- Email: [email protected]
Hemispheric preference and lateral eye movements evoked by bilateral visual stimuli
,‘m,ropsycholog!a. Vol. Prmlsd
m Great
31, No. 12, pp. 1299-1306.
1993 8
Britain
HEMlSPHERIC PREFERENCE AND MOVEMENTS EVOKED BY BILATERAL JIIJLIA WEISZ* Psychophysiology
and GY~RGY
OOZS-3932193 $6.00+0.00 1993 Pergamon Press Ltd
LATERAL EYE VISUAL STIMULI
ADAM
Research Group of the Hungarian Academy of Sciences, Department Physiology, Eiitviis Lorind University. Budapest, Hungary
of Comparative
(Received 25 January 1993: accepted 1 June 1993) Abstract-This study investigated the relationship between the direction of lateral eye shifts evoked by bilateral visual stimulation, on the one hand, and verbal vs visuospatial performance, on the other. In males a negative correlation was found between verbal relative to spatial accuracy and the ratio of leftward eye movements. No such relationship was found in females. Additionally, in males body mass index was smaller for those who displayed a greater ratio of leftward eye movements. It was concluded that visually evoked lateral eye shifts might reflect hemispheric preference, similarly as lateral eye movements obtained during the widely used questioning procedure. However, the procedure used in this study for evoking lateral eye shifts seems to be free of some methodological problems of the traditional procedure and is more easy to standardize.
INTRODUCTION years the direction of lateral eye movements has been widely used as an index of hemisphericity. Although BEAUMONT et al. [2] argued that the concept of hemisphericity lacks adequate foundation, many researchers maintain that there are relatively stable individual differences in the extent to which subjects activate one or the other hemisphere. The asymmetry of hemispheric activation is assumed to be reflected in the preferred mode of cognitive processing, in the difference of electroencephalographic and metabolic measures between the hemispheres, in perceptual asymmetries, and-which is most relevant regarding this study-in the direction of lateral eye movements. TEITELBAUM [22] was the first to observe that when asked a question which requires reflective thought, subjects move their eyes either to the left or to the right. This gaze shift is known as “conjugate lateral eye movement” (CLEM). When the questioner is face-to-face with the subject, the predominant direction of CLEMs was proposed to depend on the subject’s hemispheric preference. It is hypothesized that in subjects moving their eyes to the left (left movers) the right hemisphere is the more active one, and the opposite is true for right movers. In the “experimenter behind” condition CLEMs vary as a function of question type: verbal questions elicit rightward CLEMs, whereas spatial questions elicit leftward CLEMs. That is, CLEMs are assumed to reflect both characteristic and temporary asymmetry of hemispheric activation, depending on the experimental condition (for review, see Ref. [5]). The validity of lateral eye movements as an index of hemisphericity was tested in studies trying to find an association between CLEMs and other measures of hemispheric preference. IN THE LAST 25
*Address for correspondence: Jlilia Weisz, Department H-1088. Mlizeum krt. 4/a., Budapest, Hungary.
of Comparative
1299
Physiology,
EBtviis Lorand
University,
1300
J. WEKZand G. .bh
The most convincing evidence are of neurophysiological nature. Several studies on EEG alpha asymmetry showed that left movers exhibit a more pronounced EEG activation in the right hemisphere, while the opposite is true for right movers [13, 141. Evoked potentials were also found to be of greater amplitude in the hemisphere contralateral to the preferred gaze direction [20]. Cerebral blood flow results were consistent with the electroencephalographic findings: GUR and REIVICH [9] reported that left movers had more flow to the right hemisphere than to the left one, while for right movers there was an, albeit nonsignificant tendency in the opposite direction. Further substantiation of the eye movement-hemispheric activation hypothesis comes from studies showing a close link between cognitive abilities and CLEM direction. Right movers were often reported to outperform left movers at verbal tasks, but being inferior to left movers in solving visuospatial tasks [21,23,27]. However, in many studies of this type no difference in cognition could be demonstrated between the CLEM groups [lo, 151. Another line ofevidence regarding temporary hemispheric activation indicates that verbal processing is accompanied by rightward, and spatial processing is accompanied by leftward eye movements [12]. This effect works also in the reverse direction: voluntary gaze to the right facilitates verbal performance, while gazing to the left improves spatial performance [4, 8, 253. Lateral gaze was also shown to create an asymmetrical EEG activation with the hemisphere contralateral to gaze direction being the more activated one [l]. Taken together, converging evidence support that CLEMs reflect asymmetric hemispheric activation, and in the face-to-face condition the predominant direction of eye movements can be used as an easy-to-assess indicator of hemisphericity. However, there are a number of unresolved methodological issues as regards CLEM research, which were most thoroughly discussed by EHRLICHMAN and WEINBERGER [7], and which may, at least in part, be responsible for some negative findings. One of these problems is that different researchers use very different numbers and sets of questions to determine CLEM directionality. There is also a lack of consensus about how large an eye movement should be to score it as a CLEM, and how to describe the direction of eye movements. It is also not clear whether the subject should fixate centrally during the question, or not. Briefly, in spite of being used for a long time, the CLEM procedure resisted standardization up to now. Thus, it seemed reasonable to make an attempt for the substitution ofCLEM procedure by a method which avoids the aforementioned problems. In a study of PosNER and COHEN [17] two dots were simultaneously presented in the two visual hemifields and the subjects were instructed to move their eyes in the direction that they felt most comfortable. In the present study we assumed that this simple technique could be appropriate for revealing the characteristic tendencies of individuals to make eye movements in one or the other horizontal direction. We were encouraged to make this hypothesis by the observation of Posner and Cohen who noted that “many subjects showed strong biases in one direction or another”, although they did not try to analyse this phenomenon further. In the present paper we refer to this method as the ‘bilateral visual stimulation” (BVS) procedure, and use the terms “left shifters” and “right shifters” for those displaying a preference of leftward and rightward eye movements, respectively, during the BVS procedure. The present experiment aimed at determining whether the predominance of left or right eye shifts in the BVS condition correlated with hemisphericity. To validate the BVS method it seemed advisable to use an indicator of hemisphericity independent of eye movements. As it is widely accepted that the performance on verbal and visuospatial tasks is in close
LATERAL EYE MOVEMENTS
1301
connection with left and right hemispheric functions, respectively, we chose to use a cognitive index characterizing verbal vs spatial performance for assessing hemisphericity. Thus, our primary hypothesis was the right shifters would exhibit better verbal and worse spatial performance than left shifters. A secondary assumption of this study was that the directionality of eye shifts would be related not only to cognition but also to other individual characteristics that are supposed to be in connection with asymmetrical hemispheric activation. Specifically, we expected that obesity would also be correlated with eye shift preference. RODIN and SINGER[lS] reported that obese subjects exhibited more rightward CLEMs than normal weight subjects. This result was supported by us in a previous study where we found that right movers were more obese than left movers [26]. Accordingly, we expected that in the present experiment right shifters would turn out to be more obese than left shifters.
