives: descriptors of nouns, e.g. descriptors  ‘oin  f)  “thus” and  time,  words that express  g) e.g.  feelings,  e.g. “oh” i) articles: e.g. “a”, “the”  referential  -  the state of a word having a relationship with  another.  re—iterative  -  the  state  of  being  repeated;  re—iterative  words are words that are found more than once in a passage.  scoring method  (acceptable)  a  —  method  of  scoring  doze  passages using any word which fits the sense of the passage regardless of whether it is a synonym or even grammatically correct.  scoring  method  (sensible)  —  a  method  of  scoring  doze  passages using synonyms as alternatives of the exact—word.  18  semantic  the state of expressing a meaning rather than a  —  function.  syntactic  -  the state of the interaction of the parts of  speech.  syntax  test  —  the interaction of the parts of speech.  (discrete)  a  —  test  which  measures  the  students’  understanding at the sentence level, rather than relying on continuous textual understanding.  test  (global)  a  —  test  which  understanding of the whole passage.  test (integrative)  vocabulary  —  all  -  measures  the  students’  (see integrative test)  the same as the global test.  the words  in a passage,  be they having  lexical or functional meaning.  words (function)  —  the words which have no specific meaning  in themselves and serve to enhance lexical words by showing relationships, “the”, etc.  e.g.  “and”,  “but”,  “because”,  “if”,  “in”,  19  words  (lexical)  themselves  and  —  give  the  words  which  understanding when  carry they  e.g. “house”, “man”, “speak”, “big”, “quickly”  meaning stand  in  alone,  20  CHAPTER TWO REVIEW OF THE LITERATURE  INTRODUCTION TO THE REVIEW OF THE LITERATURE  The review of the literature in this chapter will be  presented with the current research design design  evolved  from  having  read  having had practical experience  in mind.  literature  in a)  and b) the use of color in education.  on  The  and  from  the doze procedure  Those points relevant  to the design will be discussed in terms of what previous researchers  have  said  about  them  from their  experiments.  The purpose of this discussion is to show the genesis of and need  for  the  present  research  question  and  to  show  the  reasoning behind and the organization of the research design in this paper.  Thus it can be said that each topic examined  in this chapter has its source in the literature and is an integral part in the development of the research design.  It  should be noted that certain topics are either part of the pilot project,  the main project,  future research based on  the present research, or a combination of these. In the next part of this chapter, to give the general setting for the research, the scope of the experiment will be given and then the standard (traditional) doze procedure will  be  described  researchers. deletions,  The  as  it  various  has  been  aspects  of  used this  length of blanks and passages,  by  topic  several include  and parts  of  a  21  passage left intact. looked at,  Then,  areas of investigation will be  including deletion ratios and patterns,  scoring,  ways of improving students’ scores and the education trends toward the use of discourse, context, and information gaps. In the subsequent part of this chapter the key areas of investigation  in  this  research  paper  will  be  presented.  They are:  a) the importance of discourse to the teaching of English as a second language,  b) the definition of the terms “discrete” and “global” and opposing arguments as to which of them reflects the nature of the doze procedure,  c) the linguistic nature of non—native speakers of English, and  d) the implications of these key areas of research.  In the last part of this chapter a bridge will be made from the review of the literature to the research design in which the goals and use of the previous research will be discussed.  It should be noted that an effort to link the  rationale of the design to the key points in the review will also be made throughout the second chapter.  This has been  done to firmly establish the bonds between the past research  22  and the present development of a new and somewhat complex research design.  The  original  design used  in  the  pilot  project and the more conservative revised design used in the main project were developed to discover if students could be taught with the color-coded parts of speech to look beyond sentence boundaries  for clues to help them achieve higher  scores on doze passages.  THE GENERAL SETTING This  explore  a  section  number  of  of  the  research paper will  questions  about  endeavor to  doze procedure,  the  recent trends in education, and the linguistic attitudes of non-native speakers of English.  The purpose is to show how  each is related to the use of color—coding in the present ESL experiment with the doze procedure in order to justify the effort of carrying out the research.  The questions are  as follows:  a) What is the scope of the research?  b) What is traditional doze procedure?  C)  What are some of the controversial areas of the doze procedure as they relate to this research paper?  23  d) Can color—coding be a way of improving students’ motivation for and their proficiency in doing the doze procedure?  e) What are recent education trends that have lead to the use of color—coded doze procedure?  f) How does color-coding as a trend fit into the present research?  g) How do discourse and the nature of the doze procedure overlap?  h) What is the history of the arguments in the global/discrete controversy as it relates to the doze procedure?  i) How do non-native speakers of English feel about doing the doze procedure?  j) What are the contributions of the key areas of investigation to the present research?  24  What is the scope of the research?  It  should  following  be  noted  literature  here  review  that  the  the  on  purpose  doze  of  the  procedure  and  color—coding is to set the stage for the research model of this paper. which  has  first  Although there been  undertaken  described  the  relevant  to  review.  Likewise,  the  a vast  since  doze  present  is  1953  array of when  approach,  project  have  research  Wilson  Taylor  those  areas  only been  selected  for  of the limited amount of work that has  been done on color—coding in education, only those relevant areas have been included in this paper. A selection of the areas or subtopics of doze research well summarized in Rankin’s overview (1974) will be  considered  in this paper.  of 1958 to 1974  Of Ran]cin’s  subtopics  which include “readability, reading comprehension, learning, information,  redundancy,  thinking,  aptitude,  readiness,  listening, flexibility, and context clues” (1974, p. 2) this paper will be looking at reading comprehension,  learning,  redundancy.  following  way.  thinking,  and context  clues  in  the  This research deals with adult non—native speakers of  English.  These  college  and  university  ESL  students  participated in a study using the standard doze procedure and color-coded parts of speech.  The urose of the study  25  was  to determine  if color-coded Darts  of  speech  in  doze  passages would be able to encourage and assist students in lookina beyond the immediate sentence of a doze blank to find clues as to the filler that matched the word originally deleted from the passage. The amount of research already done and which continues to be done on the doze procedure is evidence of the great contribution Taylor has made to education with his easy to make gap-filling exercise.  At the same time the interest in  so much research by so many researchers is indicative of the ongoing  controversies which  surround the  doze procedure.  This paper is a reaction to the debate which seemed to be of primary  importance  in  between  the  group  controversy  011cr was  the  whether  late and  or  the  not  “global” or “discrete”  in nature.  Chihara,  (1977)  011cr  experiment that decide  on  what  et  al  to  put  and  Shanahan  the  doze  early  1980’s  group.  The  procedure  was  To put it another way,  appeared  students did use fillers  1970’s  to  show  in  their  inter-sentential  cues to  into  blanks.  the  doze  Shanahan et al (1982, 1983, and 1984) claimed to demonstrate such was not the case.  Their reasons and arguments will be  given later. Before proceeding to this key controversy, however,  it  should be noted that the doze procedure has been shown to be effective  for teaching  reading to  group in this present research, 1971; Oiler and Conrad,  the type  of  i.e. ESL students.  1971; Oiler,  target (Craker,  1972a; Anderson,  1973;  26  Hinofotis, 1977; Streiff, 1978; and Soudek and Soudek, 1983) and with adults Rankin,  1974).  (Peterson, Also  Paradis,  before  and Peters,  proceeding,  a  1973;  look  at  and the  definition of the traditional doze procedure is in order.  What is traditional doze procedure? As  to  the  test Ashby-Davis  appearance (1984)  of  a  typical  traditional  summed it up as being a)  doze  a passage  of about 300 words with b) the first sentence left intact to give students a with  d)  every  sense of the content and c)  fifth  or  seventh  word  50  deleted  deletions to  ensure  randomness in the deletion of syntactical and lexical items. She said that answers that are e)  either originally deleted  words or acceptable alternatives are given credit. et al  (1973)  and Valmont  (1983)  Peterson  noted that the traditional  doze procedure uses a standard blank length of 15 spaces. Giving credence to standardized blank length,  Valmont went  on to elaborate on the use of standard versus non—standard length doze blanks.  He concluded that a standardized doze  blank is useful for teaching because it did not allow the length of the blank to give a clue that would detract from  the students’  need to make an effort to use the available  syntactic and semantic clues. It was this traditional doze which was the starting point of the present research,  i.e. as the basic format to  be used with the control and treatment groups.  In the pilot  project for the treatment groups the traditional doze was  27  modified  only  by  color-coding  the  words  and/or  blanks  according to the appropriate part of speech represented.  In  the major project the color—coding was used in the same way, except the traditional doze was modified.  The following  history of the doze procedure gives among other things the reasons for that modification.  What are some of the controversial areas of the doze procedure as they relate to this research paper?  Researchers deleting words It was  have  and of  found by  investigated scoring the  Foley  (1983)  various  students’  using  methods  of  replacements.  random patterns  with  deletion rates of every fifth to twelfth word that there was no  difference  several  in  researchers  difficulty  of  (MacGinitie,  replacement. 1961;  However,  Ramanauskis,  1972;  O’Reilly and Streeter, 1977) preferred the deletion of every fifth  word  because  it  sampled  meaning  and  grammar  more  thoroughly and oblectively while keeping enough information for the students to be able to comDlete the doze blanks. Such evidence supporting the every fifth word deletion pattern added justification to the use of this traditional doze pattern in the pilot project in the present research design.  This  evidence  built  on  the  benefits  of  the  The traditional doze design of approximately  300  traditional doze procedure which are as follows: a)  words had proved to be of a convenient length as it allows  28  for the passage to be typed double—spaced on no more than  two pages. b) pattern  Furthermore,  this  allowed  50  for  length  blanks  and  the  which  was  every  5th  word  convenient  for  calculating scores. c)  The traditional doze design also allowed for the  first sentence and the last sentence to be left intact to give  the  reader  a  sense  of  the  general  meaning  of  the  passage.  With regards to  every nth word deletion as a  random  pattern (as opposed to a rational pattern in which there is control in the selection of deleted items), disagreement by Taylor and Streeter  (1977).  (1956),  filling  the  adverbs.  doze  Martin  and O’Reilly  it relates to locating information blanks  deleting only lexical items, and  (1974),  In order to improve the validity of  the doze procedure as for  Rarikin  there was some  they  proposed  i.e. nouns, verbs,  (1973),  rather  rationally adjectives,  than use  the  random  “every nth” word deletion pattern, decided to delete using a rational doze form in which he chose to delete any one of the parts  of  speech  from among the  following  categories:  “(1) nouns and pronouns (2) verbs (3) adjectives and adverbs (4) prepositions (5) articles and conjunctions” Like  Martin,  “discretionary  Cohen  judgement  (1980)  modified  argued doze)  (p. 103). that  in  using  determining  which words to delete rather than simple random deletion”  29  128)  (p.  had merit.  For the present research the random  (fixed-ratio) blank design was chosen for the pilot project but was replaced by a rational doze in the main project. The change was made to take advantage of the large number of re—iterative—word Ritsumeikan  clues  teachers  in for  the  text  their  chosen  university  by  the  students.  However, whatever the deletion pattern, the contribution of the  above  researchers’  focus  credibility given to the  on  parts  speech  of  is  the  focus on parts of speech in the  present research design. Also having a bearing on this research paper was the matter of scoring. exact—word  and  There was a controversy over the use of  acceptable—word  fillers  to  replace  the  deleted words.  In her study with bilingual Hopi students,  Streiff (1978)  considered both the strong research evidence  in  favor  (Oller, yet  of  the  l972a)  reliable,  exact—word  method  for  native  speakers  arid the speculation of Oiler of the fairer, sensible  scoring method,  using  synonyms  alternatives of the exact—word, for non—native speakers. “fairer” which  Oiler  conveyed  meant the  that by same  accepting  meaning  as  alternate  the  as By  answers  original  word,  students had a chance to get higher doze scores than they might if only exact—word answers were accepted. Consequently Streiff used both sensible and exact-word methods  of  scoring  but  found  like,  Anderson  (1973),  and  Stubbs and Tucker (1974), that language proficiency could be  30  confidently  indicated  by  the  more  exact  and  less  time  consuming exact—word scoring method. On the other side of the controversy, Cohen (1980) Hinofotis  (1977a)  found  respectively  that  and  acceptable-word  scoring was more reliable and seemed to discriminate among student proficiency levels more than exact—word scoring. From  the  above  evidence  it  would  seem  that  the  controversy between the exact—word and the acceptable—word scoring methods is not over.  This, then, is one reason why  the present research design followed the example of Streiff and proposed to use both Another  reason  method  was  due  observations that  some  for  of  the to  (grammatical  and  inclusion the  Stump  students’  scoring methods of  the  separate,  and Hinofotis. answers  semantic  violated  influences)  violated long range constraints.  for  comparison.  acceptable—word  but  overlapping,  Stump  (1980)  local and  Hinofotis  found  constraints  other  answers  (l978b)  noted  that some answers were acceptable within the context of the sentence, but violated long range constraints. The above observations of Stump and Hinofotis pointed to the need to consider long range constraints in the choice of scoring methods and to thus choose the acceptable—word scoring method.  Indeed,  it was the goal of this research  paper to see whether or not the students were looking beyond the  sentence  this,  of the  doze blank.  In  fact,  to determine  it appeared a modification needed to be made to the  acceptable—word  scoring  method  to  take  advantage  of  the  31  observations  of  Stump  researcher decided  and  that  Hinofotis. answer  an  Consequently  should  be  scored  this more  precisely than ‘ust receiving a point for being acceptable and breaking  no  range level.  It was felt that if students were supposed to  constraints  either  at  look beyond the sentence for clues than  looking  the  at  local  the  local  or  long  (a more difficult task  level),  then  they  should  be  rewarded more for not breaking long range constraints than for not breaking local constraints. The observations of Stump and Hinofotis and my goal to have  students  look beyond  the  sentence  level  lead me  to  design a scoring procedure which could possibly make finer distinctions  than  either  the  exact-word  acceptable—word scoring methods. designed  to  innovative  include  technique  both which  meaning  of  the  passage,  or  the  typical  The scoring procedure  exact-word gave  according to whether it was a)  or  scoring  weight  exact, b)  to  each  an  answer  acceptable to the  acceptable  C)  and  was  only  to  the  particular sentence in which it was located as each of these related  to  the  availability  and  types  of  clues.  This  scoring procedure will be described in more detail in Tables 14 and 15 located in Chapter Three. With regards to the spelling of the fillers, et al long  (1973) as  the  Peterson  decided to allow for mistakes in spelling as intended  word  was  appropriate for the doze blank.  evident  and  otherwise  Following the above ideas  about spelling, the present research design adopted the idea  32  of  accepting misspellings  as  spellings made the word fall  correct.  However,  into a different grammatical  category, the answers were to be scored as 0. project  when  passage,  the  filler  misspellings  was  were  when the  available to  not  For the main  somewhere  accepted.  be  in In  the such  cases misspellings would indicate that the student had not been totally aware of the same—word re—iterative clues.  What are some considerations about students’ motivation and proficiency as these relate to setting up the doze tests?  Researchers have pointed out that the traditional doze is  “an  extremely  (Cranney, 1977).  1972;  difficult  Rankin,  and  1974;  and  anxiety-invoking O’Reilly  and  test”  Streeter,  Thus this section of the literature review will deal  with various researchers.  solutions to these problems given by various It will also show  how the CCC is a logical  extension of the research and explore the merits of the CCC as a possible solution. Solutions being considered in this section include a) explaining the nature of doze tests to students, b) a  guessing  strategy  for  students  to  follow,  C)  giving  supplying  clues within the text itself, and, d) providing feedback. The first solution, explaining the nature of the doze procedure to relieve the anxiety of the students, arose out of  the  efforts  of  the  researchers who  determined  that  a  doze score of approximately 43% would compare to a reading  33  comprehension multiple-choice score of (on a standardized test)  3,,  1973;  Streiff,  (Bormuth,  1977;  (approximately)  1967,  75%  1968; Peterson,  and Ashby—Davis,  1984a,).  This  information allowed the teachers in the present research to inform the students that they did not have to worry about seemingly  low  scores  on  the  doze  tests  and  should  be  satisfied with scores around 40%. The second solution came from the work of Ashby-Davis (1984b)  who,  after determining what good  they take doze tests, strategy.  readers  do when  outlined and shared their guessing  This strategy can be used to help other students  to improve their doze scores.  In fact,  it was used in the  present research for both the treatment and control groups to  make  the  task  easier  and  worthwhile  for  all  groups,  regardless of the inherent advantage or disadvantage of the color-coded or strategy,  other treatments.  as given  in Table  A copy of the guessing  1 below,  was attached to the  inside of each student’s test passage folder to be used as needed during each testing session.  34 TABLE 1 OVERALL READING COMPREHENSION STRATEGIES  Successful doze text follow these strategies.  takers  often  a) They read the first and last sentences of uninutilated text to determine the gist (main idea) of the passage.  b) Then they skim the mutilated text trying to get clues to the gist.  c) They read from the beginning to the end of the text, trying to find meanings for the omitted words by checking context clues before or after the omitted words or from general knowledge related to the text.  d) Finally, they reread the entire text when they have guessed all or most of the words, filling in words previously not guessed or correcting words already guessed in terms of the total text. (p. 587) Ashby—Davis (1985, p. 587)  The  third  solution  revolved  around various  kinds  clues being given somewhere on the doze passage paper.  of Two  types were evident in the literature: a) partial fillers and b) matching words. Partial fillers and equivalent meaning clues, including “grapho—phonic, syntactic and semantic information (Goodman,  35  1970;  Propst  &  Baldauf,  “one short underline. (Valmont,  1983)  . .  were  1981)  plus graphic clues such as  for each letter of the deleted words” suggested  as  ways  to  determining replacements for missing words. been mentioned,  the present  research  assist  in  As has already  design  followed  the  standard doze procedure which uses standard—length blanks, so Valmont’s short underlining could not be used. One grapho-phonic clue,  i.e.  the initial  letter of a  word, as mentioned by Henk (1981), Soudek and Soudek (1983), and  Valmont  (1983)  could  have  been  used  in  the  present  research to understand the effect of initial letter clues in improving doze scores as compared to the use of color—coded parts of speech.  In fact, Valmont found that using initial  letters of the deleted words as clues was effective but it appeared to me that this could cause students to increase their focus on the immediate sentence of the blank rather than encouraging them to look beyond the sentence level.  In  light of the supposed localizing effect of initial letters and other graphic aids on the search for clues to fill in the  doze  students determine  blanks, using  the  pilot project  color—coded  whether  the  blanks.  color-coded  tested The  one  purpose  blanks  are  group  of  was  to  used  by  students to find answers using local clues at the expense of the use of more distant clues. of the control (see  the  compared,  bar  When the mean scores of each  and treatment groups graph  in  Figure  1  in  in the pilot project the  Appendices)  were  it appeared that dust giving color-coded blanks  36  only as  clues had the  least positive effect and possibly  even a negative effect on doze scores.  Keeping in mind  this result of the pilot project and because there were not enough students in the main project to allow for four types of groups, the color—coded blank treatment was dropped from the research design. The  idea  of  matching  word  clues  came  from  several  researchers who provided the intended fillers somewhere in the passage.  Some fillers were included as follows:  a) at various places in the margins (Ashby-Davis, 1985), b) in a list of five words on the right side of the page for matching one (called “list”) c)  below “maze”)  the  to  one with the  appropriate  blank  (Propst & Baldauf, 1981),  doze  blanks  three  choices  (called  the  (Pikulski and Pikulski, 1977) and  d) below the doze blanks four choices (called “multiple choice doze”) Neville  and  (Aitken, 1978).  Pugh  (1971)  explained  the  benefit  matching doze procedure to the student by stating that:  matching doze parallels and measures the actual reading process in that it often forces observable regression...The simple act of rereading, erasing, writing new answers, and again rereading provides considerable evidence that the reader is editing. (p. 105)  of  37  Another benefit was pointed (1977).  said  They  procedure  that  “excessive  original  doze  considerably  out by O’Reilly and  in  using  difficulty  testing  reduced”  situation (p.  the  and  multiple-choice  ambiguity  appears  to  of  the  have  been  the  same  (Logically,  48).  Streeter  benefit could be given by the maze procedure.) Upon reflection, it seemed to me that each of the above attempts to include clues somewhere on the page of the text had the same problem.  Looking for words in the margins or  under the lines of the text is a rather unnatural process for a reader to undertake.  A more natural process,  while  maintaining the benefit of a matching type doze, might be one in which the clues could be found in the text and along the line of the sentences instead of below. color—coding appeared  to  the be  iterative words  parts  a  of  speech  reasonable  and  solution  The idea of  the in  (acting as matching words)  doze  as  much  blanks as  re  would stand out  more and possibly be identified as the necessary fillers. It  should  difficult  for  be any  admitted  here  unmodified  that  doze  it  would  passage  matching fillers for all of the deleted words. coded doze passages are rio exception,  to  be  very  contain  The color  so the color—coded  procedure cannot be classified as a true matching exercise, nor  claim  all  the benefits  of matching—doze  procedures.  But the color-coded doze procedure might be able to claim the benefits in proportion to the number of matching fillers (same—word  or  same-root  re—iterative  clues)  that  are  38  available in the passage.  (Whether or not this was true in  the present research will be discussed in Chapter Four.) In  the  research,  passages  selected  and as many as  located somewhere in the text. the  the  pilot  for the 50 deletions per passage,  matching fillers  for  for  main  project  there  fillers and as many as 40. research,  i.e.  trying  to  as  project  few as  26  32 matching fillers were In those passages selected  were  as  few  as  30  matching  For the purpose of the present determine  if  students  could  be  trained using color—coded parts of speech to look beyond the sentence  of  fillers,  it was not necessary to have matches for all the  fillers.  It was hoped that at least the students would be  able to  the  doze  blanks  for  clues  find the available matching  to  the  fillers.  required  This would  have suggested that the color—code parts of speech had had some influence on the purpose of this research.  Tables 21,  23, 25, and 27 in Chapter Four show how many matching filler blanks were filled and how many were filled correctly. It must be admitted here, notwithstanding the proposed advantages  of  the  CCC,  that  the  rationalized  color-coded  doze procedure was somewhat cumbersome to construct but in a different way from the other matching doze procedures. In their multiple-choice  doze  (matching)  design Streeter  and O’Reilly (1977) increased the amount of effort needed to prepare  the  doze  tests.  The  increase  was  a  result  of  taking care to use distractors of the same part of speech to ensure students would have to  look at meaning  instead of  39  ‘ust syntactic clues to find the blank filler.  Although the  rationalized color-coded doze skirted this extra high level cognitive effort,  it required a great deal of lower level  cognitive energy to locate the re—iterative—word clues. Another drawback  in  using  the  CCC was  that  the  CCC  required the maker to analyze the parts of speech and then code  them  with  correctly. to  defined  colors  consistently  and  Sometimes it was not as easy as one might think  assign  sentence,  the  the  appropriate  “He worked  at  a  color.  For  example,  in  the  French Canadian Hospital”,  the  word “French” can be colored “green” to indicate that it is an adjective.  But in the sentence “He worked at a Native  Hospital”  word  the  “Native”  is  not  as  easily  coded  “native” is also used in the sentence “He is a native.”  if  dilemma  is whether  “Native”  should be  show it is an adjective or “blue” Fortunately,  however,  colored  The  “green”  to  to show it is a noun.”  in the passages used in the present  research there were not very many difficulties like this. On the surface, there was the additional difficulty of physically color-coding the doze passages.  Each word had  to be given a computer code to color the word appropriately. For the CCC project this took a lot of time, especially for the  pilot  tremendously project  passages. in  But  reducing  passages.  The  the  a  revised coding  techniques  technique  time  will  be  for  helped  the  described  Chapter Three and in more detail in Appendices B and F. the  future,  with  more  advanced  computer  main  hardware  in In and  40  software, the coding process should be even much less time consuming. A  fourth  solution  to  help  students  improve  their  performance on the doze procedure is to provide feedback. In fact, Rankin  (1974)  stressed the lack of improvement in  reading comprehension when doze materials were used without feedback.  Ideally each doze test should be reviewed with  the students but time and security concerns because of the use  of  several  classes  doing  the  same  set  of  tests  at  different times made this unmanageable in the present kind of research. In the CCC project two kinds of feedback were given. The first kind of doze  procedure,  feedback was  i.e.  suggested by the matching  when there  iS  a  set  of blanks  and  another set of words to fill the blanks such that each word can only be used once and each blank only has one correct filler.  It was felt that CCC,  although it can not claim  this advantage, might be able to gain from the colored-partof—speech clues,  where the  colored blank must be matched  with a filler of the same color representing the same part of  speech.  kind  of  The  second,  feedback  was  more traditional  given  when  all  the  though partial, students  were  allowed to look at the originals of the practice passages. In the pilot project the students took about a five minute look  after  the  practice  test  in  the  subsequent  testing  session while in the main project the students looked at the  41  answer key for about five minutes right after each practice test. In Chapter Five the above CCC solutions to the problem of student doze test anxiety  will be discussed in terms of  the  of  outcome  of  the  results  the  pilot  and  main  CCC  projects.  What are the recent education trends that have lead to the use of color—coded doze procedure?  So far in this chapter on the review of the literature the  doze  procedure  has  been  defined  and  various  modifications have been presented, some of which were chosen for the present research.  Those choices were made in light  of recent trends in education including a) passage analysis, I,)  information gap, c)  educated guessing, d) vocabulary and  grammar in context, and e) redundancy. One of the most important trends at this time and one which makes analysis  a great deal  rather  than  just  of sense sentence  is the use of passage analysis.  Long  (1981)  urged students to go beyond “knowing the lexical value of individual words, individual  or of retrieving factual information from  sentences”  (p.  72).  Richards  (1980)  stressed  that “language in any authentic communications event cannot be  understood  entirely  isolation” (p. 64).  by  considering  sentences  in  42  Mohan  (1986)  has  been  concerned  with  discourse  and  emphasizes that an “important part in competence in reading is the ability to draw inferences from written text.” subsequently  defines  inference  as  “a  relation  He  between  sentences” (p. 129). Referring to that  Chihara  Bachinan,  doze and discourse,  and  (1982)  Oiler,  (1977);  Foley  (1983)  Anderson,  noted  (1980);  and  believed that doze tests are sensitive to  discourse  constraints.  procedure  could  even  Cohen encourage  (1980)  thought  students  to  that  doze  consider  the  passage as a whole rather than viewing it as being composed of isolated sentences. A second important trend in education, related somewhat to passage analysis and certainly tocloze procedure, is the idea  of  reducing  or  filling an  information  gap.  Morley  (1972) contributed to ESL with her many kinds of gap filling exercises which have motivated students to look for missing (needed)  information and thus caused them to use English to  communicate.  Mohan  (1986)  re-iterated Morley’s  idea  and  expressed the need for students to increase their “ability to  communicate  successfully  across  wider  and  wider  information distances” (p. 110). In this research paper the goal was to determine if. in fact, students were able with the doze to be trained how to  find information  (clues)  for fi1lin in blanks,  distances than just at the sentence level. acceptable  responses  to  blanks  which  at greater  Correct and/or  require  the  use  of  43  beyond-sentence clues would indicate that the students have made use of discourse clues and constraints. A third  trend  in  education  revolves  guessing,  as opposed to rote learning.  one  of  kind  expectancy”  guessing in  which  in  language  the  student  around  educated  Foley referred to  learning as  as  decoder  “crrammar  relies  (or  should) rely on the highly redundant nature of a message to make accurate restorations.  These “accurate restorations”  are a result of students processing information from within the passage. discourse  The relation of guessing to comprehension of  can  definition  also  of  be  seen  literal  in  Katz  and  comprehension  Fodor’s  (as  (1965)  mentioned  by  O’Reilly and Streeter) which refers to the understanding of grammatical  and  semantic relations both within and across  sentence boundaries. In going Fodor’s  a  little  definition,  apparent.  a  further fourth  in understanding Katz  trend  in  education  and  becomes  It can be referred to as learning vocabulary and  grammar in context. This fits in well with the idea that the discourse level for  teaching  Nilsen,  1976;  is more important than the sentence level  vocabulary and Judd,  (Chastain, 1978).  1976;  Rivers,  1968;  Also Judd commented that  learning words in context helps retention and a wider and more precise understanding of them. On the subject of grammar, Wilkins (1975) (1977)  took  communicating  opposing and  positions.  accomplishing  goals  and Janssens  Janssens to  preferred  learning  about  44  grammatical believing  rules. in  On  the  communicating  other  ideas,  hand,  Wilkins,  stressed  the  also  role  of  grammatical rules in the learning process as they are the foundation which allows language functions to successfully communicate the intended meaning of speakers or writers. From  looking  carefully  at  these  opposing  arguments  about grammar and communication, it is evident that there is possible common ground,  i.e.  students must know how to use  grammar actively as a tool in context rather than just to know about it passively. Still on the subject of grammar and its influence on the  doze  procedure,  Cohen  connection between grammar check  for  including  awareness inflectional  ungrammatical local  of  (1980) and  explained  doze.  grammatical  clues  to  In  an  important  being  relationships,  cohesion  in  a  trait  common  to  to and  discourse,  doze answers can provide evidence  reading,  able  non—native  for too— readers.  Evident in this explanation for the purpose of the present paper is that students do  (or if they do not,  should)  grammatical clues to be able to fill doze blanks,  use  grammar  being considered as a tool rather than as a nuisance in the non—natives’ attempts to become proficient in English. Aitken  (1977)  continued the argument for the value of  grammar as it refers to using redundancy to fill in doze blanks.  He argued that without knowing grammar the student  could not take advantage of the redundancy in the language and thus would have more difficulty with doze tests.  45  It is the students’ awareness and understanding of this redundancy as it is expressed through the vocabulary and the grammatical  structure  that  color—coding  the  speech was aimed to improve.  of  parts  The present research.  of  thus,  attemDted to use color to make students more aware of the vocabulary and able  to  use  structures  them  to  in the practice passaaes to be  auess  the words  to  fill  the  doze  blanks. If  Aitken  redundancy  is  proficiency,  right  an  and,  heighten this color—coded  is  indicator if  the  awareness,  doze  that of  students’ their  color—coded then  procedure  it  can  overall  clues  could be lead  awareness  to  of  language  are  shown  said  that  greater  to the  language  proficiency. Although the literature about the doze procedure that I reviewed pointed to the advantage of students being aware of redundancy,  I also found a negative reaction by Tuinman  (1972) to an increased awareness of redundancy in a passage. He discovered that when students tried to sift through a greater number accuracy blank.  in  of  non—target  finding  the  items,  correct  they  response  lost for  speed the  and  doze  This lead me to question the effect of color-coding  all the words, if all the coding does is increase the number of words to be scrutinized.  Color-coding parts of speech  does not increase the context in terms of number of words, but hopefully in the amount of awareness of context clues available.  The  result may be  the problem Tuinman noted  46  above,  i.e.  that students will be slowed down by the extra  information made aware to them.  On the other hand,  it can  be argued that color—coding parts of speech should lead to the  students  focussing  on  clues  given  by  the  particular  parts of speech rather than on all the parts of speech and therefore bring some speed back into the searching process by reducing the number of words to be looked through. The  immediate  color—coding all  question thus would be whether or not  the words  and blanks  in a doze passage  causes an information overload which in turn decreases the students’  scores?  This question will be discussed further  in Chapter Five after the research data has been collected.  Summary of language trends  Whether the recent trends in ESL deal with discourse, information grammar,  gaps,  they  all  guessing, can  redundancy,  lead  toward  vocabulary,  language  or  proficiency.  Aitken (1977) with the help of Foley (1983) ties the recent trends  together.  In Table  2  below the  fourth quote was  found in Foley’s article and the other quotes were given in Aitken’s article.  47  TABLE 2 QUOTES FROM *AITN (1977) AND **LEY (1983) ABOUT RECENT EDUCATIONAL TRENDS.  *Language proficiency more is than mastery of a specific nuniber of discrete structure points and lexical items. (p. 62)  *Reading comprehension in ESL depends on a knowledge of the meanings of words, phrases and sentences, and of arrangements of words, phrases, and sentences according to the conventions of written English. 62); Thomas (p. (1970, p. 164)  *Three layers of language meanings that must be dealt with in reading comprehension (1) Meaning carried by the lexical items. (2) Meanings carried by the grammatical structures. (3) Socio—cultural meanings.” (p. 62); (Fries, 1963)  **Because the message is normally highly redundant, the decoder should be able to make accurate restorations of many of the blanks that appear in the passage. (p. 58)  *The correspondence between the ESL student’s conception of English redundancy becomes an index of his overall language proficiency in English. (p. 65)  It would seem possible that the doze procedure, gaps  need  to  be  filled  with  information  given  in which by  clues  48  embedded  in  one  or more vocabulary  items  and  grammatical  structures located somewhere in a passage either within or beyond the sentence of the blank, could also yield an index of the non-native adults’ proficiency in English.  How does color-coding as a trend fit into the present research?  There  is  no  doubt  that  color  increasing part in our lives. and movies,  is  are out of style.  everywhere in the entertainment field, It  color—coding organizing  is  has  in been  office  discriminate  the  business  exploited  files  amongst  and  product  ever  an  Black and white television  for the most part,  education.  playing  in business,  field,  for  such  helping  sizes.  and in  however,  most,  the  Color is  Perhaps  where as  for  customers because  of  financial restrictions education has only been able to begin exploiting the use of color. films in color,  There are now instructional  especially in the video format,  and books  are illustrated in sophisticated four—color color printing. Some books are organized into chapters or topics by color and some even have key words in a passage highlighted in one color.  But  there  the  color—coding  stops.  Multi—colored  printing is expensive. Attempts language  have  to  use  been  color made  technical restrictions.  or  but  color-coding subject  For example,  to  in  teaching  financial  by at least 1926,  and in  49  the tenth  edition of her book called  English Reading”,  for  voiceless  of  of  She used “red” for short vowels, consonants,  “black”  consonants, and orange for semi—vowels. eight pages  Teaching  Nellie Dale used colors to differentiate  the graphemes of English. “blue”  “On the  colored plates  for  voiced  However, other than  (grapheme  charts)  and three  pages of colored examples, Dale’s book is in the usual black print.  In another  “Reading  with  language text,  Phonics”  the  i.e.  authors,  the Hay  1954 and  edition of Wingo,  took  advantage of improved color reproduction techniques to have several multi—colored pictures, every page.  However,  and colored words on almost  they only used one color,  supplement the basic black print. only  to  highlight  particular lesson.  the  “red”,  to  Actually they used “red”  graphemes  being  studied  in  the  Unlike Dale they did not color-code the  graphemes. The most well known user of color and color-coding for teaching  languages,  Galeb Gattegno.  especially  non—native  languages,  In his paper called “Words in Color”  was  (1964)  he explained how he used color to divide the graphemes of English other  into  words,  their his  appropriate  purpose  was  phonetic  to  help  categories.  students  spelling of words to the way they are pronounced. had started, using  not  vowel  sound  categories.  In  match  the  What Dale  Gattegno made even more detailed and colorful,  four  colors  including black,  categories Like  Dale,  and  30  colors  Gattegno  used  but for  21  colors  consonant  colored  for  sound  charts  to  50  display the  colored  categories.  print a text in color. rationale.  Rather,  However,  He,  too,  was  unable  to  cost was not given as his  from an educational point of view,  believed that students should write (and thus read)  he  in one  color. Of  the  above  authors,  the  most  influential  on  the  present research paper was Gattegno (1964).  