4;: .:.v' .MI . THE EFFECTIVENESS OF THE USE OF COLOR CUES TO TEACH LOW-ACHIEVING SECOND GRADERS VISUAL DISCRIMINATION IN PAIRED -ASSOCIATE LEARNING OF PREVIOUSLY ENCOUNTERED . SIGHT-VOCABULARY WORDS Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ALBERTA W. SIMS 1977 NWWWIN. LII? ".A R Y MiCI r1 5 4'3 Unian rsity ml... m .' I. This is to certify that the thesis entitled The Effectiveness of the Use of Color Cues to Teach Low-Achieving Second-Graders Visual Discrimination in Paired-Associate Learning of Previously Encountered Sight Vocabulary Words. ' I ~ presented by Alberta w. Sims has been accepted towards fulfillment of the requirements for Ph . D . Education I degree in I t 7/”? W Major professor Date April 8, 1 77 G7 639 ABSTRACT THE EFFECTIVENESS OF THE USE OF COLOR CUES To TEACH Low-ACHIEVING SECOND GRADERS VISUAL DISCRIMINATION IN PAIRED-ASSOCIATE LEARNING OF PREVIOUSLY ENCOUNTERED SIGHT-VOCABULARY WORDS By Alberta W. Sims Problem This study was designed to determine the effect of color as a facilitator of visual discrimination on paired—associate learning. Specifically, the purpose of the study was to determine which of three treatments, maximum color, vanishing color, or no color, was more effective in visual discrimination training when a single color was used as a cue on a variety of stimuli. It was felt that the use of a single color might eliminate the problems previously associated with the use of multiple color cues while still focusing the subject's attention on the distinctive features of the stimuli. Procedure The population for the study included students from 15 schools in one school district, who were identi- fied by classroom teachers and reading specialists as having difficulty in word recognition. A random sample Alberta W. Sims was taken to generate three groups of 40 second graders for each treatment. An instructional episode which consisted of 12 pairs of words on which some Children often make "static" and "kinetic" reversals was developed employing three sets of 120 cue cards for each treatment. Verbal directions for each group remained the same throughout the treatments. The only source of variation was the method of color highlighting applied to the stimulus word. Each subject was taken through an instructional episode individually. Three posttests—-Naming, Pairs, and Match to Form from Memory--were administered upon comple- tion of the instruction. Results The data collected were analyzed using a two-way multivariate analysis of variance. A treatment main effect of .05 level of significance was revealed. The univariate F-test was employed to specify on which dependent variable the treatment main effect was found to be significant. It was established that all three dependent variables were influenced by the treatment main effect. The univariate F-tests were found to be significant at the .01 level. Four post-hoc contrasts were constructed to find where the significant differences existed among the three Alberta W. Sims groups. As a result of using the t-test to test the con— trasts on the three dependent variables at the .01 level, vanishing color was found to be significant on the three variables. Significantly better (p < .05) achievement was found on the three tasks favoring the vanished color treatment over the maximum color and no-color treatments. Conclusions Within the limitations of the study, the investi- gator concluded that children who receive instruction that utilizes vanishing color cues learn the tasks of visual discrimination, visual memory, and association of a verbal response at a significantly higher level of achievement than children who receive instruction with maximum color cues and no color cues. Finally, it can be concluded that the use of van- ishing color cues serves to focus attention on the dis- tinctive features of the words to be learned without producing interference at transfer. Implications Teachers of reading should consider the use of vanishing hue color cues to assist children with the fol- lowing tasks: (1) learning to visually discriminate between words that are confusing, (2) learning to visually discriminate between words that have all or Alberta W. Sims many common letters, and (3) learning words that contain letters that are rotations of each other. Research should be undertaken in a school setting using materials that employ the vanishing hue color technique for one group and materials with no color for one group for a specified period of time. Comparison of the achievement of the two groups would provide informa— tion concerning the relative potency of the vanished color cues. Finally, publishers may wish to consider using the technique with beginning reading materials. THE EFFECTIVENESS OF THE USE OF COLOR CUES TO -TEACH LOW-ACHIEVING SECOND GRADERS VISUAL DISCRIMINATION IN PAIRED-ASSOCIATE LEARNING OF PREVIOUSLY ENCOUNTERED SIGHT-VOCABULARY WORDS By Alberta W. Sims A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education 1977 © Copyright by ALBERTA W. SIMS 1977 Dedicated to my son, Mathew Abrams, Jr., M.D., who made me aware of the need ii ACKNOWLEDGMENTS For their guidance and counseling throughout the doctoral program, I wish to express my appreciation and thanks to Dr. George B. Sherman, Dr. Louise Sause, Dr. Lois Bader, and Dr. Donald Nickerson, members of the committee. To Mrs. L. Brandon, the school district specific school staff, and to the second graders who participated in the research project my sincere thanks. To Dr. Joe L. Byers and his staff, and to Mrs. Suwatana Sookpokakit for her encouragement and assistance throughout the research project, my sincere thanks. Finally, an expression of thanks is extended to family and friends, especially Dr. and Mrs. Eddie Hildreth and Dr. Maryann Johnson, for their encouragement and assistance, which have made this effort worthwhile. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . Chapter I. INTRODUCTION . . . . . . . . . . . . . . Background of the Problem . . . Importance of the Study . . . Purpose of the Study . . . . . Statement of the Problem . . . Definition of Terms . . . . . Population and Sample . . Hypotheses . . . . . Limitations of the Study . . . Organization of the Remainder of the Dissertation . . . . . . . . . . . . II. REVIEW OF RELATED LITERATURE . . . . . . . Introduction . . . . . . . . . . . . . Word Recognition . . . . . . . . . . . The Place of Word Recognition in Reading . . . . . . Characteristics of the Word Recogni- tion Process . . . . . Instructional Techniques Used in Word Recognition Discrimination Training . Visual- -Perceptua1 Problems . . . . Summary . . . . . . . . . . . Word Reversal Difficulties . . . . . . . Types of Reading Reversals . . . . . . Characteristics of Reading Reversals Summary . . . . . . . . . . . . . . . Color as a Cue . . . . . . . . . . The Use of Color in Paired-Associate Learning . . . . The Use of Color Cues by Good and Poor Readers . . . . . . . . . . . . Color as an Aid to Visual Perception in Early Reading . . . . . . . . iv Page vi vii l—‘ m mmmmswwm 11 11 11 11 13 17 23 28 28 29 31 38 39 44 46 Chapter Page The Use of Color to Isolate a Single St imulus O O O O O I O O O O O O O I 47 Summary . . . . . . . . . . . . . . . . 51 III. DESIGN OF THE STUDY . . . . . . . . . . . . 53 Sample . . . . . . . . . . . . . . 53 Methodology . . . . . . . . . . . . . . 55 The Instrument . . . . . . . . . . . . 56 PilOt Study 0 O O O O O O O O O O O O O O 61 Experimental Design . . . . . . . . . 63 Research Hypotheses . . . . . . . . . . . 63 Statistical Procedures . . . . . 64 summary 0 O O O O O O O O O O O O 66 IV. ANALYSIS OF RESULTS . . . . . . . . . . . . 67 Introduction . . . . . . . . . . . . . . 67 Research Hypotheses . . . . . 68 Are There Significant Differences? . . . 69 Results of the Multivariate Analysis of Variance . . . . . . . . . . . 70 Identifying Differences . . . . . . 73 Naming O O O O O O O O O O O O O O O O 76 Summary . . . . . . . . . . . . . . . . . 83 V. SUMMARY AND CONCLUSIONS . . . . . . . . . . 85 smary O O O O O O O O O O O O O O O O O 85 Conclusion . . . . 88 Implications of the Study to Education . 88 Recommendations . . . . . . . . . . . . . 91 APPENDICES O O C O I O O I C O O O O O O O O O O O 94 A. INTRODUCTION TO AND VERBAL DIRECTIONS FOR THE INSTRUCTIONAL EPISODE . . . . . . . . 95 B. WORDS USED AND METHODS OF HIGHLIGHTING FOR EACH TREATMENT IN THE INSTRUCTIONAL EP I SODE O O O O O O O O I O O O O O O O Q 1 00 C. DIRECTIONS FOR POSTTESTS AND POSTTEST SHEETS O O O O O O O O O I O O O O C O O 105 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . 110 Table. 1. LIST OF TABLES A Multivariate Analysis of Variance Results of the Univariate F-Test for variable 0 O O O O O O O O O O O I The Results of the Four Contrasts on Match to Form From Memory Task . . The Results of the Four Contrasts on Naming Task . . . . . . . . . . . The Results of the Four Contrasts on Pairs Task . . . . . . . . . . . . Vi Page 71 74 75 80 82 Figure l. 2. 3. LIST OF FIGURES Page Distribution of the Sample . . . . . . . . . 54 Words Used for the Instrument . . . . . . . 57 Sample Booklet Showing Word and Its Paired- Associate . . . . . . . . . . . . . . . . 58 Methods of Highlighting for the Three Treatments . . . . . . . . . . . . . . . . 60 Instructional Episode: Order of Teaching and Testing . . . . . . . . . . . . . . . 62 Experimental Design . . . . . . . . . . . . 65 Observed Means on Three Treatments on Three Tests . . . . . . . . . . . . . . . 77 Histograms of Observed Means on Three Tests Tests 0 O O O 0 o O O O O O O O O O O O O 78 vii CHAPTER I INTRODUCTION Reading is a concept that seems to defy universal definition. This phenomenon is brought about because of the number of aspects involved in the reading process and the lack of agreement among educators and researchers as to what those aspects are. There is also much theory explaining how this process actually takes place. How- ever, most tend to agree that reading involves a process of interaction with visual stimuli in terms of what is known from the reader's experience and what the writer is conveying on the printed page. In other words, the reader must have experiences that can be utilized in his reaction to what the author is saying. Therefore, read- ing is a process of encoding, decoding, and recoding of information. During this information processing the child is taught to pair the correct spoken word with its printed counterpart. When he has done this he has learned an association. The accumulation of these associations creates a sight vocabulary. An adequate sight vocabu- lary is necessary if he is to become a good reader. A primary problem in teaching sight words to some beginning readers is grasping the significance of the spelling code. This code is complex and built on the assumption that no two words in English will be spelled alike. Every word, whatever its length, will have sig- nificant discriminations--letter differences—-either within word portions or inclusions. To help a child discover this insight is a major portion of a teacher's responsibility in beginning reading. Teaching techniques seem minimal for this important training. Background of the Problem The terms "kinetic” and "static" were coined by Orton (1937) to describe the reversal errors experienced by children in their early reading behavior. Many theories have been advanced by researchers as explanations for these reversal problems. However, two seem most prevalent: (1) immaturity of the learner or (2) a lack of training in visual discrimination. It appears that children who are making reversals because of maturation and developmental factors experience minimal difficulty in making progress in school because of the nature of reversal tendencies for this group. The ten— dency to make reversals seems to fluctuate in that they are not always present. However, for that group of children who are mak- ing reversals because of a lack of visual discrimination training or code insight, profound difficulties can develop. The deficiencies caused by word reversals will often result in low reading achievement and an avoid- ance of reading or participation in any activity which is related to it. Importance of the Study Because of the difficulty experienced by a large segment of children learning to read with word reversals or confusing similar words, many learners are not able to perform the reading act with proficiency and teachers are not certain about how to provide the necessary reme- diation. Therefore, many of these learners lose effi- ciency and motivation because of this deficiency even though they have the necessary cognitive ability for performing the reading task. This research hopefully will provide information as to whether two types of reading problems, kinetic or static reversals (Orton, 1937), will respond to instruction and Whether the same type of_instruction or a differential treatment is neces- sary for each of these discrimination problems. Purpose of the Study It can be postulated that the use of color cues will be of benefit to those students identified as low-achieving readers if a major problem is the lack of mature discrimination skills. Thus the purpose of this study is to analyze the effect of a single-hue color cue in assisting children experiencing difficulty with these tasks. Specifically, the purpose of this study is to identify whether maximum color treatment, vanishing color treatment, or no color treatment is more effective in visual discrimination training and to determine the effects of selective cueing of significant letter dis- crimination as an aid to accurate recognition of confused words. Statement of the Problem The research described here was an investigation of the following questions: Will there be a significant difference in the performance of children using color cues or no color treatment on the discrimination of words previously encountered as sight words when pre- sented visually? Will there be a significant difference in the performance of boys and the performance of girls in the use of color cues or no color treatment on the discrimination of words previously encountered as sight vocabulary? Will there be a significant interaction in the performance of boys and the performance of girls on sight words previously encountered whether presented with color cues or no color treatment? To answer