w M \l‘ I \~ ‘ , \ I ‘ J 1 ‘ HIM H l H «WWW {Ill 1 HI _\_3 -—|I\J \IN I THS W --- WWWW WWWW WW i V;— .5 3 1293 00814 9944 L 1 u n Michigan State University THES's This is to certify that the thesis entitled A DEVELOPMENTAL STUDY OF HAND SPECIALIZATION USING A DICHHAPTIC PERCEPTION TASK presented by Arthur Pomerantz has been accepted towards fulfillment of the requirements for “DA. deg-cc in PsyChOIOgy $0.1m; he; . an professor Date May 7. 1980 0-7639 -”~9-—.-4- ‘ . J ([fl-‘N\ r ‘ OVERDUE FINES: 25¢ per day per item RETURRIIS LIBRARY MATERIALS: Place in book return to remove 29"" charge from circulation records A DEVELOPMENTAL STUDY OF HAND SPECIALIZATION USING A DICHHAPTIC PERCEPTION TASK By Arthur P. Pomerantz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1980 ABSTRACT A DEVELOPMENTAL STUDY OF HAND SPECIALIZATION USING A DICHHAPTIC PERCEPTION TASK By Arthur P. Pomerantz This study was an investigation of age and sex differences in right hemisphere lateralization of spatial abilities as indexed by performance on a tactile discrimination task. Seven-, 11-, and l5- year-old right-handed children (N=6l) were asked to feel pairs of nonsense shaped forms, simultaneously, and then to identify them on a visual display. The results showed nonsignificant left-hand advantages for the seven- and ll-year-olds; a significant left-hand advantage for the l5-year-old girls; and a nonsignificant right-hand advantage for the 15-year-old boys. The results, for girls, support Nitelson's (1976) conclusion that in girls, right-hemisphere specialization for tactual discrim- ination develops late. The boys' results, however, fail to confirm Nitelson's finding of a left-hand advantage in boys and no hand asymmetries in girls. Differences between the current and previous studies may reflect differences in stimulus attributes, type of subject response and other procedural details. ACKNOWLEDGMENTS I would like to thank Lauren Harris, the chairman of my thesis committee and academic advisor, for his careful attention to the preparation and writing of this thesis and for his enlightening comments and discussions. I am grateful also to the other members of my committee, Ellen Strommen and John McKinney, for their helpful suggestions. I also wish to express thanks to Roger Buldain for his assistance in the data analysis and to Bill Bukowski for his sugges- tions and advice. Finally, I am grateful to my family and friends for their encouragement and support. ii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION Cerebral Lateralization: Background Studies in Hemispheric Specialization in the Somes- thetic Modality. . Studies with Neurological Patients Studies with Normal Adults Developmental Studies: Braille Developmental Studies: Dichhaptic Perception Task The Development of Cerebral Lateralization: Theory Sex Differences . . Sex Differences in Cognitive Abilities . Sex Differences in the Development and Degree of. Lateralization . . . . Place of Proposed Study in the Literature . METHOD Subjects . Stimuli Procedure RESULTS Comparison of Left and Right Hand Scores Analysis of Group Effects . Simple Difference Between the Hand Scores Phi Coefficient Scores Scores Summed Across Hands . DISCUSSION Explanation of Differences in Dichhaptic Perception Studies . . . . . Page vi Sample Size . . Task Difficulty . . Stimulus Attributes Type of Response . . . . . . . . Use of Linguistic Stimuli in Same Session Other Variables . APPENDICES A. Letter Sent to Parents of Subjects 8. Instructions to Subject REFERENCES iv Table 10. ll. LIST OF TABLES Mean Number of Correct Right and Left Hand Scores and T-tests for Each Age X Sex Group and All Groups Combined . . . . . . . . . . . . Number of Subjects Showing Superior Performance with Left or Right Hand, Grouped by Age and Sex ANOVA for Simple Difference Scores in Identification of Left and Right Hand Stimuli Mean Left and Right Hand Scores and T-tests for Each Age by Sex Group and All Groups Combined for Order 2 ANOVA for Simple Difference Between-the-Hands Scores in Frequency of First Response Mean Phi Coefficients for Each Sex at Each Age ANOVA for Phi Coefficient (corrected for accuracy) Difference Between-the-Hands Scores . . . ANOVA for Phi Coefficient Difference Between-the-Hands Scores for Frequency of First Response . Mean Phi Coefficient Scores for Hand Differences in First Response for Males and Females Under Both Orders of Presentation . . . . . . . . Mean Phi Coefficient Scores for Hand Difference in First Response at Each Age for Each Trial Block . Comparison of Procedures and Results in Studies Using the Dichhaptic Perception Task . . . . Page 36 37 42 43 45 46 47 48 49 49 56 Figure 10. LIST OF FIGURES The 5 pairs of nonsense shapes used by Witelson . Stimuli used in the present study for the nine- and ll- year-old children . Stimuli used in the present study, for lS-year-olds The testing situation for the two-hand tactile perception task, devised by Nitelson Scatter-plot of left- versus right-hand responses for seven-year-olds . . . . . . . . . Scatter-plot of left- versus right-hand responses for ll-year-olds . . . . . . Scatter-plot of left- versus right-hand responses for lS-year-olds . . . . . . . . . Relationship between difference between-the-hands scores and total scores in 7-year—olds . Relationship between difference between-the-hands scores and total scores in ll-year-olds Relationship between difference between-the-hands scores and total scores in lS-year-olds vi Page 29 31 32 33 38 39 4O 51 52 53 INTRODUCTION The purpose of the present study was to investigate age and sex differences in cerebral lateralization of spatial abilities as indexed by performance on a tactile discrimination task. I shall present a brief overview of research in cerebral lateralization and then review studies in hemispheric specialization for tactual perception. This review will include developmental research as well as studies of normal and brain-damaged adults. A discussion of sex differences in cognitive functions follows, and three theoretical positions proposed to account for such differences are outlined. Cerebral Lateralization: Background Studies of cerebral lateralization in the l9th century focused on the function of the left hemisphere and were based mainly on the effects of injury or disease in that hemisphere on speech production and comprehension. The prepoccupation with the "major" hemisphere continued (with some notable exceptions) until the post-World War II period, probably because lesions of the left side of the brain pro- duced more obvious behavioral changes (i.e., disruptions in language) than lesions of the right side. Nonetheless, investigators as far back as Hughlings Jackson speculated on the specialized role of the right hemisphere, which Jackson had characterized as being concerned with "perception" or T "visual imagery" (Jackson l932/l874). Since then, studies of patients with right-hemisphere lesions have indicated deficits in a variety of visuo-spatial tasks and in the ability to remember music, faces, and nonsense shapes (Milner, 197l). Such findings are thought to reflect right-hemisphere specialization for these capacities. The reliance on clinical populations for subjects also persisted until the middle of the 20th century when the emergence of new experi- mental techniques and devices made it easier to investigate cerebral asymmetries in normal individuals. One of these techniques is dicho- tic listening, in which different