ACQUISITION OF SKILL A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOB THE DEGREE OP DOCTOR OF PHILOSOPHY IN THE UNIVERSITY OF MICHIGAN BY WILLIAM HOWARD BATSON Published as No. 91 of the Psychological Monographs. ACQUISITION OF SKILL A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE UNIVERSITY OF MICHIGAN BY WILLIAM HOWARD BATSON Published as No. 91 of the Psychological Monographs. 33 Tiid Uaivsraliy ACKNOWLEDGMENTS In the first place the writer wishes to express his thanks to Prof. John F. Shepard for his many suggestions in planning the experiments as well as for valuable assistance in the preparation of the manuscript. Thanks are also due to those who acted as subjects,— J. E. DeCamp, C. P. Wang, H. Foulk, F. C. Dock- eray, E. L. Cole, A. Martin, Roy Feemster, Viola Coldwell, Ro- berta Scott and B. A. Hayes, whose services in the experiments are very inadequately repaid by these acknowledgments. TABLE OF CONTENTS PAGE Introduction i Review of Literature i Problem Stated 13 The Effect of a Number of Factors Working Simul- taneously 14 The Plotting of the Curves 30 Discussion of Curves 31 Discussion of Results 34 The Effect of an Isolated Factor 39 Experiment I — On Direction 40 Experiment II — On Force 49 Experiment III — On Time 58 The Effect of Several Factors Working in Succession 64 General Discussion 75 Long Period Plateaus 75 Daily Fluctuations 79 General Form of the Curves 81 Influence of Objective and Subjective Factors 85 Warming up Process 86 •The effect of Short and Long Rest Periods 86 Summary 89 Bibliography 91 Curves and Plates 92 INTRODUCTION Review of Literature The acquisition of skill in any line of endeavor presents many interesting and important problems. In recent years several studies have been made to determine the factors that influence learning in various fields. One of the first experiments carefully worked out upon this subject was the study of learning of telegraphy by Bryan and Harter.^ Their work was based upon the information gained by questioning about two hundred operators, upon reports from schools of telegraphy, and upon the records of individuals that were studying the subject. Two of these individuals were tested weekly from the time they began to practice until they had at- tained a fair degree of ability as operators. With a third person, the tests commenced about six weeks after he had begun the practice. He was tested weekly on his ability to receive, first, letters not making words ; second, letters making words but the words not making sentences; third, letters making words and the words making sentences. The curves for the first two subjects were plotted by letting the divisions on the y-axis represent the number of letters per minute sent or received and the divisions on the x-axis represent the periods between the tests. The curves for the last subject were three in number. The divisions on the y-axis for each of these represented the number of letters received per minute and the divisions on the x-axis represented the practice periods. The first of these was the letter curve, the second the word curve, and the third the connected discourse curve. As a result of the information gained from those acquainted with the work and the tests made by the experimenters, they concluded that the curves which represent the sending and re- ^ Bryan, W. L. and Harter, N. "Studies in the Physiology and Psychology of the Telegraphic Language." Psy. Rev., vol. 4, pp. 27-53 and vol. 6, pp. 345-375- 2 WILLIAM IIOiVARI) liATSON cciving- ability take different forms. The sending curves rise rapidly for some time and then gradually approach i)arallelism with the X-axis. The ordinary receiving curve takes a more irregular form and shows at least two extended flat places where there appears to be little or no progress. This was not true, how- ever, in the letter and word receiving curves. Each of these had the same general form as the sending curve. The rea.sons given for the difference in the two kinds of curves are the seeming' complexity of the language, the difference of ()l)p()rt unity for practice, the pleasure involved, and the intensity of the effort. Bryan and Harter believe that the receiving of the message constitutes a very much more complex act than the sending of a message. The learner is al)le to make ]>rogress more rapidly in controlling a series of (|uick movements which constitute the sending than in distinguishing the sounds that represent the dots and dashes and the back click of the instru- ment in receiving. Besides this, the opportunity for practice in receiving at a slow rate is much less than that for sending at a slow rate. There is no reason given as to why it should be more pleasant to send than to receive. The intensity of the effort is considered to be very important. In this connection, they make the following- statement: "One conclusion seems to stand out from all these facts more clearly than any thing else, namely, that in learning (o interpret the telegraphic language it is inten.se effort that educates."" They take the sending curve to be the typical practice curve, and their principal task is to explain the form of the receiving curve. The first ascent a])])ears because the learning of the let- ters is comparatively easy. The periods of arrest or ])]ateaus in the curve are explained on the basis of different order of habits. "A plateau in the curve means that the lower-order habits are approaching their maximum development but are not yet suffi- ciently automatic to leave the attention free to attack the higher- order habits."^ By the lower-order habits are meant those con- ' Bryan and llartcr. Psy. Rev. vol. 4, ]). 50. * Ibid., vol. 6, p. 357. • ACQUISITION OF SKILL 3 cerned with the interpretation of letters, syllables, and words. The higher-order habits constitute a system of habits associated with groups of words as they are found in connected discourse. The plateau is a period when no progress is being made in the letter and word habits. They are simply becoming automatic. As soon as they are sufficiently brought under control so that attention is free, the higher connected discourse habits begin to grow and this marks the rise from the plateau. The attention at first is confined almost entirely to the letters, then to words, and finally to larger sense groups or to the mes- sage as a whole. However, the writers do not hold that there is complete isolation of the lower and higher habits. ''The syn- chronous curves of Fig. 30 and the experience of operators agree in showing that from an early period letter, word, and higher habits make gains (a) simultaneously, but (b) not equally."^ The chief gains, however, are in the lower-order habits. In the sending curve the various order habits are formed simultan- eously and so no plateaus appear. Objective and subjective fac- tors have a great influence on the beginner but do not disturb the expert. Swift studied types of learning of different degrees of com- plexity and reached conclusions somewhat at variance with those given above.^ For the purpose of studying the improvement in what he took to be the simplest type of learning, he had six subjects toss balls. Five of these subjects practiced with two balls, receiving and throwing one while the other was in the air. The other subject practiced with three balls, using both hands and keeping one ball always in the air. The balls that he used were of solid rubber, weighing 122.6 and 130.2 grams and being 42 and 44 mm. in diameter respectively. The practice consisted of ten trials daily. The number of throws the subject could make without missing one of the balls constituted a trial. The score for the day was the sum of the catches that were made in the ten trials. * Bryan and Harter. Psy. Rev. vol. 6, p. 350. " Swift, E. J. "Studies in the Psychology and Physiology of Learning." Amer. Jour, of Psy. vol. 14, pp. 201-251. 4 WILLIAM HOWARD BATSON He found that various objective and subjective factors had a very marked effect upon the learning. The physical efficiency was very important but the subject could not always tell in ad- vance how well he would succeed. He agrees with Bryan and Harter that all advance requires effort but further states that excessive effort may cause the subject to lose rather than gain ability. The curves for this work were plotted by taking the divisions on the y-axis to represent the number of catches for each daily practice and the divisions on the x-axis to represent the practice periods. All of the curves for this work have the same general form and are concave to the y-axis. All of the curves show great irregularity of advance. In these curves there are no long period fluctuations, as was noted above. There were, however, several short periods during which there was little or no advance. No immediate rapid rise appears in these curves as appeared in those for telegraphy obtained by Bryan and Harter. Swift thinks that this first rapid rise will appear only in those cases where symbols or other devices are used for handling and pre- senting ideas. He seems to imply that the curve that is concave to the y-axis is the normal practice curve.^ Swift seems to think that the shorter plateaus that he found in this work may be caused by a combination of different factors or by a number of factors that act separately. He is not clear as to whether there is no progress or whether the progress is such that it can not be measured. He says: "This lack of energy, due to waning interest, probably has more to do with delaying the learner's progress and making plateaus than any thing else. One cannot escape a dead level in uninteresting work and after the enthusiasm that novelty stirs has spent itself the interest is dulled and effort slackens. Yet the slow progress is often only an apparent one and due to our inability to measure the ad- vance."''' In support of this last view, he points out that during the slow period the subjects occasionally made high scores and that at the end of the plateau there was a rise much above the • Swift. Amer. Jour, of Psy. v. 14, p. 228. ' Swift. Amer. Jour, of Psy. vol. 14, p. 213. ACQUISITION OF SKILL S level immediately preceding the plateau. In another place, he points out that the delay may be the result of having reached the limit of a certain method,^ In order to get a curve for a more mental type of work and one that would serve as a comparison curve with those obtained by Bryan and Harter for telegraphy, Swift studied the learning of shorthand.^ He acted as subject himself and was the only one tested. In this work, an hour and a half a day was given to study for a period of something over ten weeks. During the early part of this time the practice was confined to writing the material in shorthand, but later the daily periods were divided between writing and reading what had been written. He was given a ten minute test daily by an assistant. In both the read- ing and writing tests the number of words read or written formed the basis for the score. The material for the reading tests was the matter that he had written ten days before. The curve was plotted by letting the divisions on the y-axis represent the number of words and the divisions on the x-axis represent the practice periods. Swift decided in this work that there was no evidence of higher-order habits being separated from the lower-order habits. There were no long period plateaus. Whatever arrests there were, could be accounted for on the ground of emotional factors. In explaining the short period plateaus, he uses the following language : "In learning shorthand and presumably also in learn- ing to receive or send telegraphy, a large number of associations are formed that do not affect the speed of the work, because there is no opportunity to use them, and the learner seems to make lit- tle or no progress, not because this is the particular time for the formation of a 'hierarchy of habits', for this is going on all the time, but because the range of associated knowledge in the sub- ject is too limited to meet the demand."