Document VkazoGvnB54pQB8LZ9eXryv4

11--Human Behavior and Safety ------ 39 7%-- JJ ---68 3%--* M M liR mt 811 -- 3SD --1SD +1SD +3SD --2S0 OSD +2SD Scores, standard deviation (SD) units Fic 11--1 --The distribution of scores in a normal curve Since the normal curve has a known shape, it is possible to state the percentage of scores that lie between +1 and --1 standard de viation, or between any other two points expressed in SD units This normal curve might represent the distribution of IQ scores in the United States, m this case, the mean would be 100 and the standard deviahon would be approximately 15 Given the mean and standard deviation, therefore, the complete distnbuhon of scores in a normal distribution can be known other peison (although there are certain com mon characteristics) The fact that individual differences exist should not be news What is important is how one deals with individual differences Fig 11-1 presents the distribution of scores in a normal curve Many human traits (for example IQ) are assumed to be distributed ac cording to the normal curve distribution Note that in such a distribution, half of the persons are below the mean (the anthmetical average) and half of the persons are above the mean While the trait is relatively systematic and the extremes are relatively rare, neveitheless onlv a portion of the population has an average amount of the trait Often the manager of an enterprise or direc tor of a safety program realizes that individ ual differences exist and that it is physically impossible to handle every one individually Therefore, he will do what he feels is the next best thing--appeal to the average-man There 242 is, unfoitunately, no such thing as an average- man By definition, average means that a measured trait is possessed by 50 percent and absent in the remaining 50 percent of those measured Therefore, when an appeal is aimed at the average, by definition it misses much of the population A better approach is the use of percentiles When designing the system, instead of setting the standards to fit the average, design it to fit all but the upper 5 percent and the lower 5 percent Then the system will fit 90 per cent of the population The use of percentiles, rather than averages, has long been the technique for relating in dividual test scores to a group as a whole, and also a guiding rule m the design of man- machine relabons k Fitting the man to the job Very often when managers are replacing or relocating personnel, they strive to find the individual possessing most of the character istics deemed necessary Job specifications and employee requirements are given in terms of a minimum, but very seldom are psychologi cal factors expressed within a minimum and maximum When looking for someone, often the best qualified is considered to be the one who is most intelligent, most loyal, biggest, or what ever In other words, they look for the perfect rather than the right person How ever, a person with a below-average IQ might be more desirable for a monotonous )ob than will a person with an average or above-average IQ The amount of boredom, monotony, etc , may be less for the less bright, thereby elimi nating carelessness The same is true for loyalty and devotion The person who comes to work under any condition, whether ill or not, may be as un desirable as the one who uses every excuse to take off The worker with a high fever, bad cold, or sore back is a potential hazard, espe cially if he is under heavy medication Influences of differences on safety The problem of individual differences has been related directly to the problem of safety m industry For some years, safety profes sionals and psychologists alike explored the concept of accident proneness It was be-