« PoprzedniaDalej »
30. His data on practise effect. 31. Comparison of present experimental series in yield of false reactions. 32. Range of individual differences in liability to false reactions. 33. Speed of false reactions. 34. Specific stimuli involving false reactions. 35. Correlation in falsity and speed.
V. False reactions of pathological subjects compared with those of normal subjects. 36. Yield of false reactions in different experimental series. 37. Comparative liability to false reactions, range of individual differences. 38. Speed of false reactions, specific conditions of stimulus involving false
VI. 39. Pathological records in relation to the clinical condition of the case.
II. CORRECT REACTIONS OF NORMAL SUBJECTS. The above sections will now be taken up in order. The first 14 deal with the correct reactions of normal subjects. Reaction time measurements are given to the nearest hundredth of a second only.
1. Series 20, Reaction Speed.-In this series a pattern of five figures is shown, one of which is underscored. The subject strikes on a 5-finger key, the key corresponding to the figure underscored. Probably the speed with which this can be done varies with the finger employed, as well as with the position of the underscored figure in the pattern. These are not studied in the present report; the succession of stimuli was of course the same for all subjects.
Data on series 20 are present from 10 of the normal subjects. For the first experiment, the grand average reaction time is .84 second, with a range in central tendency from .57 in subject E to 1.10 in subject H, who are both among the psychiatrists. The three experimentalists did not show quicker times than the remainder. The normal choice time in this process is thus found to range from about .5 to i second.
2. Series 20, Work-Curve.—As the series of 25 reactions proceeds, practice makes itself evident. Averaged in groups of five reactions, the successive reaction times of the 10 subjects involved are .92, 85, 93, 174, 175. The tendency is not regular, and the lengthening of time in reactions 11-15 is a consistent feature of the group, perhaps accounted for by the fact that reactions 11-15 chance to involve only the last three fingers. The mean variation of these subjects in reaction 1-5 is .17; in reactions 20-25 it is 11. The learning has reduced the individual differences, as need not surprise us in processes of this psychomotor level. In the writer, observations were made on three different days, without the appearance of day-to-day practise.
3. Series 10, Initial Reaction Time.—In this series a pattern of five figures is shown, and the subject must strike, in order, the keys corresponding. Two measurements are involved: First, the time for starting the response (designated by r), and second, the time required for the total five reactions (designated by t). Thirteen subjects give a grand average, r-time of .87, insignificantly longer than that of series 20. The mental adjustment differs from that in series 20 in that: (1) There is no need to examine the whole pattern for an underscored figure, and (2) the process is not ended, as in series 20, by striking a single key; all the others must also be struck in right succession. The latter factor, which would lengthen the initial reaction, at least counterbalances the former, which would shorten it. There is some gathering up in the mind of the whole pattern before starting the reaction at all. There is an "overlapping” in the mental processes of the initial and the subsequent reactions. The amount of this varies among the subjects. Two subjects take markedly less time to start the process in series 10 than for the single reaction of series 20. Three take much longer to start in series 10 than for the single reaction of series 20. Two of these, however, are quicker than the average to complete series 10, showing part of this initial time well spent in fixing the pattern for the total process. The general tendency to practise improvement is less marked in these r times than in the single reactions of series 20, but it is more regular.
4. Series 10, Speed of Total Process (t).–For striking the five keys in order, the grand average time of the normal subjects is 2.95 seconds. Some overlapping of the five movements is indicated, as this is but 3-4 times the interval required for single reactions of this nature. Individual difference ranges from 1.76 to 4.81, each made by a superior personality, the shorter by a man of experimental training. A professional stenographer in the group was fourth from the fastest; the special motor training of this subject not making itself particularly effective. (The probability of interference from the figures of the typewriter keyboard is slight).
Series 20, which immediately preceded the present series in experimental routine, sets up the association between the exposed figures and certain finger movements, which series 10 develops more intensively. The association seems to be a fairly natural one for the normal subjects, as this series adds no great amount of learning to that of series 20. The average times for the successive groups of five exposures are 3.18, 3.04, 2.83, 2.82, 2.90. Mean variation for reactions 1-5 is .73, for reactions 20-25, .71. But in several subjects no learning is seen at all, nor is it the slowest operators who show the most learning. Observations on four different days with subject J and two with subject K did not show learning carried over from day to day.