METHOD Subjects Thirty-nine right-handed university students (19 males and 20 females) aged IS-25 years (x= 20.5 years) volunteered to the study, The subjects were classified as right-handed with a four-item handedness questionnaire 116, p. 34); handedness scores could vary between -4 and 4, positive values indicating right-handedness, and negative values indicating left-handedness. Materials Both cognitive tasks used were paper-and-pencil tasks. The written instruction emphasized the importance of both speed and accuracy of performance. The verbal task containing a list of 100 Hungarian words where half of the words had only one meaning whereas the other 50 words had two independent meanings. The order of the one-meaning and two-meaning items was randomized. The subject was required to put a + sign next to each word that he regarded a two-meaning word and to put a - sign next to each one-meaning word. The spatial task was a mental rotation test [24] based on the figures used by SHEPARD and METZLER [19]_ Each item contained a drawing of a three-dimensional criterion object and four highly similar alternative figures. Two of the four alternatives were identical to the criterion in structure but were rotated in space. The task was to identify and mark these two alternatives. This task consisted of 20 items, each with two correct answers. Apparatus and procedure Electrooculogram (EOG) was recorded by a Beckman Type R411 Dynograph. The electrodes were placed at the outer canthi. An AC preamplifier (time constant = 1 set) was used to obtain lateral eye movement information. The presentation of visual stimuli during the BVS procedure was controlled by a computer. The electrooculogram and a computer-generated marker signal simultaneous with the appearance of the visual stimuli were recorded on the chart. At the beginning of the experiment the subject filled out the handedness questionnaire. After that sighting dominance was determined by the point test [16, p. 1971. The experimenter stood facing the subject at a distance of about 3 m and the subject was required to keep both eyes open and to point with an outstretched hand at the experimenter’s nose, which was directly in front of him. The eye aligned with the pointing finger was noted by the experimenter, The point test was administered six times with alternating hands. Scores for sighting dominance can vary from -6 to 6, where negative values represent left eye sighting dominance and positive values reflect right eye sighting dominance. Body mass index (BMI = weight/height’) was determined by measuring the subject’s weight and height 131. After completion of the point test the verbal and the spatial tasks were administered. The order of the two cognitive tasks was counterbalanced across subjects. Each subject had 3 min to work on the verbal task and 5 min to work on the spatial one. After attaching the EOG electrodes the subject was seated in a sound-attenuated experimental chamber and left alone. Throughout the BVS procedure the subject was monitored using a closed-circuit TV system. By using an adjustable chin rest the subject looked at a TV screen located in front of him, from a distance of 50cm. The centre of the screen was at eye-level. Each trial of the BVS procedure began with a warning signal: a small flashing cross appeared in the centre ofthe screen which the subject was to fixate. The warning signal was exposed for 1.5set and at the moment ofits disappearance two small yellow dots appeared on the screen 10.5 cm to the left and to the right of fixation. The duration ofexposurewas 800 msec. The subject was instructed to move his eyes on the
1302
J. WEISZ and G. AD,&
appearance of the dots to the right or to the left, whichever seemed most comfortable. The trials were separated by 5.5-7.5 set intervals. The BVS procedure consisted of two blocks each with 36 trials. Within each block there were six “not simultaneous” trials inserted; in these trials there was a 350 msec time difference between the appearance of the two dots, the right dot and left dot each leading equal times. Every 3-5 “simultaneous” trials were followed by such a “not simultaneous” trial with left dot leading and right dot leading trials alternating. The “not simultaneous” trials were aimed at breaking the subject’s possible looking bias, caused by factors other than hemisphericity. In the remaining trials the two dots appeared simultaneously on the screen. There was a 24 min rest period interposed between the blocks. Prior to the beginning of the first block of trials each subject was given 10 practice trials.
Data analysis
There were enough verbal and spatial items so that none of the subjects finished all within the time allotted them. For the verbal task each correct answer was awarded by one point. For the spatial task an answer was regarded as correct only if both alternatives chosen were correct. For each correct answer one point was given. If only one alternative was chosen and it was correct 0.5 point was given. For each subject two measures of performance were determined for each cognitive task: the number of correct choices (correct score) and the accuracy score. The latter was calculated by dividing the number of correct choices by the number of all items attempted. Correct score depends on both the speed and the accuracy of performance while accuracy score is a pure measure of accuracy. Each subject’s verbal and spatial correct scores were converted to z-scores by subtracting from it the group mean and dividing the resulting difference by the standard deviation. A cognitive index (CI i) was computed by subtracting the spatial from the verbal z-score. This measure characterized each subject’s verbal vs spatial performance. Negative Cl, values indicate relatively better spatial performance and positive values indicate relatively better verbal performance. The same computation was performed on the accuracy scores, resulting in another cognitive index (Cl,) which reflected verbal vs spatial accuracy ofeach subject. Negative CI, values indicate relatively greater accuracy on the spatial task whereas positive values indicate relatively greater accuracy on the verbal task. For the BVS procedure the ratio ofleftward eye shifts (LES) was computed by dividing the number of leftward shifts by the sum of leftward and rightward eye shifts.
RESULTS Cognitive
measures
The mean values of cognitive measures for males and females are presented in Table 1. Correlational analyses showed no linear relationship between CI 1 and C,, either in males, or in females.
Table
1. Mean values of cognitive measures and in females
Male Verbal correct
score
Spatial
score
correct
Verbal accuracy
score
Spatial
score
CI, CI,
accuracy
in males
(S.D.)