these questions, hypotheses were formulated, a population for research identified, and a methodology was selected, using an instrument developed by the investigator. Definition of Terms The following definitions are to be used when reference is made to any of the terms below: Matching to Form From Memory: The subject selects a word from a group of four words from a test sheet that has been presented to him in a i-second flash. Naming: The student makes the appropriate ver- bal response to a word presented to him on a card. “Pairs: The student makes the appropriate ver— bal response to pairs of words which are easily confused. Discrimination: The student is able to tell the difference between two easily confused words or letters. Vanishing: A method of removing a visual prompt; specifically, removing the color highlighting by gradually reducing the area of color until no color remains. Association: When the child can pair the cor- rect sound (spoken word) with the correct signal (printed word), he has learned an association. Stimulus: In this study stimulus refers to the words; the prompting stimuli are the color highlights added to letters in the words. ‘StrephosymbOlia: A delay or difficulty in learn- 'ing to read which is out of harmony with a child's general intellectual ability. At the outset it is characterized by confusion between similarly formed but oppositely oriented letters, and a tendency to a changing order of direction in reading. Kinetic Reversals: Sequential reversal of letter order, confusion or misreading a series of letters or words in which a progressional element enters, as when a whome word is used backwards, or a transposition of letters in a word is made (ex. was, saw). Static Reversals: Orientational reversal of letter form. Confusion or mistakes in recognition of singme letters which are alike except for their orienta- tion. It is characterized by an interchange or reversal of form (ex. "iorn" for "iron," "boy" for "dog"). Highlighting: Method of adding color to the stimulus word. Population and Sample The population under investigation comprised those elementary school children in second grade who were identified by their teachers as readers who were confus- ing words. All second grade children in the Lansing school district who were making reversals, transposi- tions of letters, rotation of letters, and confusing words in context or isolation were eligible for the study. The racial and socioeconomic composition of the group was nonspecific. Fifteen elementary schools were randomly selected from a list of 20 schools to partici- pate in the study. The sample consisted of 174 students from the Lansing, Michigan, school district. From this sample 120 children were randomly selected to participate in the study. The 120 children were randomly assigned to one of three treatment groups. Each group contained 20 boys and 20 girls. An instructional episode was developed in which only the visual stimulus varied in three treatments. Treatment 1 consisted of maximum color highlights added to and continued until transfer; treatment 2 consisted of maximum color highlights which were added and then vanished prior to transfer; treat- ment 3 had no color highlights added. Upon completion of the instructiOnal episode each child was post—tested for three tasks: naming task consisted of the ability to name correctly the 24 words taught by the investi- gator, match to form from memory consisted of the child's being able to match the word presented in a é-second flash presentation to the word like it from a list of four similar words, and naming word pairs consisted of identifying correctly each set of paired words. The results for all treatment groups were analyzed to determine whether color, color with fading, or no color at all were effective as techniques in teach- ing sight words to children who were making reversals, rotations, transpositions, and confusing words. Hypotheses The results for the effect of all treatment groups—-(l) maximum color, (2) vanishing color, and (3) no color--will be analyzed to determine whether the following null hypotheses are to be rejected. The con- trol limit for the probability is set at p < .05. Hol: There will be no significant difference among three methods of teaching word dis— crimination on a lZ-item Match to Form From Memory posttest.. There will be no significant difference between boys and girls in learning word discrimination as measured by a 12-item Match to Form From Memory posttest. Ho There will be no significant interaction between sex and treatment on the learning of word discrimination as measured by a lZ-item Match to Form From Memory post- test. Ho Ho : There will be no significant difference among three methods of teaching word dis- crimination on a 24-item Naming posttest. Ho : There will be no significant difference between boys and girls in learning word discrimination as measured by a 24-item Naming posttest. There will be no significant interaction between sex and treatment on the learning of word discrimination as measured by a 24-item Naming posttest. Ho H07: There will be no significant difference among three methods of teaching word dis- crimination as measured by a lZ-item paired word Naming posttest. H08: There will be no significant difference between boys and girls in learning word discrimination as measured by a lZ-item paired word Naming posttest. Ho : There will be no significant interaction 9 between sex and treatment on the learning of word discrimination by a l2-item paired word Naming posttest. Limitations of the Study The following limitations are noted in regard to the study: ‘The study is limited to the tasks under investigation or similar tasks, the study is further limited to situations in which similar materials and environments are employed, and finally, the study is limited to populations that are similar in requirements for inclusion in this study. Qgggnization of the Remainder of the Dissertation Chapter II will present a review of pertinent literature. Particular emphasis will be on the role of color in learning and specifically in discrimination and paired-associate learning. Chapter III will contain a description of all materials and procedures employed in this study. The design of the study will be presented in detail. 10 Chapter IV will include presentation of the data colleCted, treated, and analyzed for the study. Chapter V will provide a summary of the study and appropriate conclusions. Implications of this study and suggestions for future research are also included. CHAPTER II REVIEW OF RELATED LITERATURE Introduction The review of literature is organized under three major headings. These are (l) word recognition, (2) word reversals and confusions, and (3) color as an aid to visual discrimination training. Research findings in these areas provide the basis for the research ques- tions and hypotheses of this study. Word Recognition The Place of Word Recoggition in Reading Reading is a process that involves reacting to printed symbols. Words and their immediate context are symbols in that they stand for or represent concepts. Prevalent reading theories suggest that specific skills are required in order to be able to read. A primary skill in learning to read is the acquisition of an ade- quate sight vocabulary. According to Samuels (1970), "whatever one's ultimate definition or criterion of reading may be, word recognition must be included as one of the primary and major component skills." 11 12 Carroll (1968) described the reading process as one of recognition of words by a procession of rapid fixations which merge together with material recognized in such a way as to build up an impression of a meaning- ful message. The components for learning to read include recognition of printed words from whatever cues can be used--total configuration, the letters composing them, the sounds represented by those letters, and/or the meanings suggested by the context. Travers (1975) conducted research on a new theory of word recognition which holds that word recog- nition is accomplished by filtering visual feature information from the printed word through a hierarchy of letter, letter-cluster, and word detectors. Of the studies conducted, two studies demonstrated that skilled readers draw visual information from all the letters in a word at once, rather than from one letter at a time; and that statistical co-occurrence of letter sequences affects the perceptibility of those sequences, independent of their pronouncability. This information appears to suppOrt in part the observations made by Carroll. Goodman, Burke (1973) defined reading as a pro- cess in which written language conveys meaning between writer and reader. The reader uses graphic, syntactic, and semantic cues to get to the message. In this 13 definition and letter combinations graphic refers to the letters of which words are composed, it would appear then that theories as diverse as those coming from read- ing researchers to psychologists include the recognition of visual graphmic components as an integral part of the reading act. Characteristics of the Word Recognition Process Marzano and Distefano (1975) indicated that the whole-word characteristics are strong predictors of recognition difficulty at all levels; sound symbol char- acteristics are significant predictors up to the fourth grade. This was interpreted as evidence that beginning readers use sound symbol and whole-word cues but grad— ually become less reliant on sound symbol information. In an effort to increase the predictable variance in word recognition difficulty, another predictor (previous exposure to a word) was experimentally entered into the regression equation. It was found that exposure to a word was the strongest predictor of word recognition. This conclusion is in contrast with research reported by Bond and Dykstra (1967), who stated that the child's ability to name letters is one of the best pre- dictors of beginning reading performance. A study by DeHirsch, Jansky, and Langford (1966) yielded similar results. 14 However, that letter-naming is a valid predictor is not without controversy. Samuels (1969) differed with the import assigned to letter-naming and argued instead that knowledge of letter sounds is the more relevant variable. Research by Ohnmacht (Samuels, 1969) demonstrated that letter—trained subjects per- formed no better on transfer tasks than control subjects and that the letter- and sound-trained group performed significantly better than either group. Johnson (1970) and Samuels (1970) both reported the relative- ness of letter-naming knowledge on transfer tasks. Knowledge of both letter names and letter sounds appears to be indispensable for further acquisition and develop- ment of hierarchical word recognition skills. According to Gibson (1962), identifying graphemes and letters accurately involves a process of focusing on critical features or dimensions of differences. In a study using letter-like forms, Gibson showed that normal Children between four and eight years old learn something which may make possible better discrimination of letters. She analyzed the visual characteristics of actual let— ters (printed capitals, upper case) in terms of the number of strokes, straight versus curved lines, angles, 15 open versus closed forms, etc. This analysis provided a set of characteristics or rules on the basis of which one could generate artificial graphemes that follow these rules. Twelve such artificial graphemes were constructed according to the same constraints which appear in printed capitals. From these standards, four types of transformations——line to curve or curve to line, rotation reversals, perspective transformations, and topological or close and break transformations--were constructed for each standard form. Subjects' errors were classified according to the kind of transformation mistakenly matched with the standard. Gibson found that errors decreased with age but Av: .e that more errors were made on some transformations than pgw‘,gv‘ , .r" z . *s" .’~~ 47‘ ‘ others. Errors were greatest among four-year-olds;‘ . (; Al even though there was a slight decrease with age, they remained high across all grade levels. The other three transformations showed significant decreases in errors with age, so that eight—year-olds made few errors. Thus, certain kinds of transformations in let— ters seem to become more distinctive as children grow older, allowing them to discriminate among letter forms with increasing accuracy. Perspective transformations, on the other hand, are not distinctive, and must be tolerated in order to discriminate letters which appear at different angles of sight. Gibson concluded, 16 . It is the distinctive features of grapheme patterns which are responded to in the discrimina- tion of letter-like forms. The improvement in such discrimination from four to eight (in normal chil- dren) is the result of learning to detect these invariants and becoming sensitive to them. Reading specialists and psychologists have been interested in whether words are processed letter by letter or in larger units. Research by Williams (1970) in which strategies children use when they recognize words was explored. To measure the effectiveness of two different methods of training children to attend to the critical features of letters, first grade urban children were presented two pairs of letters (similar and dissimilar) simultaneously or successively. It was found that with highly similar stimuli the successive problem was less difficult than the simultaneous problem; while with dissimilar stimuli the successive presenta- tion was more difficult. On the basis of this finding, a more complex experiment was designed which combined highly similar letters into trigrams. Similar results were obtained as in the first experiment. From two addi- tional studies it was found that children with some reading training used the initial and then the final letter of a word as the most important cues in word recognition. A study of the cues used by young children in identifying a word aurally revealed that the final and the initial consonant syllables were chosen more 17 frequently than any of the given five cues. This study indicates that visual word recognition experimental techniques are feasible in studying aural modality. Kuenne and Williams (1972) further supported the above findings in a study conducted to investigate a series of hypothesized cues using aural stimuli (nonsense syl- lable trigrams) by adapting to the oral mode an experié mental technique used successfully in visual word recog- nition studies. It was found that: (1) children in visually discriminating between two similar-looking words of one syllable tend to use the initial grapheme, (2) in aurally discriminating between two similar- sounding one—syllable words, Children tend to use the final segment or rhyme, (3) the first unit of recogni— tion used visually is the single-letter segment, and (4) the unit of recognition used aurally is a two-phoneme segment made up of a vowel and consonant combination. Instructional Techniques Used in Word Recognition Discrimination Training It has not been determined whether the kinds of discrimination skills described by Gibson are important for more complex reading requirements like word recog- nition. Vernon (1957) has indicated that children who have reading problems show perceptual difficulty in their inability to recognize significant details, distinguish 18 one letter from another, and have confusion about the direction of letters and words. Frostig (1964) postu- lated that: (a) children who could not recognize words often had disturbances in figure—ground perception, (b) children who had difficulty in recognizing a letter when written in different sizes or colors had poor form constancy, and (C) children who had rotations or reversals indicated difficulty in perceiving position in space. But Gibson has shown that as children grow older errors of this type seem to disappear. In order to find out if this finding was agreeable with those of read- ing experts, Ash and Buckland (1974) conducted a study to investigate the visual discrimination abili- ties of children who varied in their ability to recog— nize letters and words. Word recognition skill was selected as the variable to be related to visual dis- crimination performance because it is a fundamental reading behavior that is not confounded with other, more cognitive reading skills. Three important questions were posed by this study. First, what is the overall relation between grapheme discrimination skills and word recognition ability? Second, do qualitative differences in grapheme discrimination exist between children of varying word recognition skills; do they make the same kinds of errors? Third, do the relations between grapheme 19 discrimination skills and word recognition ability vary with age? First, second, and third graders were the sub— jects used in the study. Measures of word recognition and intelligence were obtained on 87 children. All subjects then performed a visual discrimination task that uses artificial graphemes as stimuli. The task required subjects to match the standard grapheme with an identical form. Errors on this task were classified into six categories. A two—way multivariate analysis of covariance was performed. In the analysis intelli- gence test scores were covaried. The main effect for grade was significant, while neither word recognition skill nor grade x word recognition skill interaction approached significance. The procedure used by Gibson was replicated. Conclusions of the researchers are: visual dis- crimination skills of the type needed to discrimi- nate between artificial graphemes do not seem essential for the word recognition aspect of read- ing. Apparently the time and expense currently devoted to visual training in the school should be re-evaluated especially if the purpose of such training is to enhance letter and word recognition abilities. In a recent review of the literature relating visual perceptual abilities to school learning, Larsen and Hammill (1974) discovered that in the vast majority of such studies IQ was inadequately controlled. Analy- sis of the present data without using IQ as a covariate 20 obtained large significant relations between word recog- nition ability and visual discrimination skills. How- ever, when the results of IQ variation were controlled, no relation was obtained. This result carries with it some important implications for public school personnel who are engaged in the diagnosis of learning difficul- ties. The efficacy of placing Children in special learn- ing environments designed to improve alleged visual perceptual deficits is called to question. Considering the outcomes of the Larsen and Hammill review and the present study, one can conclude that visual skills are not independently related to achievement, but rather are confounded with other. more global measures like IQ. Jansen (1972) reported research on the effects of visual and auditory perceptual aptitudes and letter discrimination pretraining on word recognition in which alternate methods were used with 70 kindergarten Children. The children were trained to recognize eight printed words in a vocabulary list by a mixed-list, paired-associate method. Four of the stimulus words had visual response choices (pictures) and four had auditory response choices (spoken). Both methods involved a memory component; the stimulus and two response choices were presented sequentially. Method of instruction was also varied by two types of letter dis- crimination pretraining in which response choices were 21 presented auditorily. The predictive measure of visual perception was the Frostig Developmental Test of Visual Perception, and the measure of auditory perception was a composite of the Memory subtest of the Test of Auditory Perception and the Sentences subtest of the Weschler preschool scale of intelligence. The results indicated that three of the four predicted interactions between measures of perceptual ability and method of letter discrimination pretraining were not significant. One significant interaction was between visual perceptual ability and method of pretraining. In contrast, Venezky (1971) questioned the assumption that the learning of letter names in their proper sequence is a prerequisite for literacy. There is disagreement over the value of early letter-name training. It is variously said to aid in letter or word discrimination, to aid in attach- ing sounds to letters, and to interfere with both of these tasks. An analysis of the letter names and of experimental and pedagogical evidence lends little support to the claims of letter-naming benefits. Only 16 English letter names begin with a sound they represent, and of these, seven (the five vowel letters, plus 0 and q) do not begin with the sound introduced first in most reading programs. In several countries including the 22 UniteT States, the Soviet Union, and Israel, letter— name knowledge has been found to interfere with learning to attach sounds to letters. But letter—name knowledge has also been shown to be one of the best single predi- cators of reading success, and no matter what can or cannot be shown experimentally about the utility of letter names, they are efficient labels for the letters and an inseparable element in the popular concept of reading instruction. Assuming that letter names are an aid to word recognition, success in early reading, it becomes necessary to study how Children are taught or how they learn letter names. Gibson (1970) reported research in which children given a stylus that leaves a trace spent significantly more time in voluntary scribbling than children with a nontracing tool. This was interpreted as evidence of visual stimulation as a reinforcer, or source'of motivation. Gibson Schapiro, and Yonas (1968) reported that letters are distinguished from one another by virtue of their dis- tinctive patterns. This research yielded data showing the significance for discriminability of dimensions such as verticality, horizontality, diagonality, curvature, openness, or closeness. Gibson concluded that children can discriminate the presence or absence of two specific letters that are pointed out to them (song, ring, team, read, chop, chin) but that this is not the ability that 23 leads to detecting common spelling patterns among items. Instead, she proposed that a child learns to search for invariant patterns and a discovery of them is necessary for transfer to new problems. The evidence comes from a series of studies, one of which included a searching set condition in which pairs of words were presented to children, along with instructions that induced a general hint, specific hint, or no hint. The result suggests that it is better to get only a general hint to search for invariant features in a stimulus array than it is to get "specific” help. Lott, Smith, and Cronnell (1968) conducted research that appears to support the notion of search for invariant patterns or regularities. Visual—Perceptual Problems To the more recent advocates (Bender, 1957; Frostig, 1967; Anapolle, 1967; Cruickshank, 1972), visual perceptual problems have been trumpeted as a primary source of difficulties in learning to read. The perceptual deficit hypothesis operates under the assump- tion that reading disability is a resulting effect of inaccurate perception of letters and words. In contrast, research by Vellutino, Stegar and Kandel (1972) and Vellutino, Smith, Stegar and Kaman 24 (1974) was conducted in two experiments that indicate disabled readers accurately perceive letters, symbols, and words, but incorrectly label them because of diffi- culty in making verbal associations. Subjects ranged from second to eighth grades. Poor readers achieved better performance in visual recall of words presented tachistoscopically than they did in pronouncing those same words. Their performance was similar when they were required to graphically reproduce the stimuli. This was possible even though they had a large number of spatial and sequential errors in oral reading of the stimuli (ex. was/saw, bin/din, cob/cod, lion/loin, snug/sung). Allington et a1 . (1975) conducted a study to further examine the hypothesis that a perceptual deficit is not the major difficulty for poor readers. Poor and normal third graders were presented with high-frequency, low-discriminability words in four varying perceptual tasks. It was predicted that poor and normal readers would differ significantly in their ability to read tachistoscopically presented words in isolation but that their achievement would not differ significantly on the remaining tasks. An analysis of the results suggests that (1) poor readers have developed the ability to identify and match the necessary distinctive features, whatever they may be, for successful completion of this 25 task; (2) reading disability cannot be attributed to a deficiency in visual memory; and (3) a perceptual deficit cannot adequately explain reading disability. Lahey et a1. (1975) questioned the effectiveness and appropriateness of perceptual training programs. They proposed the following questions: (1) Can visual perception be trained at all? and (2) What success have we had in such projects? Perceptual motor training pro- grams emanate from the overtly expressed or covertly implied assumption that perception a la Frostig, Getman, Garsch, etc. can be developed or remediated. But this belief is by no means conclusively demonstrated by empiri- cal research. This study considers the problem of training perceptual skills in young children. However, it approaches this problem from a behavioral point of View rather than from the point of View of traditional perceptual-training programs. The study attempted to isolate a single academically relevant behavior and directly modify it rather than attempting to deal with the problem indirectly through the underlying process of perception. In Experiment I 40 Ss from Head Start preschools, private preschools, and children enrolled in a county-wide summer remedial reading program who had completed the first, second, and third grades were studied. At the 26 elementary levels these children were selected by their teachers as being the poorest readers in each school in a predominantly rural Florida county. The test consisted of 50 match-to-sample items made up of a sample letter and six choice letters: C I iocmly SDRBTI The items were printed 10 to a page in standard type. Subjects were given repeated simple directions and additional help if the experimenter observed that a sub- ject was not following instructions (no additional help was given to subjects who were simply making incorrect choices; i.e., help was given only to those who were circling every letter, the sample letter, etc.). There- fore, the study dealt with a fundamental aspect of visual discrimination: the discrimination of single letters presented as printing on paper under untimed conditions. In all groups as in previous studies by Gibson, Smythe, Stennett, Hardy, and Wilson, most of the errors involved two letters which differed only in orientation (d and Q) or a single feature (1 and i). There is considerable evidence suggesting that good readers differ significantly from poor readers throughout the elementary years on visual discrimination tasks where the stimuli are presented tachistoscopically 27 for very brief periods of time. The evidence thus sug- gests that these children may need training in speed of discrimination. These data put into questions theories that assume deficiencies of this type of visual discrimi- nation persist into the elementary grades as a major source of reading difficulties. Two operant procedures were used in Experiment II, one with and one without a fading technique. Thirty low-income preschool children were selected for the study. They participated individually in 30-minute sessions for 8 consecutive school days. The Ss were presented with a series of match-to—sample items. The results of these experiments suggest two things. First, there are large individual differences in the accuracy of letter discriminations among children entering the first grade. Second, these problems in letter discriminations which have proved refractory to traditional interventions can be quickly and effectively modified through operant techniques. Teachers need not and should not wait for changes in an inferred autonomous developmental process (Piaget, 1969). Nor should they wait for the possible neurological or pharmacological interventions. There are two possible explanations for the significant changes in letter discrimination. First these