Although he was  primarily  level,  working  at  the  grapho-phonemic  color—code parts of speech. 1964  He referred to this  he  did  in his  article called “Words In Color” where he stated that  there is a card for each word, the word’s part of speech.  the card being the color of  He further noted that if a word  had the potential of being used as more than one part of speech, then the card would be of the appropriate number of colors.  However, he did not mention which color meant what  part of speech. Gattegno’s extremely  research  important  for  and  use  the  present  of  color—coding research  is  paper.  Concerning color blindness, even when it existed, he saw in his teaching that there was no interference.  He stated that  students could still differentiate the shades of the colors. Even more significant is what he said about his method being analytic and synthetic at the same time, while other methods were either one or the other.  He pointed out that color—  coding could help students differentiate the graphemes from each other and put them into their appropriate pronunciation group simultaneously.  In “Words in Color” Gattegno provided  51  some testimonials and Dodds readers,  (1966)  from teachers  readers  Cunliffe  had  good  results  illiterate  adults.  On  the  in her unpublished critique of  pointed out the concern of McHugh  that the positive effects of Gattegno’s color-coding  a  lot  competence with  and  (1986)  Gattegno’s colored words, (1968)  obtained  stated that this was true with beginning  remedial  other hand,  who  the  to of  do  with  the teacher.  color—coded  interact  more  Therefore,  the  with  procedure  passage  than  of the  noted here  the  student  with  it would appear that McHugh’s  intervening effect  motivation  It should be  doze the  dedication,  the  and that  has  to  teacher.  criticism of  teacher should not detract  the from  the value of the CCC. In the “silent way”  1960’s  Gattegno’s  method was  color-coding as part of his  popular  in  schools  in  the  United  States and was still being used in the late 1970’s to teach other  languages  to  Peace  (Canadian) volunteers.  Corps  (American)  and  CUSO  Being one of the latter volunteers,  the author of the present research started using color—coded parts of speech in teaching English to inmigrants in Canada and to nationals in Saudi Arabia. In Upper Volta in 1975 Montety (1977) way  and  colored—coded  lines  with  used the silent  colored  chalk  on  the  blackboard to elicit past tense sentences from his students as part of their review.  For use without color when it was  not available he devised a set of various shaped lines to do the equivalent.  Although he coded words,  it was not until  52  later that other teachers started to use color—coded parts of speech.  However,  at this time, there is only one other  teacher, known to the author, who is using color—coded parts of speech to teach English as a Second Language,  and the  distribution of colors is a little different from the system used by this researcher. defined  her  This  categories  of  is partly because  parts  of  speech  she has slightly  differently and partly because she has labelled some of her categories  with  different  colors  from  those  used  by  the  present researcher. From my  review  of  literature  and my  experience with  color-coding in teaching English it can be seen that using color—coding  is  not  a  recent  trend.  potential but has been defeated,  at least  cost of the necessary technology, technology.  of color-coding. reasoning  has  shown  in part by the  and even by the. lack of  the time may be right to revive the use It should be pointed out that, because of to  Gattegno’s  black-on-white when they write, phased out.  it  However, with ever newer and cheaper technology  in color printing,  similar  Yet  that  students  should  use  the color-coding should be  This means that color-coding might be used at  the. beginning of a course but by the end there should be no color—coding  at  incorporated  this  all. idea  The and  present  thus  the  research  post-test  design  was  left  uncolored for the treatment groups as well as the control groups. coding  Meanwhile it was felt that the advantage in color that  Gattegno  found  in  both  differentiating  and  53  categorizing  for  graphemes  also  could  analytic and synthetic techniques color—coded doze procedure. distinguish  the  Gattegno’s 51,  major  apply  to  these  in finding clues  in the  By using only five colors  grammatical  categories)  instead  (to of  it was thought the load on the memory could  be reduced and decoding made less confusing. To  summarize  trends and thus,  how  color—coding  fits  into  the  present  into the doze procedure research,  it was  hypothesized that by being given color-coded parts of speech students speech  would  from  be  each  helped other,  to and  differentiate then  to  the  mentally  parts group  of the  similar parts of speech in order to look through the passage for  the words  gaps.  in  Sometimes  the  same  (from  group  52%  to  to 64%  fill of  the  the  information time  in  the  passages analyzed for the pilot research and 60% to 80% of the time  in those  for the main CCC research)  the  fillers  found in the text were exactly the same or had the same root as the deleted word. passage speech.  were  Sometimes the key clue words in the  synonyms  However,  as  or it is  antonyms  in  the  same  part  of  obvious that all the key clue  words cannot always be in the same part of speech group as the deleted word, so the strategy cannot be foolproof. this  strategy  at  best  can  only  be  classified  as  guessing but an educated rather than a wild one. because  the  color-coded  guessing  strategy  Thus  one  of  This is  allows  the  students an organized method of searching through the text. With the time and energy saved by looking through the same  54  part of speech, the student can look systematically through other parts of speech for roots of words which could have the  required  correct that  meaning  part  of  were  and  speech.  highlighted  relationships  were  then  adjust  Given  grammatical  by  expected  the to  that  word  relationships  were  the  relationships  color-coding,  assist  the  hoped  to  such  students  finding words to fill the doze blanks also. grammatical  to  in  Such use of  be  reflected  in  higher doze scores for the CCC groups than for the control groups. To put the above briefly, the proposed advantage of the color—coded students  doze  would  be  procedure able  to  was  that  make  a  by  more  using  it  CCC  organized  and  consequently a more thorough if not faster search for clues to fill in the missing words.  If the proposal proved true,  it would seem reasonable that it should be true for inter— sentential  searching,  searching.  With the benefits of the guideposts provided by  the  color-coded parts  not  of  just  speech,  for  intra—sentential  it was  anticipated  clue  that  students would find it less intimidating and also easier to look beyond the sentence of the blank for clues.  Therefore  they would be able to fill in more blanks correctly,  thus  improving  that  their  doze  scores.  It  was  also  hoped  students would be able to internalize the parts of speech so that the color-coding could be removed and still allow the students to use the same parts of speech guessing strategy across sentence boundaries.  55  WHAT IS THE KEY AREA OF INVESTIGATION UNDERLYING THIS  RESEARCH PAPER? To answer this question there are four subtopics to be looked at.  They are:  a) the link of discourse to the nature of doze procedure, b) the nature of doze procedure——global or discrete, c) the chronology of the global/discrete controversy d)  the motivational nature of non—native speakers of English, and  e) the role of color in assisting non—native students to locate more global clues with the aim of improving the students’ doze procedure scores.  How do discourse and the nature of doze procedure overlap? Shanahan  and  Kamil  (1982,  1983,  and  1984)  in  their  research on doze procedure tried to show that students use intra—sentential  rather  finding  for  fillers  considered  was  nature.  Their  than  doze  whether  contemporary opinion  blanks.  doze  starting  inter—sentential  is  point  The  global was  clues  question or  discrete  opposite  to  in they in the  which was dominated by Halliday and  Hasan (1976), and Oiler et al (1978) who were convinced from their own research that doze was global;  that is,  test  across  the  students’  boundaries.  use  of  context  clues  it did sentence  Their research will be discussed briefly later  in Chapter Three.  56  Given the recent trends, especially about being able to comprehend text at the discourse level, Shanahan and Kamil’s research had significant influence on this paper. it  provided  question,  a  starting  i.e.  can  point  students  for be  the  In fact,  present  trained  in  research  the  doze  procedure such that they look beyond the immediate sentence of  the  doze  blank?  blank  to  find  clues  to  help  them  fill  the  An answer to this research question will add to the  controversial  answers  supplied  by  the  review  of  the  literature which follows immediately below.  Is doze essentially global or discrete? On the global side of the arguauent according to Foley (1983, p. use  59) were Halliday and Hasan who saw a need for the  of both  (1975)  clues within  and beyond the  and Chavez and Oller  (1977)  sentence.  Oller  believed students could  be trained with the doze procedure to use global reading comprehension skills.  Cohen  (1980)  even defined doze  as  being able to test the ability to read and write cohesive English. Foley (1983) doze by  gave credence to the global nature of the  explaining the mechanism of this global  proficiency in the following way:  language  57  doze In procedure all sorts of deletions, whether they be content words or connecting devices, carry with them constraints which may range backwards and forwards across several sentences. This places a strain on the short—term memory which presses the student’s grammar expectancy into operation, the accuracy with which the student is able to supply the correct or acceptable response can therefore be taken as an index of the efficiency of the student’s developing ‘grammatical’ system. (p. 59)  Ashby-Davis  (1984)  also  gave  the  argument  credence  by  explaining that doze:  ...draws once at on the overall grammatical, semantic, and rhetorical knowledge of the language. . . . and students have to understand key ideas and perceive interrelationships within a stretch of continuous discourse, and they have to produce, rather than simply recognize, an appropriate word for each blank. The focus of the task involved is more communicative than formal in nature, and it is therefore considered to reflect a person’s ability to function in the language. (p. 99)  Streiff  (1978)  proficiency Oller  and  Hanzeli that  of  doze.  Conrad  (1977)  doze  listed The  (1971),  supporters list  and  included  Oller  who apparently had  could  lead  to  students  (1973  shown  the  global  Darnell  (1970),  of  and  1975)  and  in their research  having  global  English  proficiency. Stump by  (1978)  reporting  the  tried to prove the global nature of doze ability  of  the  doze  procedure  to  “so  58  accurately  predict  scores  on  both  the  Lorge—Thorndike  Intelligence Tests and the Iowa Tasks of Basic Skills that)  all are essentially measuring the same thing  language proficiency” kind  of  proof  validity  (of  (p.  when  she  doze)  (in  global  -  57). Ashby-Davis offered the same reported  ...as  an  “substantial  integrative  concurrent  test  of  overall  proficiency in English as a second language (Oiler & Conrad, 1971;  Oiler,  Tucker,  1974;  1972;  Irvine,  Hinofotis,  Atai  Oiler,  &  1980)....(on  1974;  tests)  Stubbs  such  as  UCLA English as a Second Language Placement Examination”  &  the (p.  99) Although through the years a great deal of support was given  to  the  idea  of  doze  being  global  in  nature,  even  proponents such as Streiff admitted some contrary evidence. Streiff admitted the observation of Klare et al  (1976)  that  doze answers are most likely to use constraints that occur within a range of only four or five words before or after the  blank  Anderson’s level  in  reservation  skill,  discrete,  question.  (1980)  suggesting  that is,  Likewise,  it  Ashby-Davis  about the was  not  included  doze being a  low  integrative,  but  working only at the sentence level with  regards to finding clues to fill in the doze blanks. Foley  (1983)  came to  a compromise between the  and discrete position seeing the doze as  an  global  intermediate  skill “able to assess the student’s knowledge of the syntax, lexis and rhetorical devices...(rather than)  the extent the  59  student has understood the conceptual world of the writer” 61).  (p.  If  many  researchers  favored  were convinced that the the notion was Kaiuil  (1982,  looked back  at  and  Jongsma  (1980).  the  Leys  1984)  among  global  effective  others.  In  Shanahan,  plus  the  procedure  inter—sentential  and  for teaching, Shanahan 1986,  Kamil  and  Reutzel  and  literature  He noted the researchers’ doze  argument  attacked by  research of (1983),  traditional  measure  doze was  be vigorously  1983,  (1982),  the  to  the  Tobin  review  of  conclusion that  could  not  comprehension,  effectively  and  Jongsma’s  conclusion that doze procedure was rather ineffective  for  teaching reading comprehension. The debate involving the global versus discrete nature of  doze has  been  fervent.  Although the maor points  of  view have been identified and described in this paper, to do justice to the main researchers, will basis  now be for  given.  the  a chronology of the battle  The chronology will  research  questions  in  the  also provide present  paper and for the design of the methodology.  the  research  60  What is the chronology of the global/discrete controversy?  1971: Halliday and Hasan Halliday  and  Hasan  (1971)  wrote  that  “the  doze  procedure transcends the confines of a single sentence”  (p.  51).  1972: Carroll; Chihara & Oiler (1977)  Chihara, original  key  Oiler  et  players  al  in  (1977) the  pointed  to  one  global/discrete  of  question  the as  being Carroll in 1972 who wrote that the “doze procedure is one of the techniques that has been proposed as a possible basis  for  1972).  investigating  discourse  constraints”  They went on to present Carroll’s  linguistic  clues  are  usually  blank and that grammatical  in  the  same  1972  (Carroll, claim that  sentence  as  the  clues play a greater role than  semantic clues in doze scores. Chihara  and  Oiler  (1977)  using  “two  passages  of  prose...selected from texts written for non—native speakers of English” The  later  (p.  94)  were  tested native and non—native speakers.  either  at  advanced proficiency level.  the  “the  doze  procedure  intermediate,  or  They were given scrambled and  regular forms of the passages. that  basic,  Chihara and Oiler concluded is  sensitive  to  discourse  61  constraints ranging across sentences” higher  the  proficiency  level  of  the  and that the  68)  (p.  students,  the  more  likely they are to take advantage of the constraints. Chihara and Oiler put global  nature  different  of  doze.  the  proficiency  forward a By  levels,  they  strong case testing  were  able  for the  students to  use  at the  proficiency factor to explain that any lack of influence by discourse constraints is a result of the students, doze procedure itself.  not the  They found that the more proficient  the ESL students were (as determined by their grade levels) the higher were their doze scores.  This observation will  be pursued later as it has some bearing on the nature of the present research paper. If Chihara and Oller were correct, then the color—coded doze  procedure  should be  able  proficiency as indicated by a)  to  increase the  students’  improved doze scores and by  b) more correct fillers, fillers of the kind which depend on clues from other sentences beyond that of the doze blank. The  increased proficiency should come as  a  result of the  assistance given by the color—coded parts of speech.  The  color—coding should be able to make the students more aware of  the  parts  of  speech  inter-sentential levels. provide fillers. as  clues  by  a  students that they  at  the  intra-sentential  and  The parts of speech in turn should  constraints  and  For example,  indicated  both  as  to  the  required  blank  a blank requiring a noun to fill it  “blue”  line,  would  indicate  to  ought to look through the passage  the for  62  “blue” words,  i.e. nouns and pronouns.  Because in at least  half the instances in the present research’s passages, matching  filler  (the  same-word  located somewhere  in the text,  students  good  stand  a  chance  as  the  deleted  word)  the is  usually inter-sententially, of  taking  advantage  of  the  color-coded parts of speech and blanks to easily find the matching fillers. doze  scores.  Such discoveries would improve students’  The  ability  to  find  matching  first through the use of color-coded parts  of  fillers, speech,  finally through the use of non—colored parts of speech,  at and is  what was being explicitly tested in this research project. An analysis of the doze test scores  (calculated using both  the exact—word and the acceptable—word scoring methods) also might show whether or not students could be trained to use color—coded and non—colored parts of speech to find fillers which were graphically different to any of the words in the passage.  The  clues  necessary  to  locate  these  kind  fillers are shown in Table 3 at the end of Chapter 2.  of The  table outlines the various types of cohesion in passages.  1979: Yamada Chihara challenged by  and  Oiler’s  Shanahan and  Jun Yamada in 1979.  findings, Kamil,  however,  among  Yamada built on:  were  others,  to  be  including  63  a)  the work of Carver  (1975  —  1976),  who “stated that  sentence order would probably have little effect on doze scores (p. 70); b)  Chihara,  Oiler  et  ai,  who  compared  scrambled  to  unscrambled doze; and Halliday and Hasan who had developed a  C)  “theory of  inter—sentential cohesion” in 1976.  Yamada tested the hypothesis that the location of cohesive clues,  either within or beyond the sentence of the blank,  must cause differences in doze scores. Yamada’s Oiler’s  that  findings  agreed with  scrambling  the  those  sentence  of  order  Chihara’s does  make  and the  doze significantly more difficult but he continued that:  doze scores are highly correlated with discrete-point grammatical test scores (due to the nature of the cohesion in the discourse). (p. 76).  Yamada suggested that most of the time the sequential doze procedure  tests  sentence)  ability”  inter-sentential  “the (p.  subjects’ 76)  as  intra-sentential opposed  to  (between-sentence) ability”  “the  (within subjects’  (p. 76) because  the number of intra—sentential cohesive items is far greater than the number of inter—sentential cohesive items.  Thus to  all intents and purposes the students were encouraged by the nature of the cohesion  in the passages themselves to look  64  for cohesive clues primarily in the sentence of the blank for both the scrambled and unscrambled doze tests just the same as they would in a discrete-point grammar test. Yamada thus concluded that: a doze test is primarily a discrete grammar test rather than an integrative reading comprehension test. (p. 76) To restate the point in a slightly different way, it should be noted here also that Yamada, unlike Chihara et al,  found  that the primary factor in the debate was the nature of the cohesion,  i.e.  passage.  The primary factor, he thus believed, was not the  on the availability of cohesive ties in the  proficiency of the students.  This point, like the opposite  made by Chihara and Oller,  had a bearing on the design of  the present  one  research.  On  hand,  Chihara  and  Oller’s  ideas led me to examine the nature of students as they have related to the doze procedure, i.e in terms of proficiency level.  On  instrumental  the in  other  the  hand,  analysis  and  Yamada’s  comments  subsequent  were  selection  of  doze passages in terms of availability of matching fillers to put in the doze blanks.  1980: Thomas Susan Thomas’ research, doze  results  students.  But  on  the  like Yamada’s, put the onus of  doze  instead of  itself  looking  rather  than  on  the  at cohesion within and  between sentences, she looked at content and function words  65  and their predictability in lesser and greater amounts of context.  She restated that function words are easier than  content words to predict, but she found that:  greater context facilitates doze inferences for content words but inhibits doze inferences for function words. (p. 52)  She carefully explained this by saying that:  global predictions involve(d) general and more long range estimations of what is coming next in a reading passage....In contrast, focal predictions arise out of specific events, deal with more local anticipations, and are dispensed with more quickly... (and) a larger context actually inhibits the correct prediction of function words...(as) the reader may not be particularly attentive to (nearby surrounding) clues because of other redundant clues in the remaining context...Therefore, the availability of a large amount of contextual information may simply provide more alternative clues than are necessary...This kind of ‘information overload’ is forced upon the reader and inhibits. ..exact function word predictions. (p. 52 53) -  Given the  above  ideas,  Thomas was  able to explain a  possible reason why students tend to use such local clues (constraints)  in order to guess what words should be put in  the  The  blanks.  cause  can  likely be  discovered  in  easy  materials which by their very nature contain relatively more  66  function words  than more  difficult materials do.  Thomas  supposed that:  the greater the proportion of function words in a given message, the smaller the influence of greater context for the test as a whole should be. (p. 53)  Thomas hypothesis debate.  presented to test  for  further  her belief  study  in the  a  global  subsequent side  of  the  She hypothesized that:  the use of materials with more difficult readability levels and/or the restriction of deletions to content words will expand the boundaries of contextual constraints operating in doze tests. (p. 53)  Thomas’ the  (1980)  proponents  description of the tug of war between  of  content  and  function  words  plus  her  subsequent hypothesis called for a solution for which the color-coded doze  (CCC)  procedure seemed to fit.  seemed appropriate because  it distinguishes  parts of speech from each other (blue red  =  verbs,  conjunctions, and  green  adjectives, purple  prepositions,  highlights  purple)  =  the  =  content  (most of)  the  nouns and pronouns, =  interjections, words  The CCC  (blue,  as opposed to the function words  adverbs, and “wh” red,  black  words)  green,  (black).  =  and  At the  same time the color—coded doze procedure still apparently  67  allowed  for  the  benefits  of  the  difficulty  of  random  every  nth  word  deletion pattern. In  terms  of  passages,  the  present  research pilot passages were identified at an intermediate level  at  two  structures,  ESL  schools  and  the  type  of  vocabulary,  and length of sentences follow the curriculums  of those levels at those schools.  An IBM computer program  called  Version  “Readability  Calculations,  1.00”  used  for  rating readability, calculated the readability levels of the selected passages  in  the pilot  study  being  as  as  low as  grade six and as high as grade eight on the Flesch scale and between grade six and nine on the Fry graph.  A trial at one  school in a mid—intermediate class with the easiest passage using the traditional doze format showed that the passages were  not  too  distinguish students.  easy  the  nor  too  different  difficult  language  and  were  proficiencies  able of  to the  The above data about the passages used in this  research plus conversations with teachers having used doze tests with students pointed to the need to find ESL students who were as proficient as possible. students possible.  should be Due  to  studying the  at  number  availability of such numbers,  This meant that the  as high of  a  students  grade  level  needed  and  as the  the upper-intermediate level  was agreed upon for the pilot doze passages.  Although it  would have been preferred if the students were at a higher proficiency level  for the selected doze passages,  Thomas’  hypothesis calling for more difficult readability levels to  68  expand the boundaries of contextual constraints,  presumably  gave support to the selection of the particular proficiency levels of the students in the pilot project.  Her hypothesis  also  of  seemed  to  give  credence  to  the  use  even  more  difficult passages with the university level students in the Ritsumeikan  Program.  Due  to  the  requirements  of  the  university program the selection of the doze passages was limited to two  of  the  texts used  in the  regular program.  Passages were thus chosen from the easier of the university level  texts  in  which  the  sentences  complex with many technical terms.  still  tended  to  be  The seeming advantage of  these passages was that many key semantic terms and several function words were repeated once or more. iterative nature  of the words  for inter-sentential clues, intra— 62%  and  for  students  inter—sentential  inter-sentential finding  intra-  This highly re  in the passages,  in the main project re—iterative  clues) and  especially (60%  clues,  and  -  44%  appeared to bode well  inter-sentential  80% —  for  re-iterative  words to fill the doze blanks.  1982: Shanahan and Kamil In  1982  Shanahan  sequential/scrambled findings  that  the  and  Kamil  passages doze  test  and is  used  intact/doze!  “confirm(ed) insensitive  earlier to  the  integration of information across sentence boundaries”  (p.  69  207).  However, they made some concession to their opponents  in the global/discrete debate but qualified this as follows:  It is not that students fail to integrate information across sentence boundaries when completing a doze exercise, but only that the information does not aid in the sentence completion activity. (p. 207)  In  1983  Shanahan  and  Kamil  explained  their  justification  further in that:  sub)ects perform no better on typical random nth word deletion doze tests than they do on the same tests in which the sentences have been placed in random order. (p. 123)  1982: Bridge & Winograd Bridge and Winograd in 1982, using verbal (think-aloud) protocols  from  ninth  graders  (both  good  and  poor),  found  that students did use inter-sentential clues but the amount of use depended on the ability of the type  of  cohesive  ties.  They  found  readers  that  and on the  conjunctive  and  referential items were more likely to be determined by the use  of  inter—sentential  clues.  Lexical  items  determined by the use of intra—sentential clues.  were  Based on  the result that good readers used within-sentence clues 54% of the time while poor readers 38% of the time,  Bridge and  70  Winograd  believed  discerned  from  that  deleted  lexical  within—sentence  items  clues.  could  be  Furthermore,  believing that most ties in language are lexical ones,  and  thus also occur in the traditional every fifth word deletion doze  procedure,  tendency  is  Bridge  toward  and  using  Winograd  concluded  intra—sentential  that  clues.  the  Their  subsequent solution was to use doze passages “in which the researcher  selectively  deleted  items  which  integration of information across sentences” The  significance  of  the  require  the  (p. 310).  contribution  of  Bridge  and  Winograd’s research to the present research paper is that it explained how the nature of the passage controls the way in which  students  blanks. the  try  to  find  the  to  It also explained in what direction,  intra—sentential  somewhat  towards  clues  the  ——  cohesive  1 ties  the  lexical  inter—sentential  referential and causative items. of  solutions  Bridge  and  the i.e.  doze toward  items,  and  ——  the  clues  By setting out the types Winograd  prompted  in  the  present research paper the analysis of a number of passages to find the distribution of ties and subsequent selection of the passages with the blanks which had their replacements somewhere in the text.  (The details of this and other text  analyses will be presented in Chapter Three.)  The present  analysis chose to find same—word and same—root re—iterative , 2 ties  i.e.  lexical and syntactic clues,  as they seemed to  be the ones most easily found, once highlighted, in order to be  used  to  fill  in  the  doze  blanks.  Other  types  of  71  lexical, referential, and conjunctive ties were not analyzed selection purposes,  for passage they  were  evident  color—coded Bridge  than  and  but  anticipated  Winograd  and  in the main treatment  found  should be pointed out again here that same—word  and  same—root  to  re-iterative  be  them  even  more  be.  It  to  in the analysis clues  in  the  of  pilot  passages and the main passages, the number of deleted items appeared at least once elsewhere in the text as high as 64% and as low as 52% of the time and 80% and 60%, respectively. (This  means  clues,  most  find  that  if  students  using  the  color—coded  of them beyond the sentence of the blank,  appropriate  the  were  re-iterative  items,  the  to  color-coded  scores were expected to show as much as a similar proportion of  improvement  traditional)  over  the  doze test.  control  (non—color—coded  Any other improvement would have  to have come from the influence of color—coding on the other types of clues. Because of Bridge and Winograds suggestion about using a  rational  (non—random)  deletion  doze,  for  each  of  the  chosen passages an attempt was made to increase the number of blanks with re-iterative clues by deleting a different set  of  words.  However,  for  the  pilot  passages,  as  the  increase was not very great, and due to the large number of treatments available, also be  already  and  the  limited  number  of  it was left for the main experiment.  noted  that  it  is  beyond  the  students It should  constraints  of  this  research to have included an analysis of all the types of  72  cohesion mentioned by both Halliday & Hasan,  and Bridge  &  Winograd.  1983: Leys The  following  year  Shanahan  &  1982  students  instead  Kamil’s of  (1983) study  adults.  Leys using  The  Leys  et  al  junior  replicated high  study  asked  school three  questions:  1. Does prior knowledge of the passage topic aid doze performance? 2. Does the sequential as opposed to the random ordering of passage sentences aid doze performance? 3. Does sequential context as opposed to random or no context aid performance on a target doze sentence? (p. 112)  Ley’s answer to each question was, “No”, and therefore, she seemed to confirm... Shanahan and Kamil’s (1982) findings...(that) the traditional doze format (i.e., every fifth word deleted and exact replacement criteria) does not appear to be sensitive to the integration of information across sentence boundaries. (p. 113)  To explain the results, and  Kingston  (1983)  that  Ley drew on the idea of Weaver  the  traditional  doze  tends to measure the language skill of readers their understanding of the passage).  procedure  (rather than  73  The  first  contribution  of  Shanahan  et  al  and  Leys’  research to the present research paper was the pressure to take  another,  sentential)  but  closer  versus  the  look,  global  at  the  discrete  (inter-sentential)  (intra—  question.  The second was the pressure to look again at the scrambled doze method.  Thus,  the  research  present  treatments method  could  with  a subsequent study is needed, based on  be  paper,  given,  color—coding  in  one  and  which  using  the  two  a  scrambled  other  sentence method without color—coding.  secondary  using  doze  the  same  The scores on both of  these methods could be compared with scores on the colored— and uncolored—unscrambled  forms  Shanahan  did  al  et  and  Leys  of  not  the  passages.  find  higher  Whereas  scores  for  unscrambled passages over scrambled ones, and whereas it was thus  interpreted  sentence  clues  that  there  if  unscrambled treatments,  in  that  students  both  are  varieties,  higher  treatment  were it  scores  when  only can  for  compared  using be  the  to  within-  hypothesized CCC  the  for  other  the  three  then this would show that there is indeed some  influence of the  color-coding on the students to  look for  inter—sentential clues to fill in the doze blanks.  1983: Shanahan and Kamil In look  at  1983  their  Bridge  and  global  nature  Shanahan  earlier  Winograd and  the  and  study (1982) high  Kamil  in and level  light  took  a  of  the  findings  (1982)  about  Rankin of  retrospective  comprehension  of  of the the  74  doze procedure, agree with  respectively.  their  opponents  Shanahan and Kamil had to  that  students  did  use  inter-  sentential information in doze tests, but would not accept the criticism about their own lack of proper attention to inter—sentential 310). Kamil  cohesive  clues  (Bridge  and  Winograd,  p.  Through argument and research, however, Shanahan and tried  to  information  show  was  that  only  the  students’  minimal  use  (infrequent)  of  and  global trivial  (processed at a low level). First, Winograd’s  Shanahan and Kamil pointed out Bridge and  admission  that  students  used  inter-sentential  information minimally to effect doze scores.  Second, they  applied Rankin’s ideas of literal recall to their previous scrambled doze experiment to test what was happening inter sententially.  Their experiment was meant to address Bridge  and Winograd’s and  refute  call to make use of  Rankin’s  idea  that  order comprehension processes,  “doze  tests  measure  such as inferencing,  exclusion of lower level processes, (Shanahan et al. 1983, p. 123).  inter—sentential clues high  to the  such as literal recall”  Shanahan and Kamil focussed  on the level of importance and quantity of ideas recalled in both a) unscrambled intact passages versus unscrambled doze passages and b)  scrambled doze passages versus unscrambled  doze passages. Their new study showed that students were able to better recall unscrambled  ideas,  doze  more  ‘ust  of them,  as  with  and the  in order with thern unscrambled  intact  75  passage. do  This suggested to Shanahan and Kamil that students  “integrate information across  doze  exercises”  (p.  127).  sentences when completing  However,  because  scrambled  and  unscrambled doze scores were similar rather than different, they concluded that inferencing can take place unrelated to or despite the doze procedure.  Thus it can be said that  Shanahan and Kamil had shown the inverse to what Rankin had said about inferencing and literal recall as these high and low  order  comprehension  procedure.  Shanahan  sentential  inferencing  processes  and  Kamil  pertain concluded  took place unrelated  to  the  doze  that  inter—  or  despite  to  the doze procedure and consequently that doze was a lower order process.  1984: Shanahan and Kaiuil The following year Shanahan and Kamil continued to try to  challenge  Bormuth’s  their  data  of  critics by doing 1962.  They  an  archival  re-analyzed  the  study  on  data  to  determine “the number of comprehension test items requiring within-  and  across-sentence  information”  (p.  254).  concluded that:  (the doze procedure) predicts performance best in those situations in which within—sentence comprehension dominates. doze is less useful in situations in which students are expected to integrate information across sentence boundaries. (p. 255)  They  76  Although Shanahan and Kamil showed the lower order nature  the  of  doze  procedure  in  their  1983  and  1984  research, they did not really answer the criticism of Bridge and Winograd. and  other  cohesive  challenge. study  to  doze  Shanahan should have looked at the lexical  Thus look  blanks,  items  to  respond  it was the goal  at  cohesive  items  especially  those  properly  to  the  of the present research clues  as  that  to  filling the  occurred  inter—  sententially. For reasons of ease with color—coding, the focus was on same—word re—iterative clues.  The researcher included re  iteration of both lexical and syntactic words.  It should be  noted that this was different from Bridge and Winograd who suggested  the  opposing  referential items and b) higher  intra—sentential  influences  of  a)  conjunctive  and  lexical items, the latter favoring scores.  divided the lexical items into i)  The  present  the collocative,  research such as  the “United States” where one word usually is located beside another one, It is  and ii)  the same—word re—iterative—word items.  obvious that collocative words  are  intra—sentential  and it was found by careful analysis of the passages used in the present research that most of the same—word re—iterative items were inter—sentential.  Furthermore,  it must be added  here that it was beyond the scope of the present research to consider  comprehension directly.  Therefore,  Shanahan  and  Kamil’s conclusion about the lower order of the doze could not be debated but left for future research.  However, even  77  having such a debate rested on whether or not students could get beyond the  first  step  of  learning to  look beyond the  sentence of the doze blank for clues to find the That  step  itself  depended  on  the  filler.  inter-sentential  clues  being available.  1985: Henk Henk bias  joined  toward  the  the  inter/intra-sentential  within-sentence  side  of  debate  the  with  debate.  a He  agreed with Shanahan et al that:  readers use local redundancy when answering doze items and tend to neglect inter—sentence cues and prior knowledge as sources of information. If, indeed, doze tests fail to tap the multifaceted nature of text processing, their usefulness as measures of reading comprehension ability becomes seriously limited. (p. 213)  Like with  Bridge the  and Winograd,  standard  Henk used  (traditional)  think-aloud protocols  doze  and  found  subjects used five types of passage cues:  1) within-sentence cues only, 2) beyond sentence but within text, 3) combinations within and beyond sentence in text, 4) prior knowledge/beyond text, 5) combinations within and beyond text. (Henk p. 215)  that  78  However,  found  he  intra-sentential sentential clues, that  doze  that  cues.  for  the  most  part  He  did  find  the  students use  of  used  inter-  but very infrequently and thus concluded  procedure  was  not  dynamic  “a  measure of reading comprehension”  and  sensitive  (p. 217).  Yet Henk did leave the door open for more research in the  present  searching  debate.  for  clues  Admitting for  some  that  content  inter-sentential  word  deletions  is  necessary, he emphatically suggested:  examining various contextual factors about deletions such as position in sentence, linguistic function and supporting syntactic and semantic context, cohesive importance and relevance to other text elements. (p. 217)  In  other words,  Henk  proposed  the  idea  that  perhaps  the  motivating force for students to cross sentence boundaries was dependent on the nature of the passage, the  passage’s  readers.  This  idea  is  as well as on  important  for  the  present research because it is another reason for analyzing the passages for cohesive clues.  Certainly it would be a  waste of time and energy to color—code a passage hoping to lead the students to cohesive clues that did not exist.  79  1985: Bachman Bachman,  like Henk,  reviewed proponents of both sides  of the inter/intra-sentential debate, but rather than take a stand on one  side of the other,  inconsistencies  in  results  as  he tried to  partly  caused  explain the by  deleting  words in the fixed ratio (every-fifth-word deletion) manner. Thus  Bachinan  proposed  to  compare  this  fixed-ratio  (random) selection of words with a rational selection. the  rational  according  selection  to  Halliday  he  tried  and  to  Hasan’s  For  develop  criteria  framework.  These  included: 1) syntactic, which depended only on clause—level content, 2) cohesive, which depended on intercausal (intra-sentential) and inter—sentential cohesive content, as described by Halliday and Hasan (1976); and 3) strategic, which depended on ‘longrange patterns of parallel patterns (coherence). (p. 63)  However, despite no problems with the rational deletion procedure,  Bachman  found  the  identification  impractical  for  teachers  and  counteracting  the  desirable  ease  test of  process  writers,  construction  was thus  of  the  standard doze. To using  solve  the  deleted  this  following  words  problem two  rationally  Bachman  designed  forms.  For  and  classified  following four context levels:  the  his  first them  research form into  he the  80  1) 2) 3) 4)  For  the  within clause; across clause, within sentence; across sentences, within text; and extra textual...(where type 2 and 4 deletions were maximized). (p. 539)  second  form  he  deleted  every  11th  described them in terms of the above levels. answer  keys  were  made  which  included  word,  then  For both forms  exact  words  and  acceptable alternatives. Baclunan’s results for his rational doze tests showed that: a) test developers, using the above four levels, were appreciably consistent in identifying deletions, b) his criteria were practical for teachers, c) (scores) were comparable in both in reliability and concurrent validity to those on the fixed—ratio test (p. 549),  d) difficulty of closure increase(d) as the amount of context required for closure increased (p. 549), e) the development and use of an answer key make scoring entirely objective and greatly increase scoring efficiency (p. 550), f) the use of a rational deletion procedure allows the test developer much greater flexibility in revising specific items on the basis of both the content specification of the test and the item statistics. (p. 550)  Although  Bachman  did  not,  in  fact,  solve  the  intra/inter-sentential debate, he set the stage for the next attempt to determine whether students could be trained to  81  find clues beyond the  sentence  of the doze blanks.  One  would expect that next step to be a rational doze where the text was deleted to maximize the number of inter—sentential clues. paper  At  first  would  it  it was thought  be  better  to  for the present research  investigate  the  color-coded  doze procedure with the traditional every-nth-word deletion first in order to set a baseline for further research with a rational deletion pattern.  