Font Size: a A A

The Impact Of Information Formats And Contents Of Problem On The Teenagers' Bayesian Inference

Posted on:2011-12-18Degree:MasterType:Thesis
Country:ChinaCandidate:Z Z PanFull Text:PDF
GTID:2155330338488964Subject:Development and educational psychology
Abstract/Summary:PDF Full Text Request
Bayesian reasoning is an important project of probability inference field, but also a traditional and hot thesis in the field of thinking and problem solving in cognitive psychology. Bayesian method has great potential in psychological research. The studies of Bayesian reasoning are helpful to reveal the people's cognitive processes and processing rules of probability inference, so as to provide a theoretical basis and practical guidance for improving people's thinking activities ultimately such as probability judgments, reasoning and decision-making.In all studies on Bayesian reasoning, there has been the hot focus on natural frequency effect and content effect, but few developmental researches of young people on Bayesian reasoning. Purpose of this study is to explore the impact of information formats and contents of problem on the China Youth Bayesian inference, and to have further understanding on youth development mechanism of Bayesian reasoning by studying trend of Bayesian inference of young people from the development perspective.The classical mammography problems and winning tourism issues were used as the experimental materials in this study. The students of sixth grade in elementary school, second in junior middle school, second in high school and second in college were subjects . Posterior probability estimates, the accuracy of posterior probability estimates and frequency distribution of used strategies were tested as the dependent variable, using 4×2×2 mixed design. Purpose of this study was to study the impact of year, information format and content on the Bayesian reasoning, and to explore the complex cognitive process of Bayesian inference by analyzing the subjects' problem solving strategies. The main conclusions of the study were as follows:1. When the information format was changed from probability into natural frequency, the subjects' probability estimate value of Bayesian inference reduced significantly, accuracy increased significantly, and also much more accurate Bayesian inference strategies were used. Therefore, the natural frequency format was more suitable for Bayesian inference than other format.2. The content of the problem may affect probability estimate value significantly, but it didn't affect the accuracy of probability estimate. It illustrated that people may adjust overestimated probability value on different levels based on the contents of the problem in the process of Bayesian inference.3. In the probability format, ability and accuracy of Bayesian reasoning had apparent improvement with rising year growth from sixth grade in elementary school to second in high school sophomore (the sixth grade and other grades had significant differences). In particular, subjects of high school sophomore had best Bayesian performance, but the sophomore had emerged decline. However, in the natural frequency format, there were no significant differences between subjects of various grades and no trend of Bayesian reasoning gradually improving with grades.4. People's strategies of problem solving had a certain impact on the Bayesian inference process. Problem of Bayesian inference was solved by the Bayesian strategy, non-Bayesian strategies and other strategies. Three non-Bayesian strategies used widely were joint event strategy, evidence-only strategy and conservative strategy. Few subjects used Fisher strategy. Strategies to solve the problems of Bayesian inference by subjects in different grades were significantly different.
Keywords/Search Tags:Bayesian reasoning, information format, content of the problem, development trends, strategy
PDF Full Text Request
Related items