procedures focused directly on changing the 28 behavior in question instead of attempting to train an inferred "process" of perception. Second, the studies employed the systematic application of operant rein- forcement. Summary The review of related research in this section has been concerned with word recognition and how it comes about. There is an enormous amount of research on word recognition but there tends to be little agree— ment as to how this ability is developed. However, all agree that in order to read one must first learn to recognize and distinguish between letters of the alpha- bet because that is the basis upon which words are formed. Word Reversal Difficulties Because of the difficulty experienced by a large segment of the school population with word reversals or confusing similar words, many learners seem never able to perform the reading act effectively. This causes them to lose efficiency and motivation even though they have the ability to perform the reading task. These learners are not able to apply a specific technique of word recognition to their learning style. An individual will never become a good reader who can get meaning from the 29 printed page until he develops some efficient method for achieving this task. Interest in reversals of letter order and orien- tation in reading and writing stems primarily from the work of Orton (1937), who viewed childhood dyslexia as one element of a developmental syndrome that has as its basis an anomaly of cerebral dominance. Orton wished to establish a causal link between two observations in forming this neurological conception of reading disabil- ity: first, that Children with reading disability tend to have poorly established or unstable lateral prefer- ences, and second that they tend to reverse letters and words in reading and writing. Such difficulties were seen as related to manifestations of a failure of one cerebral hemisphere to become dominant. This conception has been Challenged by some workers in the field (Schonell, 1948; Burt, 1950; Vernon, 1960) and supported by others (Zangwill, 1960; Critchley, 1964). Types of Reading Reversals The term "reversals" is used to describe a variety of different kinds of errors in orientation. They are among the most troublesome errors made by dis- abled readers. The problems of full reversals, part reversals, axial rotations of letters, and other orien— tational confusions in word recognition interfere with 30 successful reading development. Any child who makes an extreme number of such errors has a limiting condition that must be corrected before continued growth in read- ing can be expected. Strephosymbolia, meaning ”twisted symbols," is a term used by Orton (1937) to designate the striking tendency to distort order in the recall of letters. He called individuals exhibiting this type of behavior in reading "strephosymbolics." Confusions of single letters similar in configu- ration are called static reversal errors. 'Examples of this type of error include saying pig for dig, bad for pad, but for put, and big for pig. Confusions on b, p, d, and q are most common; n and u are also often confused, the individual saying mouth for month. Static reversals represent reversed or inverted orientation of letters. Confusions characterized by a reversal of the sequence of letters in a word are called kinetic rever- sals. Word recognition errors of this type include saying pa for pg, gay for was, tub for pgp,‘dgp for Egg, and ggd for dgg. In other words, a right to left sequence is used for the reversible words. Kinetic reversals, static reversals, and transposition of words in a sentence are sometimes associated with facility in mirror reading. Reversals are not the commonest types 31 of errors in word recognition, as they are found among the errors of only about one out of ten reading disability cases. However, they are very significant and deserve careful analysis and treatment when they persist. Perhaps reversals are very prevalent in young Children beginning to read because they have learned to recognize people, animals, landscapes, and objects both from first-hand visual experience and from viewing pic- tures and diagrams. From these experiences they tend to think that the difference in position of letters such as p and d, or the order of letters in words like Egg and ggy is not important. They generalize that they can recognize letters and words about as well when upside down or sideways as when they are right side up. Characteristics of ReadingrReversals In a study of reversal tendencies in reading, Gates and Bennett (1933) concluded that known visual defects existed with twice the frequency in the rever- sals as in the nonreversals group. Further study of the degree and type of visual defects of such pupils is indicated as a promising means of revealing causes of reversal tendency. Indeed, considering merely these base data, it would appear that visual defects of some sort are the most conspicuous characteristic of the is. 32 reversal group thus far. These findings were corrob- orated in a study by Hildreth (1933), which indicated that defective vision may be a contributing factor in reading disability, both on account of distortion or vagueness of visually perceived word and letter forms, as well as tension, fatigue, strain, or nervousness resulting from compensatory effort. Barger (1953) recently reported some spectacu- lar results in remedial reading with children showing strong reversal tendencies, by having them read mate— rials by looking into a mirror that reverses the material both vertically and horizontally. He believed that this corrects a functional neurological peculiarity that causes these children to see the print normally. Leavell (1955) developed a method of overcoming direc- tional confusions in cases of crossed dominance by train- ing the nondominant eye to coordinate with the dominant hand. Blau (1945) attempted to show that left—sidedness in general is a form of neurotic negativism or defiance. While as a general explanation of left-sidedness his theory seems farfetched, it may very well apply to a few cases. Parks (1953) reported two cases in which a strong directional confusion in reading seemed to be a dis— guised form of negativism. This information lends cre- dence at least in part to Blau's theory. 33 Rubenstein (1975) conducted a study to investi— gate the effects on oral reversals of various contextual conditions ranging from no context to highly predictable contexts among white, middle-class male first graders. It was hypothesized that increasing the predictability of a reversible word by manipulating contextual condi- tions would result in a decrease in reversals. The rationale for the study was provided by Shannon's (1948) theory of communication, which enun- ciated the mathematical concept of redundancy. Smith (1971) and Weaver (1963) have attempted to apply redun- dancy to the reading process. The present study was designed to demonstrate the relationship between language redundancy and reversals in a beginning reading popula- tion to suggest an alternate to the organismic view posited by Orton (1937). A list of 11 reversible words was derived from a pilot study that was devised to determine the ambiguity of potentially reversible words. The resultant Contex- tual Reversals Test consisted of four "conditions" rang- ing from low redundancy to high redundancy. The condi- tions were: (1) Reversible Word in Isolation, (2) Rever- sible Word in a Semantically Ambiguous Sentence, (3) Reversible Word in a Semantically Ambiguous Embedded Paragraph, and (4) Reversible Word in a Semantically Unambiguous Sentence. 34 Reversible words were held constant across all conditions. Condition I consisted of reversible words in isolation; Condition II of the same reversible words in ambiguous sentences; Condition III of the same sen— tences from Condition II embedded in paragraphs designed to increase redundancy; and Condition IV of the rever- sible words embedded in unambiguous sentences. The results demonstrated that as redundancy increased, reversal frequencies declined in the average reading population used in the study. It was concluded that the manipulation of contexts with varying amounts of redundancy would have an effeCt on reversal tendencies among beginning readers, within the limitations of the study as a remedial as well as a preventative procedure. The study also demonstrated a practical reading application of Shannon's (1948) theory of communication by showing the extent to which the processing of redun- dant information enabled subjects to choose the correct form of a reversible word. In addition, Smith's (1971) assertion that as redundant context increases, readers use less visual information, was supported; and his theory was extended downward to include the immature readers as well. Finally, this study suggested that beginning readers do use bilaterial context while read- ing, an implication contrary to Weaver's conclusions. 35 Liberman (1971) conducted a study of letter confusions and reversals of sequence in the beginning reader. The subjects for the study consisted of 18 children selected from the second grade of an elementary school system. The purpose of the study was to provide a more systematic approach to the questions of relation- ship of sequence and orientation reversals to each other and to different aspects of reading mastery as well as the nature of the general error pattern in the disabled reader. The subjects were given a list of 60 real- word monosyllables including agroup of primer-level sight words, a group of nonsight words, and word— forming reversals of both types of words, where such were possible. Responses were recorded on tape as well as being transcribed during administration, to check on the accu- racy of transcription. Each child's responses were analyzed for reversals of sequence and of orientation, for consonant and vowel errors, and for total errors. The word list was administered twice to each subject-~once at the end of the school year and again in the first week of the following school year. Data from the two presentations were combined in scoring the responses of each subject, but were available separately for assessment of test-retest reliability. The follow- ing testing procedures were used: 36 Gray-Oral Reading Test, Form A. Administered by standardized procedure. Raw paragraph scores based on Gray's system of weighting time and number of errors were used to evaluate the subject's performance, rather than grade-level equivalents. Single-Letter Presentation (Tach) List of 100 items in which a given letter was to be matched to one of a group of five including four reversible letters in manuscript form (b,d,p,q) and one nonreversible letter (e) that was added as a reliability check. There were 20 such items for each letter. The order of the 100 items was randomized as was the order of the multiple choice sequence for each item on the answer sheet. A brief training session was provided for each child. The results revealed that for most of the Chil- dren reversals accounted for only a rather small propor- tion of the total of misread letters; thus only some poor readers reverse. Certainly it is important to explore the other differences among children who do and do not make reversals of sequence and orientation. It may be that reversals loom larger in importance in certain Chil- dren with particularly severe and persisting reading difficulty. Examination of the interrelations among various reading errors showed that the two types of reversals 37 are wholly uncorrelated. This is a finding of consider- able interest since both were considered by Orton and subsequent investigators to be manifestations of an underlying tendency to reverse the direction of the scan. That View cannot easily be reconciled with two additional findings: First, among reversible letters, vertical reversals occurred with as great frequency as horizontal reversals. Second, confusions among reversible letters rarely occurred when these letters were pre- sented singly, even when briefly exposed. The relationship between both types of rever- sals and other errors in reading syllables showed indi- vidual error rates on vowels and consonants correlated highly with each other. An analysis of the nature of substitutions among reversible letters (b,d,p,q) showed that the possibil- ity of generating another letter by a simple l80-degree transformation is a relevant factor in producing a rela- tively high rate of confusion among letters, a finding that is in agreement with conclusions reached by Davidson (1935) and by Gibson and her associates (1962). Other observations indicated the importance of linguistic determinants: differences in pattern of confusions among p,d,p, and g in real words and nonsense words showed that misperceptions even of reversible letters 38 are content dependent and not merely an automatic conse- quence of optical reversibility. Moreover, the substi- tutions tended to differ from the presented consonant in one phonetic feature. Finally, relatively few confu- sions of these letters occurred when they were presented in isolation rather than in word context. All of these observations point to the conclusion that the character- istic of reversibility is not by itself a sufficient condition for confusion. Summary Research has been done on reading reversals in an attempt to clarify the causes and to develOp a method or technique by which they could be eliminated. Attempts have been made to establish whether both types of rever- sals, kinetic or static, were correlated. Color as a Cue Researchers have been interested in the role of color in the general area of learning for a number of years. Therefore a considerable amount of literature exists on the application of color in education. Color has been most commonly used as a contextual cue in a paired-associative learning task in the majority of these studies. There is a need for a more accurate means of word recognition for some learners, especially those 39 individuals who are experiencing problems of sequencing and orientation of words and letters associated with reversals. Because of the requirement for clearly discriminated stimulus as a key to accurate word recog- nition more effective techniqueinn relation to this need seem to offer an avenue to this need. Vanished