fit 'J‘i MIChigch L. --«Qrb@ Universxty r mrl’ .-" ' a 1293 10171 7738 .H E‘ ,W This is to certify that the thesis entitled Functional Hand Asymmetry and Braille Reading in Blind Readers presented by Jean Wilkinson has been accepted towards fulfillment of the requirements for Masters degree in Psychology 1%jor professor Date 71/7) y/7£> 0—7 639 mania: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records FUNCTIONAL HAND ASYMMETRY AND BRAILLE READING IN BLIND READERS BY Jean M. Wilkinson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1978 ABSTRACT FUNCTIONAL HAND ASYMMETRY AND BRAILLE READING IN BLIND READERS BY Jean M. Wilkinson Some research suggests that braille, because of its configurational- spatial design, will be more efficiently processed by the right (non— language) than left (language) cerebral hemisphere in right-handed indi— viduals. Therefore, left (non—dominant) hand performances would be expected to be better. Alternatively, any reading task should be better processed by the left cerebral hemisphere suggesting better right-hand performance. Twenty—three right—handed blind students (10—20 years old) were given letter identification and paragraph reading tasks. Only fast-reading females showed a left—hand advantage for reading paragraphs, but no hand or sex differences were found for Single letter identification. For bOth males and females who initially preferred their left hand for read- ing, left hand superiority was found for both tasks, though the reverse was not found for subjects who initially preferred their right hand. ‘ The results support the idea that braille reading has an important 1 configurational—spatial component, but hand differences also are related to an initial reading skill, sex of subject, and initial hand preference for the task. — The very time I thought I was lost, my dungeon shook and my chains fell off. James Baldwin ii ACKNOWLEDGMENTS I would like to thank my chairman, Dr. Lauren Harris for his sup- portive yet critical review of my research ideas as they have developed over the years, and for the encouragement to apply my theoretical formu- lations to developmental psychology in the area of hemispheric specializa— tion. He was an astute and patient teacher who fostered my research skills throughout my ordeal. I am particularly grateful to my committee members, Drs. Ellen Strommen and Joseph Reyher for their though-provoking and constructive suggestions for improvements in the thesis. I wish to express my deep appreciation to Paulette Hatchett, whose thinking enabled me to improve and clarify several sections of the thesis. The love and friendship she shared, sustained me throughout the frustra- tions and joys of the thesis. I here express my admiration as well as my thanks to her, for she is indeed an extraordinary and sensitive person. Three other individuals- Mr. Lou Tutt, Mr. Francis Hetherington, and Dr. Nancy Bryant— encouraged and facilitated my efforts to do research at the Michigan School for the Blind. A special thanks must go to Mr. Jim Borough, who helped me with the technical aspects of understanding the various braille materials and reading styles, and the readers who participated in the study. I would especially like to thank the Michigan School for the Blind students and their parents for their enthusiasm and interest in my research. I would also like to thank my mother and sister for the support and iii patience they have given me. Finally, I would like to thank my undergraduate assistants and my friends who were helpful to me at various stages of my work. iv Chapter I. II. III. IV. TABLE OF CONTENTS INTRODUCTION REVIEW OF RELEVANT LITERATURE . Asymmetrical Processing Differences in the Human Brain Left-Right Hand Differences in Braille Reading . The Effect of Phonological Cues on Tactual Perception and Teaching Methods . . . Age and Sex Differences in the Development of Hemispheric Specializations and Cognitive Skills . Hemispheric Specialization and Inter—hemispheric Competition for Control of the Motor Pathways Contributions of the Current Study . . . Experimental Hypotheses and Tests . Rational for Using the Middle Finger METHOD Subjects Materials . . Research Design . Procedure . RESULTS . Test Trials for Letter Identification and Paragraph Reading . . Verbally— —stated Hand Preference versus the Actual Hand Used Hand Preference . Individual Differences Preference Contrasts Type of Blindness . The Left Handed Subjects DISCUSSION Interpretations and Implications Summary of Major Findings . Toward a Theory of Functional Hand Asymmetry in Braille Reading Modes of the Acquisition Process V Page AC»! l4 17 21 26 27 29 31 31 31 34 35 38 38 42 45 48 51 54 55 57 57 57 69 72 Hand Preference and Dual Coding in Braille Reading APPENDICES . REFERENCES . vi Page 76 84 . 117 Table Gl. LIST OF TABLES Results of t-tests between the silent and aloud conditions for the right and left hand in the paragraph reading task . Results of t-tests between right and left hand performance in the silent and aloud conditions for the paragraph reading task as a function of sex and rate of reading . Results of t-tests for mean times and mean errors during the letter identification task by behavioral hand performance Results of t-tests between the mean difference scores in seconds for the left and right hand paragraph reading conditions as a function of hand preference . . . . . . . . . Results of t-tests between right and left hand performance in the paragraph reading task as a function of verbalization, hand preference and rate of reading Mean reading times and percent difference for 12 paragraphs combining 'aloud' and 'silent' reading conditions as a function of behavioral hand preference in the single letter identifi- cation task Results of the analyses of variance for total reading times by congenital and adventiously blind males and females in the single letter condition Results of the analyses of variance for number of errors during the letter identification task by sex of subject and hand . Results of t-tests for mean times and errors during the letter identification task by sex of subject and hand vii Page 40 41 45 47 48 50 SS 104 Table 62. H1. 11. J1. K1. K2. L1. M1. Results of t-tests between the mean difference scores in seconds for the left and right hand paragraph reading conditions by sex . Results of t-tests between right and left hand performance in the silent and aloud condition of the paragraph reading task Results of t-tests between the mean difference in seconds for silent and aloud conditions for the right and left hand performance on the paragraph reading task as a function of sex and rate of reading Results of t-tests between silent and aloud conditions in the paragraph reading task as a function of verbalization, hand preference and rate of reading . Results of t-tests for effect of hand used and verbalization and hand for the subjects who preferred each hand equally Results of t—tests for effect of verbalization and hand for the subjects who preferred each hand equally . Mean reading times and percent differences for 3-columns of braille letters in the aloud reading condition as a function of behavioral hand preference in the single letter identi- fication task Results of the analyses of variance for the paragraph reading difference scores of congenital and adventitious blind males and females viii Page 104 106 108 110 112 112 114 116 LIST OF FIGURES Figure Page 1. Comparison of stated hand preference and the actual hand used for braille reading by males and females . . . . . . . . . . . . . 44 2. Mean difference scores in seconds for the paragraph reading task for the verbali- zation conditions as a function of hand preference . . . . . . . . . . . . . . . . . . . . . . 53 ix Appendix A. B. LIST OF APPENDICES Handedness Questionnaire . . . . . History of Blindness Questionnaire Instructions for Experiment I . Experiment I Code Sheet . Instructions for Experiment II Experiment II Code Sheet T-tests for Letter Identification and Paragraph Reading . Right and Left Hand Differences for Verbalization Trials on Paragraph Reading . . . . . . Mean Difference Scores for Paragraph Reading as a Function of Sex and Rate of Reading Mean Difference Scores for Paragraph Reading as a Function of Verbalization, Hand Preference, and Rate of Reading . . . T—tests for Subjects Preferring Hand Equally Individual Differences Experiment I . ANOVAs for Type of Blindness and Sex Page 85 87 90 93 95 102 104 106 108 110 112 114 116 I. INTRODUCTION Braille reading, like visual reading, involves the reception and interpretation of graphic symbols. By contrast, the most obvious differ- ences between the two forms of reading are the type of symbolizations and the organic system used. Braille is perceived through the fingers and print is perceived through the eyes. These differences mean braille must be read at a much slower rate than print. Despite this, the reading of braille involves many of the same factors necessary for successful visual reading such as perceiving, understanding, conceptualizing, and experiencing the graphic forms (Rex, 1970). The study of the process of learning to read braille is of particu— lar interest because reading braille requires the accurate perception of spatial patterns having linguistic labels. Since tactual spatial skills involving the mental rotation and transformation of objects perceived in space are found to be highly correlated with right hemisphere func— tioning and the development of linguistic skills is related to left hemisphere functioning, the question is, how is braille learned? For instance, one technique for reading braille is for the left and right forefingers to start moving across the axis of a punctographic line of cells to the middle of the page. Then the right finger sweeps across the rest of the line while the left forefinger helps to