Labels Effect In Category Learning: Evidence From Eye Movements And Event-Related Potential

The category learning is always a part which paid many attention to of cognitive psychology, as well as a vital, energy-consuming problem. Category learning is inextricably linked with general learning. The ability to extend is the central part of learning which get knowledge from living examples to extraneous examples. Label effects largely influences classification learning, but currently there isn't a system to analysis different linguistic labels which take an important role in category learning. Label effects means that labels can promote category learning. Hence this article puts emphasis on two major labels—linguistic labels and situational labels, then attempts to clear their role and the relationship between them, finally, a conclusion has been made in this paper.In chapter 2, there are mainly two traditional types of theories in categorical learning—similarity-based categorical learning and category-based categorical learning. Category effects, especially label effects has been gradually discovered during the argument of two types of theories in category learning, but there isn't a completed system in analysis different linguistic labels. Linguistic labels take an important role in category learning. Four experiments were carried out and participants'eye movements were recorded with a SR Research EyeLink II eyetracker that monitored the position of the right eye every two milliseconds. In experiment 2, there are three labels: similar in appearance and same species, similar in appearance and different species, dissimilar in appearance and same species. In experiment 3,we create four artificial linguistic labels: similar in appearance and dissimilar labels, similar in appearance and similar labels, dissimilar in appearance and similar labels, dissimilar in appearance and dissimilar labels. In experiment 4, we employ four real linguistic labels: similar in appearance and familiar labels , similar in appearance and dissimilar labels, dissimilar in appearance and similar label, dissimilar in appearance and dissimilar label. From four experiments above, we found that in no-label condition category learning is based on an apparent similarity of compared objects, with artificial linguistic labels category learning is based on similarity, with real linguistic labels adult use the category-based category learning. In chapter 3, the similarity-based stimulus were presented at different times to investigate the evidence which mechanism of similarity-based category learning from ERP. With the measurement of ERP, students'EEG was recorded when they sorted these objects. Our results at present times are remarkably different. When presented stimulus longer than 200ms, the P300 was more active, it shows that we process stimulus more profoundly or more far-ranging. The latency of N400 which was activated by the categorical and semantic stimulus is shorter than that of N400 which was activated by the picture. We made a conclusion that linguistic labels promote category learning and semantic stimulus was activated in picture classification.In chapter 4, in order to investigate whether and how the situational information contribute to categorization and inference on objects, three experiments has been performed. We presented different stimulus including situation-based information and information without situation. With the measurement of ERP, students'EEG was recorded when they classified the objects. Our result shows that there are significant difference between two stimulus mentioned above. Situational information improve the accuracy of classification and boost subjects'self-confidence. N2 and P300 was activated more by situation-based information than information without situation, and the latency of it which was activated by information without situation is longer than that of N2 and P300 which was activated by situation-based information. These findings indicate that situational labels about settings and events which helpful to classify objects are stored in memory. The results show that situational labels improve category learning. If the labels available in categorization match situational labels stored in memory about category, the accuracy of categorization will increasing gradually.