auditory stimuli are presented simultaneously to the two ears and the subject is tested for recogni- tion or recall. Since contralateral connections are stronger or more direct than ipsilateral ones, a right ear advantage for verbal stimuli, for example, implies a left-hemisphere superiority for pro- cessing this kind of material. An analagous rationale underlies the method of half-field presentations of visual stimuli. Researchers employing these techniques usually find a right ear or right-visual field superiority for the recognition of words, dig- its, and stimuli presented in a temporally patterned sequence. Left- ear advantages have been found for the perception of animal sounds and environmental sounds (Knox & Kimura, l970). For musical stimuli, ear asymmetries of different direction have been found depending on the precise stimulus attributes and the kind of processing required. When rhythmic and time factors are stressed, a right-ear advantage is more likely to appear, whereas presentation of melodies tends to produce left-ear advantages (Gates & Bradshaw, l977). Left hemi-field superi- ority has been demonstrated for the visual perception of spatial con- figurations (Kimura, l969) and faces (Geffen, Bradshaw & Wallace, 1971; Pirozzolo & Rayner, l977), for the judgement of depth (Durnford & Kimura, l97l), and for identification of line orientation (Kimura & Durnford, l974). The dual nature of the two hemispheres has been characterized in many different ways. The left hemisphere has been described as logi- cal, verbal, analytic, symbolic, propositional, and concerned with temporal processing. Its counterpart has been conceptualized as visuo- spatial, perceptual, nonverbal, synthetic, holistic, appositional, and involved in parallel processing (Bogen, 1969). To characterize the left hemisphere as verbal and the right as nonverbal, however, is simplistic since both hemispheres in most individuals are probably capable of both types of processing, to some extent. Furthermore, dichotic listening and visual half-field presentation studies indicate that certain aspects of the verbal or nonverbal stiumli produce the ear or hemi-field asymmetries. Schwartz and Tallal (1980), for example, discovered that the time dependent, acoustic qualities of speech but not the phonemic characteristics bring out a right-ear advantage. Studies of Hemispheric Specialization in the Somesthetic Modality Studies with Neurological Patients In contrast to experiments involving vision or audition, there has been little research dealing with lateral asymmetries in tactual perception, and most of this research has been with subjects suffer- ing from neurological disorders. Semmes, Weinstein, Ghent, and Teuber (1960) tested war veterans with penetrating brain injuries on point localization, sense of pas- sive movement, two-point discrimination, and touch-pressure thresholds. When comparing subjects with right-hemisphere lesions with those with lesions of the left side, they found that the number and severity of deficits in the contralateral hand was nearly equal for the two groups. Right-handed deficits were correlated more highly with damage to the left-hemisphere sensori-motor region than were left-hand deficits to localization of right-hemisphere lesions in the corresponding region. This result led the authors to propose that right-hemisphere repre- sentation of sensory and motor capacities is more diffuse than that of the left hemisphere. Subjects with left-hemisphere lesions in this study showed greater left-hand sensory impairment than the group with right- hemisphere damage showed for the right hand. The experimenters therefore concluded that the left hemisphere has more ipsilateral sensori-motor influence than the right. This conclusion is supported by reports from neurologists that left-hemisphere lesions often produce bilateral manual defects (such as bilateral finger agnosia) whereas right hemisphere lesions usually do not (Critchley, 1953). Corkin, Milner and Rasmussen (1964), on the other hand, found that the two hemispheres did not differ significantly with respect to control over ipsilateral sensation. 8011 (1974) administered a battery of tactile-perceptual tests (tactile finger localization, fingertip number writing perception, and tactile form recognition) to patients with lateralized brain lesions. Patients with right-hemisphere lesions made more errors than those with left-side lesions. Right-cerebral lesions produced greater deficits in the contralateral hand than left-cerebral lesions and in the ipsilateral hand as well. This finding supports the view that the right hemisphere is specialized for tasks of this kind. It also conflicts with the earlier reports of greater or equal left-hemisphere control over ipsilateral sensori-motor functions. This discrepancy may be par- tially the result of Semmes et. al.'s (1960) use of subjects who had suffered penetrating head wounds, whereas Boll's (1974) study also included people with cerebro-vascular disease. 8011 suggested this as a possible confounding factor, along with the complexity of the tasks involved and the precise location of the lesions. These studies illustrate some of the problems inherent in interpreting research with neurologically atypical subjects. Semmes (1965) reported that brain-damaged individuals with astereognosis (same sample of veterans in 1960 study) sometimes showed impairment of tactual shape discrimination without deficits in tactual acuity, sensitivity, or localization or in discrimination of texture or size. She felt that a general spatial factor entered into those performances which require discrimination of spatial arrange- ments--regard1ess of modality. This explanation was bolstered by the finding, in a second part of the experiment, that subjects who did well on a test of visual-spatial orientation (walking through paths represented on maps) discriminated shapes tactually as well as did controls, while those who did poorly on the test were also impaired in discriminating shapes. DeRenzi and Scotti (1969) had brain-damaged subjects palpate the edges of a wooden geometrical shape and then visually identify its equivalent among an array of five other shapes. Reaction time was used as the dependent measure. Subjects with right-hemisphere damage were significantly slower than those with lesions on the left side. Fontenot and Benton (1971) gave subjects with unilateral brain damage a task in which they were required to specify the direction of moving tactile stimulation applied to the palms of their hands. Patients with lesions of the right hemisphere had significantly lower scores with both hands than control subjects without cerebral disease, whereas the subjects with left-hemisphere damage showed significant deficits (relative to the control group) only in the right hand. The results were seen as providing support for the role of the right hemi- shpere in the mediation of spatial perception in the tactile modality. Similar conclusions were reached in studies of patients who had undergone cerebral commisurotomies. Levy-Agresti and Sperry (1968) asked these patients to match three-dimensional forms held in the left or right hand with drawings of these shapes as they would appear if they had been unfolded and placed onto a flat surface. Left hand performance was "much superior" to the right, and the authors reported that the method used by the subject was different depending upon which hemisphere was being