However,  due to the decidedly  difficult task of locating enough suitable ESL students,  it  was decided after the pilot project to move right away to a rational  form of  the color—coded doze where as many re  iterative clues as possible were used for deletions. decision  was  advantage  made  that  an  to  take  increase  advantage of  of  the  re—iterative  This  suspected  clues  could  bring. Bachman’s influence on the present research design was significant and resulted in two contributions. contribution was the idea of content levels. an attempt to identify various  The  first  It resulted in  locations of the  same—word  and same—root re—iterative clues and their distance from the blank. answer  The key  second  contribution was  which  included  It  inspired  alternatives.  exact a  the  inclusion  words  two-style  and answer  of  an  acceptable key,  the  difference being that the present research key was to use weighted scores for acceptable responses. the  weighting  Chapter Three).  formula  is  to  be  given  (An explanation of in  more  detail  in  82  Bachman’s  contributions  were  significant  and  the  present project was over ambitious in trying to incorporate many of them.  The modified weighted scoring scheme which  had been proposed for the pilot project,  was suspected of  being biased, tried and found to be too time-consuming, and dropped during the pilot project in favor of a simplified method  which  was  scoring process. described  in  discovered  unintentionally  during  the  The revised scoring methodology will be  Chapter  Four.  Simplified  along  with  and  partially because of the revised scoring methodology was the location of clues indicating what word should fill each of the doze blanks.  The exact location of each clue to the  doze blank was noted in preparation of the pilot and main projects’ to  any  doze passages but not considered quantitatively  great  degree.  This  kind  of  information  could  be  considered in future projects, but for the present research the essential details were whether the words needed to fill the doze blanks were located a) outside of the passage, b) within the sentence of the doze blank,  or c)  somewhere in  the passage beyond the sentence of the doze blank. As debate  a as  result  of  the  participated  in  ongoing by  discrete—versus—global  Chavez  and  Oiler,  Yamada,  Thomas, Shanahan and Kamil, Leys, Bridge and Winograd, Henk, and Bachman, the stage was set for the present CCC project. The  key  question  which  availability of clues  emerged  dealt  with  for students to be able to  a) find  the in  83  order to fill the doze blanks and b)  how much the students  were paying attention to the inter—sentential clues.  What is the nature of non-native speakers of English? In the review of the literature so far, been primarily on the nature of the doze somewhat  on  indicate  the  the  nature  need  for  of a  the focus has  itself and only  the  students  closer  look  but  at  enough  that  to  nature.  Following, then, is what various researchers have said about students and how they interact with the doze procedure. Some of the problems with the doze have something to do with the students themselves and their orientation toward the doze procedure.  Looking at the research literature the  following problems have been identified: a) non-linguistic interference (Streiff), b)  reluctance to guess (Brutten and Tuimuan),  c)  inattention to redundancy, and shorter memory span (Kalivoda),  d)  lack of knowledge of rules of grammar and redundancy (Aitken),  e) and reading ability of poor students compared to good students (Goodman, Neville & Pugh, Propst, and Schwartz & Stanovich). Streiff, talking about children doing doze tests, said that extra—lingual concerns, task,  lack of attention,  such as uncertainty about the  momentary forgetfulness,  lack of competence in their English prevents  not iust  a child  from  84  filling some doze blanks.  It is conceivable that under the  stress of the doze test,  some of the above problems could  apply to adults.  Talking about non-native speakers, Brutten  (1981) mentioned that some ESL students are reluctant to use context clues to guess because of their limited vocabulary and resultant insecurity.  Tuirunan  (1972)  pointed out that  the students in his study seemed to make only one attempt on 75%  of  the  Kalivoda  doze  blanks,  despite  talked  about  (1980)  redundancy”  (p.  2)  and  that  shorter in a second language” that  the  receiver  amount  of  (student)  language)”  does  not  type  students student’s  a (p.  redundancy  the  3).  has  “memory  span  Aitken (1977)  the  is  argued  effect  “if  the  structure  (of  the  (p. 65).  Looking  at  the  differences  between  good  readers and drawing on the ideas of Goodman as research of Neville and Pugh Baldauf  deletions.  “disregarding  little  know  of  (1981)  (1976  through  argued  that  (1970)  1977),  poor  as well  Propst and  pointed out that one difference between good  and poor readers is that of editing, forth  —  and  the  passage  good  readers  Neville and Pugh  (1976  -  as  they  use  1977)  i.e.  looking back and  read.  context  Goodman clues  (1970)  optimally.  found that poor readers only  used context clues occurring before the blank as opposed to before and after by good readers. (1981)  Schwartz  stressed that poor readers tried to  got entangled in the details of a passage.  and Stanovich find clues but  Yet,  Bridge and  Winograd found that poor readers were able to orally explain  85  cohesive  relationships,  lexical  examples,  ones.  Meanwhile,  readers  were  though  able  relationships.  especially to  Bridge to  a  Likewise,  limited  and  explain  for  extent  Winograd all  Henk  referential  the  (1985)  conjunctive  found types in  and  that of  his  good  cohesive students’  verbal protocols did find the use of inter—sentential clues, but only to a very limited extent. To summarize the literature on the nature of students and doze procedure it appears that there is a contradiction in  the  research  information.  over  the  amount  clues,  Poor readers have the most  especially inter—sentential ones,  of the text. grammar  information  redundant  in that they do not try very hard to look for  in the blanks.  and  of  But looked at from a different angle it could  be argued in the following way. difficulty,  use  of  to help them fill  Or, if they do, they get lost in the details Good readers apparently know more vocabulary  rules to  be  and  are  able  to  able see  to  extract  the  whole  the  necessary  picture  more  clearly.  What is the possible role of color in the improvement of non—natives’ doze reading scores? A  solution to the  poor readers’  problem of  becoming  bogged down in reading a passage seemed to be to help them to find the appropriate amount of use of the context.  This  was Gattegno’s solution to teaching phonics, by grouping the sound—spellings into color—coded groups of similar sounds,  86  helping the students analyze (categorize>  (discriminate)  and synthesize  or vice versa at the same time,  depending on  whether they were reading a sound-grapheme chart or reading a story. the  What Gattegno did at the sound-grapheme level,  color—coded  doze  tries  to  sentence level with the parts color—coding part  of  is  speech  to  the  at  the  of speech.  to help  the  details  (analysis),  extraneous information, clues  do  students  morpheme  and  The purpose  of  focus  thus  on particular  eliminating  some  while helping them to relate other  isolated  detail  (synthesis)  to  find  the  appropriate fillers. To  explain  students’  further,  color—coding  should  counteract  reluctance to get involved in a search for clues  by a) making the parts of speech more readily evident to the students  and by b)  redundancy. more  making  it  easier to  pay  attention  But color—coding should also make the  focussed  information,  on  reduce  relevant  redundant  the memory  load,  to  search  (part—of—speech)  waste  less  time  and  leave more time for more attempts to find the appropriate clues involved in the cohesion of the passage. color-coding  should  make  the  syntactic  In addition, (grammatical)  relationships of individual sentences easier to see and use in the search. advantage ceiling  from  effect  Both poor and good readers should gain an color-coding, for the  although  really good  there  readers.  could  be  a  Furthermore  there is no good reason foreseen why good readers should get  87  confused  by  the  increased  information,  nor  why  the  poor  attractive  than  readers should either. Intrinsically, black and white. this color  research for  color  is  usually more  Color is more motivating, too.  experiment  its own sake.  is  not  It is  primarily  However,  interested  in  interested in color as  a  means to guiding students to an awareness of the parts of speech in the text which in turn may point toward the words to fill the blanks. needs  more  In other words,  than motivation to help  scores on the doze tests.  the present research improve  the  students’  To determine the effect of color  alone as compared to the effect of color plus the teaching point, the researcher provided a control where the passages were  randomly  colored  rather  than  according  systematic form of the treatment passages.  to  the  It was proposed  at the beginning of the present research project that if the colored—treatment  forms  both result  in higher mean  scores  than the standard-doze control form, then it could be said that color is a motivating factor. scores  for  the  color—coding  Furthermore, if the mean  are  higher  than  for  the  randomly-colored forms,  this would suggest that the color-  coding has  effect.  a  positive  This  is  to  say  that  the  students are making use of the intended assistance of the color—coded  parts  of  speech.  Furthermore,  if  the  mean  scores of the randomly—colored forms are lower than those of the standard—doze control group and even lower than those of the  color-coded treatments,  it could be  said that the  88  random colors this  very  effect.  are motivating the  use  of  wrong  students to use them but  information  is  having  a  negative  Such a situation would indicate the positive value  of color—coding the parts of speech.  What are the contributions of the key areas of investigations to the present research?  From Carroll, Oiler,  through Halliday and Hasan,  Shanahan and Kamil, Leys,  Winograd,  Henk,  global/discrete resulted  nor  to debate  was  a  Yaluada, Thomas,  Bachman raged.  satisfactory  Chihara and  the Yet  Bridge and  debate no  over  extreme  solution  the  position  forthcoming  suggestions were made to get closer to the answer.  but  Streiff,  Brutten, Tuinman, Kalivoda, kitken, Goodman, Propst, Neville and Pugh, and Schwartz and Stanovich had all put forth ideas about the nature of students and their reactions to doze procedure.  The  global/discrete  present debate  research  and  had  experiments  its and  roots the  research methodology evolved from these efforts,  in  the  present selecting  and recombining various strategies of the above researchers, leaving some of the other strategies for future experiments. In Chapter Three the bits and pieces of the methodological strategy accompanying the contributions of the researchers above will be combined into a description of the research design.  89  RESULTS OF THE REVIEW OF THE LITERATURE  Introduction  The research design result of This  a  design  in this paper came  review of the suggested  and  largely as  a  literature on doze procedure. tested  a  method  of  helping  students to use cohesive cues that were located beyond the immediate sentence of the doze blank.  Overview of the method proposed in relation to the literature  The goals of the present doze research The original goals were the  following but because of  negative results in the pilot project and because there were so few students to draw on the third goal below was omitted in the main project.  The goals were:  1) to train students using the color—coded doze method to be aware of inter—sentential as well as intra sentential clues toward the point where they no longer need color-coding to help them in the use of part of speech clues, 2) to compare the major treatment group which had practised with several color—coded doze passages to a  90  control group which had practised the same number of uncolored doze passages to determine if color—coding helped students to look beyond the sentence of the blank,  3) to compare the major treatment group to a control group which used passages with only the doze blanks colored, to determine whether the students were using the color—coding of the blanks alone to find clues or whether they were using the beyond—sentence color— coded words as clues,  4) to compare the major treatment group to a control group which had practised in the same way except that the words in the passages were colored randomly.  This  was to determine whether or not students were just being motivated by the colored words to try harder in the test rather than being motivated to take advantage of the instructional information provided by the color—coding.  The use of previous research Many  of  literature  the used  modification to  experiments the  found  standard  in  doze  the and  it which was then tested.  review  of  then  made  the a  For the pilot  part of the present experiment the standard doze was used but replaced by a rationalized doze in the main part.  Both  91  the pilot and the main projects then modified the passages in no other way except by coloring the parts of speech. The pilot project used three modifications and the main project only  two,  below. 1)  the main project dropping the use  of  number  2  The control and the modifications were as follows: In order to determine a base line to test for any improvement in scores as a result of using the allcolor—coded doze, the standard doze form, without any color—coding, was used.  2)  In order to determine (not  together  with  if color—coded blanks alone the  color—coded  sufficient for students to intra-sentential clues,  words)  find inter- as well  was as  only the blanks were color-  coded. 3)  In order to determine if color and not the color— coding was effecting the scores of the students, all the words and blanks were colored randomly to make sure  the  colors  had  no  connection  to  grammatical  meaning. 4)  In  order  sentence  to  help  level,  students  all  the  find  words  standard doze were color—coded.  £4  clues  and  beyond  blanks  of  the the  92  Implications of Review of Literature for the Present Research Paper  In summarizing the  research on doze  that there were two goals paper.  The  first  goal  of  was  it can be  importance to this  to  improve  doze  seen  research  scores;  the  second to see if students could extend their use of context clues beyond the sentence to help determine blank fillers. The more recent studies  seemed to look down on the  first  goal as dust improving lower reading comprehension grammar skills.  It was expressed that, although the doze has some  educational procedures  value, to  help  it  does  not  students  in  have  as  discourse  much  as  other  comprehension  inferencing, i.e. higher level reading comprehension skills. However,  given what Goodman and Tuinman said about readers  using the least amount of effort to determine doze blank fillers, it appeared to me that clues at the sentence level, such as blank length and initial letters of deleted words could short circuit the students’ sentence for clues. color-coded blanks.  need to look beyond the  The same could be true for the use of If  indeed  this  short  circuiting did  take place, then it could be argued that the weakness in the doze  procedure  was  due  to  rather than to the students’ sentence level.  the  short  circuiting  itself  inability to look beyond the  When there is no need to look for inter—  93  sentential clues because intra—sentential clues are readily available, why should anyone waste time searching for inter— sentential ones?  Sometimes, as was shown in an analysis of  the passages for this research project, many clues as to the appropriate doze blank fillers,  are to be found somewhere  in the text beyond the sentence of the doze blank but not within the boundaries  of the same  sentence of the blank.  This appeared to be a good environment to test to determine whether or not students would be able to locate the intersentential clues,  especially the same—word or same—root re  iterative ones. It was the goal of this research project to discover whether or not the procedure of color—coding all the parts of  speech  and  all  the  doze  blanks  in  a  passage  could  provide the necessary stimulus and pathway for students to make the effort to  look beyond the sentence of the  blanks to  appropriate blank  find the  fillers.  I  doze  thought  that such an effort could lead to higher doze scores and decrease  the  strength  of  the  researchers  opposed  to  the  traditional doze procedure. In the present students’  doze  procedure. determine increase  The if  research the goal was to  scores  using  intention  color—coding,  their  use  of  of in  the  the fact,  color—coded standard doze.  color—coded  present  paper  can help  inter—sentential  improve the  clues  doze was  to  students  to  using  the  Note that same—word and same—  root re-iterative type clues were identified in the analysis  94  of the research passages, but other cohesive clues may have assisted the students. The review of the doze literature  suggested strongly  that it is important to know the nature of the doze passage and how much effect the passage has on the doze scores. reading  ESL  by  students  tends  to  be  too  localized  If the  question to be answered is whether increasing the number of inter—sentential scores, distant  (re—iterative)  Obviously, clues,  if  then  there  there  intra—sentential clues. must be  is  clues are  a  can  more  strong  increase  local bias  the  clues  than  toward  using  Therefore, the effect of the latter  neutralized.  In  a  rational  doze,  words  can be  deleted to increase the number of distant clues and decrease the number of local clues.  However, in a standard doze, it  appears to be the proportion of local to distant clues is a matter  of  chance.  To  circumvent  this  problem,  doze  passages can be and, in fact, were analyzed for location and proportion of within sentence clues to beyond sentence clues and  the  most  favorable  passages  were  retained.  pilot project only exact answers were accepted.  For  the  Instead of  using the complex weighted scoring as suggested earlier in this paper,  in addition to the scores out of 50 blanks, re  iterative inter-sentential blank types were tallied.  For  the  and  major  project,  these  two  types  inter—sentential type were used. explained  in  Chapter  Three  when  plus  the  intra—  These procedures will be the  research  described and the scores given in Chapter Four.  design  is  95  Research Question 1 From questions the  overall  arising out of the  question  for  this  literature review,  research  paper  would  be  subdivided as follows.  1) Does color—coded doze lead students to look beyond the immediate sentence for clues to be able to fill in the doze blanks?  In other words, are mean  scores higher on passages which are completely color—coded than on passages which have less color, random color, and no color at all?  2)  Are the students’ coded  passages  mean scores on completely—color-  lower  than  those  mean  scores  of  students who do the standard doze passages?  Hypothesis 1 Out of these questions the following hypothesis for this paper was put forward: lh)  On the practice tests there will be no significant difference  between  the  treatment  group  which  has  trained using the color-code doze and the control groups which have trained on either the non—colored colored-doze,  the  colored-blanks  randomly-colored doze practice tests.  only,  or  the  96  Research Question 2 3)  students  Can  transfer  any  acquired  skill  learned  from their practice with color—coding as it pertains to  inter—sentential  using  clues,  to the  and  intra—sentential  solving of uncolored rational  doze  passages? Hypothesis 2 2h)  On  the  post-test  difference  between  there the  be  will  treatment  no  significant  group  which  has  trained using the color—coded doze and the control groups which have trained on either the non—colored colored—doze,  the  colored—blanks  only,  or  the  randomly-colored doze practice tests. Research Question 3 3)  Does  the  color—coded  confidence  in  doze  give  filling doze blanks  students than  the  more other  forms of the doze? Hypothesis 3 3h)  On  the  post-test  difference  on  there  confidence  will  be  scores  no  significant  between  the  treatment group which has trained using the color— code doze and the control groups. Research Question 4 3)  Does the color—coded doze lead to more productive confidence  in  filling doze blanks  forms of the doze?  than  the  other  97  -  Hypothesis 3 3h)  on  the  difference the  post-test  there  in productive  treatment  group  will  be  confidence  which  has  no  significant  scores  trained  between  using  the  color—code doze and the control groups.  Footnote 1: Same—word re—iterative clues were chosen for analysis and subsequent choice of the passages because those clues were the most obvious. They fit into the color—coded doze procedure because they are the same part of speech as the corresponding blank. According to Bridge and Winograd, re-iterative clues fall into the lexical category and the category of clue which they said lead to intra-sentential searches rather than beyond. However, in the analysis of the passages for the present research, by far most of the time the same—word re—iterative clues were either before or sometimes even after the sentence of the blank rather than being within. All other factors being equal, one would expect the students, given the color—coding, to look beyond the sentence. Footnote 2 It may be that the color-coded doze helps improve conjunctive relationships in as much as the color— coding makes them stand out. For a definition of conjunctive and other types of cohesion, please see Table 3 below. It is an adaptation and extension of ideas presented by Halliday and Hasan (1976) and quoted by Bridge and Winograd (1982).  98  TABLE 3 VARIOUS KINDS OF COHESION o  °  COHESION TYPES  KINDS  o  0  SUBKINDS/ DEFINITIONS  o  o  0  °References  °persona].  °pronouns,  0 demonstrativ e adjectives  : 0  °  EXAMPLES  0 0  °I, you adj°My, your  0 0  :  :tI5, that:  •comparative  °adjectives °adverbs  0  o  o  °bigger °faster  0  0  0  °Lexical  ore_iteration °same words °car/car °synonyius °house/hoiue° e esuperordinates0pear/fite °general words °home/ dwellinge that refer to: : :  o  0  o o  0  °collocation  eUnited °words that together° States  oconjunctive °additive • •adversative :causal  °logical °and relationships °yet :relationships :°  cEllipsis  0  0  °  0  0  °word in paral—°He has a ].el structure° house, which is left° car, and out to preboat. vent unwanted° 0 repetition  °  °  o  0  0  0  0  °  0 °  °  °  0  Adapted from information cited by Bridge and Winograd (1982) and belonging to Halliday and Hasan (1976).  0 0  99  CHAPTER THREE RESEARCH DESIGN AND METHODOLOGY  INTRODUCTION Based doze  on  review  a  procedure  and  the  of  the  use  literature  of  concerning  color—coding,  the  the  present  research project involved training adult upper—intermediate and  university  students. and  level  English  as  Second  a  Language  (ESL)  The research design was a blending of customary  unique  ways  procedure.  of  training  The purpose  of  and  testing  with  the  doze  the design was to discover  if  students could increase their purview of the text and make use of inter—sentential clues available to them in order to give them a more precise understanding of a whole passage (elsewhere referred to as a piece of discourse). The  first  innovation  was  the  combination  traditional  doze  importantly,  the color-coding of the parts of  procedure  and  the  of  coloring,  and  the more  speech in a  piece of discourse of approximately 300 words in the pilot study  and  determine  330  in  whether  the or  main not  project.  The  systematically  purpose  was  highlighting  to the  parts of speech with color could lead to the greater use of grammatical and semantical clues across sentence boundaries to locate the correct fillers for the doze blanks. present research the principal and same-root  focus was  re-iterative clues,  as  on the  they are  In the  same—word  explicit  in  100  the text and thus are the most objectively assessed.  Any  improvement by the students would have been shown by their higher  doze  compared  to  scores  on  the  post—treatment  their pre—treatment  (same)  doze  doze  test  test  as  and  as  compared to the scores of control groups who had practised with a randomly colored doze or a non—colored traditional (n the pilot project) or non-colored rational doze (in the main project). Being used along with the above training methods were two scoring methods.  Both were used after the answers had  been inserted into a spreadsheet.  The first method was the  traditional exact—word method where only the replacement of the original word received credit.  This method was used for  both the pilot and the main projects. blanks were considered and scored.  In this method all 50 The second method was  intended to be a more sensitive measure and give more credit to the students answers where at least partial credit was due.  But the second method was intended to do more than the  traditional acceptable—word scoring method in that it was to indicate also the kind of  clues the students had used to  find the doze blank fillers.  Of primary importance were  the inter—sentential re-iterative—word clues. project this  concern was manifested by the weighted—score  method which looked good at  first glance but was  time consuming and subjective. for  the  For the pilot  pilot  project  the  far too  In scoring the practice test researcher  discovered  alternative way to focus on inter—sentential clues.  an  It was  101  much more efficient to label each answer on the spreadsheet as being either of the external type, type,  or the intra—sentential type,  the inter-sentential  to assign each cell on  the spreadsheet a “1” for a right answer, an acceptable answer, to  the  deleted  containing blanks.  or a “0” if the answer was unrelated  word  external  a half point for  and  then  type blanks  to and  delete  the  columns  intra-sentential  type  Table 16 at the end of Chapter Three illustrates  this scoring procedure. This revised method,  an innovation in doze research,  was used for the main project.  It took advantage of the  ease of deleting columns on the spreadsheet, containing  external  word  fillers,  intra—sentential word fillers. and a cell given  by  and  i.e.  columns  columns  containing  As each blan] had a column  for each student’s answer with .a score already 1 the  first  method,  it  was  easy  to  delete  the  unwanted columns and automatically get the total score for each of the columns and rows plus the overall totals for the rows and columns. weighted—scores  The advantage of this method over the  method  was  that  with  this  method  easier to give exact—word points, delete columns, graphs to types, tuning, answers,  it  was  and make  show the scores given to inter—sentential blank  all that was  in is  unbiased,  an to  done  say  straightforward  giving  later.  points  There was  to  way.  Fine  acceptable-word  still  the  danger of  biased scoring but the scoring was more transparent than the weighted  scores  could  be.  The  only  regret  was  that  in  102  scoring  no  correct  acceptable—word  restraints  specific  or  not.  attention  was  answers  This  paid  violated  awareness  of  to  whether  the  inter-sentential possible  beyond—  sentence constraints and their violation can be built into future projects.  Because the key expected answers and the  students’ answers are on the spreadsheet the scorers will be able to check for violation of the sentence constraints and assign partial points if desired.  THE SUBJECTS  The classes  subjects of  male  (mostly and  female)  female  were  adult  students  (ESL)  students  in  the  pilot  four  students.  students represented a variety of nationalities from Pacific Rim countries,  from  The  (primarily  some from South America).  project  class  were  The  enrolled  Columbia College in Burnaby, British Columbia.  at  Their upper-  intermediate program consisted of classroom work covering 14 weeks,  five hours/day,  main project,  all  five days/week.  The students in the  from one university  in Japan,  were from  three classes in the Ritsumeikan University English Program at  the  University  These students, were studying  in  of  British  Columbia  between the ages  of  in  19  Vancouver,  and  22  an ESL program based on their  interest, often business and culture.  years  B.C. old,  fields  of  All the students had  done the TOEFL test and were short of the 550 which was the normal standard for foreign students studying at UBC.  The  103  class profiles given in Table 4 below show how the students were  divided  for  their  UBC  studies.  For  the  present  research the classes were each assigned one of three doze formats.  Class one was given the standard doze,  the randomly—colored doze, doze.  Note that the  class two  arid class three the color—coded  assessments done by the Ritsumeikan  staff in order to assign the students to their appropriate proficiency proficient,  levels Class  3  showed  that  Class  less proficient,  1  was  the  most  and Class 2 the  least  proficient.  TABLE 4 CLASSIFICATION OF STUDENTS BY TOEFL AND THE LANGUAGE PROFICIENCY INDEX AS DETERMINED BY THE RITSUMEIKAN STAFF FOR PLACEMENT INTO THE RITSUMEIKAN PROGRAM o  TOEFL Range .. Class 1 53 (500 — 553) °Class 3 34 (483 — 517) °Class 2 40 (483 — 523) 0  °  °  • LANG. PROFICIENCY INDEX Avg. ° Range....... Avg. 527.80° .1 (1 9 — 2) ° 1 94 505 750 1 0 (1 0 — 2) 1 80 494.80° .5 .86 ( .5 — 1) °  THE TREATMENT AND CONTROL GROUPS In  the  pilot  project  18  students  from  the  Columbia  College ESL class were divided into basically four groups of four students. based  on  the  The selection of students for each group was pre-test  scores  such  group was the same on the pre—test. 1 in Chapter 4.)  that  the  mean  of  each  (Please refer to Figure  104  It should be noted here that the process  of dividing  the students into groups was changed from the intended plan due to the choice of some students not to participate in the study and due to the more difficult  fact that the pre-test proved to be  for the students than had been anticipated.  It had been intended to divide the 20 students in the pilot class to be  divided  into  four groups  of  five  and to make  sure there would be, as determined by the researcher ranking scores  on  the  pre—test  strong member,  doze,  a  an average member,  a weak member.  strong member,  a  fairly  somewhat weak member,  and  The students were to have fallen basically  into three groups.  The strong group was to be defined as  getting a score of at the independent reading level (43% and above).  The average group was to be defined as getting a  score at the instructional reading level weak group was score fact,  at the all  to be  defined as  those  frustration reading  level  the  students  in  the  (38%  43%).  -  students (below  pilot  38  The  getting a  %).  project  In got  frustrational level scores on the pre—test. As a result of the pilot project,  and keeping in mind  Oiler’s finding about the proficiency levels of non-native speakers, more proficient ESL students were looked for.  The  University of British Columbia provided three classes of 20 Japanese students each.  Based on the available numbers of  classes plus the results of the pilot project in Chapter Four) only  three  types  (to be given  a new research design was chosen in which of  doze  tests  (two  controls  and  one  105  treatment) were given. the three tests. assigned, group As  profiles  turned  assessed group,  In fact,  out, the  by  the  the  second  class  highest  with  teachers became  highest  the  became  the  leader.  rating  the  as  control  color—coded  doze  and the weakest group became the random-  color control group. somewhat  from the Ritsuiueikan project  Ritsuineikan  treatment group,  was  at the time the test forms were  the researcher had not been given the students’  level  it  Each class was randomly given one of  The matching of classes to test types  intuitive  and  perhaps  a  touch  biased,  but  indeed it was a fortunate choice because all but six of the students  the  in  weakest  group  project after the pre—test. even  those  class  had  dropped  out,  assigned  that  Fortunately the  the  control  main  related doze  to  the  procedure  the  would  project.  group  continue. rational  have  of  the  If  main  CCC  devastated  the weakest  treatment the  stayed  helping  of  in  to  the  students  test  and  research  color—coded treatment group  effectiveness in  out  After the first practice test  six decided not to been  dropped  and  the  question  color—coded  rational  to  look  beyond  the  sentence of the doze blanks for color—coded words to fill the blanks.  106  THE TREATMENT AND CONTROL PASSAGE FORMS  The nature of the forms  The purpose of the present research was to determine whether or not color-coding parts of speech could help adult ESL students learn to use inter-sentential clues that appear in the test passage. treatment  group  Therefore, given  was  Form  in the pilot project the 4  below  and  each  control  group was given one of the three remaining forms of doze procedure: a)  (Form  color—coded  4)  standard  doze  passages  in  which all the words and blanks were coded according to the  following non-random procedure.  pronouns were blue; verbs:  red;  Nouns  adjectives:  and  green;  adverbs: purple: and conjunctions, prepositions, and interj ectioris: black. b)  (Form 3) standard doze + randomly colored words and blanks.  c)  (Form 2) standard doze + color—coded blanks only,  d)  (Form 1) standard doze  Note that standard doze means the deletion of every fifth word,  regardless  blanks. 300  of  type  of  word,  with  standard  length  The first and last sentences of the approximately  word  traditional  passage doze  are was  left the  intact.  reference  The point  standard for  the  or  pilot  project in the present experiment and thus was used for the control group and for the basic design of the color—coded  107  doze  and  the  other  controlling  variations  in  treatment.  Each of these forms of doze is described in Table 5 below.  TABLE 5 CONFIGURATIONS OF THE STANDARD CLOZE PROCEDURE AND VARIATIONS (PILOT PROJECT) °TYPE OF CLOZE  °  0  0  0  0  0 nth  °  e  0  0  0  0  0  0  0  0  0  °Forml) °STANDARD doze  °  DELETION =  Yes  5  0  °  BLANKS  • •  °  0  0  Yes  0  0  0  0  0  °Form2) °Standard doze °ALL BLANKS °COLOR-CODED  0  0  0  0  +0  0  o  •  0  Yes  0  0  Yes  0  0  °Form4)  0  Standard doze +0 0 °all blanks/words °COLOR-CODED 0  0  Yes  •  0  Yes 0  0  Yes  0  0  0  0  0  0  o  e  0  0  0  0  0  0  0  0  0  0  Yes  Yes  0  o  0  °Foriu3) °Standard doze +0 °all blanks/words° °RANDOMLY ° 0 °colored  °  0  1st and last °sentence  0  0  INTACT  0  300+/—6  50  Yes  WORDS  0  Yes  °  Yes  °  0  0  0  0  0  0  0  0  e  0  0  •  e  0  •  0 0  Yes  0  Yes  Yes  0  Note: In the main project form 2 was not used and the other three forms used a rational doze in which as many re iterative words were deleted as possible.  0  0  0  108  The method of coloring the passages  The process of preparing the doze passages for color printing on an a printer compatible with the IBM personal computer will be given in the Appendices.  Please refer to  Appendix B.  THE STORIES USED FOR THE CLOZE PASSAGES  Introduction to the selection of the passages Originally  it  had  been  intended  to  use  passages throughout the CCC research project.  the  same  For the main  project, however, due to circumstances beyond the control of the researcher,  a new set of passages had to be chosen from  a text provided by the institution involved. The  doze  passages  for  the  pilot  project  were  taken  from a recent ESL text called CANread by Patricia Raymond (1987).  It  contains  ten  stories  about  Canada.  After  a  database analysis of parts of speech plus the frequency and location of deleted words, seven stories were chosen.  Based  on the number of same—word and same-root re—iterative clues available, on the Flesch readability formula (1948), plus on the Fry  readability graph  (1977),  the stories were ranked  and ordered according to their increasing difficulty. pre—test/post—test  selection,  “Terry  Fox”,  appeared  The to  be  109  ranked in the middle of the difficulty range.  See Tables 6  and 7 for more details about this range. Used for the main project was a university introductory sociology text called Canadian Society, A Macro Analysis by Harry  H.  Hiller  usually  one  section  where  per  (1991).  Several  passages  chapter,  usually  from  more  re—iterative  were  the  words  chosen,  introductory  were  available.  Passages with the most number of sets of re—iterative words were selected and written down.  Then a  square was drawn  around all repeated words and one of each multiple marked for  possible  highlighted, be  deletion.  Words  to  be  deleted  were  keeping in mind that deleted words should not  contiguous  as  two blanks  difficult for the students.  in  a  row would be  extremely  Deletions were also identified  in such a way that the remaining multiple would sometimes be before  and  position  sometimes  of  retrieval  the  of  clues  the  identification,  after in  order  would  deleted  have  to an  word.  determine influence To  help  if  the  on  the  in  the  all the words to be deleted were marked as  to their location relative to their multiples elsewhere in the passage.  i.e. before or after, and within or beyond the  sentence  the deletion.  of  All  these words  were  given  a  symbol to indicate where they were located in relation to the  undeleted  before/after  the  multiples sentence  which of  the  were  within/beyond  deleted  passage had 50 sets of re-iterative words, deletion had to be selected.  words.  As  or no  other words for  The researcher tried to choose  110  the easiest ones for the students to think of, while at the same time making sure that there was a reasonable distance between deletions.  The rationale for the type of text analysis performed on the pilot project The text analysis was done: a) to identify the parts of speech being deleted by the traditional every nth word procedure; and b)  to determine the proportion of the total number of 50 deletions each speech part category involved.  The purpose of this text analysis was primarily for future reference  in  understanding  the  test  results  explaining them in light of the previous research. of  more  immediate  existence  and  goals,  location  the of  process  of  and  in  In terms  determining  the  same—root  re  same—word  and  with  greatest  iterative clues was used to: C)  select  the  same—word  passages  and  same—root  the  re—iterative  number  clues,  of  thus  eliminating the supposedly more difficult passages; and d) weight the scores according to the availability and location beyond or within the sentence of the same— word and same—root re—iterative clues.  111  The presentation of the text analysis for the pilot project Table 6 and Table 7 below present a partial analysis of  the  stories.  story passage, gives  the  iterative  A  lists  all  the  words  of  each  indicates which words have been deleted and  location clues  deletion). speech  Appendix  of  (within  the  same—word  or  beyond  Appendix A groups  categories,  shows  the  the  and  the  sentence  deletions  number  of  same—root  into  deletions  re  of  the  part in  of  each  category and gives the percentage of the total 50 deletion total  per  story.  distribution clues,  of  the  Appendix same-word  A  also and  shows  same-root  a  suimiaarized re—iterative  as well as indicating the possible values for answers  based on the location of the these re-iterative clues.  TABLE 6 NUMBER OF BLANKS/50 DELETIONS WHOSE FILLERS DO NOT APPEAR AT LEAST ONCE SOMEWHERE IN THE PASSAGE AND THE NUMBER OF WORDS IN THE PASSAGE. (STORIES ARE IN THE ORDER AS THEY APPEAR IN CANREAD.) °STORY °NUMBER  TITLE  °BLANKS/50°WORDS IN° °PASSAGE  o  COMMENTS C  °Story 10 The Donnellys°  18/50  °  303  (pre/post—test)°  °Story  18/50  °  300  (no change)  20  Louis Riel  °Story 3° Sasauatch °Story  40  °Story  50  Bethune Japanese  °Story 6° Terry Fox °Story  70  Insulin  °  24/50  °  295  °  (2 changes)*  19/50  °  305  °  (no changes)*  303  0  (no changes)  0  23/50  0  21/50 24/50  °  0  297 300  (2 changes)* 0  (2 changes)*  *changes from original text made to simplify text  112  TABLE 7 AN ANALYSIS AND ORDERING OF THE CLOZE PASSAGES BASED ON A) THE NUMBER OF NON-RE-ITERATIVE CLUES FOR THE BLANK FILLERS AND B) READABILITY FORMULAS. 0  0  °STORY° 0  ç  TITLE  0  0  01:  °Terrv Fox  0  0  0  2:  °The Donnellys  °  0  0  0  03: 0  °Louis Riel 0 Bethune  04:  21/50 mid 18/50 low 18/5 low 19/50 low 14/50 high 23/50 high 24/50 high  0  0  0  °5:  °Sasquatch  0  0  0  06:  °Japanese 0  0  0  0  BLANKS/50 (non-re-iterative)  0  0  07:  Olnsulin  0  0  °8. 0  °Terrv Fox 0  FRY Grade  0  0 0 0 ° 0 0 0 ° 0 0 0 0  0 0  FLESCH Grade  8 mid  0  21/50 mid  0  0  0  READABLILITY  6  7  —  8  mid 6 low  7 low 7 low 8 mid 7 low 8 mid 9 high —  7  —  0  8  0  mid 7  —  8  mid 6 low 7  —  0  0 0  8  mid 7  —  8  0  mid  8 mid  0  7—8 mid  0 0  (low, mid and high = the relative difficulty and the passages, when compared with each other)  The rationale for the type of text analysis performed on the main project passages The text analysis was done to: a)  select  the  same—word  passages  and  with  same—root  the  greatest  re—iterative  number  clues,  of  thus  eliminating the supposedly more difficult passages; and, b)  give  as  many  iterative  clues  intraas  and  possible  inter-sentential for  the  students  re to  locate with or without the benefit of color-coding.  113  The presentation of the text analysis for the main project Tables B and 9 below are a sunmtary of the data gathered about the deletions in the passages from the text called Canadian Society, A Macro Analysis. TABLE 8 NUMBER OF BLANKS/50 DELETIONS WHOSE FILLERS DO NOT APPEAR AT LEAST ONCE SOMEWHERE IN THE PASSAGE; AND THE NUMBER OF WORDS IN THE PASSAGE. ‘STORY ‘NUMBER ‘Story  ‘  TITLE  *  °BLANKS/50’ WORDS IN’ 0 ° PASSAGE  10  Perspectives’  e  o  ‘Story 2’ Society  ‘  20/50  ‘Story  30  ‘  13/50  °Story  40  °Story  50  0  Ethnicity Uniqueness  0  10/50  •  337  COMMENTS 0  ‘2 words repeated’ bymistake  ‘  °  °  333  °  328  °  15/50  °  335  °  Polities  ‘  13/50  ‘  349  ‘  ‘Story 6’ Identity  ‘  12/50  •  329  ‘  ‘Story 7’ Perspectives’  10/50  337  ‘  0  0  0  0  U  0  0  M  ‘.o 01 0  0  o  U  I--’ 01 O  o 0  01 0  0  01  O  01  U -J  U 0 •%  0  0  0)  10  i  l  1!;!