Abstract Composites created using only the left side of a face and only the right side of the face were judged by participants as one being more intense in emotional expression than the other. Roughly half the participants looked at composites expressing negative emotion, and the other half looked at composites expressing positive emotion. The left side composites were rated above chance as being more intense in emotion than the right side composites. The group observing negative emotion expression chose composites of the left side of the face at a rate significantly greater than the group observing positive emotion expression. The findings are consistent with previous research suggesting the right hemisphere of the brain is more involved in emotional processes than the left hemisphere.

The results also suggest that emotional expression is assymmetrical in the face being more intense in the left side of the face and that this effect is stronger for the negative emotion condition. NEW PAGE There is a basic crossing in the nervous system, particularly concerning sensory input and motor output. The right and left hemispheres of the brain have also been shown to be assymmetrical and differ in their function, which is called hemispheric specialization. There have been studies conducted in this area using patients with brain damage in different hemispheres, where movies shown to elicit emotions were viewed. Both the left brain-damaged patients and the participants without brain-damage often looked away from the screen when unpleasant material was shown, whereas right brain-damaged patients rarely did, suggesting that the right side of the brain is patricularly involved in the degree of emotional involvement or manner of coping with stress, which was reduced in these patients (A. Mammucari, C.

Caltagirone, P. Ekman, W. Friesan, G. Gainotti, L Pizzamiglio and P.

Zoccolotti, 1987). There is also other evidence to suggest the right side of the brain is more involved in the processing and expression of emotions than the left hemisphere. It was proposed by Wolff (Wolff, 1943 cited in Sackeim and Gur, 2003) that the right side of the face is perciever as more similar to the whole face than the left side, however it was later discovered by Gilbert and Bak an (1973, cited in Sackeim and Gur, 2003) that rather than as symmetries in facial expression, these results were produced by a bias of the perciever having greater ability in the right hemisphere of the brain in facial recognition than the left hemisphere. However the possibility that facial expression is assymmetrical was not disprove n.

Therefore the question arose of whether the expression of emotions in the face is assymmetrical and as there is a basic crossing in the nervous system, if the emotions are expressed more intensely on the left side of the face. It is also thought the right hemisphere could possibly be more involved in processing negative information than the left, and a second question was generated of whether this effect was stronger for negative emotional expressions as compared to positive emotional expressions. To test this, composites of faces were constructed using the left side of the face and its mirror image referred to as left-left composites (LL) and the right side of the face and its mirror image referred to as right-right composites (RR). Participants were divided into two groups, one group observing negative emotion composites of LL and RR and the other observing positive emotion composites, and were asked to choose which composite expressed emotion more intensely. It was hypothesized that the LL composites would be chosen more often that the RR composites as more emotionally expressive for both the positive and negative emotion condition. If supported, this would suggest that emotions are expressed more intensely in the left side of the face than the right.

A second hypothesis was generated that the effect would be stronger for the negative emotion condition than the positive emotion condition, and if both hypotheses were supported it would suggest the right hemisphere is particularly involved in processing negative information. Method Participants The participants were 484 undergraduates at the University of Western Australia. A coin was tossed by each participant which determined which group they were allocated to. One group looked at faces that expressed negative emotions which consisted of 238 students, and the other group looked at faces conveying positive emotions which consisted of 246 students. Apparatus Sixteen composites of faces were constructed for negative emotions and 16 composites constructed for positive emotions. They were created using black and white photographs of faces, the models of who deliberately conveyed a negative or positive emotion.

These photographs from the study conducted by Ekman and Friesan (1975) on recognition of emotions in faces, and the mirror image of these photographs were cut vertically through the midline of the face, and the left side and its reversal were joined to make a LL composite, and the right side of the face and its reversal were joined to make a RR composite, creating symmetrical composites using only one side of the face. Half of the faces were created from male models and the other half using female models. The faces were presented on computers and the data was entered into the computers. Procedure The group of participants determined by a coin toss to look at negative emotion composites of faces were presented with a slide consisting simultaneously of two faces being the LL composite and RR composite constructed from the original face and its reversal. These were positioned vertically to eliminate perciever bias. The faces remained on the screen for 10 seconds and the viewer was required to choose which face expressed the emotion more intensely.

A 20 second inter-trial interval then followed and the process was repeated using different composites of faces expressing negative emotion. Eight slides of two composites were shown in total. The process was the same for those participants in the other group except positive emotion composites only were used. A forced choice format was used, requiring the participants to make a choice between the top or bottom composites and 'neither' was not a given option.

This was to reduce the number of responses so a measurable outcome was produced. The gender of the model in the photographs and the position of the faces (if LL or RR was on top) were ordered non-systematically. Results The data was screened for outliers to generate statistics more accurately representing the participants. For the purpose of conducting a general comparison, statistics for the entire sample size of participants were calculated as well as for both the groups who looked at negative emotion composites and positive emotion composites. The sample size of participants in the study overall was 484, with 246 of those participants in the 'positive image' group and 238 in the 'negative image' group. The mean number for the entire group out of 8 in which the participant chose the LL composite face was 4.

74, with 4. 64 being the mean number chosen for the 'positive image' group and 4. 95 being the mean for the 'negative image' group. All groups performed above what would be considered chance.

In order to measure the dispersion in the distribution of scores, the standard deviation was calculated. For all the participants a standard deviation of 1. 09 was calculated, the 'positive image' group had a standard deviation of 1. 15 and the 'negative image group' a standard deviation of 1. 00.

To determine if the means were significantly different from each other, the confidence interval was calculated as shown in table 1. The confidence interval indicates a 95% probability that a given result lies within that range. TABLE and GRAPH TABLE 1. Statistics for the Number of Times Left-left Composites were chosen out of Eight in each Group.

Group Sample Mean Standard 95% CI 95% CI Size Deviation Lower Bound Upper Bound Overall 484 4. 79 1. 09 4. 69 4.

89 Negative 238 4. 95 1. 00 4. 82 5.

07 Positive 466 4. 64 1. 15 4. 50 4. 79 Discussion The hypothesis that the left side of the face expresses emotion more intensely than the right side was supported by the results of the experiment as the number of times the LL composites were chosen was significantly greater on average at 4. 79 out of 8, than the chance level of performance at 4 out of 8.

This is also consistent with findings that right hemisphere is dominant in emotional processes. The lower bound of the 95% confidence interval of the overall sample mean was greater than four indicating a significant effect, implying that the left side of the face is more expressive of emotions. The second hypothesis that the lateralization of emotion effect is stronger for the expression of negative emotions compared to positive emotions was also supported by the data collected. This was determined using a sample t-test. The mean number of times the LL composites were chosen in the negative condition was significantly greater than, t (482) = 3. 09, p <.

01 (2-tailed). Both the hypotheses were supported. These findings have implications on the social adapted ness of emotional expression, as when face to face with another person, it is likely that the left side of the face is opposite the right side of the other persons face, meaning that the hemisphere inferior in emotional processes recieves information from the side of the face more expressive in emotion, which may be advantageous or disadvantageous depending on the situation (Sackeim and Gur, 1978). There were several limitations of the experiment. The models in the photographs used were asked to express certain emotions in their face however they may not have felt the actual emotions and this may have an effect on how well the research applies to a realistic situation. Only eight trials were used which seems limited considering the wide range of emotional expression possible.

Black and white photographs were used which again is further from a realistic situation than desired. It is also a matter for future research to observe whether the right or left-handedness of models or participants have an effect on the results, and to apply the experiment cross-culturally.