identify unusual collections of symbols or goes to the next line to establish a continuity in the reading material. Since this is only one style or technique used 1 2 to identify the symbols, it is essential to note that readers must explore, touch, and move their fingers up and down and around the cell or adjacent cells to discriminate the arrangement of dots until they become clear. For these reasons, braille has been used as one vehicle for under- standing the complexity of multi-dimensional tactual stimuli, that is, understanding the influence of pattern, number and direction of a stimulus as it relates to the functioning of the two cerebral hemispheres. Consequently, it was the intent of this study (1) to identify the hemi- spheric specializations which may facilitate the cognitive understanding of a complex informational system, such as the Braille Code, and (2) to determine whether any of the observed hand differences in braille reading can be attributed to functional cerebral asymmetry. II. REVIEW OF RELEVANT LITERATURE Asymmetrical Processing Differences in the Human Brain This chapter begins with a brief outline of recent research on hemi— spheric asymmetries so as to provide a framework from which to view the braille literature on hand asymmetries. The braille literature is then looked at in terms of the various stages or strategies in which braille is learned and the cognitive processes involved in these strategies. It is widely accepted that the two cerebral hemispheres have dif— ferent functions. The left hemisphere, which controls the right side of the body, appears for the most part to control speech and language mechanisms. The right hemisphere controls the left side and primarily subserves spatial functions. These characteristics have been inferred from clinical studies of deficits produced from damage to the right and left hemispheres (e.g., McFie, Percy, and Zangwill, 1950; Semmes, Weinstein, Ghent, and Teuber, 1955; DeRenzi and Spinnler, 1966); and split brain patients, whose hemispheres are prevented from transferring information across the corpus callosum (this literature is reviewed by Nebes, 1974). Additional support for hemisphere asymmetries has come from studies with normal subjects on dichotic listening tasks (Kimura, 1961) and tachistoscopic recognition tasks (Kimura, 1966; Kimura, 1969; White, 1972). In terms of the auditory and visual modalities, the superiority 4 of the right or left hemispheres for processing is inferred when, for instance, a stimulus presentation to the right visual half-field or ear is recalled or recognized faster or more accurately than the left. As a consequence, it has been suggested that the contralateral connections are better at relaying information between the two hemispheres and they also inhibit the ipsilateral connections (Kinsbourne, 1973). In this connection, Cohen (1973) has shown that the time necessary to mediate verbal material requires a more serial processing procedure and because of this, it is better processed by the left hemisphere. Con- versely, by implication the right hemisphere is better able to process nonverbal or spatial material in a more parallel manner (Kimura, 1973; White, 1969; Smith and Nielson, 1970). This dichotomy has also been characterized as requiring "propositional” as opposed to "appositional" thinking (Bogan, 1972). The former term refers to logical-analytic pro- cessing of the left hemisphere and the latter to the gestalt-synthetic approach of the right hemisphere. In summary, these findings have led to characterizing the cognitive processing of the left and right hemisphere as serial or temporal v. parallel processing; propositional v. apposi- tional processing; and verbal v. nonverbal processing. Left—Right Hand Differences in Braille Reading Functional hand differences in blind braille readers. Several studies have attempted to demonstrate hand superiorities in braille reading by examining hand differences when using letters, sentences, and varied directions of reading (Villey, 1931; Smith, 1929; Fertsch, 1947; Foulke, 1964; Hermelin and O'Connor, 1971a, 1971b). Observations have 5 suggested that the left hand is the preferred hand for braille reading implying that, it is also superior for speed and accuracy (Critchley, 1953; Grasemann, cited in Villey, 1931; Bfirklen, 1917, cited in Lowenfeld, Abel, and Halten, 1969). Most of this evidence is inconclusive, since the procedures rarely controlled for reading style, sex or grade differ- ences, but they also suggest that while there may be hand superiority for reading, there also may be a demonstrated hand preference for the same or Opposite hand suggestive of functional hand differences (Villey, 1931; Smith, 1929). For example, Critchley (1953) observed that the practiced braillist is able to read with both index fingers effectively, even though many readers eventually chose a master finger. This distinction suggestive of a learned manual preference is not altogether obvious, since each hand may be accurate in its decoding of the braille symbols and the two hands appear to discriminate the dot configurations simultaneously. Critchley suggested, however, that for a majority of braille readers the left index finger is the finger of choice even though the bi—manual reading may involve two distinct identification processes. Therefore, the finger chosen to be the "master finger" must give the reader addi- tional information or insight into the nuances of the code that are not perceived by the other finger or fingers involved in the reading process. Specifically, Critchely speculated that a bi—manual technique of reading involving swift small—range searching movements facilitated a relationship such that one finger (right) is devoted to the ”service of recognition" and the other (left) the ”service of control" (p. 22). This process would enable the braillist to develop a sense for the constituent dots and to identify the collection of symbols as a whole-tactual gestalt. 6 The recognition process is facilitated by the fluctuation in the stimuli which intensifies each configuration and adds to the degree of sensory perception (Critchley, 1953, p. 24). In other studies, the finger relationship leading to left hand pre- ference of superiority in reading is not so clear. For instance, Foulke (1964) had experienced adult braille readers use their eight fingers (excluding the thumbs), alternating preferred and nonpreferred fingers, to determine the reading rate for randomized letters in a horizontal direction. The lines consisted of five, equally spaced, five letter groups. Although some readers were better with one hand than the other, this as a whole showed no consistent relationship between reading speed and hand used. These data suggest that (a) additional practice received by the preferred hand and (b) that there is a substantial loss in reading ability from the index finger to the little finger. The latter finding may be related to the differences in the supply on sensory nerve endings of the fingertips or the cortical representations for each fingertip. Using blind subjects, Fertsch (1947) compared good and poor readers to determine which hand was more efficient. The subjects were students (30 girls and 33 boys) in grades 3 - ll attending a school for the blind. Fertsch did report the number of boys and girls in each grade. The sub— jects were divided into those who scored in the upper and those who scored in the lower third in comprehension on a standardized silent reading test. Each subject read a paragraph from the grade one and a half level of the Shank Test of Reading Comprehension. The paragraph was transcribed into braille and then divided into two parts. Each part was equal in length, number of lines (6), number of cell units (175), and number of 7 signs (32). First, the subjects silently read part one of the paragraph with their right hands alone. Then they read part two with their left hands alone. Comprehension was tested at the end of each half by asking three questions about the text of each part. The questions were prepared for the Shank Test. Handedness was assessed by grip strength and ball throwing. Sixty of the subjects were designated as right handed and three as left handed. The blind readers were classified then into three groups (1) right dominant, right hand more effective than the left hand; (2) left dominant, left hand more effective than the right hand; and (3) hands equal, both hands equally effective. The best readers were those who read equally well with either hand and used both together. Fertsch also found that the speed scores for the right hand dominant readers were higher than those for the left hand dominant group. The poor readers characteristic— ally kept right and left hands close together. Consequently, they read less material with each hand independently. It should be noted that Fertsch did not consider errors per cell unit, sex or grade differences, or interaction effects. Thus, the evidence supporting right hand superiority is somewhat weak. However, the results imply that hand pre— ference and superiority in reading are related to accuracy and speed in reading ability. Two experiments by Hermelin and O'Connor (1971a, 1971b) with blind children and adults specifically addressed the issues of handedness, differences in braille materials, and the functional asymmetry of the brain with respect to braille reading. In the first experiment, Hermelin and O'Connor (1971a) tested 16 children, all blind from birth, between the ages of 8 and 10. Handedness was determined by a seven-item test. Fourteen children were right handed and two were ambidextrous. Twelve children read spontaneously with their index fingers close together. Four read with the index finger of their left hand. Each child's index and middle fingers were tested separately to determine "finger” reading speed. Individual symbols were scored, instead of words or letters, because many abbreviations and contractions were used. The left index finger reading times proved to be significant- ly faster than the right index finger reading times. Similarly, the left middle finger reading times were significantly faster than the reading times for right middle finger. Hermelin and O'Connor (1971a) attributed the differences between the left and right hand in reading speeds partly to the left to right direction during braille reading; partly to the left to right direction during braille reading; partly to the greater convenience of motoric scanning, and partly to the fact that many of the children read primarily with their left hands. To control for the possibility of a motoric scan bias, Hermelin and O'Connor (1971b) arranged a random set of alphabetical letters into vertical columns. The subjects were 15 totally blind adults, age 25 to 65. Although speed and accuracy were both stressed, speed did not prove to be a significant factor and did not, therefore, differentiate between the hands. However, 11 of the 15 subjects made significantly fewer errors with their left hand even though the right hand had read more letters (based on the combination of correct and incorrect letters scored). The authors suggest that two factors contributed to finding different reading speeds for the group of adults and the group of children. First, many of the adults were self-taught, no longer valued speed of reading at the expense of accuracy, and switched to the left hand because 9 it seemed more efficient. Second, the difference in the material, letter v. sentences, made it difficult to compare the two groups with regard to reading speed, which suggested that the type of material being read might affect reading speed. Hermelin and O'Connor interpret their data as indicative of the cortical asymmetry for verbal and spatial-tactile stimuli and argued that the brain operates to interpret braille as spatial—tactual items first, to be subsequently analyzed more accurately by the right hemisphere. If this hypothesis is correct, it might be expected that hemispheric processing would be affected by (l) the varying amounts of practice and experience reflecting different stages in the acquisition of braille reading skills (e.g., Fertsch, 1947); (2) the methods of teaching braille; and (3) the means of coding the different braille characters (e.g., tac- tual and/or phonological). The present study was designed to consider the above mentioned fac— tors for a braille reading task and to determine at what stages, if any, left hand superiority in braille reading would occur. Stages in acquiring braille skills. As was alluded to earlier, there are many techniques for reading as well as for acquiring efficient, braille reading skills. From the data presented thus far, several fac- tors could affect braille reading: (1) the amount of training and ex- perience often interpreted as practice (Foulke, 1964; Hammill and Crandell, 1969); (2) the serial perception and complexity of the task (Flanagen, 1964; Weiner, 1963); and (3) the tactual and phonological similarity of braille characters (Ashcroft, 1960; Nolan and Kederis, 1969; Harley and Rawls, 1970; Millar, 1975a). And although the intelligence of the blind person is cited as a prerequisite for minimal tactual perceptive ability 10 above the minimum needed for learning to read print. These are important concerns which will be considered shortly. However, what seems to be essential for determining how the complex braille configurations are read is further understanding of the development or stages involved in the attainment of competent braille reading skills. Kusajima (1974), for instance, kymographically recorded individual finger movements of blind children reading braille. The progressive variations in the relationship between the two index fingers were noted when the braille reader moved from beginner to intermediate and inter— mediate to advanced reading stages of competence. The study revealed that the beginning readers who were unfamiliar with the dot configurations tended to read letter by letter, predominately with the right hand. They characteristically made numerous up and down, zig-zag movements, with many fixations on the ”spatial” forms of the words. At this stage the letters must have appeared to be arranged randomly until the distinct spatial arrangements could be identified. The readers in the intermediate stage read with their fingers light— ly pressed together. Their reading fingers moved together from the be— ginning to the end of the line. Eventually, however, some of the inter- mediate readers separated their fingers so that sharing or exchanging of roles occurred. In other words, even though these readers may have started reading the lines with their fingers pressed together, they eventially separated them towards the end of the line so that the reading material appeared to be divided into two parts. This stage suggests that the introduction of the left hand improved the processing of the spatial forms which lead to greater hand independence and an implied improvement of reading skills. 11 The advanced readers were competent with both left and right fingers and usually read half of each line with the left and right index fingers, respectively. Their finger movements were smooth, symmetric, and easy. The results from Kusajima's study clearly show that changes in the hand(s) used seem to reflect patterns or strategies for decoding the con— figurations based on practice or experience. That is, increasing the level of practice should predict independent fluid finger movements and increases in good reading skills, whereas less practice and experience on a complex sequential tactual task should predict more dependent fixated finger movements and poor reading skills. Furthermore, the shifts in reading modes mark the initiation of changes in the reader's ability to perceptually discriminate or cognitively process increasing amounts of information simultaneously and successively. Consequently, it is suggested that these developmental changes in hand preference for reading (e.g., the beginners with an apparent right hand preference for looking for spatial cues change to intermediate readers using both hands) are indica- tive of the congitive strategies used to identify the braille dots (parallel-nonverbal v. serial—verbal). And it is also suggested that the level of practice or experience of the reader (beginner v. advanced skills) will determine the hemispheric specialization and thus the hand that is superior for reading. Several studies of hand asymmetries on complex tactual tasks (Gardner, 1942; Provins and Glencross, 1968) and simple successive finger movements (Denckla, 1973; Ingram, 1975; Wolff and Hurwitz, 1976) substan- tiates these inferences, in part, and provides a dichotomy from which to compare the tactual discrimination strategies and skills found in braille reading. 12 Reading stages compared with similar adjustments in normal sighted adults and children. In the following studies it should be noted that the hand asymmetries are divided along dimensions of practice (training) and type of finger movement (non-serial or serial v. parallel). To be clear, serial means successive ordering of movements and non-serial or parallel means that the order of movements are necessarily successive. In this sense, Provins and Glencross (1968), studying trained and untrained typists found that the discrimination of letters that formed words, random letters, and the letters of the home keys (those in the middle row where the fingers rest) yielded left—right asymmetries similar to what was demonstrated by Kusajima's braille readers. That is, the untrained or less practiced and least skilled typists showed a right hand superiority for random letters and home keys (assumed to be less familiar to the typist) and no hand difference for the words. The trained or more prac— ticed and skilled typists demonstrated a left hand superiority for the words and home keys (assumed to be more familiar to the typists). Similarly, studies by Gardner (1942) on hand sorting (right hand superiority) and tracing (left hand superiority) tasks with adults; and Ingram (1975) with successive finger tapping (right hand superiority) and with copying various hand postures and fingers spacings, (left hand superiority) with children, have found right and left hand superiorities for tasks requiring the same type of processing. One explanation for these lateral hand differences in somatosensory functions may be that the untrained or less practiced individuals process unfamiliar and successive tapping tasks in a serial-temporal manner. The evidence above shows that the right hand or left hemisphere might be better for this. Likewise, because the trained or practiced individual 13 is familiar with the task, finger movements may not have to be successive because the focus of attention is on a different aspect of the informa- tion which is better processed in parallel, by the left hand or right hemisphere. Another explanation of the left hand superiority found on tasks such as tracing and copying is that they generat mental images of the figure or object whereby the processing is parallel in a spatial sense (Paivio, 1971), and image—symbols are more likely to be processed by the right hemisphere or left hand (as suggested by Hughlings Jackson, 1874, cited in Penfield and Roberts, 1959). Together these explanations suggest that the stage of reading depends on the relationship between practice or level of reading skill and the type of motor activity needed to process the tactual stimuli. Consequently, this relationship must also determine hand asymmetry and hemisphere specialization for reading. Moreover, less skilled braille readers, who