Classics in the History of Psychology

An internet resource developed by
Christopher D. Green
York University, Toronto, Ontario

(Return to index)

Psychology and Industrial Efficiency

Hugo Münsterberg (1913)


The problem of securing fit motormen for electric railways was brought to my attention from without. The accidents which occurred through the fault, or at least not without the fault, of the motormen in street railway transportation have always aroused disquietude and even indignation in the public, and the street railway companies suffered much from the many payments of indemnity imposed by the court as they amounted to thirteen per cent of the gross earnings of some companies. Last winter the American Association for Labor Legislation called a meeting of vocational specialists to discuss the problem of these accidents under various aspects. The street railways of various cities were represented, and economic, physiological, and psychological specialists took part in the general discussion. Much attention was given, of course, to the questions of fatigue and to the statistical results as to the number of accidents and their relation to the various hours of the day and to the time of labor. But there was a strong tendency [p. 64] to recognize, as still more important than the mere fatigue, the whole mental constitution of the motormen. The ability to keep attention constant, to resist distraction by chance happenings on the street, and especially the always needed ability to foresee the possible movements of the pedestrians and vehicles were acknowledged as extremely different from man to man. The companies claimed that there are motormen who practically never have an accident, because they feel beforehand even what the confused pedestrian and the unskilled chauffeur will do, while others relatively often experience accidents of all kinds because they do not foresee how matters will develop. They can hardly be blamed, as they were not careless, and yet the accidents did result from their personal qualities; they simply lacked the gift of instinctive foresight. All this turned the attention more and more to the possibilities of psychological analysis, and the Association suggested that I undertake an inquiry into this interesting problem with the means of the psychological laboratory. I felt the practical importance of the problem, considering that there are electric railway companies in this country which have up to fifty thousand accident indemnity cases a year. It therefore seemed to me decidedly worth while to undertake a laboratory investigation.

[p. 65] It would have been quite possible to treat the functions of the motormen according to the method which resolves the complex achievement into its various elements and tests every function independently. For instance, the stopping of the car as soon as the danger of an accident threatens is evidently effective only if the movement controlling the lever is carried out with sufficient rapidity. We should accordingly be justified in examining the quickness with which the individual reacts on optical stimuli. If a playing child suddenly runs across the track of the electric railway, a difference of a tenth of a second in the reaction-time may decide his fate. But I may say at once that I did not find characteristic differences in the rapidity of reaction of those motormen whom the company had found reliable and those who have frequent accidents. It seems that the slow individuals do not remain in the service at all. As a matter of course certain other indispensable single functions, like sharpness of vision, are examined before the entrance into the service and so they cannot stand as characteristic conditions of good or bad service among the actual employees.

For this reason, in the case of the motormen I abstracted from the study of single elementary functions and turned my attention to that mental process which after some careful observations [p. 66] seemed to me the really central one for the problem of accidents. I found this to be a particular complicated act of attention by which the manifoldness of objects, the pedestrians, the carriages, and the automobiles, are continuously observed with reference to their rapidity and direction in the quickly changing panorama of the street. Moving figures come from the right and from the left toward and across the track, and are embedded in a stream of men and vehicles which moves parallel to the track. In the face of such manifoldness there are men whose impulses are almost inhibited and who instinctively desire to wait for the movement of the nearest objects; they would evidently be unfit for the service, as they would drive the electric car far too slowly. There are others who, even with the car at high speed, can adjust themselves for a time to the complex moving situation, but whose attention soon lapses, and while they are fixating a rather distant carriage, may overlook a pedestrian who carelessly crosses the track immediately in front of their car. In short, we have a, great variety of mental types of this characteristic unified activity, which may be understood as a particular combination of attention and imagination.

My effort was to transplant this activity of the motormen into laboratory processes. And here [p. 67] may include a remark on the methodology of psychological industrial experiments. One might naturally think that the experience of a special industrial undertaking would be best reproduced for the experiment by repeating the external conditions in a kind of miniature form. That would mean that we ought to test the motormen of the electric railway by experiments with small toy of electric cars placed on the laboratory table. But this would be decidedly inappropriate. A reduced copy of an external apparatus may arouse ideas, feelings, and volitions which have little in common with the processes of actual life. The presupposition would be that the man to be tested for any industrial achievement would have to think himself into the miniature situation, and especially uneducated persons are often very unsuccessful in such efforts. This can be clearly seen from the experiences before naval courts, where it is usual to demonstrate collisions of ships by small ship models on the table in the courtroom. Experience has frequently shown that helmsmen, who brave found their course a life long among real vessels in the harbor and on the sea, become entirely confused when they are to demonstrate by the models the relative positions of the ships. Even in the naval war schools where the officers play at war with small model ships, [p. 68] a certain inner readjustment is always necessary for them to bring the miniature ships on the large table into the tactical game. On the water, for instance, the naval officer sees the far-distant ships very much smaller than those near by, while on the naval game table all the ships look equally large. On the whole, I feel inclined to say from my experience so far that experiments with small models of the actual industrial mechanism are hardly appropriate for investigations in the held of economic psychology. The essential point for the psychological experiment is not the external similarity of the apparatus, but exclusively the inner similarity of the mental attitude. The more the external mechanism with which or on which the action is carried out becomes schematized, the more the action itself will appear in its true character.