38.2 (12.4) 8.2 (3.4) 0.86 (0.05) 0.82 (0.17) -0.46 (1.33) -0.02 (1.02)
Female 38.7 (11.1) 5.6 (2.2) 0.83 (0.07) 0.74 (0.22) 0.44 (1.09) 0.02 (0.95)
LATERAL
EYE MOVEMENTS
I303
In order to examine whether the values of cognitive indices are sex-dependent, separate one-way analyses of variance were conducted on CI, and CI, with sex as the independent variable. It was found that CI, was smaller for males than for females, i.e. males exhibited worse verbal vs spatial performance than females [F (1, 37) = 5.37, P=O.O26]. To determine whether this difference could be attributed to a sex difference in spatial performance or verbal performance, or both, spatial correct scores and verbal correct scores were analysed separately. A sex difference emerged only for spatial correct scores, with males displaying better spatial performance than females [F (1, 37) = 8.46, P=O.O06]. Verbal correct scores did not significantly differ between sexes [F (1, 37)=0.02, P=O.886]. For CI, (verbal accuracy vs spatial accuracy) no sex difference could be observed. The relationship between CI,, on the one hand, and handedness and sighting dominance, on the other was examined by conducting stepwise multiple regression analyses separately for males and females where the dependent variable was CI 1, and the independent variables were handedness and sighting dominance scores. (The criterion for the independent variables to be added to the regression equation was set at P=O.l.) For males it was found that both independent variables have significant influence on CI 1. Partial regression coefficients for the handedness score and the sighting dominance score were 0.52 (P=O.O16) and 0.46 (P=O.O31), respectively, and the two variables explained 32.5% of the total variance of CI,. That is. in males the greater the degree of right-handedness and right-sightedness, the better the verbal vs spatial performance was. No such relationship was found for females. The other cognitive index used, CI,, did not depend on handedness and sighting dominance scores, either in males or in females.
Latera/
eye sh[fts
The mean value of the ratio of leftward eye shifts for males was 0.46 (S.D. =0.21), and for females 0.49 (S.D. = 0.22). The difference between the sexes did not approach significance. Both visual inspection of the LES histogram and statistical analyses indicated that the ratio of leftward eye shifts is approximately normally distributed (kurtosis: -0.213; skewness: 0.009; Kolmogorov-Smirnov Z=O.667, P=O.766). Correlational analyses showed that in males the ratio of leftward eye shifts is in negative connection with CI, (r= -0.5705, n= 19, P=O.Ol l), that is, the greater the ratio of the leftward eye shifts, the smaller the verbal vs spatial accuracy (see Fig. 1). Although verbal accuracy and spatial accuracy scores themselves were not significantly related to LES, there was a tendency for verbal accuracy scores to be negatively correlated with LES (1.= -0.3367, n = 19, P= 0.159) and the opposite tendency was found for spatial accuracy scores (r= 0.3606, n = 19, P= 0.129). No significant linear relationship was found between CI, and LES. In males LES was found to be significantly correlated also with BMI (r= -0.5271, n = 19, P= 0.020), i.e. more obese males displayed less eye shifts to the left eye (see Fig. 2). In females neither the cognitive variables, nor the BMI were related to LES. To check for the reliability of the ratio of leftward eye shifts and also to examine whether a smaller number of trials would be sufficient to assess eye shift laterality, two ratios of leftward eye shifts (LESl and LES2) were computed separately based on the first and the second 30 “simultaneous” trials, respectively. The correlation between the two ratios was 0.7721 (.P=O.OOl) for the whole group, 0.8875 (P-cO.001) for males and 0.6902 (P
1304
J. WEISZ and G. ADAM
0.0
0.2
0.4
0.6
0.6
1.0
LES Fig.
I. The relationship
between Cf, (verbal accuracy Z-score minus spatial LES (the ratio of leftward eye shifts) in males.
16
’ 0.0
0.2
0.4
0.6
0.6
accuracy
Z-score)
and
1.0
LES Fig. 2. The relationship
BMI and significant
between
BMI (weight/height2) males.
and LES (the ratio of leftward
LES2 was also significant (-0.5545, n= 19, P=O.O14), correlation was found between BMI and LESl (r=O.4347,
eye shifts) in
while a marginally n= 19, P=O.O63).
DISCUSSION Previous research related to the connection between handedness and sighting dominance, on the one hand, and cognitive skills, on the other are equivocal (for review, see Ref. [16]). In the majority of the studies which reported an association between handedness and cognitive performance, right-handers were found to outperform left-handers both in verbal and in spatial tasks. However, Porac and Coren reported that college students majoring in literature or langauge are more strongly right-handed than those specializing in science or in visual arts [16]. As to sighting dominance, Kerschner and Jeng found that right-eyed subjects were better in processing verbal material, while left-eyed subjects were better in the perception and recall of nonverbal material [ 1 I]. In another study, left-eyed subjects showed superior performance in a mental rotation task [ 163. Our results also support that, at least in males, right-sidedness is associated with a better verbal vs spatial performance.