attempt to decode configurations with their right hands in a serial or verbal processing manner should be limited in their reading ability until their left hand is introduced enabling the non~ verbal spatial forms to be processed in parallel. Therefore, increased reading skill would mark a change in what the reader was doing. The skilled reader must attend, integrate, and process the braille informa- tion coded by the two hands, perhaps in the way Critchley (1953) suggested through the adaptation of recognition and control roles for both hands. However, since braille configurations appear to always be a successive series of letters, when both hands are capable of reading braille, what processing strategy would then dominate (serial—left hemisphere or paral- lel-right hemisphere)? Accordingly, it seems that the braillist would 14 still have to decide which features (verbal or nonverbal) to encode. Therefore, it is still a question as to which finger will be the one of choice or superior for reading. The Effect of Phonological Cues on Tactual Perception and Teaching Methods Millar's (1975) study of how blind children three to ten years old encode nonverbal and verbal features used three lists of successive braille letters to determine at what skill level phonological and tactual cues were coded. The three item sets used were: (1) letters similar in feel, but dissimilar in name sound, (2) letters similar in name sound, but dissimilar in feel, and (3) letters dissimilar in both name sound and feel. The results, contrary to Kusajima's (1974) findings, suggested that less skilled blind readers tended to encode tactual features, while readers tended to encode verbal or similar sounding features. Overall, however, letters similar in feel and different in name sound were more easily recalled by both skilled and unskilled readers. Although hand and sex differences were not assessed for this study, the results also imply that the memory recall process for tactual and verbal features varies with ability and learning levels as well as with the hand used to haptically identify the dot features. These findings are of some importance for identifying characteristics of potentially ”good” and "poor" readers within each developmental stage, as well as for teaching braille readers how to improve their reading strategies. As an example, two early studies, one by Bfirklen (1917) on the legibility of the braille characters, the other by Maxfield (1928) 15 on observing teaching methods of braille, were concerned with these pro- blems. In comparison with poor readers, readers who were characterized as having "good" reading styles were found to make fewer repeated move- ments, to apply slight and uniform pressure on the dots, to proceed in a straight horizontal line, and to read ahead with the left hand before finishing the preceeding line with the right hand. And ”poor" readers repeated horizontal finger movements, used heavy unsteady pressure on the dots with excessive up and down movements on individual letters, and frequently used lip movements and inner speech, both of these being pro— perties believed to retard reading. These characterizations are very close to how Kusajima (1974) described beginning and advanced readers. Though Bfirklen (1917) did not indicate the skill levels of his sub- jects, three fourths of his subjects read better with their left hands. Consequently he argued that the reading difficulties encountered were problems in the identification of word forms: the size, shape, and spac- ing of the letters. By implication Bfirklen suggests the reader should be taught to use the left hand. However, Maxfield (1928) delineated of three methods of teaching children braille which significantly influence the comprehension of "word forms" in the braille code: the letter method, the letter—word method, and the word method. From her observations, Maxfield implied that the right hand was better for overcoming reading difficulties and should be used for braille reading. Maxfield (1928) assumed the letter method and the letter-word method only enabled the child's finger to encounter dots or letters one at a time in only a passive mode of perception. The word method, on the other hand, used contractions (of dot symbols) which represented several letters and phonological combinations, i.e., ‘tion', 'ment', 'bl', 'ch', 'sh'; 16 therefore, these symbols could be encountered and understood in a "sweeping movement across the line," in an active mode of perception as opposed to the passive letter, letter-word mode of perception. Maxfield advocated use of the word method because the others essentially repre- sented the passive modes of perception and reflected the major differ- ence between "good" and "poor" readers as a consequence of their respec- tive styles. Moreover, since the right hand was far more capable of sweeping movements, it was therefore best for reading. Whatever teaching method is taught (the letter method or the word method which are assumed to use the left and right hand respectively, both seem to overlook the possibility that verbal and tactual features are subject to different, yet necessary coding processes that may change according to the individuals ability to integrate the information coming from both hands. Consequently, to evaluate the learning of braille, one must consider: the level of practice, the task difficulty, the memory load, and the strategy best suited to the individual for coding tactual and phonological (verbal) features of the code. In summary, the studies of blind braille readers generally demon— strated that braille reading skills are acquired in stages that vary with ability and can change with practice, experience, or training; yet there may be constraints on the ability to progress through learning levels and to recall verbal and nonverbal features, constraints that are related to the task, the type of hand movements, and the cognitive pro— cesses used. In this regard, it seems that these factors would predict hand asymmetries based on the type of features by which the readers would find hardest to discriminate (Millar, 1975). This is an important l7 predeterminant for hand asymmetries, since even very skilled readers have difficulty recalling the tactual features of braille letters (Millar, 1975). Consequently the left hand must be essential for reading braille competently. One purpose of this study was to further substantiate this contention. There were two important developmental problems with the preceding research, though. First, the data on hand asymmetries very rarely consi— dered the age where the change-over took place, and second, the differ- ences between the sexes for processing braille configurations. Perhaps both children and adults who were unfamiliar with braille configurations would show even stronger left hand superiorities for the learning of braille dot configurations, and could provide additional insight into age and sex differences for processing braille. The research on somato— sensory functioning and motor asymmetries with sighted subjects is rele- vant to the discussion of these issues. Age and Sex Differences in the Development of Hemispheric Spgcialization and Cognitive Skills Somatosensorygdifferences in normal subjects on complex motor tasks. Several studies emphasizing the ability to discriminate stimuli tactually and the serial ordering of motor movements have revealed conflicting re- Sults for hand asymmetries. Figures such as simple geometric forms have been used to determine the degree of sensitivity and the left-right hand aSYTmnetry for normal individuals. For example, Becker (1931) applied geonmetric figures to the palms of adult subjects and discovered that the left palm required fewer presentations for recognition (3.94) than the rigth: palm. k 18 In one of two studies, Gardner (1942) used nonsense syllables made up of alphabet letters to test 30 sighted college students on a tracing task. The material consisted of heavy cord stitched to cardboard to form raised Latin-type letters. Fifty groups of these letters were traced from left to right and right to left. The unpracticed subjects identi— fied the letters from left to right and right to left, and made fewer errors with their left than their right hands suggesting they were pro- cessing the spatial forms rather than the verbal features of the letters. In a cork sorting experiment, Gardner (1942) found that the right hand was better when the corks were sorted into groups of cylinders and cubes, and the left hand was better when the corks were sorted according to size. There were no hand differences when the corks were sorted according to different sizes and shapes. One explanation for these differences that was suggested earlier may be that the determination of the size of an object like tracing of an object or area facilitates the generation of mental images or pictures that do not have to be processed in serial order. Consequently, haptic perception in the case of some sorting demands can focus more on the location of spatial features which are most adequately performed by the right hemisphere. These speculated hand differences also appear to be present in very young children. On a finger tapping task, which required three to five year old right—handed children to tap a key as quickly as possible, Ingram (1975) found the right hand to be better as early as three years. In children five to eight years of age, Denckla (1973) measured simple repetitive tapping movements (index finger against thumb) and successive tapping movements (each finger in turn against thumb) for each hand. Though right hand superiority was found on the tapping task for Denckla's 19 subjects, it decreased with age and no hand asymmetry was found for the successive finger movements. The girls, however, performed more rapidly than the boys on the