Today, the economic and social importance of occupational accidents is undeniable worldwide. Hence, research aimed at reducing this type of accident is considered a discipline of great interest for society in general. In this environment, working conditions play a fundamental role in the occurrence of accidents, and from their study, results can be obtained that provide information for decision-making that guarantee optimum conditions for the development of the employees’ tasks. Organizing the conditions of work execution is also a task that constitutes an essential aspect for a firm’s productivity, therefore, affecting their viability and results. In this work, a model is proposed for the study of different groups of working conditions and their influence on the probability of occupational accidents, in accordance with the data provided by the 7th National Survey of Working Conditions (VII NSWC). The survey sampled 8892 workers active in all sectors of national production and is the last nation-wide survey administered in Spain. Bayesian networks (BNs) are used to generate a network that analyzes working conditions in all areas (27 variables have been included in addition to those corresponding to the sector and accident), and then, more specifically, the relationship that is established between ergonomic factors in the workplace, psychosocial factors of the worker, and the probability of an accident. The results are achieved through the network obtained by highlighting some of the proposed variables. The dependencies generated by the chosen variables are analyzed, and subsequently, the probability of accident for each of the productive sectors is determined. It is concluded that the ergonomic risks associated with physical strains in the workplace, together with the lack of job satisfaction on the employer’s behalf, both pose a very significant increase in the probability of being involved in an occupational accident, above the other variables of study.