LATERAL
EYE MOVEMENTS
1305
The main finding of this study was that in males, eye movements evoked by bilateral visual stimulation are related to cognitive performance. Specifically, males with a preference of leftward eye movements are relatively more accurate on a spatial than a verbal task, while the opposite is true for males preferring rightward eye movements. This result, in itself, strongly indicates that eye movements are in close connection with hemispheric preference. The direction of the effect is in accordance with expectations, based on the results of brain stimulation, brain damage and CLEM studies. The other findings of this study seems to offer additional, albeit indirect, support for the hypothesis that eye movement direction during the BVS procedure reflects hemispheric preference. In males, body mass index was found to negatively correlate with the ratio of leftward eye movements, a result similar to those found in CLEM studies [lS, 261. CLEM studies were interpreted as indicating that there is a connection between hemispheric preference and obesity, left hemisphere preferent individuals being more obese than right hemisphere preferent ones. We expected that in our experiment this connection should be reflected in a relationship between eye movements and BMI. For males this expectation proved to be true, and the direction of the relationship was the same as that found in CLEM studies. It was suggested by Rodin and Singer that obese individuals are more responsive to external stimuli than normal weight subjects, and this “external responsiveness of many overweight individuals might be a predisposition to process information in the externally oriented, verbal (left hemisphere) mode” [ 181. The tendency toward right eye shifts for males with a greater BMI, found in this study is in harmony with this hypothesis. However, some aspects of the results are not easily interpretable. First, we have no ready explanation for the fact that only males displayed the expected effects while none ofthem was found in females. It must be mentioned that such sex differences are often reported in laterality research, and they are usually explained as being the consequences of the different functional organization of male and female brains. As regards lateral eye movements specifically, Dunn pointed to the fact that “the strongest findings supporting CLEM theory are more consistently found using male than female subjects”[6]. Another problem was the lack of the expected relationship between the directionality of eye movements and one of the cognitive indices (CI i). It can hardly be argued that CI, is not an appropriate index of verbal vs spatial abilities, because its dependence on sex, on handedness and on sighting dominance speaks in favour of its validity. In our study electrooculographic recording and the great number of the trials used made the procedure more difficult as compared to the usual CLEM method. However, it seems very probable that EOG recording could be replaced by the technically simpler videorecording. Considering the relatively high amplitude of the eye movements and the fixation of the head during the procedure, with an appropriate placement of a video-recorder the gaze shifts could be easily recorded and scored without a loss of accuracy. Besides, the fact that using only the trials of the first half of the experiment (LESI) did not notably change any of the results indicates that a shorter procedure would equally do. Taken together, our study can be regarded only as the first step in a process of validating the direction of BVS-evoked eye shifts as an index of hemispheric preference. Other studies using other types of cognitive tasks, and also EEG, cerebral blood flow and metabolic studies are needed to confirm the usefulness of the BVS procedure. However, ifthe validation would succeed it would be highly recommendable to substitute the CLEM procedure with the BVS procedure, because the latter could be much more easily standardized and could result in more comparable data.
1306
J. WEISZ and G. .&vAu
REFERENCES I. AUTKET. A., AUVEKT, L., LAFFONT, F. and LARMANUE, P. Electroencephalographic
spectral power and lateralized motor acivities. EIecrroenceph. c/in. NeurophysioL 60, 228-236, 1985. 2. BEAUMONT,J. G., YOUNC;,A. W. and MCMANUS, I. C. Hemisphericity: A critical review. Cognit. Neuropsycho[, I, 191-212, 1984. 3. BENN, R. T. Indices of height and weight as measures of obesity. Br. J. Prev. Sot. Med. 24, 64-68, 1970. 4. CASEY, S. M. The influence of lateral orientation on cerebral processing. Cortex 17, 503-514, 1981. 5. CHARLTON, S., BAKAN, P. and MORETTI, M. Conjugate lateral eye movements: A second look. Int. J. Neurosci. 48, l-18. 1989. 6. DUNN, B. R., BARTSCHER, J., TURANICZO, M. and GRAM, P. Relationship between conjugate lateral eye movements, brain organization, and cognitive style. Brain Coynit. 10, 171-188. 1989. 7. EHKLICHMAN, H. and WEINUERGER,A. Lateral eye movements and hemispheric asymmetry: A critical review. Psych& Bull. 85, 1080-l 101, 1978. 8. GROSS, Y., FRANKO, R. and LEWIN, I. Effects of voluntary eye movements on hemispheric activity and choice of cognitive mode. Neuropsychologia 16, 653-651, 1978. 9. CUR, R. C. and REIVICH, M. Cognitive task effects on hemispheric blood flow in humans: Evidence for individual differences in hemispheric activation. Brain Lang. 9, 78-92, 1980. 10. HISCOCK,M. Eye-movement asymmetry and hemispheric function: An examination ofindividual differences. J. Psychol. 97, 49-52, 1977. I1 KEKSHNER, J. R. and JENG, A. G. Dual functional hemispheric asymmetry in visual perception: Effects of ocular dominance and postexposurai processes. Neuropsychologia IO, 437-445, 1972. 12. KINSBOURNE, M. Eye and head turning indicates cerebral lateralization. Science 176, 539-541, 1972. 13. NEUBAUER, A., SC:HULER,G. and PFURTSCHELLER, G. Lateral eye movements as an indication of hemispheric preference: An EEG validation study. ht. J. Psychophysiol. 6. 177-184, 1988. 14. NEWLIN, D. B., ROHRBAUGH, J. W. and VARNER, J. L. Lateral eye movement tendency related to alpha asymmetry. [Abstract] Psychophysiology 19, 338, 1982. 15. OWENS, W. and LIMBER,J. Lateral eye movement as a measure ofcognitive ability and style. Percepr. Mot. Skills 56, 711.-719,1983. 16. PORAC. C. and COREN,S. Lateral Pwfirences and Human Behaviour. Springer, New York, 1981. 17. POSNER. M. 1. and COHEN, Y. Attention and the control of movements. In Tutorials in Motor Behavior, G. E. STELMACK and J. REQUIN (Editors), pp. 243--258. North-Holland, Amsterdam, 1980. 18. ROVIN, J. and SINGER, J. Eye-shift, thought, and obesity. J. Pen. 44, 594-610, 1976. 19. SHEPARV, R. N. and METZLER, J. Mental rotation of three-dimensional objects. Science 171, 701-703, 1971. 20. SHEVRIN, H., SMOKLER, I. and KOOI, K. A. An empirical link between lateral eye movements and lateralized event-related brain potentials. Biol. Psych. 15, 691-697, 1980. 21. SWINNEN,S. Some evidence for the hemispheric asymmetry model of lateral eye movements. Prrcepl. Mol. Skills 58, 79-88, 1984. 22. TEITELBAUM, H. A. Spontaneous rhythmic ocular movements: Their possible relationship to mental activity. Neurology 4, 350-354, 1954. 23. TUCKER, G. H. and Sum; M. R. Conjugate lateral eye movement (CLEM) direction and its relationship to performance on verbal and visuospatial tasks. Neuropsychologia 16, 251-254, 1978. 24. VANUENHEKG,S. G. and KUSE, A. R. Mental rotations, a group test of three-dimensional spatial visualisation. Percept. Mot. Skills 47, 599604, 1978. 25. WALKER, E., WADE, S. and WALVMAN, 1. The effect of lateral visual fixation on response latency to verbal and spatial questions. Brain Cvgnir. 1, 399404, 1982. 26. WEISZ, J., BAL~ZS, L. and A&M. G. Hemispheric preference and obesity. Neuropsychologia 28,883-887, 1990. 27. WEITEN, W. and ETAUGH, C. Lateral eye movement as related to verbal and perceptual-motor skills and values. Percepl. Mol. Skills 36, 423-428, 1973.
m Great
31, No. 12, pp. 1299-1306.