successive movement task with both hands demonstrat- ing that the coordination necessary for this type of sensorimotor skill depends on inter- and intra—hemispheric cooperation. Evidence of sex-related differences in haptic percepEion. Witelson (1974) used a dichhaptic presentation technique to determine whether left- right asymmetries would be demonstrated for three dimensional nonsense shapes and embossed letters. This study, although done only with boys six to 13, found left hand superiority for the discrimination of nonsense shapes as early as six. No asymmetry was found for the discrimination of letters. On a similar task just using nonsense shapes, right handed boys and girls ages six to 13, were tested (Witelson, 1976). Left hand superi- ority was again reported for the boys but no hand differences were found for the girls. One explanation Witelson gives the for hand and sex differences is that girls may have bilateral representation for processing spatial stimuli and this would increase the probability of finding no hand asymmetries. In contrast, boys appear to have right hemispheric specialization for spatial stimuli by about the age of six. Witelson (1976) also suggests that the bilateral representation in females may be a function of hemi— spheric processing differences in females. Semmes (1968) proposed that the neural structures may be organized to function differently and this may change as the individual matures. Semmes (1968) proposed explanation in terms of a maturational effect for the change in processing differences is noteworthy. Earlier it was stated that the left hemisphere was dominant for verbalizing and analyzing 20 information serially and that the right hemisphere was dominant for establishing a spatial gestalt. In the context of language acquisition, Semmes has suggested that fine sensorimotor control exhibited by the left hemisphere for articulation would require an increase in the localization of similar sets of functional grammetical units, such as the ordering of phonemic elements. This would represent a focal or pre- cise coding of input and output elements in temporal and sequential ordering. By contrast, Semmes suggested that since the integration of encoding elements in the right hemisphere does not appear to be special— ized for language directly, it must be more diffuse or less forcused, thereby more conducive for visuo-spatial perception. This would allow the right hemisphere to grasp and integrate fragmentary or dissimilar units of information as a whole or as a total gestalt. Semmes speculates that these two processes would become more differentiated as the indi- vidual matures. An important implication can be drawn from Semmes' (1968) suggestions. First, specialization in females and males delineates two distinct coding processes that w0u1d undoubtably tend to dominate the person's perception of the world. In other words, since females may develop an earlier focal- ly-organized hemisphere which codes in an analytical and serial fashion, this mode of hemispheric dominance may supersede right hemispheric func- tioning, i.e., spatial coding, even when the task is more global or gestalt oriented. In this respect, males who do not show this type of "language” specialization might be at an advantage in terms of perceiving stimulus configurations, gestalts or wholes (Harris, 1975b). Thus, the specialization of the hands demonstrated by the previous studies reflects sensorimotor control consistent with right and left hemisphere 21 development in males and females and the possibility of interhemispheric competition for control of the motor pathways. The research on hemispheric commissurotomy patients and motor functioning seems appropriate for dis- cussion at this point. Hemispheric Specialization and Interhemispheric Competition for Control of the Motor Pathways Motor adaptations. The hemispheric specialization of commissurotomy patients has provided insights into the basis dissimilarities between the two hemispheres (Sperry and Gazzaniza, 1967; Levy, Nebes, and Sperry, 1970). The method of processing sensory information based on the commis— surotomy data has produced two general types of hemispheric processing. First, the right hemisphere is instrumental in spatial recognition and in completing fragmentary or partial information. If this capacity is impaired the individual will have trouble structuring and organizing his spatial environment. Second, the sensory input is monitored according to a competing-interacting response hierarchy, with the hemisphere whose specialty the tasks requires, will gain control over the motor pathways that are needed to perform the task. Levy, Trevarthen, and Sperry (1972) tachistoscopically presented conflicting stimuli simultaneously to both hemispheres. The commis- surized patients had to report one of two ways: (a) pointing with the right or left hand to the correct item among an array of complete stimuli or (b) naming or verbally describing the stimulus. The stimuli were left and right half chimeras of faces, antlers, and common objects. The re- sults were that the subjects pointed to the stimulus seen in the left visual field and named the stimulus seen in the right visual field. 22 Levy interprets the results not as an interference or transitory effect as was previously assumed but as the product of a competition effect. Here, two hemispheres have responded to a hierarchy programmed for the specialty necessary for the completion of the task. Trevarthen and Kinsbourne (cited in Levy, 1972) also noted that in all three categories of stimuli tested, each hemisphere tended to com— plete its own half-field percept. Evidence for this was demonstrated when without warning a manual response was blocked and the subject was requested to verbally describe the image seen. Since the object described was that projected to the left hemisphere, each subject described that image correctly. Consequently, the right hemisphere in the right handed subjects seemed to use a strategy that (l) perceptually apprehended the shape, (2) stored the visual material as a visual code without recourse to verbal labelling, and (3) dominated the response when no verbal label— ling was required, even with the ipsilateral right hand. The specialty necessary for the completion of these tasks lay in the perceptuo-cognitive realm of cerebral functioning, requiring oculomotor search and retrieval as opposed to mechanisms governing the motor expression in pointing. Bryden and Allard (1976) suggest that the perceptual processing prior to naming represents a global preprocessing operation, whereby the stimu- lus is normalized or stored and attention is focused on "relevant charac— teristics of the target" (p. 198). Because this type of hemispheric pro— cessing is more closely associated with the right hemisphere it may be that this ability is a structural by-product as Semmes (1968) suggested. It is also indicative of task complexity. The research by Bryden and Allard illustrates this latter point. Ten different typefaces of Roman letters were tachistoscopically presented 23 to college students. The more difficult script—like letters exhibited a left visual field or right hemisphere advantage, while the others were best recognized in the right visual field. The authors argue that the greater global preprocessing capabilities of the right hemisphere were necessary to evaluate the more complex features of the script like type— face. Essentially, the studies cited above illustrate that the two hemi— spheres can accomplish the same task but by characteristically different strategies. On the one hand, in reference to visual stimuli, there appears to be a two part process wherein (1) when visual material needs to be preprocessed and visually coded without a verbal label, the right hemisphere will dominate the motor system whether on the left or right side; and (2) when the visual material requires a verbal and/or concept— ual symbolic transformation, the left hemisphere will dominate the motor system. Braille stimuli on the other hand still pose a question of whether braille will require a similar two-part decoding process, since the braille reader is likely to use multiple strategies in the course of completing the spatial-verbal material. Consequently, which hemisphere controls the motor system leading to different braille identification strategies will most likely be governed by cortical influences and their past response pathways. Braille letter learning in sighted subjects. There have been several studies with blind children that have suggested that the right hand is better for reading braille, at least initially (Fertsch, 1947; Kusajima, 1974; Maxfield, 1928), while other studies have implied that the left hand is superior for reading braille, a complex tactual motor 24 task (Hermelin and O'Connor, 1971a, 1971b; Gardner, 1942). Subsequently, several studies were designed to determine whether left—right hand dif- ferences related to braille reading were a function of age, sex, practice, and/or hemispheric specialization strategies (Rudel, Denckla, and Spalten, 1974; Rudel, Denckla, and Hirsch, 1977; Feinberg and Harris, 1974; Wagner, 1976). These latter studies used sighted subjects and resulted in findings that are still somewhat conflicting with respect to the degree to which each variable is likely to contribute to hand asymmetries, but they do pro- vide greater insight into the problem of determining hand superiorities for reading braille. Inexperienced sighted children, ages seven to 14 years old, were used by Rudel et a1. (1974) to study single—letter —raille recognition in a paired associate task. The children's index fingers were guided over the braille letters by the experimenter to begin with, in order to orient the children to how the letters felt. Then the children were en— couraged to feel