کاهش یا بهبود؟: تفاوتهای مرتبط با سن در بازشناسی حالت چهره

Abstract We examined age-related differences in facial expression recognition in association with potentially interfering variables such as general cognitive ability (verbal and visuospatial abilities), face recognition ability, and the experiences of positive and negative emotions. Participants comprised 34 older (aged 62–81 years) and 34 younger (aged 18–25 years) healthy Japanese adults. The results showed not only age-related decline in sadness recognition but also age-related improvement in disgust recognition. Among other variables, visuospatial ability was moderately related to facial expression recognition in general, and the experience of negative emotions was related to sadness recognition. Consequently, age-related decline in sadness recognition was statistically explained by age-related decrease in the experience of negative emotions. On the other hand, age-related improvement in disgust recognition was not explained by the interfering variables, and it reflected a higher tendency in the younger participants to mistake disgust for anger. Possible mechanisms are discussed in terms of neurobiological and socio-environmental factors.

Results Fig. 1 provides the mean number of correct identifications by age and emotion in the task of Facial Expression Identification. An analysis of variance was conducted on the number of correct identifications of the six basic emotions with two factors of age and emotion. The results showed a marginally significant main effect of age (F(1, 66) = 11.669, p = 0.078) and a significant main effect of emotion (F(5, 330) = 66.802, p < 0.001). Importantly, the main effects were qualified by a significant interaction between age and emotion (F(5, 330) = 5.000, p < 0.001). Therefore, the simple main effects of age were examined for respective emotions, indicating a significant age-related improvement for disgust (F(1, 396) = 8.768, p = 0.003) and a significant age-related decline for sadness (F(1, 396) = 10.207, p = 0.002). Age-related declines for surprise and anger (for both, F(1, 396) = 2.939, p = 0.087) were only marginal, and age-related differences were insignificant for happiness and fear (p > 0.1). Mean number (±S.E.) of correct identifications in the task of Facial Expression ... Fig. 1. Mean number (±S.E.) of correct identifications in the task of Facial Expression Identification by age and emotion. HA: happiness, SU: surprise, FE: fear, AN: anger, DI: disgust, SA: sadness. Figure options The performances of the older and younger participants in the other four tasks are summarized in Table 1. The scores the two WAIS-R subtests were consistently lower in the older participants than in the younger participants, and statistically, the age-related decline was marginal in Information (t(66) = −1.764, p = 0.082) and was highly significant in Picture Completion (t(66) = −4.563, p < 0.001). The results probably reflect steeper age-related decline in visuospatial ability than in verbal ability ( Howieson et al., 1993 and Kaufman et al., 1989). The age-related difference was also significant in Facial Identity Matching (t(66) = −3.245, p = 0.002), indicating age-related decline in face recognition ability in general. With regard to General Affect Scales, age-related decrease in Negative Affect (t(66) = −2.713, p = 0.008) as well as the maintenance of Positive Affect (t(66) = 0.264, p = 0.793) was observed, which is consistent with the earlier literature ( Carstensen and Charles, 1998 and Mroczek, 2001). Table 1. Means (±S.E.) for the scores of two WAIS-R subtests, Facial Identity Matching, and General Affect Scales Older Younger WAIS-R subtests Informationa 17.85 ± 0.75 19.65 ± 0.69 Picture Completiona 11.53 ± 0.56 14.53 ± 0.34 Facial Identity Matching 23.35 ± 0.60 25.76 ± 0.44 General Affect Scales Positive Affect 26.03 ± 0.71 25.76 ± 0.70 Negative Affect 16.00 ± 0.77 19.12 ± 0.85 a Raw scores. Table options Table 2 shows correlations between the number of correct identifications of respective emotions in Facial Expression Identification on one hand and the remaining five measures on the other hand for each age group. Happiness and surprise were excluded from the correlation analyses because ceiling effects and poor variances are indicated in Fig. 1. The score of Picture Completion had the most consistent relationship with the performances in Facial Expression Identification across age and emotion. Specifically, in the younger participants, the positive correlations between the two tasks were significant or marginally significant for any emotion, whereas in the older participants, all the correlations