1993 8
Britain
HEMlSPHERIC PREFERENCE AND MOVEMENTS EVOKED BY BILATERAL JIIJLIA WEISZ* Psychophysiology
and GY~RGY
OOZS-3932193 $6.00+0.00 1993 Pergamon Press Ltd
LATERAL EYE VISUAL STIMULI
ADAM
Research Group of the Hungarian Academy of Sciences, Department Physiology, Eiitviis Lorind University. Budapest, Hungary
of Comparative
(Received 25 January 1993: accepted 1 June 1993) Abstract-This study investigated the relationship between the direction of lateral eye shifts evoked by bilateral visual stimulation, on the one hand, and verbal vs visuospatial performance, on the other. In males a negative correlation was found between verbal relative to spatial accuracy and the ratio of leftward eye movements. No such relationship was found in females. Additionally, in males body mass index was smaller for those who displayed a greater ratio of leftward eye movements. It was concluded that visually evoked lateral eye shifts might reflect hemispheric preference, similarly as lateral eye movements obtained during the widely used questioning procedure. However, the procedure used in this study for evoking lateral eye shifts seems to be free of some methodological problems of the traditional procedure and is more easy to standardize.
INTRODUCTION years the direction of lateral eye movements has been widely used as an index of hemisphericity. Although BEAUMONT et al. [2] argued that the concept of hemisphericity lacks adequate foundation, many researchers maintain that there are relatively stable individual differences in the extent to which subjects activate one or the other hemisphere. The asymmetry of hemispheric activation is assumed to be reflected in the preferred mode of cognitive processing, in the difference of electroencephalographic and metabolic measures between the hemispheres, in perceptual asymmetries, and-which is most relevant regarding this study-in the direction of lateral eye movements. TEITELBAUM [22] was the first to observe that when asked a question which requires reflective thought, subjects move their eyes either to the left or to the right. This gaze shift is known as “conjugate lateral eye movement” (CLEM). When the questioner is face-to-face with the subject, the predominant direction of CLEMs was proposed to depend on the subject’s hemispheric preference. It is hypothesized that in subjects moving their eyes to the left (left movers) the right hemisphere is the more active one, and the opposite is true for right movers. In the “experimenter behind” condition CLEMs vary as a function of question type: verbal questions elicit rightward CLEMs, whereas spatial questions elicit leftward CLEMs. That is, CLEMs are assumed to reflect both characteristic and temporary asymmetry of hemispheric activation, depending on the experimental condition (for review, see Ref. [5]). The validity of lateral eye movements as an index of hemisphericity was tested in studies trying to find an association between CLEMs and other measures of hemispheric preference. IN THE LAST 25
*Address for correspondence: Jlilia Weisz, Department H-1088. Mlizeum krt. 4/a., Budapest, Hungary.
of Comparative
1299
Physiology,
EBtviis Lorand
University,
1300
J. WEKZand G. .bh
The most convincing evidence are of neurophysiological nature. Several studies on EEG alpha asymmetry showed that left movers exhibit a more pronounced EEG activation in the right hemisphere, while the opposite is true for right movers [13, 141. Evoked potentials were also found to be of greater amplitude in the hemisphere contralateral to the preferred gaze direction [20]. Cerebral blood flow results were consistent with the electroencephalographic findings: GUR and REIVICH [9] reported that left movers had more flow to the right hemisphere than to the left one, while for right movers there was an, albeit nonsignificant tendency in the opposite direction. Further substantiation of the eye movement-hemispheric activation hypothesis comes from studies showing a close link between cognitive abilities and CLEM direction. Right movers were often reported to outperform left movers at verbal tasks, but being inferior to left movers in solving visuospatial tasks [21,23,27]. However, in many studies of this type no difference in cognition could be demonstrated between the CLEM groups [lo, 151. Another line ofevidence regarding temporary hemispheric activation indicates that verbal processing is accompanied by rightward, and spatial processing is accompanied by leftward eye movements [12]. This effect works also in the reverse direction: voluntary gaze to the right facilitates verbal performance, while gazing to the left improves spatial performance [4, 8, 253. Lateral gaze was also shown to create an asymmetrical EEG activation with the hemisphere contralateral to gaze direction being the more activated one [l]. Taken together, converging evidence support that CLEMs reflect asymmetric hemispheric activation, and in the face-to-face condition the predominant direction of eye movements can be used as an easy-to-assess indicator of hemisphericity. However, there are a number of unresolved methodological issues as regards CLEM research, which were most thoroughly discussed by EHRLICHMAN and WEINBERGER [7], and which may, at least in part, be responsible for some negative findings. One of these problems is that different researchers use very different numbers and sets of questions to determine CLEM directionality. There is also a lack of consensus about how large an eye movement should be to score it as a CLEM, and how to describe the direction of eye movements. It is also not clear whether the subject should fixate centrally during the question, or not. Briefly, in spite of being used for a long time, the CLEM procedure resisted standardization up to now. Thus, it seemed reasonable to make an attempt for the substitution ofCLEM procedure by a method which avoids the aforementioned problems. In a study of PosNER and COHEN [17] two dots were simultaneously presented in the two visual hemifields and the subjects were instructed to move their eyes in the direction that they felt most comfortable. In the present study we assumed that this simple technique could be appropriate for revealing the characteristic tendencies of individuals to make eye movements in one or the other horizontal direction. We were encouraged to make this hypothesis by the observation of Posner and Cohen who noted that “many subjects showed strong biases in one direction or another”, although they did not try to analyse this phenomenon further. In the present paper we refer to this method as the ‘bilateral visual stimulation” (BVS) procedure, and use the terms “left shifters” and “right shifters” for those displaying a preference of leftward and rightward eye movements, respectively, during the BVS procedure. The present experiment aimed at determining whether the predominance of left or right eye shifts in the BVS condition correlated with hemisphericity. To validate the BVS method it seemed advisable to use an indicator of hemisphericity independent of eye movements. As it is widely accepted that the performance on verbal and visuospatial tasks is in close
LATERAL EYE MOVEMENTS
1301
connection with left and right hemispheric functions, respectively, we chose to use a cognitive index characterizing verbal vs spatial performance for assessing hemisphericity. Thus, our primary hypothesis was the right shifters would exhibit better verbal and worse spatial performance than left shifters. A secondary assumption of this study was that the directionality of eye shifts would be related not only to cognition but also to other individual characteristics that are supposed to be in connection with asymmetrical hemispheric activation. Specifically, we expected that obesity would also be correlated with eye shift preference. RODIN and SINGER[lS] reported that obese subjects exhibited more rightward CLEMs than normal weight subjects. This result was supported by us in a previous study where we found that right movers were more obese than left movers [26]. Accordingly, we expected that in the present experiment right shifters would turn out to be more obese than left shifters.