-  (t 1  I’’  I Ict  0 i i  r10  0  01  :.  0  0  cn ci-  UI  O  .  o  0)  10  I”  10 Ict  r  j  III  iPI  0  •  j0  O  -  :  o  0  o ci0  o  rn ci-  01 0  ‘.  .J  0  0  UI  %  U 01  0  u)  I  i I’ I  t  I  I::I  I  0  .  0  0  cn rt  0  0  01 0 01 0  ,  M  ts3  -J  0  0  Ui  U 0  1  r  0  0  01  -  U  0  I  I i’ 1  I 0  10  10  IN  frt I’  0  M  Ii ‘<  0  0  rn ci-  0  U  1<  II  0  0  o ci-  o  01 0  I—’  U  0  0  01  .b 0  0  0 10)  1<  I’•  10 10 I  ‘  I  rtJ10  0  I—’  Ii ‘<  0  0  o ct  0  0  0  t N  H  0  ‘  N  0  o  a  o  a  H U  •  a  H Ui  o  01  ‘.0  a  H (31  0  0)  U L)  0  0  01  U  F-’  0  0  F-’  0  0  H 01  0  0  01  0  01  I-’ 0  0  10)  10  Ii-’  ct  a  0  10)  IJ 10 P1  0  0 ti <  0  rn ci-  )  0  U U U  i  ci-  IU) 10 10 ii-’  0  0 ‘1 <  0  cn ci-  0  o  01  ‘  (1)  Z  0  H  0  0  Z  ‘W  oe  0  NH  0  ‘Cl)  Cfl  00  0  0  WJ  10  C1-3  0  Zcn  0  0  I3  H 1.3  o  U U i  U  0  0  01  01  H  0  <  in-  10 in  10  ‘  IN Kt 1Z’ i ii-’  0  w  1<  0 II  0  o ci-  10  i  II-’-  I::  Ic  0  0 ‘1 <  0  cn ci-  Z ‘-3  I—’ 01  0  U  t3 .D  0  0  Vi  U  F-’  0  I’-’It rn  ic’-  IF-’ Ii-’-  10  II  0  01  1<  0 ti  ci-  0  Nt-I  Cl)  -  U -.J  U  0  0  0 0  Vi  1)  F-’  1<  I’—’•  10 i i-i-  IPi  Il-i  0  O  0 ti  0  o ci-  VI  0  I-’  0  Ill)  lIT)  ii-’-  0  1t1  0 P1  I’  0  .J  1<  0 t.i  0  cn (I  (1)1-3  4Z  (1) N i  W  0  CflI  I  W  N  WI-I =4111  HO  Nt-’  l3W  H  N I3  0 ITj  0  Z 0  lJ  1-3 F-  Me.ZI  1-3  NZ Z  OH  0  0  N’•<  0 W1  zrn i-3  OH  N t  Cl)  H  N Cl)  i-3 ‘  t’1  00  N’ 0(1) (1)  N  HItllt N  W H  zz  -  Cl)  11  HZ  115  THE ADMINISTRATION OF THE CLOZE PASSAGES FOR FIVE CLASSES  Overview of the distribution of the test and practice doze passages in the pilot project In the pilot project each group of students was given the  same  test.  standard doze passage as the pre—test  The passage was not modified in any way.  six more  passages  were  more the first week,  given  to  each  of  two the second week,  the  and post—  A series of classes,  one  and two the third  week, with the sixth and post-test coming during the fourth week.  In  students  each  had  a  class  each group  different  of  variation  four of  (sometimes  the  five)  passage.  The  distributions of doze tests were as follows in Table 11.  TABLE 11 DISTRIBUTION OF THE TREATMENT, VARIATIONS CLOZE TESTS, WITH VARIATIONS APPLIED (EVERY NTH WORD) CLOZE PROCEDURE FORMAT a  AND CONTROL TO STANDARD  Pilot Class  °(FORM 4)  Color—coded  °all blanks and words color-coded °blanks and words colored randry  -  -  0  0  (FORM E)  random  °  (FORM 2)  blank colored °only blanks color-coded  0  (FORM 1)  standard  0  °  °no color given  Note: each cell had four students.  116  Length of time required for pilot project Each contact session with a class was a maximum of 50 minutes.  That allowed 30 minutes for the students to fill  in the doze passage and 20 minutes to be divided up between initial instructions and feedback for the previous session’s doze passage.  Given a total of eight sessions over a four  week period that meant 6.6 hours of class time per class. In other words, a total of 33 hours spread over five classes was  required  of  the  ESL  institutions’  hours at Columbia College.  teaching  time,  6.6  In fact, from the second session  onward the students took less and less time.  As a result  less time was given to the students in the main project to do the  doze passages;  (20  minutes  for the test  and  five  minutes for looking at the answer keys)  Overview of the distribution of the test and practice doze passages in the main project In the main project each group of the  same  test.  students was given  standard doze passage as the pre—test  The passage was not modified in any way.  and post—  A series of  five more passages were given to each of the classes,  two  the second week, and two the third week, with the fifth and post—test being administered during the fourth week. of the three classes potential  of  20  Each  in the main project was made up of a  students.  Table  20  shows  students in attendance for each session.  the  number  of  Each class had a  117  different color variation of the passage.  The distributions  of doze tests were as follows in Table 12.  TABLE 12 DISTRIBUTION OF THE TREATMENT, VARIATIONS AND CONTROL CLOZE TESTS, WITH VARIATIONS APPLIED TO RATIONAL CLOZE PROCEDURE FORMAT IN WHICH DELETIONS WERE CHOSEN TO MAXIMIZE NUMBER OF INTRA- AND INTER-SENTENTIAL CLUES 0  Class 1  Class 2  0  0  Class3 0  °(FORM 4 ) o  Color—coded  (FORM 3 )  °  random color 0  0  0  -  °(FORN 1)  no color —  Note: Each class had a maximum attendance of 20 students, but the scores for only 13 students each from class 2. and class 3 were used for statistical purposes. Only the above 26 students completed the pre-test, post-test, and most of the practice tests. The majority of class 2 dropped after the pre-test and the rest after the first practice test.  Length of time required for the main project For  the  remaining  pre-test  six  tests  25  20  minutes  minutes  were  were  given.  given  for  For  the  completing  each of the them and five minutes for reviewing the answer key of the previous test. instructions  The time used for presenting the  for the pre-test was much more  planned  for:  However,  subsequent sessions were on schedule.  approximately  45  minutes  than had been  instead  of  15.  118  The roles of the researcher, teachers In  order  to  administer  the  and the students  doze  practice  and  test  passages it was necessary for all participants to understand their roles correctly. of  the  researcher,  described in detail.  Therefore,  the  in this study the roles  teacher,  and  the  students  are  As it was necessary for the teachers  to understand the roles of everyone and the students to know their own roles, the necessary instructions for carrying out the experiment were included in handout fori and were given to the appropriate  participants.  The originals of these  handouts are included in the appendices of this thesis.  The  role of the researcher is given in Table C, the role of the teacher in Table D, and the role of the students in Table E and  Table  F.  Table  E  contains  instructions for all the groups.  general  and  specific  The specific instructions  were read out by the teacher as the students read along in their test folders. the  groups  using  instructions were  Table F adds special instructions for color—coded  included  passages.  in the  doing the color-coded passages.  The  special  folders of the students In the pilot project the  special  instructions were on the  folder.  For the main project,  inside  left page  thanks to a  of the  suggestion by  Vera Wojna of the Language Institute of the University of British Columbia, the top sheets.  of  the  the special  instructions were placed at  first page of the  two page doze  passage  This adjustment was made because it was believed  that students would not have to look so far to find out the  119  meanings  of  color—coded words  pilot and main projects the read The  out  loud  groups  by  the  doing  and  blanks.  both  the  special instructions were not  supervising  tests  For  with  researcher  randomly  or  teachers.  colored words  or  no  coloring were not provided with any special instructions. In  order  teacher’s  to  role  give  and  an  part  idea of  described at this time.  the  of  the  proceedings,  researcher’s  role  the are  In the pilot project the teacher  was not required to take part, as the researcher carried out the teacher’s role main  project  the  in administering of the tests. teachers  were  charge  in  of  In the  their  own  classes. In general,  each day  some exceptions. the  doze  students’  followed the  On the first day,  folders anonymity  and  pencils.  while  same routine,  the teacher handed out  In  providing  order the  to  ensure  researcher  information to allow the tracking of the students’ from test to test,  asked to  open the  folder to the  inside left page of the folder. instructions  with  the with  results  the students wrote a self—chosen secret  code on the paper on the outside of the folder. were  with  the  Next,  students.  Then they  instructions  on the  the teacher read the These  instructions  contained Ashby—Davis’ suggestions for doing doze passages. When finished this,  the teacher told the students to begin.  While the students were doing the doze, the teacher watched to see there was no talking or other forms of compromising the  test.  Approximately  every  ten  minutes  the  teacher  120  indicated on the board how much time was left. specified time was up, pencils  down  and  After the  the students were told to put their  close  their  folders  and wait  teacher collected the folders and pencils.  while  the  Then the teacher  continued with the regular classwork if there was time.  In  the pilot and main projects there was no time left before the end of the class. For  the  pilot  project,  during  researcher collected the folders the  tests.  The  pre-test  the  same  and pencils,  scores  were  used  day  the  then marked to  group  the  students by assigning the scores so that the sum totals of each group were equivalent. given  number  a  to  Then, each student’s folder was  identify the  indicate the group,  These  Appleworks’ passage  in the  the form of the test,  student within the group, pre—test.  student  sets  and the of  were  score on the  typed  onto  spreadsheet and the scores recorded.  two  tests  were  stapled  into  a  to  the rank of the  student’s  numbers  class,  an  Then the  duplicate  set  of  identified folders. On day two through the last day the procedure was the same. keys.  The  only  exception  On day three  was with  regards  onward after the  to  the  answer  folders and pencils  were collected the teacher handed out the undeleted text of the previous  day’s  review individually.  passage  for  the  students  to  read  and  After about five minutes the teacher  collected these passages.  121  Day eight was the  last session of the pilot project.  Due to time constraints, the  main  project.  At  session seven was the last one for the  end  the  researcher  teacher  or  thanked the students for their cooperation and a small gift was given to each of the classes.  The routine for the main project was similar except for two procedures. test  data  on  First,  the  it was not necessary to enter the  spreadsheet  and  analyse  it  right  away.  This was because the assignment of test forms did not depend on  the  pre—test  assignment  had  class lines.  scores. been  This,  in  turn,  quasi—randomly  Second,  was  because  pre—determined  the  along  the students in the main project were  given the answer keys to the day’s test right after the test session  instead  of waiting until  after the  following test  day.  THE SCORING OF THE CLOZE TESTS AND DATA ANALYSIS  Spreadsheets responses computer  for were  were  each  used  on  Using  the  typed  in all  (see Table  test.  an  to  collect  “Appleworks”  sufficient  memory was needed was  used  for  the  pilot  and  386  Appleworks’  computer  spreadsheet  the students’  5 below).  answers  for  an  students’ Apple  project  for the major project.  IBM  the  the  program into a  but  Thus,  more  “Quattro”  latter the  lic  project.  researcher  formatted table  As planned the exact-answer marking  122  system  was  used.  However,  developed to weight scores and Table A  there  was  more  a  a  formula  had  been  (see Table 6 and Table 7 below  in the appendix),  word system was dropped. that  although  this particular acceptable-  This was because of the valid  and  less  insight  cumbersome  way  to  analyze the answers in relation to the goal of finding out if  students were  looking beyond the  immediate  sentence  of  the doze blank rather than just within the blank. The  new  spreadsheet  method the  of  analysis  columns  was  containing  to  the  delete types  from  of  the  fillers  which either required external knowledge, followed by intra sentential re-iterative fillers. To analyze the data for each doze passage two methods were used.  First,  (exact—answer) “Quattro” Second,  bar  was  to get found  graphs  a general  for  (see  idea,  the mean  score  each group and graphed using Figures  3,  8,  two statistical programs were used.  13,  and  18).  The first was  the T—test to see if the two classes involved in the main project were significantly different from each other on the pre-test  and  on  the  post-test.  (Note  that  class  2  had  completely dropped out by the end of the second test because  the uncolored pre—test had proved to be too frustrating for the  majority  and  the  randomly—colored  second  test  too  difficult for the remaining six more determined students.) The  second  Repeated  statistical  Measures  analysis,  ANOVA”  was  called then  “One  applied  Between to  give  and a  sensitive analysis and to check for trends across the series  123  of test passages.  These statistical programs were used only  for the main project where the cells were larger than for the pilot project. It  should be  pointed  “Quattro” were made  out here  that  bar  graphs  using  for a) the mean scores for the full set  of the students in a class who wrote the particular test and b)  a  subset  written  the  passages.  of  students  pre-  and  (n=14  for  post-test  each  and  most  class) of  the  who  had  practice  The graphs of the full set were made to give a  general idea of what the trend was for the different forms of  doze  tests  and  the  subset  were  to  give  a  more  valid  picture of what had happened in the tests. To use statistical  the  “One  program  Between and Repeated Measures ANOVA”  the  score  data  for  the  five  practice  tests were typed into a data file and run on the Main Frame Computer at the University of British Columbia.  The purpose  of the statistical analysis was to determine whether or not the treatment (completely—color—coded doze practice) groups had made any significant improvement compared to the control (standard  doze  practice)  groups  and  to  the  other  color  treatment groups and also to see what effects the passages had on the group scores. The data were analyzed and interpreted in Chapter Four, and  recommendations  Chapter Five.  and  further  research  suggested  in  124  TABLE 13 ANSWER COLLECTION AND BLANK SCORE DETERMINATION CHART AS IT LOOKED ON A SPREADSHEET (PILOT PROJECT) Page 1 TOTAL PRETEST  TEST 2 TOTAL  MEAN EXACT MTH  MEAN EXACT MTH  MEAN ACCEP MTH  MEAN ACCEP MTH  Page  1  SCORES FOR BLANKS  POINTS FOR......... EXACT WORD  Exact Mth  =  EXPLANATION.  CLUE LOCATION  same wrd  : diff wrd  T1:not in passage  T:1O/9/8  .. .....  POINTS FOR  TYPES OF CLUE...... T2: beyond senten T3: within senten  T:5/4/4  :  T:7/6/5  :  1  8/7/6  6/5/5  Cloze Form 4  lA4a: Student 1 Class A Group 4 Level a  (everything  2A4b: Student 2 Class A Group 4 Level b  color—coded)  3A4c: Student 3 Class A Group 4 Level c 4A4d: Student 4 Class A Group 4 Level d 5A4e: Student 5 Class A Group 4 Level e  MEAN EXACT MTH MEAN ACCEP MTH  125  doze Form 3 (everything.  6A3a: Student 6 Class A Group 3 Level a .... . . .  7A3b: Student 7 Class A Group 3 Level b  color-coded  8A3c: Student 8 Class A Group 3 Level c  randomly)  9A3d: Student 9 Class A Group 3 Level d lOA3e: StudentlO Class A Group 3 Level e  MEAN EXACT MTH  MEAN ACCEP MTH  Cloze Form 2  llA2a  (blanks  l2A2b  color-coded)  l3A2c l4A2d l5A2e  MEAN EXACT MTH MEAN ACCEP MTH  doze Form 1  l6Ala  (Standard  l7Alb  doze)  . . . . . .... .  .l8Alc l9Ald 2OAle  MEAN EXACT MTH MEAN ACCEP MTH  126  Cloze Form 4  21B4a 22B4b 23B4c 24B4d 25B4e  MEAN EXACT MTH  MEAN ACCEP MTH (For more see Appendix D.) page  1  BLANK1  B2  B3  his 0004  son 000(1)  He 10002  same  same  same  T:10/9/8 T:7/6/5  T:7/6/5  127 TABLE 14 CRITERIA FOR DEVELOPMENT OF WEIGHTED SCORES  FOR THE PILOT PROJECT A. Availability of clue 1. clue is not in passage at all 2. clue is beyond sentence of blank 3. clue is within sentence of blank B. Relation of clue to deleted word 1. clue is same as deleted word (re—iterative) 2. clue is same part of speech (synonym) 3. clue is same part of speech with same root (friend.. .friendship 4. clue is different part of speech, but with same root (eq. except. . .exception) 5. clue is a phrase equivalent to the deleted word (eq. in the world...universal) 6. clue is a pronoun or antecedent (Bill...he) 7. clue is a referent (eq. the boy...Bill) 8. clue is a cohesive pair (eg. salt and pepper) C. Correctness of answer 1. answer is exact and same, i.e. same word as blank 2. answer is acceptable to meaning of passage 3. answer is acceptable to meaning of only the sentence of the blank  TABLE 15 CALCULATION OF CLOZE PASSAGE SCORES FOR THE PILOT PROJECT  To  calculate  the  scores  categories,  use  the  procedure.  1. Start with category A: availability of clue.  The highest possible score is 10. If clue is not in the passage,  Al.  =  10  If clue is beyond sentence,  A2.  =  8  If clue is within sentence,  A3.  =  6  following  128  2. Then consider category B in conjunction with A.  This  gives three types of scores as follows.  TYPE 1:  (no same-word (re-iterative) clue within  passage) pvAl, if Cl (answer is same as blank)  10  =  Al, if C2  (acceptable to passage)  =  9  Al, if C3  (acceptable to sentence)  =  8  TYPE 2:  (clue beyond sentence) (clue same as deletion)  A2,  if  Bi  A2, =  7  andCl  =  8  or  =  6  or  C2  =  7  =  5  or  C3  =  6  C2  (clue within sentence) (clue same as deletion)  A3,  if B2 through B7  andCl  or TYPE 3:  (clue different)  if Bl  (clue different) A3,  if B2 through B7  andCl  =  5  andCl  =  6  or  C2  =  4  or  C2  =  5  or  C3  =  4  or  C3  =  5  129  TABLE 16 ANSWER COLLECTION AND BLANK SCORE DETERMINATION CHART AS IT LOOKED ON A SPREADSHEET (MAIN PROJECT) °NAME OF PASSAGE °  TYPE OF BLANK° (EXTERN) 0 B1  0 0  0  0  KEY WORD  0  ***  ° 0  INTERB2  ° 0  0  0  0  0  INTRA B3 0  °STUDENT °IDENTITY  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  °JNK °DAV :APR  0  1  0  0  0  0  AAAAA  0  0  Classl  0 T otal: °Mean C  : 0  1 2 .66  e  °Class3 0  0  °TM  °  0  0  :  :  °Total: °Mean:  0 0  (empty) 1 (empty) 1 •33  : 0 0  1 (empty)  :  1  (empty) 1  :  1 33  0  0  0  0  0  33  0  a  0  0  0  °  0  0  : • °  1 1 (empty) 2 .66  0  :  (empty)  •  o  0  0  :  130 CHAPTER FOUR  ANALYSES AND DISCUSSION INTRODUCTION Chapter  In  project were  One  the  given.  reasons  for  the  present  research  Chapter Two presented the historical  background which underpinned these reasons and suggested the framework for the research design.  Chapter Three set out  the design of the pilot project and mentioned some of the results which led to the design of the main project. now the design  role  Chapter Four to  (please see Table 17  problems 18),  of  and  solutions  in  summarize the  below), the  It is  changes  in  point out some of the  projects  (located  in  Table  and in light of all of these to present the research  data.  The data are of two kinds.  at the scores for the right answers.  One kind includes a look The other looks at the  same information as it interacts with the confidence levels of the students  (the confidence levels being determined by  the  doze  number  Finally,  in  of  Chapter  blanks  Four  some  filled  by  conclusions  the will  students). be  drawn  about the effectiveness of the color—coded doze procedure in helping students  look for inter-sentential reiterative  word clues  in  for use  filling the  doze blanks.  Chapter  Five will discuss implications of the research and suggest a variety  of  possible  color—coded doze.  future  designs  for  research  with  the  131 CHANGES MADE TO THE CLOZE PROCEDURE AFTER THE PILOT PROJECT  As a result of observations made in the pilot project a number  changes  of  were undertaken.  They  are  outlined  Table 17 below. TABLE 17 CHANGES TO CLOZE PROCEDURE AFTER PILOT PROJECT 0  TYPES OF °CHANGES  0  PILOT PROJECT  ° Prof ici— ency level of • students  •  0 °  0  00  upper intermediate (as defined by the participating college)  ° ° 0  0  e  0 0  °  °  0  0  °  °  Assign— • ment of • • students to groups°  according to pre— test results: the students were placed° so the groups’ mean scores were the same°  0  0  0  No. of four • forms of • the doze 0 0 0 °  0  0  0  0  0 °  •  No. of tests  0  °  •  0  0  0  0  0  0  0  0  °  0  0 0  °  “Canadian Society” an introductory university text on ° Canadian issues  0  0  0  0  0  °  0  •  0  0  °  330 words (approx.) (to keep the sense of the passage)  0  0 0 0  -  Length of time to dothe doze tests  30 minutes for each test  -  0  300 words (approx.)  seven (to save time and reduce boredom)  0  °  —  0  0  0  eight  0 0  0  0  ° Source of° “CANRead” an ESL passages e anthology about 0 0 Canadian heroes 0 0  V  0  °  °  °  °  each of the three classes was given a different form of the doze test  °  0  Length ° of the ‘ passages  0  upper advanced (as defined by TOEFL and Language Proficiency Index, three levels mdi— cated) to make doze easier  three (to increase the no. of students 0 0 in each cell to lessen the factor 00 of chance)  0  0  0  0  0  0  0  MAIN PROJECT  0  °  0  0  •  0  o  0  0  25 minutes for pre-° and post-tests; 20 minutes others (in pilot only 15 minutes needed) 0 0  0  0  0  in  132  Deletion ° pattern 0  0  every fifth word  0  0  0  0  0  0  0  0  0  0  0  0  0  xeroxed passages; underlined and counted the reiterative-words; wrote the passages 0 into a notebook; 0 highlighted reiterative—words; deleted some from ° before, others • from after, some from between other members of the set.° Blanks were never side by side.  °  0  °  0  typed passages onto spreadsheets and deleted every fifth word  0 °  °  0 0 0  0  0  0 0 0  0  e  0  0  0  0  0  0  0  0  0  0  0  0  0  ° °  0  0  0  0  0  0  0 Location on the inside of of color— 0 the front cover ocoding key° 0 Feedback  gave answer keys after the following test  0  0  0  0  0 0  ° 0  0  0  0  0  0  0  0  0 ° 0  0  0  0  °  0  0  0  0  0 0 0  0  graphed mean scores for a) all 50 blanks 0 and b) beyond re° iterative Nouns, • Verbs, Adjectives, Adverbs 0 0 0  °  0  °  0  0  0  0  used Quattro (IBM) spreadsheet; used rows for students, columns for key answers for each doze blank and each student’s anonymous ID listed within class in no particular order; labelled each blank° as to type (intra-,° inter—sentential, external (nothing)  0  0  0  0  0  0  0  0  gave answer key immediately after its practice test  0  0  Analysis  0  0  °  0  0  on top of page 1 of 0 the passage (for easier reference 0  °  0  0  0  •  Data used Appleworks ° Collect— • spreadsheet; used 0 ion rows for students, 0 0 columns for key answers for each 0 0 doze blank and each° student’s answer; 0 0 assigned each stud— ent a ranking number° 0 0 (Student 1. Class A, 0 0 Group 1, Level a) to° attach to their own anonymous ID code; 0 0 each answer weighted 0  0 0  0 0 0 0  0 0 0  °  0  0  0  °  0  0  0  0 Method of° deleting words 0 ° 0 0 0 0 0 0 0 0  o  rational deletion (to maximize no. of reiterative— word blanks)  0  •  graphed mean scores° for a) all 50 blanks, b) intrainter—sentential, c) intra-sentential° and d) inter-sent0 ential blanks °  +0  0  °  0  0 0  133 PROBLEMS AND SOLUTIONS IN PILOT AND MAIN PROJECTS Table 18 which follows looks at the changes in terms of the problems faced in the pilot project.  Changes are given  as intended solutions to those problems.  Problems in the  main project are presented as a guide to solutions in subsequent doze procedure research. TABLE 18 PROBLEMS AND SOLUTIONS IN PILOT AND MAIN PROJECTS  • PROBLEMS and 0 SOLUTIONS 0  °  °  PILOT PROJECT  ° Schools: 0  0  ° 0  tight schedules 0 to finish their own curr— ° iculuius, but they volunteered one °class 0  0  0  0 0 0  0 0 0  0  0  0  ° 0  0  Test ° Location:° 0  0  0  a 0  0  0  0  0  0  Students: quantity  :  o  a  o  a  0  number insufficient for statistical purposes (4/cell)  0 a few students mortality o missed tests, most wrote them another day °  0  a  0  0  0  0  °  ° °  ° ° °  °  °  °  MAIN PROJECT  0 0  tight schedules to finish their own curriculums, but volunteered three classes  0  0  0  00  ° did not accept pilot project 0 passages, but 0 0 offered a text 00 fromtheir °°program °  0  0  0  0  0  0  0  0  00  0  0  0  0  0  °  °  a  0  0  a  0  0  a  0  0  0  0  a  0  0  0  0  0  a  0  a  a  0  e  0  0  a  0  0  0  0  °  poorly lighted, but teacher of treatment group put strong lamp onatablein middle of group  0 0 0  0  0  0  number barely sufficient (20/cell max) (13/cell wrote most tests and so were suitable for doing statistics) a few students missed tests and did not do them later, but their scores were based on the mean of  0 0 0 0 0 0  a 0 0  a  0  S9.  • • • • .  : a  0  •  •  o  UOTSS3S. O. SA e  et; uo epuz e.xaz SeTdODo;otd—oToD vq ‘AtToT;..xe apoo—otoo o. ubeq .oeço.xd et; jo pue e; .xeu .xe;u.xd—ioioo et.  pu -axd et. uo .ieio eq o. se.xoos P9SrLD .xO.x.ze StT.fl. T  UTd  •  0  0  c.eA)  0  0  0  .  00 0 0  0 •  0  0  0  .  •  .  0  0  e.tez4et; ;t’ s;uepn;s pro; atesai ;nq ‘sp.xo, jo dno ietoun euo  0  0  0  0  0  0  0  a  0 0 0  0  0  0  •  0  0  0  bUTdI;  .  0  ioie  • 0 0  bUT;UtId  .UOT.OflpOJd.  .tflDTJTP LeA  .tflDTTP LeA  00  0 dno.xb ;ueiu;ei; et; ..zo tox;uoo UT eti; ;ou ;nq ‘;rto pecIdop SSTD etot s;se; eoT;oxd et. tt ;o sueut  a  •  • •  :: 0 0  UOU(OU ST 0 T) 0 e a (;uTod;o. 0 s;uapn.s tie; 0 uTodeuos 0O pe;deoo s ietisu 0 etqssod Au ;nq 0 0 0 ‘se;—;sod/-e.xd 0 0 0 9 UT UOTSflUOD 0 0 . o. buTpet UT 0 0 a pesn .xet;ou pU 0 0 • • U3 je St4 et; 0; ieisu et;  •  0  .  00  0  • o a o O 0 0 0  •  0  ase  0  :s;se;. ezo.  0  a  0  •  0  a  e  o  0  0  •  •  0  0  a  o  a  0  e  o  0  0  o  o  0  e  e  o  0  0  o  o  ‘Ct  135 THE RESEARCH DATA The Pilot Project When the test means were graphed for the four groups the data appeared as on the graphs immediately below. apparent  from  treatment group  these  sample  graphs  did not do as well  that as  overall  the  It is  the  CCC  standard—doze  control group but did accelerate and slightly overtake the main control group on tests 5 and 6, once  again  removed.  in  the  post-test  when  losing the advantage the  color—coding  was  Yet, because the number of subjects in each group  cell was so small  (N  =  4),  it would have been difficult to  state any statistical conclusions with very much confidence. The results  needed to be confirmed by research done with  more students.  Even though the results of the pilot project  could prove nothing, they did suggest that the color-blankonly control was the weakest and could be eliminated,  and  thus allow more students per group in the main project. Using the pilot graphs as a (very rough) guideline and given the limited opportunity to obtain students to do doze research, it was decided to move on and try out ideas uncovered in the review of the literature.  The researcher  wanted to know what would happen when the students were at a higher language proficiency level and when more inter sentential reiterative words were available to be found to fill the doze blanks.  As a result of these questions these  changes were introduced into the main project for all the groups.  It was hoped that more proficient readers could  136  make better use of the color—coded parts of speech to find the reiterative clues that were known to the researcher to be somewhere in the passages.  It was thought that these new  conditions, applied to all the groups, just might improve the treatment group’s scores over the control groups who had everything but the color—coded clues.  To make the color—  coded parts of speech easier for the treatment group to learn, the color—coding keys were moved from the inside of the test folder to the top of page one of each of the practice doze passages.  r  S  W  PILOT: Mi 50 Blanks (N  12  13  1  14  q....  15 —  4)  16  I  Ti  I  =  18  4)  _______ __  GROUP4  CROUP3  CROLJP2  .21 St  1. Standard /2. Color— Blank/S. Random/4. CCC 40 7C JJ  r  2 ” a 20I .J 15  10 5...  0 L _  Ti  tests (for group 1  I.  -fr  (  ZO1[  91  91  ca LL  UI) S4LE1flEJ[  [1  IIV IO1Id  /  ci  sdnf:ijb .ioj) Sj.Sq.  P3P01—J[0i0)  LL  (-t7 pIJI)  S  C;’  —  .-.  If.’  ii..  m  I-’ 4-.) 0,  139  The Main Project An overview The kinds  data  and  for  pertain  the  main  to  groups  project  are  (classes)  comprised 1  and  3,  of  two  group  2  having completely dropped out due to frustration by the end of the  first practice test.  The  first data summarize the  scores for the exact—word and acceptable—word answers given by the students in the control group (Cli) and the treatment group  (C13).  The  summaries  are  in tabular,  graphic,  and  statistical forms for the exact-word scores and tabular form for the acceptable—word scores. the  same  levels  information but  of  the  students.  as (In  it  The second data summarize relates  to  the  confidence  finding confidence  levels,  it  should be noted that for absentee students the actual N was used instead of absentee  including the mean of the means  students  summaries.)  as  was  confidence  filled.  the  other  kind  The  of  interaction of exact—  (only exact—word answers were focussed on) levels  yielded  scores  “productive  confidence”.  This  tabular  graphic  only.  and  for  Confidence levels were defined in terms of the  number of doze blanks word answers  done  for those  forms  under data  the  was  While  and  heading  of  summarized  in  summarizing  and  finding trends in exact—word answer scores were the main and original  concerns  in  the  present  CCC  study,  productive  confidence information emerged during the research as a way to describe and discover what effects the CCC was having on the students’ motivation to use the color-coding to find the  140  same—word  re—iterative  inter—sentential  clues.  Together,  the outcome of these two kinds of data analyses along with the answers from a post—session questionnaire was meant to indicate •the  effectiveness  and value  of using  the  color-  coded doze procedure with adult ESL students.  Summary of the data To determine any improvement in the doze ability the results of each doze test were subjected to four analyses for answer scores as follows: a) all 50 doze blanks, b) intra— and inter—sentential (Erra) doze blanks, C)  intra—sentential doze blanks, and  d)  inter—sentential doze blanks.  Each of these analyses examined exact—word answer scores and acceptable—word included for regards  to  answer  scores.  exact—word answers  Raw—score for all  acceptable—word answer  tables  the tests.  scores,  as  a  are With  matter  of  economy, raw—score tables are only included for the pre— and post tests.  In Appendix H mean—score tables summarize and  compare data for both the exact- and acceptable-word types. Then,  to  determine  any  improvements  in  motivation  (confidence), four more analyses were made: e) the total number of blanks filled, f) the number of blanks filled to the possible number of blank types, g) the total number of blanks filled correctly, and  141  h) the nwnber of correct blanks as a proportion of the nuier of blanks filled. Analysis (e) was meant to indicate the confidence level and (g)  the  each  level  test.  of productive confidence For  the  motivational  of  each group  analyses  it  sufficient to present only the exact—word data.  was  for felt  The reason  for this was that the motivational data was determined by including the score data in their calculation. acceptable—word  scores  increased  the  Because the  distance  between  the  control group and the treatment group scores in favor of the control group,  this meant that the  resulting motivational  scores would be greater for the control group. of  the  motivational  treatment control  group group,  analysis  could  close  there  was  was  the no  to  As the goal  discover  gap  between  point  in  acceptable—word analysis any further.  it  if  the  and  the  pursuing  However,  the  the pursuit  may be worthwhile in subsequent research if the treatm ent group does better than the control group. For  the present  research many tables  and  graphs  are  still provided.  In order for them to be conveniently viewed  and understood,  the data  for answer scores and confidence  scores are grouped together so that all summaries  for  (a)  the 50 blanks are shown, all the summaries for b) the intra and inter-sentential (c)  and  (d).  discussion Following  After  section these  are  (Erra) these  blanks are shown, tables  expressing the  T-test  and  the and  etc.  graphs apparent  ANOVA  through  there  is  a  findings.  tables  and  an  142  explanation.  Included  next,  the  discussion  of  the  motivational data intends to add information to the previous discussion to help decide whether or not the rationalized color—coded doze procedure is worth all the extra effort to produce the color—coded passages.  The motivational data are  presented in that particular location in order to help in the interpretation of the data and in the decision on which to  rely  data.  on  more,  Finally,  discussion  are  the  tabular/graphic  or  the  statistical  a sumary table of the questionnaire and a included  to  further  formulate  the  answer  about the value of color—coding to the doze procedure.  143 TABLE 19 FOR THE MAIN PROJECT TEST DATA: EXACT-WORD (AND SOME ACCEPTABLE-WORD) °CLASS 1 °EXACT-WORD SCORES °  °CLASS 1  PRE TEST Erra  °JNK° °DAV° °APR° °ABC° °JPN° °BOX° °MAT° °USA° °MCN° °TAK° °YAS°  105 10 0 90 6 0 12 5 12 5 10 0 11 0 13 0 11 0 7 0 9 0 7 5  75 6 0 60 5 0 10 5 9 5 8 0 10 0 10 0 10 0 4 0 7 0 5 5  PRE- °(N = cell) TEST 0 (Erra = Inter Intra- + 4.0° Inter— 2.0° sentential 4.0°word blanks) 3 0° 6 0° 7 0° 6 00 6 0° 8 00 8 0° 3 00 5 00 3 50  eTotol2g 0  99 0  65  °ABC°  °Mn  °  °  0  °JNK° °DAV° °APR° °ABC° °JPN° °BOX° °MAT° °USA° °MCN° °TAK° °YAS° °SSK° °ABC°  7 6  5 00  19.9  15.2  10.1°  12 6 10 0 5 0 7.0 13.0 9 0 9 0 3 0 11 0 8 0 8 0 9 0 12.6  ° N ° PRE°CL1° TEST ° ° All °JNK° °DAV° °APR° °ABC° °JPN° °BOX° °MAT° °USA° °MCN° °TAK° °YAS° °SSK° °ABC°  PRE TEST 0 Inter  205155 17 0 12 0 150 110 7.0 6.0 13.5 12.5 14 5 10 5 210 100 18 0 14 0 17.0 13.0 17.0 15.0 11.0 7.0 15 0 11 0 12.5 8.5  120° 8 0 90° 3.0 8.0 7 0 70° 10 0 10 0 12.0 6.0 8 0 6.5  (Tot=Total) °Tot° 199 • 0 (Mn = Mean °Mn Score) °  °  °  146.0 106.5°  153  112  820  30.6  22.2  16.4°  T3 Erra  T3 Inter  10 7 9 0 4 0 6 0 11 0 8 0 8 0 3 0 10 0 6.0 7 0 8 0 10.7  11 0 12 0 7 0 11.0 12 0 5 0 14 0 9 0 14.0 3.0 10 0 10 0 11.0  8 0 6 0 6 0 7 0 6 0 4 0 10 0 5 0 9.0 2 0 5 0 7 0 9.0  17 0 15.0 ° 16 0 ° 12 0 • 15 0 ° 8 0 ° 15 0 ° 12 0 ° 12 0 ° 11 0 ° 16 0 0 15 0 ° 14.0  50.4°146.0 129.0  84.0  0178.0 146.0 109  6 70 4 00 2 00 2 0° 5 00 4 00 3 0° 2 0° 6 0° 3 0° 3.0° 3 0° 67°  13 0 13 0 9 0 12 0 13 0 7 0 15 0 10 0 14 0 4 0 12 0 12 0 12.0  °  9.0  7.8  3.9° 11.2  9.9  6.5  °  18.0  15.6  7.8° 22.5  19.8  13.1  °  PRETEST Erra  T2 T2 ° T3 Erra 0 Inter All  °Tot°117.2 101.4 °Mn  50  9 9  ° N ° T2 °CL].° All  0  °  N ° PRE°CL1° TEST ° ° All  5(O  SCORES  ° °  T4 All  ° °  0  °  T4 T4 ° Erra 0 Inter 15 0 13 0 14 0 9 0 12 0 6 0 13 0 9 0 10 0 8 0 14 0 13 0 10.0  12.0° 10 0° 11 0° 6 00 9 0° 4 00 9 0° 7 0° 8 0° 5 0° 11 00 10 0° 7.0°  13.7  11.2  27.4  22.5 16.8°  8.4°  144  °CLASS 1 EXACT-WORD SCORES (cont’d) N ° T5 T5 T5 T6 T6 °CL1° All Erra Inter All Erra  T6 Intere  °JNK° °DAV° °APR° °ABC° °JPN° °BOX° °MAT° °USA° °MCN° °TAK° °YAS° °SSK° °ABC°  10 0 8 0 9 0 6 7 6 7 8.0 7 0 6 7 17 0 6 7 6 7 6 7 13 0  o  15 12 12 12 15 16 15 6 17 11 18 13 17  0 0 0 0 0 0 0 0 0 0 0 0 0  13 0 11 0 10 0 10 0 15 0 14.0 13 0 6 0 15 0 11 0 18 0 12 0 14 0  11 0° 8 0° 7 0° 7.0° 12 0° 11.0° 10 0° 5 0° 12 0° 9.0° 14 00 10 0° 11 0°  14 0 13 0 13 0 12 6 12 6 13.0 13 0 12 6 27 0 12 6 12 6 12 6 22 0  11 0 11.0 11 0 10 7 10 7 10.0 9 0 10 7 23 0 10 7 10 7 10 7 18 0  0  0  °Tot°179.0 162.0 127.0°190.6 157.2112.2° e  0  0  13.8  12.5  27.6  24.9  9•70  14.7  12.1  8.6  19.4° 29.3  24.2  17.3  EXACT-WORD SCORES  ACCEPTABLE-WORD SCORES  N °POST- POST- POST-° °CL1°TEST TEST TEST ° °A11 Erra Inter 0 °JNK° °DAV° °APR° °ABC° °JPN° °BOX° °MAT° °USA° °MCN° °TAK° °YAS° °SSK° °ABC°  (Erra = intra- + 6.0° inter2.5° sentential 6.0°word blanks) 6 0° 7 5° 8.0° 7.0° 5 0° 13.0° 11 0° 8 0° 10.00 5.0°  N °POST- POST- POST °CL1°TEST TEST TEST ° °A11 Erra Inter °JNK° °DAV° °APR° °ABC° °JPN° °BOX° °MAT° °USA° °MCN° °TAK° °YAS° °SSK° °ABC°  95.0°  °Tot°235 5 182 0 151 0  °  105 7 5 12 0 11 0 11 5 13 0 90 110 15 0 17 0 12 0 15 0 10.0  85 5 5 10 0 9 0 9.5 11 0 70 75 14 0 15 0 11 0 13 0 8.0  °Tot°154.5 129.0 V  °Mn  °  11.9  9.9  7.3°  °  23.8  19.9  14.6°  °  0  °Mn  °  0  0  21 5 12 5 180 190 15 5 18 0 21 0 110 19 0 23 0 21 0 21 0 15.0  17 5 6 5 120 120 12 5 15 0 15 0 105 16 0 20 0 16 0 17 0 12.0  13 0 3 5 80° 90° 10 5 11 0 0 14 0 80° 14 0 16 0 23 0 13 0 8.0  18.1  14.0  11.6  36.2  28.0  23.2  145  ‘CLASS 3 ‘EXACT-WORD SCORES N °PRE‘CL3’TEST ° ‘All  °  PRETEST Erra  °TMI’ 7.5 ‘MMl° 9.5 ‘TM2° 11.0 °HN 0 8 0 ‘MW 0 10 0 °GST° 6 0 ‘YK ° 8 5 °MM2’ 20 ‘HG ° 13 5 ‘HO ° 14 0 ‘MI ‘ 14 0 ‘TI 0 3 5 ‘NN ° 6.0  4.5 7.0 9.0 7 0 8 0 4 0 6 5 0 10 5 10 0 13 0 2 0 4.5  ‘Tot°1l3 5  86 0  ‘CLASS 3 ‘ACCEPTABLE-WORD  PRE- ‘ TEST ‘ Inter’  57  00  87  62  44°  ‘  17.5  13.2  8.8’  0  N ‘ T2 °CL3° All  T2 Erra  °TMI° 0 ‘NM].° 6 0 °TM2° 14 0 ‘HN ‘ 7 0 ° 4 0 °GST° 7 0 ‘YK 0 9 0 MM2’ 10 0 0 ‘HG ° 4 0 ‘HO ‘ 4 0 ‘MI ° 7.0 ‘TI ‘ 6 0 °NN ° 8 0  0 5 12 6 3 6 6 7 3 3 6 6 6  ‘Tot° 86.0  69.0  = cell)  N °PRE °CL3°TEST ‘All -  0 3 7 4 0 3 2 3 1 0 3 4 2  °  0’ 0° 0° 0’ 0° 0’ 0’ 0° °  0° 00  0°  PRETEST Erra  11.5 15.5 14.0 20 0 14 0 9 0 13 5 40 20.5 21 0 16 0 8 5 8.0  PRE TEST Inter  (MN = Mean ‘Mn’ Score)  6.5 11.0 9.0 9 0 10 0 5 0 8 5 10 16 5 15 0 14 0 5 0 4.5  ‘  ‘  4 18 7 10 9 8 11 12 7 2 4 9 9  T3 Erra 0 0 0 0 0 0 0 0 0 0 0 0 0  32.0°l10.0  4 15 6 10 7 7 10 11 5 2 4 8 9  0 0 0 0 0 0 0 0 0 0 0 0 0  T3 Inter 3 9 5 6 5 5 8 7 2 2 3 5 5  0 0 0 0 0 0 0 0 0 0 0 0 0  98.0  65.0  ‘Mn  °  6.6  5.3  2.5°  8.5  7.5  5.0  °  13.2  10.6  4•90  °  16.9  15.1  10.0  135  88  27.0  17.6  ‘ °  ‘  T4 All  4 15 ‘ 11 ‘ 9 0 15 0 9 • 9 14 • 12 ‘ 13 ‘ 9 ‘ 10 0 16 •  0  -  (Tot=Total) ‘Tot’ 175.5 115.0  ‘ T3 T2 Inter’ All  0 0 0 0 0 0 0 0 0 0 0 0  °  (Erra = intra- + 3.0° inter— ‘TM].’ 5.0’ sentential ‘MM].’ 6.0°word blanks)’TM2° 6 0° ‘uN ‘ 6 0’ ‘MW ‘ 3 0’ ‘GST° 1 0° ‘YK 0 0 ‘ ‘MM2’ 6 0° ‘HG 0 7 0° ‘HO ‘ 9 0’ ‘MI ° 1 0° ‘TI ° 4.0° ‘NN ‘  ‘Mn° ‘  (N  RE’ 0 0  5.0 0 9.0 6.0 8 0 8 0 4 0 3 0 0 10’ 12 0 0 12 0 10 0 4 0 4.0 86.0  0  66’ 13.2  0 T4 T4 Erra 0 Inter  0 0 0 0 0 7 0 0 0 0 7 0 0  2 12 9 7 12 8 7 12 10 10 8 8 13  0 0 0 0 0 3 0 0 0 0 3 0 0  1 9 6 4 9 5 3 8 7 7 5 6 9  0’ 0° 00 00  0’ 7’ 00  0° 0’ 0° 70  0’ 0’  ‘147.4 118.6 80.4° 11.3 °  22.7  9.1  6.2°  18.2 12.4°  146  °CLASS 3 EXACT WORD SCORES (continued) • N 0 T5 ° T5 T5 T6 T6 T6 °CL3° All Erra Inter All Erra Inter 0  °TMI° 0)]O  8 0 15 0 11 0 14 0 9 0 10.0 22.0 14 0 16 0 11 0 13 0 9 0 17 0  6 0 13 0 10 0 14 0 9 0 8.0 19.0 13 0 16 0 9 0 9 0 9 0 16 0  2 0 5 7 11 0 5 7 5 7 6.0 12.0 7 0 5 7 12.0 7.0 11.0 5 7  °Tot°169 0 151 0 121 0°l65 5 139 5  96 5  00l3O  116  26.0  23.4  °  °  0  N °POST- POST- POST-° °CL3°TEST TEST TEST °A11 Erra Inter 0 6 5 13 0 12 0 105 95 5 5 19 0 90 110 13 0 12.0 3.5 11.0  4 5 10 0 11 0 95 75 3 5 16.5 70 70 10 0 10.0 2.0 9.0  °Tot°135 5 107 5  930  127  107  74  18.6° 25.5  21.5  14.8  75 0°  104  83  58°  °  20.8  16.5  11.5°  0 0  0  ° ° 0 0 0 0  ° N °POST- POST- POST °CL3°TEST TEST TEST ° °A11 Erra Inter  (Erra = intra- + 4.0° inter— °TM1° 8 5 6.0° sentential °MM1° 21 0 7.0°word blanks) °TM2° 21 0 ° 0 8.0° 22 5 6.50 °MW° 165 2 00 °GST° 5 5 10 00 °YK ° 27.0 55° °!012° 120 6 0° °HG ° 19 0 7.00 °HO ° 19 0 6.0° °!4I° 140 2.00 °TI° 55 5.0° °NN 0 18.0  °Mn° °  4 0 9 7 19 0 9 7 9 7 13.0 19.0 14 0 9 7 17 0 15 0 16 0 9 7  4 0 8 3 15 0 8 3 8 3 10.0 17.0 11 0 8 3 15 0 11 0 15 0 8 3  °TM2° ‘HN • °MW ° °GST° °YK ° °1012° °HG ° °HO ° °MI 0 °TI ° • °  °TMI° °MM1° °TM2° °HN° °MW° °GST° °YK ° °MM2° °HG° °HO ° °MI ° °TI 0 °NN °  6 0° 12 0° 7 0° 11 00 7 0° 7.0° 15.0° 10 0° 12 0° 7 00 7 0° 7.0° 13 0°  5 5 13 0 17 0 17 5 125 3 5 25 5 100 14 0 15 0 100 30 14.0  5 0 8 0 0 12 0 15 0 95° 2 0 15 0 85° 12 0 ° 11 0 60° 30° 10.0  °Tot°226 0 160 5 117 0 °Mn° 174  123  34.8  24.6  °  90° 18.0  0  0  •  a  a  o  •  •  a  H  a  La)  a  U  a  H 0)  -J  ‘.0  H  a  La)  J  H La)  0  o  0)  -J  H U  I-I  0  0  Lii  0’ •  •  (43  01  0  a  (43  0  0  I.  I WIM  o  Ui  a  o  —1 •  ‘.0  w  •  a  > 1%)  a  > M  a  0  >1> WIJ 0IW  *  -  La) La) Ui  La) .) 0)  a  IW  I-J 13  UI  La)  a  o  0  0  a  Lii  ‘-3  ‘-3 0’  0  0  a  o  o  It  (I)  o  0  a  0  •  a  H —1  a  H 0)  0)  H  I-’ 0’  a  iU  H U  0  ‘.0  •  0  fl  0  .  •  t3  > 1%.)  a  La) M t.J  0  I-’ 01  a  W  ‘-3  0  a  I-’  ,-.‘  H U  0  -J  •  (43  0  .  a  —I  t4.)  >  a  La) La) t.)  a  1-’ Ui  0  .  ‘-3  0  a  ‘.0  H  •  U  J  I—’  H  0  P.3  •  w  0  0  •  ê-’ 0)  >  a  La) La) La)  0  e  I-’  a  1)  ‘-3  0  I-b 0  0’  It  t-1  s.Q  I-’  Ill  ‘d  ‘1 0  U)  i-’-  It  0  1  (D  ° 0 II  •U)  I CD  0  ft  ‘101  It (t 0  ow  (DI  (t  >  o  -J  w  0  I-’ Ui  a  1  0  I-  0  a  a  •  *  a  a  a  0  (4.)  ‘.0  H  a  La)  W  H  a  0  I (43IH -ii-. •I• 0101 I  0  i  aiw  a  a  0  0  >1> Wlla OI’.O • I•  o  a  0  0  a  a  0  a  a  0  Cn  a  a  a  a  a  I  Tj  H  a  Cl)  )J  a  a  Cfl  Z  •  tJ 0 Z  0  a  II  U 1 t  0  a  0  0  a  0  0  o  0  0  0  •  0  1-3  0  U a  OH  0I  o  a  0 ‘-3t I1 M 0 Z HIt ‘-3  Zc  LTJ  WI-.M  ‘-s’-<’o  I-i’ 01  OH  0  It’iO*01o QIXPI O oi 010W WM ‘dIrt i-i-I  0  0  ‘dIrt  i-ti  a  a  t’ ‘-3WM0t!i  a  a  HP’J  a  a  QIXPI 0101 0IC)H  a  0  a  a  ‘ILi1O  a  0  a  0  Cl)  WM  MO  H  tM  00  T.1’  co  _rj 0  30  15””  5..  0’ PRE—  .  -V  -  S  12  .  I  N  ..  -  -  =  14  13  -.  I•  15  I  16  -____  All 50 Blanks  - -  I  1.3  tests (fix classes 1 and 3)  -  I  -  -  V  .  POST—  3  149 TABLE 21 FOR THE MAIN PROJECT PRODUCTIVE CONFIDENCE SCORES FOR ALL 50 CLOZE BLANKS PCS °TEST°LEVEL NO NO TOTAL TOTAL°% OF TOTAL COR o C C CBLANKSC NO OF OF • OF ° NO NO • RECT° o C CFILLEDC OF °READ-°BLANKS°STUD-° OF OF °BLANKS° C C °ABIL-° PER °ENTS .BLANKSeBLANKSC TO °BLANKS° TO C C e C CITY CFILDCPOS S_ • TEST IN FILLED° NO ° C C ° • C °EACH IBLE °COROF C 0 C C °TEST 0 °RECTLY°FILLED° o 0 C 0 C C • In % C C C In % C C C C C C • ° AxB= D/C= F/D= C C C • C C CD CE °A B C • G • °Grl C CPre_C 15 C C3 C C  A  C  50  C  CGr1  C  C  C  C  C  0  CGr3  C  o T3 C C o T4 C C  14  °Grl 15  13  0  °  13  C  °Gr3  C  C °Grl • 15  0  °Grl 13  650  for  °  50 13 13  C  • 353  0  this  550  C  group  543 is  344  CA62  309  C  C  650  C  650  C  650  C  650  C  CA496  CA76  C  0  C  367  C  C  514  CA79  550  C  13  C  650  CA478  CA73  C  C  C  • e T6  °Grl • 12  C  CGr3  C  CGr1  C  •  0  0  C  50  7  C  °  C  650 350  C  C  8  C  13  50  15  C  C  13  13  than  395  C  71 8  C  397  C  C  275  CA78  6  C  C  650  CA432  0A66  C  C  379  63 5  C  C C  C  5  for  26 7 8  29 4  •  CA3O  0  C  178  CA34  6  C  C  128  C  32 4  CA18O  C  37 7  C  C  C  169  C  115  C C  CA42  6  41 8  117  CA154  C  C  C  C  C  C  110  C  C  CA46  C  5C  C  58.3  C  0A27  C  °  32.2  86  o  C  • 254  C  5  61 1  400  650  1  C  C  CA146 C  2  92  C  0  C  3  C  C  C C  113.5C  higher  56 5  11  C  C  C  C  50  C  47•5  C  C  0A33  C  C  CGr3  •A129 V  5  C  50  7  C  C 0  CA59  C  C  CGr3  °  CA388 C  C  °  CpostC  650  C  °  C  CGr3  50  11  C  o T5 C  0  13  that the score the other group.  