Perceptud and Motor Skills,1983, 57, 683-686. @ Perceptual and Motor Skills 1983ACCURACY AND LATENCY OF JUDGMENTOF FACIAL EXPRESSIONS OF EMOTIONS1GILLES KIROUAC AND FRANCOIS Y. DORfiUniversitC LuualSummary.-The aim of the experiment was to study the relation betweenaccuracy of judgment of facial expressions of emotions and time for judgment.The results for 34 college students confirmed previous data showing high performance in identification of all emotions, although there were some importantdifferences between emotions. Also, times for judgment were longer for theemotions which were more difficult to identify.Empirical data have recently indicated that human subjects can accuratelyand reliably identify specific and fundamental emotions as they are expressedin facial stimuli (Ekman, 1982; Ekman & Oster, 1979; Izard, 1971, 1977).Numerous authors have even concluded that this is a universal or pan-culturalfeature of human functioning (Bouchet, 1979; Ekman, 1982; Shimoda, Arg~le,& Ricci-Bitti, 1978). However, some studies have reported consistent sexdifferences in the encoding and decoding ability of nonverbal signals, includingfacial expressions of emotions (Hall, 1978). As a rule, female subjects havebetter performances than males. In addition, in most experiments, the variousemotions did not receive equivalent recognition levels (Boucher & Carlson,1980; Ekman, 1982; Kirouac & Dork, 1982), and fear is more difficult toidentify than any other emotion.The present experiment assessed these sex and emotion differences bymeasuring two indices of the efficiency of facial recognition, accuracy andlatency of judgment. It should be emphasized that latency of judgment andits relationship with accuracy have not been investigated in the past.METHODSubjectsThe sample of 34 university students, 18 females and 16 males, wereFrench-speaking citizens of Quebec. They were volunteers and received $3.00CAN for their participation.MaterialThe facial stimuli were Ekman’s Picturer of Facial Affect ( 1976). Allslides were used except the neutral ones. Those 96 slides represent 14 Caucasian adults from the United States, six men and eight women, displayingsix fundamental emotions. The frequency of the stimuli is as follows: happi-‘This research was supported by Grant EQ-1717 from Fonds FCAC (Gouvernement duQuebec). The authors are grateful to Mr. Marius Morin and Pierre Boivin for theircooperation in the conduct of the study. Reprint requests should be sent to the firstauthor, Ecole de psychologie, Tour F.-A. Savard, UniversitC Laval, Quebec, QuC., CanadaG l K 7P4.684 G. KIROUAC & F. Y. DORBness ( 18), surprise (14), fear ( I S ) , sadness ( 17), disgust (15) and anger( 17). A micro-computer (Rockwell) and an event-recorder were used toregister the emotions selected by the subjects and to measure the latency ofthese judgments. The event-recorder was a panel of seven keys, one controlled the onset of the slides and the six others identified the six fundamentalemotions.ProcedureThe subjects had to press the “control” key to release the presentation ofthe next slide. They observed the facial stimulus and once their judgment wasmade, they selected by a pressure on one of the six “identification” keys, theword which best described the emotional expression they had seen. Thisresponse and its latency (as measured by the duration between the pressureson the “control” and on the “identification” keys) were recorded. At thebeginning of the experiment, the subjects were familiarized with the task bya series of 10 preliminary trials.RESULTSAs shown in Table 1 the judgments of facial stimuli were highly accurateover-all. Happiness received the highest judgment rates whereas fear receivedthe lowest. These accuracy data were submitted to a two-way analysis ofvariance (sex X emotion) with repeated measures on the last factor. Theanalysis indicated a significant effect of sex (F = 4.24, df = 1/32, p < 0.05)and emotion (F = 16.82, df = 5/160, p < 0.01), but no interaction (F =1.58, d f = 5/160). A measurement of the strength of that effect by u2showed that sex accounted only for 8.70% of the variation in the dependentvariable, while emotion accounted for 31.27% of the variance. Consequentlyit was decided to restrict further analysis to this last factor. The significance ofthe differences was assessed using the Tukey B test. Happiness received betterrecognition levels than the other emotions ( p < 0.01), except for surprise.The latter was judged more accurately than sadness and fear ( p < 0.01). Thejudgment of disgust was also more accurate than sadness ( p < 0.01) and fear( p < 0.05) and finally, anger and sadness had better recognition levels thanfear ( p < 0.01).TABLE 1MEAN ACCURACY OF JUDGMENT(EXPRESSED