METHOD Subjects Thirty-nine right-handed university students (19 males and 20 females) aged IS-25 years (x= 20.5 years) volunteered to the study, The subjects were classified as right-handed with a four-item handedness questionnaire 116, p. 34); handedness scores could vary between -4 and 4, positive values indicating right-handedness, and negative values indicating left-handedness. Materials Both cognitive tasks used were paper-and-pencil tasks. The written instruction emphasized the importance of both speed and accuracy of performance. The verbal task containing a list of 100 Hungarian words where half of the words had only one meaning whereas the other 50 words had two independent meanings. The order of the one-meaning and two-meaning items was randomized. The subject was required to put a + sign next to each word that he regarded a two-meaning word and to put a - sign next to each one-meaning word. The spatial task was a mental rotation test [24] based on the figures used by SHEPARD and METZLER [19]_ Each item contained a drawing of a three-dimensional criterion object and four highly similar alternative figures. Two of the four alternatives were identical to the criterion in structure but were rotated in space. The task was to identify and mark these two alternatives. This task consisted of 20 items, each with two correct answers. Apparatus and procedure Electrooculogram (EOG) was recorded by a Beckman Type R411 Dynograph. The electrodes were placed at the outer canthi. An AC preamplifier (time constant = 1 set) was used to obtain lateral eye movement information. The presentation of visual stimuli during the BVS procedure was controlled by a computer. The electrooculogram and a computer-generated marker signal simultaneous with the appearance of the visual stimuli were recorded on the chart. At the beginning of the experiment the subject filled out the handedness questionnaire. After that sighting dominance was determined by the point test [16, p. 1971. The experimenter stood facing the subject at a distance of about 3 m and the subject was required to keep both eyes open and to point with an outstretched hand at the experimenter’s nose, which was directly in front of him. The eye aligned with the pointing finger was noted by the experimenter, The point test was administered six times with alternating hands. Scores for sighting dominance can vary from -6 to 6, where negative values represent left eye sighting dominance and positive values reflect right eye sighting dominance. Body mass index (BMI = weight/height’) was determined by measuring the subject’s weight and height 131. After completion of the point test the verbal and the spatial tasks were administered. The order of the two cognitive tasks was counterbalanced across subjects. Each subject had 3 min to work on the verbal task and 5 min to work on the spatial one. After attaching the EOG electrodes the subject was seated in a sound-attenuated experimental chamber and left alone. Throughout the BVS procedure the subject was monitored using a closed-circuit TV system. By using an adjustable chin rest the subject looked at a TV screen located in front of him, from a distance of 50cm. The centre of the screen was at eye-level. Each trial of the BVS procedure began with a warning signal: a small flashing cross appeared in the centre ofthe screen which the subject was to fixate. The warning signal was exposed for 1.5set and at the moment ofits disappearance two small yellow dots appeared on the screen 10.5 cm to the left and to the right of fixation. The duration ofexposurewas 800 msec. The subject was instructed to move his eyes on the
1302
J. WEISZ and G. AD,&
appearance of the dots to the right or to the left, whichever seemed most comfortable. The trials were separated by 5.5-7.5 set intervals. The BVS procedure consisted of two blocks each with 36 trials. Within each block there were six “not simultaneous” trials inserted; in these trials there was a 350 msec time difference between the appearance of the two dots, the right dot and left dot each leading equal times. Every 3-5 “simultaneous” trials were followed by such a “not simultaneous” trial with left dot leading and right dot leading trials alternating. The “not simultaneous” trials were aimed at breaking the subject’s possible looking bias, caused by factors other than hemisphericity. In the remaining trials the two dots appeared simultaneously on the screen. There was a 24 min rest period interposed between the blocks. Prior to the beginning of the first block of trials each subject was given 10 practice trials.
Data analysis
There were enough verbal and spatial items so that none of the subjects finished all within the time allotted them. For the verbal task each correct answer was awarded by one point. For the spatial task an answer was regarded as correct only if both alternatives chosen were correct. For each correct answer one point was given. If only one alternative was chosen and it was correct 0.5 point was given. For each subject two measures of performance were determined for each cognitive task: the number of correct choices (correct score) and the accuracy score. The latter was calculated by dividing the number of correct choices by the number of all items attempted. Correct score depends on both the speed and the accuracy of performance while accuracy score is a pure measure of accuracy. Each subject’s verbal and spatial correct scores were converted to z-scores by subtracting from it the group mean and dividing the resulting difference by the standard deviation. A cognitive index (CI i) was computed by subtracting the spatial from the verbal z-score. This measure characterized each subject’s verbal vs spatial performance. Negative Cl, values indicate relatively better spatial performance and positive values indicate relatively better verbal performance. The same computation was performed on the accuracy scores, resulting in another cognitive index (Cl,) which reflected verbal vs spatial accuracy ofeach subject. Negative CI, values indicate relatively greater accuracy on the spatial task whereas positive values indicate relatively greater accuracy on the verbal task. For the BVS procedure the ratio ofleftward eye shifts (LES) was computed by dividing the number of leftward shifts by the sum of leftward and rightward eye shifts.
RESULTS Cognitive
measures
The mean values of cognitive measures for males and females are presented in Table 1. Correlational analyses showed no linear relationship between CI 1 and C,, either in males, or in females.
Table
1. Mean values of cognitive measures and in females
Male Verbal correct
score
Spatial
score
correct
Verbal accuracy
score
Spatial
score
CI, CI,
accuracy
in males
(S.D.)