hue color cues appear to be a valuable technique for helping the learner focus on the discriminator of words to be learned. The Use of Color in Paired—Associate Learning As early as 1926, Pan (1926) had subjects asso- ciate names and faces presented on picture postcards and found that they made more recall errors when the postcard backgrounds were different than when they remained the same. However, Dulsky (1935) conducted the most care— fully conceived research. By varying the color of the cards on which paired-associates were placed, he found that recall is reduced when the color background present in the learning situation is Changed for the retention test. His subjects were presented with pairs of nonsense words on two types of backgrounds. They were classified as (l) homogeneous, that which featured a different color for a pair or the same color for all pairs of words; (2) heterogeneous, that which featured the stimu- lus half of each word card colored and the response half 4O gray and vice versa. Three conditions were observed for learning: replication of the learning conditions, interchange of colored backgrounds, and change to all gray backgrounds. Recall was most accurate when the gray stimulus backgrounds and colored response back— grounds remained the same as in the learning trials. The learning colors were Changed then when stimulus or total backgrounds were changed. Weiss and Margolius (1954) conducted research to determine the effectiveness of the retentive support provided by context when the primary stimuli were changed and to determine whether or not the presence of a dis- tinct nondistracting context stimulus would affect the learning of a response. This consideration differs from the above in that it is concerned with the relationship between environmental stimulation in the learning and retention situations. The question now being raised is: Regardless of the condition under which retention will occur, might there be any advantage for learning if certain context stimuli are presented throughout the learning session? Since the learning of complex responses requires a series of trials, it is conceivable, in accord with behavior theory, that the presence of an extraneous stimulus to which a desired response could provoke the arousal of that response in the early trials, before it is strongly associated with the appropriate primary 41 stimulus. Weiss and Margolius stated the following con- cerning their findings. The research confirms the theoretical prediction that the representation of context stimuli, which had been present consistently while learning was occurring, would aid retention of the learned responses. The retentive support provided asso- ciated context stimuli occurs not only when the original primary stimuli are present, but also when generalized primary ones are introduced. Thus the findings of Dulsky are confirmed and extended. . . . Presence of distinct context stimuli facili- tates acquisition of the stimulus-response associa- tion. The Weiss and Margolius procedure has served as a model for subsequent research designed to determine the functional component in paired-associate learning. Underwood et a1. (1962) reported that subjects could not recall a list of low-meaning trigrams that had originally been learned on different-colored cards; but, when meaningful words were learned instead of low-meaning trigrams familiar colors became the functional stimuli. But because the adult subjects were more accustomed to responding to words than to colors the familiar words were the functional stimuli. Hill et a1. (1962) felt that the form and color components of a stimulus might summate to evoke a response, whereas either presented singly would not. After speculation they had subjects learn nonsense word/color-common word pairs in a nonanticipation 42 sequence in which pairs were presented together in learning trials and stimuli were presented alone in the alternate testing trials. The best results were attained by the sub- jects who learned the components separately and then responded to a combination in final testing. Because many subjects responded correctly to only one component before a combination was formed, the researchers rejected the summation theory. They concluded that two cues were more helpful than one because each subject was free to choose his functional stimulus. Experimental research has shown that paired- associate learning could be facilitated by surrounding each verbal stimulus with a different color. Saltz (1963) conducted research designed to determine if such facilitation occurs when the colors cannot readily be used as cues. Critical manipulations involved alternat- ing learning trials and test trials. Availability of color as cues was reduced by presenting them during learning trials but not test trials, or during test trials but not learning trials. The results indicated that color differentiation can significantly facilitate learning even when color cannot readily be used as a cue. Concerning the findings, Saltz commented: 43 . [The] principal contribution of the present study is that it operationally distinguishes between the effects of cognitive and sensory differentiation and indicates that both may facilitate verbal learn— ing. . . . The greatest amount of facilitation occurred in the color—color condition. . . . Only condition in which color could readily act as a cue. . . . Stimulus selection as well as differen- tiation acted to facilitate learning. . . . Only one in both cognitive and sensory differentiation operate simultaneously. The possibility remains, therefore, that the superiority of the color-color condition may be accounted for by assuming a simple additive relationship between cognitive and sensory differentiation. This study indicates that color can facilitate learning even when the color cue is not readily accessible as a cue, depending upon the manipulation of treatments. Whereas, Birnbaum (1966), alternating study and test trials in paired-associate learning, provided secondary color cues on the study trials only. Subjects were presented with 10 consonant-vowel-Consonant trigrams (CVC) paired with nonsense syllables, each trigram out- lined with distinctive color during the study trials. After intervening tasks, she presented half of her subjects with the same task and the others with the same stimulus-response pairs but different secondary color cues. The latter group did less well on the post- test in spite of the fact that the stimulus-response pairs were unchanged. This is an indication of the potency of the color as well as the need for careful planning if it is to be used to facilitate learning. 44 The Use of Color Cues by Good and Poor Readers Otto et a1. (1968) conducted a series of four studies dealing with good and poor readers' utilization of selected cues in paired-associated learning. Spe- cific cues considered were color, order of presentation, and verbal mediators. Answers to two basic questions were sought: Do the selected cues have a facilitative effect upon children's paired-associate learning? Is the learning of good and poor readers affected differ- ently by the additional cues? Color was shown to have a positive effect upon Iearning when intralist similarity was high, but there was no reliable differential effect for good and poor readers. Serial as opposed to scrambled order of presentation was shown to enhance both initial learning and recall. Instructions to use verbal media- tors also enhanced learning; but again there was no differential effect for good and poor readers. In the four studies reported, it was shown that certain cues affect the rate of learning a paired- associate list, but each subject's selection from available cues is also an important determiner of learning speed. The first three experiments indicated that color cues may facilitate learning and/or reten- tion if other, more powerful cues are absent or ignored 45 by the subjects. In Experiment II, color enhanced learn- ing, while in Experiment III, color did not affect the learning rate, but the removal of color had an adverse effect on subsequent performance; thus, color cues were better than no cues at all. It was demonstrated in Experiment IV that verbal mediators are extremely power- ful cues regardless of type of materials, grade, and reading ability. The questions that still remain per- tain to the relationship among all possible cues. Otto asked: . . . Can they be placed on a continuum of facili- tation? If so how do they stand in relation to each other? If a certain type of cue aids initial learning, does it also aid retention or relearning? Differences in rate of learning by good and poor readers were apparent in Experiments I and II. In Experiment IV, it was suggested that good readers may be more apt to provide their own verbal mediators than poor readers, had more learning experiences and verbal experi- ences on which to base future learning strategies, and have more successful verbal learning experiences than poor readers. The material in the first two studies, where good and poor did differ, the stimulus materials and tasks were much more reading-like than the fourth study, which may have enabled the good readers to capi- talize on their more successful and broader verbal experiences. 46 The trends noted in Experiment I indicate that the good readers and the sixth graders benefited more from color cues, but the opposite was noted in Experi- ment II. The increased intralist similarity in Experi- ment II may have affected the direction of these trends; as shown in Experiment III, the second graders learned faster with low-intralist similarity. Color as an Aid to Visual Perception in Early Reading Jones (1965) examined color as an aid to visual discrimination of words and letters among nursery school children. He reasoned that nursery school children are old enough to take visual matching tests but too young to have acquired reading habits that would influence their responses. The task comprised a pair of matching tests, one black and one with color, of six English reversal letters (p,q,u,n,d,b) followed by a second pair of matching tests in black with and without color of six English words transposed into an unfamiliar script to control for learned reading responses. On the basis of the data, Jones concluded that without color the task was "at least three times" as difficult as with color, even when possible color matching was considered. Furthermore, he noted that the subjects strongly 47 preferred the colored test materials. The implication seems to be that color may have value both as an aid to discrimination and as a motivational device in early reading. The Use of Color to Isolate a Single Stimulus Hedwig von Restorff (1933) demonstrated with her research that more rapid learning of an item in a list which was different from other items in the list was possible. Her findings have been confirmed by other research. The facilitation in learning a particular item in a list produced by isolating that item or making it distinctive from other items has come to be called "the von Restorff effect." Many variables can be cited which should influence the von Restorff effect in a somewhat similar manner as does meaningfulness in formal stimulus similarity. Formal similarity refers to the degree of visual distinctiveness among items. Samuels (1968) used first graders in ExperimentIPuOdeterminethe effect stimulus similarity and stimulus isolation have on the von Restorff effect when only a single stimulus within the list is isolated by color. It was predicted that when stimulus similarity is high, rate of PA learn- ing will be faster for the S-R pair in which the stimulus term is isolated. When similarity is moderate or low, there will be no difference in rate of PA learning 48 between S-R pairs having an isolated stimulus term and S-R pairs having nonisolated stimulus terms. During learning trials an isolated stimulus term was presented in red, while the nonisolated stimulus terms were pre- sented in black. After each five learning trials, a transfer test was given to determine if Ss were able to give the correct response when a formerly isolated (red) stimulus term was presented in black. During learning trials in the high-similarity list more correct responses were given to the stimulus in red. At transfer there was a reversal of significance, and fewer correct responses were given to the stimulus formerly in red. On medium- and low-similarity lists the difference in correct responses to red and black stimuli was not sig— nificant at learning or at transfer. Samuels stated: . . While significant differences during learning in number of correct responses were found in one of the three lists favoring the isolated over the non- isolated words, the critical issue was one of trans- fer or ability to give the correct response when the word formerly in red was presented in black. . . . Ss responded to isolated words on the basis of color. On transfer the color cue was absent, and Ss did poorly with the formerly isolated words. . . Lists where discrimination on the basis of letter form was easier, color was a less potent cue. For these lists it appears that 88 tended to use letter form as a cue for responding even for iso— lated words. Samuels conducted Experiment II using adult subjects because of the possibility that Ss in Experi- ment I did not have sufficient experience with the 49 transfer task to realize that color was an irrelevant cue and letter shape the primary cue. The purpose of Experiment II was to determine whether it would be pos— sible to facilitate PA learning by (l) isolating with color one of the stimulus terms, (2) providing enough transfer tests for £3 to realize that in order to respond correctly at transfer he would have to focus on letter shape during learning, and (3) using Ss with SOphisticated learning strategies. Thirty college Ss learned a high—stimulus similarly paired-associate list. During learning trails one stimulus was in red. The other stimuli were in black. 