Visual-Field Bias in the Judgment of Facial Expression of Emotion
Hari S. Asthana
ABSTRACT. The left and right hemispheres of the brain are differentially related to the processing of emotions. Although there is little doubt that the right hemisphere is relatively superior for processing negative emotions, controversy exists over the hemispheric role in the processing of positive emotions. Eighty right-handed normal male participants were examined for visual-field (left-right) differences in the perception of facial expressions of emotion. Facial composite (RR, LL) and hemifacial (R, L) sets depicting emotion expressions of happiness and sadness were prepared. Pairs of such photographs were presented bilaterally for 150 ms, and participants were asked to select the photographs that looked more expressive. A left visual-field superiority (a right-hemisphere function) was found for sad facial emotion. A hemispheric advantage in the perception of happy expression was not found.
Key words: bias, emotion, face, visual field
ATTEMPTS HAVE BEEN MADE to examine the issue of hemispheric bias in the perception of facial expressions of emotion with both neurologically intact brain subjects and brain-damaged patients (see a review by Borod, 1992; Borod & Koff, 1989; Borod, St. Clair, Koff, & Alpert, 1990; Mandal, Asthana, Tandon, & Asthana, 1992; Sackeim & Grega, 1987). Split-field studies with intact-brain subjects indicate a left visual-field (Lvf, a right-hemisphere function) superiority in the perception of facial expressions of emotion (Ley & Bryden, 1979; Mandal & Singh, 1990; Strauss & Moscovitch, 1981). Studies with focal brain-damaged patients indicate that the right hemisphere-damaged patients are significantly more impaired than left hemisphere-damaged patients in tasks requiring identification or recognition (or both) of facial expression of emotion (Borod, Koff, PerlmanLorch, & Nicholas, 1986; Cicone, Wapner, & Gardner, 1980; Dekosky, Heilman, Bowers, & Valenstein, 1980; Mandal, Tandon, & Asthana, 1991).
Contemporary research also has suggested that the hemispheres are specialized for the valence (positive-negative) of emotional stimuli (See reviews by Borod & Koff, 1989; Ehrlichman, 1987; Tucker, 1981). For example, there is a right visual-field (Rvf, a left-hemisphere function) superiority for the emotion of happiness and a Lvf superiority for the emotion of sadness (Reuter-Lorenz & Davidson, 1981; Reuter-Lorenz, Givis, & Moscovitch, 1983). Other studies contradict the findings, indicating a bilateral involvement for the processing of positive emotions and a right-hemisphere involvement for negative emotions (see Borod, Koff, Perlman-Lorch, et al., 1986).
The present study was conducted to reexamine the issue with certain methodological variations. Unlike other studies that used either full face or chimeric faces as stimuli (for example, Ley & Bryden, 1979; Mandal & Singh, 1990), the present study involved judgments about expressed intensity of a left hemiface, a right hemiface, and a hemifacial composite of photographs (right-right, left-left), presented tachistoscopically. It was expected that a valence effect would reflect a Lvf superiority for sadness and a Rvf superiority for happiness in the judgment of expressed intensity of emotion. No hypothesis on the nature of interaction between visual field and facial composite in a given emotion was drawn (but see Borod et al., 1990), although any such interaction would further specify the nature of a valence effect in terms of hemispace (visual field) or hemiface (facial composite).
Eighty right-handed male participants were selected for the study. They were between 15 and 25 years of age. Mean age was 19.4 years (SD = 8.6), and mean education was 12.8 years (SD = 2.5). The participants were mostly of middleclass socioeconomic status. The hand preference for each participant was assessed using the Edinburgh Handedness Inventory (Oldfield, 1971). All participants were right-handed (M = +75). Right-handed persons with a score below +25 (maximum = +100) on this inventory were excluded from the study. Participants with (a) a history of hand change and (b) impaired sighting performance (determined by the participant’s ability to identify photographs of familiar objects shown briefly with an electronic tachistoscope) were excluded.
A total of 18 photographs depicting posed facial expressions of happy (10 photographs, 5 of each sex) and sad (8 photographs, 4 of each sex) emotions were selected from a series of standardized photographs developed by Mandal (1987). These photographs were standardized in three steps.