used. This was inferred from the degree of difficulty the subject experienced with different types of prob- lems. Levy-Agresti and Sperry concluded that the left hemisphere is specialized for "logical, analytic computer-like processing" and the "minor" hemisphere excels at Gestalt, wholistic perception. Franco and Sperry (1977) presented a similar kind of visuo- tactile matching test [as used by Levi-Agresti and Sperry (1968)] to commisurotomized patients, but varied the stimuli to include geomet- ric discriminations in Euclidean, affine, projective, and topological space. Left hand superiority depended on the type of stimulus being felt. The two hands were nearly equal with Euclidean shapes, for example, whereas the greatest left-hand advantage appeared for the topological forms. Milner and Taylor (1972) gave commisurotomized patients a delayed match-to-sample task involving nonsense-shaped wire figures and found clear left-hand superiority in six out of seven subjects. Milner and Taylor viewed this as a demonstration of right hemisphere specialization for the perception and recall of spatial patterns. Control subjects with cortical lesions but intact commissures performed at a superior level with both hands, indicating that both hemispheres are necessary for such tasks, with the right having a predominant role. Studies with Normal Adults There are problems in generalizing from clinical populations to normal individuals, For example, the injury or disease producing the lesion may well have a wide range of effects besides impairing the function associated with the cortical area of interest. "Split brain” patients, for instance, not only undergo surgery to sever the corpus callosum, but usually have a history of very severe Grand Mal epi- lepsy as well. Thus, it is of obvious value to use neurologically intact individuals in this area of research. A relatively small number of studies have been conducted to determine hand asymmetry in tactile perception in normal subjects. Gardner (1942) had adults sort corks of varying sizes and shapes using one hand at a time. There was a modest right-hand advantage when size was the criterion. Gardner also arranged letters made of cord upon a cardboard background and tested for manual asymmetries in speed of reading with the fingers. He found a left-hand advantage, suggesting that the letters were being processed as spatial stimuli. Nebes (1971b) had college students palpate an arc and then esti- mate the size of the complete circle from which it came. He found no significant differences between the hands for college students, though in an earlier study, commisurotomized individuals had shown a left-hand advantage for this task (Nebes, 1971a). Benton, Levin, and Varney (1973) found a left-hand superiority for the perception of the direction of tactile stimulation applied to the palms of the hands. This stimulation was of brief duration and consisted of lines very close in orientation. In a similar study which suggested that the complexity of the task is a crucial factor in some cases, Umilta et a1. (1974) presented subjects with a visual equivalent to this line orientation task on a tachistoscope, but at several levels of difficulty. Left-visual fieldsuperiority was exhib- ited only in the most difficult task in which the slopes of the lines differed very little from one another. This difficulty factor may help explain why manual asymmetries in tactile perception often do not show up with normal subjects when one hand is tested at a time. Using the same task as Benton, Levin, and Varney (1973), Varney and Benton (1975) showed that the presumed right-hemisphere role in spatial perception in the tactile modality does not hold for left- handers and that familial handedness is a significant and independent factor. They found that right-handers having no familial sinistrality (FS-) performed significantly better with the left hand, whereas right—handers with familial sinistrality (FS+) showed no difference in performance between the hands. As for left-handers, FS- subjects performed at equal levels with the two hands whereas FS+ subjects performed significantly better with the right hand. A high percent- age of right-handed subjects, however, did not conform to the pattern of a left-hand advantage (34%), and familial handedness did not fully account for these deviations. lO Developmental Studies: Braille There have been a few developmental studies of hemispheric spe- cialization for tactual perception, the majority being studies of manual differences in reading braille letters. As early as 50 years ago there was speculation as to which hand is superior for braille reading in blind people (Smith, 1929). There was no universal agreement as to which hand was "the eye of the blind” in these early and, for the most part, inadequately reported studies. Hermelin and O'Connor (1971) rediscovered this avenue of research and found that right-handed blind children were faster and more accurate when reading braille with the left hand than with the right. They had subjects use their middle fingers in order to reduce the effects of practice, since the index finger is predominantly used by the blind for braille. With adults, no differences in speed were found, but fewer errors were made with the left hand. Even though the stimuli were linguistic in this instance, Hermelin and O'Connor reasoned from the fact of left-hand superiority that the braille characters were treated as spatial configurations which must first be analyzed by the right hemisphere prior to being analyzed by the left hemisphere. The use of the less practiced middle finger did not eliminate the possibility that the left-hand advantage resulted from greater practice with this hand. Therefore, at least three studies were car- ried out using naive, sighted individuals to circumvent this compli- cating factor. ll Rudel, Denckla, and Spalten (1974) trained right-handed, sighted 7- to 14-year old children to read braille, and discovered a left- hand superiority by age 11 in boys, but only by age 13 in girls. Boys younger than 11 showed no hand differences while girls age 7-8 showed a right-hand advantage. The authors suggested from this latter result that girls, at least at this age, employ processes associated with the left hemisphere (e.g. verbal, analytical) for tactile discim- inations. Feinberg and Harris (1975; cited in Wagner and Harris, 1976) conducted a braille study with sighted adults and found small left- hand superiorities, but no sex differences in magnitude. A possible sex-difference in strategy, however, was indicated by the finding that females who showed a large right-hand superiority tended to have higher overall learning scores, while the reverse was true for males. Harris, Wagner, and Wilkinson (1976) found left-hand superior- ities in sighted 8-13 year-old children as well as in college stu- dents. Among the eight year-olds, left-hand superiority was more marked for boys than for girls, but no other sex differences were found. Subsequently, a re-analysis of these data failed to confirm even this sex difference (Wagner & Harris, 1979). Braille studies, then, generally disclose a left-hand advantage in recognition of these stimuli, but the sex difference found by Rudel et al. (1974) has not been found by others. 