The Role of Knowledge, Skill and Ability in the Prediction of Unsafe Behaviors in the Staff of an Industrial Company

Aims: Previous research have shown many organizational factors as determinants of unsafe behaviors. One of the most prominent is the lack of adequate knowledge, skill and ability in the field of safety. So, the aim of the present study was to examine simple and multiple relationships between safety knowledge, safety skill and safety ability with unsafe behaviors in an industrial company.
Methods: The population of this descriptive – correlation study included all line employees who work in an industry company (N= 1160). Among this population, 265 employees were selected by a stratified random sampling method. Participants in this study were measured by using the unsafe behaviors questionnaire of Hofmann & Stetzer (1996) and the Salleh (2010) safety competency scale. In order to analyze the data, the Pearson’s correlation coefficient and regression analysis (by use software SPSS v19) and in the validation stage, the confirmatory factor analysis (by use software AMOS-21) were applied.
Results: Findings indicate that the correlation coefficients between safety knowledge and unsafe behaviors were significantly negative (r=-0.259, P<0.0001). Also, the correlation coefficients between safety skill and unsafe behaviors were significantly negative (r=-0.411, P<0.0001). Likewise, the correlation coefficients between safety ability and unsafe behaviors were significantly negative (r=-0.323, P<0.0001). The stepwise multiple regression analyses among the three independent variables, revealed that only two variables of safety skill and safety ability qualified for the inclusion criteria of the regression equation (R2=0.181, P<0.0001).
Conclusion: The results of this study showed the importance of safety skill and safety ability as predictors of unsafe behaviors. Therefore, in order to select people for high risk environments such as the petrochemical industry and also in training courses we should rely on these variables.