AND THEIR MENTAL CORRELATES It is a matter of common observation that individuals differ greatly in decision time, accuracy, and constancy of or consistency in subjective decisions. It is generally supposed that the slow in decision are both more constant and more accurate than the quick. "Slow but sure" is a popular maxim. The chief purpose of this investigation was to determine the actual nature of these various decision types—to see, for example, whether "quick but sure" and "slow but unreliable" are not equally true of certain classes. In other words, it was desired to ascertain the relation between decision time and constancy, between time and accuracy, and also between the subjective and objective decision types thus determined. A further purpose was to determine the correlation subsisting between these decision factors (viz.: time accuracy and constancy) and other mental traits such as memory, association time, suggestibility, attention, etc. The experiments were conducted in the Harvard Laboratory during the years 1911 to 1913. In all, nineteen subjects took part; three of them (A, C and L) continued through the two years' work. Twelve participated in the first year's experiments reported in the first part of this paper, and ten in the second year's experiments reported in Chapter II. All were members of the laboratory and trained in experimental methods. Five (D, F, I, O, and P) were women. Three of these were in the first year's investigation and two in the second. In the first year's experiments on decision the materials used were cards upon which were printed the names of the objects to be decided upon from various points of view. Of these there were forty pairs, fifty groups of three each, and fifty groups of five each.1 The following are some representative examples. 1 All the pairs, twenty-five groups of three and twenty-five groups of five were selected from material prepared by G. W. Peckham, who had begun work upon this problem in the Harvard Laboratory, but found it necessary to discontinue the investigation. Pairs: violin music, piano music; Hamlet, Macbeth; orange marmalade, baked apple; Hume, Kant; gunpowder, printing; camel, elephant. Threes: Poe, Kipling, Hawthorne; wealth, strength, power; literature, science, philosophy; baseball, football, tennis; Dante, Goethe, Shakespeare; X-rays, photography, telegraphy; fountain pen, penknife, watch. Fives: music, sculpture, painting, drama, poetry; peaches, pears, apples, grapes, cherries; chrysanthemum, lily, dahlia, carnation, rose; Rembrandt, Holbein, Titian, Rubens, Ingres; temperance, wisdom, justice, courage, truth; bicycle, safety razor, automobile, telephone, balloon. The material used for subjective decision in the second part of the investigation, and that used for accuracy of decision, and in the several tests, will be described under the sections dealing with these various experiments. RELATION OF TIME AND CONSTANCY The first experiments performed were to determine the relation between time and constancy in relatively subjective decisions. These experiments can be divided into three series. In the first series, consisting of forty experiments, the pairs of cards, described above, were exposed by tilting back a shutter behind which they had been placed. The subject was directed to decide as quickly as was compatible with sincerity, and to raise the right or left hand according as the decision was for the right or left card. The basis of decision, which varied with the nature of the concepts, was in each case given by the experimenter; thus: decide according to your preference, their greatness, their importance, or their usefulness. The time was taken with a stop watch, which was started at the same instant that the crank of the shutter was turned back, and stopped at the signal of the subject. The time and the decision were recorded. After five weeks the experiment was repeated in the same way. The decision and time were again recorded and compared with those of the first presentation. The results of this series of experiments are shown in Table I. The number of changed decisions, the average time for the first presentation, the average time for the second presentation, the combined average, and the number of first decisions remembered at the second presentation, are given for each of the nine subjects. The number of decisions changed varies from 1 for subject F to 13 for subject G; and the decision time from 1.46′′ for C to 3.30" for H. There is, however, no evidence of a correlation between time and constancy. The figures in the last column give some idea of the part played by memory in this experiment. A. T. I 2.96 2.34 2.65 31 4(-1) 2.74 2.56 2.65 9 7 3.30 3.12 3.21 18 8(-1) 2.49 1.99 2.24 12 No. C. Number of changed decisions. Those in brackets doubtful. A. T. No. R. Number of decisions remembered from first to second presentation. = In order to eliminate the memory factor as far as possible, a more complex series of experiments was devised. This second series consisted of the fifty groups of three cards each. These were arranged by the subject in order of preference, greatness, importance, or usefulness. The method was otherwise the same as in Series I. The motor activity of arranging the cards was rendered approximately constant in the following manner: On tilting back the shutter the cards were exposed in an upright position between two slats. The subject removed each card and laid it down whether the presented arrangement was maintained or changed. This series was also repeated in four to five weeks, except in the case of subjects K, J and L; here the time between first and second presentation was seven to eight weeks The same directions for arranging the cards were given the subject on each occasion; but a change of the basis of decision sometimes occurred under the category specially mentioned,-e.g., importance might mean importance to self, to the world, to science, etc. The possibility of such changes was as far as possible avoided ; but all differences due to these changes or to lack of knowledge of the material were eliminated. The results of this series are shown in Table II. The number of changed decisions, and the average, and the median for the first and for the second presentations are given. Also, in columns marked "Above M2" and "Below M2", the changes in decisions whose times are above or below the median in the second presentation are given. There are always more changes in decisions above the median. This may be explained by supposing that the uncertain decisions take the longest time, and are also the more liable to change. = Above M2 and Below M2 Number of changed decisions above and below the median in the second presentation. In Brackets, doubtful decisions to be subtracted. Table III-A shows how the subjects might be classified according to type. The types, however, grade into each other and some |