Boosted neural network modeling of psychological and social factors of work affecting safety performance and job satisfaction in the process industry.

Psychological and social factors of work were found to influence workers' safety performance and job satisfaction. This study aimed to assess the effects of psychological and social factors of work affecting safety performance and job satisfaction of control room operators in a process industry in Iran. The general Nordic questionnaire for psychological and social factors at work containing 99 items covering 13 factors was used to collect data on psychological and social factors at work. A hybrid intelligent approach was developed in three stages to model the obtained data. In the first stage, the Haar wavelet transform was used to extract the level of each factor. In the second stage, a boosted neural network (BNN) was used to model safety performance and job satisfaction based on extracted factors in the first stage. In the third stage, the marginal model plots obtained by the BNN model were applied to characterize the marginal relationships between each psychological and social factor of work, safety performance, and job satisfaction. The findings from the marginal model plots, which assessed the relative importance of psychological and social factors at work, indicated that leadership was the most important factor affecting safety performance of operators. Additionally, the BNN model results suggested that among all psychological and social factors, group work, predictability at work, and work motives strongly influenced safety performance. Regarding job satisfaction, the marginal model plots revealed that organizational culture was the most influential factor. The BNN model further indicated that mastery of work and group work played a key role in job satisfaction among all psychological and social factors examined.