C T2  ° C  means  C  C  13  C  1  35 9  C  135.5  C C  35.8  0 In H  a:’  b C) C -•  .  T2  All 50 Blanks  1.3  14  .  .  15  .  16  POST—  N=13 (T2N[=ii, T4N2=11, T6N1=?,N2=8) 600  500  ..jL)L)  200  PRE—  tests (for groups 1 i:Jfld 3)  CLASS 1  7  rI  w  L  J  -  . .  JI.  R$  -  Ij  T2  M[il. 30 Blanks  T.3  —[  .  14  .  I RSi I  LHI  F  T6  —•  •1  T5  —_____  POST—  .  1  N=13 (T2NI=ii, r1l4N)Zii rp(3NI.7Np.8) I-)  I-si  PRE—  tests (for clusses 1 and 3)  CLL\SS3  I  0  Ci  0  0  g 0  -—  T2  All 50 Blanks  I  T3  I  14  I  -  T5  I  16  I  POST—  -  N=13 (T2NI=11, T4N2=11, T6NI=7,N2=8) 200  •1 L(’ i  i00”  50----  0 PRE---  tests (for classes 1 and 3)  CLASS 1  CL/SS3  If) ‘-I  0  ..- -“  0 0  j_J  Ii  T2  AU 50 Blanks  .  T.3’  T4  I  T5  113  POST—  F  N=13 (T2NI=ii, T4N2=11, T6NI=7N2—8)  40  2O” J0  0  PRE—  tests (for classes 1 and 3)  CLASS 1  CLASS .3  o  0  0  o  o  o  ‘-3 Ui  ‘-3  H W  H  H  (3  0  a  0  o  0  o.()  I  0  H  H  a  W  a  0)  -.1  H W  H  •  a  UI  0  OIUI  a  •  I •I•  o  F)  I  •  OI’.O  H 0)  -  H  a  W  H  H W  a  I3  •  tJ W  0  ‘.0  . .  > t.3  >  op.  0ID  0  -  0  W W UI  W I...) 0  0  0  0  0  Ui  0  O•  0  0  0  r1  o w  0  >1>  o  o 0  0  H -.1  Os  H  H W  a  H  •  H 0)  0  UI  •  > J t’J  0  W W t3  0  H UI  0  i  0  H 0)  0)  H  a  W  H  H W  a  H  •  H Ui  0  0)  •  > H ‘.0  0  W 3 t.3  0  Ui  H  ‘-31-3  0  0  ‘.0  H  .  H  a  W  H  H H  a  Os  •  0  H  0  I-b  ID  rI  0  I-’.  (1)  ‘ti  o  ‘.Q  in  I-’-  CF  0 ‘1 0  I-b  Oi  0  F  0 0  •11)  rF Ob $0 d  CF  CF Ct (D  0(I)  0W  (t  >  •  > H UI  0  W W W  0  .  H  0  t.3  ‘-3  0  I  0  -J  W  0  I—’ UI  0  0  0  I•  a  0  0  0  •  I  a.()  0  0  1%.)  H ‘.0  0  W  H  H L.3  a  OIFJ  0  I I HIH JIW • t•  0  t3I)  •  MIU1  >1>  0  0  0  0  MIH  0  0  0  0  0  0  0  0  0  0  0  0  0  0 0  0  ‘-3  ‘-3  0  0  0  0  0  t-’  tn  0  ((3  0  • 0  ‘J0  0  0 0Z  0  II  <F-’xi<  F-44  0  0  0  0  0  o  0  0  o.()  0  0  a  0  0  w  0  1<  OH  ID  HrF  0 Ii 0  0  0  t-3 (1)  Z  [uj  tEl t-3 Zc  CO  Z  ‘  w H  t-’  H  1-3  •  U)00)  ID  wi-a-Lu  1-3 I—I  1-3 Lii  0  GH ti’<’0 m’nl  0  H.,.  0) Lii  I  WtI)  0 (tJ tiLl  I—I Z .3  0  “i  tn  0  fl  ((3  Z  tn  Z 0  i’:z’  GH  0  0IX 01W olow ‘IrF rIP  ItxiØ*I-Ifl  0  H ‘IMC 0I<Ii rFtTi 01W Ii OIOF-’ ‘Ir1HW (F1  0  0  0  0  H VI  Ji  0  T2  1.3  -  15  N=13  T4  •  T6  POST—  •  (I  i_i AC  ]{ntra— and Int..er sentential Blanks -2;fl  —  LU  :;  PRE—  tests (fc..ir classes 1 and 3)  156 TABLE 23 FOR THE MAIN PROJECT PRODUCTIVE CONFIDENCE SCORES FOR INTRA- AND INTERSENTENTIAL BLANKS PCs °TEST° LEVEL° o 0 OF  0 0 NO NO TOTAL° TOTAL°% OF °TOTAL °COR0 C OF OF NO. ° NO °BLANKS° NO RECT° 0 °READ-°INTRA-°STUD-° OF OF °FILLED° OF °BLANKS° ° °ABIL—°INTER-°ENTS °INTRA-°INTRA-° TO °BLANKS° TO 0 ITY 0 °SENTIN °INTER-°INTER-°POSS- °FILLED° NO 0 ° C °ENTIAL°EACH °BLANKS°BLANKS° IBLE °COROF 0 °BLANKS°TEST °FILLED° ‘RECTLY°FILLED° 0 0 0 0 C 0 0 In % In % • 0 0 0 0 0 0 0 AxB= D/C= F/D= 0 C ° 0 A B° C • D °E F • G 0  0  0  o 0  0  0  0 0 0 0  °Gr]. °Pre— 15 0 °Gr3 A  T2  0 0  •  T3  °  • T4 ° 0  0 °  13 40  0  0  °  °Grl 14 °Gr3  0  °  0  0  °Gr1 15 °Gr3 °Grl C 15 °Gr3  °Grl T5 • 13 CGr3  °Grl T6 12 0 °Gr3 0  °Grl  °Post° 15 ° °Gr3  30  0  520  0  13  °  0  0  37  ° 0  35  °  11  0  C 0  °  0  0  C  0  520  •  0A318  0A61  0  0  0  • 52.7  0  0  274  2  0  A99  86  0  31 1  0  0  °“31.4  0  means that the score for this group is higher than for the other group.  0  0  °  13 13  0  •  13  0  13  0  390  • 481  •  282  • 455  •  371  0  0  0  • 282  0  0A380  0A79  0  0  0  13  0  481  ?39  0 *  °  °  °  481 273  0  8  °  13  ° 0  0  7  40  312  520 °  0  316  0  213  0  0  13  50 3  0A79  0  0  °  0  0  0  196  0A381  c 37  520  °  7  2  •  0A81  69  0  0A78  1  0A129  0  33 9  0  98 146  •  0A34  8  0A39  4  0  106  37 6  0A163  42 9  0  * 0  0  0A478  0  °  °  151 93  °  °  43 7  0  98  CA49  2  9  0A673  CA129  •A36  0  0  0  • 59.2  107.5°  0 0  0A350  308  0 0  0 C  0  0  0  63 8  0  0  C  73.2  0  4  0  35 2  5  65.7  •  0A36  80  0  58 6 •  0  199  0 0  0  385  °  •  481  0  C  0A66 0  0  °  11  13  220  C  0  0  0  330  0  0 C  349  I— In  t1I  -  -  -_____  -  .  POST—  CLASS.3  LLI-kcc 1  Int.ra- and Inter—sententjai Blanks  -  N=13 (T2NI=ii, T4N2=11, T6NI=7,N2=3) 400—  _‘f.’  )uU  --“  200--”  1IY)  PRE—  T3 T4 T5 T6 tests (for classes 1 and 3)  ‘  UI  (/‘Cfl  (j)  UI  cj o 00  II  C\1  z  WIL’/IIII//I!Ii/!’/iIIIIJI//ifII/III/L///I  -‘  H  r• .•.  W?iifflJJiflZlZJI7ZlIIij I I  -:  -  H  J Li  ,  7/Iflh7i’Vi/ii))1/ill/Z7Z’1ii 1 L -  tjj  1  N  Li) H-  C 0 (i) jI \  -  C)  H =:f=; —  (J3 LI  z  .  -  .I,--  -fl  77il//,’///////,’1jjJj/Jli//z1//,)//J/,llj,  LU 0  z  C) C)  C)  C)  C) -i  :E:ZE.Ej (1  C)  IjF.i.’1 IC  (0  LLiC:_i.._  89t  0•i In I-I  U I)  -D  E _s  (T2NI=ii,  13  T5  T4N2=11,  T4  T6N1=7,N2=8)  POST—  zz: zzz 16  ‘LF4  .IAC  3  7  II  CUSS  Intra— and Int..er— slente ntial Blanks N=13  Q’ I L)L)  1  16C on  •140 i  100 80”  T2  zzJz zz PRE—  tests (for classes 1 and 3)  0 I’D I-I  ‘  U  C...)  C  : 13  T5  .  16  ....  POST—  içc  CLASS .3  C1  Intra— and Inter- sentential Blanks  •  T2  14  N=13 (T2NI=i1, T4N2=11, T6N1=7,N2=8)  3Q....  20””  •  PRE—  tests (far classes 1 and 3)  a  a  a  0  o  0  o  0  H  I-’  a  (A)  a  0  (A)  H  (0  a  0  IH  H W  0  0)  H  H ..J  H -  01  H  H (A)  H W  o  0  01  •  0  J  0  (A)  (0  UlI  I  •  •  •I.  0  t.J  )IH  0  0  > (0  >1>  o  I-Io I  •  Ui  01  0  •  0  •  0  t%)  (A)  (A)  0  H Ui  01  0  (A) (A) Ui  0  H W  0  ‘-3  (DIm  0  o  O  Li)  t#) t.j  0  0  H (%)  I-’ Ui  0  Ui  ‘-3  0  >  a  o  0  0  01  ‘-3  0  >  0  o  0 U)  0  I I  o  o  0  0)  H  0  (0  H  H  a  W  H H (A)  H  H  0  -.J  •  UI  0  0)  •  > 01  0  (A) (.) (A)  0  H  0  H (A)  0  H  •  (31  0  (0  •  > 01  (A) M t%)  0  H Ui  0  ‘-3  0  0  0  0  0  0  Li) I’)  0  H 01  0  I•  I  I  0  0  0 Ii  0 0  0  t%) 0  0  0  0  0  0  0  0  0  til 0  t-  0  H  0  0  a  0  0  0  rtI  0  0  0  0  •  0  Z  0  G  0  lj  0  (A) 0  0  0  Co  ‘-3  t4 Z  ‘-3I  Ijl_3  WCflti)  WJ’ij  0  ‘CI) U)  HO  ‘1’dO  H”  lj  GH  0  OIXli DIP) 010W  ILG*Wfl  0IXII DIP) CDl0Ili’ IrI HP)Z (ti ZI 0 0 0 Cl)  0  xjo Cfl  0  0 dOZ  0  II  t  F-4P’ -‘ I-3WLTiOLTj  0  GH  H (0  a  W  H  (A)  H  0  0  0  0  I[ZjØ  0  0  0  1b  0  a  I  0  Hit  0  a  0  011H  •  0  0  ‘-3  0  CD  0  ‘tJ  0  ‘-3 ti  0  I >1> UiI • I• 0)1(0 I  0  0  0  0  W  it  0  I-’-  U)  o  I-• U)  rt-  0  CD  0 0  •U)  Ø  O  lirt  r1  liP)  OW r1 rt  >  HO  tj  ZH  z p31-3  0 H  o  ITj  01 I-’  H  rj  lb  C.-) 0)  12 1L)  B  T.3  T5  N=13  T4  T6  POST—  Intra—sent..en[ial iBlanks  T2  : 1 ELLLi PRE—  tests (far classes 1 and 3)  ‘m  C’I ACC LLJ-JJ  7  3  163 TABLE 25 FOR THE MAIN PROJECT PRODUCTIVE CONFIDENCE SCORES FOR INTRA-SENTENTIAL BLANKS PCS 0 0 0 °TEST° LEVEL 0 NO NO TOTAL° TOTAL°% OF °TOTAL COR0 0 0 ° ° 0 OF OF OF NO NO BLANKS NO 0 RECT° 0 0 °READ-°INTRA-°STUD-° OF OF °FILLED° OF °BLANKS° 0 °ABIL-° SENT- °ENTS °INTRA-°INTRA--° TO °BLANKS° TO 0 0 °ITY °ENTIAL° IN °BLANKS°BLANKS°POSS- 0 F ILLED NO 0 0 0 0 °BLANKS°EACH 0 °FILLED° IBLE °COROF 0 0 • PER 0 0 °TEST 0 °RECTLY°FILLED 0 0 0 0 0 0 0 0 0 0 TEST 0 0 0 0 0 0 0 0 0 In % 0 0 0 0 0 0 AxB= e D/C= 0 F/D= o o 0 0 0 ° 0 0! A B C D F 0 G 0  0  °Grl °Pre—° 15 0 °Gr3 A  T2  °  ° 0  0  T3  e T4 0  0  °  0  9  13  °  0  °  °Grl 14 °Gr3  °  T6  0  0  °Grl 15 °Gr3  0  °Grl 15 °Gr3  °Grl 12 °Gr3 0  °Grl °Post 15 0 °Gr3  8  °  117  0  0  13  0  A9j  0A77  0  117  0  70  011  0  0  0  0  13  °  104  °  74  13  0  130  0  A95  0  0  0  °Grl T5 13 °Gr3  °  °  0  8  0A32  0  0  °  0  59.8  0  5  35 7  0  29.0  0  0A414  means that the score for this group is higher than for the other group.  ° 0  °  0  10  0  •  °13 6  0  0  7 °  0  91  °  ‘63  013  °  91  0  70  0  72  °  117  0  13  0  54  0  49 48  0  “90  0  117  50 5  0A47  4 0  04710  °440° 0  06670  33  °‘69 2  °  35  0  0  0  0A7500 °  55 6  0  05930  30  05560  0A7780  21  °429°  667°  30  •A6250  0  “34  0  076  9  0  79  0  0  0  0  °  0  0  0  0  °  °  84  0  0  13  0 0  0  0  A45  37  0  0  °  0A9550  78  0  °  1  33  44  °  0A73  °130  0  8  71 2  0  05380  66  •  37  0  0  070  0  7  °  0  0  11  °  0A5730  0  0  0  13  43  0A8520  0  0  e  °  75  °13  °  9  °  0  0  10  88  °  0  67.5  0  0  0  0  37 8  0  32.5°”41.1  0 0  1)  -o n  n  Ifntr a— sen ten tial Blanks  I  I  V  I  I  .  POST—  N=U3 (T2NI=1I, T4N2=11, T6NI=7,N2=E3) 100  20”” 0 PRE—  12 T3 14 15 16 tests (for classes 1 and 3)  CLASS 1  (‘I ACC  I  0’, Li  r’4J1.’  £ S9[10  91  ci  PLJr: .  fri  sssL)I.:) r  Jcij) ssq  00  ifl  C’9  Crfr  0  S)1Efl]  lR.!1.U[ -Y41U[ S _1EUI1TJ[1[  (9ZN’ZTNaL 1 ‘TTz:Nf’ ‘TT=TNEa) TN J  —-ISC’.d  (c  I—.  Cr  if;’  ()  0  CL  U)  cr  —  ()  ‘I:’  -1: (t  1.0  o a-)  a-) (I)  ‘-4-  Intrasententia1 Blanks  •  •  T2  1.3  ...  14  —  15  b: 16  •  POST—  S  N=13 (T2N1=I1, T4N2=i1, T6N1=7,N2=8) 50 4C 30  IC  PRE—  tests (for ciass 1 and 3)  CLJ3 1  CLI6S 3  ,..  C.  ‘4—  Int..ra—sententja1 Blanks  T2  T.3  T4  -  .  ...  16  [il T5  POST—  -—  N=13 (T2N1=ii, T4N2=11, T6NI=?,N2=E3) L)U  t)1.J  PRE—  tests (for classes 1 and 3)  (‘I ACC  CLASS 3  0  0  o  0  o  0  H  H  0  0  to  H  to  0  -J  H W  H to  0  0  0)  H  -j  H  0  H to  ,H tO  0  0  0  o  •  H  0  ‘.0  •  H  HIH  • I. 0101 I  0  0  0  Ui  G o  •  •  > H  0  (‘0  •1.  0  UI  W  W  0  0  0  o  W  W 03  0  0  W  H  >  0  o  1%)  H  H Ui  0  0 0  i-3 Ui  1-3  o  0  0  0  o  o  ft  rn  0  0  >1> t’JIH WI  o  o  0  -  H  H a,  0  H H  tO  H  0  01  •  H (‘3  0  0)  •  ai  > H  0  W W  0  01  )i  0  ,.  1-3  0  o  H 0)  H  0  H to  H tO  e  0  •  H  0  ‘.0  •  > H  0  H 01  0  w  ‘.0  H  H  0  H to  H H  0  ‘.0  •  0  -.3  •  > -J  0  W W w  0  .  H  0  r.i  0  )-3’-3  0  0  0  0  0  0  W  0  01  H  0  Ii 0 I  0  0  0  0  0  0  0  Ii  I-b 0  ft  0 ‘1  I-’-  0  0  0  (‘3 0  H to  0  I-’ to  Cl) 0  ‘-3  0  t’  1:1  XJOZ  0  0  H  0  ftti (D Hli  0  ITjo Cl).  0  0  II  0  0  0  e)  0  0  0  0  ØH  0  0  0  0  •  0  0  to 0  G)H  1  a Hft  G  0  0  Cl)  ‘-3  ‘-3t) 1:1  U) Cl) Cl) 1:11-3 Zcl  li”’O  1  H to  0  0  ‘-3  I<Hxj<  w  0  I  0  0  0 0 0 (1) P’IMG)dPW() OiXli HO aip, 0(0W ‘dIrt ‘Cl) rti W  ftl  0  0  ‘-3Wt40111  0  0  Hi  0  0  ‘dirt  0  0  0  0  ‘IMØ OIXli Qip, OIC)H  0  0  0  0  H’  0  •  (‘310)  I HI wi •I•  0  >1> HIH Ol0 •I• IH I  0  0  0  0  I-,.  ‘d  0  i-’U)  ft  H) 0 Ii  Ii 0  O fØ  ‘j  ft  O  ft  ft  ow  (Dp,  (1•j 0  >  C.f)  t’tl  HO  Fl  Z H  :I’xi  0)  0  I-’  j *  1)  çD U)  25  N.= 13  Inter—sentential Blanks  V  12  13  V  14  15  16  POST—  .  20• Li  1G  PRE—  tests (for classes 1 and 3)  I  _  ______  c-i cç  170 TABLE 27 FOR THE MAIN PROJECT PRODUCTIVE CONFIDENCE SCORES FOR INTER-SENTENTIAL BLANKS PCS 0 ° o °TEST°LEVEL° NO NO TOTAL° TOTAL°% OF °TOTAL °CORo ° ° 0 OF OF ° OF NO ° NO °BLANKS° NO • RECT° • 0 °READ-°INTER-°STUD-° OF OF °FILLED° OF °BLANKS° o • °ABIL-°SENT °ENTS °INTER-°INTER-° TO °BLANKS° TO 0 • °ITY °ENTIAL° IN °BLANKS°BLANKS°POSS- °FILLED° NO o ° 0 °BLANKS°EACH ° °FILLED° IBLE °COROF o 0 ° 0 PER °TEST °RECTLY°FILLED° 0 0 0 0 0 0 0 TEST 0 o o 0 0 0 0 • In % ° ° In % • o o 0 0 °AxB=° °D/C=° o a 0 0 0 ° 0 •A B c D E °F G  o  0 °Grl °Pre—° 15 • ° • °Gr3  A  T2  0 0  T3  ° 0  ° °  T4  °Grl 14 °Gr3 0  °Grl 15 °Gr3 0  °Grl T5 0 13 °Gr3  °Grl T6 12 0 °Gr3  0  31  0  0  °  °  0  0  °Grl °Post° 15 ° °Gr3  22  0  0 °Grl 0 • 15 0 °Gr3  °  0  13 13  • 403  0A227  °‘56 3  0  0  0  •  403  0  204  ‘65 5°’28 9 0  • 50.6 •  0  57  °  27.9  means that the score for this group is higher than for the other group.  °  °  ° °  °  27  °  °  0  0  13 13  °  0 0  30  °  0 0  ?29  °  0  °  13 13  0A81  0  0 0  0  351  • 212  0  377  °  0  390  0A310  079  0  0  0  7  °  203  0  0 °  0  291  0  164  0  13 13  °  0  232  0  403  8  0  238  °  5  60 4  0A78  0  0  74 6 5  74 6  0A80  8  0  151  •  °  32  0  A84  °  2  29.4  65  0A30  7  0  109  0A36  7  0  0  0  69  0  29 0 41 0  0A127  0 0  • 121  0A41  72  0  6  43 9  0  0A45  0  °  A95  0A36  5  064  • 403  0  • 56.8 •  0  0  75  0  0  0  0 0  68  0  0  0  °  0  0  0  0  0  0  0  °  0A260  5  0  0A26  0  0  65 1  • 25 5 0  0  °  229  37  0  319  0  8  42 7  0  390  °  297  0  0  9  0  0A286  °  °  0  351  13  0  0A59  122  0  13  • 145 0  286  0  °  31  242  0  °  0  0 0  0 0  29  11  0  0  32.8  0  I-I  QJ  .  inter—sentential Blanks  I  13  I  T4  :  I  15  I  16  I  POST—  .  -  N=13 (T2NI=Ii, T4N2=i1, T6N1=7,N2=8)  3ft)  t2  25G 200”” 15C  luO 5C C  PRE—  tests (for c.Iasses 1 and 3)  ULAS 1  CLASS 3  172  pr-i-t of blanks fiiled to posble  -  0  S.-.—  CD  Cr)  C)  CD  C)  -fl El  ‘  . 1 I  .,L.  L)  ‘AL.&LL4J.LJ  z  -  (D  H  vj -H  /II//I/I/II,iIII/I/////h,i////Il/iiii,,’/Iii fiiIi  H  ff!ul/IJ/J1ffJ//fItuIlIIJIiIullllffJ/4777  z  1  i),,  L.LLA4..LI.A.LLL4  1177/,lfl)  (.0  o  D  fl  1 —1 C)]  I............ TI!Jl!!II!JIlfJIIIII!!IlfJ!if !1A’J’J!JfIJf!1JJfJlIIIfi  H H  4Fø44c4.r”  z  -  CD  CJJ  H  -a  U  -  YffL1I1A  II Un  !I!I  CO  H  (•1J  Inter-sententiai Blanks  T3  14  15  POST—  ii 16  N=13 (T2N1=I1, T4N2=11, T6NI=?,N2=B) 140 ILL!  100  80  PRE—  tests (für classes 1 and 3)  I  CSS 3  ‘ro  0  ci -U  Inter—sentential Blanks  12  .  13  .  14  15  ........  113  POST—  N=13 (T2NI=11, T4N2=11, T6N1=?,N2=8)  40  30  20””j  PRE—  tests (for classes 1 and 3)  CLASS 1  CLASS 3  175  An explanation of the tables and graphs The above graphs are all very similar and show that the control group did better on the tests than the treatment group.  If this were the only fact given, then it could be  concluded  that  the  rationalized  CCC  had  no  significant  effect on the treatment group over and above the effect of the non—color-coded doze on the control group and therefore the  treatment  would  have  to  be  considered  a  failure.  However, another look at the graphs appears to indicate that there was a tendency for the treatment group to catch up to the control group as long as the crutch of color—coding was available.  As  color—coding,  the  the  only difference  acceleration  caused by the color-coding, themselves.  can  in the be  seen  tests was to  have  the been  if not by the parts of speech  Therefore, it can be said that the color—coding  had reached the objective of the research, that is, to helD iluDrove doze scores and to increase the students’ awareness of inter—sentential reiterative—word clues. A closer look When  the  inter-sentential  blanks  were  considered  by  themselves it was seen that both groups had made progress. Interestingly, increase clues.  in  the nuiber  However,  increases of  approximately  inter—sentential  followed  the  reiterative—word  as long as the color—coding was present,  the treatment group tended to improve more than the control  176  group  in  being  able  locate  to  and  use  reiterative—word  inter—sentential clues. Part of that greater improvement can be explained by the  fact  that  proportionally first  much  practice  rebound.  the  treatment  worse  test,  Actually,  than  giving  group  the  the  seemed  control  group  treatment  all else being equal,  more  to  do  on  the  room  to  there should not  have been such a difference in test two, given that the pre test  showed  the  two  groups  as  being  much more  equal  proficiency, at least in terms of doing doze tests.  in  It can  be assumed that the color—coding was the cause because it added information that the students had to consider, adding  an  encumbrance,  if  not  confusion.  Then,  thus  as  the  students in the treatment group became more familiar with the  color-coding they were  able to  narrow the difference  between themselves and those students in the control group. This were  suggests not  confident  that those  only and  becoming able  in  students less the  use  in the  confused, of  the  treatment group but  also  more  color-coding  to  discover the inter—sentential reiterative words to fill in the doze blanks. the  greater  Thus it was thought that another part of  improvement  of  the  treatment  group  could  be  attributed to the color-coding of the rational doze. Four additional points should be made the graphs.  First,  from examining  the main project showed that to  some  extent students’ were able to locate inter—sentential clues, with or without the color-coding as a crutch.  This seemed  177  to validate Chihara and Oiler’s contention that proficient ESL students do search globally for clues and do use them. Second, both groups tended to improve scores on all blank, intra—/inter—sentential,  and  inter-sentential  analyses  on  both out—of—50 mean scores and percentage—of— possible mean scores.  This suggests that practice with the doze did and  does lead to improvement. to that improvement.  Third, there was a ceiling effect  Fourth, despite the evidence for the  valuable role of color-coding,  it is difficult to say that  the students were using the parts of speech to help find the clues as to what word to put in the doze blank.  Because  the gap between the control and the treatment group appeared to widen  in the post—test on inter—sentential blanks when  compared  to  the  pre—test  color—coding  had  not  sensitive to parts  difference,  done  of  as  hand,  without  to  seems  make  that  the  the  students  speech as had the practice by the  control group with no special other  much  it  the  color—coded clues.  color—coding  as  treatment group still did seem to improve.  a  On the  crutch,  the  The improvement  must have come from the same source as did the improvement of the control group.  As the pre—test and the practice test  were the same, part of the improvement could have come from having  done  the  test  before.  However,  some  of  the  improvement for both groups could have come from the parts of  speech,  coding. been  an  with  or  without  the  influence  of  the  color  On the other hand, some of the improvement may have illusion  as  some  of  the  tests  had  a  lower  178  readability level.  Table 9 shows that the readability of  tests 2, 5, and 6 are easier than the rest and in some cases the graphs rise in an apparent reflection of this.  However,  this conclusion is rather tenuous as others of the graphs do not behave in this fashion, possibly due to other influences such as boredom as in test 6, the passage with the easiest readability level.  See test 6 on the All 50 Blanks graph  where both groups fall, the decrease occurring primarily on the inter-sentential blanks. To summarize the  above points  it could be  both forms of the rational doze procedure, not,  were  useful  as  a  function  of  said that  color—coded or  practice.  It  is  interesting and necessary to note, however, that despite all the increases, for  doze  the mean scores were still quite low,  scores.  frustration  No  level.  mean  This,  score  of  ever  course,  went  could  beyond have  even the  been  a  result of the difficulty of the passages. There  is  special attention. analysis  to  use  another observation that deserves  some  It was an afterthought in the research the  intra—sentential  scores.  This  was  especially so because the focus of the research had been on inter—sentential clues rather than on intra—sentential ones. It was also because it was easier to get total scores on the spreadsheet  by  first  deleting  the  non-reiterative-word  blanks and then the intra-sentential blanks.  It would have  taken another step to have gone back to the intra-/ inter sentential copy of each of the spreadsheets and deleted the  179  inter—sentential blanks. was  done  later  Instead, an arithmetic calculation  subtracting  by  the  inter—sentential  scores from the intra—/inter—sentential mean scores,  mean and a  graph drawn. The intra-sentential graph showed that the mean scores were not consistent in improvement. test both groups  improved on the  but then declined. the  On the first practice intra-sentential blanks,  By test four (the third practice test)  treatment group  started to  improve while the  control  declined until test five when the mean scores were about the same.  On test six both made gains, but the treatment group  made more.  Both  declined  on  the  post—test,  the  control  group taking the lead once more. In terms  of the number of available  intra—sentential  blanks the above can be explained in the following way.  For  test two both groups improved when there were less intra sentential clues and for test three declined when there were more. ups  After that the control group followed the downs and in  the  number  treatment group  of  intra-sentential  continued to  last practice test.  clues,  but  improve until  test  while  the  six,  the  The treatment group was not affected  very much by the decline in word  blanks,  did,  like  intra-sentential the  control  reiterative group,  react  positively to the presence of the greater number of intra sentential post-test  reiterative-word clues, when  no  color—coding  even more  was  present  so. both  In the groups  declined, with the treatment group falling the most and even  180  losing  its  advantage  over  the  control  group.  It  is  important to note that on the post-test both groups still had made improvements when compared with what they had done on the  pre-test.  intra—sentential  From the blanks,  above  it  can  observations  be  concluded  about  that  the  color-  coding appeared to have even a more powerful influence on finding  intra-sentential  fillers  than  on  finding  inter-  sentential fillers but there were not enough occurrences of intra—sentential blanks in the passages to be certain, given the  somewhat  means.  inconsistent  nature  of  the  intra—sentential  It is important to emphasize here that despite the  influence of the color—coding on the intra—sentential scores this  does  not  negate  its  influence  on  inter—sentential  scores.  A statistical look at the data The amount However,  above of  description  merit  should  and  statistical  of  methods  validated over the years.  the  graphs  be  considered  have  been  has  a  certain  carefully.  designed  and  Without them any research can be  misleading, even though Dr. Boldt, the statistics advisor to this research project,  cautioned that statistics themselves  are only one of the tools in understanding the data.  When  T-tests  “One  Between  (reported and  in  Tables  28  and  One Repeated Measures  29  below)  Factor ANOVA”  and  analyses  (see Table 30 for the steps in the program and Tables 32 40 for the results)  were applied to the data,  -  a different  181  picture  from  the  graphs  dependent-variable  T-tests  treatment  groups  pre—test.  were  emerged.  First  of  indicated that the  not  significantly  all  the  control  different  and  on  the  The difference shown On the graph was merely a  chance difference according to the T—test.  Nor were they  significantly different on any of the practice tests except for Test 3, where the difference was in favor of the control group.  Even  on  the  post—test  there  was  difference indicated between the two groups. is  that  the  significant  rationalized improvement  no  significant  The conclusion  color—coded treatment  for  the  treatment  led  group  to no  over  the  control group on the doze scores. T—tests (above) done on class 1 comparing the pre-tests and post-tests  showed a  significant difference  that the practice had led to  improvement.  indicating  Meanwhile,  significant differences were found by the T-tests for class 3 from the pre-test to the post-test.  no  (above)  This latter  finding is contrary to what was interpreted from the graphs. (Somewhat  puzzling,  contradiction.  If  however, there  was  is no  a  seeming  significant  statistical difference  between class 1 and class 3 on both the pre— and post—tests, how could there be a significant difference between class 1 from pre- to post—test, but not one for class 3?  There must  have been some data error or a flaw in the interpretation of  the statistical findings. to  have  been  that  At first the answer was thought  dependent  variable  t—tests  were  done  throughout instead of using independent T-tests when the two  182  different groups were compared. statistical  program  called  mistake was corrected.  To check this assumption a  “Minitab”  was  used  and  the  However, the previous results were  confirmed. The problem was  probably  that  the  statistical  tests  lacked the power to pick up the subtleties in the data.  The  statistical  the  sample  tests  size  were  available  too to  conservative the  given  researcher  was  that so  small.  Likely there was a significant difference from the pre- to the post test for the treatment group. that  this  was  the  case.  The graphs suggested  Furthermore,  the  t-tests  had  indicated that the groups were not significantly different on the pre— and post—tests from each other and that there had  been  significant  Therefore, significant  logically  improvement speaking,  improvement  pre- to post-test.  for  the  for  there  the  control  must  treatment  have  group  group. been  from  a the  183 TABLE 28 T-TESTS FOR THE MAIN PROJECT :Pre-test (Exact-Word Scores) 1 vs Class 3 0  degrees of freedom  12  =  Mean Diff  SD Diff  T-score  1.192  4.294  1.001  1.000  4.619  •  P  = 0  ‘All 50 blanks  0.337  •  0 0  ‘Intra—/inter— ‘blanks  0.781  0.450 0  0  0  ‘Inter—blanks  :Test 2  0.654  3.555  0.663  0.520  :  (Exact-Word Scores) 1 vs Class 3  0  degrees of freedom  Mean Diff  0  SD Diff  0  =  12  T-score  P  = 0  ‘All 50 blanks  2.394  5.846  O  1.477  0.166  0 0  ‘Intra—/inter— °blanks  2.494  4.891  1.838  0.91  0  0  ‘Inter—blanks  1.408  3.330  0  1.524  0.153 0  3  :  (Exact-Word Scores) 1 vs Class 3  •  degrees of freedom  =  12  Mean Diff  SD Diff  2.769  3.940  2.534  0.026*  2.385  3.664  2.347  0.037*  1.462  2.757  1.911  0  T-score  P  = 0  ‘All 50 blanks •  0  ‘Intra—/inter— ‘blanks •  .  ‘Inter—blanks *  Significant at the .05 level  0.080  184  :Test 4  :  (Exact-Word Scores) 1 vs Class 3  0  degrees of freedom  Mean Diff  •  SD Diff  12  =  T-score  P  = .  All 50 blanks  2.348  4.533  1.867  0.086  •  0  lntra—/inter— °blanks  2.114  4.672  1.631  0.129  2.208  3.884  2.050  0.063  0  0  °Inter—blanks 0  0  T55t 5  :  (Exact-Word Scores) 1 vs Class 3  0  degrees of freedom  =  12  Mean Diff  SD Diff  T-score  0.769  4.850  0.572  0.578  0.846  5.129  0.595  0.563  P  =  0  0  °All 50 blanks  0  0  °Intra—/inter— °blanks  a  0  0  °Inter—blanks  0.462  4.196  0.39.7  0.699  O  0  :  6 (Exact—Word Scores) 1 vs Class 3  degrees of freedom  Mean Diff  o o  SD Diff  =  12  T-score  P  = 0  °All 50 blanks  1.897  7.300  0.937  0.367  1.382  6.277  0.794  0.443  •  0  1ntra—/inter— °blanks  0 0  0  0  lnter—blanks  1.204  5.134  0.845  0.414 0  185  :Post_test (Exact-Word Scores) 1 vs Class 3  0  degrees of freedom  12  =  V 0  •  Mean Diff  SD Diff  T-score  1.462  5.387  0.978  9  0  °All 50 blanks  0347  V  V  lntra—/inter—  1.654  5.398  1.105  0291  0  V 9  Inter—blanks a  1.538  3.761  1.475  0.166  :d1a 1 (Exact—Word Scores)  0 0  vs Post—test  degrees of freedom  =  12  V V  •  Mean Diff  SD Diff  T—Ratios  P= •  °All 50 blanks  —1.962  2.883  0.030*  —2.453  V  0  lntra—/inter °blanks  —2.308  2.983  0 016*  —2.789  0  0  0  • Inter-blanks *  —2.269  1.867  0.001*  —4.383  Significant at the .05 level  :d1a55 3  (Exact—Word Scores)  0 0  :Pre_test vs Post-test  degrees of freedom  =  12  0 V  •  Mean Diff  SD Duff  0  T-Ratios 0  50 blanks  —1.923  4,476  —1.549  0.147 a  lntra—/inter— • blanks  —1.654  3.815  —1.563  0144 0  —1.385 V  •2 .966  —1.683  0.118 0  186  TABLE 29 SIGNIFICANT DIFFERENCES (As determined by t-tests) P < .05 Ni PRE-TT: Class 1 vs Class 3  CLASS 1: Pre-test vs Post-test  ALL  ERRA  not sign.  not sign.  ALL  ERRA  . •  °Exact—word  °  0  13  N3  13  =  TNTEI  0  not sign.  a  INTER  0  °Exact-word  °significant significant  °CLASS 3: Pre-test vs Post-test  •  ALL  ERRA  not sign.  not sign.  ALL  ERRA  not sign.  not sign.  significant INTER  .  °Exact—word °POST-TEST: Class 1 vs Class3  0  °Exact-word  •  not sign. INTER  not sign.  For the practice tests a “One Between and One Repeated Measures  Factor”  ANOVA  was  used.  This  ANOVA  is  a  more  sensitive tool than the T—test and was warranted because the T-tests on the pre-tests groups  as  determine  not what  showed the treatment and control  significantly effects  interaction of the two,  different.  It  was  used  to  and  an  if any, had on the doze scores.  It  the  subjects,  the  tests,  was also used to analyze the trend across the five practice tests. a) and  This ANOVA was applied to the data which considered  all the 50 doze blanks, inter-sentential  blanks,  b)  a combination of the intra and  C)  the  inter-sentential  187  blanks.  The ANOVA computer program is given  in Table  30  following and the ANOVA summary tables right after. TABLE 30 STATISTICAL PROGRAM: TREND ANALYSIS ONE BETWEEN AND ONE REPEATED MEASURES FACTOR ANOVA 01  • 2 04  5 07  10 °l1 °12 °13 0  014 015 016  17 ‘18 19 20 21 • °22 ‘23 ‘24 •25  $run sas:sas sercom=—log sprint=—a 1=howdata par=;size=4000k data; infile filel; input subjects a b score; proc sort; byab; proc means mean std n; byab; var score; title ‘rep. meas.’; proc anova; class sujJects a b; model score= a b subects(a) a*b b*subjects(a) / ss4; means. b / snk e=b*subjects(a); test h=a e=subects(a); test h=b a*b e=b*sub ects(a); means a / snk e=subects(a); title ‘rep. meas’; proc glm; class subjects a b; model score = a b subjects(a) a*b b*subjects(a) / ss4; contrast ‘b linear’ b —2 —1 0 1 2 / e=subects (a); contrast ‘b quadratic’ b 2 -1 -2 —1 2 / e=subjects(a);’ contrast ‘b cubic’ b -l 2 0 -2 1 / e=subjects(a); title ‘trend anal.’;  0  0  0  0  0  0  188 TABLE 31 DATA COLLECTION TABLE FOR ANOVA °A(GROUPS)° ‘  B (TESTS)  °Subjects’Test 2 ‘Test 3  ‘Test 4  ‘Test 5 ‘Test 6  0  0  0  ‘  ‘  0  0  0  0  0  0  0  0  0  0  0  0 0  ‘ 1. JNK 0 ‘ 2. DAV ‘ 0 3. APR ‘ 0 ‘ 4. ABC 0 0 5• JPN ‘ ‘ O 6. BOX 0 °Classl°7.MkT° 0 8. USA ‘ 0 9. MCN 0 ‘ 10. TAI( ‘ 0 ‘11. YAS 0 0 ‘12. 55K 0 0 ‘13. ABC 0  14 ‘ 0 0 ‘  0 Class 3  •  015. 016.  17 . ‘18. ‘19. °20. 021.  0  022.  0  ‘23. ‘24.  0  025.  TN TM MM TM HN MW GST 1414 HG HO MI TI NN  0  ‘26.  Note: A  Group, B  0 ‘  0  0  0  0 *  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  *  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 0 0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0 *  =  0  0 0  0  0 0  Test, variables used in ANOVA program.  0  189 TABLE 32 SUMMARY TABLE: ONE BETWEEN-GROUPS FACTOR AND ONE REPEATED MEASURES FACTOR FOR ALL 50 BLANKS °Source • ‘Between °subects • A 5(A)  ‘Within ‘subjects • B • AB B x 5(A) °  ‘TOTAL  Sum of squares  Degrees of freedom  134.64 684.69  a—1= 1 a(n—1)= 24  668.92 15.51 867.35  b—1= 4 (a—l)(b—l)= 4 a(n—l)(b—l)= 96  2371.11  abn—1=129  Mean square  134.64  F  4.72*’ 28.53  18.5l*° 0.43 9.03  0  * Statistical significance at the .05 level Result 1: There was a significant difference between subjects, i.e. between the control and treatment groups. Result 2: There was a significant diference within subjects, i.e. between some tests.  TABLE 33 SUNMARY TABLE: ONE BETWEEN-GROUPS FACTOR AND ONE REPEATED MEASURES FACTOR FOR ALL INTRA- AND INTER-SENTENTIAL BLANKS ‘Source  :  ‘Between ‘subjects A • S(A) O  Within subjects 0 B • •AB B x S(A) TOTAL  Sum of squares  Degrees of freedom  110.51 545.12  a—l= 1 a(n—l)= 24  Mean square  110.52  F  4.87*’ 22.71  :  0 ,0  504.51 12.98 733.86  b—l= 4 (a—1)(b—l)= 4 a(n—1)(b—l)= 96  1906.98  abn—1=129  16.50*’ 0.42 7.64  Statistical significance at te .05 level Result 1: There was a significant difference between subjects, i.e. between the control and treatment groups. Result 2: There was a significant diference within subjects, i.e. between some tests. *  0  190  TABLE 34 SUMMARY TABLE: ONE BETWEEN-GROUPS FACTOR AND ONE REPEATED MEASURES FACTOR FOR INTER-SENTENTIAL BLANKS °Source D  Sum of squares  Degrees o freedom -  •  •  Mean square -  F 0  subjects A  Within subjects • B AB • B x S(A) °TOTAL *  •0  59.10 290.77  a-1= 1 a(n—1)= 24  611.58 10.15 464.64  b-1= 4 (a—i) (b—1)= 4 a(n—1) (b—1)= 96  1436.24  abn—1=129  59.10  4.88*° 12.12  :  .  0  31.59*° 0.52 4.84  Statistiàai significance at the .05 level Result 1: There was a significant difference between subjects, i.e. between the control and treatment groups. Result 2: There was a significant diference within subjects, i.e. between some tests.  0  191 TABLE 35A STUDENT-NEWMAN-KEULS TEST FOR VARIABLE SCORE (A) FOR ALL 50 BLANKS • This test controls the type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses  o  Alpha0.05 DF=96MSE=9.03 °NumberofMeans 2 3 4 • Critical Range 1.65 1.98 2.18 GROUPING 0 °  • o °  MEAN  NUMBER  0  5 2.32 B (TEST)  A A A A A  13.69  26  6  13.38  26  5  12.52  26  4  B  9.85  26  3  c  7.82  26  2  DIFFERENCE  Not Signif. Significant  • o  : Not Signif.  0 0  0  Significant  * Means with the same letter are significantly different. Results: When the scores of class 1 and class 3 were considered together, from test 2 to test three, and from test 3 to test 4 there was significant improvement.  TABLE 35B STUDENT-NEWMAN-KEULS TEST FOR VARIABLE SCORE (B) FOR ALL 50 BLANKS • This test controls the type I experimentwise error rate • under the complete null hypothesis but not under partial null hypotheses  :  °  Alpha — 0.05 DF = 24 • Number of Means 2 Critical Range 1.93 GROUPING • •  MEAN  MSE  =  :  28.53 C  NUMBER  A  12.47  65  B  10.43  65  A (Group)  DIFFERENCE Significant  3  * Means with the same letter are significantly different. Results: When class 1 and class 3 were compared on the total of all there score for the practice tests, there was a significant difference, with class 1, the control group getting the higher scores.  : :  192 TABLE 36A STUDENT-NENMAN-KEULS TEST FOR VARIABLE SCORE FOR INTRA- AND INTER-SENTENTIAL BLANKS  (A)  • This test controls the type I experimentwise error rate under the complete null hypothesis but not under partial null hypotheses  °  : °  Alpha 0.05 DF 96 °NumberofMeans 2 Critical Range 1.52  :  GROUPING  O  • •  B B B  o O  •  0  A A A C C C  MEAN  MSE = 7.644 3 4 5 1.83 2.00 2.13 NUMBER  B (TEST)  12.04  26  5  11.41  26  6  10.17  26  4  8.73  26  3  D *  6.55  DIFFERENCE  :  Not Signif. Not Signif. Not Signif.  • o  :  0  Significant 26  2  0 0  Means with the same letter are significantly different. Results: When the scores of class 1 and class 3 were considered together, from test 2 to test three there wa& significant improvement.  TABLE 36B STUDENT-NEWMAN-KEULS TEST FOR VARIABLE SCORE (B) FOR INTRA- AND INTER-SENTENTIAL BLANKS 0  This test controls the type I experimentwise error rate • under the complete null hypothesis but not under partial null hypotheses  :  °Alpha—O.05 DF=24 • Number of Means 2 • Critical Range 1.93 • GROUPING o  • •  MEAN  NUMBER  A (Group)  A  10.70  65  1  B  8.86  65  3  DIFFERENCE Significant  Means with the same letter are significantly different. Results: When class 1 and class 3 were compared on the total of all there score for the practice tests, there was a significant difference, with class 1, the control group getting the higher scores. *  :  MSE=28.53  :  193 TABLE 37A STUDENT-NEWMAN-KEULS TEST FOR VARIABLE SCORE (A) FOR INTER-SENTENTIAL BLANKS • This test controls the type I experimentwise error rate • under the complete null hypothesis but not under partial null hypotheses.  :  ‘Alpha—0.05 DF=96 NuinberofMeans 2 Critical Range 1.21 GROUPING  : o  • • • • •  !4SE=4.84 3 4 1.45 1.60  MEAN  NUMBER  A  9.54  26  5  B B B  8.01  26  6  7.28  26  4  C  5.73  26  3  D  3.17  26  2  *  B (TEST)  DIFFERENCE Significant  Not Signif. Significant  0  o  5 1.70  Significant  Means with the same letter are significantly different. Results: When the scores of class 1 and class 3 were considered together, from test 2 to test three there was significant improvement. From test 5 to 6 there was a significant decline. TABLE 37B STUDENT-NEWMAN-KEULS TEST FOR VARIABLE SCORE (B) FOR INTER-SENTENTIAL BLANKS  • This test controls the type I experimentwise error rate • under the complete null hypothesis but not under partial null hypotheses.  :  Alpha=0.05 DF=24MSE=12.12 • Number of Means 2 Critical Range 1.26  : :  GROUPING  • • *  MEAN  NUMBER  A  7.42  65  1  B  6.07  65  3  A (Group)  DIFFERENCE Significant  Means with the same letter are significantly different. Results: When class 1 and class 3 were compared on the total of all there score for the practice tests, there was a significant difference, with class 1, the control group getting the higher scores.  : : :  194 TABLE 38 GENERAL LINEAR MODELS PROCEDURE FOR ALL 50 BLANKS  ° °  a  °  Test of hypotheses using the type iv MS for subjects (A)’ as an error term. CONTRAST DF SS F VALUE PR > F  0  .  Linear  1  608.88  21.34  0.0001  Significant’  Quadratic  1  51.26  1.80  0.1926  Not sign.  Cubic  1  3.71  0.13  0.7215  Not sign.  Results: For the linear contrast the F Value is higher than 2.0 and the Probality is less than .05, so there is a linear tendency of improvement from test to test. TABLE 39 GENERAL LINEAR MODELS PROCEDURE FOR INTRA- AND INTER-SENTENTIAL BLANKS • Test of hypotheses using the Type IV MS for subjects (A)° as an error term. • CONTRAST DF SS F VALUE PR >F  o  0  0  :  0  Linear  1  440.13  19.38  0.0002  Significant’  Quadratic  1  50.13  2.21  0.1504  Not sign.  Cubic  1  8.09  0.36  0.5562  Not sign.  0  o  0  Results: For he linear contrast the F Value is higher than 2.0 and the Probality is less than .05, so there is a linear tendency of improvement from test to test. TABLE 40 GENERAL LINEAR MODELS PROCEDURE FOR INTER-SENTENTIAL BLANKS • Test of hypotheses using the type iv MS for subjects (A)° • as an error term. • CONTRAST DF SS F VALUE PR > F • 0  : °  Linear  1  473.04  39.04  0.0001  Significant°  Quadratic  1  104.14  8.60  0.0073  Significant’  Cubic  1  20.02  1.65  0.2109  Not sign.  Results: For the linear contrast the F Value is higher than 2.0 and the Probality is less than .05, so there is a linear tendency of improvement from test to test. There is a slight quadratic trend.  195  To summarize the ANOVA results for the five practice tests  can  it  difference  seen  between  differences their  be  the  between  scores  were  that two  there  was,  classes,  earlier  tests  pooled,  and  a  significant  some  significant  a)  b)  for  both  that  classes  there  C)  interactive effects of the students and tests. also showed that, linear,  i.e  practice  d)  there  test  to  when  were  no  The ANOVA  the trends of the tests scores were  was  generally  practice  test  steady with  improvement  the  exception  from of  a  quadratic tendency in the inter-sentential analysis results. This quadratic tendency by definition takes precedence over the linear trend but the much higher significance of the linear trend suggests that despite the quadratic downturn in the inter—sentential results, the general overall direction of the scores was upwards.  Point (a): A significant difference between classes Point (a) terms  of  the  above is interesting but not very helpful in research  illuminate the fact,  question.  The  ANOVA  does  not  like the graphs do, that the treatment  group did poorly at first and then improved.  Most likely  the initial drop while the treatment group was getting used to  the  color—coding  is  what  difference between the two classes. test,  the  ANOVA  possibly  caused  the  significant  When compared to the T—  overstates  the  significant  difference between the groups, being that the T-test stated that only test 3 was significantly different.  196  Point (b): Some significant differences between tests Point the  (b)  students  about the differences between tests for all taken together  is  not  very  helpful  as  this  research project is more interested in the groups compared to  each other.  The graphs  appear to  clearly  show which  groups did better on which tests, the T-tests showing which differences were real and which probably were by chance. fact,  after  looking  at  the  graphs,  the  T-tests  and  In the  ANOVA, there was a lot of “noise” as Dr. Boldt referred to the ambiguity in the information generated by the research design.  Point (c): No interactive effects of students and tests Point  (c)  is  the  most  interesting,, revealing,  and  disappointing as far as those people who support color—coded doze procedure are concerned.  If the rationalized CCC had  been really effective, then there would have been something in  the  tests  that  interacted  with  something  within  the  subjects to create a new dynamic which was different from both the tests and the students.  Perhaps the color—coding  would have sparked an understanding of the doze exercise which  would  have  generated  motivation  to  get  even  more  involved in the process of finding inter-sentential clues. to fill in the doze blanks.  197  Point (d): Linear trend given by the scores of the practice tests Point  (d)  is  encouraging  for the  doze procedure  that there was benefit gained from the practice. were  never  above the  frustration  level,  were headed in the right direction.  the  last  practice  test when  the  but  in  The scores  the  students  The exception was in  inter—sentential  results  declined and the gap between the treatment group and the control group increased although not significantly.  For the  inter—sentential analysis there was also a slight tendency of the trend to be quadratic.  This quadratic tendency is  shown on the graphs and in the statistics as a decrease in the scores from tests 5 to 6 as far as the inter—sentential blanks  were  concerned.  The  downturn  suggests  that  the  ceiling of accomplishment had been reached for both groups by  test  5  clues.  in  finding  inter-sentential  It should be pointed out here,  reiterative-word however,  that for  practice test 6 the attendance was very low for both the control and the treatment group.  Only seven students in the  control group and eight students in the treatment group took the test.  So  The others had to go on field trips.  far  in this  discussion  of  the  different ways  of  analyzing the data, exact—word scores have been concentrated on.  Acceptable-word scores should be talked about because  they  were  Indeed,  discussed  however,  the  in  the  review  discussion will  of be  the  literature.  short because  in  198  terms of the present research results they did not improve the effects of the rationalized CCC.  They only widened the  differences and brought in more error from assigning points to words that may or may not have been acceptable.  The Hypotheses Hypothesis 1 On the practice tests there will be no significant difference  between  the  treatment  group  which  has  trained using the color—code doze and the control groups which have trained on either the non-colored colored-doze,  the  colored-blanks  only,  or  the  randomly—colored doze practice tests. Result 1 According to the data as analyzed by the T-tests and ANOVA the above null hypothesis was not rejected. Hypothesis 2 On  the  post-test  difference  there  between  the  will  be  treatment  no  significant  group  which  has  trained using the color—coded doze and the control groups which have trained on either the non—colored colored-doze,  the  colored—blanks  only,  or  the  randomly—colored doze practice tests. Result 2 According  to  differences, rejected.  