38.2 (12.4) 8.2 (3.4) 0.86 (0.05) 0.82 (0.17) -0.46 (1.33) -0.02 (1.02)
Female 38.7 (11.1) 5.6 (2.2) 0.83 (0.07) 0.74 (0.22) 0.44 (1.09) 0.02 (0.95)
LATERAL
EYE MOVEMENTS
I303
In order to examine whether the values of cognitive indices are sex-dependent, separate one-way analyses of variance were conducted on CI, and CI, with sex as the independent variable. It was found that CI, was smaller for males than for females, i.e. males exhibited worse verbal vs spatial performance than females [F (1, 37) = 5.37, P=O.O26]. To determine whether this difference could be attributed to a sex difference in spatial performance or verbal performance, or both, spatial correct scores and verbal correct scores were analysed separately. A sex difference emerged only for spatial correct scores, with males displaying better spatial performance than females [F (1, 37) = 8.46, P=O.O06]. Verbal correct scores did not significantly differ between sexes [F (1, 37)=0.02, P=O.886]. For CI, (verbal accuracy vs spatial accuracy) no sex difference could be observed. The relationship between CI,, on the one hand, and handedness and sighting dominance, on the other was examined by conducting stepwise multiple regression analyses separately for males and females where the dependent variable was CI 1, and the independent variables were handedness and sighting dominance scores. (The criterion for the independent variables to be added to the regression equation was set at P=O.l.) For males it was found that both independent variables have significant influence on CI 1. Partial regression coefficients for the handedness score and the sighting dominance score were 0.52 (P=O.O16) and 0.46 (P=O.O31), respectively, and the two variables explained 32.5% of the total variance of CI,. That is. in males the greater the degree of right-handedness and right-sightedness, the better the verbal vs spatial performance was. No such relationship was found for females. The other cognitive index used, CI,, did not depend on handedness and sighting dominance scores, either in males or in females.
Latera/
eye sh[fts
The mean value of the ratio of leftward eye shifts for males was 0.46 (S.D. =0.21), and for females 0.49 (S.D. = 0.22). The difference between the sexes did not approach significance. Both visual inspection of the LES histogram and statistical analyses indicated that the ratio of leftward eye shifts is approximately normally distributed (kurtosis: -0.213; skewness: 0.009; Kolmogorov-Smirnov Z=O.667, P=O.766). Correlational analyses showed that in males the ratio of leftward eye shifts is in negative connection with CI, (r= -0.5705, n= 19, P=O.Ol l), that is, the greater the ratio of the leftward eye shifts, the smaller the verbal vs spatial accuracy (see Fig. 1). Although verbal accuracy and spatial accuracy scores themselves were not significantly related to LES, there was a tendency for verbal accuracy scores to be negatively correlated with LES (1.= -0.3367, n = 19, P= 0.159) and the opposite tendency was found for spatial accuracy scores (r= 0.3606, n = 19, P= 0.129). No significant linear relationship was found between CI, and LES. In males LES was found to be significantly correlated also with BMI (r= -0.5271, n = 19, P= 0.020), i.e. more obese males displayed less eye shifts to the left eye (see Fig. 2). In females neither the cognitive variables, nor the BMI were related to LES. To check for the reliability of the ratio of leftward eye shifts and also to examine whether a smaller number of trials would be sufficient to assess eye shift laterality, two ratios of leftward eye shifts (LESl and LES2) were computed separately based on the first and the second 30 “simultaneous” trials, respectively. The correlation between the two ratios was 0.7721 (.P=O.OOl) for the whole group, 0.8875 (P-cO.001) for males and 0.6902 (P
1304
J. WEISZ and G. ADAM
0.0
0.2
0.4
0.6
0.6
1.0
LES Fig.
I. The relationship
between Cf, (verbal accuracy Z-score minus spatial LES (the ratio of leftward eye shifts) in males.
16
’ 0.0
0.2
0.4
0.6
0.6
accuracy
Z-score)
and
1.0
LES Fig. 2. The relationship
BMI and significant
between
BMI (weight/height2) males.
and LES (the ratio of leftward
LES2 was also significant (-0.5545, n= 19, P=O.O14), correlation was found between BMI and LESl (r=O.4347,
eye shifts) in
while a marginally n= 19, P=O.O63).
DISCUSSION Previous research related to the connection between handedness and sighting dominance, on the one hand, and cognitive skills, on the other are equivocal (for review, see Ref. [16]). In the majority of the studies which reported an association between handedness and cognitive performance, right-handers were found to outperform left-handers both in verbal and in spatial tasks. However, Porac and Coren reported that college students majoring in literature or langauge are more strongly right-handed than those specializing in science or in visual arts [16]. As to sighting dominance, Kerschner and Jeng found that right-eyed subjects were better in processing verbal material, while left-eyed subjects were better in the perception and recall of nonverbal material [ 1 I]. In another study, left-eyed subjects showed superior performance in a mental rotation task [ 163. Our results also support that, at least in males, right-sidedness is associated with a better verbal vs spatial performance.
LATERAL
EYE MOVEMENTS
1305
The main finding of this study was that in males, eye movements evoked by bilateral visual stimulation are related to cognitive performance. Specifically, males with a preference of leftward eye movements are relatively more accurate on a spatial than a verbal task, while the opposite is true for males preferring rightward eye movements. This result, in itself, strongly indicates that eye movements are in close connection with hemispheric preference. The direction of the effect is in accordance with expectations, based on the results of brain stimulation, brain damage and CLEM studies. The other findings of this study seems to offer additional, albeit indirect, support for the hypothesis that eye movement direction during the BVS procedure reflects hemispheric preference. In males, body mass index was found to negatively correlate with the ratio of leftward eye movements, a result similar to those found in CLEM studies [lS, 261. CLEM studies were interpreted as indicating that there is a connection between hemispheric preference and obesity, left hemisphere preferent individuals being more obese than right hemisphere preferent ones. We expected that in our experiment this connection should be reflected in a relationship between eye movements and BMI. For males this expectation proved to be true, and the direction of the relationship was the same as that found in CLEM studies. It was suggested by Rodin and Singer that obese individuals are more responsive to external stimuli than normal weight subjects, and this “external responsiveness of many overweight individuals might be a predisposition to process information in the externally oriented, verbal (left hemisphere) mode” [ 181. The tendency toward right eye shifts for males with a greater BMI, found in this study is in harmony with this hypothesis. However, some aspects of the results are not easily interpretable. First, we have no ready explanation for the fact that only males displayed the expected effects while none ofthem was found in females. It must be mentioned that such sex differences are often reported in laterality research, and they are usually explained as being the consequences of the different functional organization of male and female brains. As regards lateral eye movements specifically, Dunn pointed to the fact that “the strongest findings supporting CLEM theory are more consistently found using male than female subjects”[6]. Another problem was the lack of the expected relationship between the directionality of eye movements and one of the cognitive indices (CI i). It can hardly be argued that CI, is not an appropriate index of verbal vs spatial abilities, because its dependence on sex, on handedness and on sighting dominance speaks in favour of its validity. In our study electrooculographic recording and the great number of the trials used made the procedure more difficult as compared to the usual CLEM method. However, it seems very probable that EOG recording could be replaced by the technically simpler videorecording. Considering the relatively high amplitude of the eye movements and the fixation of the head during the procedure, with an appropriate placement of a video-recorder the gaze shifts could be easily recorded and scored without a loss of accuracy. Besides, the fact that using only the trials of the first half of the experiment (LESI) did not notably change any of the results indicates that a shorter procedure would equally do. Taken together, our study can be regarded only as the first step in a process of validating the direction of BVS-evoked eye shifts as an index of hemispheric preference. Other studies using other types of cognitive tasks, and also EEG, cerebral blood flow and metabolic studies are needed to confirm the usefulness of the BVS procedure. However, ifthe validation would succeed it would be highly recommendable to substitute the CLEM procedure with the BVS procedure, because the latter could be much more easily standardized and could result in more comparable data.
