Abstract: In order to overcome the shortcomings of traditional assessment method of situational awareness (SA) reliability, a more robust SA causality model was built in this paper. Firstly, the organization-oriented SA error analysis framework or method was adopted to analyze nuclear power plant incidents and 132 samples were obtained. Then, the correlation analysis method was used to identify the correlation relationships between factors of influencing SA and factor analysis method was used to identify the scenes triggering SA errors, including operator’s mental level, operator’s work attitude, stress level and system situation display level. Finally, based on the above analysis results, a data-driven SA causality model was established. The results show that the data-based SA causality model can identify the scenes triggering SA errors. It is useful to improve the accuracy of quantitative assessment of SA reliability because of considering the causality relationships of performance shaping factors.