In the first step, 29 right-handed posers (15 male, 14 female) were photographed for expression of happy and sad facial emotions. The hand preference for each person was assessed by using the Handedness Inventory (Oldfield, 1971). These photographs were shown to 630 observers for recognition. Only those photographs on which there was a consensus of more than 75% of the observers were selected.
In the second step, 100 observers were asked to rate the selected photographs on a 7-point rating scale, with neutral or no emotion at one end of the scale and the intended emotion at the other. The judgment for a particular photograph was derived from the highest rating the observer gave for those two emotions (happy and sad). Only those photographs were retained on which at least 70% of the observers agreed. The procedure followed for standardization of photographs was similar to that used by Ekman and Friesen (1976).
In the third step, using a 5-point scale, 50 observers were asked to make a pair-wise comparison, within the photographs of an emotion, of the degree of expressiveness. The selected photographs had approximately equal intensity ratings for expressiveness.
The selected photographs (10 happy expressions and 8 sad expressions) were randomly divided into two sets, 5 happy expressions and 4 sad expressions in each set. One set was used for the facial composite task and the other set for the hemifacial task.
For the purpose of experimental manipulation, each selected photograph was printed twice, once in normal orientation (RL) and once in mirror-reversed orientation (LR). The facial composite and hemifacial sets were prepared by cutting the RL and LR prints along the vertical midline of the photograph. The midline of each photograph was determined by using the following points: (a) midpoint between the internal canthi of eyes and (b) midpoint of the upper lip (Rhodes & Lynskey, 1989). Each photograph with RL and LR was bisected along the vertical midline of the photograph. The left-left (LL) and right-right side (RR) composites were prepared from the lateralized half of one side of the face and its mirror-image (Sackeim & Gur, 1983). The hemifacial photographs were the vertical halves of the same face. The hemifacial set was prepared by pasting the two hemifaces of a photograph to the left and right of the central line on the photograph, separated from each other by 1 cm.
An electronic tachistoscope capable of exposing photographs from 1 ms to 1000 ms was used. The intensity of light reflected from the exposure field was 1.9 MW/[cm.sup.2] and from the surface of the stimulus photograph in the exposure field it was 2.1 MW/[cm.sup.2]. Exposure duration of a photograph was 150 ms, which was “below the time required for the eyes to refocus at the peripheral stimulation point to the central fixation point” (Wexler, 1980).
Participants were divided into two groups, 40 in each, for their judgment of the experimentally manipulated photographs. The two groups did not differ (p [greater than] .05) on age, education, and other demographic variables. One group judged the photographs displaying expressions of the emotion of happiness, whereas the other group judged the photographs displaying expressions of the emotion of sadness. This procedure was used to control the relativity effect in the judgment of a facial emotion on the judgment of the following emotion (Russell & Fehr, 1987). On this basis, we presumed that a participant’s judgment of a given emotion (for example, happiness) would be affected to some extent by the emotion (for example, sadness) judged earlier.
Participants were asked to look through the central eyepiece of the tachistoscope and fixate their eyes at the central point, following a ready signal. Each participant was given 10 practice trials. For the facial composite task, the photograph of an expressor was presented in two forms: in one, the RR composite was presented in Lvf and the LL composite in Rvf; and in the other, the pattern was reversed: the RR composite was presented in Rvf and the LL composite in Lvf. The task of the participant was to judge the intensity of expression for the two trials. Five photographs of a happy emotion gave rise to a total of 10 trials (5 photographs x 2 trials). The identical procedure was followed for the photographs of sad facial expressions with the instruction to pick the one that looked “sadder.” Each participant gave 8 (4 photographs x 2 trials) judgments of facial expressions.
In the second task, hemifacial photographs (R-L/L-R) of full faces depicting expressions of happy emotions were presented simultaneously for 150 ins, with one hemiface in the Lvf and the other in the Rvf. Participants were asked to judge, as quickly as possible, the hemiface that looked happier. The experimental condition was identical to the facial composite task condition. The presentation order was randomized and counter-balanced. The identical procedure was followed for the photographs of sad facial expressions with the instruction to pick the one that looked sadder.
