There has been many researches about conjuction fallacy, which people tend to evaluate conjunctive statement more probable than component statements. Conjunction Fallacy shows that people obviously violate probability rules, which is conjuctive state cannot be higher than component statement. However, we in our experiment, we tried to explain this fallacy by using conditional probability. If people used bayesian updating rule, then the fallacy cannot be the evidence of people's irrationality. In our result, people showed some pattern that they had used bayesian updating rules, but the result was not statistically significant.