12 Developmental Studies: Dichhaptic Perception Task Braille configurations are not purely spatial stimuli, since they comprise an alphabet. "Dichhaptic" presentation of nonsense shapes represents a more direct attempt to test lateral specialization for nonverbal, spatial, tactual functions. Witelson (1975) devised this tactile version of the dichotic listening technique in which subjects felt two different nonsense shapes simultaneously with the two hands. (The objects were hidden from view.) Subjects then were asked to identify the stimuli from among several similar shapes in a visual display. Witelson assumed [by generalization from the structural (Kimura's) model for dichotic listening] that having the subjects process objects with both hands at the same time would create a competition between the hemispheres, thereby maximizing input to the contralateral hemisphere. The depen- dent measures were accuracy and frequency of first response in inden- tifying the objects. Witelson does not explain why she used frequency of first response as a dependent measure. Perhaps it was because a first response measure would be less likely to reflect the effects of an attention set. The subjects were normal, right-handed boys and girls, ages 5-13 years. The results showed sex and age differences as well as a sex x age interaction. For boys there was a significant (p<.05) left-hand advantage at ages 6-7 and 10-11. At ages 5 and 12-13 there was a left-hand advantage significant at the p=.10 level but no difference 13 in accuracy between the hands for the 3 and 4 year-olds. On first response, boys showed a left-hand advantage (p<.05) at ages 5-11. No differences were found at ages 3, 4, and 12-13 for this measure. For the girls there were no significant hand differences except for a right-hand advantage for the 4-year-olds. (Rudel et al., 1974, also found a right-hand advantage in one of their groups of girls, but the children in this case were somewhat older than Witelson's 4-year- olds.) The oldest girls in Witelson's (1975) study showed a nonsig- nificant left—hand advantage. In an earlier study, this time with boys only, Witelson (1974) found a significant (p=.05) left-hand advantage in accuracy for non— sense shapes for children at ages 6.4-7.4, 9.5-ll.4, and 12.0-14.3. These were significant differences despite small sample sizes (N's of 7, 5, and 14, respectively). Witelson gave the same boys a test of tactile perception of letter shapes and failed to find hand dif- ferences at any age tested (6-14 years). The procedure in Witelson's studies was to have the subject identify the correct stimuli in the display by pointing with the hand earlier designated (during pretest trials) by the experimenter as the response hand. When a right hand response was called for, there were no significant differences between the two hands, but with a left-hand response, identification of objects perceived with the left hand was better than with the right. Witelson (1974) suggested that instruc- tions to respond with one hand or the other focuses attention on the corresponding perceptual field so that instructions to respond with 14 the left hand could have enhanced right-hemisphere processing for the task, whereas use of the right hand would have a balancing effect by enhancing left-hemisphere activation. LaBreche, Manning, Goble and Markman (1977) used Witelson's dich- haptic task to measure hemispheric asymmetry in congenitally and pro- foundly deaf children whose average age was 15 years. The authors reasoned that because the congenitally deaf depend greatly on non- verbal learning functions and employ linguistic systems that are visuo-spatial in nature, deaf individuals may have a greater degree of bilateral representation of spatial functions than hearing people. They do not discuss the more plausible possibility that cerebral organization is the same, but that there might be differences in strategies employed by deaf and hearing people. Therefore, it was thought that the deaf children would show either a left-hand advan- tage or equal performance with both hands. The results did not con- firm these expectations. The lS-year-olds showed no significant difference between the hands, what difference there was being in favor of the right hand. What is more, a comparison group of hear- ing children (average age 17 years) showed a significant rightrhand advantage. The authors suggested that verbal mediation might have occurred (particularly since the subjects were older than Witelson's, 1975) in trying to explain these results. They concluded that, at least within the tactual modality, the cerebral organization of con- genitally deaf and hearing individuals is not differentially influ- enced by such ever differences in experience as might be associated with deafness or normal hearing. 15 Cioffi and Kandel (1979) presented seven-, nine-, 11-, and 12- year-old children with tactile stimuli of three kinds: words, bigrams, and nonesense shapes. The stimuli were presented simultaneously to the two hands as in the Witelson testing procedure. The result was that both boys and girls identified nonsense shapes better with the left hand and words with the right hand. A right-hand advantage was found among girls and a left-hand advantage among boys for recognition of the bigrams. The authors inferred that the bigrams were generally processed by boys as shapes whereas girls tended to process them as words. So, Witelson's (1975) finding of a sex difference did not show up on the nonsense-shapes task, but there was a sex difference in processing of the bigrams. This latter result is in accord with the view that cognitive processes are organized differently in boys and girls. Flanery' and Balling (1979) repeated Witelson's (1975) study with first, third, and fifth graders and adults. In addition to using Witelson's dichhaptic procedure (two hands simultaneously), they had a second condition in which subjects felt a nonsense figure with a single hand. The fifth graders and adults were more accurate in identifying objects presented to the left hand, but no hand differ- ence appeared for the younger groups. Unlike Witelson, they found no sex differences. The two procedures, dichhaptic and single hand, yielded similar results. Cranney and Ashton (1980) administered Witelson‘s dichhaptic task to deaf children, as well as to hearing children and adults. 16 They found no significant differences between the left-hand and right- hand scores of any of the groups. In all of the groups, except the younger group of deaf children, the average right-hand score was greater than the average left-hand score. Finally, Dawson, Farrow and Dawson (1980) tested first and sixth graders and undergraduates on the dichhaptic task, using Witelson's (1975) nonverbal stimuli, presented in four different orientations. Shapes were identified by having subjects call out the number of the labeled choice stimulus. Like Witelson (1975), they found a signifi- cant sex x tactual field interaction. Ten-year-old boys, but not girls, tended to have higher scores for the left hand. The six dichhaptic studies reviewed here present an inconsistent pattern of findings. Most, though not all of these studies, show a left-hand advantage for perception of nonsense shapes and among these two show such an advantage for boys and not girls. The Development of Cerebral Lateralization: Theory It generally has been assumed that at birth, the two cerebral hemispheres have equal potential for the sub-serving of cogni- tive functions. A correlary to this view was that lateralization is absent or largely absent during infancy and develops