We fitted up a modest psychological laboratory near the offices of the medical examiner and the supervisor of employment and training. When men came to take a physical examination, some psychological tests were also given, designed to measure speed and accuracy of response, consistency of serial action, flexibility, oscillation, perseveration, resistance to distraction, range of peripheral vision, and similar attributes. Some of the tasks consisted of pressing the right buttons as various electric lights flashed momentarily. Other tests made use of the familiar street-car controls - the electric power handle, air-brake lever, gong, etc. - all connected with suitable recording apparatus.

These psychological tests were used for research and as clinical aids to supplement, in certain instances, the analysis of a man's difficulties made by studying the detailed descriptions of his accidents, interviewing him, and observing him at work. A number of supervisors were specially trained to ride with the motormen and note the points on which they individually needed further training or supervision. A man's accident record was first examined to see whether there was a particular type of accident to which he was prone. One man had had five collisions in as many months. The supervisor who observed him in action reported that he was competent, skillful, and careful.

A comparison of the five collisions showed that they were all alike in two respects: they were all of one relatively infrequent type, namely, front-end collisions; and all five of them happened on Sunday afternoons. Another search of the records showed that all the various cars operated by this motorman through the week were medium-weight cars, but his schedule on Sunday afternoons called for operation of a large, heavy car. Other motormen with similar schedules had no accidents, but this man lacked the necessary flexibility of habit. In other words, he was a perseverator. This peculiarity of mental make-up is measurable in the laboratory. After his schedule was changed so that all his cars were medium weight, he stopped having front-end collisions.

Consideration of the details of a man's accident record, observation of his actual operation, and personal interview with him all help in diagnosing his difficulties. For instance, a man of sixty, with an excellent record over many years, began to have a good many collision accidents. Reference to the records showed that eight out of eleven of them happened just as he was starting his car. Was he simply getting careless? Not at all. Observation and inquiry showed that he was so conscientiously careful about one part of his work that it was interfering with another part. He had become slightly deaf, and had formed the habit of listening for the starting signal from the conductor so intently that he neglected to look about in front of his car before throwing on the power. By a little retraining he learned to look quickly from side to side instead of starting his car without doing so. Since that time he has had no accident when starting.

Some of the cases of accident proneness turned out to be medical, some were traceable to wrong mental attitude, while others needed specific information or training. No two cases were exactly alike, so it was necessary to study them individually and to deal with each man differently in order to help him to overcome his suceptibility to accidents.

Although each case of accident proneness is unique and complex, and the essence of our procedure in combating accidents is individual attention, we have made certain group comparisons from time to time, to determine the relationship of various isolated traits to the practically important trait of accident susceptibility. One special report of a study by C. S. Slocombe and O. M. Hall covers eighty-six motormen, half of whom have had more than the average number of accidents. Forty-three of them, then, we may call high-accident men; forty-three, low. They were selected and paired for age, length of service, and operating conditions. These two groups, differing in ability to avoid accidents, have been compared with reference to twenty-four other variables. The results are summarized in the accompanying table.

Reference to the table shows that among these eighty-six motormen, seven had hernia, and all of these are in the high-accident group. Sixteen had a somewhat abnormal blood pressure, and fourteen of these are found in the high-accident class. Of eight who did relatively poorly in a serial-action test, all but one are high-accident men. Thirteen had on their records instances of insubordination; of these, two were in the low-accident group; and so on.

Here is evidence that accident proneness is associated with personality defect or uncooperative attitude, as measured by records of insubordination, operating delinquencies, and failure to report for duty. Thirty-nine per cent of the high-accident men, but only 5 per cent of the low-accident men, were defective, as measured by this record of uncooperative behavior. Forty per cent of the high men and 12 per cent of the low showed lack of aptitude, as measured by test S. Forty-nine per cent of the Association between Accident Proneness and Various Conditions Among Forty-three Motormen with High Accident Records and Forty-three with Low Accident Records, Paired for Length of Service, Age, and Operating Conditions.

Explanation of Table:

Psychological Trait S means relatively poor performance in a serial-action test.