Abstract:

Present study aimed to explore the priority issues between expression and gender information processing in face perception and further to reveal the laws of the human brain processing multidimensional stimuli. Three behavioral experiments were implemented. And each one included two types of task set (Facial Expression judgment task and Facial Gender judgment task) randomly assigned for the stimuli. The number of trials in different task was the same. The color (green or red) of image border gave the participants the clue to perform which task. Furthermore, there were four (in Experiment 1 and 3: 12ms, 50ms, 250ms and 500ms ) or five (in Experiment 2: 12ms, 50ms, 100ms, 150ms, 250ms) kinds of duration time for stimuli presentation in each task set. E-prime was adopted to compile the program and collect data. The stimuli presented on 17 inches of computer screen with 85 Hz refresh rate. Repeated measures ANOVA was used to analyze the data of reaction time and accuracy with the kind of duration time (Duration Time ) and the type of task (Task) as within factor tasks. Experiment 1 shows that no significant results on reaction time. But there are two significant main effects (Duration Time and Task) on accuracy analysis. That indicates making judgment on expression is more accurate than on facial gender (t(29) = 4.742, p <. 001) in 12 ms conditions. And so does on the condition of 50 ms (t (29)= 2.256, p <. 05). But no similar findings are found in the other duration. Experiment 2 reveals that the reaction time is inverse proportion with the duration time of stimuli in the conditions of 50ms, 100ms and 250ms (p <. 05). Then the results of accuracy analysis show significant main effect of Duration Time and significant interaction between Duration Time and Task. Further analysis indicates that the accuracy of judging expression is significantly higher than that of facial gender (t(29) = 3.647, p <. 05) under the condition of 50 ms, however, the comparison for the other time windows do not reach significant difference. Experiment 3 shows that the main effect of Duration Time is significant on reaction time analysis. Analysis on accuracy finds out significant main effect of Duration Time, significant main effect of Task and significant interaction between these two factors. In summary: (1) the accuracy of the condition 12ms is the lowest among the four kinds of duration time; and the accuracy under the 50ms condition is lower than that of 200ms and 500ms conditions. (2) judging expression is more accurate than judging facial gender under the conditions of 12ms (t = 3.647, p <. 01) and 50ms (t = 2.103, p <. 05). In conclusion, the study revealed that the efficiency of expression recognition has advantage over that of facial gender judgment in early automatic processing stage. And in the automatic stage of facial information processing, the processing extent is proportional to the duration time of stimuli presentation.

摘要：

为了探讨面孔知觉时表情或性别信息加工优先性的问题，要求被试对不同时长窗口中呈现的一系列面孔刺激的性别或者表情做出判断。两类任务的排列设置分为完全随机（实验一、二）和随机区组（实验三）。结果一致发现在面孔刺激呈现时间为12ms和50ms时，面孔表情信息识别比面孔性别信息识别更准确。由此推论，面孔知觉中面孔表情信息加工优先于面孔性别信息加工。