gradually with the development of language (Lenneberg, 1967). This viewwas based partly on reports that damage to the left hemisphere in childhood produces less severe and more transient language deficits than similar injury during adulthood (Lenneberg, 1967). It has been thought that the earlier the age at which the 17 injury is suffered, the less severe is its effect upon language development. Part of the problem with such evidence is that the kinds of brain damage suffered by children are usually not equiva- lent to the adult cases and most researchers have used adult aphasia symptoms as the basis for comparison of cerebral injury in children and adults. Thus, a valid comparison is not made. Further evidence for bilateral language representation in child- ren comes from reports of right hemisphere damage producing language deficits in pre-school children (Lenneberg, 1967). Until very recently, most investigators have agreed that hand- edness was not clearly or permanently established in early childhood and this was thought to reflect incomplete lateralization in the first several years of life. A more current view is that cerebral lateralization is present from birth. That early left hemisphere insult produces less severe aphasias than later damage may reflect the greater plasticity of the less complex, immature brain rather than less complete lateralization. Moreover, it is by no means certain that age, per se, is related to greater degree of recovery from central nervous system injury (St. James-Roberts, 1979). It is more likely that age interacts with a constellation of other factors in influencing recovery from brain damage. Reports of right hemisphere damage causing aphasia in children (Lenneberg, 1967) have been criticized on several grounds (Kinsbourne, 1975). For example, the damage often may not have been limited to 18 the right hemisphere and language disruption was not always reliably reported. In any case, the hemispheric equipotentiality theory does not adequately account for the greater disruption of language develop- ment caused by left-hemisphere injury in infancy as compared to right- hemisphere injury. Furthermore, for tests such as dichotic listening and tachisto- sc0pic presentations of visual stimuli, the magnitude of the asym- metries found has generally not been found to increase as a function of age. Evidence for early cerebral lateralization comes from a number of other sources as well. Turkewitz (1976) found that infants turn towards visual, auditory, and tactile stimuli presented to the right side of their perceptual field more frequently than stimuli presented to the left side. Caplan and Kinsbourne (1976) and Hawn (1978) reported that infants as young as two to three months of age hold a rattle or wood bar-bell in the right hand for a longer period of time than in the left. Previous studies which reported shifts in handedness in infants (Gesell & Ames, 1947, for example) may have involved unreliable or age-inappropriate techniques for measuring predisposition to handed- ness in young children. Corroborating evidence for lateralization in infancy also comes from studies showing that verbal stimuli produce greater electrical cortical responses from the left hemisphere of infants than the right 19 (Molfese, 1973) and ear differences in dishabituation to dichotically- presented sounds (Entus, 1975; Glanville, Best, & Levenson, 1977). If cerebral lateralization is present from birth, how do pro- ponents of this view account for findings, such as Witelson's (1974), that boys show a left-hand advantage on the dichhaptic task at age 6 but girls do not? Harris and Witelson (1977) note a growing ten- dency among psychologists to interpret all such tasks as absolute measures of the degree of cerebral specialization for the spatial or verbal function that the test is presumed to represent. These tasks, however, may not simply index changes in lateral organization, but reflect changes in the types of processes individuals will invoke to solve particular problems. Laterality may be more or less con- stant during childhood--cognitive competence for various types of skills is not. The results of different tests of laterality also depend importantly on the unique qualities of the sensory mode involved. The haptic sense, for example, may be inherently more spatial in operation than other senses because of its limitation in taking in information, thus requiring a temporal-spatial mode of analysis best served by right-hemisphere systems. Sex.Differences The inconsistent reports of sex differences in the tactile per- ception studies behoove us to take a closer look at the topic of sex differences in cerebral organization. 20 Sex Differences in Cognitive Abilities Women generally exhibit better performance than men on a variety of language-oriented or predominantly left-hemisphere tasks, whereas men excel at visual-spatial skills, skills usually linked to the right hemisphere. Women are, on the average, superior to men in areas of verbal skill such as speed of articulation, fluency, and grammar. Throughout childhood, girls reach various milestones of linguistic development ahead of boys and reveal fewer problems with and better average skill in reading (Harris, 1977). Women are also better at fine motor coordination, including precise temporal regu- lation of motor repetitions (finger tapping, Wolff & Hurwitz, 1975), and rapid processing of detailed perceptual information (Maccoby & Jacklin, 1974). Males outperform females on such spatial tests as visual and tactual mazes, the rod-and-frame test, and map-reading, as well as recall and detection of shapes and mental rotations (see Harris, 1978, for review). Broverman, Klaiber, Kobayashi and Vogel (1968) propose that the type of tasks males excel at are those requiring inhibition of immediate response and they have developed a hormonal theory to explain the sex differences along this activation-inhibition dimension. Sex Differences in the Development and Degree of Lateralization Several theories have been advanced to account for sex differ- ences in performance on cognitive and perceptual tests. A brief 21 discussion of the more plausible of these theories (after Harris, 1978) follows. Bilateral language_representation in females. One explanation for the well-established pattern of sex differences in cognitive skills posits that language functions are lateralized to a lesser extent in females than in males. The supposition in this theory seems to be that spatial skills suffer by the intrusion of linguistic pro- cesses into the right hemisphere. Language function in the right hemisphere, according to this view, may interfere with the supposed diffuse organization of that hemisphere which Semmes (1968) saw as essential to spatial analysis. Characterization of the right hemi- sphere as more diffuse in organization than the left is suggested by studies showing that impairment on tests of spatial orientation is related to locus of injury only in the left hemisphere and not in the right (Semmes, Weinstein, Ghent, & Teuber, 1960). If this theory is valid, then other groups known to be less well lateralized with respect to language should also show deficits in spatial ability. Left-handers are such a group. Levy (1969) tested male graduate students on the WAIS and found that the left-handers performed significantly worse on the performance scale than the right-handers, but approximately equal on the verbal part. In Nebes' (1971b) study with college students, left-handers had more difficulty than right-handers matching a segment of an arc to the circle to which it belonged. 