5. When learning in this type of series takes place, it is of interest to know how different portions of the process are affected; how it alters the time of different parts of the total process. The main point of comparison is between the initial reaction time r, and the time for the remainder of the process; i. e., total time minus initial time, designated by s. Fusion is shown in a decrease of the time s, more marked than decrease in the time r. In general, this fusion is questionable for series 10. The time s for the successive 5-exposure groups is 2.25, 2.15, 1.98, 1.96, 2.07; with m. v.'s increasing towards the end of the series, -57, 65, 61, 64, .69. The per cent of t time included in r time, is, for these successive 5-exposure groups, 29.3, 29.3, 30.1, 30.5, 28.6. If an equal time were spent on each unit of the pattern, this figure would be 20 per cent. The first reaction takes disproportionately long, but this proportion changes little as the series progresses. Individually, subject C shows much learning, but it affects the initial time as much as the total process. K, on the other hand, decreases the t time from 3.79 to 2.70, while r remains about 1.60. L shows marked decrease in s, 2.04 to 1.58, while the r time is nearly constant at 1.05. These two cases alone show notable fusion.
6. Series 50, Individual Differences in “y” Speed.-Series 50 is the same as series 10, except for using a more difficult type of association. Instead of the natural association between the five fingers and the digits 1-5 it calls for association between the five fingers and the alphabetical order of the five vowels, a eiou. The results differ correspondingly from series 10. The process takes longer and there is more learning. The general average r time is 1.41, the decrease through the 5-exposure groups being 1.61, 1.51, 1.41, 1.31, 1.23; m. v.'s about .30 throughout. The range in central tendency is from .97 in subject T to 2.02 in subject H, physicians of about equal standing. Substantial learning for the r time in the first experimental session is shown in all subjects except K, L and R, two experimentalists and a medical man who has done experimental work. The stenographer is somewhat slower than the average in reactions 1-5, and improves to as much faster than average in reactions 20-25. Repeated experiments in subjects E, J and K show considerable learning carried from day to day in the first two subjects.
7. Series 50, Individual Difference in "t" Time. This is the longest of any process in the present study. Its general average, for 13 subjects, is 4.53 seconds. Its course through the 5-reaction groups is 5.76, 4.79, 4.29, 4.08, 3.72; with m. v. of 1.13 and .68 at beginning and end respectively. Its range in central tendency is from 3.78 to 7.20, both results from persons of the upper levels of distinction. The former is a trained experimentalist, and his record is nearly equaled by the stenographer. For a touch typist, as is this subject, different fingering from the typewriter keyboard is used for the reaction to a, e and u, but the same for i and o. She, with the college student, also a woman, show the most improvement through the series. All records are marked by fairly regular learning in this series.
8. Fusion: Properties of the Time s.-In series 10 there was little progressive fusion of the reaction processes as the series proceeded. In series 50, on the other hand, the initial reaction takes not only a larger proportion of the time than in series 10, but this proportion becomes larger as the series progresses. The general averages for the time s in the successive 5-exposure groups are 4.15, 3.27, 2.88, 2.77, 2.49, showing marked decrease in the length of the t time as compared with the r time. M. V. of these quantities is respectively 1.01, -79, 158, -72, -36. Thus a more equal relation between r and t is also indicated as the series progresses. Introspection might disclose many different types of dealing with the mental task at the outset, of which practise eliminated the less efficient ones. Again significant is the percentage which the initial reaction is of the total process time. These percentages for the successive 5-exposure groups are 27.9, 31.5, 32.8, 32.3, 33.1. Progressively more of the entire adjustment becomes incorporated in the initial reaction. More than with the easier series to it is necessary to fix this unaccustomed pattern in the mind before an effective beginning can be made. The progressively greater tendency to this preliminary fixation of the pattern goes hand in hand with increased effectiveness in carrying it out.
9. Series 110, Individual Difference in Reaction Speed.—Simple mathematical sums are exposed; reaction is with the right hand if the sum is correct, with the left hand if it is incorrect. The general average time of these reactions in the 14 subjects is 1.17 seconds. The range is from .79 to 1.73 in a chief attendant and experimentalist respectively. The 5-exposure groups average 1.37, 1.17, 1.01, 1.28, 1.03; practise beyond a short initial stage being questionable.
10. Series 110, Reactions to the Different Items.—The separate sums exposed in this series, with the average reaction time to them, and the ratio of the m. v. to this time, are as follows:
In this series, 13 of the sums are correct and 12 are incorrect. The average reaction time to the correct sums is 1.11 seconds and that to the incorrect sums is 1.26 seconds.
11. Time in Series 110 as Affected by Degree of Incorrectness.It has been established in the sensory field that the magnitude of difference bears a close relation to the ease of discrimination. In