تشخیص ابراز هیجانی در نمایش صورت به سرعت در حال تغییر در با اضطراب اجتماعی زنان بالا و پایین

Abstract Facial information and attention to facial displays are distributed over spatial as well as temporal domains. Thus far, research on selective attention to (dis)approving faces in the context of social anxiety has concentrated primarily on the spatial domain. Using a rapid serial visual presentation (RSVP) paradigm, the present study examined the temporal characteristics of visual attention for happy and angry faces in high- (n=16) and low-socially anxious individuals (n=17), to test whether also in the temporal domain socially anxious individuals are characterized by threat-confirming attentional biases. Results indicated that presenting angry faces as the first target (T1) did not aggravate the detection of the emotional expression of the second target (T2). Yet, participants generally showed superior detection of the emotional expression of T2, if T2 was an angry face. Casting doubt on the role of such attenuated attentional blink for angry faces in social anxiety, no evidence emerged to indicate that this effect was relatively strong in high-socially anxious individuals. Finally, the presentation of an angry face as T2 resulted in a relatively hampered identification of a happy-T1. Again, this “backward blink” was not especially pronounced in high-socially anxious individuals. The present anger superiority effects are consistent with evolutionary models stressing the importance of being especially vigilant for signals of dominance. Since the effects were not especially pronounced in high-anxious individuals, the present study adds to previous findings indicating that socially anxious individuals are not characterized by a bias in the (explicit) detection of emotional expressions [Philippot, P., & Douilliez, C. (2005). Social phobics do not misinterpret facial expression of emotion. Behaviour Research and Therapy, 43, 639–652].

Results Interference effects of T2 on correct identification of T1 Mean number of correct identifications of T1 is shown in Table 1 as a function of the emotional expressions of T1 and T2, Lag, and level of social anxiety. Table 1. Mean (SE) percentage of correct identifications of T1 as a function of the emotional expression of T1 and T2, Lag, and social anxiety High anxious Low anxious Angry-T2 Happy-T2 Angry-T2 Happy-T2 Lag 2 Angry-T1 91.0 (2.3) 84.0 (4.0) 98.2 (2.3) 93.1 (3.9) Happy-T1 78.5 (4.0) 93.8 (2.1) 86.4 (3.9) 93.8 (2.1) Lag 3 Angry-T1 90.6 (2.5) 89.9 (3.0) 95.9 (2.4) 93.8 (2.9) Happy-T1 88.7 (3.4) 92.2 (2.4) 83.8 (3.3) 94.5 (2.4) Lag 8 Angry-T1 92.2 (2.1) 85.2 (3.4) 95.3 (2.1) 94.2 (3.3) Happy-T1 82.4 (3.7) 88.2 (2.6) 92.3 (3.6) 93.8 (2.6) Table options A 2 Emotion-T1 (happy vs. angry)×2 Emotion-T2 (happy vs. angry)×3 Lag (2 vs. 3 vs. 8)×2 Group (high vs. low socially anxious) ANOVA revealed no significant main effect of Group indicating that overall both groups showed a similar accuracy in identifying T1 (F(1,31)=2.20, p=0.15, η2=0.07). There was no main effect of Lag (F(2,30)=1.33, η2=0.08), indicating that correct identification of T1 was independent of the time-lag between T1 and T2. Most important for the present context, there was a significant main effect of Emotion-T2, indicating that participants were relatively inaccurate when T2 displayed an angry face (F(1,31)=7.26, p=0.01, η2=0.19), mean percentage of accurate identification being 89.6% and 91.3% for the angry- and happy-T2, respectively. This effect appeared independent of social anxiety (F(1,31)<1, η2=0.002). It was most pronounced for Lag 2 and virtually absent for Lag 8 as is evidenced by a borderline significant Emotion-T2 by Lag interaction (F(2,30)=3.18, p=0.05, η2=0.18). Accordingly, posthoc t tests indicated that the