The Effect of Intuitively Adjust Deviation Biases on Bayesian Reasoning

Shi Zifu,Zhang Qinglin
DOI: https://doi.org/10.16719/j.cnki.1671-6981.2011.04.037
2011-01-01
Abstract:Kahneman and Tversky(1972) assumed that people's intuitive reasoning was far from Bayesian Theorem,even they didn' t follow the rule at all;Gould(1992) suggested that people reasons not following the rule by nature;Slovic and Lchtenstein(1971), however,considered that people's intuitive reasoning was almost coincident with the result of computing in affirmation or negation direction,only precision was inadequate;Peterson and Beach(1967) thought also that,human being as a intuitive statistician,his probability's reasoning way was commonly similar with compute process of Bayesian rule.So,what caused so different conclusion, and what effect Bayesian reasoning? On the basis of the findings above,this study aimed to explore cognitive biases which existed in the process of Bayesian reasoning,and to analyse the effects of these biases on Bayesian reasoning.Besides,this study investigated the extent of biases in different task.The present study also discussed the difference on reasoning results of participants majored in arts or science caused by their difference on mathematical logic capability. The experiment was collectively conducted with totally 168 arts and science students,among which 79 students majored in commerce and 89 in Chinese and History.All participants were not familiar with Bayesian inference rule.168 questionnaires were collected respectively.This research adopted 'the study and explore department problem','the office problem' and 'the model workshop problem' as experimental materials,and the tests for the extent of 'availability to false report rate' in Bayesian reasoning task was done.Meanwhile,a variance analysis of the absolute value of difference between posterior probability and normal value was done. The results indicated that:(1) Available heuristic was closely related with 'intuitively adjust deviation' bias in the process of probability's evaluation;(2) The 'anchoring reference error' bias and the 'intuitively adjust deviation' bias could all be found among arts and science students;(3) There was remarkable negative correlation between the evaluation of post-probability and the subjective probability of proposition including false report bate. The results of this study suggested that heuristic cognitive bias in the process of people's Bayesian reasoning affect the precise of probability's evaluation.It also indicated that the extent of bias was different with reasoning task,the bias of participants majored in arts or science had no difference.All these findings offered psychological evidence to realize,find out and alarm common mistakes in the process of people's Bayesian reasoning so as to learn the skill of probability's thinking and improve ability of Bayesian reasoning.
What problem does this paper attempt to address?