the letters on their own. In general, the left hand was superior to the right, but the effect varied by age and sex. In this respect, Rudel et al. (1974) suggested that the order in which the hands were tested may have had a greater affect on the girls than the boys. The boys were superior with their left hand regardless of the order of testing (right hand or left hand first), and significantly so by age 11. But, the girls achieved left hand superiority only after using their right hands first and then, only after the age of 12. The left hand superiority found on this letter recognition task by the girls may be the result of a shift from early verbal coding strategies to the development of lateral representation specific for processing spatial cues, or, perhaps that when the left hemisphere is activated by 25 right hand reading, the right hemisphere may be activated in such a way as to ameliorate the sensitivity of the left hand. In this regard, both Rudel et a1. (1974) and Witelson (1974) have argued that the performance of one hemisphere may be the result of prior activation or stimulation of the other. In another study by Rudel et al. (1977), subjects were asked only to make a same-different judgement about braille letters presented, pre- sumably to minimize left hemispheric functioning. The results strongly paralleled the findings of the earlier Rudel et a1. (1974) study, but the hand order effects were not as strong. Feinberg and Harris (1974) tested undergraduates and had them learn braille letters by feeling them actively, to simulate the searching move- ments often observed in braille readers. Two sets of eight letters were used: four to each hand. There were a total of 40 trials per hand in— stead of six trials as used by Rudel et a1. (1974). When the hands were tested in straight alternation over the 80 trial periods, the left hand was found superior for right handers. These findings were confirmed by Harris and Wagner (cited in Harris, 1975a). Wagner (1976) was also able to demonstrate an overall left hand superiority for sighted children and adults similar to that found by Rudel et a1. (1974). However, in contrast to Rudel's et a1. (1974, 1977) findings, the pattern of hand asymmetry revealed lateralization for the boys (age nine) and no consistent demonstration of right hemispheric pro- cessing in girls until they reached college age. Wagner and Harris (1977) attribute the age and sex differences found, to invariant lateralization in boys, that is, age did not necessarily determine specialization for spatial perception by the right hemisphere, and, possibly to an increase 26 in memory demands for girls. Consequently, Wagner and Harris (1977) con- cur with Millar's (1975) speculation that braille is a complex task that is capable of being decoded by a variety of strategies which depend on practice, sex and age of the individual, but conclude that braille is predominately a right hemispheric task which is suggestive of left hand superiority. Contributions of the Current Study This study proposes to determine whether the left or right hand is superior in braille reading by blind readers under two conditions: (1) controlled motoric scanning, and (2) verbalization. There were several remaining questions about the previous studies which suggested the neces- sity for a new study, for example: (1) will the speed and accuracy of braille reading be affected when a subject reads letters as opposed to sentences; (2) will speed and accuracy of braille reading be affected when vocalizing, i.e., reading aloud; (3) do good (fast) and poor (slow) readers use different strategies when reading; (4) does the time of onset of blindness affect hand asymmetry? The assumption underlying the two experimental conditions, controlled motoric scanning and verbalization, is that each facilitated a different cognitive process. Two factors that directly affect cognitive processing in braille reading are: (1) level of reading skill and (2) the task of decoding the braille symbols (Hampshire, 1974). The type of stimuli selected for this study, i.e., letters and paragraphs, may appear to be similar, but it is likely that each requires different hemispheric strate— gies for decoding tactual-spatial input. Consequently, in a single letter identification task in which the right or left direction of motoric scan 27 is controlled, the left hand or right hemisphere should be better at pro- cessing the letters, regardless of whether speech is involved. Paragraph reading which involves complex sentences, i.e., understanding the basic contractions as well as syntax and meaning, should require a different type of processing not only because of its complexities but also because of the way the individual choses to encode the information (Millar, 1975). Therefore, the cognitive processing of braille symbols might vary according to the level of braille reading skills, sex, and decoding strategies. Also, since the development of hand asymmetries appeared to be re— lated to age, to what extent does length of blindness affect the develop— mental level of skill in reading and the acquisition of different modes or strategies of processing. Experimental Hypotheses and Tests The first hypothesis is that the left hand or right hemispheric will be superior in identifying single braille letters despite having to say them aloud. This is primarily expected because each letter must be pro- cessed as an individual tactual configuration without phonological struc— turing or scanning aids. The first hypothesis is in direct contrast to the commissurotomy studies of chimeras by Levy et a1. (1972), in which verbal labelling of a nonverbal stimulus was found to be a left hemispheric skill for right handers. In experiment I, the letter identification experiment, it is proposed that the identification of a specific spatial configuration appears to warrant the initiation and continuation of the tactual-spatial strategy of the right hemisphere which at that time might dominate the motor pathway. However, if right hemispheric cognitive processing 28 persists during word or sentence reading, slower and poorer readers will develop. In other words, these readers will have continued using a let- ter by letter or ”static” reading style. Usually they read with their two fingers close together. The major consequence encountered when using a "static" reading style is a decreased level of comprehension. The second hypothesis is that poor or slow readers will show left hand superiority for paragraph reading while good or fast readers will read equally well with both middle fingers. Evidence, thus far, does not appear to show any clear developmental level for simultaneous improvement of left developmental level for simultaneous improvement or impairment of left or right hand reading performances. But, two factors apparently influence the development of the better reader. One factor is inherent in the reading style which it is more "dynamic." The two fingers appear to separate and begin to show a ”division of labor,” much the way Critchley (1953) described earlier. The right finger sweeps across the line more freely, indicating that the left hemisphere has developed an adequate level of nonverbal—verbal transfer. The second factor is that there appears to be a sex difference in the acquisition of right or left hand superiority. It could be argued that females achieve an earlier right-visual and right-tactual superiority because their left hemispheres mediate both verbal and nonverbal stimuli. This particular kind of cerebral representation during childhood could very well afford females a slight advantage in expressive language skills (McGlone and Davidson, 1973) and early right hand superiority (Rudel, 1974; Feinberg and Harris, 1974), which would allow for an easier transition from ”static" to ”dynamic" reading styles. This kind of arrangement would clearly benefit females unless competition effects within the same hemisphere were to make it 29 impossible to process nonverbal stimuli in the left hemisphere. The third hypothesis is that there will be interhemispheric effects when left hand superior subjects read sentences aloud. Verbalizing will obscure the left hand effect, i.e., right hemispheric cognitive proces— sing. This finding would be consistent with the findings for commissurized patients which demonstrated the competing response hierarchy (Levy et al., 1972). The process of verbalizing will probably lengthen reading times and increase the number of errors, but left hand superior subjects on the letter identification task should show less of a left hand effect. Finally, the fourth hypothesis is that congenitally blind readers will read better than partially signted or adventitiously blind individuals. In the case of the congenitally blind child, birth stress may be reflected in the grequency of the left handedness implying right hemisphere superi— ority. Wortchell ((1951) suggests that congenital and adventitiously blind individuals might differ in their methods of translating somatosen- sory cues because of the latters' propensity to form visual images. The adventitiously blind individual might rely more on propositional thinking or visual translations instead of tactual—spatial means of interpreting the symbols. Thus, handedness and the time of onset or length of blind- ness may have some relevance for determining good or poor braille reading skills. Clearly accurate and efficient braille reading depends on the strategy used to comprehend the material. Rationale for using the Middle Finger There are a variety of methods of teaching braille reading. However, the technique most frequently used involves variations of the two index fingers moving across a line of braille type together. The remaining 30 fingertips although available are generally not used. In an experiment to determine the