t-tests so  the  there null  were  no  hypothesis  significant was  not  199  Hypothesis 3 On  post-test  the  difference  there  will  confidence  on  be  no  scores  significant between  the  treatment group which has trained using the color— code doze and the control groups. Result 3 Raw data was kept and the use of graphs.  informally compared through  These graphs  patterns to the mean score data. was suggested,  followed similar No significance  so again the null hypothesis  could  not be rejected. Hypothesis 4 On  post-test  the  there  will  be  no  significant  difference in productive confidence scores between the  treatment  group  which  has  trained  using  the  color—code doze and the control groups.  Result 4 From the statistical data it appears that there were no significant differences between the control group which  used  procedure  the  and  the  non-colored treatment  color-coded version. practice  tests  and  This the  post  rationalized group which  was  used  the  for  the  Therefore,  the  the  test.  doze  case  null hypotheses could not be rejected.  200  As  is typically the case when the null hypothesis is  not rejected it does not mean that the treatment failed.  It  just means that it was not proved to be successful, in which case more research was suggested. improved  from  supported  pre—test  futher  a  to  look  a  look  research,  at  a)  b)  the  and  confidence data.  c)  was  the  taken the  encouraging  test  Therefore,  statistical  information  questionnaire,  post—test  within  future research projects. is  The fact that both groups  results  and  and in  immediately following  design from  the  of  the  confidence  present  post—session  and  productive  In Chapter Five there are some suggestions  future research designs that are intended to address some of the questions raised in the present research.  Problems with the present research design Having looked at a number of data analyses it can be seen that the results of the research are not very clear. There  was  technique  a  lot  did  of  not  noise  choose  differences  on  differences,  statistically  Certainly,  the  created the  graphs  because  students could  different,  been part  given the weak effect or even the  results of the rationalized CCC treatment,  sampling  randomly.  have or  the  of  The chance each.  inconclusive and combining  the burdens of the expense and time involved in color—coding the parts of speech, there is no way one could justify going to the trouble of color—coding doze passages for classroom use.  201  Hope given by the questionnaire Yet, there are some things about the present research which should keep the rationalized CCC alive for at least a short time.  First,  Table 41 below) the  ma)ority  in the questionnaire  (please refer to  given after all the tests had been written of  the  students  in  the  treatment  group  indicated that although they did not like doze procedure they would prefer the color—coded form. that  they  felt  doze exercises. the  nuaiber  productive above. clues  of  that  they had  learned  Also they reported something  from  the  Furthermore, the treatment group increased guesses  confidence  they  made  scores  as  and  increased  indicated  in  on  the  their graphs  Finally, the students reported that they looked for inter—sententially  as  the  clues  appeared  after the doze the doze blank in question.  before  or  202  TABLE 41 POST-SESSION QUESTIONNAIRE 0  Class 1 (N=l4) Yes No  • o  Class 3 (N=15) Yes No 0  1. doze experience in • a) Japan: b) Canada: Enjoyed?  1  0  :  Any difficulties?  13 1 13  13 3  2 1 0  13 14 3  1  14  1  °2.2 More difficult • project beginning: • project middle: project end:  7 3 4  11 2  °2.3 Easiest o project beginning: • project middle: project end:  2 4 8  3 10 1  °2.4 Gain confidence?  5  8  8  2.5 Problems seeing words?  7  7  10  2.6 Improved guessing?  6  8  8  7  14 12 12 13 3  0 2 2 0 10  15 13 11 12 1  0 1 2 2 13  3  11  5  10  12  3  :  : o  : 0  : 7  °2.7 Looked 0  •  a) within sentence: b) before sentence: c) after sentence: Guessed: Did something else:  °  • •  Want to do doze again: °3.l Used color—coded doze? •  :  •  3.2 CCC parts of speech helpful?  03•3  • • •  • o  : 0  :  CCC parts of speech made passages more difficult? a) passages 2 and 3? b) passage 4? c) passages 5 and 6?  13  2  2  13  1  : 0  203  parts of speech made passages easier?  03•4  : o  13  a) passages 2 and 3? b) passage 4? C) passages 5 and 6?  •  :  2  6 5 2  :  helped find clues  : o  a) within sentence b) before sentence c) after sentence  °  : •  fl 12 10  of the blank?.  •  0  :3.6. CCC helped?  :  7  3.7. Type of doze preferred ° in the future: 0  0  °  a) non—colored b) colored doze  °  1 14  0  Whether the students in the treatment group were helped or hindered by the color-coding, long,  with  intra—sentential  iterative-word clues, an active part  in the short run or the or  inter—sentential  re  it is exciting to know that it played  in the present research.  Fortunately,  it  helped to show that students did look beyond the éentence of the  doze  important  blank  for  clues  to  find  fillers.  Even  more  is that the color-coding did aid  in the global  process, at least as long as it was present.  Even when the  color—coding was group  still  made  not used progress  in  the post—test the  over  the  pre—test.  treatment Finally,  despite the difficulties of the doze procedure (color—coded or not)  and a certain dislike of it,  the students in the  treatment group reported that they thought they had learned  204  something and, most satisfying of all,  the majority of the  students  indicated  in  the  questionnaire  that  treatment in  the  group  future  they  would  color—coded doze form to the non—colored one.  on  the  prefer  the  To rule out  that their answer was ‘ust by chance, when group two  (only  six of whom had experienced one randomly—colored doze test) answered  the  same  question  coding got very few votes. such  a  limited  on  the  questionnaire,  color-  These votes were so few and had  experiential  basis  that  they  could  be  considered to be given by chance unlike the firm choice by group three. The conclusions drawn from the analysis of the productive confidence scores A  look  at  the  questionnaires  indicated  that  the  treatment group preferred the color-coded doze to the noncolored  form  but  was  this  conclusion  students efforts on the doze tests? were  re—analyzed  students  had  in  filled  terms out  of (the  the  supported  by  the  To find out the tests number  confidence  of  blanks  level)  and  the  the  number of right answers as a proportion of that confidence (the productive confidence level).  It should be noted here  that unlike the previous analysis  absentee student  scores  were not interpolated. The reasoning behind the productive confidence tables (Tables 21, 23, 25, and 27 above and Table 42, below) was to find out  whether  or  not  the  color-coded parts  of  speech  205  treatment was encouraging students to fill  in more blanks  and more blanks more accurately than the control group. A quick look at the tables will show that the control group did better than the treatment group on most of the measurements and most of the time. exact—word answers for a)  This was true for the  all the doze blanks,  and inter—sentential blanks,  and c)  b)  inter— blanks.  also true for intra- blanks through test 5.  intra It was  In test 6 and  the post—test the treatment group earned higher mean scores for exact answers. With regards to the productive confidence scores there are two columns  in the tables to look at.  The first  is  Column E  (% of possible blanks filled in) and the second,  Column G,  the correct blanks as a proportion of the number  of blanks filled in.  Column E gives the confidence level  and G, the amount of success the group had in the number of blanks filled out. the  kind  Group 1 was ahead in Column F despite  of blank used,  i.e.  the  control  confident throughout the research project. productive got  confidence  higher  scores  understandable  scores,  than  group was more However, on the  the treatment group usually  the  and predictable  control in  light  group. of the  conservative risk takers make less mistakes. as  the  treatment  improvement statistical advantage  scores  made than  by  were the  real.  over the  control  only  Furthermore, group  higher,  groups if  fact  is that  In addition,  slightly  treatment  This  there  was was  the more any  gained by the treatment  206  group on the practice tests for inter- blanks, in  the post-test.  For  intra- blanks,  it was lost  the treatment group  began on the pre-test lower on the number of blanks filled but  were  a  little  productive  answers.  productive  confidence  generally  lower  higher  than  During scores  than those  the of  of  the  control  practice  the the  treatment regained the lead on the post—test.  on  sessions  treatment control  group  group  group  the were  but  the  Nevertheless,  the treatment group did not do as well as on the pre—test, while the control group reduced the difference slightly. TABLE 42 TREATMENT GROUP’S CONFIDENCE LEVEL AS A % OF CONTROL GROUP’S CONFIDENCE LEVEL FOR THE MAIN PROJECT 0  TEST  INTRA+INTER  INTRA  Pre—  •  91.0 %  •  86.1 %  •  76.9 %  T2  •  760%  •  75.4%  •  836%  0  0  T3 0  T4 0  °  ALL BLANKS  0  • •  74.0 %  •  90.1 %  • . 0  0  T5  .  T6  Post—  74.0 %  83.1%  •  80.8 %  •  87.7  °  89.8 %  42  %  above  0  shows  73.6 %  •  83.2%  0  69 8 %  88.0 % the  •  0  •  71.3%  •  0  74.1 %  •  0  94•7 %  0  85.7%  0  81.7 %  89.3 %  0  0  0  •  •  •  0  0  0  • •  Table  0  INTER  85.7 %  87.8 %  students  in  0  • •  938%  •  80.6 %  0  the  0 0  88.1 % treatment  group did not surpass the confidence levels of the control group on any of the levels.  The only major improvement in  207  confidence  from  pre—  to  post  test  came  on  the  intra—  sentential measurement. It appears that color—coding helped improve confidence more on the intra—sentential level than the targeted inter— sentential deletions  level. were  made  This in  is  order  surprising to  maximize  given the  that  the  number  of  beyond—sentence clues and given that there were only a low number of intra—sentential re—iterative-word clues.  208 CHAPTER FIVE IMPLICATIONS OF THE RESEARCH INTRODUCTION  I wonder if the inventor of the doze procedure, Wilson Taylor,  had  a  sense  of humour besides  his  genius.  What  could be simpler than deleting every fifth word from a 300 word passage and multiplying the number of correct answers out of 50 by two to get the students scores in percent.  I  wonder if Taylor realized at first how difficult it would be for  gaps  in  difficult  the  for  doze ESL  procedure  students,  to  but  be  filled,  also  not  only  difficult  for  researchers in the context of methodological questions. review of the literature showed a challenge for both.  A  There  are so many seeming possibilities on both levels to consider in  dozing the  encounters”  gaps  (to  that there will  borrow  the  name  of  be many more a  series  of  “doze books  containing graded cJ.oze exercises). The  original  enthusiasm  and  effort  expended  doze procedure by students and researchers,  on  the  both in terms  of moral support and the great amount of research, has waned over the years.  It is curious to note that ‘ust as students  have been criticized for not looking beyond the sentence of the doze blank for clues, researchers can be criticized for not trying to develop the doze procedure beyond a discrete exercise.  In fact,  the case is somewhat the opposite as  indicated in the present research and in a review of the doze literature that showed that students and researchers  209  do  look  and  act  more  globally,  sometimes  Chapter Four pointed out how much sententially.  successfully.  students  looked  inter  Chapter Two told of Chihara and Oiler who  claimed to have shown that the more proficient ESL students they tested were able to use inter—sentential clues. it  told  of  others  who  devised  beyond—sentence clues. other hand, the  (See  ways  Bachman  to  test  and  for  Also use  Henk.)  On  of the  Shanahan and Kamil and others claimed to show  opposite.  Working  on  the  discrete  level  some  researchers tried to improve c].oze scores by focussing on intra—sentential lengths  equal  clues  to  such as  that  of  the  initial  letters  missing words.  and  blank  Meanwhile,  others tried to bridge the seeming opposite point of view (discrete-versus-global) by dissecting the doze passages to account for what influenced the students to look beyond the sentences,  i.e.  to the presence of inter—sentential clues.  Apparently, few existed. fillers  and  various  Still others tried multiple-choice  kinds  of  matching  which  required students to draw on inter—sentential  sometimes  information.  The present research made use of color-coding in an attempt to help  students  find intra- and  inter-sentential  (explicitly of the reiterative-word type).  fillers  The choice of  color—coding was borrowed from the use of color—coding in business and in teaching from Gattegno who taught ESL/EFL students  to  read  various spellings.  by  color—coding  the  sounds  and  their  210  At the heart of the present research, then, so  simple  issue  of  “discreteness”  versus  is the not  “globality”.  Regardless of the difficulty of the doze passages in this research, using  students  the  using  the  rational  color—coded  rational  find  and  ability to  intra-  doze  doze  and  improved  inter-sentential  students in  clues  their of  the  reiterative—word type.  FUTURE RESEARCH DESIGNS Table 43 below suggests many variations on the colorcoded doze that can be tested to see under what conditions the  CCC  could  possibly  help  students  look  beyond  the  sentence of the doze blank to find the appropriate filler. (Note the  “MP”s  ‘ust  to  the  right  of  the  table.  These  indicate that these characteristics were used in the “Main (Research) Project”. There are two ways to test the effects of the color— coded doze procedure.  The first is against the non-colored  doze forms and the second is the color—coded doze against various forms of itself. In  using  constant. Group  2,  the  Group the  I  former,  the  should use  randomly-colored  color—coded (*)  groups  the  should  non—colored  (#)farm;  and  form of the doze procedure.  be  kept  (I)  Group  form; 3  the  If the niunber  of students available is limited, then the random form could be omitted. be  left  out,  (In later studies, so  the  the non-colored form could  random-color  form could be  left  in.  211  There is a danger that the students doing the random—colored doze would drop out as they did in the present research project.)  If two classes are used, one being judged to be  better than the other, then the non—colored doze and color— coded  doze  could  be  used,  each  half  of  receiving one or the other of the forms.  the  each  class  This technique of  splitting the classes should partially compensate for the bias  caused  other much  at  by  the  larger  students  one  class  start.  being more proficient  Complete  classes  to  in each class.  allow  than  the  compensation would require for  However,  a  random  selection  of  such large classes are  rarely available. In using the latter, the differences within the various categories  (read down left column)  could be compared with  each other. All the above variations could be repeated by comparing ESL speakers to native speakers of English and by comparing young ESL students to adult ones. Even though the possibilities for doze pro)ects seem limitless, research  the  still  two  questions  apply.  First,  so  important  “Can  and  in  do  the present  students  look  beyond the sentence of the doze blank to find reiterative— word clues to fill  in the doze blank?”  Second,  “Do the  color—coded parts of speech help students in their inter— sentential clue search more than when the words were not color—coded?”  212  The second question also leads to another question and another variation in the research. happens  when  the  color—coding  is  The question is, removed?”  The  “What  present  research appeared to show that the advantage of the crutch called  “color—coding”  removed. advisor, see  if  coding  disappeared  when  the  color  was  To counteract this result, it was suggested by my Dr.  Slade,  that the color could be phased out to  diminishing  the  could  them  help  students’ focus  reliance  on  the  on  the  color-  parts  of  speech  themselves rather than on the color—coding.  The phasing  out of the color could be done by reducing the intensity of the  color  from  the  first  practice  test  to  the  last.  Actually this tended to happen in the present research as the printer  ribbon  started to wear out.  However,  a  new  ribbon had to be used before all the tests had been printed, so the colors became more  intense again.  Another way to  phase out the color might be to make the first practice all in color, the second one one-fifth in color, fourth in color, and so on.  the next one—  An added benefit of this would  be that the test scores could be broken down into color— coded and non—color—coded parts and these compared to each other.  (With  a  good  computer  program  patience this could be done successfully.  and/or  a  lot  Otherwise,  process would be too time consuming and tedious.)  of the  213 TABLE 43 SUGGESTIONS FOR POSSIBLE FUTURE RESEARCH DESIGNS = done in Major Project  “HP”  0  TYPES OF CLOZE °1. Standard 2  •  0 GROUP I • GROUP II GROUP III •) 0 I * • non—colored random—colored color—coded°MP  •  •  Rational  .  maximum  1 • intra-type a inter-type  • no. of • re-iteratives o (all parts ‘ of speech) intra/inter’  :  1 no • of intra—type 0 re—iteratives • (nouns inter-type a verbs, 0 adjectives ‘ intra/inter 0 adverbs) •  °PASSAGES  •  inter-type intra/inter  0  intra—type  0  °2. Order increasing °(according to ‘ no. of re— decreasing oiterative_word • blanks)  0  inter—type intra/inter  *  indépeñdeñt’  instruction  0  frustration  • frustration °MP  increasing  0  decreasing  instruction°  mp  increasing  different •  0  0  a  lsame+ • lsame+ ‘1 different 1 different  •  ‘  unrelated •  decreasing 0  •  related  °  intra—inter°  0  0  a  4. Relatedness’ of subject • matter •  0  indedendent  0  •  inter—type 0  3. Pre-/Post • tests • (no—color)  same  intra-type a  0  0  O  0  intra/interMP *  •  l. Readability’ independent’ Level by formula • instructiono doze results frustration’  *  intra—type inter-type  1  —  O  a  maximum  •  o  .  intra—type  *  same  •  same  0  different  °MP 0  • different 0  •  lsame+  °l different  related  •  related  unrelated  •  unrelated  ‘HP  214  STUDENTS Di.  •  •  Proficiency° Level  e  medium 2. Experience • with doze .  0  •  some  0  .  none SCORING  :e  o  Spelling  •  exact  0  0 0  •DATA (steps)  :  exact  Collection (1 student! row)  some  0  :  0  0  0 Evaluating answers  •-right —accept _wron  :  exact  •  *  0  exact  0  =  —  0  acceptable •MP exact  °MP  0  0  ::  incorrect • incorrect •M? if not same if not same° as re-iter ative blank O ative blank° word • word O  •  as  re_iter_0  *  spreadsheet (1 word/ column/ student) -intra-type  0  -inter-type -neither  0  0  :  spreadsheetMP (1 word! coiumn/ student) 0  :  °MP -inter-type °MP °—neither •MP °-intra-type  0 0  =  °)  0  °  •  MP 0  none  •  spreadsheet° (1 word! coiumn/ student)  °MP  0  0  0  :  •  exact  Assignment -intra-type of type of °-inter-type re-iterative neither °word 0  some  acceptable  • incorrect o if not same° • as re—iter—° • ative blank° °word  0  °  medium  acceptab  0  0  •  0  0  0  medium  0  0  high  0 0  0  •  .  high  *  none  O acceptable  0  :  high  0  0  o  •  0  1 -right .5 —accept O -wrong  Isolation (delete of type of • columns re—iterative containing word • other two • types)  °  0  — = =  1 .5 0  (delete columns containing other two types)  0  °-right °—accept -wrong  1 •)( =  •5°MP  =  0  (delete • columns 0  containing  • other two  types)  : mp  215 .  Summation ‘—across for • of points student’s • • to find • score total scores ‘—down for • for each blank’s • student, score/group’ means for ‘—compare a each blank! sum of down’ group, and • totals to • means for sum of • each group across • totals • ‘—do averages’ 0  ‘  ‘  ‘  ‘  °  ‘  ‘  ‘  ‘  0  °  ‘  ‘  Graphs  —across for student’s score —down for blank’s score/group —compare sum of down totals to sum of across totals —do averages  a ‘—use bar —use bar graphs in graphs in spreadsheet’ spreadsheet program program  ‘  0  ‘ ‘  ‘  Statistics Program  ‘ ‘  ‘—use “One —use “One Between and’ Between and ‘One Repeated’ One Repeated Measure Measure ‘Factor ANOVA° Factor ANOVA ‘  •  0  ‘  ‘  ‘—across for MP student’s score ‘—down for ‘HP blank’s score/group’ ‘—compare ‘MP • sum of down’ • totals to • sum of • across • totals ‘—do averages’MP • •  ‘—use bar ‘HP graphs in spreadsheet’ • program ‘  ‘—use “One ‘HP Between and’ ‘One Repeated’ • Measure ‘Factor AVOVA’ ‘  Note: “MW’ means a similar version done in Major Project. The  above table  suggest many variations  done in the present research project.  on what was  Many variations are  what could and should have been done in the main research if there had been enough students and enough time.  However,  even a lot more could be done to find out if students are looking inter-sententially for clues fill in doze blanks. Think-aloud protocols could be recycled to include the new idea of color—coding. of  participants  for  Furthermore, to increase the number statistical  purposes,  many  students  could use the language laboratory where their thoughts could be recorded to determine which and how many inter—sentential reiterative-word clues they are able to find.  216  For  those  available, era  in  i.e.  research  laboratory  computers  in  a  presented to  with  even  more  sophisticated color computers,  doze  computer  researchers  is  with  awaiting. a  minimum  network.  The  a whole new  Needed of  20  doze  resources  would  IBM  be  486  passage  color  would  each student on his/her computer.  a  be  Half the  students would have the color-coded version, the other half the  non—colored  answers.  form.  The  students  could  type  in  the  A computer program (yet to be written) would keep  track of every key stroke to determine where the student was looking  for  answers,  and record  every  answer  spreadsheet, keeping statistics throughout.  on  a  group  The researcher  would know at every instant how everyone was doing.  Both  sets of information could be produced in hard copy to make the  flow of  results easy to  see.  The research could be  designed so that the students would get immediate feedback after the test. Once the methodology has been polished, teachers could make use of the instant feedback to do what researchers by definition cannot do.  Teachers could make any changes that  the feedback indicated. spreadsheet  and  With the sophisticated computers,  analysis  programs,  and  peripherals  like  scanners, teachers could easily create more doze passages. Teachers  could  choose  stories,  scan  them  program,  easier  into  a  passages, word  more  interesting  processing/spreadsheet  check to confirm the readability level,  find the  number and location of reiterative words, delete as many of  217  those words as possible, color—code the blanks and remaining words  and  make  network.  a  master  copy  for  use  over  the  computer  Such computer technology could give the teacher  the insights of a researcher without restricting the teacher to  the  inflexibility  benefit  because  of  the  pure  research.  teacher  could  Students  quickly  would  adapt  the  materials to the needs and interests of the student.  The  result  the  of  ease  in  preparation  and  analysis  plus  relevancy to students of the doze passages could increase the motivation of both the students  and teachers to take  full advantage of the intended value,  i.e. the globality of  the doze procedure. The coding, given  above  vision  about  computer  and the doze procedure the  capabilities  programers, project,  plus  the  the  dream  of  of  a  is a dream,  computer  insights  technology,  the  computerized  procedure is a reachable one.  indeed.  designers,  from  color— But  computer  present  research  color—coded  doze  Even though the color—coded  treatment group in the present research did not overtake the non-colored control group, overall  and  there was enough improvement in  inter—sentential  scores  and  in motivation  to  support the value of more research into the use of color coded doze passages. the  door  has  been  Taylor, would have been excited that opened  wide  once  more  on  his  doze  procedure and his dream for a simple global teaching/testing method is closer than ever.  218  From  discrete,  to  global,  to  the  universe...One’s  imagination can run on endlessly and that’s exciting for the researcher. with  the  gritty.  The actual research is much more down to earth,  research  and  the  students  buried  in  the  nitty  Getting out of the discrete level with the doze  procedure, color—coded or otherwise, is like breaking the G force  of  gravity  student with  in  lots  a  of  rocket  ship.  It  determination to  takes  take  a  special  off with  doze test and reach some distant destination.  the  The doze  procedure is not really for the masses who want only to be spoon-fed.  Nor  is  it  for those  researchers who want to  isolate one variable in a universe of interacting forces, while students dance around the candy shop of fancy packaged knowledge.  The doze procedure and its color—coded crutch  are better left to the classroom situation where they can be pulled out when a) the students are in the mood for them, b) when the teacher can discuss the guessing strategies with the  students,  and  Unfortunately  none  C)  when  of  these  there  is  time  conditions  do  for  feedback.  coincide  with  traditional research which by its very nature is discrete, not global.  Perhaps,  though as humankind has managed to  analyze the universe with billions of discrete  questions,  almost simultaneously using sophisticated computers, maybe, as suggested above, we can bring the power of computers to bear on researching the doze procedure going beyond the sentence level.  as  it relates to  Maybe in the process and  under more favorable circumstances we can see if the color-  219  coded doze procedure is as good as on the surface it would appear to be.  220  BIBLIOGRAPHY  Aitken, K. G. Using doze procedure as an overall language proficiency test. TESOL Quarterly, 1977, j,, 59-67. Aitken, K. G. Constructing a multiple-choice test of ESL. RELC Journal, 1978, (2), 105—108. Alderson, J. C. The doze procedure and proficiency in English as a foreign language. TESOL Quarterly, 1979, (2), 219—227. Alderson, 3. C. Native and non—native speaker performance on doze tests. Lariauage Learning, 1980, (1), 59—76. Anderson, 3. The use of readability formulae in the Territory. Papua and New Guinea Journal of Education, 1969, 17—22. .,  Anderson, 3. The application of doze procedure to English learned as a foreign language in Papua and New Guinea. Encrlish Lanquage Teaching, 1973, 22, 66—72. Anderson, D. C. Cohesion and the doze test. Unpublished M. A. thesis, University of Illinois, 1980 Anderson, 3. The application of doze procedure to English learned as a foreign language in Papua and New Guinea. En1ish Lariauage Teachinc, 1973, 21, 66-72. (supported exact scoring over alternative response method p. 71) Ashby-Davis, C. A. Levels of certitude for educated guessing in strict doze passage. Journal of Reading, 1984, 22, 96—108. Ashby-Davis, C. A. doze and comprehension: A qualitative analysis and critique. Journal of ReadincT, 1985, 585—589. ,  Bachman, L. P. The trait structure of TESOL Quarterly, 1982, i(1), 61—70.  doze  test  scores.  221  Baclmian, L. F. Performance on doze tests with fixed ratio and rational deletions. TESOL Quarterly, 1985, (3), 535—556. Bormuth, J. Comparable doze and multiple choice R. comprehension test scores. Journal of Reading, 1967, 291—99. Bormuth, J. R. The doze readability procedure. English, 1968, 4., 429—436.  Elementary  Borrnuth, J. R., Carr, J., Manning, J., & Pearson, P. D. Children’ s comprehension of between-and—within—sentence syntactic structures. Journal of Educational Psychology, 1970, i, 349 —357. Bridge, C. A, Thomas, S., & Evers, L. The effects textual cohesion on doze comprehension scores. 1980, , 184—190.  of  Bridge, C. A., & Winograd, P. N. Readers’ awareness of cohesive relationships during doze comprehension. Journal of Reading Behavior, 1982, j(3), 299—312. Brutten, S. An analysis of student and teacher indications of vocabulary difficulty. RELC Journal, 1981, .(i), 6671. Carroll, J. B. Defining language comprehension: Some speculations. In J. B. Carroll & R. 0. Freedie, (Eds.), Lanauae comprehension and the acguisition of knowledge. New York: Halsted, 1972. Carver, R. P. Measuring prose difficulty using the Rauding scale. Reading Research Quarterly, 1975-1976, U(4), 660—685. Chastain, K. Developing second-language skills: Theory to practice, 2nd edition. Chicago: McNally, 1976. Chihara, T., Oiler, J., Weaver, K., & Chavez-Oiler, M.A. Are doze items sensitive to constraints across sentences? Lanauaae Learning, 1977, aZ(l), 68—73.  222 Cohen, Andrew D. Teaching Newbury House, 1980.  language  ability.  Rowley:  Craker, H. U. Clozentropy procedures as an instrument for measuring oral English competencies of first grade children. (Doctoral dissertation, University of New Mexico, 1971) Dissertation Abstracts International, 1972, fl, 6286A—6287A. (University Microfilm No. 72—13, 781) Cranney, A. The construction of two types of reading tests for college students. Journal of Reading Behavior, 1972, , 60—64. Cunliffe, M. Reading programs using Unpublished manuscript, U.B.C., 1986. Dale,  N.  colour  On the teaching of Enlish reading.  cues.  Edinburgh:  Turnbull & Spears, 1926.  Darnell, D. K. The relationship between sentence order and comprehension. Speech Monoarah, 1963, ,Q, 97—100. Darnell, D. K. The development of an English language proficiency test of foreign students using a clozentropy procedure. Speech Monographs, 1960, fl, 36-46. (ERIC ED 024 039) Dodds, W. G. Jr. A comparison traditional basal reading.  of  words  Fleech, R. A new readability yardstick. Psycho1ocy, 1948, 221—233.  in  Fry,  reading.  and  Journal of Alied  Foley, J. A. More questions on assumptions testing. RELC Journal, 1983, li(l), 57—68. Fries, C. C. Linauistics and Rinehart and Winston, 1963.  color  New  about  York:  doze  Molt  E. Fry’s readability graph: Clarification, validity, and extensions to level 17. Journal of Reading, 1977, 249.  223  Gattegno, C. Words in color: Background Chicago: Encyclopedia Britannica, 1964. Gattegno, C. silent way.  and  orinciples.  Teaching foreign lanauaae in schools N.Y.: Educational Solutions, 1972.  the  Goodman, K. S. Psycholinguistic universals in the reading process. In P. Pimsteur and T. (Eds.), g Quinn sycholoav of second lanuae learnina. Cambridge: Cambridge University Press, 1971, 135-142. Halliday, M.A.K. and Hasan, London: Longman, 1976.  R.  Cohesion  in  English.  Hanzeii, V. E. doze tests in French as a foreign language error analysis. In E. J. Briere and F. B. Hinofotis, (Eds.), Concepts language in testing: Some recent studies. Washington, D.C: TESOL, 1979, 3—11. Hanze].li, V. E. The effectiveness of doze tests measuring the competence of students of French in academic setting. French Review, 1977, .Q, 865-874.  in an  Hay, J. & Wingo, C. Lippincott, 1954.  B.  E.  Reading with phonics.  N.Y.:  3.  Henk, W. A. Effects of modified deletion strategies and scoring procedures on doze test performance. Journal of Reading Behavior, 1981, (4), 347—357. Herik, W. A., Helfeldt, 3. P., & Rinehart, S. D. A metacognitive approach estimating to intersentential integration in doze tests. , 1985, , 213—218. Hiller, Harry H. Canadian Society, A Macro Scarborough, Ontario, University of Calgary, Hall Canada Inc., 1991.  Analysis. Prentice-  Hinofotis, F. B. doze as an alternativie method of ESL placement and proficiency testing. In J. W. Oiler, Jr. and K. Perkins, (Eds.), Research in lanauaae testing. Rowley, Mass: Newbury House, 121-128.  224  Hinofotis, F. B. and Snow, G. B. An alternative doze testing procedure: multiple—choice format. In 3. W. Oiler, Jr. and K. , (Eds.), Research in lanage testing. Rowley, Mass: Newbury House, 129-133. Hymes, D. H. On communicative competence. Holmes 3• (Eds.), Sociolinquistics. Penguin, 1974. Irvine, P., Atai, P., and Oiler, 3. W. the Test of English as a Foreign Learning, 1974, 2.4, 245—252.  In 3. B. Pride & Harmondsworth:  doze, dictation and Language. Lanauaae  Janssens, R. Stimulating motivation through audio—visual aids based on ‘English by Radiovision’. Enalish Language Teaching Journal, 1977, fl(l), 43—49. Jongsma, E. A. doze instructor research: Newark, Delaware: IRA, 1980.  A second look.  Judd, E. . vocabulary teaching and TESOL: A need for reevaluation of existing assumptions. TESOL Quarterly, 1978, 14(1), 71—76. Jupp, T. C. & Hodlin, C. Industrial English: An examDle of theory and practice in functional lanuae teaching. London: Heinmann, 1975. Kalivoda, T. Learning to listen what can be done. Teachina Forum, 1980, , 2-7.  Enalish  Katz, 3. and Fodor, 3. The structure of semantic theory. Language Learning, 1963, , 170—210. Kibby, N. W. comprehension. 299—312.  Intersentential processes reading in Journal of Reading Behavior, 1980, ,  Kirby, C. L. Using the doze procedure as a testing technique, reading diagnosis and evaluation International Readina Association Conference Proceedings, 1968, (4), 68-77. Newark: International Reading Association, 1968.  225  Klare, G. R. formulas.  A second look at the validity of readability Journal of Readina Behavior, 1976, , 129—151.  Leys, N., Fielding, L., Herman, P., & Pearson, P. D. Does doze measure intersentence comprehension. A modified replication of Shanahan, Kamil, and Tobin. , 1983, .3., 111—114. Little, J. Canadian ESL readers for adults: An annotated bibliography. TESL Talk, 1988, .2(l), 104—111. Long, N. N. A new look at discourse in the teaching reading. RELC Journal, 1981, 12(1), 72-77.  of  Lord, F. N. and M.R. Novick. Statistical theories of mental test scores. Reading, MA Addison-Wesley. MacGinitie, W. Contextual constraint in paragraphs. Journal of Psycholoqy, 1961,  English prose 121—130.  ,  Martin, C. J. Comprehension of telegraphic prose 11. Effect of a grammatical scheme. , 1973, 2.2., 101—111. Reaction to using colour in early reading McHugh, W. instruction, Decade of Innovations Aoroaches to Readinc, Beginnin Newark: International Reading Association, 1968. —  Mohan, B. Lanquage and Content. Wesley, 1986.  Reading,  Mass:  Addison—  Montety, P. de Silent teacher, talking pupils. Teaching Forum, 1977, 2, 11—14. Morley, J. Improvina aural comprehension. University of Michigan Press, 1972.  Ann  Enalish  Arbor:  Neville, N. H., & Pugh, A. K. Context in reading and listening: variations in approach to doze tasks. Readina Research Quarterly, 1976 1977, , 13-31. —  226 Nilsen, D. L. F. Contrastive semantic vocabulary instruction. TESOL Quarterly, 1976, j, 99-104. Oiler, John W., Jr. Contrastive analysis, difficulty, predictability. Foreign Lanuae Annals, 1972, 106. ,  and 95-  Oiler, John W., Jr. Scoring methods and difficulty levels for doze tests of ESL proficiency. Modern language Journal, 1972, 151—158. ,  Oiler, John W., Jr. doze tests proficiency and what they measure. 1973, 21, 105—118.  of  second  language  Language Learning,  Oiler, John W., Jr. Research with doze procedure in measuring the proficiency of non—native speakers of English: an annotated bibliography. Arlington, Va.: ERIC Clearinghouse of Languages and Linguistics, Center for Applied Linguistics, 1975. (ERIC ED 104 154) Oiler, John W., Jr. doze, discourse, and approximations to English. In M. K. Burt & H. C. Dulay (Eds.), New directions in second language learning, teaching, and bilingual education: Selected papers from the Ninth Annual TESOL Convention. Washington, D. C.: TESOL, 1975. Oiler, John W., Jr. and Conrad, C. A. The doze technique and ESLJ proficiency. Language Learning, 1971, 2J, 183196. Oller, J. W., Jr. & Perkins, K. (Eds.). Lanquage in education: Testing the tests. Rowley, Mass.: Newbury House, 1978. Oller, J. W., Jr. lanauage testina.  & Perkins, K. (Eds.). Research in Rowley, Mass.: Newbury House, 1980.  O’Reilly, R. P. & Streeter, R. E. Report on the development and validation of system a measuring literal comprehension in a multiple—choice doze format: Preliminary factor analysis results. Journal of Reading Behavior, 1977, (1), 45—69.  227  Pack, A. 1973,  C. ,  doze testing and procedure. 1—2.  TESL ReDorter,  Peterson, J., Paradis, E, & Peters, N. Revalidation of the doze procedure as a measure of the instructional level for high school students. , 1973, , 144—149. Peterson, J., Peters, N., & Paradis, E. Validation of doze procedure as a measure of readability with High school, trade school, and college populations. 1972, , 45—50. Pikuiski, J.J., & Pikulski, E,C. doze, maze, and teacher judgment. The Readinc Teacher, 1977, Q, 766-770. Propst, I. K. Jr., & Baldauf, R. B. Jr., A psycholinguistic rationale for measuring beginning ESL reading and matching doze tests. RELC Journal, 1981, (l), 85-89. Publication manual of the American Psychological Association. Baltimore, Maryland: Garamond/Pridemark Press, 1979. Ramanauskas, S. The responsiveness of doze readability measures to linguistic variables operating over segments of text longer than a sentence. Reading Research Quarterly, 1972, (l), 72—91. Ramanauskas, S. Contextual constraints beyond a sentence on doze responses of mentally retarded children. American Journal of Mental Deficiency, 1972, fl, 338—345. Rankin, E. F. The doze procedure revisited. In P. L. Nacke (Ed.), Interaction: Research and vractice for col1ee—adu1t reading, , 1974, , 1—8. Rarzkin, E. F. Characteristics of the doze procedure as a research tool in the study of language. ç, 1978, 27,149—153. Raymond, P. CANread: Reading skills in ESL. Scarborough, Ontario: Center for Second Language Learning, University of Ottawa, Prentice-Hall Canada Inc., 1987.  228 Reutzel, D. R. dozing in on comprehension: The doze story map. The Reading Teacher, 1986, 6, 524—528. Richards, D. laboratory.  Communicative interaction in the RELC Journal, 1980, 11(2), 64-76.  language  Rivers, W. K. Teaching foreiqn—lanquage skills. The University of Chicago Press, 1968.  Chicago:  Rye, J. doze procedure and the teaching London: Heinemann Educational Books, 1982.  reading.  of  Schwartz, R. K. and Stanovich, K. E. Flexibility in the use of graphic and contextual information by good and poor readers. Journal of Reading Behavior, 1981, (3), 263269. Shanahan, T., & Kamil, N. L. The sensitivity of doze to passage Organization. Zj, 1982, fl, 204—208. (In J. Niles & L. Harris (Eds.), New Inquiries in Reading Research and Instruction. Thirty-First Yearbook of the National Reading Conference, 1982, 204 208.) -  Shanahan, T., Kamil, N L., and Tobin, A. W. doze as a measure of intersentential comprehension. Reading Research Quarterly, 1982, )j, 229-255. Shanahan, T •, & Kamil, K. L. A further comparison of sensitivity of doze and recall to passage organization. 1jQ, 1983, , 123—128. Shanahan, T., & Kamil, K. L. The relationship of concurrent and construct validity of doze. ZjR, 1984, , 252—256. Schoenfeld, F.G. Instructional uses of the doze procedure. The Reading Teacher, 1980, 4., 147-151. Soudek, M and Soudek, ELT Journal, 1983,  L. I.  New uses in language teaching.  .3.Z, 4, 335—339.  229  Relationships among oral and written doze and Streiff, V. written doze scores and achievement test scores in a In J. W. Oiler & K. Perkins (Eds.), bilingual setting Lanauacie in education. Rowley, Mass: Newbury House, 1978. 65—102. Stubbs, J. B. and Tucker, G. R. The doze test as a measure of English proficiency. Modern Lanauaae Journal, 1974, , 239—241. doze and dictation tasks as predictors of Stump, T. A. intelligence and achievement scores. In J. W. Oller & K. Perkins (Eds.), Lanuae in education. Rowley, Mass: Newbury House, 1980. Taylor, W. L. readability.  doze procedure: a new tool for measuring Journalism Quarterly, 1953, , 415-433.  Recent developments in the use of doze Taylor, W. L. procedure. Journalism Quarterly, 1956, , 42—48, 99. Thomas, S. Contextual constraints validity of the doze procedure.  and the constraint 1980, 2, 47—55.  Tuinman, J. J. Speculations on doze as a search procedure. 1972, , 74—84. Valmont, W. J. Cloze deletion patterns: How deletions made makes a big difference. The Readina Teacher, 1983, fl(2), 172—175. Weaver, W. & Kingston, A. A factor analysis of the doze procedure and other measures of reading and language Journal of Communication, 1963, j3., 252—261. ability. The relationship Ya]nada, J. order and doze difficulty Journal, 1979, (l), 70—80.  between scrambled sentence among EFL students. RELC  230  APPENDICES FOR PILOT PROJECT APPENDIX A: PILOT PASSAGES  The Original Texts with Deletions Indicated  Pre-test:  Terry Fox  Test 2:  The Black Donnellys  Test 3:  The Doctor (Norman Bethune)  Test 4:  The Leader,  Test 5:  Creatures of the Wild (Sasguatch)  Test 6  Insulin  Post-Test:  Terry Fox  (Louis Riel)  Pre—test: A Young Man’s Dream  Terry Fox, born in Winnipeg, Manitoba on July 28, 1958, was Betty and Rolly Fox’s second child . The family of six l)moved to Port Coguitlam, near 2)Vancou ver when Terry was seven. 3)At school, Terry was always 4)involve d in sports and began 5)cross-country running as early 6)as the eighth grade. His 7)keen interest in all sports 8)led him to Simon Fraser 9)University where he began studying lO)kinesiology, the study of human ll)movement.  In November of  1976,  12)he was involved in a  crash which injured his 14)right knee.  13)car  Never a complainer,  231  l5)he one day collapsed in l6)front of his mother because l7)his pain was so intense.  18)At hospital he was diagnosed  l9)as having osteogenic sarcoma, a 20)rare, malignant tumour that develops 21)mostly in human males between 22)the ages of ten and 23)25. 25)knee  where  Eventually  renders  30)to  the  breaks  muscles,  bloodstream is  it  27)it  surrounding  cause  It is a bone 24)cancer that begins at the  be  through  sending carried  32)unknown.  26)bone  soft  the  bone  29)cancer all  Terry’s  over  leg  28)to  the  into  the  cells  3l)the  would  mushy.  and  body.  have  Its  33)to  be  amputated dust above 34)the right knee in order 35)to stop the cancer.  All 37)spoke  36)through elementary of  perseverance,  his  drive  tenacity,  school,  and  Terry’s  teachers  38)determination,  and 39)mental  toughness.  his  A hard  driving, 40)gutsy guy who never gave 4l)up, Terry decided to look 42)upon the loss of his 44)He  decided  46)leg  as  he  he had  48)his amputation,  could  be  been  43)leg as a  45)just  as  47)with two.  new challenge.  positive  with  one  The night before  he formulated his 49)idea to run across  Canada 50)to raise money for cancer research.  Later, he underwent a series of gruelling chemotherapy treatments to destroy the cancer cells in his blood. did not complain.  