Is 'invisible gorilla' self-reportedly measurable? Development and validation of a new questionnaire for measuring cognitive unsafe behaviors of front-line industrial workers.

The most complicated problem in detecting workplace hazards is the dynamic condition of industrial settings and the unpredictability of workers' behavior. A newly developed method focusing on cognitive differences between individuals is required to evaluate unsafe behaviors of workers. This study aimed at development and validation of a new questionnaire for measuring cognitive unsafe behaviors of front-line industrial workers. A new questionnaire with 61 items was developed and the main measurement characteristics (validity and reliability) were analyzed. Content validity analysis showed that 61 items had an excellent content validity index (<0.78) and content validity ratio (>0.42). Good internal consistency (Cronbach's α = 0.95) and stability (intra-class correlation coefficient = 0.98) were found for the new instrument. The result of confirmatory factor analysis indicated that the selected model was satisfactory. The new instrument appears to be a valid and reliable tool to assess cognitive unsafe behaviors of front-line industrial workers.