In the redundant system, common cause failure (CCF) can result in failures of more than one module. Human error is one of the most important causes among all the causes of CCF. Considering the diversity of human factors is an effective method against CCF due to human error. The current methods of human reliability analysis (HRA) almost never take into account the impact of the diversity of human factors on system reliability. Therefore this paper proposes a new method which can quantitatively describe the impact of the diversity of human factors on system reliability from the perspective of component degradation level. The key idea of the new method is to build a fuzzy logic model based on an extended CREAM for the estimation of the probability of an erroneous human action which is then mapped into component degradation valve (CDV).The paper also studies the relationship between CCF rate and the diversity of human factors by performing CCF analysis based on impact vectors which can be calculated from CDV. Finally, the paper presents an example to illustrate the effectiveness of the novel method.