Trait confirmation and disconfirmation: the development of attribution biases.

In the present study, a new method for conducting person perception research, the confirmability paradigm, was adapted for use with children. In the confirmability paradigm subjects are asked how many behaviors they require before attributing positive and negative traits to other people. This new procedure was used to study developmental differences in the amount of information required for trait inference, as well as negativity and egocentric biases in trait attribution. Subjects were 70 children in grades 3 through 5 and 46 college students. The study revealed that traits differ in the amount of behavioral information required for their inference. In addition, a developmental decrease in attribution criteria was detected for various traits. Evidence for a negativity bias in the attribution of several traits was also obtained. Subjects exhibited an egocentric bias in their trait ratings but not in their attribution criteria. The utility of this new approach for the study of person perception is discussed.