22 More recent studies have not supported these early findings of a spatial deficit in left-handers. Briggs, Nebes, and Kinsbourne (1976) found that left and mixed handed college students achieved lower full-scale 1.0. scores (WAIS) on the average than right- handers. The difference in average scores was small but significant. Unlike Levy (1969), they did not find that left handed individuals do worse on the performance part of the WAIS than right handers but not on the verbal part. Heim and Watts (1976) gave verbal, numerical, and perceptual tests to a large sample (N=2165) of children and adults. The authors meant to improve on methodological flaws in Levy's (1969) study, including the small, select group of subjects used (graduate science students), and the "cognitive dichotomy" set up between performance (visuo-spatial) and verbal (including numeri- cal) test groupings. Heim and Watts (1976) found no evidence that left-handers perform less well on tests of visuo-spatial skill. Others (Newcomb & Ratcliff, 1973; Fagan-Dubin, 1974; Hardyck, Petro- vich, & Goldman, 1976; Hardyck, 1977) have also found no deficit in left-handers on spatial or performance tests. Some investigators have looked for differences in spatial skill as a function of strength of lateralization among right-handers only. Zoccolotti and Oltman (1978) investigated the relationship between degree of lateralization and spatial ability in right-handed men, ages 18-30. Men who performed well on embedded figures tests and the rod-and-frame test showed right hemi-field superiorities in recognition of tachistoscopically presented letters, but those who had poorer scores on the spatial tests did not show visual field 23 asymmetries. Among right-handers, Kail and Siegel (1978) found a positive correlation between strength of handedness and more accurate recall of spatial information in the right visual field for women, but a negative correlation for men. It seems, then, that the relationship between degree of laterali- zation and spatial skill is not firmly established. Therefore, this avenue of research does not unequivocably support the theory of bilat- eral language representation in females as an explanation of sex dif- ferences in cognition. Some studies with neurological patients can be interpreted as lending support to this theory. Lansdell (1961) found that women's performance on a proverb comprehension test did not change after left- hemisphere surgery, but men's scores declined. In a second study, Lansdell (1962) administered the Graves Design Test to men and women before and after surgery to remove the right or left temporal lobe (for purposes of relief of epilepsy). The Graves Design Test was designed to measure art appreciation but has a strong spatial compo- nent. Among those who underwent left-hemisphere surgery, men's scores rose and women's declined. For those who had the right temporal lobe removed, men's scores declined, women's rose. McGlone (1978) tested right-handed patients with unilateral brain damage on the WAIS. Males showed a decrement on the verbal sub- tests after left-hemisphere damage and lower scores on the perfor- mance subtests after right-hemisphere injury. In females, verbal and performance 1.0. scores were not significantly different after left or right hemisphere injury. 24 In dichotic listening studies, right-handed males have shown a greater right-ear superiority than right-handed women (Lake & Bry- den, 1976). Men also display a stronger right-field superiority for verbal stimuli presented in a tachistoscope (Kimura, 1969; Kail & Siegel, 1978). Earlier right-hemisphere specialization in males. The previ- ously mentioned studies by Rudel et a1. (1974) and Witelson (1975) suggest that boys may develop earlier right-hemisphere specialization for spatial functions than girls. Studies measuring evoked potential recordings from the two hemi- spheres (Molfese, 1973) and dishabituation to dichotically presented stimuli (Entus, 1975; Glanville, Best & Levenson, 1977) have indi- cated right-hemisphere specialization for the perception of nonverbal sounds in infants. No sex differences were reported in these studies. Shucard et a1. (1979), however, measured evoked auditory potentials from three-month-old infants while musical and verbal stiumli were played and did find a sex difference. With both types of sounds, seven of eight boys showed a higher amplitude of right than left response. The pattern was reversed in girls, among whom seven of eight produced higher amplitude responses from the left than the right hemisphere. The few studies which have compared boys and girls on asymmetries in visual half-field or dichotic presentations of nonverbal sounds have found no significant differences. Marcel and Rajan (1975), for example, reported no sex difference for recognition of faces in seven- 25 to nine-year-old children. Piazza (1975) presented environmental sounds dichotically to three- to five-year-old children and found a similar left-ear advantage for boys and girls. Research with monkeys indicates that man may not be the only primate in whom males and females follow different developmental schedules for higher nervous system organization. Goldman, Crawford, and Stokes (1974) studied the performance of Rhesus monkeys with bilateral, prefrontal lesions on an object discrimination reversal task and on spatial, delayed-response problems. Males were impaired on these behavioral tests at 2% months of age, but females did not show comparable impairment until 15-18 months of age. Strategy_differences. An alternative theory to the ones which propose differences in lateralization is that females rely primarily upon linguistic modes of processing whereas males prefer right- hemisphere strategies. Of course, these different theories are not mutually exclusive--a greater degree or earlier onset of laterali- zation may predispose an individual to favor one hemisphere (cogni- tive process) over another. Many of the aforementioned studies, cited in support of a neu- rological model, also can be interpreted to conform with the strat- egy theory. This is particularly true of some of the braille studies, since braille configurations apparently can be treated as either spatial or verbal stimuli. In the study by Rudel et a1. (1974), for example, the right-hand advantage for the youngest 26 girls could mean that they are processing the braille characters using linguistic skills. The Lansdell (1962) experiment also employed stimuli which are readily processed by either verbal or spatial means. The drop in score among women who had had left-hemisphere surgery could mean that they depend more on this hemisphere (language) to make aes— thetic judgments of art, and that males use more purely spatial means of analysis. Mellone (1944) gave seven-year-old children a variety of osten- sibly spatial tests, such as block counting and identifying the mirror image of a picture. Factor analyses suggested that the girls were drawing upon verbal capacities to solve the same problems that the boys were processing (sometimes more efficiently) spatially. In summary, there is evidence to support theories of differences in the degree and ontogeny of neurological organization as well as differences in preferred strategies as underlying the observed sex differences in cognitive abilities. Place of Prgposed Study in the Literature The purpose of the present study was to provide additional evidence bearing on the question of hand differences in tactile perception of spatial stimuli in normal children of different ages. More specifically, this study was intended to clarify some of the conflicting results obtained by Witelson (1975) and other researchers using the dichhaptic perception task, especially with regard to the question of age and sex differences. 