amount of transfer of braille reading to the normally unused fingers (excluding the thumbs), Foulke (1964) found that the middle finger took significantly less time and made fewer errors than the ring and little fingers. He attributed the differ— ences to the musculature associated with the ring and little fingers and the greater degree of control and manipulation in the middle and index fingers. In addition, Hermelin and O'Connor (1971a, 1971b) found that braille readers were able to use their middle fingers and adequately com- prehend the material. More importantly, the asymmetry effects were stronger for the middle fingers than the index fingers. It could be concluded that the middle fingers will serve as an ade— quate substitute for the index fingers and the relatively unpracticed middle fingers will make slightly more errors and be slower in performance than the index fingers. But, using the unpracticed middle fingers will more clearly indicate the functional asymmetry of the two hemispheres by minimizing the effect of finger sensitivity due to years of experience across the age groups while synthesizing the information. In summary, the four hypotheses are: (1) In the task of letter identification there will be left hand superiority for both sexes, despite having to say the letters aloud; (2) Slow or poor readers will demonstrate left hand superiority on the paragraph reading task while fast or good readers will read qually well with either hand; (3) Vocalizing while read— ing paragraphs will obscure or mask the left hand superiority effect for those who demonstrate it; and (4) Congenitally blind subjects will read better than adventitiously blind or partially sighted readers. III. METHOD This study has two experiments with two preceding questionnaires. The questionnaires helped to determine the handedness and the history of blindness for each subject. In experiment I, the letter identification experiment, the Controlled Motoric Scan Measure was used to evaluate hand asymmetries for simple, elementary reading material. For experiment 11, the paragraph reading experiment, the Gates-MacGinitie Reading Test, Form D—2 was used to determine hand asymmetries for complex reading materials. Subjects The sample consisted of 33 subjects, 16 females and 17 males, who were braille readers attending Michigan School for the Blind. The stu- dents were in grades 5-12. All used braille as their primary means of reading, although their degree of blindness varied. Their ages ranged from 10 to 20 years. Materials The Handedness Questionnaire. The handedness questionnaire shown in Appendix A, consisted of 20 items. Twelve items were from the Edinburgh Inventory for Handedness (Oldfield, 1971), and eight new items were made to account for characteristics specific to the blind. 31 32 History of Blindness. This questionnaire contained in Appendix B, consisted of nine items designed to assess the history of blindness and the acquisition of braille reading skills. Controlled Motoric Scan Measure. This measure was devised to test speed and accuracy under conditions controlling for a left to right reading and scanning bias. (A copy of the measure is provided in Appendix D.) There were eight lists of 26 randomized letters typed in braille on a standard sheet of braille paper (30 x 25 cm). The braille dots were of standard height and size, since they were transcribed using a Perkins Braille Typewriter. The practice list was horizontally arranged and the control and six experimental lists were vertically arranged to be read from the top of the page to the bottom. The control and experimental lists included a total of 182 letters. There were no abbreviations or contractions on this measure. If an error was committed while identifying a letter, the letter was crossed out. The number of mis-identified letters for each column was indicated at the bottom of each column. The Gates—MacGinitie Reading Test, Form D-2. Part I of the Gates— MacGinitie Reading Test for grades 4-6 was used to measure the dependent variables, reading speed and accuracy (See Appendix E). This test con- sisted of 30 short paragraphs, each 2—3 sentences long and containing be— tween 28 and 30 words. Four words followed each paragraph. One of these was the correct response for the paragraph, i.e., completed a sentence or answered a question appropriately. If the subject answered with the wrong word the error was marked on the paragraph sequence number above the space where the paragraph time was recorded. The 30 paragraphs in experiment 11 were divided in three parts. Part I of this experiment consisted of six paragraphs with approximately 33 174 words each. Parts II and III contained 12 paragraphs with approxi- mately 348 words each. The paragraphs were distributed among the three parts in such a way that each part had a equivalent level of difficulty. Part I constituted the control condition of the second experiment, and Parts 11 and III the repeated measures of that experiment. Each paragraph was transcribed by a skilled braille typist using the appropriate braille abbreviations and contractions. Three paragraphs were typed on one page, a standard sheet of braille paper (30 x 25 cm). The line lengths approximated those in braille reading classes. The braille dots were of standard height and size because they were trans— scribed using a Perkins Braille Typewriter. The experimenter's manuals were set up to parallel the subject's test booklets, i.e., three para- graphs to a page. Data Record Sheets. Data recording sheets were designed to facili- tate the recording of the data. The data recording sheet for the first experiment contained each of the alphabetical lists used (see Appendix D). It was used by the experimenter during the testing sessions and at the time the lists were coded from the taped recordings of the first experiment. This provided a check against experimenter error and protec- tion against mechanical failure of the recording equipment. A second data recording sheet provided spaces for recording paragraph reading times, sequential errors, total times and total errors. It also provided spaces for summarizing the data from the second experiment (see Appendix F). Additional testing apparatus included a stop watch, a cassette tape recorder, and cassette tapes. 34 Research Desigp Each Subject in experiment I read a horizonal list of 26 randomly arranged alphabet letters in his normal reading style for practice. A vertical column of 26 randomly arranged alphabet letters was then pre- sented as a control list, to be read by each subject with his index fin— gers or in his normal reading style. The six experimental lists or trials that followed were read with the middle finger of each hand in an alternating finger procedure. Half of the subjects began with the middle finger of the right hand and the other half began with the middle finger of the left hand. Following the six trials for experiment 1, three sample paragraphs were read by each subject in his normal reading style. The subjects were asked to select an answer for each paragraph that best completed the paragraph or answered the question. The subjects, then, read and answered three paragraphs silently and three paragraphs aloud in their normal read- ing style. The subjects were told they would be reading the remaining 24 para- graphs with the middle fingers of each hand in an alternating finger pro- cedure (three paragraphs were read with the right or left middle finger followed by three read with the left or right middle finger, etc.). Again, three sample paragraphs were presented to be read either silently or aloud but this time with the middle finger of the subject's choice. The first 12 paragraphs were read silently and the second 12 para— graphs were read aloud depending on the condition. The results for each silent and alound condition were compiled separately and compared to the three silent and three aloud control paragraphs. 35 Procedure Subjects were tested during two sessions. The first session con— sisted of an interview with each subject by the major experimenter. The Handedness Questionnaire and the History of Blindness Questionnaire was administered. The objective of the first session was to determine handed- ness prior to the experimental testing session so that subjects could be randomly assigned to the experimental conditions. This interview was also arranged and conducted in order to maintain confidentiality regarding each subject's history of blindness. During the second session, trained aides (three undergraduate women) administered the Controlled Motoric Scan Measure and the Gates-MacGinitie Reading Test. The major experimenter introduced the subject to the aide. The subject was seated Opposite the aide, who gave the subject a test booklet containing the braille copies of the measures. The experimenter had a similar booklet containing the non-braille copies of the instruments and the appropriate instructions for their administration. Experiment I. The experimenter began by giving the instructions for the Controlled Motoric Scan Measure, experiment I (for instructions see Appendix C). The experiment began with the subject reading the practice list. Then each subject was asked to read the control list in his normal reading style. The experimenter timed the subject and coded which hand the subject used. After asking for questions, the experimenter stressed speed and accuracy, started the tape recorder and stop watch, and checked the errors committed as the subject read aloud. After the subject completed the control list, the experimenter stopped the tap recorder, recorded the times elapsed, and gave the instructions 36 for the six experimental trials. After checking for questions, the ex- perimenter indicated the appropriate starting finger and reminded the subject to say the letters out loud and to read