the room but is carried away through a pipe by means of suction. Accident prevention work should be guided by experience. Otherwise it is likely to be misdirected, for intuitive conclusions on the importance of various hazards and the efficacy of methods of prevention are subject to a high degree of inaccuracy. Small plants must rely on the indications of statistical experience FIG. 4.-Improper and proper cutting heads for planer. gathered from large plants or groups of plants, but a large establishment may find dependable indications in its own record. An analysis of accidents by cause is of great use in formulating plans for accident prevention. One result of many such analyses is a demonstration of the fact that the less spectacular hazards are often the more serious. The diagram below (Fig. 6), based on the experience of a large plant, shows the relative importance of various causes of accidents in terms of accident frequency.1 While an analysis of accident frequency is serviceable to some degree, it is much less significant than an analysis of accident severity, measured in terms of time lost. The results of accidents are of more significance than their mere occurrence. An analysis of relative severity by causes will indicate which causes deserve closest attention. Fig. 7, which is based on the experience used for Fig. 6, presents such an analysis. Measured in terms of severity, the relative importance of the various causes shows marked changes. For example, the 1 This and the following diagram were prepared for a manufacturer by his insurance company as indications of the conditions calling for preventive attention in his plant. apparent importance of "machinery" is reduced, while that of "handling objects" is greatly increased. Much more might be written to show the need of careful study and of reliance on statistical and engineering experience. For example, certain causes yield much more readily to prevention work than do others. The probable relative results of a given amount of time and effort should be given due weight. The extent to which safeguarding induces carelessness and the relative efficacy of different methods of attacking the same hazard are likewise fruitful subjects for consideration. Accident prevention is a scientific problem worthy of the application of scientific method. APPENDIX IV-A RELATIONSHIP BETWEEN NEW EMPLOYEES AND ACCIDENT FREQUENCY The chart below illustrates the close relationship between the number of accidents in a given plant during the period of one year, and the number of new employees. These new employees are often not properly trained in the operation of machines, and are more likely to be injured than are older and experienced employees. Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 1922 Months 1923 40 20 0 |