Evaluation System of Human-Factor Training Effect Based on Hierarchical Model

Abstract Nuclear power enterprises attach great importance to staff training and have invested a lot of resources to establish training and authorization systems based on Systematic Approach Training (SAT), including demand analysis, outline design, material preparation, training implementation and evaluation feedback five parts. However, the lack of effective training effect evaluation means in practice makes it difficult to continuously improve the training system. This paper takes human-factor training in nuclear power plants as an example, analyzes the characteristics of human factors training, such as complex training object, rich content and diverse methods, compares the Kirkpatrick model, Context, Input, Process, Product (CIPP) model and Context, Input, Reaction, Output (CIRO) model. The kirkpatrick four-layer model is selected as the theoretical basis to establish the evaluation system for the effect of nuclear power human factor training, including a closed-loop training evaluation process, five evaluation methods and corresponding tools. After a period of practice in a nuclear power operation team, this system has been proved to be able to achieve effective evaluation and continuous improvement of the human-factor training effect in nuclear power plants, and effectively improved the long-term performance of the operation team, the non human-factor mistake record has been more than 900 consecutive days. In view of the problems found in the system practice, the paper also puts forward suggestions for further improvement.