27 Four of the studies using the dichhaptic procedure were pub- lished only after the present research began (Flanery & Balling, 1979; Cioffi & Kandel, 1979; Cranney & Ashton, 1980; Dawson et a1. 1980). The relationships among these studies and the present one and the implications of their disparate findings will be analyzed in the Discussion section. At the time the current experiment was designed, only Witelson's (1975) and one other study had been done, and the main concern of the current study therefore was to use stimuli that were less verbalizable than Witelson's (1975) and to eliminate the specification of the response hand to the subject from the procedure. METHOD Subjects The subjects were first, fifth and ninth graders from a public elementary and high school in a local suburban area. There were 10 first-grade boys and 11 girls, 8 fifth-grade boys and 13 girls, and 9 ninth-grade boys and 10 girls. The average age for the first-grade boys was 6:8 and for the first-grade girls, 7:0. For the fifth- grade boys and girls, average ages were 11:1 and 10:9, respectively, and for the ninth-graders 15:1 for the boys and 14:9 for the girls. Only right-handed subjects were included in the study. Handed- ness was determined by asking the children to perform five common actions (Annett, 1970) and noting which hand was used. Only subjects who reported using their right hand for at least four out of five of the tasks were included in the study. Informed consent forms were given to the parents of prospective participants along with a statement explaining the purposes and pro- cedures of the study. (See Appendix A.) Stimuli The stimuli consisted of irregularly shaped styrofoam forms, approximately 1% x 1% x 3/16 inches in size. Pairs of stimuli, spaced 4" apart, were glued to an 8" x 10" cardboard backing. Witel- son stated that her stimuli were sufficiently meaningless and unfamil- iar so as to make verbal mediation almost impossible (see Figure 1). 28 29 PROD! ‘ — O W 9 Figure l.--The 5 pairs of nonsense shapes used by Witelson (l974). 30 Some of the forms, however, were quite familiar (e.g. an oval) and some of the pairs differed in terms of features that could be easily encoded verbally. Therefore, for the current study, a set of shapes was designed so as to minimize the chance for linguistic processing, thereby maximizing the likelihood of gestalt perception. For the seven- and ll-year olds, the same stimuli were used (see Figure 2), but two new pairs of stimuli were substituted for two of the old ones with the 15-year-olds (see Figure 3). This was done after pre-testing suggested that the forms used with the younger children would be too easy for the lS-year-olds. Procedure Subjects were asked to feel a stimulus with each hand, simultane- ously, for 10 seconds. The nonsense shapes were concealed from the subject's view behind a box-like construction (see Figure 4). In light of evidence that there is no ipsilateral control for fine move- ments of the digits (Brinkman & Kuypers, 1972), the subjects' arms were contrained by placing them through the holes pictured in the testing apparatus to prevent gross movements which could produce ipsilateral feedback. Also, subjects were asked to use only the mid- dle and index fingers of each hand. The experimenter demonstrated the procedure to the child. Immediately after the lO-second exploration period, subjects were asked to point to the correct stimuli from a visual display. The forms in the display were arranged randomly, and the positions of the correct stimuli varied randomly from trial to trial. 31 thlfi GGRQV Figure 2.--Stimuli used in the present study for the nine- and 11- year-old children. Figure 3.--Stimuli used in the present study, for lS-year-olds. 33 Figure 4.--The testing situation for the two-hand tactile perception task, devised by Witelson (1974). 34 There were five choice stimuli in the display, the two correct ones and three distractors. No time limit was given for each response, but a child who did not know the correct stimulus for a particular hand was asked to guess. As in Witelson's (1974) study, knowledge of results was not given. Earlier I suggested that in Witelson's (1974) study, the required use of a particular hand as the response hand may have created a bias in attention to the corresponding perceptual field, resulting in enhancement of left hemisphere processing in the case of a right-hand response or right hemisphere processing for a left-hand response. Witelson (1974) reported that although all her subjects were right- handed, when they were instructed to use the right hand, they often spontaneously tried to use the left. In the present study it was thought preferable to let an intitial perceptual asymmetry influence motor output, rather than to let the influence work in the reverse direction (as Witelson's procedure seems to have done). Therefore, subjects were not restrained in the use of one hand or the other as the response hand in the current experiment. Each subject was given 10 trials. Each stimulus pair was pre- sented twice, once to the left hand and once to the right. The stimulus pairs were’ counterbalanced so that half the subjects were given one set of stimuli to the left hand for the first five trials, with the presentation reversed for the others. Subjects were given 10 practice trials to teach them to explore the two shapes simultane- ously with just two fingers of each hand while keeping the wrists 35 immobile. The stimuli used in the practice trials differed from those used during the main part of the study. Knowledge of results was provided. RESULTS Comparison of Left and Right Hand Scores The average scores for left and right hand for the dichhaptic task, over 10 trials, are summarized in Table 1. For all groups Table 1: Mean Number of Correct Right and Left Hand Scores and T-tests for Each Age X Sex Group and All Groups Combined Males Females Age Left Right Left Right 7 7.10 6.40 6.09 5.50 t L L > R R > L L > R 7 4 5 3 6 11 5 4 4 5 15 4 3 l 7 Analysis of Group Effects Simple Difference Between the Hand Scores A multivariate ANOVA was performed using overall difference scores between the hands and difference between the hands in first response as dependent measures. These were the same measures used 38 .mupoucmmxlcm>mm Low mmmcoammc u:m:-u;mwc msmcm> -ucmp we uo—nncmuumumlu.m meamwg mcoum ccmzlummA mmpmsmd m m u o q1 d a . .0 n. a. .0 mmpmz m w m o m e m - n 4 d d d 4 op 94095 puvH-iufila 39 q .muFoucowxl—F com mmmcoammc ucmg-u:mwc mzmcm> Hemp we poFaucmuumum--.m meam_d mmoum oz<= ham; mmpcemu mmsz m m A m m q a q q i! - J \ .I n m 34035 PUPH-lufilfl 4O .mupolcmmxlmP Lee mmmcoammc ucmsupcmwc mzmcm> upmmp co popalcmpumumlu.N mgamwu maoum oz.lO) In summary, when phi coefficient scores were used, the signifi- cant main effects of Sex, Order and the interaction of Age and Sex found for the simple difference scores (when difference between the hands scores was the dependent variable), did not appear. In fact, for phi coefficient scores there were no significant effects at the .05 level. However, marginally significant (p<.10) interactions of Sex x Order, Age x Block and Age x Sex x Block were found for the dependent variable of hand differences in initial response, whereas there were no significant main or interaction effects for this second dependent variable when simple difference scores were analyzed. Scores Summed Across Hands Overall accuracy in hand scores did not differ for boys and girls (t R. Hand .5 T C) CD R.H. L H c: I 1 I 1 I 1 1 1 1 I I 1 1 I I 1 I I 1 I 1 1 I 1 I 1 C3 1 1 I I Cf. I I I 1 I 1 R. Hand > L. Hand [I 1. 