^^ Swift seems to place his explanation more on a quantitative than a qualitative basis. A certain amount of material must be accumulated and properly associated before it becomes effective, and the accumulation of * Ibid., p. 214. * Ibid., p. 224. ^ Swift Amer. Jour, of Psy. vol. 14, p. 224. 6 ii'ii./.i.iM iiow'.iKi) n.rrsoN lliis (KdiiiU- (|iiaiility ol" iiiatciiiil exijkiins what lakes pkicc dur- iiii; tlic aiTi'Sl pi'iiod. In liis sliuly of typcwritinj;," he practiced one hour per day with the sij^ht method for forty-four practice days. The score was kept by uotiuj^ the nuuiher of words written ])er hour. The ciu"ves for ihi- vvoik were plolli-d hy lellini; the divisions on the y-axis represent the nuniher of words ]}er h(jur and the (H visions on tlie X-axis represent the niinil)er of practice periods. Swift and Schuyk-r'- later carried on another experiment on typewritini;. In this experiment, Schuyler was the subject and practiced thirty minutes daily by the touch system, 'riic work extended over sixty-six practice days. The material copied was at first taken from (liant's typewriting manual, and later lec- tures and essays were co])ie(l. In this work the number of strokes were used as a basis for kee|)in};- the score. I'\)r a few practice |)eriods the same sentence was repeated for the whole lialf hour. In both of tlu'se curves the proj^ress was irrej^nlar. The change of copy, physical condition, subjective stale, and perhaps other factors contributed tit this irregidarity. There were no long period plateaus as in the ctu-ve for telegraphy. Jioth simple and complex factors reveal themselves to introspection from the be- ginning of the work. Ilowever, the simpler elements are more active in the early period. The shorter plateaus such as are found here have two causes, "t'ousidered from the standpoint of auto- mati/aliou, they are icsting places. The learner has overshot his permanent power and nmst wait until automatization is ])er- fected. They are also i\\w to slump in enthusiasm."''' The steps of improM'uient herc> as in llu> ball tossing were made uncon- sciously and then adopted. Swift holds that this unconscious advance is an argument for the idea that the associations are being made automatic during the apparent rest periods. The k'lpse of attention which accompanies these ])lateau periods is a result of the accuuudation of loosely connected associations. "S'vvifl. "'riic Ai(|Mi.sili()ii of .Skill in TyiH'wriliiiK." I'sy. l^nl. vol. T, p'P. ".Swifl aiuj .SclutyliT. "Tlir l.r;iriiiiij; l'n>coss." I'.sy. Hut. vol. 4, l)p- '"Svvifl. Psy. Ihil. vol. 1, p. 305. ACQinsrnoN oi' skill 7 111 (he study ot Iclci^rapliy, ;i Morse iiislnnnent was used and James' "Talks to Teachers" riiniished the nialerial. The prac- tice consisted of four hours on four successive days given to the study of the alphabet and following- this were thirty-nine practice days. JCach of these practice periods consisted of a half hour of practice and live niiiui(i-s lest. The eiuve for this work was of the same general form as for hall-tossing .and for tyi)ewri(ing. He holds that the delays are the result of the time it takes the associations to become automatic.*' Swift also studied the learning of a language and secured a curve for the improvement. The Russian language was used as the basis for this work I)ecause the sul)ject wotdd be least assisted by his ])ast ex])erience. The practice continued from March 30 until June 14, 1905, and consisted of thirty minutes study and fifteen minutes reading each day. The number of words read was made the basis for the score. The curve obtained from this work showed three periods of advance and foni' plateaus, lie holds here that the plateaus do not represent places of no prog- ress but that the progress cannot be measured. The lower-order and the higher-order habits go along together.*''' Swift has also studied the problem of relearning in ball-tossing and typewriting. In both cases, he finds that during a long rest ])eriod there is little, if any, loss of skill and that the subject in a very short time is able to go beyond bis best previous record.'" Hook has made a more elaborate study of learning to type- write. His aim was to obtain learning curves that might be ex- plained in detail by the aid of introspections of the subjects. Eleven subjects took part in (he work. Both the sight and touch methods were used. The score was kept in terms of strokes. The curves were plotted by having the divisions on the y-axis represent (he nnniber of strokes and the divisions on the x-axis represent the i)rac(ice periods. ".Swift. "l.carniiiK lo TclcKiupIi." P.sy. Bill. vol. 7, pp. 147-15.3. '".Swift. "I'oKiiiiiiiifj; a I-an|.;iiaKC ; A iContrihulion to the P.sycholoKy of Learning." Studies of I'hilosopliy and I'sycliolof^y hy I'\)rnicr Sttidciifs of 'Charles iMlward Carman. Pp. 297-313. ".Swift. "Moniory of a Complex Skillful Act." Amcr. Joiir. of Psy., vol. 16, 'PP. 131-133. "Memory of Skillful Movements." Psy. TUil., vol. 3, pp. 185-187. "Relearning a Skillful Act." Psy. P>ul., vol. 7, 147-153. 8 WILLIAM HOWARD BAT SON He found that the curves obtained showed three types of fluctuations: the first was the short fluctuation from day to day; then there were slightly longer periods of arrest which he called "breathing places" ; lastly, there were long periods during which there was practically no progress at all, which he called plateaus. The daily fluctuations he explained on the basis of objective and subjective factors. The principal objective factor was the difference in the material to be copied. On the subjective side the amount of re-learning and warming up, fluctuations in at- tention, and changes in emotional attitude were very important factors in determining the daily fluctuations. Fatigue from general causes was another important element. The breathing places were largely the result of lapses in spon- taneous attention. In some cases strong application of voluntary effort caused them. Sometimes the subject on good days made an unusually high score and for several days it took all his at- tention to maintain this level. In this case Book's statement that "It takes some time for the new way of writing to become sufficiently automatic to allow part of the attention to forge ahead in quest of more economic methods,"^^ seems almost to agree with the interpretation of Swift who speaks of giving the associations time to become automatized. Besides these short period arrests, some long period arrests were noted. The curves of two of the subjects, one that had used the sight method and the other the touch method, showed pro- nounced long period plateaus. Both of these curves extended over long periods of time. Book thinks that these plateaus be- long to rather definite levels of attainment. There is no regular order for the development of special habits in typewriting. These long period plateaus "do not represent periods of incubation, where certain elementary habits make substantial gains, pre- paratory to their organization into higher-order habits, they are: (a) Resting places in the learner's interest and effort; or (b) 'breakdown' stages caused by excessive effort wrongly ap- " Book, W. F. "The Psychology of Skill : with Special Reference to Its Acquisition in Typewriting." University of Montana Publications in Psy- chology: Bui. No. 53, p. 155- ACQUISITION OF SKILL 9 plied, "^^ Here Book disagrees with both Swift, and Bryan and Harter. The rise from the plateau, instead of being the result of old associations which have become automatic, is the result of re- newed effort by means of which new short cuts have been dis- covered. He holds that in simple forms of work where only one or a few simple associations are to be formed no plateau will appear. Even in cases of complex work, it is possible so to direct and control effort that they will not appear. Some of his subjects were able to do this complicated work in typewriting without the occurrence of any plateau. In regard to the warming up and re-learning, the evidence is quite conclusive. The records showed that the last half of the practice period in almost every case gave better results than the first half. Besides the daily revival of the associations already learned, there is a process of getting into a certain "set" which he terms a "typewriting psychosis," and as this "set" comes many of the former difficulties disappear. Book also gave a number of memory tests in this work. He found that after a period of six months from the time of the last regular practice, ten daily tests showed an average slightly lower than the last ten tests of the regular practice. One year later a second series of ten tests showed a gain over the last ten regular practice tests in the number of strokes, but a slightly higher percent of errors. Book feels that the increase in the number of errors for the memory tests shows that more effort was put forth than had been in the regular practice. This, how- ever, does not account for the increase of skill in the second memory test over the first one nor does it fully explain the gain made over the regular practice. The author attributes that gain, "to the disappearance, with the lapse of time, of numerous as- sociations, bad habits of attention, incidentally acquired in the course of learning, interfering habits and tendencies, which, as they faded, left the more firmly established typewriting associa- tions free to act."^^ '^Book. "The Psychology of Skill," p. 157. " Book. "The Psychology of Skill," p. 80. 10 WILLIAM HOWARD BATSON Several less extensive studies have been made touching upon the subject of the acquisition of skill. Among these may be mentioned the work of Scripture, Smith and Brown ; Johnson ; Partridge; Bair; Wells; Starch; and Whitley. Scripture, Smith and Brown studied the effect of practice upon one's ability to insert a needle into a small hole. They reached the conclusion that the improvement was due to the training of attention and that it was psychical rather than physical. Any distraction or mental or physical fatigue lowered the result. The best results were obtained when the attention was fixed upon the hole rather than upon the needle.-" Johnson had his subjects practice at tapping at the corners of an e(juilateral triangle. In this work, he found that all of the curves of improvement follow the same form. He reached the conclusion that transition from a state requiring constant atten- tion to a state of automatic control follows a law as exact as any physical law. But because the subjects are never mentally and physically equal at two different tests and because we do not have scientific results from allied subjects, it is difficult to determine the law. The closer the personal factors are to the normal and the more accurate the measurements, the nearer the curve will show the law of habit. The daily variations in the curve can be explained by the variations in the subjects.-^ Partridge studied the reflex wink. The work consisted of having a little hammer hit a glass plate immediately before the eye. The two adult subjects used were able to improve greatly their ability to inhibit the wink and their curve of improvement shows a steady rise from the beginning with the usual daily fluctuations in the successive tests. ^- Bair studied the acquisition of control over a voluntary muscle and found tliat the power was gained by associating the given muscle in a group with others. After the power was gained to " Scripture, K. W., Smith, T. L., and Rrown, E. M. "On the Education of Muscular Control and Power." Studies from the Yale Psy. Lab., vol. 2, pp. 1 14- 1 19. "Johnson, W. S. "Researches in Practice and Habit." Studies from the Yale Psy. Lab., vol. 6, pp. 51-103. ^ Partridge, G. E. "Experiments upon the Control of the Reflex Wink." Amer. Jour, of Psy., vol. 11, pp. 244-250. ACQUISITION OF SKILL li move it with others, the abihty to move it separately was obtained by concentrating the attention upon the particular act. He also studied the nature of the improvement of subjects practicing on throwing shot into a tumbler and manipulating a modified form of typewriter. These tests were continued only for a few days and do not give much idea of the curve of improvement.