The percentages of trials in which one visual field was preferred over the other in intensity discrimination of facial composite (RR, LL) depicting happy and sad emotions are shown in Table 1. The dependent measure for analysis was the frequency of preference for a given photograph in the two visual fields. Data were analyzed with a 2 (visual field) x 2 (facial composite) within-subject design for each emotion separately. Emotions (happy and sad) were not considered a between-subjects factor because that would be confounded with the valence factor. Analysis of the data for happy facial composites revealed that the main effects of visual field, F(1, 39) = 2.69, p [greater than] .05, and facial composite, F(1, 39) = 0.08, p [greater than] .05, were nonsignificant. The interaction of Visual Field x Facial Composite was also nonsignificant. A 2 x 2 (Visual Field x Facial Composite) within-subject factorial design analysis of data for sad facial composites revealed that the main effects of visual field, F(1, 39) = 6.57, p [lesser than] .005, and facial composite, F(1, 39) = 13.96, p [lesser than] .001, were significant. Participants judged the photographs as more expressive in Lvf (56.87% of trials) than in Rvf (43.13% of trials), and LL (66.63% of trials) facial composites were judged to be more expressive than RR (39.37% of trials) composites. The interaction effect of Visual Field x Facial Composite was nonsignificant.
The percentages of trials in which a visual field was preferred in intensity discrimination of hemifacial photographs depicting happy and sad emotions are shown in Table 2. Data were analyzed with a 2 (visual fields) x 2 (hemifaces) within-subject factorial design for each emotion separately. For happy emotion, the main effects of visual field, F(1, 39) = 0.019, p [greater than] .05, and the interaction of Visual Field x Hemiface, F(1, 39) = 0.04, p [greater than] .05, were nonsignificant. The main effect of hemiface was significant, F(1, 39) = 17.05, p [less than] .001. Left hemiface (M = 59.75% of trials) was judged more expressive than right hemiface (M = 40.23% of trials). An identical analysis was carried out with sad facial emotion. The main effects of visual field, F(1, 39) = 4.45, p [less than] .05, and hemiface, F(1, 39) 4.19, p [less than] .05, were significant. Participants judged the photographs as more expressive in Lvf (55.03% of trials) than in Rvf (44.37% of trials), and left hemiface (5 5.10% of trials) was judged more expressive than right hemiface (45% of trials). The interaction effect of Visual Field x Hemiface was nonsignificant.
Although the prime objective of the present study was to examine visual-field bias in the perception of facial emotion, the role of facial asymmetry during expression of emotion and the interaction of Perceiver Bias x Expressor Asymmetry were also ascertained. Results indicated that (a) the left hemiface, in general, was judged more expressive than the right hemiface (except for the composite photographs of happiness); (b) the interaction of Visual Field x Facial Asymmetry was not significant; and (c) sad facial emotion was perceived as more expressive in Lvf, although such an effect was not documented for happy emotion.
The present finding that sad facial emotion was perceived as more expressive in Lvf than in Rvf confirms the first part of the hypothesis. The finding, however, did not corroborate some earlier evidence that showed an overall Lvf superiority for the perception of faces expressing emotions (for example, Sackeim & Grega, 1987). Instead, the hypothesis that the positive emotion of happiness would induce bilateral involvement (see Borod, Koff, Perlman-Lorch, et al., 1986) was substantiated by the present findings.
Tucker (1981) reviewed some studies and remarked that hemispheric advantage depends on the valence of (positive–negative) emotional stimuli–for example, negative emotions induce a right-hemisphere advantage and positive emotions induce a left-hemisphere advantage (Reuter-Lorenz & Davidson, 1981; Reuter Lorenz et al., 1983). Ley and Bryden (1979) found that a sad face produces the maximum right-hemisphere advantage and a happy face the minimum. Mandal and Singh (1990) used facial composites of negative emotions (sadness, fear, anger, and disgust) and found an overall Lvf superiority for negative emotions, in which sad emotion was judged with greatest accuracy. Sad faces were also rated as relatively sadder in Lvf than in Rvf (Sackeim, Putz, Vinigaho, & Colman, 1980); such a visual-field asymmetry was not evident for happy emotion.
The difference in processing of positive and negative emotions has been found in lesion studies as well. Right-hemisphere-damaged patients were found to be significantly more impaired in perceiving negative emotions than were left hemisphere-damaged patients and normal control subjects (Borod, Koff, Perlman-Lorch, et al., 1986; Mandal, Mohanty, Pandey, & Mohanty, 1996; Mandal, Tandon, & Asthana, 1991). Although the correlation between a specific impairment and focal brain damage is not sufficient to infer the normal brain function (Sergent, 1988, p. 71), these studies have clearly established the salience of the right hemisphere in emotion processing.