influence of T2 was significant for Lag 2 (t(32)=2.67, p<0.05), but not for Lag 3 (t(32)=1.7, n.s.), nor for Lag 8 (t(32)<1). The interfering influence of an angry-T2 was especially pronounced for trials depicting a happy-T1 as was evidenced by a significant Emotion-T1 by Emotion-T2 interaction (F(1,31)=32.88, p<0.01, η2=0.52). Subsequent analyses indicated that there was no difference between congruent trials (i.e., happy-T1 and happy-T2 vs. angry-T1 and angry-T2) (t(32)=1.11, p>0.1), whereas a significant difference was evident between both types of incongruent trials (happy-T1 and angry-T2 vs. angry-T1 and happy-T2) (t(32)=2.96, p<0.05), indicating that correct identification of T1 was especially hampered when an angry-T2 followed a happy-T1. Finally, there was a main effect of Emotion-T1 (F(1,31)=5.67, p<0.05, η2=0.16), indicating that, overall, participants were more often correct when T1 displayed an angry than when T1 displayed a happy facial expression, which means being 91.8% and 88.7%, respectively (see also Table 1). Attentional blink Mean percentage of correct identifications of T2 (given accurate identification of T1) are displayed in Table 2. Table 2. Mean (SE) percentage correct identifications of T2 as a function of the emotional expression of T1 and T2, Lag, and social anxiety High anxious Low anxious Angry-T2 Happy-T2 Angry-T2 Happy-T2 Lag 2 Angry-T1 14.1 (6.0) 4.7 (3.8) 25.9 (5.8) 9.1 (3.7) Happy-T1 9.1 (5.5) 14.1 (5.8) 23.5 (5.3) 24.0 (5.6) Lag 3 Angry-T1 42.4 (7.7) 14.1 (6.5) 59.7 (7.4) 22.3 (6.3) Happy-T1 17.8 (6.8) 29.2 (7.9) 34.9 (6.6) 43.5 (7.7) Lag 8 Angry-T1 83.2 (3.9) 78.4 (4.8) 94.2 (3.8) 86.2 (4.6) Happy-T1 81.8 (4.9) 84.2 (3.9) 90.9 (4.8) 88.7 (3.7) Table options A 2 Emotion-T1 (happy vs. angry)×2 Emotion-T2 (happy vs. angry)×3 Lag (2 vs. 3 vs. 8)×2 Group (high vs. low socially anxious) ANOVA revealed a no significant main effect of Group indicating that there was no convincing difference between high- and low-socially anxious individuals’ ability to identify the emotional expression of T2 (F(1,31)=3.00, p=0.09). In addition, there was a main effect of Lag (F(2,30)=234.9, p<0.05, η2=0.94), indicating that participants showed an attentional blink when the time-lag between T1 and T2 was small (Lags 2 and 3), whereas the blink almost disappeared when the time-lag was large (Lag 8) (see Table 2). Most important for the current context there was a main effect of Emotion-T2 (F(1,31)=23.71, p<0.05, η2=0.43). As can be seen in Table 2, this effect indicates that, overall, the percentage correct identifications were larger for angry (M=48.1%) than for happy (M=41.5%) expressions. This angry superiority effect varied as a function of lag as is evidenced by significant Emotion-T2 by Lag interaction (F(2,30)=6.20, p<0.05, η2=0.29). As can be seen in Table 2, the effect was relatively large for Lag 3, and similarly small for Lags 2 and 8. Meanwhile, posthoc t tests indicated that for all Lags the effect was significant. For Lag 2 (t(32)=2.4, p<0.05), for Lag 3 (t(32)=4.9, p<0.05), and for Lag 8 (t(32)=2.1, p<0.05). In addition, there was a tendency suggesting that the anger superiority effect was most pronounced in low-anxious participants, yet the Emotion-T2 by Group interaction did not reach the conventional level of significance (F(1,13)=3.8, p=0.06, η2=0.11). Furthermore, there was evidence of a congruency effect: accurate detection of T2 was enhanced when a congruent emotional expression was displayed on T1 (F(1,31)=59.97, p<0.05, η2=0.66). This congruency effect was similar for both groups (F(1,31)<1, η2=0.01). Finally, there was no main effect of Emotion-T1 (F(1,31)<1, η2=0.01), indicating that the type of facial expression presented during T1 did not generally influence the magnitude of the attentional blink.