In the method of my experiments with the motormen, accordingly, I had to satisfy only two demands. The method of examination promised to be valuable if, first, it showed good results with reliable motormen and bad results with unreliable ones; and, secondly, if it vividly aroused in all the motormen the feeling that the mental function which they were going through during the experiment had the greatest possible similarity with their experience on the front platform of the electric [p. 69] car. These are the true tests of a desirable experimental method, while it is not necessary that the apparatus be similar to the electric car or the external activities in the experiment be identical with their performance in the service. After several unsatisfactory efforts, in which I worked with too complicated instruments, I finally settled on the following arrangement of the experiment which seems to me to satisfy those two demands.

The street is represented by a card 9 half-inches broad and 26 half-inches long. Two heavy lines half an inch apart go lengthwise through the centre of the card, and accordingly a space of 4 half-inches remains on either side. The whole card is divided into small half-inch squares which we consider as the unit. Thus there is in any cross-section 1 unit between the two central lines and 4 units on either side. Lengthwise there are 26 units. The 26 squares which lie between the two heavy central lines are marked with the printed letters of the alphabet from A to Z. These two heavy central lines are to represent an electric railway track on a street. On either side the 4 rows of squares are filled in an irregular way with black and red figures of the three first digits. The digit 1 always represents a pedestrian who moves just one step, and that [p. 70] means from one unit into the next; the digit 2 a horse, which moves twice as fast, that is, which moves 2 units; and the digit 3 an automobile which moves three times as fast, that is, 3 units. Moreover, the black digits stand for men, horses, and automobiles which move parallel to the track and cannot cross the track, and are therefore to be disregarded in looking out for dangers. The red digits, on the other hand, are the dangerous ones. They move from either side toward the track. The idea is that the man to be experimented on is to find as quickly as possible those points on the track which are threatened by the red figures, that is, those letters in the 26 track units at which the red figures would land, if they make the steps which their number indicates. A red digit 3 which is 4 steps from the track is to be disregarded, because it would not reach the track. A red digit 3 which is only 1 or 2 steps from the track is also to be disregarded, because it would cross beyond the track, if it took 3 steps. But a red 3 which is 3 units from the track, a red 2 which is 2 units from the track, and a red 1 which is 1 unit from the track would land on the track itself; and the aim is quickly to find these points. The task is difficult, as the many black figures divert the attention, and as the red figures too near or too far are easily confused [p. 71] those which are just at the dangerous distance.

As soon as this principle for the experiment was recognized as satisfactory, it was necessary to find a technical device by which a movement over this artificial track could be produced in such a way that the rapidity could be controlled by the subject of the experiment and at the same time measured. Again we had to try various forms of apparatus. Finally we found the following form most satisfactory. Twelve such cards, each provided with a handle, lie one above another under a glass plate through which the upper card can be seen. If this highest card is withdrawn, the second is exposed, and from below springs press the remaining cards against the glass plate. The glass plate with the la cards below lies in a black wooden box and is completely covered by a belt 8 inches broad made of heavy black velvet. This velvet belt moves over two cylinders at the front and the rear ends of the apparatus. In the centre of the belt is a window 4½ inches wide and 2½ inches high. If the front cylinder is turned by a metal crank, the velvet belt passes over the glass plate and the little window opening moves over the card with its track and figures. The whole breadth of the card, with its central track and its 4 units on either side, is visible through it over [p. 72] of 5 units in the length direction. If the man to be experimented on turns the crank with his right hand, the window slips over the whole length of the card, one part of the card after another becomes visible, and then he simply has to call the letters of those units in the track at which the red figures on either side would land, if they took the number of steps indicated by the digit. At the moment the window has reached Z on the card, the experimenter withdraws that card and the next becomes visible, as a second window in the belt appears at the lower end when the first disappears at the upper end. In this way the subject can turn his crank uninterruptedly until he has gone through the 12 cards. The experimenter notes down the numbers of the cards and the letters which the subject calls. Besides this, the number of seconds required for the whole experiment, from the beginning of the first card to the end of the twelfth is measured with a stopwatch. This time is, of course, dependent upon the rapidity he crank is turned. The result of the experiment is accordingly expressed by three figures, the number of seconds, the number of omissions, that is, of places at which red figures would land on the track which were not noticed by the subject; and, thirdly, the number of incorrect places where letters were called in spite [p. 73] of the fact that no danger existed. In using the results, we may disregard this third figure and give our attention to the speed and the number of omissions.