0n transfer trials all stimuli were in black. Learning and transfer trials were alternated. The results were similar to those found in Experiment I for the high—similarity list. It appears that the magnitude of the von Restorff effect is influenced by stimulus similarity. Samuels concluded: . . During learning trials, when color cues were present, significantly more correct responses were given to the isolated than to the nonisolated S terms. During transfer, when color cues were removed, significantly more correct responses were given to the nonisolated S terms. Correct responses to iso- lated S terms on learning trials and failure to do so at transfer indicated that Ss were attending to stimulus color and not stimulus shape. The fact that subjects focused primarily on the irrelevant dimension of color rather than the relevant dimen- sion of shape for the isolated S—R pair is surpris- ing first, because the number of alternations between learning and transfer trials was sufficienttx>indicate that color cues were being removed at transfer and second, because Ss were told that some nonsense syllables would appear at learning and at transfer. 50 It appears that the learning strategy of college Ss was similar to that of the children in Experiment I, because they also found it difficult to focus on a less salient but critical one in the presence of a more salient cue, 22192:- In contrast, Allington (1973) conducted research using kindergarten Children to determine whether sub- jects who received one of three treatments of color cues in an instructional program differed significantly on three learning tasks. Four letter-like transforma- tions and their up-and-down, left-to-right transforma— tions were selected for use in the study. These eight figures were presented in one of three treatments: (1) no color added to letter-like figures, (2) maximum color added to the letter—like figures, and (3) maximum color added and then vanished from the letter-like figures. A program sequence was developed employing an audio-flash card reader. The verbal directions of the program sequence remained identical for all subjects, while the treatment of color cues varied. Subjects were pretested and posttested on their ability to match to fomn and nmtch from memory, and to associate a low-meaningfulness C-V-C trigram with each figure. 51 Significant achievement on all three tasks was found for the vanished color treatment over no-color treatment. From these results it was concluded that vanished color treatment enhances the learning of visual discrimination, visual memory, and paired— associate tasks when compared to the no-color treat- ment. This effect seems to stem from improved attention to the distinctive feature of a stimulus. Summary The studies reviewed on color as a one often demonstrated that color facilitated rate of initial learning but produced undesirable complications on a transfer task. This phenomenon was one of the major characteristics of the study conducted by Samuels. His research provided insight into the potency value of color and provided some necessary cautions for the use of color cues. The utility of color under varying con- ditions was investigated by Otto. His research demon- strated that color enhanced learning but that other cues, such as order of presentation, could have a more facilitative effect. Other research studies focusing on color as an attentional factor attempted to clarify the roles of attention, distinctive features, and color cues in dis- crimination and paired-associate learning. Many studies 52 demonstrated that poor readers and children who had difficulty with the task of visual discrimination were inferior in their ability to attend to and identify distinctive features of letters and words. Research by Jones showed that color aided his subjects in helping them to pick out distinctive features of letters and words. CHAPTER III DESIGN OF THE STUDY In this chapter, the description of the sample, the methodology, the development of the instrument, the testable hypotheses, and the pilot study are discussed. Sample The Lansing school district was selected for the study. Fifteen elementary schools were randomly selected from a list of 20 schools. The population under investigation comprised those elementary school children in second grade who were identified by their teachers as low-achieving readers who were confusing words. The total sample population for this study consisted of 174 second graders from the 15 Lansing elementary schools selected. A simple random sample was used to select 60 boys and 60 girls to participate in the study. The subjects from each of the two groups (boys, girls) were randomly assigned to one of three treatment groups. (See Figure l.) The total number of children in the sample was 120. 53 54 T1 T2 T 3 n 40 n 40 n 40 Number of subjects per cell for the treatment effect Boys 60 Girls 6O Number of subjects per cell for the sex effect T1 T2 T3 Boys n = 20 n = 20 n = 20 Girls n = 20 n = 20 n = 20 Number of subjects per cell for the treatment x sex interaction effect Figure l.--DiStribution of the sample. 55 Methodology An instructional episode was used as the data- gathering method for the study. An instrument was developed which consisted of 120 words that were paired, and bound into booklets, for each of three treatment groups. The instrument measured the effectiveness of vanishing hue color ones on the paired-associate learn- ing of previously encountered sight vocabulary words by low-achieving second graders. The investigator administered all treatments in November and December, 1976. It took approximately 15 to 20 minutes for each subject in the study. One day was designated for each school. The treatments were administered randomly in each building. The investi- gator returned to schools where some subjects had been absent. The experiment took place in small rooms in all schools. The investigator sat at a low table, which contained a tape recorder, the instructional episode booklets, 12 flash cards, and the posttest sheets. The subject sat on the right of the investigator. The instructional booklets were placed in front of the learner as needed. The cards were turned by the inves- tigator. 56 The Instrument In developing the research instrument the gen— eral idea was to use sight words that children reverse, transpose, rotate, and confuse. A review of literature revealed some of these words. The results of the Slossen and Dolch lists, which had already been administered to children, were analyzed for look-alike words that chil— dren were miscueing. These word lists had been admin- istered to children by graduate students enrolled in a diagnostic class, Reading 830E, at Michigan State Uni- versity. The list was compiled and a tally was made of the words most often missed by the children who were tested. The list contained word pairs, that is the word missed and the word it was mistaken for. The list was validated by reading experts at Michigan State Univer- sity as those that are often confused. From this list was generated a second list of words--those words most commonly missed that contained letter differentiations that are critical for‘some learners. Twelve words were paired with a word of high similarity or one that had been identified in the lit- erature as being one children confuse or miscue. The paired word contained no highlighting for any group. (See Figure 2.) 1. want 13. went 2. no 14. on 3. every 15. very 4. boy 16. dog 5. always 17. away 6. where 18. were 7. saw 19. was 8. those 20. these 9. sit 21. its 10. been 22. be 11. they 23. that 12. many 24. may Figure 2.--Words used for the instrument. For each treatment a matched set of 12 booklets was used. The only variation was the color highlight- ing. Each word was presented to the learner five times. The booklet contained 10 words. The first two presen- tations of the words were used as the instructional episode. The pairs of words were randomly arranged throughout the booklet. The words were presented 10 times throughout the episode. (See Figure 3.) Three sets of 120 3 x 5 cards were prepared for each treatment in the instructional episode. The first set consisted of the cards printed in black with the stimulus letter or letters printed with maximum no no on on no on no Figure 3.--Sample booklet showing word and its paired— associate. Words were placed randomly in the booklets. 59 highlighting on trials one through eight, while trials nine and ten presented both words in plain black letters. The second set of cards consisted of the word printed in black on white with maximum color on the stimulus letter or letters in trial one and subsequently vanished in trials two through eight, with the criterion word in plain black in trials nine and ten. The third set of cards consisted of all the letters of the word printed in black throughout the treatment. (See Figure 4a-d.) In addition to the 360 cards printed for the instruc- tional episode, 120 plain cards were prepared and placed randomly in the booklets so that only one word was revealed at a time. Figure 2 showed the pairs of words used in the study. The words were in spiral—bound booklets. There were 12 booklets for each treatment. Each booklet con- tained two paired words placed in random order. The booklets were 4” x 11” in size. The cards were bound in the booklets as a means of keeping them in the same order for each treatment. The booklets were color coded for easy identification by the investigator. They were numbered from 1 to 12, for consistency in administration. Each learner was given directions prior to the beginning of the instructional episode. The words were randomly placed in the booklets to prevent serial learn- ing. The instructional episode was presented to the 6O .mpaosuswnp amaze me» no“ waspewssemHn «0 meoESD=II.e masses xomam Q0 guess so suede so Roman ed Roman cm x Hm? O... U C ‘° H van sense sauna” 11:31 It Be sumac. S183 o c 0.: 0: on d c + 7 a Roman scans ewe we swoflm sow: sums Roman CNS 0: 0:. 0:, F 0:, oc, .//eQm gleam A s/emm. fi/emm. mMH 3v MOAOU EDSHxHue>somsoo Osos sons cums OBV .HO>OH Ho. pm ucsowwwswsm we amen one* Avaa we show House one new Eocmmsw mo Ccswocv *Ho.v Hooo. woum.¢m www.® mwov.HbH m mesmm *Ho.v Hooo. mmmv.ua mmm.vm mmmo.¢m¢ m Ewab Hwowsswmw.a msmwnuoawm we Cancwsd> .manmflnm> coco noH owcuim Opssss>flss one mo mpHSmCMII.N Canes 75 .Hm>ms Ho. one as pesosesewsm ms Once one* . . . . . m I *Hoo o sHH oem m msem mass H mmflmm o> wee. o.sHH see.I seam. omsH.I oz I oz *oso. o.sHH Hom.m seam. ommo.H oz I o> *moo. o.sHH meo.m Heme. ooom.H o: I o> spssflnsnonn .e we msse>Is eemmmmmm owmmmwwmmmemm» .xmsp macaw: Bosh Show Op sows: map so mpmsspsoo 950% can we prsmms onBII.m OHQSB 76 at .010, which is less than .01 (the critical level). Therefore there is also a significant difference between the two groups. Accordingly the magnitude of the differ- ence between the two groups is 1.0. Thus, on the average, the VC group can Match to Form approximately 1.0 words more than the NC group. In checking the comparison between MC and NC the t-value is -.444. However, the .658 level of sig- nificance exceeds the .05 control level established for the study. Thus there is no significant difference between the MC group and the NC group. Further, the table shows the comparison between the VC group and the average of the combined means of the MC group and the NC group with a t-value of 3.260 is significant at .001, which is less than .01, the critical level. Therefore, there is a significant dif- ference between the VC group and the other two groups. The means of three treatments on Match to Form from Memory, Naming, and Pairs are displayed in Figure 7a-C. The histograms in Figure 8a-c display their mag— nitudes. Naming Table 4 shows that the comparison between the VC and MC groups has a t—value of 3.623 andijssignificant 77 THE OBSERVED MEANS OF THREE TREATMENTS ON MATCH TO FORM FROM MEMORY (12) (a) T1 (MC) T2 (VC) T3 (NC) 7.8 9.0 7.98 I III I 1.2 1.0 THE OBSERVED MEANS OF THREE TREATMENTS ON NAMING (24) (b) T1 (MC) T2 (VC) T3 (NC) 17.15 21.15 15.70 I II II 4 5.45 THE OBSERVED MEANS ON THREE TREATMENTS ON PAIRS (12) (0) T1 (MC) T2 (VC) T3 (NC) 7.5 10.2 6.2 L1 II I 2.7 4.0 MC = Maximum Color, VC = Vanishing Color, NC = No Color Figure 7. --Observed means on three treatments on three tests . 78 THE OBSERVED MEANS OF THREE TREATMENTS ON MATCH TO FORM FROM MEMORY (a) L A j 1, (MC) (VC) (NC) THE OBSERVED MEANS OF THREE TREATMENTS ON NAMING (b) 21 .L 19 .. 17 di- 15 a-l- .I- T l 1 i T r (MC) (VC) (NC) THE OBSERVED MEANS ON THREE TREATMENTS ON PAIRS (c) 10 I i- . .T—I j 1 (MC) (VC) (NC) Figure 8,..-Histograms Of‘ observed means on three tests. 79 at .000, which is less than .01. Therefore there is a significant difference between the two groups. More- over, the magnitude Of the difference between the two groups is 4.0. Thus on the average the VC group can name approximately 4.0 words more than the MC group. The comparison between the VC and NC groups has the t-Value Of 5.842; this is significant at .000, which is less than .01. Therefore, there is a significant difference between the two groups. Moreover, the mag- nitude of the difference between the two groups is 6.4. Thus on the average the VC group can name approximately 6.4 words more than the NC group. The table shows that the comparison between the MC and NC groups has the t-value Of 2.219, significant at .028, which is greater than .01 (critical level). Thus with 99% confidence there is no difference between the MC and NC groups. The table also shows the comparison between the VC group plus the average of the means of the MC and NC groups has a t-value of 5.464, which is significant at .000, less than .01 (the critical level). Therefore there is a significant difference between the two groups. Moreover, the magnitude of the difference between the two groups is 5.2. Therefore on the average the VC group can name approximately 5.2 words more than the other two groups combined. 