-^ Wells did some work with special reference to fatigue phe- nomena but also touching upon the problem of improvement. The unit of measurement used was the taps upon a telegraph key. The subject selected his own method of tapping but did not change it during the experiment. The subject commenced at a given signal and continued at maximum speed for thirty sec- onds. After a rest of two and one half minutes this was re- peated. Five trials of thirty seconds each were given each day. Two of the subjects practiced for thirty days. The work of these two subjects is the only part that is important here. The curves show a constant improvement until near the end, when progress gradually diminishes. There is no rapid rise at the beginning of the curve. In a later test on one of the subjects there was an improvement over the best score at the end of the practice. There was evidence of warming up for both subjects, and as the practice continued this warming up process increased. After an intermission there was a loss in the right hand in all cases and a loss in the left hand in the first case but a gain in the second and third for Subject i. Subject ii in all three cases with each hand showed a gain. With both subjects, the first interval of five seconds showed a gain but fatigue came in and made the total score lower in the cases stated above. The only explana- tion he offers for the initial increase is that it is the result of a renewed "Neuigkeitsantrieb" but he seems to think that the practiced co-ordination parts become more firmly set during a period of rest. He further points out that the daily fluctuations in this work vary more after a week than at the beginning.^^ ^Bair, J. H. "Develajunent of Voluntary Control." Psy. Rev., vol. 8, pp. 474-510. "The Practice Curve." Mon. Sup. to Psy. Rev. No. 19. ^ Wells, F. L. 'Wormal Performance in the Tapping Tests." Amer. Jour, of Psy., vol. 19, pp. 437-483- 12 WILLIAM HOWARD BATSON Starch studied the trial and error method by having a subject trace a six pointed star as it is seen in a mirror. He finds that the advantages of using a figure in the form of a star rather than some other form are, that it gives frequent changes in di- rection, the divisions are of equal length, and they are sufficiently long to be difficult and yet not cause undue fatigue. The curves were obtained by taking one hundred records made at the rate of one a day on consecutive days. Both an error and a time curve were plotted. The error curve falls rapidly at first and then gradually until near the end of the practice where it seems to have reached a dead level. The time curve shows a rapid fall at first but after about fifty practices it remains nearly parallel to the horizontal axis until the error curve ceases to fall then it takes another drop. Near the end of the practice it again approaches parallel- ism with the x-axis.^^ Whitley gave a number of tests to determine individual dif- ferences. She emphasizes the importance of a number of tests in order to get a true estimate of a person along any line. She also points out the danger of mis-interpreting measurements. She believes that in the same field, improvement follows the same law for different individuals. The higher mental functions are more susceptible to practice than the sensory functions.^^ A large number of other improvement tests have been made, but, since very little of the motor element was involved, they have little connection with this problem and are not discussed. In all the tests mentioned except the first three the period of practice was rather short and very little can be said about the longer period plateaus. In every case the short fluctuations ap- peared. * Starch, D. "A Demonstration of the Trial and Error Method of Learn- ing." Psy. Bui., vol. 7, pp. 20-23. "Whitley, M. T. "An Empirical Study of Certain Tests for Individual Diflferences." Archives of Psychology, No. 19. ACQUISITION OF SKILL 13 Problem Stated From this work have grown a number of interesting problems and not a few conflicting opinions are held in regard to the questions raised. The influence of the objective and subjective factors on different types of learning and on the same type of learning at different stages of development, the daily fluctuations in the curve, the warming up process, the general form of the curve, the effect of short and long rest periods — all present fields for further study. One of the most troublesome problems is that of the plateaus. Are they the result of different habit levels, are they due to the automatization of associations, are they depend- ent upon attention, or may there be some other explanation ? Are they necessary for all or any types of learning? The purpose of this study is : first, to get more information in answer to the questions stated above; second, to gain such in- formation on the other points pertaining to the curve of learning as the data will afford. The plan of the work has been so to devise experiments that the elements determining the curve of learning can be analysed. In the first set of experiments, the factors affect the process of the learning simultaneously. In the second set, the work is reduced to its simplest form and only a relatively simple sensori- motor reaction is required. In the third set, several factors in- fluence the same learning process but they act in succession. On the basis of this work, the curves are plotted and the explanation of their form is given. The study ends with a general discussion of these results in their relation to the results that have been ob- tained in other experimental work in this field. THE EFFECT OF A NUMBER OF FACTORS WORKING SIMULTANEOUSLY Method of Procedure and Results This part of the work is a continuation of Swift's ball-tossing experiments. The only apparatus used was two hollow rubber balls and two pieces of hard rubber. The balls weighed 16.7 and 1 7. 1 grams and were 1.6 inches in diameter. The pieces of hard rubber were cylindrical in form, weighing 8.91 and 8.94 grams and being .5 inches in diameter and 2.1 inches long. Neither the difference in weight of the pieces of hard rubber nor of the balls could be noticed. Five subjects took part in the work. Subjects D( Decamp), B(Batson), W(Wang), and F(Foulk) were all graduate stu- dents taking research work in Psychology. Subject H( Hayes) was a junior in college taking Psychology. Experiment I. Subject D. Subject D. was instructed simply to toss and catch the ball so that one ball would be always in the air. He had had no previous experience. The right hand was used. The practice consisted of ten trials daily at 10 A. M. The work was commenced on October 8, 1912 and continued until November 23, 1912. After the first three days the same room was used for practice. The subject knew his score daily as well as that of each of the other subjects doing the work at that time. There was no limit put upon the time between trials. At first scarcely any time was taken but as the success became greater and the work was more fatiguing a minute or two was taken for rest between successive trials. Table I shows the result of this practice in detail. ACQUISITION OF SKILL IS TABLE I Subject — D. Experiment — Tossing and catching two rubber balls with right hand. Trials I 2 3 4 5 6 7 8 9 10 Sum Oct. 9 I 2 2 2 3 3 3 I 2 2 21 10 2 2 4 7 3 3 7 I 3 6 38 II 3 I 8 3 3 3 I 2 3 2 29 12 2 3 2 3 2 2 5 2 2 3 26 13 2 3 3 II 4 I 3 2 2 5 36 14 7 I 2 9 3 I 2 I I 4 31 15 5 4 3 4 2 2 2 I 3 3 29 i6 3 6 3 3 2 2 3 3 2 2 29 17 24 2 2 4 10 2 5 3 5 3 60 i8 6 12 4 4 5 8 3 8 3 3 56 19 10 7 4 10 2 5 I 2 4 I 46 20 18 4 2 3 4 3 4 8 46 21 2 9 5 I 4 3 3 2 4 II 44 22 5 5 4 3 6 17 3 9 2 54 22, 13 4 9 5 4 I 2 4 9 3 54 24 4 14 7 8 33 2 3 14 85 25 22 3 6 2 5 3 7 4 5 57 26 6 12 4 5 9 13 II II 5 6 82 27 9 9 16 22 3 7 2 II 3 82 28 ID 20 9 7 24 29 22 2 35 26 184 29 10 I 16 35 7 3 I 27 I 4 105 30 23 23 4 31 21 2 28 17 15 5 169 31 17 3 35 40 4 17 18 7 18 27 196 Nov. I 19 I I 9 I 8 2 8 32 18 99 2 21 10 40 19 35 40 I 6 12 14 198 3 2,7 II 34 6 12 32 62 5 56 45 300 4 52 4 20 II 32 58 9 28 2 61 277 5 52 25 47 49 45 5 59 3 45 6S 395 6 17 4 67 43 6 6 4 10 29 114 300 7 14 13 32 29 5 34 34 9 I 171 8 18 23 22 16 18 10 39 48 96 54 344 9 84 23 4 60 80 22 16 21 7 33 350 10 24 30 6 27 9 33 48 37 5 5 224 II 2 51 25 24 27 67 13 72 16 24 321 12 8 22 38 9 81 27 24 57 28 66 360 13 2,7 42 64 51 88 8 64 18 60 31 471 14 42 15 21 26 37 64 102 77 30 24 438 15 57 5 39 29 27 23 16 62 76 12 346 16 100 2 2,2 35 46 3 43 42 17 60 380 17 48 55 27 48 10 3 9 30 5 235 18 22 25 lOI 23 35 65 37 123 17 53 511 19 8 26 88 265 16 95 82 19 32 78 709 20 36 151 20 29 74 79 58 30 48 62 588 21 3 151 259 170 58 70 247 181 265 1404 22 116 88 72 94 80 53 179 222 47 124 1075 23 113 30 137 178 65 74 352 96 24 1069 Total 1 134 943 1357 126 1 1 169 958 1 175 1616 1 1 15 1396 12124 When these trials are divided on the basis of those that fall below twenty, those between twenty and ninety, and those that are ninety or above, the following results are obtained : i6 iril./.I.IM l/oir.lh'l) liAlSON li(l((W JO JO to H 52 Tliinl lliiid of table .^7 (/> 24 '\\\v. lioi i/oiit.'il coluniiis opjjositc tlic dates show bow many calclics were made each trial. The vertical colmiiii to the rif^dit inaikcd "Sum" shows the iiiiiiihcr of calchcs that were made during the ten trials ol" each day's practice. The last set of lij.;iires at the Ix^ltcjin of each vertical colmim shows the sum of the columns. This j^ivcs liic sum of all liic lirst (rials, second liiais, elc. 'liic luimhcr at the hasc of the coinmii marked Sum shows the lolal luimhcr of cite lies that ihe suhjccl made in all the practice that was taken. 'J'he subject was tested a^aiu under the same conditions on AujMist I 1^ and 17, K)!^, ;uid the foiiowiuL; lij.;urcs wiTe ol)taiued : Triul.s I Ji .\ ,| 5 6 7 « <> 10 Sum Auk. is ^.'1 <>o «-| loH K(> .s I'**' J<)7 -'"^ 'U -'5,^ JSi I.V 7S .|^~' -'S.SO Ihe balls foi this (est wcie similar to those used above but wei^^hed 20.;^ and iX. 1 grants respectively. 'I'hc test was made in a smaller loom but it is not liki-ly that this bad any effect upon Ihe throwing;. I he subject had not practiced fiom the time Ihe rcf^iilar practice bad ended up to the time of the test of Aiioiist 15, i<)i,^ r.clvvccn that tim<- :uid the test in July I';i4, be bad occasionally tossed dilTereiit objects but was sure that all Ibis practice would nol amount to a half hour. I'.x n-.in M i:n !' II. Siii(|i;(f 1'. Tn this experiment the work was the same as in the precedin}*" one, that is to toss and catch two balls. Tbi- method was to center the attention on some one phase of the work. The sul)- jcct aiiuc(| to lix his attention oii the luauucf of pitchino the balls. 'Ihe balls were piUhed in a circle from rij;bt to left. Only u very little incidental practice in this work had taken place ACQUISITION ()!■ SKILL 17 before the experimcnl. Tlie ri}j;lil liaiul was used. 'Ilic practice was in the forenoon and i onsislcd of ten trials daily. 'Ihc work coninienced on October 15, 1912 and continned nntil N(;venib(T II, \(ji2. All the practice was in the same room. 'Ilie daily score was known. As much time was (aken between the separate trials as the subject desired. Table ll.i shows the residt of this work. Table I la is arran}j;ed the .same as Table I and the ex- planation of the first table may be used for this one. TAI'.IJ-: I fa Subject— B. ■Experiment — To.ssiiiK and catcliing two riililHr l)alls willi riglil hand. Trials i 2345678910 Snm Oct. i.t; 7 H 6 6 T 6 4 II 9 '7 7S i() I 10 1.^ u ;^ 7 4 4 s <> S4 17 .3 l^ 7 20 10 2,i 12 10 10 17 125 iH 12 J 3 I 5 22 22 s .3 7.S i<) 10 27 2 ;i^ I 15 16 K H 10 132 20 12 H () 30 4 II 24 13 31 35 177 21 6 I 32 3 47 24 25 10 2S 176 22 7,7 7 I 41 35 .14 4 i5.3 3« H 2^21 23 13 19 23 6 49 i« 14 36 2H 23 229 24 I 23 2 41 12 4 3« 31 22 2<> 203 ^.i? n 7 12 28 6 37 213 35 46 14 219 26 4 6 2K 70 45 \U 19 52 17 18 275 27 " *J ^.S 3.S 4.S 24 53 46 24 lo 312 28 34 77 .38 7 o 20 1 1 79 28 70 364 29 20 40 24 42 64 i.s 79 (^x) <;<> ^>7 .Sfo 30 08 .S8 52 142 31 52 33 \() <> 8 4 3« 7 72 3 i.S 31 3.S 3« .3.S .S4 4' .33' ■Nov. I 29 75 Tk; 21 UK; 48 112 ^;i yy (/> O39 2 /}4 16.S 148 92 19 .S^> K^ 40 34 .S9 765 3 ' ^'3 110 41 13 II 27 34 31 I dy ,}<>8 4 3 "4 24 II 63 22 117 79 (x) 13 y/) 5 25 20 47 23 29 5'^ 75 f<») I.S 184 577 6 23 19 104 63 F23 118 92 42 y:^ 137 793 7 183 31 no 163 153 m 142 (>.