The necessary condition for carrying out the experiments with this apparatus is a careful, quiet, practical explanation of the device. The experiment must not under any circumstances be started until the subject completely understands what he has to do and for what he has to look out. For this purpose I at first always show the man one card outside of the apparatus and explain to him the differences between the black and the red figures, and the counting of the steps, and show to him in a number of cases how some red figures do not reach the track, how others go beyond the track, and how some just land in danger on the track. As soon as he has completely understood the principle, we turn to the apparatus and he moves the window slowly over a test card and tries to find the dangerous spots, and I turn his attention to every case in which he has omitted one or has given an incorrect letter. We repeat this slowly until he completely masters the rules of the game. Only then is he allowed to start the experiment. I have never found a man with whom this preparation takes more than a few minutes.

[p. 74] After developing this method in the psychological laboratory, I turned to the study of men actually in the service of a great electric railway company which supported my endeavors in the most cordial spirit. In accordance with: my request, the company furnished me with a number of the best motormen in its service, men who for twenty years and more had performed their duties practically without accidents, and, on the other hand, with a large number of motormen who had only just escaped dismissal and whose record was characterized by many more or less important collisions or other accidents. Finally, we had men whose activity as motormen was neither especially good nor especially bad.

The test of the method lies first in the fact that the tried motormen agreed that they really pass through the experiment with the feeling which they have on their car. The necessity of looking out in both directions, right and left, for possible obstacles, of distinguishing those which move toward the track from the many which move along the track, the quick discrimination among the various rates of rapidity, the steady forward movement of the observation point, the constant temptation to give attention to those which are still too far away or to those which are so near that they will cross the track before the approach [p. 75] of the car, in short, the whole complex situation with its demands on attention, imagination, and quick adjustment, soon brings them into an attitude which they themselves feel as identical with that in practical life. On the other hand, the results show a far-reaching correspondence between efficiency in the experiment and efficiency in the actual service. With a relatively small number of experiments this correspondence cannot be expected to be complete, the more as a large number of secondary features must enter which interfere with an exact correlation between experiment and standing in the railway company. We must consider, for instance, that those men whom the company naturally selects as models are men who have had twenty to thirty years of service without accidents, but consequently they are rather old men, who no longer have the elasticity of youth and are naturally less able to think themselves into an artificial situation like that of such an experiment, and who have been for a long time removed from contact with book work. It is therefore not surprising, but only to be expected, that such older, model men, while doing fair work in the test, are yet not seldom far surpassed by bright, quick, young motormen who are twenty years younger, even though they are not yet ideal motormen. Moreover, the standing in the [p. 76] company often depends upon features which have nothing to do with the mental make-up of the man, while the experiment has to be confined to those mental conditions which favor accidents. It is quite possible that a man may happen to experience a slight collision, even though no conditions for the accident were lying in his mental make-up. But we may go still further. The experiment refers to those sides of his mind which make him able to foresee the danger points, and that is decidedly the most essential factor and the one from which most can be hoped for the safety of the public. But this does not exclude the possibility that some other mental traits may become causes of accidents. The man may be too daring and may like to run risks, or he may still need discipline, or he may not be sufficiently acquainted with the local conditions. Any such secondary factors may cause some slight accidents with the man who shows rather fair results in the experimental test of his foresight. Finally, we must not forget that some men enter into such tests under a certain nervous tension and therefore may not show so well at the very first test as their mental equipment should allow. Hence it is decidedly desirable not to rely on the first test, but to repeat it. If those various interferences are taken into account, the correspondence [p. 77] between efficiency and the results of the tests is fairly satisfactory. It justified me in proposing that the experiments be continued and in regarding it as quite possible that later tests on the basis of this principle may be introduced at the employment of motormen.