as fast as possible with accuracy. Then the experimenter started the tape recorder again and the timing for the first trial began. After each trial, the experimenter stopped the subject and reminded him to say the letter aloud and switch to the middle finger of his opposite hand. Then the experimenter turned the page, instructed the subject to begin and started the timing. This procedure was the same for five additional trials, a total of seven - one control and six experimental lists. Experiment 11. Following the completion of the alphabetical lists, the experimenter gave the instructions for the sample paragraphs of the reading test (see Appendix E). After the subject completed the samples, the experimenter gave the instructions for Part I, the control condition, in which the subject used his normal reading style. After checking for questions, the experimenter turned the page, instructed the subject to read the first paragraph and say his answer aloud. The subject gave his response and stopped. The experimenter recorded the time expired and whether or not an error was committed. If an error was committed the experimenter noted the error by marking the paragraph sequence number above the space where the paragraph time was recorded. Then the experi- menter indicated when the subject was to start each successive paragraph. After the first three paragraphs were completed, the subject was in- structed to read the last three aloud. When the subject completed Part I, the experimenter gave instructions for Part II, which explained the alternating finger procedure. The ex— perimenter then turned back one page and instructed the subject to use 37 the middle finger of one of his hands to read the first, second, and third successive paragraphs using the same finger. Then the subject was instructed to read silently or aloud according to condition. Each sub- ject read 30 paragraphs, 15 silently and 15 aloud. There were 30 re— corded times for each subject. The subject changed fingers at the end of each page, or every three paragraphs. The experimenter reminded the subject when to change fingers and when to read aloud. After completing section two, the experimenter gave the instructions for Part III, which used the alternating finger procedure but in a dif- ferent sequence. After checking for questions, the experimenter indicated the appropriate finger to begin with, reminded the subject to say the words aloud, stressed speed and accuracy, turned the page and started a procedure similar to the second part. The elapsed time was again recorded for each response along with errors. 'At the end of each page, or every three paragraphs, the experimented reminded the subject to change to the middle finger of the opposite hand and to read silently or aloud. Upon the completion of the 30 paragraphs, the subject was thanked for his cooperation and told he did a good job. IV. RESULTS Six subjects (three boys and three girls) who were determined to be left handed by the Handedness Questionnaire were excluded from the analyses because there were not enough subjects per cell to make a statistical com— parison. Consequently, the total number of subjects in these analyses was 27. A brief description of the six subjects' performance of both experiments will be presented after the data on the 27 right-handers are presented. First, the results are presented according to the analyses done by sex of subject. Both single letter and paragraph reading results are described. In the analyses based on sex, the speed of reading paragraphs is the focus since the subjects had very little difficulty in accurately answering the paragraph questions or paragraph statements. Subsequent analyses for hand preference are then presented, with a comparison of dif- ferences between individuals. The results from the preference analyses were also contrasted to determine the extent of preference group differ— ences. Finally, comparisons between sex and type of blindness are pre- sented. Test Trials for Letter Identification and Paragraph Reading Sipgle braille letters. In order to test for Hypothesis 1, left hand and right hand performances in the reading of braille letters, were 38 39 compared by subtracting the control reading time from the mean for the three trial reading times for each finger. An error rate was obtained by subtracting the mean number of trial errors for each hand from the control error rate. After grouping by sex of participant, t-tests were performed contrasting left and right hand reading speed scores and errors for each sex. There were no significant differences found (see Table Gl, Appendix G). Paragraph readipg, To determine the difference between sex and right and left hand performances (hand differences) on the paragraph reading task difference scores were computed by subtracting the mean of the three control trials read silently from the mean of the experimental silent trials for the right finger and the left finger, respectively. Similarly, the mean for the three control trials read aloud was sub- tracted from the mean of the experimental aloud trials for the right and the left finger, respectively. The differences between right and left fingers were tested using t~tests for dependent samples. (The results are presented in Table GZ, Appendix C.) These analyses failed to dis- close any significant effects for either the male or female groups. When all the subjects were combined, the difference between the right and left fingers was also nonsignificant. To evaluate the effect of verbalization on the differences between right and left hand performances, additional t-tests were performed (see Table H1, Appendix H). Table 1 presents a summary of the results for the effect of varbalization as a function of handedness. A significant re— sult was obtained in the comparison between the silent-right and aloud— right conditions for the male group and the combined group with the better performance occurring in the silent reading condition. There was 40 a trend in the same direction for the left hand. The combined groups comparison between the left versus right hand aloud condition also showed a trend toward better left hand performance. Table 1. Results of the t-tests between the silent and aloud conditions for the right and left hand in the paragraph reading task. Difference Means (in seconds) Hand Group Silent Aloud t df Boys 152.50 195.14 -2.25* 13 (0-154.3l) (0-102.02) Right Girls 106.61 113.61 - .68 12 (o: 67.38) (o= 78.94) Combined 135.59 155.88 —2.20* 26 (o=121.70) (o=lZ7.6S) Boys 94.35 110.14 -1.83 13 (0:102.59) (0-102.02) Left Girls 79.23 80.38 — .05 12 (e: 67.69) (0: 91.62) Combined 87.07 95.81 - .76 26 (e: 86.23) (0: 96.48) (*p < -05) Speed of reading. Next, to test for Hypotheses II and III and to assess the different effects of reading speed, subjects were grouped according to verbalization condition, sex, and reading speed (fast or slow). Fast and slow readers were determined by selecting the subjects, 41 with the fastest total (left and right hand) reading times (7 boys and 7 girls) to be in the fast reading group and those with the slowest total reading times (7 boys and 6 girls) were placed in the slow reading group. Comparisons were made for right and left hand reading performances, and Table 2 presents the results of the eight comparisons. Table 2 shows Table 2. Results of the t-tests between right and left hand performance in the silent and aloud conditions for the paragraph reading task as a function of sex and rate of reading. Difference Mean Rate of (in seconds) Condition Group Reading Right Left t df Fast 74.14 52.14 1.50 6 (o: 37.99) (0: 16.55) Boys Slow 250.85 136.57 1.09 6 (o=178.70) (o=135.54) Silent Fast 67.71 46.28 4.84** 6 (0: 27.23) (0: 26.83) Girls Slow 152.00 117.66 .58 5 (o: 73.00) (o= 82.71) Fast 100.28 77.00 1.27 6 (e: 50.55) (0: 35.09) Boys Slow 290.00 143.28 1.37 6 (o: 50.55) (0: 35.09) Aloud Fast 71.71 41.42 3.76** 6 (o= 42.34) (0: 28.18) Girls Slow 162.50 125.83 .58 5 (e: 86.48) (o=lZO.78) (**p < .01) 42 that for females, left hand performances were better for the faster read- ing group in both the silent and aloud conditions, respectively. In the comparison between left and right and silent reading for males, there was a trend toward faster left hand reading. The males also demonstrated a significant result for silent right hand reading (74.14 seconds) when com— pared with aloud right hand reading (100.28 seconds) (see Table 11, Appen— dix I). Verbally-stated Hand Preference Versus the Actual Hand Use To determine which finger was the "finger of choice" for braille readers, the 27 right-handed subjects were grouped according to stated hand preference and the actual hand used. Initially, the participants were divided into three groups according to the hand that they perceived as best for reading. This was their stated hand preference as determined by the History of Blindness Questionnaire, i.e., (1) those who felt they read better with their right hand; (2) those who felt they read better with their left hand; and (3) those who felt they read well with both hands - had no hand preference. Next, subjects were grouped according to their actual hand (behavioral) used, i.e., their preference as indi- cated by choice of index finger used during an initial reading performance. Figure 1 presents a comparison of the actual (behavioral) hand used and the verbally-stated hand preference. The graph shows an increase in the number of subjects preferring their left hand a decrease in the num- ber of subjects with preferences for their right hand or for neither hand. The males who stated a right hand preference or no hand preference chose to use their left hand more often. For the females, the percent of those 43 Figure 1. Comparison of stated hand preference and the actual hand used for braille reading by males and females. muzmdmmmmm QZ