\ 74 1120 8 f.S9 107 112 33 26 58 53 95 18 42 703 9 31 46 124 131 42 39 65 84 74 245, 881 10 114 67 43 46 262 134 149 96 84 197 1192 11 29 206 184 65 179 33a 14 7 I2X 134 1271 Total 927 1359 12/14 ' '^> '37-^ M'^» '.34<' '24.S "'7.S '7.^2 12919 When the trials of Table I la are divided (jn the basis of those that fall below twenty, those between twenty arul ninety, and those ninety or ab(;ve, the following.; residts are obtained : Hclow 20 20 to 90 "r more First third of tahle 64 26 o Scconfl third of table 27 58 s Thirrl tJiird of table 12 47 41 l8 WILLIAM HOWARD BATSON After the subject had been able to make an average of one hundred catches per trial on successive days with the right hand, the practice was changed to the left hand. The method was to give attention to all the particulars, — the force of the throws, the direction, the manner of catching the ball, the position of the body, etc. There had been no previous practice with the left hand. The practice consisted of ten trials daily in the forenoon. The work commenced November 12, 191 2 and continued until January 15, 191 3, with an intermission from December 20 to January 8 and one practice period on November 28. A part of the practice was in a smaller room than that in which the work commenced, but the walls were the same in color and the lighting was about the same. The daily score was known. No time limit was set between trials. The result of this practice is shown in Table lib. TABLE lib Subject — B. Experiment — Tossing and catching two rubber balls with left hand. Trials i 2.'? 4 S 67 8010 Sum Nov. 12 2 4 2 4 4 2 6 2 5 31 13 3 2 2 3 7 5 6 3 I 8 40 14 6 6 7 9 8 8 9 6 8 2 69 15 21 2 20 3 2 2 2 4 10 66 16 13 13 I 7 10 17 I 9 7 78 17 3 7 6 2 6 2 8 10 6 50 18 6 3 8 21 3 7 4 6 9 4 71 19 S 17 4 4 7 8 7 9 2 10 IZ 20 I II 17 7 3 4 8 II 9 13 84 21 9 4 2 12 II 9 10 25 10 92 22 28 12 4 18 7 14 6 14 22 19 144 23 2 17 14 20 3 19 12 8 II 15 121 24 3 7 II 4 12 II 6 4 9 II 78 25 16 8 II 10 16 5 3 II 8 5 93 26 7 II 7 7 r> 21 10 19 14 98 27 3 5 14 28 13 6 I IS 8S 28 29 2 10 6 5 14 21 15 6 23 4 106 30 6 6 2 21 12 13 20 42 13 14 149 Dec. I 9 15 28 12 20 22 2 20 8 ^1 163 2 22 14 8 12 13 15 I 7 8 13 113 3 10 23 4 13 16 IS 21 20 II S 138 4 28 16 6 9 19 20 27 2 26 3 156 5 15 II 14 12 25 22 10 29 12 24 174 6 9 31 19 12 23 16 29 5 7 39 190 7 10 3S 10 40 20 20 17 32 6 23 213 8 22 17 17 28 40 27 17 29 3 14 214 9 21 50 8 49 5 21 26 41 13 7 241 10 31 40 35 34 5 24 37 II 26 S 248 ACQUISITION OF SKILL 19 TABLE lib Subject — B. Experiment — Tossing and catching two rubber balls with left hand. Trials 123456789 10 Sum Dec. 1 1 5 6 35 39 17 30 46 21 II 45 255 12 26 56 II 4 8 27 24 9 24 52 241 13 19 71 3 53 63 23 7 21 36 41 227 14 74 90 34 30 5 7 60 7 I 39 24-7 15 60 18 8 2,^ 140 83 29 39 59 83 551 16 3 5 63 178 80 38 3 104 45 96 615 17 no 56 69 79 35 64 88 48 45 29 623 18 2^ 5 25 50 204 43 20 32 42 459 19 31 34 12 28 75 II 22, roi 46 74 435 20 35 91 I 21 22 27 H 35 6 252 Jan. 8 14 I5'3 40 66 85 5 91 45 42 67 608 9 lOI 61 16 46 127 108 9 19 23 6 516 10 58 103 62 31 44 41 10 95 51 16 511 II 83 i7 57 24 9 149 85 31 45 76 596 12 45 44 118 30 48 12 17 94 II 108 527 13 118 65 73 30 8 133 60 123 117 119 846 14 35 130 289 48 166 119 133 162 153 115 1350 15 42 73 268 241 156 10 161 26 52 141 1170 20 to 90 90 or more 7 63 88 38 Total 1210 1495 1470 1416 161 I 1241 1233 1332 I 143 1466 13617 When these trials of Table lib are divided on the basis of those that fall below twenty, those between twenty and ninety, and those that are ninety or more, the following- results are obtained : Below 20 First third of table 143 Second third of table 87 Third third of table 34 The subject was tested again on July 17, 1914. The condi- tions were the same except that the balls were slightly heavier, weighing 20.3 and 18.1 grams respectively. The difference in weight, however, could not be noticed. With the right hand, the following score was made : Trials 123456789 10 Sum July 17 329 4 7 142 o 86 100 I 227 228 1 124 July 18 63 82 55 201 130 257 169 60 2^ 384 1437 The left hand was tested on July 17, 191 4, and again two days later. The score was: 20 WILLIAM HOWARD BATS ON Trials 123456789 lo Sum July 17 10 17 57 7i 14 7« 4i 98 52 49 492 July 19 2Ti 70 59 52 88 46 68 130 114 52 702 The subject had tossed tennis balls a few times after the regu- lar practice had stopped but not over fifteen minutes had been spent in such work. There had been no practice with the left hand. A rest of two minutes was made between each trial of this last practice. Experiment III. Subject W. The subject was instructed to toss and catch the two balls after the same manner as the other subjects. The practice consisted of ten trials daily at eight A. M. The work was commenced Oct. 30, 1 912, and continued until Feb. 12, 1913. During this time, the following daily practices were missed : Nov. 3, 5, 6, 21 ; Dec. 7, 8, 12, and the period from Dec. 20 to Jan. 8. After a few practices, the subject rested one minute between trials. He had had no previous practice in the work. The results of his work are shown in Table III. TABLE III Subject — W. Experiment — Tossing and catching two rubber balls with right hand. Trials 1234 5 6789 10 Sum Oct. 30 1 I 1 4 I 2 3 I 3 3 20 31 2 2 3 3 2 2 3 4 2 2 25 Nov. I 2 2 2 2 2 3 2 I I 3 20 2 4 I 3 4 3 I 2 1 3 2 24 3 4 2 I 2 3 I 2 3 4 I I 20 5 6 7 I 2 2 3 3 5 5 4 3 28 8 2 2 3 I 4 5 6 2 7 5 37 9 4 5 2 5 3 6 7 5 5 3 45 10 5 5 4 8 3 5 3 6 6 45 ] I 5 7 3 3 5 3 2 6 34 12 4 14 2 8 2 2 4 2 5 3 46 13 5 4 3 6 4 3 7 2 I 5 40 14 6 4 7 3 II 2 6 6 2 4 51 15 4 3 4 5 2 4 6 3 6 6 43 16 5 7 18 8 9 3 4 I 8 I 64 J7 5 3 7 5 8 2 5 I 36 Nov. i8 2 4 4 8 3 3 I 2 7 34 19 II 6 7 2 14 2 9 16 9 76 20 10 6 I 3 2 11 2 9 4 8 56 21 22 2 3 4 13 I 2 7 2 II 45 ACQUISITION OF SKILL 21 TABLE III— {Continued) Subject — W. Experiment — Tossing and catching two rubber balls with right hand. Trials 123456789 lo Sum 23 3 15 17 4 3 7 5 12 13 2 81 24 4 2 5 9 8 7 9 12 8 15 79 25 4 22 7 20 4 13 9 4 8 3 94 26 3 8 7 20 ID 7 4 II 9 3 82 27 12 8 II 22 7 6 18 I 4 24 113 28 2 4 8 5 2 6 8 8 4 4 51 29 II 2 7 I 8 5 I 7 42 30 15 15 4 II 26 9 12 4 17 5 118 Dec. I 5 4 10 9 4 10 8 5 5 2 62 2 4 10 5 6 9 23 6 17 12 3 95 3 8 29 21 3 19 16 22 7 II 16 152 4 II 25 7 7 5 25 7 4 29 10 130 5 3 15 28 32 17 5 16 24 12 152 6 7 I 7 II II 52 2 6 12 12 114 8 9 I 9 2 9 24 24 16 8 19 17 129 10 29 20 34 10 32 7 10 52 31 5 230 II 12 5 17 8 17 21 25 53 6 22 34 208 13 7 14 29 9 66 5 7 39 178 14 12 23 6 29 35 25 2 6 42 9 189 15 8 I II 10 2 28 9 12 15 96 16 23 14 29 13 39 13 40 28 14 213 17 3 12 16 31 35 19 31 36 12 49 244 18 28 II 19 4 63 10 6 24 13 178 19 33 12 29 81 21 23 26 43 13 62 343 20 96 52 5 58 7 4 16 55 39 18 350 Jan. 8 12 10 II 25 55 II 17 18 47 32 238 9 21 8 23 80 19 8 23 44 II 18 255 ID 22 25 8 49 40 8 21 4 15 84 276 II 10 66 5 12 10 8 17 17 4 149 12 8 35 40 63 3 7 II 8 22 50 247 13 30 52 56 50 13 62 34 63 I 361 14 32 30 III 39 80 42 13 21 23 28 419 15 59 27 20 55 30 24 13 24 24 276 16 8 84 36 24 3 16 27 33 32 31 294 17 10 7 13 87 24 37 24 58 19 68 347 18 II 24 16 41 12 28 50 23 32 6 243 ^ 28 36 26 23 19 39 23 51 40 26 311 20 26 27 25 46 13 18 45 14 25 10 249 21 8 32 10 18 56 II 19 32 26 66 278 22 24 51 55 25 27 10 14 50 t6 10 282 23 3 19 35 50 55 91 20 75 33 42 423 24 23 79 15 78 60 15 3 13 6 22 314 25 4 2^ 22 38 28 24 27 73 30 50 322 26 3 74 64 32 63 50 39 38 68 17 448 27 9 41 33 28 62 67 20 43 38 4 342 28 21 7 8 9 no 74 54 124 9 50 466 29 61 93 95 60 7 17 36 I 9 18 397 30 19 10 27 15 53 21 20 15 57 50 287 Second trial on same day 30 27 19 39 23 43 14 49 27 40 30 311 31 54 69 33 97 29 66 33 25 13 43 462 Feb. I 32 30 II 2 16 46 28 48 89 37 339 2 87 36 5 23 68 72 12 20 26 22 371 22 WILLIAM HOWARD BATS ON TABLE lll~{Continucd) Subject — W. Experiment — Tossing and catching two rubber balls with right hand. Trials 123456789 lo Sum Feb. 3 43 51 62 14 8 Z7 36 25 48 88 412 4 3 29 89 22 64 105 36 79 90 41 558 5 19 51 9 32 loi 50 29 41 126 27 485 6 50 9 53 97 54 58 91 I33 127 26 698 7 4 41 43 3 83 II 52 84 79 127 537 8 26 24 105 Z7 2,3 79 81 133 16 48 582 9 72 136 106 2y 186 61 12 52 72, 16 741 10 67 9 106 100 44 71 114 III 70 152 844 11 19 77 118 160 159 172 88 66 23 117 999 12 181 no 132 59 72 103 178 116 36 118 1 105 Total 1485 1997 2090 2120 2381 2093 1842 221 1 1923 2038 20180 When the trials of Table III are divided on the basis of those that fall below twenty, those between twenty and ninety and those ninety or above, the following results are obtained : Below 20 20 to 90 90 or more First third of table 265 5 o Second third of table 158 no 2 Third third of table 68 176 36 Subject W. was tested again under the same conditions on June 26 and 27, 191 3. He had had no practice in the meantime. The results were: Trials 123456789 10 Sum June 26 2 62 27 68 72, 2,7 167 92 66 116 712 June 27 53 51 173 230 36 170 67 130 181 215 1306 On July 3, 1 914, the subject was tested again in the same room and under the same conditions except that the balls weighed 20.3 and 18. 1 grams respectively. The scores were: Trials i 23456789 10 Sum July 3 146 55 116 115 105 118 201 205 156 145 1362 The subject had thrown a few times three or four months before with tennis balls but with no intention of making a high score. He had also tested himself once before with these same balls but did not think that over twenty minutes had been spent in such practice. Experiment IV. Subject F. Part I In the first part of this experiment, the subject was to toss and catch the two pieces of hard rubber described above in the same manner as the other subjects caught the balls. No specific directions in regard to the manner of attending the work were ACQUISITION OF SKILL 23 given. The subject had had some Httle experience in tossing and catching balls. The right hand was used. The practice con- sisted of ten trials. The work was commenced on Oct. 9, 191 2 and continued until Jan. 12, 191 3. The practice was on alter- nate days from Oct. 9 to Nov. 19 and after that it was daily. The following practices were missed : Nov. 24 and from Dec. 20 to Jan. 8. The same room was used for all the work except a few practice periods near the end. The last room was not very different from the former one and no change was noted as a result of this. The subject knew his score. He took as much rest between trials as he desired but it was never over two or three minutes. The results of his work are shown in Table IV which is arranged on the same plan as those for the ball tossing. TABIE IV Subject — ^F. Experiment — Tossing and catching two pieces of hard rubber. Trials 123456789 10 Sum Oct. 9 2 7 9 8 12 2 6 9 7 6 68 II I 4 6 5 2 8 II 10 9 9 65 13 3 3 9 19 5 9 7 13 7 3 78 15 2 2 3 6 10 13 6 4 15 5 66 17 I 5 8 3 7 26 4 12 13 18 97 19 17 3 14 14 9 II 7 6 3 32 116 21 17 2 10 9 4 18 7 6 12 5 90 23 II 14 6 3 8 15 12 II 9 89 25 33 10 2 13 3 6 4 6 4 10 91 27 20 3 9 4 22 8 30 9 12 9 126 29 13 3 6 47 12 II 22 2 II 127 31 25 17 9 10 2 9 3 5 II II 102 Nov. 2 7 3 9 10 14 2 26 5 3 7 86 4 3 12 13 35 12 19 6 II 8 3 122 6 13 II 7 5 21 39 40 12 18 17 183 8 6 2 II 29 7 5 13 22 6 lOI 10 17 8 4 16 3 2 15 14 9 88 12 7 24 II 7 10 5 15 20 7 52 158 14 5 6 9 17 12 57 7 64 6 6 189 16 24 9 25 15 37 12 I 47 32 56 258 18 9 13 3 18 14 4 7 39 41 4 152 19 18 28 21 17 16 S3 28 17 12 14 224 20 5 21 10 7 9 26 39 12 8 4 141 21 7 15 32 5 31 21 20 13 22 166 22 20 8 53 22 40 II 21 3 35 3 216 23 14 16 12 31 20 19 3 28 II 154 24 25 17 44 5 7 22 12 10 35 2 20 174 26 7 30 14 25 2 12 37 41 29 9 206 27 34 7 II 32 18 46 26 9 7 25 215 28 8 16 30 47 4 19 4 13 8 9 158 29 41 4 26 22 4 14 2 42 26 7 188 30 3 10 24 5 13 14 I 70 24 WILLIAM HOWARD BATSON . . TABLE IV Subject — F. Experiment — Tossing and catching two pieces of hard rubber. Trials 123456789 lo Sum Dec. I 34 2 13 22 75 32 15 4 8 18 223 2 24 45 5 n 13 8 12 I 36 78 233 3 13 3 12 34 14 7 9 13 13 53 171 41120 o I 16 12 5 6 3 22 96 5 19 18 3 28 39 36 3 49 44 9 248 6 20 34 9 S7 52 54 44 46 13 26 335 7 14 15 5 31 