There are certain explanations for the finding that negative emotions are relatively strongly lateralized whereas positive emotions are less lateralized. Diener and Larsen (1984) reported that “the levels of positive affects are more situationally determined whereas the level of negative affects are more cross-situationally consistent” (p. 880). Borod, Koff, and Buck (1986) also had a similar explanation. Their opinion was that negative emotions are more lateralized in the brain because they are associated with survival mechanisms. It is presumed that these behaviors are associated more with gestalt and synthetic processing than with discrete analysis (Borod & Koff, 1984). The gestalt processing is associated more closely with the right hemisphere than with the left hemisphere (Bryden, 1982). Although positive emotions are considered to be less tied to the neural system (Ehrlichman, 1987), it is possible that the bilateral involvement of positive emotion is a function of emotion as well as the nature of stim ulus. It was found that the left hemisphere is quite accurate with a simple, discrete stimulus, such as a smile (Loretta & Danny, 1985). Furthermore, positive emotions have been suggested to be more communicative or linguistic (LH function) than affective (Borod, 1992).
Because the present study included male participants only, the findings need to be substantiated with people from both sexes before a conclusion is drawn. In addition, future studies should be directed toward examining the motoric component (approach-avoidance) rather than simply the hedonic valence (positive-negative) of emotion categories as the basis of lateralization patterns.
Borod, J. C. (1992). Interhemispheric and intrahemispheric control of emotion: A focus on unilateral brain damage. Journal of Clinical and Consulting Psychology, 60, 339-340.
Borod, J. C., & Koff, E. (1984). Asymmetries in affective expression: Behavior anatomy. In N. Fox & R. Davidson (Eds.), The psychobiology of affective development. Hillsdale, NJ: Erlbaum.
Borod, J. C., & Koff, E. (1989). The neuropsychology of emotion: Evidence from normal, neurological and psychiatric populations. In E. Perecman (Ed.), Integrating theory and practice in clinical neuropsychology (pp. 175-215). Hillsdale, NJ: Erlbaum.
Borod, J. C., Koff, E., & Buck, R. (1986). The neuropsychology of facial expression of emotion in normal and brain-damaged subjects. In P. Blanck, R. Buck, & R. Rosenthal (Eds.), Nonverbal communication in the clinical context. University Park, PA: Pennsylvania State University Press.
Borod, J. C., Koff, E., Perlman-Lorch, M., & Nicholas, M. (1986). Expression and perception of facial emotion in brain-damaged patients. Neuropsychologia, 24, 169-180.
Borod, J. C., St. Clair, J., Koff, E., & Alpert, M. (1990). Perceiver and poser asymmetries in processing facial emotion. Brain and Cognition, 13, 167-177.
Bryden, M. P. (1982). Laterality: Hemispheric specialization in the intact brain. San Diego: Academic Press.
Cicone, H., Wapner, W., & Gardner, H. (1980). Sensitivity to emotional situation in organic patients. Cortex, 16, 145-158.
Dekosky, S. T., Heilman, K. M., Bowers, D., & Valenstein, E. (1980). Recognition and discrimination of emotional faces and pictures. Brain and Language, 9, 206-214.
Diener, E., & Larsen, R. J. (1984). Temporal stability and cross-situational consistency of affecting behavioral and cognitive responses. Journal of Personality and Social Psychology, 47, 871-883.
Ehrlichman, H. (1987). Hemispheric asymmetry and positive negative affect. In D. Ottoson (Ed.), Duality and unity of the brain (pp. 194-206). London: Macmillan.
Ekman, P., & Friesen, W. V. (1976). Pictures of facial affect. Palo Alto, CA: Consulting Psychologists Press.
Ley, R. G., & Bryden, M. P. (1979). Hemispheric differences in processing emotions and faces. Brain and Language, 7, 127-138.
Loretta, S., & Danny, W. (1985). Superiority of the left hemisphere in the recognition of emotional faces. International Journal of Neuroscience, 25, 219-223.
Mandal, M. K. (1987). Decoding of facial emotion, in terms of expressiveness by schizophrenics and depressives. Psychiatry, 50, 371-376.