Terry  232  Test 2: The Black Donnellys  Arrival in Canada  James  Donnelly  arrived  in Lucan,  Ontario  from his  native Ireland in 1487.  He was accompanied by l)his wife  Johannah and his 2)sons,  James Jr.  married Johannah in 4)Ireland. woman-—extremely masculine,  and William.  9)beard  with  6) large hands  and even smoked a  Johannah ll)loved to fight. insignificant,  was  13)good  had  She was a strange-5)looking  shoulders 7)and certainly no beauty. grew a  3)He  In 8)later years she  l0)pipe.  Any dispute, reason  for  and broad  a  Both Jim and  12)no matter how brawl.  14)They  hated guns of any 15)kind and preferred clubs and 16)their own fists.  The Beginning of Trouble  Arriving in 17)Canada, where land grants were 18)easily obtained,  they  preferred  20)privately owned land.  to  19)settle  on  40  hectares  They simply 2l)took it over.  of  They  were 22)thus called “squatters”.  In eight 23)years Johannah  bore  24)sons-—John,  Patrick,  25)and Thomas --and a daughter,  26)Jenny.  five  more  Michael,  Robert  After ten years  the 27)land changed hands and the 28)new owner, whose name was 29)John Farrell, took James Donnelly 30)to court. had  to  3l)surrender  nearly  20  hectares  of  James  32)land  to  233  Farrell.  From  then  33)on  there  was  constant  battling.  34)Farrell found his cows poisoned 35)and his barn burned. Moreover,  36)one  day  while  sitting  Farrell felt a 38)bullet go by. than one 40)of the Donnelly boys. sin 42)as their father.  in  37)his  kitchen,  It could 39)be none other They 41)were as black in  Thus they 43)were called the black  Donnellys.  Prison  44)James was a hard drinker. one  46)day,  he  had  too  quarrelled 48)with him. with an 50)iron bar.  much  45)At a public gathering  47)to  drink.  John  Farrell  James hit Farrell 49)on the head It took Farrell three days to die.  James ran into the forest surrounding his home and hid there for two years. escaped.  The people  of Lucan thought that he  had  234  Test 3: An Exceptional Leader  The Metis  Louis Riel grew up on the east side of the Red River in St.  Boniface.  He  belonged  distinct people  2)of Native  The Metis were  semi—nomadic,  to  the  l)Metis--a  culturally  Indian and French 4)buffalo  3)ancestry.  each year  5)traveled West for their annual 6)buffalo hunt. 7)high point  of  the  year  8)for  them because  9)meat for the upcoming winter. ll)certain captains,  military soldiers,  l4)perform,  orders,  who  without  permission,  Catholic.  a  l2)appointed  had  a  job  to  l5)required to obey  Any participant who l9)did not follow  20)lagged was  behind  or  punished  advanced 22)because  Afterwards,  or  21)hunted  food  for  the  24)all meat from  25)was shared equally among the  family,  provided  it  No buffalo 17)could be hunted on  entire 23)year was at stake.  Louis’  were  l3)Each  and all participants were  the l8)Sabbath day.  the hunt  There  guides.  the rules l6)of the hunt.  It was the  lO)The buffalo hunt had  precision. and  in June  like most 27)Metis families,  26)participants. were fervently  28)Louis excelled in school and 29)was chosen for  religious studies.  30)In 1858,  when he was 31)13,  he was  sent East 32)to a Jesuit college in 33)Montreal to study for the 34)priesthood.  He remained in school 35)for six years.  Then he 36)left and took a job 37)as a clerk in a 38)law  235  office.  A year  39)he  later,  was  back  in  40)River,  Red  actively participating in the 41)Metis community.  Trouble in the Red River  At this time, 42)Red River belonged to the 43)Hudson’s Bay Company. 45)sell 47)and  The Hudson’s  Red River to white  Canada.  44)Bay Company had decided to  Canada.  settlers  there  46)No  one  48)they  if  asked the Metis wanted  to  ioin  49The) Metis feared that their 50)long-established  farms would be divided up because of new land surveys.  The  Canadian  new  government  sent  William  McDougall  the  as  lieutenant governor of Red River.  Test 4: The People’s Doctor  Norman Gravenhurst,  Bethune  Ontario.  Toronto Medical  was  born  March  on  3,  1890,  in  He enrolled at the l)University of  School  2)in  1909.  He  took  a  3)two-year  break from medical 4)school to work by day 5)at lumber camps and railroad 6)construction sites.  It was here 7)that he  first encountered the  8)harsh life of the workers.  returned  school  to  l1)surgeon in  medical 1916.  l0)and  He married  was  qualified  l2)and moved to  9)He as  a  Detroit,  Michigan 13)where he set up his 14)first medical practice.  236  In  Detroit,  15)Bethune  16)enormous  gap  became  between  the  acquainted  level  with  17)of  health  available to l8)the rich and to the 19)poor.  the care  Most of his  patients 20)were barely able to pay 21)him.  As  his  established,  medical  Bethune  practice  22)was  23)developed  becoming  tuberculosis.  securely In  1920  people 24)feared tuberculosis as they fear 25)cancer today. He was forced 26)to stop working and spent 27)a year at a sanitorium, painting. 31)his  28)recuperating.  At  this  time,  he  29)began  His despair during 30)this year is reflected in  paintings,  for  he  was  32)sure  he  was  dying.  To  33)make matters worse, his wife 34)divorced him.  Once fully recovered,  35)Bethune went to Montreal in  36) 1928 where he worked as 37)a tuberculosis (TB) specialist in 38)an English hospital.  He conducted 39)his operations  at high speed 40)in order to minimize operating 41)time and designed  several  new  42)surgical  surgical 43)procedures.  instruments  Thus he brought both  to  improve  44)critical  skill and creative style 45)to surgery.  Bethune hospital. poorer English  then  took  At this time,  49)and Canada.  less  well Just  46)  his  work  at  a  47)Quebecois  48)French Canadian hospitals were equipped as  he  had  than  50)those  noticed  the  serving glaring  disparity between the rich and the poor of Detroit, he now  237  saw the difference between health care in French and English Canada.  And he was determined to do something about it.  Test 5: Creature of the Wild  The earliest references to the Sasquatch are found on the carved totem poles and masks of the coast Indians of British  Columbia.  These  Indians  say  (meaning “wild man 2)of the woods”) ancient  race  However,  more than 750  4)which  has  that  l)the  Sasguatch  is the 3)remnant of an  managed  to  6)sightings of  avoid  5)capture.  the beast or 7)its  foot prints have been 8)reported in the past 100 9)years. These sightings have given 10)us a good idea of ll)what the Sasguatch might look l2)like.  In some ways the l3)Sasquatch seems to resemble human l4beings  Sasguatch  A  is  approximately  15)360  centimetres  tall, weighing between 16)270 and 360 kilograms.  It 17)is  completely covered with short l8)hair except for the palms 19)of the hands and the 20)soles of the feet. 21)humans have no hair on 23)hair on the 25)on  the  27)black,  rest  22)their palms  Sasguatch head 24)is of  the  26)body.  or  Similarly, soles.  longer than the hair  The  Sasquatch  face  and the eyes are 28)larger than a human’s.  29)Sasquatch  head  is  said  to  30)be  The  larger  at  the  is The  back  238  3l)than at the front. 33)ears,  Its 32)ears are comparable to human  its nose, broad and 34)flat.  Unlike a human,  35) Sasquatch has a very short 36) neck.  the  The arms are longer  37)than a human’s and reach 38)below the knees.  The hands  39)are  pendulous  massive.  The  females  have  40)large  breasts.  Like a 4l)huinan, the Sasguatch walks upright 42)on two legs, places its 43)heel down first, and swings 44)its arms as it walks.  45)It is extremely agile and 46)mobile for its  huge  47)The  size.  centimetres long,  Sasquatch  foot,  which  is  48)about  37  differs 49)from a human foot in 50)that  its ankle bones are enlarged to support its great weight. Yet there is much flexibility in the toes.  239  Test 6: A Different War  Before 1941 there were 23000 Japanese people in Canada --22000 in British Columbia and 1000 elsewhere.  Although  World War II did l)not really touch Canada, the 2)Japanese people  living  there were  Canadian government. 6)but,  all  families,  the  3)in  a  personal  There were  same,  5)no  war with  bombs  and  4)the  no  guns  they 7)lost everything--homes,  )obs,  8)pride —-because the Canadian government 9)took  it all away from 10)them.  The Japanese had been since  12)1877.  lumberjacks  found  They  on  the  14)west  ll)coming to British Columbia work coast  as of  15)sending later for their families. in 17)Vancouver, hours, “first  these  on Powell Street.  Japanese  20)generation”  who valued family honor. did something wrong,  British  and  Columbia,  16)Many of them lived Working 18)long,  l9)immigrants, in Japanese),  13)fisherinen  or  Issei  were good  hard  (meaning  21)citizens  22)If one person in the 23)family  the 24)vhole family suffered.  There  was 25)very little crime among Japanese 26)families for they would not 27)stand for it.  The Issei 28)did not like to be 29)in debt.  Often, if  they 30)did not have enough money 31)to buy something, they waited 32)until took good  they had enough money  34)care  of  their elderly,  33)saved up.  refusing  They  35)to place  240  them in old They  36)people’s homes to die,  considered  their  38)old  people  poor 37)and alone. to  be  the  39)most  important members of the 40)Japanese family who could teach 41)the young a lot about 42)life.  The Xssei also  looked  43)after their own poor, saying 44)that it was up to 45)them to ensure that they 46)had enough to eat and 47)a place to stay.  The 48)Issei preferred to live in 49)groups with other Japanese.  They  Canadians.  They wanted to keep their Japanese way of life,  50)did  not  want  to  change  to  be  like  so they cut themselves off from other people.  Test 7: The Discovery of Insulin  Looking back,  it must have  seemed  like  Banting was ‘ust 29 in l)that summer of 1921, not long out of 3)medical school. only  a  5)recent  laboratory,  graduate  grudgingly  loaned  University 8)of Toronto, poorly.  a 2)surgeon  Best was a 4)boy of 22, arts.  for  Their  7)three  was dark and  They lO)were not paid.  inexperienced young men, 13)to  of  a miracle.  6) research  months  9)humid.  by  the  They ate  Calculate the ll)odds: two  l2)badly  equipped,  change the face of l4)medical history.  with  90  days  241  As the world 15)knows well what Sir Frederick l6)G.. and  Banting  summer——l8)a chemical  Charles  Dr.  crude  derived  17)Herbert  extract from  of  20)the  Best  discovered  precious pancreas  that  19) insulin, and  capable  a of  21)controlling diabetes meflitus, a killer 22)disease as old as China.  23)The existence of insulin had 24)been suspected for more  than  25)a  decade  and  the  attempt  26)to  extract  isolate it 27)occupied researchers around the world. Banting and Best,  mongrel  famous  32)Marorie;  ——  33)insulin  roused  35)years.  Six months later,  first  28)But  experimenting 29)with diseased pancreases  removed from 30)diabetic dogs, were the first. one  and  her  37)verifying  from  its  a  a  34)coma  31)They made  shot  of  and  she  on 36)January 11,  safety  with  large  unpurified lived  for  1922——after 38)doses  on  themselves--the first 39)human diabetic was given insulin: 40)14-year-old hospital. another  Leonard  With 12  Thompson  42)insulin,  years  before  41) at  Thompson  44)dying  of  Toronto  went causes  on  to  General 43)live  unrelated  to  45) diabetes  The world at large 46)proclaimed them, Banting 47)and Dr. had  J.J.R. l4acleod,  although only  the physiologist 48)who  reluctantly made the 49)lab space available and then  50)left for a summer holiday, were cited for the 1923 Nobel  242  Prize in medicine.  Banting graciously shared his award with  Best.  If  Best  accorded him,  was  slighted  at  the  lack  of  recognition  and vexed by Macleod’s tendency to want all  the credit for his own, he nursed his bitterness privately.  .. .. ........  •...o................ ..........................  zq3 The Black Dc3nnellys  Arrival  in Canada  James Donnelly arrived in Lucan, native Ireland in 1487.  Ontario from his  He was accompanied by  wife Johannah and his 2)  1)  James Jr. and William. 3)  had married  Johannah in 4) 5)  .  She was a strange—  woman——extremely masculine, with hands and broad shoulders 7)  6)  certainly no beauty. a 9)  Both Jim and Johannah 11)  .  to fight.  she grew  and even smoked a  -  10)  In 8)  Any dispute,  12)  matter how  insignificant, was 13) brawl.  114)  reason for a hated guns of any  and preferred clubs and 16)  15)  own fists.  The Beginning of Trouble  Arriving in 17) 18) 19)  owned land. They were 22) 23) 24)  ,  obtained,  where land grants were  they preferred to  on ‘40 hectares of 20) They simply 21)  it over. called  squatters’.  Johannah bore five more Patrick, Michael,  Robert  In eight  25)  Thomas ——and a daughter,  26)  After ten years the 27)  changed hands and the 28)  owner, whose Farrell,  name was 29) court.  30)  James had  took  to 31)  nearly 20 hectares of 32)  to Farrell.  From then 33) 34)  there was constant battling. found hi8  -  his barn burned.  COWS  poisoned 35)  Moreover, 36)  sitting in 37)_  day while  kitchen, Farrell felt a go by.  38)  It could 39)  none other than one 40)  the Donnelly  They 141)  boys.  Donnelly  James  as black in sin  42)  their father.  Thus they  143)  called the black Donnellys  Pr i son  ‘44)  was a hard drinker.  45)  a public gathering one ‘46)  he had too much 47) quarrelled 48)  him. the head with an  49)  bar.  drink.  John Farrell  James hit Farrell 50)  It took Farrell three days to die. James ran  into the forest surrounding his home and hid there for two years. he had escaped.  The people of Lucan thought that  An Exceptional Leader  The Metis  Louis Riel yre up on the east side of the Red River  in St.  Bonifacs.  1)  He belonged to the  culturally distinct people  2)  Native  Indian and French 3)  The Metis ere semi—nomadic, 4) in June 5)  each year  West for their annual hunt.  6)  It as the 7)  point of the year 8)  them because  provided 9)  for the upcoming winter. buffalo hunt had  10)  military precision. soldiers,  There were  and guides.  14)  ,  15) hunt.  a 11)  12>  captains..  13)  had a job to  and all participants vere to obey  the rules  No buffalo 17)  16)  day.  19)  not follow orders.,  Any participant NHo ho 20)  behind or advanced or 21)  without permission.  as punished 22>  food for the entire  23)  as at stake.  24)  —  Afterwards,  meat from the hunt 25)  shared equally among the 26) like most 27) 28)  the  be hunted on the  18)  Catholic.  it  .  families, excelled  Louis’  were fervently in school and  family,  29)  chosen for religious studies. 1858,  30)  when he was 31)  he was sent East 32)  —  in 33)  to study for the  He remained in school 35)  .  34)  six years. job  Then he 36)  and took a a clerk in a 38)  37)  office.  A year  39)  later,  Red ‘40)  was back  in  actively participating in the  ,  community.  41)  Trouble  in the Red River  At this time,  42)  the 43)  River belonged to Bay Company.  The Hudson’s  Company had decided to 45)  44)  Red River to Canada. Metis 47) 48)  a Jesuit college  46)  one asked the white settlers there  wanted to join Canada.  if  49)  Metis feared that their 50) would be divided up because of new  farms land surveys.  The Canadian government sent William McDougall as the new  lieutenant governor of Red River.  27  The People’s Doctor  Gravenhurst,  School  of Toronto Medical  the  He enrolled at  Ontario  1) He  1909.  2)  break from medical  took a 3)  to work by day 5)  ‘4)  he first encountered the  was here 7)  life of the workers.  8)  returned to medical  school in  11)  was qualified He married  1916.  he set up his  13)  practice.  acquainted with the  9)______________  10)  moved to Detroit,  12)  medical  It  sites.  lumber camps and railroad 6)  as a  in  1890,  Norman Bethune was born on March 3,  Michigan  114) became  15)  In Detroit. 16)  —  gap between the  level  health care available to 18)  17)  rich and to the  Most of his  19)  barely able to pay  patients 20)_________ 21)  becoming  s his medical practice 22) securely established,  tuberculosis as they fear  1920 people 2’4) today.  25)  tuberculosis.  E1ethune 23)  He was forced 26) year at a  stop working and spent 27) sanitorium, 29)  28)  .  painting.  At  this time,  he  His despair during  In  30)  f  .  ca intl nos  tor  he  rj  j  r  32)  3 he es ciy i ncj  matttre. vJorse  his  -i  e  Hi m  1. v re c:c’ver e ci ::; a  I  35  •  ie r 1:  a I••i a  or ka  3.;7 )  t. be i c::u I. cs is  38)  imi ish hosp :1  39)  c:  or  cler  to ml  sever a I  rLi.L ze  a r.:ce ate .1 is t:. He  :t  cordtc..tei .) )  operati nq 41. )  end deal qned natru.ments to  imorc::ive sur q ice].  Thus he brou.qrit both 44j  43) 1  T:B )  atic:ns at h:i. gh speed  new 42)  to Hontr so I  1. 3.  onc: creat. 1  St\/  VS  Is: 4j  OLtl  then toc:d 47  his 52.5. S .L r-  t this t. I.  Cane di an hc:s p1 to I a e re poorer  48)  no E.Enq i.  :i.  sh  p1 ar I n p d : a r.:: [ t;etr c: :1. t  cii  ]as  —  node  he nc::w saw  as h o ci not : c:o ci t  J  t.y be tw n  the  t ‘i.c::  it.  and t. he pc::o r  di ffe enc:e set  n French anci Enpl i sh Cenecie cic:sc:meth :1 ng about  E1LIJF4td  h i  Lb  c:a re  ndhs was determ :1 ned  zLt1  Creature of the Wild  The earliest references to the Sasquatch are found on the carved totem poles and masks of the coast Indians of British Columbia,. 1)  Sasquatch  These  ( meaning ‘wild man  the woods’)  2)  of an ancient  race  is the 3)  14)  has managed to  avoid 5)  .  However,  more than 750  of the beast or 7)  6)  prints have been 8) 9)  look  approximately  idea of  seems to .  15)  and 360 kilograms.  19)  20)  is weighing between  17)__ except for  the hands and the of the feet.  Similarly,  21)  palms or soles.  The 23) 214)  It  18)  have no hair on 22)  on the Sasquatch head longer than the hair 25)  the rest of the 26> is 27)  Sasquatch  1  centimetres tall,  completely covered with short the palms  the  13)  l’4)  16)  11)  12)  In some ways the resemble human  100  These sightings have given a good  Sasquatch might  foot  in the past  .  10)  28)  Indians say that  The Sasquatch face •  and the eyes are  than a human’s.  The  ___  250  head larger at the baoi.:. Its  ‘  —  is sa i d to  31)__________  i ts roan  hum  broad and 24)  the .3)  nsa. a vary  short  The arms arc: a  the front  aL  compo ab) a  —  Un ii k a a human  30  :I.c:nqer  hi..rransanc:  she knees  The hanis 29)  T he toma 1 as. have 40)  mass i  pendu :L Qua  breasts  Like a 4.!. 42) dc:wn  •  to +i  rat  ieqs  pisces  arms as  for  %as.qusl:c::h foc:t  ii ci  its huqe siza .s 18  d:i.fsrs 49)  47)  •___________.  27  a huiran foot  •••••t  in  i ts an k 1. a bores are en L ar qed  n the  I is  _c.€.o;  it  is a::<trema I y aqi La and  4é)  suqpc:r t  ciht  its 4%)  and swi nqs 44)————.—  4%)  lDng  the Oz:c!:;..I..ch walks nor  qreat wc I riOt  Yet there  ta  :1. a mucri f lexi bi I :i  e  251  t  Different War  Before  19141  there were 23000 Japanese people  Canada ——22000 in British Columbia and lthough World War Canada,  II  did  really touch people  living there were  a personal war with  Canadian government.  1000 elsewhere.  1)  the 2)  3)  in  14)  There were 5)  no guns 6)  ,  all  bombs and  the same,  they  7)  everything——homes,  8)  -—because the Canadian government it all away from  9)  The Japanese had been  later for  16)  of  ,  on Powell  18)  ,  hard hours,  lsssi  Japanese  Street.  them  lived in  Working  these Japanese  ( meaning “first 20) )  ,  22)  23)  family suffered.  who valued one person  did something wrong, There was 25)  among Japanese 26)  19) in  were good 21)  family honour.  27)  114)  15)  17)  or  They found work as  lumberjacks on the  coast of British Columbia, their families.  to British .  and  10)  11)  Columbia since 12) 13)  families,  Jobs,  in the  the 214) little crime  for they would not for  it.  _  The  Issei  28)  not  29)  debt.  Often,  like to be  if they 30)  not have enough money 31)  buy something,  they waited 32)  they had up.  33)  of  their elderly,  They took good 34)  refusing 35)  in old 36) 37)  enough money  place them  homes to die, —  alone.  people to be the 39) the ‘40)  poor  They considered their 38) important members of  family who could teach young a lot about 42)  41)  Issei also looked  143)  saying 44)  .  their own poor, it was up to  ‘45)  to ensure that they 46)_  enough to eat and 47)  The ‘48)  place to stay.  preferred to  live  149)  with other Japanese.  50)  not want to change to be  Canadians. life,  in  They like  They wanted to keep their Japanese way of  so they cut  themselves off from other people.  The  ______ _______  23  The Discovery of  Insulin  Looking back, Banting was 1921,  it must have seemed  just 29 in  l)_______________ summer of  a 2)  not school.  3)  of 22.  like a miracle.  long out of  Best was a  only a 5)  4)  graduate of arts.  Their 6)  laboratory,  grudgingly  loaned for  months by the University 8)  7)  Toronto,  was dark and 9)  poorly.  They  .  10)  11)  not paid. :  change the face of  Frederick  Calculate the  two inexperienced young men, equipped,  12)  As the world  They ate  with 90 days  14)  13)  history.  15)  well what Sir  16)  .  Banting and Dr.  Charles  17)  Best discovered that summer——  18)  crude extract of precious ,  19)  a chemical derived from 20)  pancreas and capable of 2l)_,_ a killer 22)  23) 24)  diabetes mellitus,  as old as China.  existence of  insulin had  suspected for more than 25>  decade and the attempt 26)  extract and  2.53 Cs.  :i sc3:L.c.e  i  :  Z)  —c:r :rr s .rs:’:1 Snd  Dflt.L1 C  diseased  do cis one no nn  do  .L  panc:reesec-:.  iatE•:r.  1 ivacJ  on  an  Ia  di. as  Loorsar ci  3. VOO  r  1 L3 c  3.  t:  .  .  .11.  .  th  t [35 t  1 r  L*:S::.—t?  iarqe  st.  4(1)  ——\/e r  41  Toronto henoral  I ci  riospi tel  Thom3scjn went on to  •  1  another  .  .  V55”  a S  i  riocna.o  nsu 1. i n  With 42) 43)  fr  its  .  a a hot of  -  ...  cv rat  -rom  —  rc..sod and aiio  munt:ho  removed  ..2)  -—  unpur if-i ed  Si::<  e;<1er inenti nq  d25t  rc the -F :i. r s t,  .,  famous  ur i. d  1:h  ._S::c____.i.  year  S  causes unrelated to $ñ  -::  rI—E.  ares  Rit!if  45)  onl’’ Ianti rq c:: 1.  Dr 33. P.  1 at 48.)  had  Tiac:leod  re.L uc tent :t  •  made the  ava lable end then f:f) a. saiio r Nc:ib5  I  P  ho 1:1 iiN/  .i. ;:‘ei n  mc  cii  •  we r a a i. na  j:j  I3e n t i np  .f:c_...  the  I  qr a c: i ous.l.  V  SPr ad  h:a  SOS  a. ii cjhter  If Dr-si: cc:c:r dad hi in all 1:3  i  t hec:r  ci f  end  sexed  :.r  1: tar ness pr I vats 1.  [35  hi a sw n  .:  the  I ask  reren I Li on  Meal sod a tenc:ienc:v he  nor as c: c—  15  to wa it  TheBlackDonnellys  rrival  !HHH!!!  in Canada  James Donnelly arrived native 1)  Ireland  JL1i  and William. in  3)______  __.5.Qfl5_ had married  She was a strange—  woman——extremely masculine,  0 ) 6 J b j  no  beauty.  9)b€&r__ Pny  insignificant,  In 8)  Both Jim and Johannah  dispute, was  she grew  and even smoked a .  to fight.  with  7)____fl_  hands and broad shoulders  certainly  brawl.  He was accompanied by  4)_Xff]4$ji_.  5)Jo4cJ  a  Ontario from his  w i fe Joha nnah and hi s 2)  James Jr. Johannah  1467.  in  in Lucan,  l2)__fl______ matter  13)__  14)  ——  reason for  how a  hated guns of any and preferred clubs and  15)<Ij own fists.  The Beginning of Trouble  rriving  in  17)_CO.flê&....., obtained,  land.  They simply  land grani were  they preferred to  on 40 hectares of owned  where  21).hfIk  2O)ft1.VDifl —  it over.  They were 22)__j___ called “squatters’. 23)_Q_L___ Johannah bore five more 24)_jê5___——John,  Patrick,  Michael,  Robert  In eight  25)____ Thomas ——and a  26)_ny___.  fter  changed hands and the  daughter,  ten years the  28)__fleIA____ owner, whose  name was 29)_..Ah___ Farrell, 30)______  James had  court.  33___tL  From then  34)1O!.tCLII__ his  barn  sitting  burned.  none other boys.  Moreover,  go  ——  I t  by.  day while  3S)On..e kitchen,  Farrell  cou 1 d  __b..e  than one 40)______ the  their black  battling.  found his cows poisoned 35)_fl.c[  father.  Thus  felt a  Donnelly  in sin  black  as  They 4l)jAJ€f.  called the  to Farrell.  there was constant  in 37)_j  38) _L!?JJ€.±’  to 3l)_.$fV!ey  IOF%ti  nearly 20 hectares of 32)  James Donnelly  took  they  Donnellys.  Pr i son  44)__.f____ as a hard drinker. 45)_  a public gathering one 4b)__y____  he had too much 47)___O quarrelled 48)__J  him.  It took Farrell  three days  into the -Forest surrounding hi for  drink.  John Farrell  James hit Farrell  the head with an  49) bar.  _  —  two years.  he had escaped.  to die.  James ran  home and hid there  The people of Lucan thought  that  255  The Cloze Passages and the Distribution of the Deleted Words.  1. Terry Fox (ord  =  order of words in the passage)  ord word  location  1A 2 Young 3 Man’s 4 Dream 5 Terry 6 Fox, 7 born 8 in 9 Winnipeg, 10 Manitoba 11 on 12 July 13 28, 14 1958, 15 was 16 Betty 17 and 18 Rolly 19 Fox’s 20 second 21 child. 22 The 23 family 24 of 25 six 26moved 1: 27 to 28 Port 29 Coquitlam, 30 near 31 Vancouver 1: 32 when 33 Terry 34 was 35 seven. 36 At 3:: 37 school, 38 Terry 39 was 40 always 41 Involved 2: 42 in 43 sports 44 and  bef bef aft aft  0000  o  144,(111)  83  0 0 0  0 0 0 1(1  0 0 0 1  256  45 began 46cross1: 000 0 47 country 48 running 49 as 50 early 51 as 6:: 49, 237,247,252,(1 16) 0 1 0 3(1 52 the 53 eighth 54 grade. 55 His 56keen 1: 0000 57 interest 58 in 59 all 60 sports 6lIed 1: oooo 62 him 63 to 64 Simon 65 Fraser 66 University 1: o 0 0 0 67 where 68 he 69 began 70 studying 71 kinesiology, 1: 0 0 0 0 72 the 73 study 74 of 75 human 76 movement. 1: 0 0 0 0 77 In 78 November 79 of 80 1976, 81 he 9:::68, 113,243,253,263,278,(96,241 1 0 0 5(2 82 was 83 involved 84 in 85a 86car 1: oooo 87 crash 88 which 89 injured 90 his 91 right 2: 192 0 0 0 1 92 knee. 93 Never 94a 95 complainer, 96 he 9:::68, 113,243,253,263,278,(96,241 1 0 0 5(2 97 one 98 day  257  99 collapsed 100 in lOifront 1: 0 0 0 0 102 of 103 his 104 mother 105 because 106 his 10::55,90, 103, 208,212,235,265,292, (261 2 1 0 5(1 107 pain 108 was 109 so 110 intense. 111 At 3:: 144,(36) (1) 0 0 1 112 hospital 113 he 114 was 115 diagnosed ll6as 6::49, 237,247,252(51) 1(1 0 0 3 117 having 118 osteogenic 119 sarcoma, 120a l2lrare, 1: 0 0 0 0 122 malignant 123 tumour 124 that 125 develops l26mostly 1: 0 0 0 0 127 In 128 human 129 males 130 between 131 the 15:::22,52,72/145,150,159,162,169,198,232,25 3 0 0 9(2 132 ages 133 of 134 ten 135 and 13625. 1: 0 0 0 0 137 It 138 is 139a 140 bone 141 cancer 5:: 199,275,289,(166) 0 0 0 3(1 142 that 143 begins 144 at 145 the 146 knee 3:92, 193 1 0 0 1 147 where 148 it 149 renders 150 the 151 bone 3: 140, 160 0 1 0 1 152 soft  258  153 and 154 mushy. 155 Eventually 156it 3: 137,148 2 0 0 0 157 breaks 158 through 159 the 160 bone l6lto 3:27,63 2 0 0 0 162 the 163 surrounding 164 muscles, 165 sending 166 cancer 5:: 199,275,289,(141) 0 0 0 3(1 167 cells 168 into 169 the 170 bloodstream 171 to 7:::: 229,267,286(186,196,271) 0 0 0 3(3 172 be 173 carried 174 all 175 over 176 the 1 5:::22,52,72, 145,150/159,162,169/198,232,25 5(1 3 0 4(1 177 body. 178 Its 179 cause 180 is 181 unknown. 1: 0 0 0 0 182 Terry’s 183 leg 184 would 185 have 186to 7:::: 229,267,286,(171,196,271) (1 0 1 3(1 187 be 188 amputated 189 just 190 above 191 the 15:::22,52,72,145,150,159,162,169/198/232,25 8(2 0 1 3 192 right 193 knee 194 in 195 order 196to 7:::: 229,267,286,(171,186,271) (1 (1 0 3(1  197 stop 198 the 199 cancer. 200M 201 through 2: 158 202 elementary 203 school, 204 Terry’s 205 teachers 2o6spoke 1:  1 0 0 0  0 0 0 0  259  207 of 208 his 209 drive 210 and 211 determination 1: 212 his 213 perseverance 214 tenacity, 215 and 216 mental 1: 0 0 0 217 toughness. 218A 219 hard220 driving, 22lgutsy 1: 0 0 0 222 guy 223 who 224 never 225 gave 226up, 1: 0 0 0 227 Terry 228 decided 229 to 230 look 23lupon 1: 232 the 233 loss 234 of 235 his 236 leg 3:: 183, (251) 1 0 237 as 238 a 239 new 240 challenge. 241 He 9::: 68,113,243, 253,263,278,(81,96 242 decided 243 he 244 could 245 be 2:189* 246 just 247 as 248 positive 249 with 250 one 251 leg 3:: 183,(236) 252 as 253 he 254 had 255 been 256with 2: 249 257 two. 258 The 259 night 260 before  0 0 0 0  0  0  0  0 00 0  0 (1  2(2 0 1 3  0 0 0 1  1(1 0 0 0  0 1 0 0  260  261 his 10::55,90,103,208,212,235/265/292,(106) 262 amputation, 263 he 264 formulated 265 his 2661dea 1: 0 0 0 0 267 to 268 run 269 across 270 Canada 271 to 7::::229, 267, 286,(171,186,196) 1 272 raise 273 money 274 for  275 cancer 276 research. 277 Later, 278 he 279 underwent 280 a 281 series 282 of 283 grueling 284 chemotherapy 285 treatments 286 to 287 destroy 288 the 289 cancer 290 cells 291 in 292 his 293 blood. 294 Terry 295 did 296 not 297 complain.  6(1 1 0 1  1 0 1(3  261  An Analysis of Same-Word Re-Iterative Clues TABLE A AN ANALYSIS OF RE-ITERATIVE CLUES: LOCATION AND WEIGHTED VALUES FOR SCORING THE CLOZE PROCEDURE 3. An Exceptional Leader: (50 deletions, 299 word passage) no. of clue words...BEFORE= AFTER ....sentence of doze blank B W:W B nit: (clue word) not in passage e i:i e bey: (clue word) beyond sentence y t:t y bed: (different clue) beyond sentence o h:h o win : (due word) within sentence n i:i n wid: (different clue) within sentence d n:n d SCORES FOR CLUE LOCATION  NOUNS and PRONOUNS:  VERBS  1) Metis 1 00 3(2)= 7/6/5 bey 5) traveled— = 10/9/8 nit 3) ancestry = 10/9/8 nit 14) perform = 10/9/8 nit 6) buffalo 0002 = 7/6/5 bey 15) required— = 10/9/8 nit 9) meat 000 1 = 7/6/5 bey 17) could = 10/9/8 nit 13) Each 10/9/8nit 19)did = 10/9/8nit 18) Sabbath = 10/9/8 nit 20) lagged = 10/9/8 nit 23)year 2002 = 7/6/5bey2l)huntedldood = 7/6/Sbey 26) participants2 00 d = 7/6/5 bey 25) was 3004(1) = 7/6/5 bey 28) Louis 2d0 00 = 7/6/5 bey 29) was 4003(1) = 7/6/5 bey 33) Montreal = 10/9/8 nit 36) left = 10/9/8 nit 34) priesthood = 10/9/8 nit 45) sell 10/9/8 nit 38) law = 10/9/8 nit = = = = = = = = = = = = = = = = = = = = 39) he 5000 = 7/6/5 bey 11 verbs/50 deletions 40) River 1 004 = 7/6/5 bey = 22% 42) Red 3002 = 7/6/5 bey 43) Hudson’s 000 1 = 7/6/5 bey 44) Bay 1 000 = 7/6/5 bey 48) they = 10/9/8 nit —  —  —  —  —  —  —  —  —  18 nouns/50 deletions = 36%  —  —  —  —  = =  262  ADJECTh’ES  ADVERBS  7) high = 10/9/8 nit 10)the xOOx = 7/6/5bey 11) certain = 10/9/8 nit 0 adverbs/SO deletions 12) appointed = 10/9/8 nit = 0% 24) all 1000 = 7/6/5bey 27)Metis 2(1)002(1)= 7/6/5bey 31)13 =10/9/8nit 41)Metis 4002 = 7/6/5bey 46) No 1 000 = 7/6/5 bey 49) the x 001 = 7/6/5 bey 50) long = 10/9/8 nit —  —  —  —  —  11 adjectives/SO deletions = 22%  NIT BEY(D) WIN(D) NOUNS + PRONS = 36%> 14% 22% 0% 2) of 1 003(1) = 7/6/5 bey 4) and 1 005 = 7/6/5 bey VERBS = 22%> 16% 6% 0% 8)for 1013(1)= 7/6/5bey 16) of 2(1>002 = 7/6/5 bey ADJS = 22%> 10% 12% 0% 22) because 1 00 1 = 7/6/5 bey 30) In 30 1 5 = 5/4/4 win ADVS = 0%> 0% 0% 0% 32)to 1002 = 7/6/5bey 35) for 5(1)000 = 7/6/5 bey OTHER = 20%> 0% 18% 2% 37) as = = = = = = = = = 000 1 = 7/6/5 bey = = = = = = = = = = 47) and 4(1)000 = 7/6/5 bey all = 100% ) 40% 58% 2% OTHER  10 other/SO deletions =20%  = =  = =  = =  = = = = = =  263  APPENDIX B: PRODUCTION OF THE PILOT CLOZE PASSAGE FORMS  The following procedure sets out a reasonably efficient way to make the various modifications of the doze passages the present  research.  As  technology  improves  the  in  process  should become even easier.  TABLE B METHOD OF COLOR-CODING AND PRINTING PASSAGES FOR FIVE  CLASSES How to make the standard doze procedure passages 2. 1.  —  Create a  Type  for doze  8. file and call  passage, e.g. 2.  Format  out  program,  it by the name  of the doze  “Fox”.  the  doze passage using  entitling  it the  name  a word processing  of the passage,  e.g.  Fox Passage. 3. Make a copy of the passage within the file and label the copied portion (Fox)*blank. 4. Using (Fox)*blank, a) put an b)  then  *  go  before every word which is to be deleted; to  replace the C)  the *  search  replace  function  and  with a 15 space blank;  press “all” and all the *s will be replaced with a 15 space line;  d)  and  save the amended file.  264  5. Make a copy of the completed (Fox)*blank and call the new portion (Fox)Form—l. 6. Using (Fox)Form-l, a)  delete one by one the words  immediately after the  blank; b)  then  start  at the  first blank  and number  it  two  spaces before as “1)”; c)  continue  numbering  “3)”, etc.  the  subsequent  blanks  “2)”,  (This will give you the standard doze  procedure Format);. d)  save the amended file.  ((Fox)Form-l is now ready to  be printed.) B. How to make Form 4 (the standard doze Format with colorcoded doze blanks and words) for passages 2 7. —  1.  Within the file,  copy  (Fox)Form-l.  and call the copy  (Fox)—Form 3/4. 2. Using (Fox)Foriu-3/4, a)  before  each  word  or  blank  designate  a  color  according to the part of speech represented by that word or blank; b)  (Color designations made with a Star color-printer require intended  a  printer color  code  change.)  to  be using  coding from Table B.l below,  placed the  before  the  appropriate  type the appropriate  abbreviated symbol before each word and blank.  265  TABLE B.l MEANINGS OF THE SYMBOLS USED FOR CODING  Part of Speech Symbol  Color  Abbreviated  Printer  NOUN  blue  ((C))2.....  VERB  red  ((C))l  ADJECTIVE  green  $  ADVERB  violet  OTHER  black  C)  3.  Code  ((C))6 ((C))3  A  ((C))O  save the file.  To make Form 4,  copy  (Fox)Form-3/4 and call the copy  (Fox) Form-4. 4. Using (Fox)Form—4, a) use the search—and-replace function to replace each abbreviated symbol with the printer code necessary for the printer to color the words and blanks;  b)  when all the printer codes are in place, file.  (Passage  (Fox)Form-4  is  now  save the  ready  to  be  printed on the Star color—printer.) D. How to make Form 3 (the standard doze procedure with all blanks and words colored randomly). 1.  Make  another copy within the  file  of  (Fox)Form-3/4,  before each word or blank randomly designate a color using the codes for the Star color printer. 2. Call the new copy (Fox)Form-3. E.  How to make Form 2 (the standard doze with all blanks color coded) for passages 2 7. -  266  i. copy (Fox)Form-3/4 and call the new copy (Fox)Form-2 2.  In  order  to  set  the  up  printer  codes  to  color  the  blanks, a) after each abbreviated symbol for the blanks insert a “b”; b)  then  go  to  search—and—replace  the  function  and  replace each (symbol)b with the appropriate printer code as in the code table above; c) 3.  save the file.  Next to remove the remaining unnecessary abbreviated symbols, a)  go to the search-and-replace function and replace each symbol with a  b)  save the file.  “  (Passage  (Fox)Form-2  is now ready  for the Star color—printer.) F.  How to print the four variations of the standard doze passages for passages 2 —7. 1.  doze  passage  Form  1  can  be  printed  on  any  IBM  compatible printer. 2.  Forms 2,  3 and 4 can be printed on the Star NX1000R,  NX1000CL, or LC1OCL color printers.  U’) t  268 G. The number of copies of the Forms to be made. Form 4  =  25 copies/passage  all color coded  Form 3  =  25 copies/passage  all randomly colored  Form 2  =  25 copies/passage  blanks  Form 1  =  25 copies/passage  nothing colored  Subtotal  =  100 copies/passages  x  Subtotal  =  6 passages (passages 2 to 7)  600 copies  + 200 copieEd(lOO Form 1  Total  =  colored  pe1&.ett)  800 copies for 100 students  269  APPENDIX C: DISTRIBUTION OF THE RESEARCH TASKS The Role of the Researcher TABLE C ROLE OF THE RESEARCHER 1. Before day 1 —  the researcher will prepare the tests, booklets,  and the pencils.  the the test  He will also orient the  teachers as to the duties during the test sessions.  2. Day 1: a)  before  the  class  the  researcher will  give  the  test  booklets and the pencils to the teachers. b)  after the test the researcher will  collect the test  booklets and pencils. c)  then the researcher will type test  answers  data base score sheet for each student, scores for each blank  the  exact  appropriate  score,  group  and  calculate the  (both exact and weighted),  the scores to get totals, to  into the  add  rank the students according  assign rank  the  students  within  the  to  group,  their and  choose the passage form accordingly. d)  then the researcher will take out test one  from the  students’ booklets and replace with the assigned form of test 2.  270  Day 2: a)  before  the  class  the  researcher will  give  the  test  booklets and the pencils to the teachers. b)  after the test the researcher will  collect the test  booklets and pencils. C)  then the researcher will type test  answers  data  scores  base  score  sheet,  calculate  into the for  each  student. d)  next the researcher will take out test one from the students’ booklets and replace with the assigned form of test 3.  4. Day 3 through day 7: a)  the researcher will give to the teacher 20 copies of the undeleted passage of the previous test  b) the researcher will repeat the same routine as for day 2, except he will go to the subsequent test.  5. Day 8: a)  before the class the researcher will give the teacher the test booklets  .b)  after the class the researcher will thank the students and teacher.  C)  then he will collect the booklets, pencils.  questionnaires and  271  6. Finale: a)  the  researcher will  take the data and calculate  it  all, analyze it, and draw conclusions. b) the researcher will meet with the teachers and discuss the data and conclusions.  272 Instructions for Teachers TABLE D INSTRUCTIONS FOR TEACHERS  ADMINISTERING THE CLOZE RESEARCH  I • GENERAL STATEMENT  This  research  requires  eight  sessions,  two  per week.  Each session should take about 45 minutes to administer the day’s doze passage, which includes 30 minutes for the test and  15  minutes  for  instructions,  collection of the tests.  distribution,  and  The task of the teacher is to hand  out pencils and test booklets,  to read the instructions to  the students given in their booklets, watch the students to see  they  informed  are of  doing the  their  remaining  own work, time  collect the booklets and pencils, test  onward  to  hand  out  the  to  (every  keep  the  five  minutes),  students to  and after the third doze  previous  sessions  undeleted  passage which after five minutes is to be collected again.  II. DETAILED STATEMENT  Day 1: a)  Please explain to the students that they are going to help  in  will  be  research  an E.S.L. anonymous, as  an  research project, and  that  they  official  test  and  that their work must act  treat  according  the to  273  standard  test  behavior,  i.e.  no  collaboration  with  other students. b) Tell the students that there will be eight tests, each of which they will be given half an hour.  for Also  tell them that if they can get 20 blanks correct out of 50, then they are doing reasonably well. C) Hand out the pencils. d)  Hand out the booklets, them  closed  until  telling the students to keep  they  are  asked  to  turn  to  the  students have a test booklet tell  the  instructions and suggestions on page 1. e)  When all  the  students  to  initials  of their mother or  write  their booklet.  a  three-figure  number  father on the  plus  the  front  of  Tell them to keep a copy of the number  and initials in their purse, wallet, etc. for the next sessions when the booklets will be returned to them. f) Read the instructions to the students as they silently read along and make sure they understand them.  (Try  to remember their questions for subsequent recording.) g) Read the guessing strategy to them as they read along silently.  Answer any questions they have about the  strategy.  (Try  to  remember  their  questions  for  subsequent recording.) h) Tell the students to open their booklets and to begin. Remind them they have 30 minutes to do the best they can on the test and that 20 blanks correct out of 50 is a reasonable score.  274  i)  Watch the students to work.  j)  see they are doing their own  Do not allow them to use their dictionaries.  While they are doing the test, i) write down on the back of this sheet the questions the students asked you about the instructions and suggestions, ii)  write down the names absent,  of  the students that  are  (or their number-initial code), and  iii) write on the board every five minutes the amount of time left for the test. k)  After  30  minutes  tell  the  students  to  put  their  pencils down and to close their booklets. 1)  Collect  the  booklets  and  pencils  and  remind  the  students to remember their number—initial code. m) Thank the students for participating in the day’s test and continue with your regular lesson. n)  Keep the pencils and booklets for the researcher who will pick them up on the same day.  0)  When the researcher comes, ask any questions and give any suggestions or concerns you might have about the test and give him the booklets and pencils.  Day 2: p)  Please follow steps c, this  session  the  d,  and f through o.  researcher  will  give  Before you  the  275  students’ booklets. the  students  Display them on a table and ask  one by one to tell you quietly his/her  number-initial  code  and  then  hand  him/her  the  q)  Please follow steps c through 1.  r)  Hand out copies of the undeleted doze passage  from  corresponding booklet from the table.  Day 3:  the previous day. s)  Let students look at the undeleted passage for about five minutes.  Remind them that if they got 20 blanks  right out of 50, then they are doing reasonably well. t)  Collect all the undeleted passages and keep them for the researcher.  U)  Follow steps m through o.  Day 4 through 7: v)  Please follow the routine of day 3.  Day 8 w) Please follow steps c through 1.  (Do not hand out the  undeleted passage from day 7.) x) Hand out a questionnaire to each student.  276  y)  Give the students about  10 minutes to  fill them out  and then collect them. z)  Follow steps m through o to complete your part in the research project.  Thank you very much for your co—operation.  You will receive  a summation of the results when this thesis is completed.  277  Instructions for Students TABLE E INSTRUCTIONS FOR STUDENTS  Thank you for participating in this research project. Today you are going to have one of eight doze tests that will be given to you in a period of four weeks. is to of.  Your task  fill in the blanks with the best word you can think Do your best.  If you can get 20 blanks correct out of  50, then you will have done reasonably well.  A. GENERAL INSTRUCTIONS: 1. Please do not put your name on the booklet.  Instead,  think of any three figure number, eq. 123 and write it down  in the upper  this  booklets  left hand corner  cover.  Before  your  initials of your father or mother, George Smith. This  method  on the  number write eg,  FGS  The example looks like this: will  identify  your  front  booklet,  of the  for Frank FGS  123.  but  will  protect your identity. Please make will  need  it  a  note  for  the  of your other  identity code  tests  in  this  as  you  research  proj edt.  2.  For this doze test you will have 30 minutes.  Every  ten minutes your teacher will write on the board how much time there is left for you to complete the test.  278  3. As this test is for research purposes it is important that you do your own work.  