A difficulty is presented by the valuation of the numerical results. The mere number of omissions alone cannot be decisive, as it is clear that no intelligent man would make any omissions if he should give an unlimited amount of time to it; for instance, if he were to spend fifteen minutes on those 12 cards. But this is the same thing as to say that a motorman would not run over any one if he were to drive his car one mile in an hour. The practical problem is to combine the greatest possible speed with the smallest number of oversights, and both factors must therefore be considered. The subject who makes relatively many mistakes but uses a very short time must be acknowledged to be as good as the man who makes fewer mistakes but takes a longer time. In the results which I have gathered in experiments with motormen, no one has gone through those 12 cards in a shorter time than 140 seconds, while the longest time was 427 seconds. On the other hand, no one of the motormen made less than 4 omissions, while the worst ones made 28 [p. 78] omissions. I abstract from one extreme case with 36 omissions. On the whole, we may say that the time fluctuates between 180 and 420, the mistakes between 4 and 28. The aim is to find a formula which gives full value to both factors and makes the material directly comparable in the form of one numerical value instead of the two. If we were simply to add the number of seconds and the number of omissions, the omissions would count far too little, inasmuch as 10 additional omissions would then mean no more than 10 additional seconds. On the other hand, if we were to multiply the two figures the omissions would mean by far too much, as the transition from 4 mistakes to 8 mistakes would then be as great a change as the transition from 200 to 400 seconds, that is, from the one extreme of time to the other. Evidently we balance both factors if we multiply the number of omissions by 10 and add them to the number of seconds. The variations between 4 and 28 omissions are 24 steps, which multiplied by 10 correspond to the 240 steps which lie between 180 and 42O seconds. On that basis any additional 50 seconds would be equal to 5 additional omissions. If of two men one takes 100 seconds less than his neighbor, he is equal to him in his ability to satisfy the demands of the service, if he makes 10 mistakes more.

[p. 79] On the basis of this calculation I find that the old, well-trained motormen come to a result of about 450, and I should consider that an average standard. This would mean that a man who uses 400 seconds would not be allowed to make more than 5 omissions, in 350 seconds not more than 10, in 300 not more than 15, in 250 not more than 20, under the condition that these are the results of the first set of experiments. Where there are more than 20 omissions made, mere quickness ought not to be allowed as a substitute. The man who takes 150 seconds and makes 30 mistakes would come up to the same standard level of 450. Yet his characteristics would probably not serve the interests of the service. He would speed up his car and would make better time than any one else, but would be liable to accidents. I should consider 20 mistakes with a time not longer than 250 as the permissible maximum. Among the younger motormen whom I examined, the best result was 290, in which 270 seconds were used and only 2 omissions made. Results below 350 may be considered as very good. One man, for instance, carried out the experiment in 237 seconds with 11 mistakes, which gives the result 347. From 350 to 450 may be counted as fair, 450 to 550 as mediocre, and over 550 as very poor. In the case of old men, who may [p. 80] be expected to adjust themselves less easily to artificial experiments, the limits may be shifted. If the experiments are made repeatedly, the valuation of the results must be changed accordingly. The training of the men in literary and mathematical work or in experimentation may be considered, as our experiments have shown that highly educated young people with long training in experimental observations can pass through the test much more quickly than any one of the motormen could. Among the most advanced graduate students who do research work in my Harvard laboratory there was no one whose result was more than 275, while, as I said, among all the motormen there was no one whose result was less than 290. The best result reached was by a student who passed through the test in 223 seconds with only 1 mistake, the total therefore being 233. Next came a student who did it in 215 seconds with 3 mistakes, total, 245; then in 338 seconds with 2 omissions, total, 248, and so on.

I recapitulate: With men on the educational level and at the age that comes in question for their first appointment in the service of an electric railway company, the test proposed ought to be applied according to this scheme. If they make more than 20 mistakes, they ought to be excluded; if they make less then 20 mistakes, the number of [p. 81] omissions is to be multiplied by 10 and added to the number of seconds. If the sum is less than 350, their mental fitness for the avoidance of accidents is very high, between 350 and 450 fair, and more than 550 not acceptable under any conditions. I submit this, however, with the emphasis on my previous statement that the investigation is still in its first stage, and that it will need a long coöperation between science and industry in order to determine the desirable modifications and special conditions which may become necessary in making the employment of men partly dependent upon such psychological tests. There can be no doubt that the experiments could be improved in many directions. But even in this first, not adequately tested, form, an experimental investigation of this kind which demands from each individual hardly 10 minutes would be sufficient to exclude perhaps one fourth of those who are nowadays accepted into the service as motormen. This 25 per cent of the applicants do not deserve any blame. In many other occupations they might render excellent service; they are neither careless nor reckless, and they do not act against instructions, but their psychical mechanism makes them unfit for that particular combination of attention and imagination which ought to be demanded for the special task of the motorman. If the many thousands of injury and the many hundreds of death cases could be reduced by such a test at least to a hall, then the conditions of transportation would have been improved more than by any alterations in the technical apparatus, which usually are the only objects of interest in the discussion of specialists. The whole world of industry will have to learn the great lesson, that of the three great factors, material, machine, and man, the man is not the least, but the most important.