Performance shaping factors dependence assessment through moderating and mediating effect analysis

Dependence assessment is an important task in human reliability analysis (HRA). The dependences in HRA include three categories, dependences between human failure events (HFEs) (CAT1), dependences between HEP and performance shaping factors (PSFs) (CAT2), and dependences between PSFs (CAT3). Among them, CAT3 dependences have not been well studied due to their complexity. According to their function mode, PSF dependences can be further distinguished as mediating effect and moderating effect. Taking the inter-dependences between PSFs for civil flight crew as examples, this paper proposes an empirical approach to assess dependences between PSFs. Firstly, 10 PSF categories including 59 lower level PSFs were identified for civil flight crew through literature review and focus groups screening. Then a 432-pilot-joined survey was carried out to collect pilots' opinions on PSF influences on performance. After that, moderating and mediating effect analysis methods were used to identify the existence and type of PSF dependences. The results show that mediating effect is much more common than moderating effect between PSFs. The method proposed in this paper can provide a deeper insight to PSF dependences, and its results serve as important inputs to other HRA researches such as the construction of complex Bayesian networks for HRA.