II 30 56 95 3 36 296 8 10 69 35 40 108 59 121 10 8 104 564 9 IS 103 53 67 85 28 no 25 24 22 532 10 7 194 193 66 157 128 IIS 46 53 6 965 11 121 119 13 133 72 44 103 71 108 57 841 12 68 34 53 23 141 130 88 146 41 84 808 13 76 91 91 113 113 122 44 66 32 61 799 14 36 145 IIS 12 96 2 38 98 18 64 624 15 44 165 Si 30 no 52 y7 24 112 78 775 16 54 45 19 121 208 71 69 4 179 71 841 17 66 265 294 195 III 59 30 15 54 197 1286 18 I 71 102 120 180 II 285 60 77 29 836 19 91 12 136 16 31 2S 159 88 128 33 719 20 10 IIS 117 137 76 98 57 5 95 36 746 Jan. 8 20 37 89 60 I 54 24 19 102 80 486 9 14 60 41 24 99 35 59 117 77 2,7 563 10 6 113 72 9 159 23 82 67 35 57 623 n 4 191 318 187 9 no 19 25 39 165 1067 12 221 484 20 391 78 123 254 109 159 116 1955 Total 1409 2800 2232 2410 2466 1885 2340 1752 1851 1951 21096 When the trials in Tal)le IV are divided on the basis of those that fall below twenty, those between twenty and ninety, and those ninety or more, the following results are obtained : Relow 20 20 to 90 90 or more First third of table 169 21 o Second third of table 114 76 o Third third of table 31 90 69 The subject was tested again on June i, 1913, and made the following score : Trials 12 3 456789 10 Sum June I 29 157 158 69 iS4 no 10 54 21 762 June 2 4 97 27 31 19 106 75 67 14 184 624 No practice had been taken between the end of the regular practice and this date. Experiment IV. Subject F. Part II In this part of the experiment the subject was requested to toss and catch two balls, keeping one in the air all the time. The same balls were used that had been used by the other sub- ACQUISITION OF SKILL 25 jects. The practice was with the left hand. The subject had had no previous practice with that hand. In this work, the subject continued each day until he had made one thousand catches. There was a rest of one minute between each hundred catches. The work was commenced on Jan. 13, 191 3 and con- tinued daily until Feb. 4, 191 3. The number of misses before a hundred catches were made in each trial and the total number of misses in each day's practice are shown in Table V. Where zero appears in the table, the subject was able to make a score of one hundred without missing. It should be noted that the number of misses counted is always one less than the number of trials for each one hundred catches. When the columns are added vertically so as to show how many misses were made in catching the first hundreds, second hundreds, etc., the following figures were obtained: 1st 2nd 3rd 4th 5th 6th 7th 8th 9th loth Sum 44 27 32 27 38 52 2,(> 26 23 35 350 The total number of catches was 23000. This represents the amount of practice the subject had. Table Va shows this work in detail. This table, with the exception of the first day's prac- tice, shows where each miss was made for each one hundred catches. TABLE V 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th Sum 100 100 TOO 100 100 100 100 100 100 100 Jan. 13 4 4 5 8 4 12 8 3 4 4 56 14 5 5 4 8 9 6 2 2 I 3 45 15 16 3 2 I 5 5 7 5 4 I 2 35 2 2 3 3 4 2 4 I 3 4 28 17 3 3 2 2 5 2 2 r 2 2 24 18 3 3 2 I 2 I 5 17 rg 2 I I I I 3 3 12 20 4 I I I I I 9 21 I 2 I I I I I 3 II 22 2 I 3 2 I I 10 23 2 3 2 2 4 I 3 2 19 24 2 I 2 4 I 4 I 2 I 18 25 2 I 2 3 I 9 26 2 I I I 2 I 9 27 I I I 3 I 2 I I 12 28 2 I I I 6 29 I 2 30 2 I 2 2 I I 10 31 I I 2 I 5 Feb. I I 2 3 2 I I 2 3 4 I I I I 8 4 Total 44 27 32 Z7 38 52 36 26 23 35 350 26 iri/.IJ.IM IIOir.lNI) HATSON TAHLIC Va 'Siil)jccl— 1'\ ICxprriiiu'iil — Tossing' and catcliiiiK two rirhbcr balls witli left liand. 'Each practico was a lliousand caUlics witli one niiinute rest between each liiuulred. Tlie lirst horizontal row in each KToup represents the number of misses in eacii Innidred. 'riie liKnres just below show the count at which the miss was made. Tlie cohnnn at (he rinht siiows the misses per one tiiousaiul catches. T rials i 23456789 lo Total JanT 13 4 4 5 8 4 ig 8 3 4 4 56 Jan. 14 5 5 4 8 9 S 2 2 i 3 45 28 20 9 II 16 13 74 64 74 39 44 4« «^ 13 20 36 (X) 4 Xi 41 -'0 .vH yb 0; 77 94 100 1" 49 51 67 87 97 Jan. 15 3 8 2 I 5 5 7 5 4 I 2 35 .50 fi7 24 42 II 31 45 32 29 17 91 60 46 21 42 71 98 33 66 92 KK) 49 67 85. 40 43 52 67 73 90 92 96 79 86 Jan. 16 2 2 3 3 4 2 4 I 3 4 27 37 55 4« 32 25 26 14 10 25 10 55 99 54 88 61 100 60 71 74 74 64 81 64 85 21 48 59 Jan. 17 3 3 2 2 5 2 2 I 2 2 24 37 6 34 14 16 44 20 4 56 26 66 .VI- .|S .IS 17 07 75 6() 47 87 .S<) 2() 77 94 I 2 I 5 Jan. 18 3 3 2 17 ,s6 42 75 51 52 69 3 68 57 ')4 61 13 81 67 40 70 86 3 .Tan. 19 2 I I I I 3 12 39 69 44 42 4 35 2 57 86 88 / 61 Jan. 20 4 o I o I o I ^ ,!. ° ^ n 78 5^ io 99 66 23 60 70 __ Jan. 21 I o 2 I I 1 I 103" 35 io 7 34 ^7 7» 00 67 79 98 99 ACQUISrnON ()!■ SKILL 27 TAllU.!'! Va (Continued) Subject — I"'. Experiment — Tossing and catching two riil)l)er halls willi left hand. I'",acli practice was a thousand catches with one minute rest between each hundred. The first horizontal row in eacli group represents the lunnber of misses in each iiundred. 'l"he figures just below sliow llie count at which the miss was ma(k'. Tlie column at flic riglit shows tlic misses per one tliousand catches. Jan. 22 2 I 3 2 I I 10 I 33 45 24 41 67 12 76 26 Jan. 23 2 3 2 2 4 I 3 2 19 80 .17 (>^ 77 .SI 9.S 46 86 66 81 87 98 36 3« 7« 76 22 91 Jan. 24 2 I 2 4 I 4 o 1 2 I i8 Fd) Feb. 2 I 2 4 I 4 iM 7.3 .4 3.3 70 46 00 83 .S8 87 90 69 7.S' 79 94 48 54 71 Jan. 25 2 I 2 3 I 9 I 92 17 14 gg 10 91 60 Jan. 26 2 I I I I 2 I 9 9 74 92 75 54 3 73 J5 66 Jan. 27 III I o 3 I 2 I 3 St 82 ^63 39 40 42 I 10 <>7 50 70 Jan. 28 2 I 1 I I 6 62 77 57 4 3 94 Jan, 29 1 I 2 2 88 75 84 62 26 45 80 10 Jan. 30 o 2 I I o 2 o 2 I I 10 85 54 95 47 32 89 52 , 87 74 70_ Jan. 31 I o 102 100005 79 5^ 60 80 50 Feb. I I 2 3 78 82 Feb. 2 I I 2 75 I 28 WILLIAM HOWARD BATSON Experiment V. Subject H. This work was the tossing- and catching two balls. No fur- ther instructions were given than that the subject should toss and catch one ball while the other was in the air. The practice con- sisted of five hundred catches, or as near as the score under the conditions could be kept to that number, daily. In the last trial that commenced below the five hundred, the subject was permit- ted to continue until he missed. This sometimes ran slightly over the five hundred. The practice was taken in the evening between seven and eight o'clock by electric light. The work commenced Feb. 3, 19 14 and continued until March i, 191 4. The following- days of practice were missed: Feb. 10, 18 and 28. The subject had had no previous experience in tossing balls. A rest period of five minutes was given as near the two hundred and fiftieth catch as it was convenient to make it without stop- ping the subject in the midst of a trial. The results of this prac- tice are shown in Table VI. TABLE VI Subject — H. Experiment — Tossing and catching two rul)bcr balls with rig'ht hand. Each practice was live hundred catches with live minutes rest near the 250th eatch. The numbers at which the mis.ses were made are given below. Feb. 3—1, 2, 3, 5, 7, 9, 10, 13, 15, 18, 22, 26, 29, 30, 32, 34, Z7, 40, 44, 46 49, 51. 55, 56, 57, 57, 61, 62, 63, 64 67, 70, 72, 75, 78, 79, 80, 84, 85, 88 93, 95, 100, 102, 102, 103, 105, III, 112, 116 123, 125, 129, 133, 135, 139, 140, 148, 149, 153 160, 163, 166, 171, 173, 176, 179, 182, 184, 186 188, 193, 195, 198, 201, 203, 205, 208, 212, 215 219, 221, 223, 224, 227, 231, rest period, 232, 233, 238, 240, 241, 251, 265, 276, 277, 281, 293, 296, 299, 301 303, 304, 308, 317, 325, 330, 334, 336, 341, 343 349, 351, 358, 365, 368, 384, 386, 396, 404, 410 414, 418, 428, 431, 434, 435, 437, 442, 445, 446 449, 454, 455, 460, 465, 468, 474, 477, 483. 488 491, 492, 494, 497, 500. Total 145. P^b. 4—4, 7, 12, 14, 16 19, 21, 23, 25, 28, 31, 34, 38, 43, 46 47, 49, 49, 58, 64, 64, 65, 71, 75, 77 81, 85, 92, 96, 100, 105, 109, 109, 120, 120 123, 128, 131, 137. 141, 143, 146, 149, 150, 155 163, 177, 181, 189, 193, 200, 208, 218, 222, 226 22i5, 229, 231, 237, 247, rest period, 252, 2S3, 258, 263, 268 280, 284, 289, 292, 293, 294, 297, 319, 326, 329 335, 336, 347, 356, 357, 357, 360. 369, 374, 377 381, 383, 394, 40O, 409, 412, 426, 432, 437, 442 445, 446, 448, 452, 458, 461, 463, 469, 475, 481 493, 496, 496, 498, 502. Total 1 10. Feb. 5—1, 3, 8, 10, 12 13, 18, 22, 22, 25, 28, 28, 28, 34, 36 2>7, 2,7, 41, 45, 46, 48, 52, 55, 57 59, 60, 60, 60, 60, 61, 64, 72, 76, 83 83, 84, 87, 88, 94, 96, loi, 103, 106, 109 no, 119, 126, 131, 135, 140, 141, 141. j^S, 151 151, 152, 152, 158, 161, 164, 168. 174, 175, 188 200, 207, 207. 212, 220 229, 237, 244, rest period, 245, 245, 251, 252, 254, 255, 256, 259, ;268, 281, 283, 301 -^ 321, m, 338, 343, 2>77, 277, 379, 384, 388, 410 425, 425, 439, 461, 481, 482, 494, 496, 508. Total 104. ACQUISITION OF SKILL 29 TABLE VI {Continued) Subject — ^H. Experiment — Tossing and catching two rubber balls with right hand. Each practice was five hundred catches with five minutes rest near the 250th catch. The numbers at which the misses were made are given below. Feb. 6 — II 14, IS, 21, 24, 2,2, 2>i, 44, 45, 47, 67 y2, yj, 82, 86, 90, 91, 92, 97, 99, 103 ■ 107, III, 130, 143, 151, 168, 174, 182, 190, 205 219, 228, 255, rest period, 261, 280, 285, 311, 326, Z2>2>, 336 385, 391, 428, 430, 431, 433, 461, 475, 485, 501. Total 51. Feb. 7 — 7, 17, 19, 21, 41, 60, 78, 96, 114, 132 135, 150, 166, i67j 175, 179, 181, 190, 195, 201 215, 222,, 225, 225, 227, 240, 258, rest period, 263, 265, 268 2yz, 276, 290, 334, 356, 364, 377, 381, 400, 414 423, 444, 453, 455, 479, 486, 504. Total 47. Feb. 8—2, 3, II 13, 17, 27, 27, 30, 36, 41, 48, 60, 70 90, 94, 105, 107, no, 121, 133, 141, 143, 150 150, 172, 203, 235, 267, rest period, 279, 286, 290, 300, 310 318, 322, 326, 334, 345, 375, 388, 400, 403, 407 435, 438, 441, 455, 489, 518. Total 49. Feb. 10 — 16, 17, 20, 21 21, 28, 47, no, 152, 160, 173, 196, 201, 202, 209, 231, 235, 238, 246, rest period, 253, 260, 264, 267, 268 300, 330, 343, 346, 376, 401, 416, 432, 449, 452 510. Total 35. Feb. II — I, 98, 129, 130, 153, 165, 193, 213, 272 rest period, 280, 287, 324, 339, 381, 421, 430, 442, 504. Total 18. Feb. 12 — 39 60, 81, 100, 122, 176, 194, 221, 281, rest period, 290, 318 450, 489, 499, 560. Total 16. Feb. 13 — 49, 93, III, 116, 121 143, 182, 249, rest period, 274, 315, 2€?>, 410, 431, 509. Total 15. Feb. 14 — 0, 42, 69, 131, 190, 203, 238, rest period, 214, 264, 284 301, 32,^, 364, 3^7, 2,72,, 389, 409, 4i8, 450, 453 460, 471, 535- Total 23. Feb. 15—6, 21, 31, 59, 78, 88, 89 116, 135, 150, 169, 174, 185, 193, 199, 204, 205 244, 246, 247, 255, rest period, 257, 272, 285, 288, 289, Z22, 227, 344, 362, 369, 379, 383, 409, 427, 428, 438 472, 488, 494, 505. Total 41. Feb. 16 — 56, 84, 124, 152, 158, 200 219, 223, 234, 242, 296, rest period, 299, 311, 350, 359, 386 391, 395, 402, 446, 497, 506. Total 22. Feb. 17 — 15, 21, 44, 83, 130, 138, 163, 216, 261, rest period, 287, 309, 407, 428, 432 492, 524. Total 16. Feb. 19 — 99, 130, 178, 208, 235, 252, rest period, 290, 328 352, 416, 428, 454, 518. Total 13. Feb. 20 — 97, 159, 200, 240, 268, rest period, 322, 352, 413, 459, 537. Total II. Feb. 21 — 31, 47, 61, 99 126, 131, 149, 150, 162, 180, 185, 216, 217, 222 243, rest period, 263, 267, 278, 227, 3^2, 272, 41 3, 4i5, 434, 455, 460, 461, 523. Total 29. Feb. 22 — 24, 54, 91, 91, 109 134, 14s, 171, 220, 247, rest period, 261, 22S^ 277, 428, 436 456, 463, 474, 476, 477, 489, 494, 517. Total 23. Feb. 22 — 15, 116 149, 186, 208, 237, 298, rest period, 340, 369, 434, 500. Total II. Feb. 24 — 22 61, 200, 205, 231, 322, rest period, 335, 449, 537. Total 9. Feb. 25 — 86, 114 189, 234, rest period, 516. Total 5. Feb. 26 — 53, 87, 187, 266, rest period, 348, 415, 435 551. Total 8. Feb. 27 — 227, rest period, 435, 509. Total 3. 