1306
J. WEISZ and G. .&vAu
REFERENCES I. AUTKET. A., AUVEKT, L., LAFFONT, F. and LARMANUE, P. Electroencephalographic
spectral power and lateralized motor acivities. EIecrroenceph. c/in. NeurophysioL 60, 228-236, 1985. 2. BEAUMONT,J. G., YOUNC;,A. W. and MCMANUS, I. C. Hemisphericity: A critical review. Cognit. Neuropsycho[, I, 191-212, 1984. 3. BENN, R. T. Indices of height and weight as measures of obesity. Br. J. Prev. Sot. Med. 24, 64-68, 1970. 4. CASEY, S. M. The influence of lateral orientation on cerebral processing. Cortex 17, 503-514, 1981. 5. CHARLTON, S., BAKAN, P. and MORETTI, M. Conjugate lateral eye movements: A second look. Int. J. Neurosci. 48, l-18. 1989. 6. DUNN, B. R., BARTSCHER, J., TURANICZO, M. and GRAM, P. Relationship between conjugate lateral eye movements, brain organization, and cognitive style. Brain Coynit. 10, 171-188. 1989. 7. EHKLICHMAN, H. and WEINUERGER,A. Lateral eye movements and hemispheric asymmetry: A critical review. Psych& Bull. 85, 1080-l 101, 1978. 8. GROSS, Y., FRANKO, R. and LEWIN, I. Effects of voluntary eye movements on hemispheric activity and choice of cognitive mode. Neuropsychologia 16, 653-651, 1978. 9. CUR, R. C. and REIVICH, M. Cognitive task effects on hemispheric blood flow in humans: Evidence for individual differences in hemispheric activation. Brain Lang. 9, 78-92, 1980. 10. HISCOCK,M. Eye-movement asymmetry and hemispheric function: An examination ofindividual differences. J. Psychol. 97, 49-52, 1977. I1 KEKSHNER, J. R. and JENG, A. G. Dual functional hemispheric asymmetry in visual perception: Effects of ocular dominance and postexposurai processes. Neuropsychologia IO, 437-445, 1972. 12. KINSBOURNE, M. Eye and head turning indicates cerebral lateralization. Science 176, 539-541, 1972. 13. NEUBAUER, A., SC:HULER,G. and PFURTSCHELLER, G. Lateral eye movements as an indication of hemispheric preference: An EEG validation study. ht. J. Psychophysiol. 6. 177-184, 1988. 14. NEWLIN, D. B., ROHRBAUGH, J. W. and VARNER, J. L. Lateral eye movement tendency related to alpha asymmetry. [Abstract] Psychophysiology 19, 338, 1982. 15. OWENS, W. and LIMBER,J. Lateral eye movement as a measure ofcognitive ability and style. Percepr. Mot. Skills 56, 711.-719,1983. 16. PORAC. C. and COREN,S. Lateral Pwfirences and Human Behaviour. Springer, New York, 1981. 17. POSNER. M. 1. and COHEN, Y. Attention and the control of movements. In Tutorials in Motor Behavior, G. E. STELMACK and J. REQUIN (Editors), pp. 243--258. North-Holland, Amsterdam, 1980. 18. ROVIN, J. and SINGER, J. Eye-shift, thought, and obesity. J. Pen. 44, 594-610, 1976. 19. SHEPARV, R. N. and METZLER, J. Mental rotation of three-dimensional objects. Science 171, 701-703, 1971. 20. SHEVRIN, H., SMOKLER, I. and KOOI, K. A. An empirical link between lateral eye movements and lateralized event-related brain potentials. Biol. Psych. 15, 691-697, 1980. 21. SWINNEN,S. Some evidence for the hemispheric asymmetry model of lateral eye movements. Prrcepl. Mol. Skills 58, 79-88, 1984. 22. TEITELBAUM, H. A. Spontaneous rhythmic ocular movements: Their possible relationship to mental activity. Neurology 4, 350-354, 1954. 23. TUCKER, G. H. and Sum; M. R. Conjugate lateral eye movement (CLEM) direction and its relationship to performance on verbal and visuospatial tasks. Neuropsychologia 16, 251-254, 1978. 24. VANUENHEKG,S. G. and KUSE, A. R. Mental rotations, a group test of three-dimensional spatial visualisation. Percept. Mot. Skills 47, 599604, 1978. 25. WALKER, E., WADE, S. and WALVMAN, 1. The effect of lateral visual fixation on response latency to verbal and spatial questions. Brain Cvgnir. 1, 399404, 1982. 26. WEISZ, J., BAL~ZS, L. and A&M. G. Hemispheric preference and obesity. Neuropsychologia 28,883-887, 1990. 27. WEITEN, W. and ETAUGH, C. Lateral eye movement as related to verbal and perceptual-motor skills and values. Percepl. Mol. Skills 36, 423-428, 1973.