NIOSH 1982 Feb:13-17
The need for a systematic evaluation, or model, of safety research was discussed. It is suggested that the systems development concept and the criteria used in systems design should be applied to industrial safety. The author points out that data analysis is an absolute requirement for utilization of accident data, but that individual human judgment has great implications not only for the coding of data but also for the design of accident report forms and the precision of data extraction. Statistical methods such as multivariate and multiple regression analyses should be used much more than they are now. Many studies in the literature are too small and are characterized by uncontrolled observations and simple summary statistics. In data collection, a broader approach is recommended, and many different types of data should be collected. For example, reports on near injuries should be evaluated with regard to preventive measures that could be instituted. It is also considered important to develop an accident injury correlation matrix which correlates types of injury with specific body parts, and to relate accident data to work performance evaluations. Finally, the author stresses the importance of designing a new accident injury report form to replace NASI Z-16.
Accident-analysis; Accident-statistics; Data-processing; Safety-research; Work-performance; Humans; Injuries; Human-factors-engineering; Industrial-safety
DHHS (NIOSH) Publication No. 82-103
Proceedings of the Symposium on Occupational Safety Research and Education, January 1981, NIOSH, U.S. Department of Health and Human Services, DHHS (NIOSH) Publication No. 82-103