'March i — 332, rest period, 550. Total 2. The table shows the number at which the misses were made, that is in the first practice the subject caught one ball and missed the second one. The next trial he caught one more and then missed. This makes two catches as represented by the two. By 30 WILLIAM HOWARD BATS ON taking the difference between any two successive numbers, the number of catches made during that trial may be found. The breaks, marked " " in the cohimns divide the work into periods of ten trials each. This was done for convenience in plotting one of the curves. The last figures marked "total" show how many misses were made in that particular practice period. The Plotting of the Curves In order that the significance of these figures may be better understood, the curves which they represent have been plotted. Curves I, II, III, IV and V (Plate I) are plotted by letting the divisions on the x-axis represent the practice periods and the divisions on the y-axis, the number of catches. In Curves VI and VII (Plate II), the divisions on the x-axis represent the number of practice periods and the divisions on the y-axis rep- resent the number of misses in the practice period. Curve VIII (Plate III), also, has practice period divisions on the x-axis and the number of catches on the y-axis. Curve VIII is obtained by taking the work of Subject H. and finding how many catches were made in each ten trials. Since he was not stopped at the end of any trial, the only difference between this and the first five curves is that the practice was differently distributed. Each practice period for Subject H. continued until he had caught five hundred or a few more than five hundred balls. In Curve VIII, each ten trials is called a practice period and given a space on the x-axis. Curves IX and X (Plate III) are obtained by taking the sec- ond part of the work of Subject F. and the work of Subject H. and using the riciprocals of the number of trials for the divisions on the y-axis and the number of practice periods for the divi- sions on the X-axis. The number of trials and the number of misses are the same for Subject H. But for Subject F. one trial must be added in each hundred as the last one was not counted a miss. This places him at a disadvantage as he was not allowed to finish out his last trial. Under the given conditions these reciprocal curves have the same form as the curves that would be obtained by taking the average number of throws per ACQUISITION OF SKILL 31 trial in the given practice period, which in these cases includes one thousand or five hundred catches for the divisions on the y-axis. Curves XI, XII, XIII, XIV (Plate II) and XV (Plate III) are plotted by taking groups of one thousand tosses for the di- visions on the X-axis and the average nuuber of catches per trial for the divisions on the y-axis. Tables I, Ila, lib. III and IV are used as a basis for the data. Each table is divided into as many groups as there are thousands of catches. Where a thousand catches ended in the midst of a trial the number that was used out of that trial was used as a fractional part of a trial. For instance, if the first one thousand included the sum of the number of catches made in the first fifty trials and a half of the number of catches made in the fifty-first trial, then the whole number of trials for the first one thousand would be fifty and one half. This divided into the one thousand would give the average num- ber of catches per trial in the first one thousand. In each of these tables there was a fractional part of a thousand left over. The curves do not show this. Another way of handling the data was to divide the catches in each table into groups of one thousand each and then to find what per cent each thousand catches was of the total number of tosses for making a thousand catches. For instance, in Table I, it is found that in order to make the first thousand catches 11 72 -|- tosses (1000 catches plus 172 trials) were necessaary. This gives 85 per cent of success. In each succeeding thousand as the number of trials per thousand catches becomes less the percent of success becomes greater. Curves were plotted on this basis and Curves Ila and IVa (Plate I) are typical examples. Curves were also plotted by taking the highest score in each one thousand group for divisions on the y-axis and the thousand groups representing divisions on the x-axis. Curves Illb and IVb (Plate I) are typical examples obtained in this way. Discussion of the Curves Curve I, II, IV, and V have a rather long period during which they rise but little. It took twenty practices for Curve I to rise above the one hundred level. Half of the number of prac- 32 WILLIAM HOWARD BATS ON tices were spent before the two hundred mark was reached. The drop in the curve at the twenty-fourth and the fortieth practices was the result of the room being cold. The drop at the thirtieth and thirty-third practices was caused by those apparently un- accountable bad days for which there was no reason that could be pointed out. Curve II shows that the practice commenced further along. It is, however, slightly concave to the y-axis. Curve III has a more typical form for this method of plotting- the work. There is a period of slow rise extending over more than half of the practice periods. The two hundred mark was not reached until the twenty-fifth practice. The thirty-sixth, thirty-seventh and thirty-eighth practices were on the three days just before the Christmas vacation and this seems partly, at least, to account for the drop here. It will be seen that after a period of eighteen days rest, the score on the thirty-ninth practice is nearly as high as it had ever been before. Curve IV shows a very slow rise but without any very marked fluctuations. It took thirty-six practices for the subject to reach the two hundred mark, and he did not rise above it per- manently until the fiftieth practice. The fluctuations between the forty-fifth and fifty-second practices represent the period just before and after the Christmas vacation. In this case there was a slight fall in the score immediately after the practices were resumed. Curve V is peculiar in that it shows a very marked decline period. This is, no doubt, the result of too much effort. The subject was very anxious to reach the mark set which was to make an average of a hundred per trial for two consecutive days. After he made the score on the forty-ninth practice, he tried very hard to make a thousand catches in the ten trials on the next day. He said that his attention was so set on making a thousand catches that he forgot how to throw. His failure caused a temporary period of despondency. He thought he never could reach the mark set. The higher level from the nineteenth to the thirty-fourth prac- tice may be explained by the fact that the practices were daily; ACQUISITION OF SKILL 33 before this time they had been on alternate days. The curve shows a somewhat higher level at first than do the other curves. This is possibly due to the subject's previous practice. Curves VI and VII show a very rapid drop at first and then a slower fall toward the end. The fluctuation in Curve VI, perhaps, would not have appeared had the subject not had to practice in a cold room and had he been in a good physical con- dition. Curve VII has two marked rises. The first marked rise at the eleventh practice is the resuU of subject's being in a bad physical and mental state. At this practice he stated that he had had a disagreeable conversation just before and that he was all "worked up." He had been up late for several nights. He stated that his head and eyes ached and that he had no con- trol of his hand. The next night he said that he was all worn out, that he had a cold and that he did not think he would ever be able to get below the mark that he had already set. The rise at the seventeenth practice was due to the subject being in a hurry to get to orchestra practice. In these cases where the practice period is long, it appears that the fluctuations under normal conditions would be rather small. Subject H. felt confident after he had finished that if he had not tried to take the practice under adverse conditions, he would have had a nearly smooth curve. When the work of Subject H. is reduced to a ten trial basis, the form of the curve does not differ greatly from the first five curves. It does not appear that the heaped up practice has any effect on the general form of the curve. In the curve only six trials are represented on the last division of the x-axis and yet the score for these six trials is more than twice as large as that for any other ten trials. When the reciprocal of the number of trials is used as a measure of the skill, the curves take the form of IX and X. As has been stated, Curve IX is defective because the last trial in each hundred is not complete. The subject was stopped when he reached a hundred. Curve X has a form similar to the first five curves, — that is it is concave to the y-axis. It has a slow rise at first and a rapid rise at the end. 34 WILLIAM HOWARD BATSON Curves XI, XIII, XIV, and XV are all to a certain extent concave to the y-axis. By taking groups of a thousand instead of ten trials as a basis for the divisions on the x-axis, the early part of the curves rise more rapidly but even in these cases the latter part of the curves shows a rapid rise. Curve XII corresponds to Curve II plotted by letting the ten trials represent the divisions on the x-axis. In this case where the groups of one thousand represent the divisions on the x-axis there is a slight convexity to the y-axis. It approaches more nearly a straight line. Curves Ila and IVa rise rapidly at first and then approach parallelism with the x-axis. Curves obtained by taking the highest score in each one thousand group as a basis for divisions on the y-axis, of which Curves Illb and IVb are examples, ap- proach a straight line in general form. In none of these curves is there evidence of an arrest period of any considerable length between two periods of rapid rise. Discussion of Results It was the intention in this part of the work, to change the conditions so as to give a chance to analyse the factors involved and thus to get a better idea of how they influence the acquisition of skill. Experiment' — I : The purpose here was to get a curve for comparison where no directions were given and the subject was free to learn as he pleased. The first difficulty he noticed was his inability to throw the balls properly. This was such a bother that it was necessary to move to a larger room on the third day. This did not seem to aid any as he made a lower score on that day than on the day previous. The subject had great confidence throughout the work and was always sure that he was going to do better than on the preceding day. He noticed after the third trial that he fixed his attention on the ball at its highest po'int. On the twentieth day, when he reached a score more than twice as large as he had ever made before, he could give no other reason for the improvement than that they just seemed to fall ACQUISITION OF SKILL 35 right. A cold room had a very marked effect upon the score as is shown by the resiiU on the twenty-fourth and fortieth practice days. Experiment — II : After two trials the subject had control of the manner of throwing to a large extent. In the fourth trial, all the attention apparently was centered on the throwing and the trouble came in the balls rolling out of the hand. In the fifth trial there was a greater feeling of confidence. The sub- ject knew that he had reached a higher level of ability. At the ninth trial there was trouble because of the balls hitting together in the air. This difficulty lasted until the end of the practice. If the thought of their hitting together came up, they were almost sure to do so. When the left hand was used, it was just the same as be- ginning all over again. The ability to grasp the ball came slower. It was noted at the eighteenth trial that this power had increased. It was also noted at that time that there was increase in ability to place the hand so as to receive the ball. Experiment — III : The subject's first object was to pitch in a curve from right to left. His eyes seemed to follow the ball in its course and then with a quick jerk he threw the second ball and attempted to catch the first. The tendency was for him to follow the ball around in a circle from right to left. He was encouraged to follow the plan of pitching rather than to aim to make a high score. In the third trial, he thought there was no hope; his