Remember that  the  test  will not be counted for your own personal grade, so do not  worry  if  you  find  some  parts  of  the  test  difficult.  4.  In this test and all the doze tests try to do your best.  The harder you try,  the more benefit you will  get from your effort.  B. SPECIFIC INSTRUCTIONS:  1.  With your pencil  fill  in  as many blanks  as  you  can  of  the  with the best answer you can think of.  2. Follow these suggestions.  a)  read  the  first  and  last  sentences  uruuutilated text to determine the gist (main idea) of the passage. b) then skim the mutilated text trying to get clues to the gist. C)  read  from the beginning to the  trying to  find meanings  end  of  the  text,  for the omitted words by  checking context clues before or after the omitted  279  words  or  from general  knowledge  related  to  the  text. d)  Finally,  reread  the  entire  text  when  you  have  guessed all or most of the words, filling in words previously not guessed or correcting words already guessed in terms of the total text.  280  TABLE F SPECIAL INSTRUCTIONS FOR STUDENTS USING COMPLETELY-COLORCODED PASSAGES (FORM 4) AND COLOR-CODED BLANKS (FORM 2) SPECIAL INSTRUCTIONS: (Teacher cannot read any of the special instructions.) (This is for the students who are practicing with the color coding.)  The find  following  clues  to be  passage able to  is  color  fill  in  coded  the  to  blanks.  help The  you  to  color  coding means as follows: BLUE  =  Nouns:  student; Bob; Tokyo; school; science...  =  Pronouns:  I, he; me, her...  RED  =  Verbs:  go,  GREEN  =  Adjectives: big; my, his; this; a, the; no...  PURPLE  =  Adverbs:  quickly; today, then; there; very...  BLACK  =  Other:  and, or; in, on; where, how; not...  .  study, think; went; will study...  281  APPENDIX D: SPREADSHEET CHART FOR DATA COLLECTION AND DETERMINATION OF SCORES For Collection of Answers an Determination of Scores TABLE G ANSWER COLLECTION AND BLANK SCORE DETERMINATION CHART Page 1 TOTAL PRETEST  TEST 2 TOTAL  MEAN EXACT MTH  MEAN EXACT MTH  MEAN ACCEP 14TH  MEAN ACCEP 14TH  Page  1  SCORES FOR BLANKS  POINTS FOR..  .......  EXACT WORD  Exact Mth  =  EXPLANATION.  .......  CLUE LOCATION  same wrd  : diff wrd  Tl:not in passage  T:1O/9/8  POINTS FOR  TYPES OF CLUE...... T2: beyond senten T3: within senten  T:5/4/4  :  T:7/6/5  :  8/7/6  6/5/5  doze Form 4  lA4a: Student 1 Class A Group 4 Level a  (everything  2A4b: Student 2 Class A Group 4 Level b  color coded)  3A4c: Student 3 Class A Group 4 Level c 4A4d: Student 4 Class A Group 4 Level d 5A4e: Student 5 Class A Group 4 Level e  MEAN EXACT 14TH  282  MEAN ACCEP MTH doze Form 3 (everything.  6A3a: Student 6 Class A Group 3 Level a ...... .  7A3b: Student 7 Class A Group 3 Level b  color coded  8A3c: Student B Class A Group 3 Level c  randomly)  9A3d: Student 9 Class A Group 3 Level d lOA3e: StudentlO Class A Group 3 Level e  MEAN EXACT MTH MEAN ACCEP MTH  doze Form 2  llA2a  (blanks  l2A2b  color coded)  l3A2c l4A2d l5A2e  MEAN EXACT MTH MEAN ACCEP MTH  Cloze Form 1  l6Ala  (Standard  l7Alb  doze)  . . ... . ... .  .l8Alc l9Ald  V  MEAN EXACT MTH MEAN ACCEP MTH  2OAle  283  Cloze Form 4  21B4a 22 B4 b 23B4c 24B4d 25B4e  MEAN EXACT MTH MEAN ACCEP 14TH  doze Form 3  26B3a 27B3b 28B3c 29B3d 3OB3e  MEAN EXACT 14TH MEAN ACCEP MTH  Cloze Form 2  31B2a 32B2b 33B2c 34 B2 d 35B2e  MEAN EXACT 14TH MEAN ACCEP MTH  “Si  285  doze Form 1  36B1a 37B1b 38B1c 39B1d 4OBle  MEAN EXACT MTH MEAN ACCEP MTH  doze Form 4  41C4a 42C4b 43C4c 44C4d 45C4e  MEAN EXACT MTH MEAN ACCEP MTH  doze Form 3  46C3a 47C3b 48C3c 49C3d 5OC3e  MEAN EXACT MTH MEAN ACCEP MTH  286  Cloze Form 2  51C2a 52C2b 53C2c 54C2d 55C2e  MEAN EXACT MTH MEAN ACCEP MTH  doze Form 1  56C1a 57 C lb 5BClc 59Cld 6OCle  MEAN EXACT MTH MEAN ACCEP MTH  doze Form 4  6OD4a 61D4b 62 D4 C 63D4d 64D4e  MEAN EXACT MTH MEAN ACCEP MTH  287  doze Form 3  65D3a 66D3b 67D3c 68D3d 69D3e  MEAN EXACT MTH MEAN ACCEP MTH  Cloze Form 2  7OD2a 71D2b 72 D2 c 73D2d 74 D2 e  MEAN EXACT 14TH MEAN ACCEP MTH and so on to lOOEle MEAN EXACT 14TH  MEAN ACCEP MTH  page  1  BLANK1  B2  B3  his 0004  son 000(1)  He 10002  same  same  same  T: 10/9/8 T:7/6/5  T:7/6/5  288  APPENDIX E: THE MAIN PROJECT PASSAGES The Main Project Passages Pre—test:  Perspectives  Test 2:  The Question of Society  Test 3:  Ethnicity  Test 4:  Uniqueness  Test 5:...  Polity  Test 6:  Identity  Post-test:  Perspectives  The Original Texts Pre-test: Perspectives There are three dominant perspectives for understanding a society. The  first  perspective  attempts  to  society holds together as an entity.  determine  how  This perspective  a is  known as functionalism because whatever aspect of society is analyzed,  the  object  is  determine  to  how  that  structural  feature contributes to the integration of the society. example,  broadcasting  can  be  contributes to national unity,  assessed  in  terms  how  it  disunity.  Do  recordings  by  Canadian artists so that Canadian culture is developed,  or  Canadian  does  the  Canada’s  radio  stations  American national  or conversely,  of  For  give  priority  music  industry  sport,  bringing  to  dominate? Canada  States  than  in  Canada,  does  hockey  together,  because there are more National Hockey League United  Is  hockey  or,  teams in the does  hockey  289  contribute to North American continental important  are  a  national  anthem  development  of  societal  unity?  regionalism  in  creating  sectional  unity?  of  these  All  aspect  of  society  integration  within  What  questions  has the  in  and  developing  on  flag  to  is  the  role  than the  either  or  in  How  a  rather  focus  society,  integration?  the of  societal  function  an  cohesion  or  preventing  such  concensus from occurring. The second perspective tends to see society less as an on-going equilibrium than as a struggle between conflicting groups. its  This is known as the conflict’ perspective because  focus  what  is  on what  groups  power  possess  determines  dominance wealth  and  and  groups have power within a less  what  power, a  society  subordination,  poverty  all  and  how  is  center  express  between people within a society.  the  use  of  that  Terms  like  periphery,  and  like. and  society,  different  relationships  For example, the fact that  persons of British descent have historically had more power in Canadian society than persons of French descent,  created  a  British  set  of  institutions  in  Canada  expressing  heritage and a form of communications dominant. both  But  language  evidence of power groups  regarding  in which English was  struggles whose  are  found  definition  Canadian society should be like will dominate.  Test 2: The Question of Society  a  of  among what  290  It may seem ironic that even though the Canadian state is  over  one  hundred  years  old,  the  precise  nature  of  Canadian society and its existence as an entity is still in question.  In fact,  the stormy years after the centennial  birthday in 1967 suggested more than ever that the concep t of a Canadian society could not be taken for granted. Quebec was  While  contemplating what degree of distance from the  rest of Canadian society was most appropriate,  The Symons  Report was concluding that Canadians knew little about their  own society, and a Federal Task Force on Canadian Unity was scouring the country for clues about ways to create a more integrated and cohesive society. prompted  by  the  Meech  Lake  More recently, the debate  Accord  whereby  Quebec  sought  recognition for its status as a “distinct society” created controversy that again suggested the fragility of nation al unity.  What  kind  of  society  is  this  that  has  been  problematic for so long? The use of the term “Canadian society” implies that it can be differentiated from other societies and that it has some measure of internal coherence. evidence  to  suggest  that  internal  Yet there seems to be coherence  society has been in continual question.  in  Canadian  Repeated waves of  immigration and emigration, British and American influences, French-English clustered  differences,  population  in  a  a vast  relatively territory,  sparse and  but  uneven  economic development are only some of the factors that have contributed to fragmentation rather than societal unity.  291  It is therefore, by no means certain that there really is such a thing as a Canadian Society.  Does the strength of  the various small scale sub—societies in Canada preclude any meaningful discussion about Canadian society as a whole? differences  in  the  resident  population  Do  overwhelm whatever  may be held in common. Canada  exists  as  a  nation  by  the  political  and  legislative degree of the British North America Act passed by  the  British  parliament  in  1867.  This  document created an independent national unity.  legislative  292  Test 3: Ethnicity Perhaps the most dominant feature of the “New World as opposed to the was  the  vast  “Old” Word,  amount  of  as  seen through European eyes,  sparsely  settled territory  in the  “New” World.  It is this basic fact of a relatively small  population  an  in  enormous  area,  compared  to  the  crowded  “Old” World,  in combination with the expansionist ambitions  of  powers,  European  that  predestined  ethnicity  to  be  a  salient issue in the building of Canadian society.  First it  was  Peoples,  the  contact of  European  cultures with Native  then the intermingling of European peoples with each other in the new land, and now more recently, immigration from new source years  areas of  such as Asia  residence  multiculturalism remember, Clearly,  in  and South America.  CaTlada,  encourages  rediscover, ethnicity  or  has  the  government  members retain  been,  of  their  and  Even  the  policy  of  society  to  ethnic  continues  after  origins.  to  be,  a  significant feature of Canadian society Ethnicity is an amalgam of objective factors relating to place of birth, and  traditions  citizenship,  which  are  mother tongue,  transmitted  through  heritage and characterize that individual. experience,  ethnicity is  and customs a  person’s  In the Canadian  frequently rooted in reference to  another nation—state which provides a “foreign” dimension to the  concept.  objective  But  traits  ethnicity such  as  does  language  not and  only  involve  customs;  it  these also  involves a subjective element pertaining to how people view  293  themselves,  i.e.,  their  difference,  then,  not  Italian or English, Canada  or  Italy,  ethnic  only  in  identity. whether  There  a  person  is  a  speaks  or whether that person is a citizen of  but  also whether the person embraces  an  identity as “Italian,” “Italian—Canadian,” Canadian, or even Canadian.  Each  alternative  tells  us  something  important  about that person in relation to the wider society. The diversity of contemporary responses people give to the objective and subjective facts of their background makes the analysis of ethnicity both a dynamic and essential for the understanding of Canadian society.  Test 4: Uniqueness One of the best ways to learn about a society is to compare  it  with  other  societies.  Such  comparisons  are  important because, while each society is unique and no two societies are identical, there is a tendency to assume that the society under review has few parallels elsewhere.  This  is particularly the case for people who live in Canada, who are caught up in the internal struggles of Canada’s societal problems, never  or who are trying to resolve them as if they had  been  experienced  anywhere  before.  A  comparative  analysis helps us to see that there are other societies with similar problems have been found. the outside,  for which  similar or different  solutions  For those studying Canadian society from  it is also important to see linkages to other  societies with which they may be familiar through the media  294  or personal experience.  In sum, a comparative focus enables  us to place a society in a global context which sharpens our ability to identify distinctives and similarities.  Not only  ar our horizons broadened, but we are able to understand the dynamics of Canadian society in a new way. attempt  No detail  the  is  made  in  this  many ways  in which  with other societies.  In fact,  remain  weakly  implicit  or  chapter  Canadian  it  exhaustively  society  contrasts  some of the contrasts will developed  because  social  scientists have not engaged in much comparative work as such work is both difficult and complex.  Yet it is possible to  identify  some  society,  to look for parallels in other societies,  outline  some  of  of  the  characteristic  their  significance  features  and  of  meaning  Canadian and to in  that  social context. We have already seen that the nature of Canada’s land surface,  settlement  regional  differences  character.  history, has  population  produced  a  distribution,  society with  its  and  own  We have also noted that the variables such as  social class,  ethnicity,  language,  religion,  add a dynamic to these formative features.  and occupation What has been  the experience of other societies where similar factors are present?  Test 5: Polity If you want to create some controversy in a group of Canadian people,  there  is no faster way to do it than to  295  bring up the issue of bilingualism.  Everyone has an opinion  about  also  bilingualism  and  everyone  from  everyday  has  stories  or  anecdotes  to  position.  Whether it is hearing a bilingual version of “0  tell  life  to  justify  their  Canada,” hearing two Beets of preflight instructions on an airplane, or seeing a simultaneous translation of a document or  instructional  sign,  bilingualism  is  an  issue  many  Canadians are still grappling with because it is foreign to the unilingual manner in which most people live their lives. It is  frequently assumed that most  with one dominant universal  language.  countries  operate  Many Canadians,  for  example, are aware that the United States has taken as many immigrant peoples utilizing numerous languages, but they not that one  language  (English)  means of communication.  is  still used as the primary  What is perhaps less well—known is  that the large influx of Spanish-speaking peoples into the southern United States in recent years has created a real battle  in  whether  some  two  endorsed.  areas  (e.g.,  languages The  successfully American norm.  (English  national  argued  California  that  and  and  Florida)  Spanish)  English-speaking unilingualism  should  over be  has  so  far  should  be  the  The facts of Canadian history and polity,  however, are somewhat different, though there is a tendency for people  to  assume  that  unilingualism  ought  to  be  the  normal pattern in Canada as well. But if unilingualism is the norm in many countries, are there countries where more than one language is officially  296  issue?  WE  majorities  may  support a language distinct from the official language.  The  recognized, have  or does Canada stand alone on this  already  seen  regional  ethnic  for example, has several regions where ethnic  Soviet Union, groups possess Moldavia).  that  It  a distinct is  language  possible  for  (e.g.,  the Ukraine  countries  to  and  officially  recognize the existence of more than one language as a basis for communication within their society as a whole?  297  Test 6: Identity It is the commonality of sharing a territory (in spite of its size)  and participating in its polity  (in spite of  its inequities) that makes the country’s residents Canadian. In other words, who  share  the  symbol  “Canadian”  creates  that  a  national  A social identity is the sum of the sentiments,  identity. cultural  it is the collective interaction of people  attributes,  and  arrangements  structural  people  share which gives them a feeling that they belong together. Individuals  and  identity,  but  identity  into  groups they  create  can  their  also  personal  (though to varying degrees).  contribute  and  internalize definition  to  the of  that  national themselves  A societal identity, then, has  a collective component as well as an individual dimension. For this reason, it is possible to speak of the residents of Canada  forming  accepting  that  a national  society,  collective  identity  and  individual  as  something  members that  is  personally meaningful. But  there  is  a  second  aspect  to  being  part  of  a  national collectivity; that is, we learn what it means to be part of Canadian society by distinguishing this society from other societies.  One study found that members of Canadian  society became more aware of their national identity through interacting with foreigners.  It could also be argued that  state negotiations with other societies make Canadians more aware  of  national  their national identity  can  interests.  coagulate  as  The point a  result  is  of  that  a  external  298  relationships, and as a consequence this external dialogue, a  society  may  become  more  aware  of  its  internal  relationships or the concerns its people share. There is also a third aspect to national identity. perception  of  distinctive  a  national  identity  (descriptive).  identity  emerges  For example,  as  is  things  merely as  One  that they  a are  no society has quite the same  relationship between anglophones and francophones as Canada does,  and for better or worse,  society its identity. prescriptive,  that is what gives Canadian  But another perspective is far more  suggesting  Canadian society should be.  vision  and  ideals  about  what  299 Color-”cc:di ng is used to hel p you fi d clue Elue =  Red reen Purple Ellack  =  =  Nouns: Pronouns: Yerbs Adjectives Adverhs Others  CANADIAN SOCIETY  i  fill  in the blanks  student; Ebb; Tokyo; school; science,.. I he; me, her... go, study, think; went; will study... big; my his; this4 a the; flOM quickly; Loday then; there; very. and, or; in, on...  The Question of Society  It may seem ironic: that even though the Canadian state is over one hundred years ci ci  the cirec:i se  nature of Canadian society and its existence as an entity  is still  in question.  En fact,  the  1)  years after the centennial  2)_  1967 suggested more than ever that the  conc:ept of a  bi rthday  3)_._ society could not be for granted.  While Quebec 5)__  contemplating what degree of 6)  from the  7)_ of Canadian society was approprate  the Symons Report 9)__.  Canadians 10)  12)_  conc 1 LLCJ i rig that  Ii tt.le about their own  ..  and a Federal  11)  about ways to  L  Task Force on Canadian  was scouring the coutry fc;r  13)_,__,  a more i nteqrated and  14).,,  15)____ society.  More recently  the 16)__  prompted by the Meech Lake Accord whereby 17)._ sought recogni Lion for  16) the fragility of 20)  its status as a  diet nct .  created controversy that again 19)  unity.  What kind of  3oo  society 21)_ for  50  this that has been 22)  1 on ci?  The 23)_  of  implies that it can be $5  the term “Canadian 24)  from other  —.  societies and 26)  it has some measure internal  ....  H  coherence  to be evidence to 29)  N  Vet there 26)  that  internal  in Canadian society has been in continual 31  Repeated waves of immigration and emigration 9  )_  and American influences  differences,,  a 3:3)  FrencriEnplish  but clustered population  in a vast 34)  and uneven economic development  are only 35)  of the factc,rs that have to fragmentation rather than societal  37)  It  9 is  therefore  there really soc iety 41. )  is 39)  m’ ninju1 4’?)  the strenqth of  the var ious  scale sub-”societ ies in Canac:Ia preclude any dbouL C anach an 4 )  ——  as a  —  DC) 44),__ in the resident 45)  overhelm whatever 46) 47)___,,, —  be held in common? exists as a nation by the  and 49)_  -  North America Act passed by the 50) i n 1667  that  a thing as a Canadian  —  40) ,,,..  whole?  by no means 36),_,._  decree of the British parliament  This 1 ep i slat i ye docume nt created an i ndependent  national unity 0  3o1  Color—coding is used to help you find clues to fill in the blanks. Blue  = =  Red Green Purple Black  The  = = = =  Nouns: Pronouns: Verbs: djectivss: Adverbs Other:  student; Bob; Tokyo; school; science... I, he; me, her... go, study. think; went: will study... big; my, his; this; a, the; no. quickly; today, then; there; very... and, or; in, on...  Issue of Ethnicity  Perhaps the most dominant feature of the  ‘New World  as opposed to the “Old” World, as seen through European eyes,  was the vast amount of sparsely settled  territory in the “New’  World.  It  is this 1)  fact of a relatively 2)  population in an  enormous 3) 4)  compared to the crowded in combination 5)  ambitions of 6)  the expansionis  powers, that predestined ethnicity be a salient issue 8)  7)  building of Canadian society.  the  with Native Peoples,  the intermingling of  European  with each other 13 )  12)  new land, and now 14) from  the  First it 9)  contact of European 10) 11)  ‘Old”  recently,  immigration  source areas such 16)  15)  and South 17)  .  residence in Canada,  the government policy of  rediscover, 21)  ethnic origins.  sia  Even after years 18)  encourages members of the 20)  19)  remember,  the  retain 22)  Clearly, ethnicity has been, and  to  to be, 24)__,_  a significant feature of  society  Ethnicity is an amalgam of 2.5)_  relating to place of birth, 26)____  factors  mother tongue,  and customs and 27)____ which are transmitted a 29L_ that  individual  In 3O)__.,.  heritage and characterize ___  Canadian exper ienceq  is frequently rooted in reference 3$)_,__ another a  “foreign”  nation—state 33)_•  dimension to the 34)___  ethnicity does not only 35)  involves a 37)  ____.  these objective  _,  j  --  it also  element pertaining to how  view themselves 9  36)  provides  ?ut  _  traits such as language and 36)_  identity  ——  There is a difference,  iNel.,  then,  their 39)__,,_______  not 40)___________  in whether a person speaks I tal ian or 41)  or whether  that 42)__  43)_____ or  Italy,  person embraces an 45) 46)____  “  ___  is a citizen of  hut also 44) as “Italian,  Canadian or even Canadian..  __  “  the  “Italian—  47)_,_  alternative tells us 4$)________ important about that  49)__  in relation to the wider 5O)_  The diversity of contemporary responses people give to the objective and suhective facts of their background makes the analysis of ethnicity both a dynamic and essential fc:;r the understanding of Canadian society..  3L3  Color—coding is used to help you find clues to fill Blue  = =  Red Green Purple Black  = = = =  Nouns: Pronouns: Verbs: Adjectives: Adverbs: Other:  in the blanks.  student: Bob: Tokyo: school; science... I, he; me, her... go, study, think; went: will study... big; my, his; this: a, the; no.. quickly; today, then; there: very... and, or; in, on...  The fuestion of Uniqueness  One of the best ways to learn about a society is to compare it with other societies. 1)  Such comparisons  important because, while each 2)  unique and no two 3) 4)  —  are identical 4  is a tendency to 5)  society under review has 6) This  that the parallels elsewhere.  particularly the case 8)  7)  people who live in Canada, 9) up in the internal problems,  1  are caught  10)  of Canada’s societal who are trying to resolve  11)  12)  as if they 13)  i’4)  anywhere before.  us to 15)  never been A comparative analysis helps  that there are other  16)  with similar problems for which  17)  or different solutions have been found.  For those studying Canadian society from the outdside, 18) to  is also 19)  25)  to see linkages  societies with which 21>  20)  be familiar 22) 23)  —  may  the media or personal •  In sum, a 24) to place a society in a global  focus enables  3o’1 context which 26)  our ability to identify  distinctives and 27)  Not only are  •  horizons broadened, 29)  28) able to 30)  we are  the dynamics of Canadian society  inanew3i)___________ No attempt 32)  made in 33)  chapter to exhaustively detail the many ways in which Canadian society contrasts with 34)  societies.  fact, some of the contrasts 38)  In  remain implicit  or weakly developed because social scientists 36) not engaged in much comparative work as such is both difficult and 38)  37)  it is possible to 39)  •  Yet  some of the  characteristic features of 40)  society, to look  for parallels in other societies, 41) outline some of 42)  to  significance and meaning in  that 43)  context.  We have already 44)  that the nature of  Canada’s land surface, settlement history, population distribution, and regional 45)  has produced a 46)  with its own character. that variables 48) 49) 50)  47)  have also noted as social class, ethnicity,  ,  religion, and occupation add a dynamic to formative features.  What has been the  experience of other societies where similar factors are present?  ;c 0• Color—coding is used to help you find clues to fill Blue  = =  Red Oreen Purple Black  =  =  Nouns: Pronouns: Verbs: Adjectives: Adverbs: Other:  in the blanks.  student: Bob; Tokyo; school; science... I, he; me, her... go, study, think; went: will study... big: my, his; this; a, the; no... quickly; today, then: there; very... and, or; in, en...  Polities With Mere Than One Official Language: an Introduction  If you want to create some controversy in a group of Canadian people, there is no faster way to do it than to bring up the issue of bilingualism. Everyone  an opinion about 2)  1)  3)  and  also has stories or anecdotes to  from everyday 5)  ‘4)  6)  position.  to justify  Whether it is hearing a bilingual  version of “0 Canada,” 7) preflight  two sets of on an 9)  8)  ,  a simultaneous translation 10) instructional sign,  a document or  bilingualism is an 11)  Canadians are still grappling 12)  many  because it is  foreign to the unilingual 13) 14)  or seeing  in which most  live their lives. it is frequently 15  operate with 16) 17)  United States 19)  that most countries dominant universal language. Many  for example, are 18) taken as many immigrant  peoples utilizing numerous languages, 20) that one 21)  that the  (English)  is still 22)  they note  3o as  the primary 23)_____  24) 1 ar ge  of communication,  What is  less well—known is 25)___________  influx o1  Span i sh”i eak i nq 26)  southern Un ted 27)__ created a real  (English and Spanish) national  into the  in recent 26)  battle in 29)  California and Florida)  the  __  has  areas  over 3O)_  (e 1 q 0 1  to languages  should 3l)__  endorsed.  _  The  __—speakxng majority has so far  3$)_,.  successfully argued that uni lingual ism 33),_,, American 34)_._  be the  The facts of Canadian history  35)  polity  36)  is a tendency for 37)__  however  assume that 36)  are somewhat different  though  to  -  ought to he the normal  pattern in Canada as 39)____  But if uni lingual ism is the norm in many 4O)___,_  are there countries where more than one is officially recognized  42)_  stand alone on this issue?  have already seen that 44)  or  does  43)_  —  ethnic majorities  45)__• support a language distinct 46)_,__  the official 47) 46)  has several  groups possess a 5O) Ukraine and Mol davia)  The Soviet Linionq for regions where 49)  language  (eugu,  It is possi bie for countries to  officially recognize the existence of more than one  language as a basis for communi c:ation within their society as a whole?  the  3c7 Color—coding is used to help you find clues to fill in the blanks.  Blue  Nouns: Pronouns: Red = Verbs: Green = Adjectives: Purple = Adverbs: Black = Other: =  =  Aspects of Societal  It  student; Bob; Tokyo; school; science... I, he; me, her... go, study, think; went; will study... big: my, his; this: a, the: no... quickly: today, then; there: very... and, or; in, on...  Identity  is the commonality of sharing a territory  (in  spite of its size) and participating in its polity (in spite of its inequities) residents Canadian.  that makes the country’s  In other words,  it 1)  the  collective interaction of people 2) symbol Canadian 3) 4)  share the creates a national  eccial identity is the 5)  of  the sentiments, cultural attributes, 6) arrangements  12)  structural  share which 8)  7)  feeling that they belong 9) 10)  .  create 11)  them a  Individuals and contribute to that  but they can also internalize the national  identity 13)  their personal definition of  14)  (though to varying degrees)  15)  identity,  16)  well as an individual dimension.  17)  this reason,  18)  of the residents of 19)  a 20)  identity.  then,  it  .  A  has a collective component  is possible to  society, and individual 21)  forming  3°f acceptinq that 22),  23)  identity as something  _.,,,,_  is personally meanincful.  But 24)__  is a second aspect to being  25)__,,,  of a national  26)_  learn what it means to he part of  collectivity;  27),,,___ society by distinguishing  from other societies.  that is  this 29)___  One study found 29)  members of Canadian society became more aware of 3O)_ national  identity through their travels outside  the country and 3i)_.._ could 32)_  interacting with foreigners,  It  be argued that state negotiations other societies make Canadians more  33)  34__,__ of their national 35)_  that a national  interests.  The point  identity 36)_  coagulate as a result of external relationshipsq 37) as a consequence 3E1)  __  external  diaiogue  society may become more aware 39)_____ i 4c)),_  internal  or the concerns its people 41)_,__,  There is also a th:ird 42) identity.  a  to national  _  One perception oF a 43)  identity is  merely that a distinctive identity emerges 44) as 45)_  are  society 47)_  4a)_  (descriptive)  .  thing  For 46)  quite the same relationship anglophones and francophones as Canada does,  and for 49).  or worse,  that is what gives  its identity. is far more prescr,:pta ve  But another perspective  suggesting vision and ideals about  what Canadian society should be.  n  3°’1  Arswer ky  Aipue tf Beieta1 Identity It is the commonality of sharing a territory (in spite of its size) and participating in its polity (in spite of its inequities) that makes the country’s residents Canadian.  In other words,  collective interaction of people 2) symbol  “Canadian” 3)  4)__A.  it 1)_._..4 C  the  aij)aPk.  share the  creates a national  social  identity is the 5)  ,  structural  share which 1 B)• . C  feeling that they belong 9) 10)  of  UrI  the sentiments, cultural attributes, 6) arrangements  identity.  •  create 11)  them a  Individuals and contribute to that  but they can also internalize the national identity 13)  their personal definition of (though to varying degrees).  A  identity, then, has a collective component 16)  well as an individual dimension.  17)  this reason,  16)  of the residents of 19)...ChflJ.L..... forming  a 20)  it i. possible to  society, and individual 21)rn,hIkef,_  3/0  b  accepting that 22)  identity as something  23),,  is personally meaningful. But 24L+lhOt  is a second aspect to being  25)...fQ.r..f_._... of a national collectivity; that is, 26)  learn what it means to be part of society by distinguishing this 28)...f.CJ..L_.  from other societies.  One study found 29)  members of Canadian society became mor. aware of 30)  national identity through their travels outside  the country and 31)  interacting with foreigners.  could 32)  It  be argued that state negotiations ether societies make Canadians more  35)  IS  of their national interests.  The point  that a national identity 36)  COtl  coagulate as a result of external relationships, 37) as a consequence 3B)  external dialogue, a  society may become more aware 39)  4o)fI±1DniL9pser  the concerns its people 41)  There is also a third 42) identity.  One perception of a 43)  merely that a distinctive identity emerges  as  45)J  he  society 47) 48) and for 49) 50)  r  •r itk.c  its internal  (descriptive).  a  to national  1  identity is  44)__#tflb1r%_..  things  For 48)  quite the sam. relationship  anglophon.s and Francophon.s as Canada does, or worse, that is what gives society its identity.  But another perspective  is Far mar. prescriptive, suggesting vision and ideals about what Canadian society should be.  no  311 Sample of distribution of reiterative words TABLE H A SAMPLE OF HOW THE PASSAGES WERE ANALYZED FOR SELECTION OF THE MAXIMUM NUMBER OF REITERATIVE WORDS AND THEIR LOCATION Perspectives There are three dominant perspectives for understanding a society. THE  first °perspectiveattempts  society°ho1ds °known°AS  together AS  to  an entity.  determine  °This°perspective  functionalism°because°whatever  is analyzed,  the object is to°determine°h that structural  broadcasting°can°be  assessed  contributes to national unity, Canadian  radio  °society.  in  terms  For  OF°°it  conversely, disunity.  stations ogiveepriority  to  Do  recordings  by  °Canadian°artists so that_Canadian°culture°is developed, DOES  THE  Canada’s  is  aspect°° society  feature°contributes°to the integration of example,  a  American  music  national  industry°dominate°?  °,  bringing  Is  hockey  together°,  0  because there are more National°Hockey°League  teams  United  does  States  than  .jj  Canada,  °does°hockey  or  I, THE  hockey  contribute to North American important development  are  aonationale anthem  of  What  regionalism°°creating unity?  °°of  and  these  sectional questions  °aspect°of  society  has  integration  within  THE  concensus fromtoccurringo.  is  rather focus  a°flag°to  on  THE than  role  THE OF  societal  THE°function°an cohesion  society,  iii  preventing  such  312  THE secondoperspectiveo tends to see society less°ran on-going equilibrium° groups.  as  °  struggle °between conflicting  This is known as  its°focus°is what groups power  °  on  what  possess  because  groups  have  less°ower°,  determines  what  a  dominance and subordination,  power and  society  within  a  THE use is  center  LIKE.  °°  society, OF  that  °Terms°like  periphery,  and  °like°wealth and poverty°fl°express different relationships °between°people within that persons  a  society.  For°example°,  THE  fact  OF British°descenthave historically had more  °POWER° CANADIAN  created  a  set  a°British°heritage  society than°persons°of French descent,  OF AND  English was°dominant°.  institutions a°form°OF  jj  Canada  expressing  communications  which  But evidence OF POWER struggles are  found among both language groups regarding whose definition OF what CANADIAN society should be like will dominate.  313  APPENDIX F: PRODUCTION OF THE CLOZE PASSAGES Step 1  Type out all the stories  Step 2  Code for parts of speech using one symbol for each corresponding color (verb  =  (noun  =  (4 (* Using  1)  =  adverb)  =  adjective)  =  other parts of speech)  the  function,  word  processor’s  a word can be  “search  and  replaced by  the appropriate symbol in front.  replace”  itself plus  Using the “all”  feature, all of the same word can be replaced in a second or so. first  story,  other.  Starting at the beginning of the replace  each  word,  one  after  the  After a paragraph or two have been done,  many words all the stories will have been coded. As  more  and  more  words  are  becomes exponentially easier. with  some  words  which  can  replaced  the  Care must be taken vary  their  part  speech by their position in the sentence. like “work” way.  task  of  Words  can change in part of speech in this  Either  be  sensitive  to  such  cases  replacing words with words and symbols, after the passages have been coded.  when  or check  314  Step 3  Choose the words  to be  deleted and replace them  with a question, number, and two short lines.  Step 4  After all the words have been deleted, replace the short blanks with the standard 15 space lines and make all characters bold.  Step 5  Replace the single symbols with the corresponding printer codes as follows: 1  &  Step 6  =  ((C))l  =  red  =  verb  =  ((C))2  =  blue  =  noun  =  ((C)) 3  =  purple  =  adverb  =  ((C))6  =  green  =  adjective  Use the “search and replace” function and replace all the printer codes with “0” to give all black print  for  the  standard  doze  for  the  control  group.  Step 7  Use the “search and replace” function and replace all the printer codes with “0”, “6”  “1”,  “2”,  3”,  or  in a random fashion to give make the random  for of the doze procedure.  Step 8  Print  with  the  Star  NX1000  (or  printer using a colored ribbon.  NX1O2O)  rainbow  315  TABLE I COLLECTION AND BLANK SCORE DETERMINATION CHART (MAIN PROJECT) °NAME OF PASSAGE ° 0  0  TYPE OF BLANK° (EXTERN) 0 Bi  0  0 °STUDENT °IDENTITY 0  0  KEY WORD  ***  °JNK °DAV :APR  INTERB2  0 °  0  0  0  0  INTRAB3  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  °  °  0  0  •  °Classl 0  0  0  :  AAAAA  1  :  1 (empty)  :  (empty) 1  :  °Total: °Mean  0  0  0  0  0  0  °Class3  0  0  0  0  0  0  0  0  TM °NM :HN  0  °Total: °Mean:  0  : 0  2 .66  1 0  0  33  0  (empty) 1 (empty) 1 •33  0 0  : °  1 1 (empty) 2 .66  1  0  33  0  0  0  0  (empty) 0  0 0  316 APPENDIX G: COMPARISONS OF EXACT- AND ACCEPTABLE-WORD MEAN SCORES TABLE J EXACT- AND ACCEPTABLE-WORD MEAN SCORES ALL BLANKS Class 1 Class 3 Pre-test • °  19.9  17.5  °exact—word scores  22.3  °acceptable—word scores  C  23.2  0  Test 2 180°132  00  C  00  o o  Test 3  23.2  •  17.0  °°  Test 5  Test 4  225169° C  33.1  °  24.6  Test 6  0  C  377  •  32.8  00  C  °°  0  Post-test °  23 8  •  0  °  20 8  0  30.6  0  ° 0  27.1  39.7  0  32.7  •  227  C  39.2  027602600029302550 o  274  00  0  344  0  317 TABLE K EXACT- AND ACCEPTABLE-WORD MEAN SCORES INTRA-/ INTER-SENTENTIAL Class 1 Class 3 Pre-test 15.2  13.2  0  0  •  °  191  °exact—word scores 0  16.0 °acceptable-word  Test 2 0  o °  Test 3  Test 4  1560  10600  0  00  0  00  0  °  00  °  •°  0  18.4  12.2  Test 5  19801510022501820  28.0  20.6  Test 6  024902340024202150 •  O  0  32.9  28.7  00  0  °°  0  Post-test °  19 9  0  °  0  16 5  0  30.6  0  ° 0  27.0  0  32.3  0  27.8  30.3  0  26.0  318 TABLE L EXACT- AND ACCEPTABLE-WORD MEAN SCORES INTRA-SENTENTIAL Class 1 Class 3 Pre-Test 0  5.1  0  0  0  4•4  °exact—word scores  5.0  °acceptable-word  0  5.8  0  Test 2  Test 3  Test 4  0  780  57°°  5.100  570  580  o  0  00  0  00  0  0  00  °  00  0  0  95  0  57  Test 5  67 7.6  6.4  Test 6  0  550  4800  690  670  0  0  00  0  0  0  0  °°  54  55  Post-test 0  530  0  0  0  0  0  0  12.5  500  12.9  9.8  9.1  0  6.0  6.5  319 TABLE M EXACT- AND ACCEPTABLE-WORD MEAN SCORES INTER-SENTENTIAL Pre-test o  • o  io.i  •  8.8  °exact—word scores  0  13.3  °  11.0  °acceptable—word  Test 2  Test 3  0  780  49°°  0  0  00  o  8.9  0  6.5  00  Test 5  Test 4  1310100 0  20.4  0  14.2  °  0  •°  0  Test 6  019401860017301480 0  °  C  27.5  °  00  23.2  00  Post-test 0  14 6  o  °  °  11 5  0  18.1  0  0 0  14.1  0  22.5  0  0  18.7  16801240  00  24.3  0  19.5  320  APPENDIX H  THE COLOR-CODED CLOZE PROCEDURE PROJECT QUESTIONNAIRE  Thank you very much for participating in the Color-Coded doze Procedure Project. Now you have had the chance to practise the doze procedure, you have the opportunity to share your ideas and feelings about the project. 1.  Before this project had you ever done doze exercises a)  in Japan?  Yes  No  b)  In Canada?  Yes  No  If yes, did you enjoy the doze exercises?  Yes  No  2.1 In the present project did you have any difficulties with exercises? Yes  No  2.2 Please check one. Did you find the doze exercise(s) more difficult at the: a) b) c)  beginning of the project, at the middle of the project, at the end of the project?  2.3 Please check one.  or  Did you find the doze exercise(s)  easiest at the: a) beginning of the project, b) at the middle of the project, c) at the end of the project?  or  321 2.4 As you got more practice did you gain more confidence in doing the doze exercises?  Yes  No_  2.5 Did you have any problems seeing the words on the exercise papers? Yes  No_  If yes, what problems?  2.6 Do you think this project has helped you improve your ability to accurately guess missing words in a passage? Yes  No  2.7 When trying to find words to fill in the blanks did you ever: a)  look in the sentence of the blank? Yes  b)  look in the sentence before the blank? Yes  c)  No  No  look in the sentence after the blank? Yes  No  d) use your knowledge of the topic?  e)  Yes  No  Yes  No  Yes  No  guess?  f) do something else?  If yes, what did you do?  322  2.8 Would you like to do doze exercises in the future? Yes_  No  If yes, what changes would you like to be made?  2.9 Please write down any comments you have about anything in this project.  323 3.1* Did you practise the doze exercises using color? Yes  No  3.2 Did the color help you fill in the blanks? Yes  No  3.3 Did the color make the exercises more difficult? Yes  No  If yes, please circle the most difficult exercise(s). a) the beginning exercises (#2 and #3) b) the middle exercise (#4) C) the end exercises (#5 and #6) 3.4 Did the color make the exercises easier? Yes  No  If yes, please circle the easiest exercise(s)? a) the beginning exercises (#2 and #3) b) the middle exercise (#4) c) the end exercises (#5 and #6) 3.5 Did the color help you to find words to fill in the blanks, words that were: a) within the sentence of the blank? Yes b)  in a sentence before the sentence of the blank? Yes  c)  No  No  in a sentence after the sentence of the blank? Yes  No  3.6 Do you think the colored exercises helped you to do the last exercise (uncolored)? Yes  No  3.7 Please circle the type of doze exercise you would prefer to do in the future? a) non-colored doze  b) colored doze  324 3.l** Did you practise the doze exercises using the color— coded parts of speech? Yes  No_  3.2 Did the color—coded parts of speech help you fill in the blanks? Yes  No  3.3 Did the color—coded parts of speech make the exercises more difficult? Yes  No  If yes, please circle the exercise which was the most difficult. a) the beginning exercises (#2 and #3) b) the middle exercise (#4) c) the end exercises (#5 and #6) 3.4 Did the color—coded parts of speech make the exercises easier? Yes  No  If yes, please circle the easiest exercise(s)? a) the beginning exercises (#2 and #3) b) the middle exercise (#4) C) the end exercises (#5 and #6) 3.5 Did the color-coded parts of speech help you find words to fill in the blanks, words that were: a) within the sentence of the blank? Yes b)  in a sentence before the sentence of the blank? Yes  c)  No  No_  in a sentence after the sentence of the blank? Yes  No  3.6 Do you think the color-coded exercises helped you to do the last exercise (uncolored)? Yes  No  3.7 Please circle the type of doze exercise you would  325  prefer to do in the future? a) non—colored doze  *  **  This section was for Class 2. This section was for Class 3  b) color—coded doze  BIOGRAPHICAL INFORMATION NAME:  Howat Alan Labrum  MAILING ADDRESS:  #241-13612 67th Avenue Surrey, B.C. V3W 6X5  PLACE AND DATE OF BIRTH: EDUCATION:  Vancouver, B.C., Canada 24/02/50  (Colleges and Universities attended, dates, and degrees)  The University of British Columbia  Bachelor of Arts  1972  The University of British Columbia  Diploma in Adult Education  1973  POSITIONS HELD: Supervisor of E.S.L teachers, United Nations Refugee Program E.S.L. teacher in Canada, Thailand, Malaysia, and Saudi Arabia PUBLICATIONS:  AWARDS:  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items