Study Of The Characteristics Of Reasoning Under The Category Uncertain

Category learning is one of the most important aspects of cognitivepsychology. And the research of feature prediction when categorization isuncertain is an important part of category learning. There are theoreticaland practical senses to inspect how people induce when categorization isuncertain.At present the arguments of the feature prediction researches weremainly between single-category theory and rational model. The essentialdifference between single-category theory and rational model was:whether people considered the information of alternative categories whenpredicted unknown features. The single-category theory advanced thatpeople paid more attention to target category during the induction, whilethe rational model advanced people induced according to all of thecategories' information.Murphy and other experts had done a lot of researches and acquiredmany experimental data to support single-category theory since 1994. Butthey found some experimental results were hard to explain bysingle-category theory in their research of 2005. So they began to thinksome prediction maybe base on features association when categorizationwas uncertain which was totally different from category-based prediction.When people predicted based on features association, they didn't considerwhich category the exemplar was belonged to, and the possibility of thepredicted feature in the categories. The more feature associations, theeasier to predict.On the basis of Verde and Murphy's research in 2005, we designedsome new experiments to explore whether people's induction was basedon categories or based on features association, when categorization was uncertain. Only by doing this, we could go further to discuss which onewas right, single-category theory or rational model. So the explorationthat people's induction was based on categories or based on featuresassociation when categorization was uncertain was the most importantthing.Our research was going to explore how people induced in twoparadigms. In one paradigm, exemplars were presented one by one. Inanother paradigm, exemplars were presented at the same time. There werefive experiments: experiment 1 repeated the experiment of Verde andMurphy in Chinese and achieved the same results.Both experiment 2 and experiment 3 explored how people induced inthe paradigm in which exemplars were presented one by one. Experiment2 controlled features association in order to find whether people'sinduction was category-based. The results of experiment 2 indicated inthe paradigm when exemplars were presented one by one, people'sinduction was not category-based.Experiment 3 controlled the factor of category to find whether people'sinduction was based on features association. The results of experiment 3showed in the paradigm when exemplars were presented one by one,people's induction was based on features association.Experiment 4 and experiment 5 discussed how people induced in theparadigm in which exemplars were presented at the same time.Experiment 4 controlled features association in order to find whetherpeople's induction was category-based. The results of experiment 4indicated in the paradigm when exemplars were presented at the sametime, people's induction was not category-based.Experiment 5 controlled the factor of category to find whether people'sinduction was based on features association. The results of experiment 3 showed in the paradigm when exemplars were presented at the same time,people's induction was based on features association.According to all the results of five experiments, we could concludepeople's induction was based on features association in the twoparadigms.