Mandal, M. K., Asthana, H. S., Tandon, S.C., & Asthana, S. (1992). Role of cerebral hemispheres and regions in processing hemifacial expression of emotion. International Journal of Neuroscience, 63, 187-195.
Mandal, M. K., Mohanty, A., Pandey, R., & Mohanty, S. (1996). Emotion specific processing deficit in focal brain damaged patient. International Journal of Neuroscience, 84, 87-95.
Mandal, M. K., & Singh, S. K. (1990). Lateral asymmetry in identification and expression of facial emotions. Cognition and Emotion, 4, 61-70.
Mandal, M. K., Tandon, S.C., & Asthana, H. S. (1991). Right brain damage impairs recognition of negative emotions. Cortex, 27, 247-254.
Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburgh Inventory. Neuropsychologia, 9, 97-114.
Reuter-Lorenz, P., & Davidson, R. J. (1981). Differential contributions of the two cerebral hemispheres for perception of happy and sad faces. Neuropsychologia, 19, 609-614.
Reuter-Lorenz, P., Givis, R. P., & Moscovitch, M. (1983). Hemispheric specialization and the perception of emotion: Evidence from right handers and from inverted and noninverted left handers. Neuropsychologia, 21, 687-692.
Rhodes, G., & Lynskey, M. (1989). Emotion in resting face: Taking sides. A reply to McGee & Skinner (1987). British Journal of Social Psychology, 28, 273-278.
Russell, J. A., & Fehr, B. (1987). Relativity in the perception of emotion in facial expressions. Journal of Experimental Psychology: General, 116, 223-237.
Sackeim, H. A., & Grega, D. M. (1987). Perceiver bias in the processing of deliberately asymmetric emotional expression. Brain and Cognition, 6, 464-473.
Sackeim, H. A., & Gur, R. C. (1983). Facial asymmetry and communication of emotion. In J. T. Cacioppo & R. E. Petty (Eds.), Social psychophysiology: A source book (pp. 300-322). New York: Guilford Press.
Sackeim, H. A., Putz, E., Vinigaho, W., & Colman, E. (1980). Lateralization in processing of emotionally laden information: Normal functioning. Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 1, 97-110.
Sergent, S. (1988). Some theoretical and methodological issues in neuropsychological research. In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (Vol. 1, pp. 69-81). New York: Elsevier North Holland.
Strauss, E., & Moscovitch, M. (1981). Perception of facial expression. Brain and Language, 13, 308-332.
Tucker, D. M. (1981). Lateral brain function, emotion and conceptualization. Psychological Bulletin, 89, 19-46.
Wexler, B. E. (1980). Cerebral laterality and psychiatry: A review of the litemture. American Journal of Psychiatry, 137, 279-291.
Percentages of Trials for Which a Visual Field Was
Preferred for Intensity Discrimination of Facial
Composites (Left-Left, Right-Right) Depicting
Happy and Sad Emotions
Emotion/ Left Right
Photo LL RR LL RR
1 26.25 22.50 27.50 23.75
2 21.25 25.00 25.00 28.75
3 22.50 25.00 25.00 27.50
4 25.00 23.75 26.25 25.00
5 25.00 18.75 31.25 25.00
M 24.00 23.00 27.00 26.00
1 28.75 26.25 23.75 21.25
2 30.75 30.00 20.00 16.25
3 37.50 22.50 27.50 12.50
4 35.00 13.75 36.25 15.00
M 33.75 23.12 26.88 16.25
Percentages of Trials for Which a
Visual Field Was Preferred
for Intensity Discrimination of
Hemifaces (Left-Right) Depicting
Happy and Sad Emotions
Emotion/ Left Right
Photo Left Right Left Right
1 32.25 22.50 27.50 17.50
2 28.75 21.25 28.75 21.25
3 32.50 23.75 26.25 17.50
4 25.00 17.50 32.50 25.00
5 31.25 16.25 33.75 18.75
M 30.00 20.25 29.75 20.00
1 30.00 23.75 26.75 20.00
2 28.75 33.75 16.25 21.25
3 32.50 20.00 30.00 17.50
4 30.00 23.75 26.25 20.00
M 30.31 25.31 24.69 19.69
COPYRIGHT 2001 Heldref Publications
COPYRIGHT 2001 Gale Group