were positive but reached the significance level only for disgust. As compared with Picture Completion, Information appeared to have a less marked relationship with Facial Expression Identification. Likewise, the correlations between the score of Facial Identity Matching and performances in Facial Expression Identification were not significant for any emotion. With respect to emotional experiences, the positive correlation between Negative Affect and Facial Expression Identification was noted only for sadness, and it reached the significance level in the younger participants. Table 2. Correlations between performances in Facial Expression Identification and the other five measures FE AN DI SA Older Information 0.160 −0.217 0.284 0.223 Picture Completion 0.137 0.212 0.378* 0.050 Facial Identity Matching 0.226 0.134 0.095 0.167 Positive Affect −0.129 0.186 0.119 −0.165 Negative Affect −0.050 0.004 −0.081 0.242 Younger Information 0.199 −0.087 0.159 0.047 Picture Completion 0.411* 0.289† 0.381* 0.543** Facial Identity Matching −0.168 −0.009 0. 141 0.101 Positive Affect 0.069 0. 058 −0.124 −0.003 Negative Affect −0.009 0. 150 −0.018 0.340* Note: FE: fear, AN: anger, DI: disgust, SA: sadness. †p < 0.10. *p < 0.05. **p < 0.01. Table options In order to clarify the unique contribution of age to disgust and sadness identification, regression analyses were conducted on the number of correct identifications of disgust or sadness as the dependent variable; the scores of Information, Picture Completion, Facial Identity Matching, Positive Affect, Negative Affect, and age (dummy variable; older = 1, younger = 0) were entered as the independent variables. Table 3 summarizes the results for the regression analyses. With regard to disgust, the addition of age resulted in a significant increment in R2 (ΔR2 = 0.144, F(1, 61) = 11.110, p = 0.001), indicating the unique contribution of age. In line with Table 2, the score of Picture Completion was another significant predictor of disgust identification. On the other hand, the addition of age did not result in a significant increment in R2 for sadness (ΔR2 = 0.004, F(1, 61) = 0.361, p = 0.550). Table 3 indicated that the score of Negative Affect was a significant predictor of sadness identification, as was expected from Table 2. Table 3. Summary of regression analyses on the number of correct identifications of disgust or sadness Disgust Sadness B S.E. B β B S.E. B β Information 0.080 0.064 0.154 0.029 0.052 0.068 Picture Completion 0.242 0.098 0.337* 0.119 0.079 0.199 Facial Identity Matching 0.089 0.085 0.131 0.055 0.069 0.098 Positive Affect 0.004 0.064 0.008 −0.065 0.052 −0.144 Negative Affect −0.038 0.056 −0.085 0.111 0.046 0.298* Age 2.081 0.624 0.473** −0.302 0.503 −0.083 R2 = 0.210 (p = 0.021) R2 = 0.254 (p = 0.005) Note: All the independent variables were forcibly entered into the regression model. *p < 0.05. **p < 0.01. Table options We then analyzed whether some age-related differences in erroneous identification underlay age-related improvement in disgust identification. A trivial possibility was that the older participants’ preferential use of the disgust label might lead to an apparent improvement in disgust identification (Calder et al., 2003). Indeed, the number of older participants misusing the disgust label was marginally significantly larger than that of the younger participants (older, M = 3.88, S.E. = 0.38; younger, M = 2.91, S.E. = 0.38; t(66) = 1.811, p = 0.075). However, the correlation between the number of correct identifications of disgust and the number of such misuse was neither significant nor positive (r = −0.139, t(66) = −1.140, p = 0.259). Another possibility was that the younger participants might tend to make some specific errors in identifying facial expressions of disgust. Table 4 summarizes the frequencies of erroneous identifications mistaking facial expressions of disgust to be other emotions.2Table 4 reveals that the younger participants misused the anger label almost twice as often as the older participants did, and statistically, the mean of such errors was significantly larger in the younger participants (older, M = 1.53, S.E. = 0.25; younger, M = 2.79, S.E. = 0.42; t(66) = −2.568, p = 0.012). As a natural consequence, the number of correct identifications of disgust and the number of erroneous identifications as anger were negatively correlated almost perfectly (r = −0.980, t(66) = −40.008, p < 0.001). Table 4. Frequencies of erroneous identification of facial expressions of disgust as other emotions HA SU FE AN SA Older 0 1 5 52 2 Younger 0 0 1 95 2 Note: HA: happiness, SU: surprise, FE: fear, AN: anger, SA: sadness.