IN PERCENTAGE)FOREACH EMOTIONAND SEX SexN Happiness Sadness DisgustAnger89.6393.4491.697.67Surprise96.4394.0595.1710.06Fear74.5884.7079.9416.12MalesFemalesTotalu16183498.60100.0099.342.2783.4691.1587.5211.6192.8393.7093.297.34 JUDGMENTOF FACIAL EXPRESSIONS685TABLE 2MEAN LATENCY OF JUDGMENT (IN SEC.) FOREACHEMOTION AND SEX– –Surprise FearSexN Happiness SadnessDisgust4.533.874.181.41Anger5.794.965.352.23Males164.033.023.501.196.264.565.361.414.213.553.861.477.085.186.072.52Females18Totalu34 The latenq of judgment (Table 2 ) was also submitted to a two-wayanalysis of variance which gave similar results. The effect of sex ( F = 6.67,df = 1/32, p < 0.05) and emotion (F = 23.78, df = 5/160, p = 0.001)were significant but there was no interaction between them ( F = 1.54, df =5/160). As in the case of accuracy of judgment, w2 showed that emotionaccounted for a moderately high proportion of the variance (37.60%) whilesex had a relatively limited influence (16.81% of the variance). Tukey B testshowed that happiness was recognized more rapidly than anger, sadness, orfear ( p < 0.01) and that surprise was also identified more rapidly than anger( p < 0.05), sadness ( p < 0.05), and fear ( p < 0.01). Latency of judgment was shorter for disgust than for fear ( p < 0.01) while those of anger,sadness and fear did not differ significantly.TABLE 3PEARSON CORRELATIONSBETWEEN ACCURACYAND LATENCYOF JUDGMENT FOREACHEMOTIONHappiness Sadness Disgust Anger Surprise FearTable 3 presents Pearson ss between accuracy and latency of judgment.Significant negative correlations ( p < 0.01, bilateral) were observed for threeemotions: happiness, disgust and anger. The other three correlations, althoughnegative, were not statistically significant.DISCUSSIONOur results do not confirm previous reports which concluded that womendisplay better performances than men in decoding facial expressions of emotions.There was a significant sex difference but it accounted for a limited proportion of the total variance, more limited for accuracy than for latency of judgment. Our results, however, confirmed a difference, frequently reported, inthe recognition levels of emotions and the fact that the judgment of fear wasless accurate than any other emotion. The latency of judgment also showeddifferences between the various emotions but they did not match exactly thedifferences in accuracy. For instance, even if fear was judged less accurately686 G. KIROUAC & F. Y. DORBthan sadness and anger, latency of judgment was not longer than the latencyof these two emotions.Latencies of 3 to 7 sec, may seem excessive considering the fact that inreal life emotional expressions have to be decoded much more rapidly. Itshould be remembered that our index was not a reaction time but a measureof delay between the onset of the slide and the report of the subject’s judgment. W e thought that less accurate judgments would provoke more hesitation and produce longer latencies, which was partially confirmed by three negative correlations. However, some experiments in our laboratory have recentlyshown that emotional expressions can be recognized with very brief presentation times (20 to 50 msec.) and that the more easily recognized emotions (e.g.,happiness) are accurately identified at shorter exposure times.REFERENCESBOUCHKR, J. D. Culture and emotion. In A. J. Marsella, R. G. Tharp, & T. J.Caborowski (Eds.), Perspectives on cross-cultural psychology. New York: Academic Press, 1979. Pp. 178-203.BOUCHER, J. D., & CARLSON, G. E. Recognition of facial expression in three cultures.Journal of Cross-Cukurul Prychology, 1980, 11, 263-280.EKMAN, P. Pictares of faciul affect. Palo Alto, CA: Consulting Psychology Press, 1976.EKKAN, P., & OSTER, H. Facial expressions of emotions. Annual Review of Psychology.1979, 30, 527-554.EKMAN, P. (Ed.) Emotion in the human face. New York: Cambridge Univer. Press,1982.HALL,J. A. Gender effects in encoding nonverbal cues. Psychological Bulletin, 1978,85. 845-857.IZARD,C . E. The face of emotion. New York: Appleton-Century-Crofts, 1971.I-, C. E. Human emotions. New York: Plenum, 1977.K~ROUAC, G., & DORE,F. Y. Identification des expressions faciales Cmotionnelles parun Cchantillon quCMcois francophones. International Journal of Psychology,1982, 17, 1-7.SHIMODA, K., ARGYLE, M., & RICCI-BITI?,P. The intercultural recognition of emotionalexpressions by three national racial groups: English, Italian, and Japanese. European Journal of Social Psychology, 1978, 8, 169-179.Accepted August 16, 1983.
- Assignment status: Already Solved By Our Experts
- (USA, AUS, UK & CA PhD. Writers)
- CLICK HERE TO GET A PROFESSIONAL WRITER TO WORK ON THIS PAPER AND OTHER SIMILAR PAPERS, GET A NON PLAGIARIZED PAPER FROM OUR EXPERTS