8O .Hm>ws Ho. one pa anaeseflewsm ms pmmp mss* *ooo. o.sHH eee.m meme. ommm.m mmHmm I o> mmo. o.sHH mam.m Heoa.H oome.m oz I oz *ooo. o.sHH mem.m HeoH.H come.e oz I o> *ooo. o.sHH mmo.m H¢OH.H oooo.e oz I o> .HOHH @UQQHQ .H o spsssnanona .e we CSHR>I9 eneeeemm mo mewmemamh .Mmsp wassmz one so mumsspooo snow one HO meHSMOH oseil.v Canes 81 Table 5 shows that the comparison between the VC and MC groups has a t-value of 4.981; this is sig- nificant at .000, which is less than .001, the critical level. Therefore there is a significant difference between the two groups. Moreover, the magnitude of the difference between them is 2.9. Thus on the average the VC group can name approximately 2.9 more pairs of words than can the MC group. The comparison between the VC and NC groups has a t-value of 6.812, which is significant at .000, less than .01. Therefore there is a significant differ- ence between the two groups. Moreover, the magnitude of the difference between the two groups is 4.0 pairs of words. Thus on the average the VC group can name approximately 4.0 more pairs of words than the NC group. Table 5 also shows that the comparison between the MC and NC groups has a t-value Of 1.831. This is significant at the .070 level, which is greater than .01 (the critical level). Thus with 99% of confidence there is no significant difference between the MC and NC groups on the Word Pairs task. The table also shows a comparison of the VC group with the combined means of the MC and NC groups. The t-value is 6.809, which is significant at .000, less than .01 (the critical level). Therefore there is a significant difference between the two groups. Moreover, 82 .zm>mz Ho. one pa pesosmsemzm ma same one* *ooo. o.szz mom.e mwom. mmee.m mmmwm I o> oso. o.szz Hmm.z memo. omso.z oz I oz *ooo. o.szz mzm.e meme. oooo.e oz I o> *ooo. o.szz Hmm.e meme. comm.m oz I o> HOHH mOQQHm H @ zpzzznenozc .s we ozze>Ie eneeeemm mo mewwwmwamh .Mmsp msflmm on“ so mummzpnoo sdow one mo measmos oneii.m canes 83 the magnitude of the difference between the two groups is 3.4. Thus on the average the VC group can name approximately 3.4 more pairs of words than the MC and NC groups combined. Summary The multivariate analysis of variance shows that the treatment main effect is significant at .05 levels; however, neither interaction between sex and treatment nor sex main effect was found to be significant. The univariate F-test was employed to specify on which dependent variable the treatment main effect was found to be significant. The conclusion drawn from the univariate test was that all three dependent vari- ables were influenced by the treatment main effect. The three univariate F—tests were found to be signifi- cant at the .01 level. Four post-hoc contrasts were constructed to find where the significant differences existed among the three groups. They were vanishing color minus maximum color, vanishing color minus no color, maximum color minus no color, and vanishing color minus the combined means of maximum color plus no color divided by two. As a result Of using the t-test to test all four contrasts on the three dependent variables at the .01 level, vanishing color minus maximum color, vanishing color 84 minus no color, and vanishing color minus maximum color plus no color divided by two were found to be significant on all three dependent variables. However, maximum color minus no color was not found to be significant on any Of the three dependent variables. CHAPTER V SUMMARY AND CONCLUSIONS A brief summary of the study is presented at the beginning of this chapter, followed by the conclu- sions, implications of the study, and suggestions for further research. M1 The subject of this research was the effect of color cues as a facilitator of visual discrimination on paired-associate learning. Specifically, the effec- tiveness of the use Of color cues to teach low-achieving second graders visual discrimination in paired-associate learning of previously encountered sight vocabulary words was explored. The study was designed to test the effect of employing a single color as a cue for a variety Of stimuli. It was felt that the use Of a single color might eliminate the problems previously associated with the use Of multiple color cues while still focusing subjects' attention on the distinctive features of the stimuli. The sampled population for the study included 15 schools from one school district. A random sample 85 86 was taken to generate three groups of 20 boys and 20 girls from second grade classes for each of the three treatments. The total number in the sample was 120 children. Three sets Of cards were prepared for use in the instructional episode. The first set had maximum color cues added to the stimulus word, the second set had maximum color cues added and gradually vanished from the stimulus word, and the third set had no color cues added to the stimulus word. This instructional episode was used with 120 Children enrolled in the second grade classrooms of 15 schools in the Lansing school district. Each subject was taken through the instructional episode individually. Upon completion of the instructional episode each sub- ject was administered a posttest on the ability to name the words, to name the pairs Of words, and to match the word to form from memory. The multivariate analysis of variance showed a treatment main effect of p < .0001. It also showed the test to be significant at the .05 level. However, it did not identify the dependent variable that produced the significant treatment main effect. The univariate F—test was employed to find out which dependent variable produced the treatments main effect. The conclusion drawn from the univariate F—test 87 was that with 99% confidence all three dependent vari- ables influenced the treatments main effect. The test was significant at the .01 level. Four post-hoc contrasts were constructed and were investigated to find where the specific significant difference existed. They were: (1) VC-MC, (2) VC-NC, MC+N 2 The t-test was employed (3) MC-NC, and (4) VC- to test at a .01 level. The univariate analysis of variance on the dependent variable Match to Form From Memory indicated a significant treatment effect at the p < .0058 level. The t-test was significant at the .01 level, favoring the vanished color treatment over the MC and NC treat— ments. The univariate analysis on the dependent vari- able Naming indicated a significant treatment effect of p < .0001 level. The post-hoc contrasts identified a significant difference (T < .01) between the vanished color and no color treatments and between the maximum color and the vanished color treatments. The univariate analysis of variance on the dependent variable Pairs indicated a p < .0001 level. The post-hoe contrast t-test indicated a t < .01 level on the Pairs test between vanishing color and maximum color and vanishing color and no color, both significant differences favoring the vanishing color treatment. 88 The results of the post-hoe analyses indicate that the vanished color treatment was superior to the maximum color and the no color treatment on each of the three learning tasks. Further, the vanishing color treatment was superior when the means of maximum color plus no color divided by two minus vanishing color were evaluated. Conclusion The conclusions drawn from the analysis of the data of this study are that children who received instruction that utilized vanished color cues learned the tasks of visual discrimination, visualmemory, and association Of a verbal response at a significantly higher level of achievement than Children who received instruction with maximum color cues or those who received instruction with no color cues. Finally it is concluded that the use of vanished color cues serves to focus attention on the distinctive features Of the words to be learned without producing interference at transfer. Implications of the Study to Education Much emphasis in the 1970's has been focused on reading and the right of all individuals to be provided with the Opportunity to learn how to read. However, much is needed in the realm of methods and techniques 89 to reach the needs of the diverse populations of which our schools are composed. In the area of orientation difficulties, teachers appear to be as confused as the researchers as to the causes of these problems and how best to meet the needs of the children who manifest them. In presenting the study to the reading special- ists of the schools involved and other individuals who were indirectly connected with its implementation, responses were positive. From the superintendent down the hierarchical ladder of authorization, interest in the study and its possible contribution to the body Of already existing knowledge in the area of reading was without exception. Because of the impression made on the Children by the treatments, some teachers in some of the schools came out Of the classroom and asked, "What are you doing to my children? They are so excited when they leave you." The results of this study indicate that attend- ing to appropriate features of visual stimuli improves performance on visual discrimination, visual memory, and paired-associate tasks. Specifically, this study demon- strates that vanished color cues facilitate attention in learning these three basic tasks. In contrast to previous studies using multiple hue color cues, a single hue color one that is vanished seems to facilitate 90 attention to the distinctive features of words without producing interference on a transfer task. Color cues were useful because they directed the learner's atten- tion to salient information. A vanishing hue elimi- nates the possibility of responding on the basis of the color cue alone. It forces the learner to sit up and analyze the words in terms of what is different rather than just looking for similarities. By printing the letter or letters that are different in a bright red color, attention can be focused on the distinctive features of the words while allowing the printed form of Ithe word to remain the same regardless of the color. Vanishing the color gradually in subsequent presenta- tions, combined with verbal directions, increases the effectiveness of the cue by forcing the subject to attend more to the distinctive feature of the stimulus and less to the color cue. The results of the study revealed a significant difference in achievement favoring the vanished color treatment over both the no color treatment and the maximum color treatment. These data indicate that the effectiveness of visual discrimination training depends upon the technique employed and the information pre— sented. Even though it is not unusual for young chil— dren to experience difficulties with discrimination tasks involving reversals or letter rotations, it is felt that 91 the use of vanishing hue color cues should assist these children in mastering the discrimination tasks more efficiently, rather than wait for some mysterious, undetermined period Of maturation. This is not to deny the importance of maturation; however, it would appear that if instruction is to be provided then it should be on tasks that have relevance to the reading process and can be transferred without confusion to the child. Thus, the use of vanishing color cues to focus atten- tion on distinctive features of visual stimuli should help eliminate those perceptual confusions that are not of organic etiology. Finally, from the analysis of the data of this study it would appear that children who are experiencing either kinetic or static reversals will both respond to the treatment of vanishing hue color cues . Recommendations The recommendations gleaned from the study include the following. Further study is needed regarding the applica- tion of vanished color cues to other instructional tasks and methods of adaptation of a technique for use in published materials for primary grades. The effect of vanished color cues needs to be investigated experimen- tally under the following conditions: 92 l. Replication Of the present study under the same conditions. 2. Present the three conditions, maximum color, vanishing color, and no color; for each treatment add another criterion trial and compare the treatments under these conditions. 3. Test the effect on other words that are con- fused using the vanishing color technique presented here with Older children who are making reversals. 4. Publishers should try the technique with beginning reading material. 5. Research should be undertaken in a school setting using materials that employ the vanishing hue color technique for one group and materials with no color for one group for a specified period of time. Comparison of the achievement of the two groups would provide information concerning the relative potency of the vanished color cues. Teachers Of reading should use vanished hue color cues to assist children with the following tasks: 1. Learning to visually discriminate between words that are confusing. 2. Learning to visually discriminate between words that have all or many common letters. 3. Learning words that contain letters that are rotations of each other. 93 4. Learning directionality, e.g., where to look first in reading words. 5. Using the procedure with kindregarten and first grade children to prevent confusions. In summary, this study has demonstrated that vanished color cues can enhance the learning achievement of children who are making reversals that are either kinetic or static reversals. However, more research is needed utilizing this technique. That attention can be appropriately focused with the use of vanished color cues has been demonstrated by this research. APPENDI CBS 94 APPENDIX A INTRODUCTION TO AND VERBAL DIRECTIONS FOR THE INSTRUCTIONAL EPISODE 95 APPENDIX A INTRODUCTION TO AND VERBAL DIRECTIONS FOR THE INSTRUCTIONAL EPISODE Introduction to Instructional Episode Instructions to subjects: "Good morning. My name is Mrs. Sims, and you're How do you like school? Do you have any sisters? Brothers? What subjects do you like best? Really! Well, do you like to read? The reason why I asked is that I have some words that I would like for you to read for me. It isn't going to be hard, because I am going to tell you what they are. All you have to do is listen closely to what I say and look carefully at what I show you and you'll be able to read the words." 96 97 Verbal Directions for Instructional Episode A pointer was used in emphasizing differences and showing the Child how to discriminate between the words. It was also used to show the direction from which the child should begin to read the words. Want—Went Investigator: This word says "want"; say want. This word says "went”; say went. The differ- ence between want and went is that want has an "a" after the "w" and went has an "e" after the "w." Now you tell me the difference. Subject: This one has "nt" at the end and "w" at the beginning, and this one has "nt" right there and "w" right here (Child points to letters). Investigator: Yes, what you said was right but you told me how they were alike. Now look again and tell me how they are different. Subject: This one has an "a" and this one has an He." Investigator: What does the one with the "a" say? Subject: I forgot. Investigator: What does this one say? Subject Want. Investigator: NO, the one with the "a" says want; say want. The one with the "e" says went; say went. Note: Subject is taken through directions again on second trial if he does not know the words. Investigator: What does this word say? Subject: Answer. Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: Investigator: Subject: 98 All right, what does this word say? Answer. What does this word say? Answer. My, but you are really paying attention, aren't you? Now, what does this word say? Answer. What does this word say? Answer. What does this word say? Answer. OK, what does this word say? Answer. My, but you are doing a good job! What does this word say? Answer. What does this word say? Answer. What does this word say? Answer. Book Two: no-on All right, let's look at this one. This word says "on”; say on. This word says "no"; say no. The difference between no and on is that no begins with an "n" and on begins with an "o.” Now you tell me the difference. Answers. 99 The same procedure for Book One is followed for the remainder of the instructional episode. APPENDIX B WORDS USED AND METHODS OF HIGHLIGHTING FOR EACH TREATMENT IN THE INSTRUCTIONAL EPISODE 100 APPENDIX B WORDS USED AND METHODS OF HIGHLIGHTING FOR EACH TREATMENT IN THE Maximum Color Highlighting INSTRUCTIONAL EPISODE l 2 3 4 Criterion' want want want want want no no no no no they they they they they sit sit sit sit sit every every every every every always always always always always many many many many many saw saw saw saw saw those those those those those boy boy boy boy boy where where where where where been been been been been lOl 102 No Color Easily Confused and Look-Alike Words l 2 3 4 Criterion went went went went went on on on on on that that that that that its its its its its very very very very very away away away away away may may may may may was was was was was these these these these these dog dog dog dog dog were were were were were be be be be be 103 NO Color Highlighting l 2 3 4 Criterion want want want want want no no no no no they they they they they sit sit sit sit sit every every every every every always always always always always many many many many many saw saw saw saw saw those those those those those boy boy boy boy boy where where where where where been been been been been Vanished Color Highlighting 104 l 2 3 4 Criterion Want want want want want no no no no no they they they they they sit sit sit sit sit every every every every every always always always always always many many many many many saw saw saw saw saw those those those those those boy boy boy boy boy where where where where where been been been been been APPENDIX C DIRECTIONS FOR POSTTESTS AND POSTTEST SHEETS 105 APPENDIX C DIRECTIONS FOR POSTTESTS Naming Task--24 words Instructions: The subject studied 24 words contained in 12 booklets. The words were arranged in pairs. After completion of the Instructional Episode the student was given a score of one for each word that he named cor- rectly on the criterion trial. Pairs Task—~12 pairs of words Instructions: The subject studied 12 pairs of words contained in booklets. Upon completion of the Instructional Episode, the student was given a score of one for each pair of words named correctly. No score was given if either of the two words was missed. Match to Form From Memogy Instructions: "I am going to place an answer sheet on your desk with some words written on it. The words will be covered with a sheet of tag board paper. Then, I am going to show you a word on a flash card for about one-half second, 106 107 like this. What did you see? Nothing? Yes, I know, because that card was blank. The next card is going to have a word on it. As soon as you see the word I want you to move the cover sheet down until you see all the words on the first line. Look very carefully at each word and find the one which looks exactly like the one that you saw on the card. Draw a Circle around the word that you think is the correct one. Be sure to look at all of the words before you make a selection. A score of one will be given for each scoring word circled." 108 NAME SEX GROUP Yes NO 1. want 2. went 3. every 4. very 5. boy 6. dog 7. always 8. away 9. where 10 . were 11. saw 12. was 13. those 14. these 15. sit 16. its 17. been 18. be 19. many 20. may 21. they 22. that 23. no 24. on 109 Name Sex Group wont went were want* on one no* only very year every* ever 80d b0y* dog box always* way away among were here where* ever as saw* sat was those* the they these its set sit* isnt be bet bee been* many* any . may man that they* hat them *Refers to words on flash cards; students were to circle on this sheet the word that matched the one shown on the flash card. B I BLIOGRAPHY 110 BIBLIOGRAPHY Allington, Richard L. ”An Evaluation of the Use of Color Cues to Focus Attention in Discrimination and Paired—Associate Learning." Ph.D. dissertation, Michigan State University, 1973. et a1. "Poor and Normal Readers' Achievement on Visual Tasks Involving High Frequency, Low Dis- criminability Words." Journal of Learning Dis- abilities 9 (May 1976). Ash, Michael et al. The Relation Between the Discrimina- tion of Letter-Like Forms and Word Recognition. 1975. ED 105380. Barges, Williams C. "An Experimental Approach to Alphasic and to Non-Reading Children." American Journal of Orthopsychiatny 23 (1953): 158. Birbaum, I. M. "Context Stimuli in Verbal Learning and the Persistence of Associative Factors." Journal of Experimental Psychology 71 (1966): 483-87. Blau, Abram. The Master Hand Monograph Series, No. 5. New York: American Orthopsychiatric Association, 1945. Bond, G. L., and Dykstra, R. "The Cooperative Research Program in First Grade Reading Instruction." Reading Research Quarterly 24 (1967): 5—142. , and Tinker, Miles A. Reading Difficulties; Their Diagnosis and Correction. New York: Appleton-Century-Crofts, 1973. Carrol, John B. New Nature of the Reading Process; Theoretical Models and Processes of Reading. Edited by Harry Singer and Robert Ruddell. Newark, Del.: The International Reading Associa- tion, 1970. Combs, ArthurVL, and Syngg, Donald. Individual Behavior. New York: Harper and Row, 1959. 111 112 Critchley, M. Developmental Dylexia. London: Heinemann, 1964. Dechant, Emerald V. Improvinggthe Teachingof Reading. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1970. De Hirsh, K.; Jansky, J. J.; and Langford, W. S. Predict- ing Readinngailure. New York: Harper and Row, 1966. Duffy, Gerald G., and Sherman, George B. Systematic Reading Instruction. New York: Harper and Row, Publishers, 1973. Dulsky, S. G. "The Effect of a Change of Background on Recall and Relearning.” Journal of Experimental Psychology 18 (1935): 725-40. Dykstra, R. "The Use of Reading Readiness Tests for Prediction and Diagnosis: A Critique." Unpub- lished manuscript, University of Minnesota, 1966. Frostig, M. The Frostig Program for tne Development of Visual Perception. Chicago: Follet, 1964. Gates, Arthur I., and Bennett, Chester C. Reversal Tendencies in Reading. New York: Bureau of Publications, Teachers College, Columbia Univer- sity, 1933. Geyer, J. J. "Models of Perceptual Processes in Reading." In Theoretical Models and Processes of Reading. Edited by H. Singer and R. B. Ruddell. Newark, Del.: International Reading Association, 1970. Gibson, E. J. "The Ontogeny of Reading." American Psychchologist 25 (1970): 136-43. ; Gibson, J. J.; Pick, A. D.; and Osser, II. "A Developmental Study of the Discrimination of Letterlike Forms." Journal;of Comparative Physiological Psychology 55 (1962): 897-906. Gibson, E. J.; Schapiro, F.; and Yonas, A. "Confusion Matrices for Graphic Patterns Obtained With Latency Measure." In The Analysis of Reading Skill. A Program Of Basic and Applied Research. Final Report Project No. 5-1213. Cornell Uni- versity and U.S. Office of Education, 1968. 113 Goodman, Kenneth 8., and Burke, Carolyn L. Theoretically Based Studies of Pattggns of Miscues in Oral Reading Performance. Washington, D.C.: Office of Education, Bureau of Research, 1973. ED 079708. Hammill, D. "Training Visual Perceptual Processes." Journal Of LearningyDisabilities 5 (1972): 552—59. Harris, Albert J., and Sipay, Edward R. How :0 Increase Reading Ability. New York: David’McKay Company, Inc., 1976. Hildreth, Gertrude H. "Reversals in Reading and Writing." Psychological Bulletin 30,9 (1933): 670-71. Hill, F. A., and Wickens, D. D. "The Effect of Stimulus Compounding in Paired-Associate Learning." Journal of Verbal Learning and Verbal Behavior 1 (1962): 144-45. Jansen, David Rainsford. "Effects of Visual and Auditory Perceptual Aptitudes and Letter Discrimination Pertaining to Word Recognition." Ph.D. disserta- tion, The Pennsylvania State University, 1972. Johnson, R. L. "The Effect of Training in Letter Names on Success in Beginning Reading for Children of Differing Abilities." Paper presented at the Meeting of the American Educational Research Association, Minneapolis, Minnesota, February 1970. Jones, J. K. "Colour as an Aid to Visual Perception in Early Reading." British Journal of Educational Psychology 35 (1965): 21-27. Kuenne, Janet B., and Williams, Johnna B. "Cues Asso- ciated With Recognition of Aural Stimuli in the Primary Grades." Paper presented at the Annual Meeting of the American Educational Research Association, Chicago, Illinois, April 1972. ED 065187. Lahey, B. B.; McNees, M. P.; and Brown, C. C. "Modifi- cations of Deficits in Reading for Comprehension." Journal of Applied Behavior Analysis 6 (1973): 475-80. Larson, S. C., and Hammill, D. D. "The Relationship of Selected Visual Perceptual Abilities to School Learning." Journal of Special Education, in press. 114 Liberman, Isabella Y. et a1. Letter Confusions and ReverSals of Sequence in Ens Beginning Reader: Implications for Orton's Theory of Developmental Dyslexia. 1971. ED 096605. Lott, D.; Smith, F.; and Cronnell, B. Functional Equiva- lence of Feature Combinations in Visual Identi- fication of Words. Inglewood, Calif.: Southwest Regional Laboratory, 1968. Marzano, Robert J., and Distefano, Philip. Elements of the Word Recognition Process: A Two-Part Study. 1975. ED 106793. Nevins, Rosemary J. "The Effect of Training in Letter Names, Letter Sounds on the Acquisition of Word Recognition Ability. Ph.D. dissertation, State University of New York at Albany, 1973. ED 084504. Orton, Samuel T. ReadingLerEing, and Speech Problems in Children. New York: W. W. Norton and Company, 1937. Otto, Wayne, and Cooper, Calvin. "Investigation of the Role of Selected Cues in Children's Paired- Associate Learning." Technical Report No. 53. Wisconsin Research and Development Center for Cognitive Learning, University of Wisconsin, May 1968. ED 036315. Pan, S. ”Influence of Context Upon Learning and Recall." Journal of Experimental Psychology 9 (1926): 468—91. Park, George. "MirrOr and Reversed Reading." Journal of Pediatrics 42 (1953): 120-28. Piaget, Jean. ”Development and Learning.” Conference on Cognitive Studies and Curriculum Development. Ithaca, New York: School of Education, Cornell University, March 1964. Reinstein, Steven S. "The Effects of Contextual Predict- ability on Oral Reading Reversals.” Dissertation Abstracts International, 1975, p. 7307-A. Rosen, H.; Richardson, D. H.; and Saltz, E. ”Meaningful- ness as a Differentiation Variable in the von Restorff Effect." Journal of Experimental Psy— chology 64 (1962): 327-28. 115 Saltz, Eli. "Compound Stimuli in Verbal Learning; Cognitive and Sensory Differentiation Versus Stimulus Selection." Journal of Experimental Psychology 66 (1963): 1-5. , and Newman, Slater E. "The von Restorff Iso- lation Effect: Test of the Intralist Association Assumption." Journal of Experimental Psychology 58 (1959): 445-51. Samuels, S. J. ”Formal Intralist Similarity and the von Restorff Effect." Journal of Educational Psychology 59 (1968): 432-37. "Letter Names Versus Letter Sound Knowledge as Factors Influencing Learning to Read." Paper presented at the Meeting of the American Educational Research Association, New York, February 1970. . "Modes of Word Recognition." In Theoretical Models and Processes of Reading. Edited by Harry Singer and Robert Ruddell. Newark, Del.: Inter- national Reading Association, 1970. ”Word Recognition and Beginning Reading." The Reading Teacher 23,2 (1969): 159-61. Schonell, F. J. Backwardness in Basic Subjects. Edinburg: Oliver and Boyd, 1948. Singer, Harry, and Ruddell, Robert B. Theoretical Models and Processes of Reading. Newark, Del.: Inter- national Reading Association, 1976. Spache, George D., and Spache, Evelyn B. Reading in the Elementary School. Boston: Allyn and Bacon, Inc., 1974. Sunderland, D. M., and Wickens, D. D. "Content Factors in Paired-Associate Learning and Recall." Journal of Experimental Psychology 63 (1962): 302-306. Travers, Jeffery R. Formal Models of Word Recognition. National Institute of Education (DHEW). Washington, D.C.: Office of Research Grants, 1975. ED 108130. Underwood, B. J.; Ham, Margaret; and Ekstrand, B. "Cue Selection in Paired-Associate Learning.” Journal of Experimental Psycholqu 64 (1962): 405-409. 116 Vellutino, F. B. An Age Difference and the Perceptual Deficit Hypothesis. Research Report 815. 1974. ; Steger, J. A.; and Kandel, G. "Reading Dis- ability: An Investigation of the Perceptual Deficit Hypothesis." Cortex 8 (1972): 108-18. Venesky, Richard L. Letter Naming and Learning to Read. Washington, D.C.: ResearChfand Development Center for Cognitive Learning, Office of Educa- tion (DHEW), 1971. von Restorff, H. "Uber die Wirkung Bererchsbildungen im Sparenfeld." In Analyses von vorgangen im Spurenfeld. Edited by W. Kohler and H. von Restorff. Psychol. Forch. 1 (1933): 18, 299-342. Weiss, Walter, and Margolius, Garry. "The Effect of Context Stimuli on Learning and Retention." Journal of Experimental Psychology 48,5 (1954): 318-22. Williams, Johnna P. "Effects of Discrimination and Reproduction Training on Ability to Discriminate Letter-Like Forms.” University of Pennsylvania, 1968. ED 017418. "Some Experiments on Visual and Aural Word Recognition.” Twentieth Yearbook of the National Reading Conference, Inc. Marquette. "IIIIIIIIIIIIIIIIIIIIIIIS