1 1 1 I 1 1 1 1 I, .1 l 1 ”6789101112131415161718 Total Score (LH + RH) Figure 8.--Relationship between difference between-the-hands scores and total scores in 7-year-olds. L. Hand > R. Hand R.H. L.H. 52 R. Hand > L. Hand Boys 0 " 0 Girls 0 1- o L (3 ‘. - (J o 8 .......................... 0----0---3--- _ C) (D t 8 h- . 0 .fl 1 1 n, 1, 1 l 1 1 1 1 l 1.44 I7 6 7 8 9 10 ll 12 13 14 15 16 17 18 Total Score (LH + RH) Figure 9.--Re1ationship between difference between-the-hands scores and total scores in ll—year-olds. L. Hand > R. Hand = RH LH R. Hand > L. Hand 53 Boys 0 Girls C’ [I 1 1 1 J 1 L, I .1, 1 _1 1 l 1 ”6 7 8 9‘101112131415161718 Total Score (LH + RH) Figure lO.--Re1ationship between difference between-the-hands scores and total scores in 15-year-olds. DISCUSSION As we saw in the Introduction, since the dichhaptic perception task was first devised, several researchers have used this procedure with widely varying results. Some have found a significant left-hand advantage for both males and females (Cioffi & Kandel, 1979; Flanery & Balling, 1979); others have found a significant right-hand advan- tage for both sexes (LaBreche et al., 1977); still others have found a left-hand advantage for males, but not for some of the female groups (Witelson, 1975; Dawson et al., 1980); and one‘failed to find any significant hand differences (Cranney & Ashton, 1980). The results of the current study do not seem to fit clearly with any of the others. The finding of no significant differences between average left and right hand scores (for all groups except one) in the present study, like the results of the Cranney and Ashton (1980) study, obviously are inconsistent with the original Witelson (1974) findings. The Cranney and Ashton (1980) scores, however, were in the direction of a right-hand advantage whereas in the present study, the average left hand score exceeded the right hand score in all groups except the oldest boys. If these left-hand advantages had reached significance, they would not be inconsistent with the results of the two groups who found overall left-hand superiority regardless of sex. In any case, the current results do not support either Witelson's (1975) or Dawson et a1.'s (1980) more recent finding of a sex 54 55 difference in hand asymmetries on this task. For the seven- and 11- year-old groups, both boys and girls had greater left than right hand scores (though the difference was not significant). For the lS-year- olds, there was a sex difference, but it was the reverse of the direction expected. In this group girls did better with the left hand (the greatest and only significant hand score difference of any group) whereas the boys did better with the right. As already men- tioned, this oldest age group probably accounts for most of the sig- nificant Sex effect in the ANOVA for simple difference scores. It is difficult to explain this pattern of results among the lS-year-olds in light of current ideas pertaining to the earlier maturation of and preference for right hemisphere modes of processing in males. The boys' right-hand advantage was slight (.34) and not significant (t “even do ma>p mmmcm>< Aa~.v __wu\m Aea.v MNHm __mo\N_ FPmU\o_ P_m6\o_ Fpmo\o_ __mo\m —FmU\o~ Fpm8\mm mgazzoz Aaampv Amampvcougmq Aomm_v Acwa_v_aeea¥ Amampvmce_Pam Aomm_v Aoum_v .Fm um msomcmmo use moccmcu Nacmcmsoa use memowu acmcwpd .Fm pm comzmo comFmpwz cmzugmmmmm mmocmcmcmwo xmm oz mocmcmmmwo xmm mmmucw>n< mmucmcmeewo new: use?“ new: “cauwcwcmwm oz amapcm>e< age: “can xmmp cowpamucma owpqm;;u_o mg“ mcwmz wmwozum cw mppsmmm ucm mmcaumuoca Go commcmasoo ”Pp mpnmh 57 Sample Size First, the absence of significant hand differences for most groups in the present study and in Cranney and Ashton's (1980) study, in contrast to other studies, might be related to differenceS'h1samp1e size. The former two studies used 10 subjects per cell, whereas three studies finding significant differences had 19 or more (Dawson et a1. (1980)--20/ce11; Witelson (1975)--25/ce11; Cioffi and Kandel (1979)--19/ce11). Other researchers, however, found significant dif- ferences with sample sizes comparable to those used in the present study (Flanery & Balling (1979)--8/ce11), so differences in sample size is a partial explanation at best. Task Difficulty The possibility that the task used in the current experiment was too difficult for some or many of the children is unlikely, since average scores for each hand over 10 trials ranged from 5.4 to 7.38, and these are significantly greater than the score expected by chance (=2). It seems equally unlikely that the task was too easy, since the shapes were generally more complex (had more sides) than Witelson's and Witelson's stimuli were used in the LaBreche et a1. (1977) and Cioffi and Kandel (1979) studies in which significant hand asymmetries were found. It also may be that the use of five choice stimuli in the current study made the task easier than the other studies, most of which used six choice stimuli. However, the scores for each hand (for 10 trials) in most of the other investigations were similar in magnitude to those in the current study. Witelson's (1974) scores, 58 for example, ranged from 4.5 to 7.0 (averages for each group); the range for each hand in the current study was 5.4 to 7.4. Stimulus Attributes The absence of a sex difference in the present study in contrast to the differences found by Witelson (1975) and Dawson et a1. (1980) also might be attributable to differences in the stimuli used. As already noted, although Witelson (1974) designed her stimuli to mini- mize the possibility of verbal mediation occurring, many of her forms seemed easy to label, either in respect to certain features or taken as a whole. The stimuli in the current study were meant to improve upon Witelson's forms by being less easily subject to verbal media- tion--more irregular and with fewer outstanding features. (These forms were designed, for the most part, however, in an intuitive, subjective way, and in at least one case, a subject was heard to say to himself that he was "trying to find the one (stimulus) with the sharp corner.") Flanery and Balling (1979) report having had the same aim in mind with their "randomly generated" figures. In neither their study nor the present one were sex differences found. So it may be that some of the girls in Witelson's (1975) experiment tended to process the forms in a manner that drew strongly on verbal media- tion, and this information-processing strategy eliminated the left- hand advantage for girls. Thus, the use of stimuli which did not lend themselves easily to dual modes of processing may help account for the failure to find a sex difference in the present study as well as in Flanery and Balling's (1979) experiment. 59 As noted previously, overall accuracy in hand scores did not differ for boys and girls (t