trouble was all in catching. He was still urged to give all his attention to throwing and get that under control first. In time, he did succeed in mastering the plan of throwing but an improvement in all the other elements of the work went along with this improvement. The subject was very sensitive to a cold room and the low score on Dec. 14 was the result of this. The missing of a few practices did not seem to effect the score either way to any great extent. After the Holidays, when the practice had been discontinued for seventeen days, his score fell slightly below what it had been the last two days before the vacation but these had been unusually high. Experiment — IV : The purpose in using hard rubber was to 36 WILLIAM HOWARD BATSON compel the subject to throw in a certain way. At first there was a tendency to throw the pieces so far away that he could not reach them. At the ninth trial on the first day, he com- menced to throw in a circle from right to left. On the sixth day, he changed his plan of throwing because the pieces had whirled end over end. He now tried to pitch them straight up and let them fall flat on his hand. On Nov. 30, the low score was made because the subject was trying out a new plan of throwing. He now let the pieces roll off the ends of his fingers. He continued to use this method but in most cases when the pieces did whirl end over end he was able to get them started properly and would thus recover. In the practice of all the subjects, a large part of the improve- ment seemed to be in their ability to get out of tight places. Ordinarily the throwing would go smoothly for some time until a bad pitch or catch would put one in a difficult position to make the next move. Early in the practice, this would mean a miss, but later the subject would be able to recover and get them going correctly. Subject F. thought his ability to get out of bad places was due to his giving attention to the next throw rather than centering all attention on the catch about to be made. In the second part of this experiment, the object was to see the effect of giving a longer practice period. These results are shown in Curves VI and IX. Experiment — V : This experiment had the same object as the second part of the experiment above. The subject was not at all given to any form of athletic work and took but little interest in physical exercise. He tried to study every phase of the work. At first he had very little confidence in his ability to do the catch- ing. In the second practice, he noticed that he recovered a num- ber of times from bad places. In the fourth practice, he found that the chief trouble was fatigue. He now felt confident that he could make a good score. Taken as a whole, the subjects were affected more by sub- jective and objective factors in the beginning of the work than after they had reached a high degree of efficiency. The sub- ject's feeling that he was going to make a high score was no ACQUISITION OF SKILL 37 guarantee that he would do so. At times both high and low scores were made without the subject being able to account for them. An extra effort had no noticeable effect in increasing the score. In the early practice, there was a great amount of useless effort. The other hand would work in harmony with the one that was being used, the feet would be shifted with every catch, and the whole body seemed to be affected. When the skill, became greater, the unnecessary movements dropped out. They would return, however, when the subject found himself in a tight place but as soon as he got control they ceased. By a study of the curves, it is seen that an attempt to give attention to different phases of the work does not modify the general form of the curves for this work. There seems to be scarcely any room for doubt that although some one factor may be picked out and emphasized, all the elements improve together. In this work having a method did not help much and nearly all the improvements were made accidentally and then adopted. The Warming up Process A good idea of this may be obtained by taking the sums of the trials separately for the different subjects. The following are the figures : 1st 2nd 3rd 4th 5th 6th 7th 8th 9th loth Subject — D. 1 134 943 1357 1261 1169 958 1157 1616 1115 1396 B. 927 1359 1244 1 189 1372 1416 1340 1245 1075 1752 B. 1210 1495 1470 1416 1611 1 241 1233 1332 1 142 1466 W. 1458 1978 2051 2097 2338 2079 1793 2184 1883 2008 F. 1409 2800 2232 2410 2466 1885 2340 1752 ' ' 1852 1951 Total 6138 8575 8354 8373 8956 7379 7881 8129 7066 8573 From the figures in the total, it is seen that the first trial falls almost a thousand below any other trial. In three cases it is below any other of the ten. In the other two cases it is below the average for the ten. Relearning None of the subjects shows any considerable loss for a long rest period. Subject D., on Aug. 15, after a rest of 265 days, was able to make a score higher than any he had ever made before excepting the first three at the end of the practice where he had made a sudden rise. Two days afterwards without any further 38 WILLIAM HOWARD BATSON practice he was able to make a score that almost equalled his last practice score. After nearly a year (s^y days) more had elapsed, he was able to make the very high score of 2550 which was al- most twice as high as he had ever made before. Here the only thing that seemed to bother him was the fatigue. Subject-B. after more than a year's rest (613 days) made nearly as good score with the right hand as he had ever made before. He was sure he could have exceeded any previous score if he had not fatigued himself so much by the first throw (score 329). After that he could not grasp the balls so well. After the beginning of each trial was made the work went along without much trouble. The next day, he surpassed any previous record with a score of 1437. Here the fatigue element, especially that of the eyes, was very annoying. The left hand, after a period of 612 days, seemed to have lost more. The process seemed very strange and the subject had no confidence that he could do it. Even in this case the score did not fall much below what it had been before the last two times in the regular practice. Two days later the subject was able to make a score of 702. This exceeded any score he had previously made except the last three of the regular practice. Subject-W., after a rest of 134 days, was able to reach a score higher than any he had before made except the last four in the regular practice. The next day he exceeded all previous records. After a period of 371 days from this practice he was able to make a score higher than any he had ever made before. The only explanation he gave for this is that he had been out playing tennis and felt especially well that morning. Subject F., after a period of 143 days seemed to have dropped to a level that preceded his last two practices. The next day he was not able to do so well. The first day he started out confident that he would exceed an average of a hundred catches per trial but after a bad run in the seventh trial he became discouraged as the figures show. Taken as a whole the results show that there remains for long periods of time after practice ceases an ability about equal to what it was at the end of the practice. In some cases it seems ACQUISITION or SKILL 39 to increase and in nearly all cases is capable of a very rapid in- crease when practice is commenced again. The Total Amount of Practice In the hrst live experiments, attention has been called to the amount of practice that passed before any great amount of ad- vance was made. This must not be taken to mean that a propor- tionate part of the practice time had passed. In the trial method, the first ten trials gave but very little practice. The amount of actual practice increased as the ability increased. This, of course, is not the case where a certain number of catches had to be made each day or practice period. The total number of catches gives an idea of the total amount of practice and may be a truer index of the subject's ability to improve. It is seen from these figures that Subject D. took the least practice (12124 catches) although he took more practice periods than did Subject B. The figures show further that Sub- ject B. improved almost as fast with the left hand as with the right hand, the catches being 136 17 and 129 19 respectively. However, the practice with the left hand is distributed over a longer period. This might mean that the left hand improves about as fast as the right hand or that the longer distribution has^ the advantage. Subject W. took 20180 catches to gain the degree of skill re- quired of the other subjects. Subject F. took 21098 catches but his work was much more difficult. THE EFFECT OF AN ISOLATED FACTOR While doing the work just discussed, it was noticed that several factors influenced the subject's progress. At least three of these could be easily picked out. It has already been noticed that the ability to throw the ball so that it could be reached was an im- portant element in the learning. This judgment of direction or the ability to throw properly is one of these elements. Again, the force with which the balls were thrown was im- portant. If the ball was not thrown high enough, it would return before the other could be thrown. Not all the subjects 40 WILLIAM HOWARD BATSOX used the same force, or in other words pitched the balls the same height, but they soon fell into some habit of pitching and when the work was going smoothly, about an equal amount of force was put into each pitch. The element of force was the second factor to be noticed. The third element is that of time. It has been stated that the eye saw the ball only at the highest point of its upward motion and a short way down on its course. The absolute eye movement was not measured but by watching the subject's eyes it was seen that they scarcely moved in a vertical line. It was necessary for the subject to learn to catch the ball by timing it from the time it left the field of vision until it reached the hand. That thi - was the case is shown by the subject's inability to grasp the ball at first when it hit his hand or by his closing the hand too soon and permitting the ball to hit the ends of his fingers. That this grasping is not simply a reflex that takes place after the ball hits the hand is further shown by taking a practiced subject and intercepting the ball after it leaves the field of vision but before it reaches his hand. It was found that the hand closes even though the ball does not touch it. These three elements, judgment of direction, judgment of force, and judgment of time, have been selected for further study. The aim was to discover the nature of the improvement of these factors when isolated and trained in as simple a form as it was possible to devise. Experiment I. On Direction Two subjects took part in this experiment: Subject D. (De- Camp) and Subject W. (Wang) had both been subjects in the ball tossing experiments. The apparatus consisted of a smooth board, a steel ball, and a small block of wood. The board was fifteen inches wide and about fifteen feet long. Forty and one half inches from one end of the board a line was drawn across at right angles to the side. On this line, inch spaces were marked oft and numbered from zero out on one side to plus seven and on the other side to minus seven. Seven and one half feet from this line another line was drawn across the board. The board ACQUISITION OF SKILL 41 was three and one half inches higher at the first mark than at the second mark. The steel ball was one inch in diameter and weighed 66.8 g. The block of wood was two and one half inches long and seven eights of an inch square. This was set up on the zero mark as a target. The subject stood at the lower mark and rolled the ball at the block of wood. His score was kept in figures as + 1, — 3, etc. accordingly as the steel ball passed through these numbers or spaces.* Subject W. commenced this work on Apr. 14, 191 3 and con- tinued it until May 18, 191 3. The practice was daily in tii2 forenoon and consisted of fifty trials. The result of this work is shown in Table VII. TABLE VII Subject — W. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. April 13. o, 2, I, o, 2, — '7, 2, o, — i, i, o, — 7, 6, 2, i, 2, i, 3, o, — 5, i, — 2, 3, o, — 2, 3, 5, o, 0, o, o, 2, 2, — 4, o, 2, — 5, o, o, — 2, 4, o, o, 3, 3, 3, o, o, o, —I = 90. Apr. 14. o, — 4, — 4, — 2, — 3, 2, — I, — I, 0, — 2, o, — 2, — 2, — I, — I, o, 0, 2, — 2, o, I, — 2, o, — I, o, o, 2, 2, I, — 2, — 3, — 2, — I, — 2, o, o, — 2, o, o, 2, 3, — 2, — I, o, o, — 2, I, — I, o, I = 63. Apr. 15. 2, 2, I, o, — I, o, o, o, o, o, o, o, i, o, i, — i, i, — ^5, 4, 2, o, — 3, — 2, I, 2, 2, I, — I, o, o, — I, o, o, — 3, 7, o, — 3, — 4, — 4, o, o, —I, —I, —3, o, —3, —4, —3, —2, —I = 73. Apr. 16. I, o, I, 2, 2, o, 2, 2, o, o, o, — I, o, o, — i, i, o, — 3, o, o, o, — 4, o, — I, o, o, o, — I, — 2, o, — I, o, o, o, — 2, — 2, 2, — 4, I, o, I, o, 2, o, o, r, — I, — 2, I, I r= 44. Apr. 17. I, 2, o, — I, — I, 2, I, 2, I, I, o, o, I, 2, 2, o, I, o, o, o, o, o, 4, 2, 0, I, o, — 3, 2, I, I, — 3, o, — I, o, o, o, o, o, o, — I, o, I, o, I, o, I, 2, — 2, — 5 = 49. Apr. 18. — 2, 2, I, o, I, — 2, o, 3, o, I, — 4, o, I, 0, I, I, o, o, o, o, 2, 0, — I, — 2, o, o, o, o, — I, — 3, I, 2, I, o, o, —2, I, — 3, — I, — I, o, — 4, 2, o, o, o, — I, 0, 0, o = 47. Apr. 19. I, I, 2, — 5, — 3, o, — 2, — 2, o, — 3, o, 5, 2, 2, o, — 4, 2, o, 0. — ^5. o, — 3. I. I, o. 2, o, I, o, o, o, — 3, — I, o, — I, o, o, 0, I, — I, 2, o, o, 4, o, 3, o, I, o, I = 65. Apr. 20. 2, 2, 2, 0, I, o, — 2, — 4, I, o, 2, o, — I, — 2, 2, o, o, 2, o, 0, o, — 2, — 2, 2, 2, 2, 2, o, o, — I, I, o, I, o, o, o, — 2, — 3, o, — 2, o, o, — I, — 2, I, 2, o, 0, o, o, = 51. Apr. 21. o, o, — I, o, I, o, I, o, o, o, i, — i, o, o, o, — 3, 2, 2, i, o, 3, 2, ■ — 2, 2, o, 2, I, 0, o, 2, o, — 2, o, — I, I, 0, o, — 3, 2, o, I, 2, — I, — I, — 2, o, — I, 3, — I, o — 48. Apr. 22. 2, — 2, 0, I, o, — I, o, o, o, 2, o, o, 0, o, — 2, — i, — 3, i, o, o, o, 1, o, 2, I, o, o, o, o, — 3, I, I, — 2, o, 0, — I, — 3, — 4, o, — I, o, I, I, I, — I, o, — 2, — I, 0, 4 — 46. * In order for these figures to correspond to the value they represent, the divisions should have been weighed according to the divisions on the base line of the curve of probability. But as there were relatively few high scores the error in the final result was negligible. 42 WILLIAM HOWARD BATSON TABLE VII (Continued) Subject— W. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. Apr. 23. o, — I, — 1,0, — 2, o, 2, I, — 4, 2, 2, — 2, o, I, o, o, o, — I, o, o, 2, I, 2, 2, — I, 2, o, — 4, o, 2, I, I, o, o, — I, — 2, 2, I, I, o, I, o, —2, o, o, I, —I, —3, I, —3 = 56. Apr. 24. — I, o, — 2, — I, o, — 3, I, o, — I, I, — I, — 3, o, o, o, o, i, o, o, 1, 0, 2, I, o, o, 2, o, o, 0, o, o, o, 2, — 5, o, I, 2, o, o, I, 2, 2, 2, o, I, o, o, I, 2, I = 43. Apr. 25. o, I, 2, o, I, o, o, I, o, I, o, — I, o, 4, i, o, i, i, — 4, i, o, — i, o, — 2, o, o, — 7, 3, I, o, — 2, — 2, o, — 2, o, — 6, 0,0, I, I, o, o, o, o, — I, o, o, —3, o, = 51. Apr. 26. 3, — 1, — 6, 2, — 3, — 2, o, — I, 1, o. — 2, I, — 2, 0, I, o, I, o, I, o, o, 3, —4, o, o, —3, —3, I, o, o, —2, —4, I, o, —3, —I, o, 2, I, o, o, I, o, o, o, — 4, I, o, 2, o, 2 = 65. Apr. 27. I, o, I, I, 2, 2, o, o, o, 2, I, I, 2, 2, 3, 1, o, — I, — 4, o, — 4, I, 2, o, o, o, I, 4, o, — 2, — 4, o, o, — 3, I, I, O, I, I, I, O, O, I, O, O, I, — 3, T, — I, O = 57. Apr. 28. o, — 3, o, 0,0, I, o, — ^4, — I, o, o, — 2, 0,0, T, o, — 2, o, o, o, o, — I, o, 5, o. — 3. o, — ^4, o, I, — 2, I, o, — I, I, o, — 4, — 2, I, o, O, I, O, — I, O, — I, I, O, O, = 44. Apr. 29. I, I, 2, o, o, o, I, 0. o. 1, o, o, o, — 3, o, o, o, o, o, — 2, O, 2, O, — 2, — 2, O, O, O, O, I, O, I, O, O, O, I, O, — I, — 2, 2, o, o, I, o, — 2, o, — I, o, — 3, o = 32. Apr. 30. o, I, o, o, o, — 2, — 2, — 3, o. i, o, 2, — 2, i, o, o, o, — i, i, — 4, O, — 2, — 2, O, I, O, O, 2, O, — 2, — 4, I, — I, — I, O, O, — I, O, I, — 2, 0, O, O, — 2, I, — 2, O, — I, O, O =46. IMay I. —I, o, o, i, o, i, — 3, i, 4, o, i, o, — 2, — 2, o, — 2, i, o, 2, o, 1, — I, — 4, o, — I, —3, o, o, — 5, o, o, o, o, o, o, —I, — I, —3, I, — 2, o, o, — I, — 4, — I, —2, — 3, o, I, o = 56. May 2. I, —4, o, o, i, — 4, o, —3, — 4, i, — i, o, — i, o, o, — i, o, 0, — 4, O, O, O, O, 0, — 5, — 2, — I, O, O, O, O, I, I, — 3, O, O, — I, O, 0, o, o, o, — I, o, 0, o, o, I, o, — I = 42. May 3. 3, o, 1, o, o, i, — 2, i, o, i, o, o, o, i, i, 0. o, i, — i, o, O, O, I, 0, O, — ,1 O, O, O, I, — 3, I, O, I, 2, O, I, O, — I, I, O, O, — I, 0, O, — 2, O, O, — I, O = 30. May 4. o, o, o, o, o, o, o, o, — i, o, o, — i, o, o, o, o, 0, 0, o, 0, — 3, — I, — 3, O, O, O, — I, O, O, O, O, O, O, O, O, O, O, I, o, o, o, o, — I, o, o, o, o, o, — I, o = 13. May 5. o, — i, o, — i, o, i, o, o, — i, o, o, o, o, o. 0, o. i, o, o, o, — 3, o, — 4, O, O, — 2, O, O, O, O, O, O, O, O. O, O, O, O, 0, O, O, — 3, o, o, O, I, O, 0,0,0== 18. May 6. i, o, o, 0. o, o, t, i, — 3, — 2, o. — 2, — 4, o, o, — i, o, — i, — i, i, — 3, — I, — I, O, — I, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, 0, o, o, o, o, — 4, 0,0 = 28. May 7. 2, o, o, o, o, o, o, o. o, o, o, i, o, o, o, — 2, i, t, o, i, o, o, 1, O, O, O, O, O, — 3, O, O, O, O, O, O, O, 0, O, O, O, O, — I, o, o, O, O, O, — I, 0,0= T4. May 8. o, o, o, o, o, — 2, o, o, o, 0,0, i. 6, — 3, o, — i, o, o, — i, o, o, O, — I, o, I, 2, o, o, o, o, o, o, o, O, — I, o, o, o, o, 3, o, o, O, O. O, O, O, 0,0,0= 22. May 9. —I, —I, —3, —I. —I. —2. —I, o, o, o, o, 2, o, o, o, 3, o, o, o, 3, o, O, O, 2, O, O, 2, O, O, O, O, O, I, — 2, — I, — I, 3, O, O, — I, O, — I, — I, O, O, — I, O, O, — 2, O = 36. May 10. o, o, o, o, o, o, o, o, i, i, o, o, i, o, o, o, o, o, i, o, o, o, O, I, T, 2, T. I, O, I, O, O, O, O, O, O, O, O, O, O, — 3, — 2, O, O, O, O, O, 0,0,0= 16. ACQUISITION OF SKILL 43 TABLE VII {Continued) Subject — W. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. May II. o, o, o, o, o, i, — i, o, o, i, 2, o, — 3, — i, o, o, o, o, o, o, o, o, — 2, o, o, o, o, o, o, o, — I, o, o, — I, o, o, o, o, o, o, o, o, o, o, o, o, o, 0,0,0= 13. May 12. o, o, o, o, o, o, — i, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, O, I, o, o, o, o, o, o, O, — I, — I, o, o, o, o, o, o, o, o, o, O, O, I, o = 5. May 13. — I, o, o, o, I, o, o, o, o, o, o, o, i, o, o, o, o, o, o, i, o, o, o, o, o, o, O, I, o, o, O, O, I, O, — I, I, o, o, o, o, o, o, o, o, o, O, O, I, o, o = 9. May 14. o, o, i, o, o, — i, i, o, o, o, o, o, o, i, o, o, o, o, o, o, 0, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, — 3, o, o, o, o, o, o, O^ o, o, I, o, — I, = 9. May 15. I, I, o, — 3, — 3, o, — 2, o, o, o, o, o, o, o, o, o, o, 0, o, o, o, — i, o, o, O, — I, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, O, O, 0,0,0= 12. May 16. o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, O, O, O, O, I, o, o, o, o, o, o, o, o, o, o, o, O, o, 0, 0, o, 0, 0,0, 0=1. May 17. — I, 3, I, o, o, o, o, o, o, o, — i, o, — i, o, i, o, o, o, o, o, o, o,. o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, O, O, O, O, — I, o, o = 9. May 18. — I, o, — ii, o, — i, o, o, o, o, o, o, o, — i, — i, o, — i, o, — 2, o, o„ O, O, O, I, o, o, o, o, o, o, o, o, o, O, — 2, o, o, o, o, o, o, O, O, O, O, O, O, 0, O, O = II. May 19. o, o, o, o, o, o, o, o, o, o, o, i, o, 0, o, i, i, o, O, I, O, O, I, O, O, O, O, I, o, O, O, I, O, O, I, O, O, I, o, O, O, O, 0, 0, 0, 0, O, I, I, — I = 12. Subject D. commenced work on Apr. 17, 191 3 and continued until May 22, 191 3. His practice was in the forenoon and con- sisted of a hundred trials. The results of his work are shown in Table VIII. TABLE VIII Subject — D. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. Apr. 17. — 2, — 2, — 3, — I, — I, — 2, o, o, o, I, o, o, o, — I, o, i, 2, o, o, 0, O, O, — I, — 2, O, O, O, Q, O, O, O, — I, — I, O, o, — 3, I, 2, O, O, o, 1, O, O, O, — I, 2, 2, O, O, I, O, O, 2, O, O, I, O, O, 3, — 2, — 2, —I, I, —4, 3, I, 3, —I, —3, I, O, 2, 2, —2, O, O, O, 2, 6, —I, —2, o, —2, o, —2, 5, o, 3, 4, —3, o, I, o, 2, o, o, —I = 104. iApr. 18. — 4, 2, o, 2, 3, 2, o, — 3, — I, 5, — i, o, o, — 5, o, — 4, 3, O, — I, — s, O, — I, — 2, — 2, o, O, — I, 4, O, O, O, 2, o, — ^3, O, — 2, 2, — I, I, 2, — 2, O, — 3, — I, 2, O, 3, 3, O, I, O, — I, — I^ o, o, I, — I, o, — I, I, o, 3, — I, — 3, o, — I, o, o, — 1, — 3, — 1» — I, I, 0, 2, O, O, — I, 3, O, 2, O, 2, O, O, 2, — I, O, — I, O,. ' — 2, — I, — I, I, — I, 2, — I, O, 2, O = 125. Apr. 19. — 3, o, 4, o, I, o, o, 2, — 7, o, o, i, 2, — i, i, 2^ 0, I, —3, 3, o, 5, o, I, I, o, I, I, o, I, —3, o, o, o, o, o, 1, O, 3, I, I, — I, I, I, 3, I, I, I, 4, O, 2, o, o, O, O, 2^ 2, 3, O, 3, O, O, O, 2, I, O, I, O, I, I, 3, O, O, 3, I, O, I» 44 WILLIAM HOWARD BATSON TABiLE VIII {Continued) Subject — D. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. 2, o, o, o, 3, o, o, o, I, 4, 4, 2, o, o, 3, 4, o, i, 2, 2, o, 0, 5 = 122. Apr. 20. 3, 2, I, I, o, I, 4, I, I, 2, I, 3, I, — I, I, — I, 3, 2, 4, I, 2, —2, —3, 2, —2, o, o, o, o, 2, 3, I, 2, —2, —2, o, — r, 1, I, 2, 2, — I, — I, o, — I, — 2, o, — I, o, I, 3, 5, — 4, — I, — ^4, — 3i — 2, o, o, — I, o, I, 2, — 2, — ^3, — I, — I, I, o, — I, — I, 3, — 3, o, I, 2, — 3, — I, I, — 4, — 3, I, I, — I, — I, — I, o, — 2, o, 1, o, o, 2, o, 2, o, o, o, 2, o = 142. Apr. 21. o, I, 3, o, o, 2, 3, 2, o, I, o, o, I, o, 2, o, i, o, i, 3, o, — 2, o, I, o, — 2, 2, I, o, o, o, I, o, 5, o, o, o, — 2, — I, O, O, I, O, I, I, 0, I, — 2, I, 4, O, O, O, — I, — I, o, o, o, o, o, — I, o, 5, —6, o, 2, o, o, 0, — 3, — I, — 2, o, — I, 2, o, o, — I, o, 2, 3, 3, o, o, o, o, — I, o, o, o, I, — 3, I, — I, — I, 4, o, o, I, o = 95. Apr. 22. 2, — 2, — I, — ^3, 0, o, — 2, 3, 0, 6, o, I, o, — 2, I, o, I, 'I '\r — 'I 'O 'O 'O 'Z 'I 'O '£ 'Z T 'O 'O 'O 'Z 'l 'l 'l — 'o — 'l 6, — 2, I, 2, I, 0, — I, — 3, o, 2, 2, 2, I, 2, 2, o, 2, — 2, I, 3, I, — I, 2, 2, o, o, — I, I, — I, 2, o, 0, — I, I, — 2, o, — I, o, o, I, — I, 2, O, — 2, O, O, — I, 2, — I, — ^I, 2, I, — I, O, — 1, O, 3, O, — I, O, O, 2 = 122. Apr. 23. 2, — 2, o, — 2, o, — 2, o, o, o, — 2, 3, o, 2, o, o, 4, O, O, O, O, 3, O, 2, I, I, O, O, 2, 2, O, O, — I, O, O, — I, O, 2, O, O, 2, 2, 3, I, 2, O, O, O, 4, I, O, O, O, 3, O, 2, 2, — 3, 2, 3, — I, I, I, I, o, o, o, I, o, o, o, o, o, o, o, — ^3, 2, o, i, 3, 4, 3, 2, O, O, O, — I, — I, — 2, O, — I, O, O, O, O, O, O, O, — ^3, 2, o, o = 96. Apr. 24. — ^3, — 2, O, — I, O, O, I, 2, I, O, 2, O, 2, 2, O, O, 2, —I, —2, 3, 3, —2, o, 3, o, o, —3, o, —3, 3, o, —2, 3, 3, 2, O, O, O, I, 4, O, O, — 3, 3, I, O, O, O, 2, — 2, 3, I, 2, O, I, o, o, — 2, o, — 3, — I, I, o, — 2, o, 3, o, — 3, 4 o, o, 3, — i, — 2, — 2, — I, I, O, I, O, 2, 5, O, O, — I, 2, I, O, O, I, 4, O, O, O, — 2, O, I, O, O, — I =: 124. Apr. 25. — I, o, o, o, 3, — I, o, o, 3, i, 4, o, 3, 3, 2, o, o, o, O, I, 2, 4, 4, O, I, — I, 2, — I, 2, I, O, 2, 2, — '3, O, — I, O, 2, O, 0, 2, O, O, O, O, O, I, O, O, O, I, O, 3, 3, 4, 2, O, — I, I, 3, 1, O, O, O, O, O, I, 2, O, — I, 2, I, O, 2, — I, O, O, 2, O, I, O, I, O, O, O, I, I, 2, 2, 3, I, O, O, O, I, 2, I, O, I, I =: I02. Apr. 26. o, — 3, 2, o, o, — 2, o, o, o, 2, o, — 3, o, i, o, o, i, O, 2, O, 2, — 5, O, 0, O, O, O, — 2, — I, 2. — I, 4, — I, O, — 2, O, — 3, — 2, — I, o, O, — ^3, — I, O, O, O, O, O, O, — 3, O, O, — 2, O, O, — 3, o, 0, 2, — I, 2, — I, I, O, — I, O, O, O, O, O, — 2, O, O, I, O, — 2, 1, I, 2, O, — 2, 3, I, O, — 2, — I, O, — I, — I, O, O, 2, O, O, O, O, O, O, — I, O = 84. Apr. 27. o, o, o, 2, 2, o, 3, — I, o, o, 0, 2, 2, 7, — 2, — I, — 2, — 3, o, — 2, — 2, o, I, I, — 3, o, — 3, I, I, — I, — I, 2, o, — I, — I, 0, 2, I, — 2, o, o, o, — 2, — I, o, — I, I, o, — I, o, I, 4, 0, — 3, 1, — I, 2, I, o, 2, o, o, 2, — I, 0, — 3, o, o, I, 3, o, o, — I, o, 2, — I, — I, — 'I, O, O, I, 2, O, — 3, — I, I, — 1, O, I, 0, 0, 2, I, o, 2, o, 3, o, o, o = 107. Apr. 28. 2, 3, 3, o, o, — I, I, o, 3. 2, o, o, o, — I, o, — 2, o, O, 2, O, O, — 2, 2, 2, O, 2, O, — I, O, 2, O, O, O, 2, O, — 2, 2, O, — 2, 3, O, O, O, — I, I, O, I, —5, 2, O, O, 3, O, — I, 2, I, — I, — 4, O, O, O, 2, O, O, O, O, O, — I, — I, O, O, O, — I, O, 2, O, O, O, I, O, O, O, 2, O, 3, O, 2. O, O, 2, — 2, — I, 2, — I, O, O, 2 = 89. lApr. 29. — I, — I, o, — I, o, o, I, 2, 4, I, o, o, o, o, 2, o, — ^i, o, o, — 4, o, o, I, I, o, o, I, — 2, 2, — I, o, o, I, — 2. ACQUISITION OF SKILL 45 TABLE VIII {Continued) Subject — D. Experiment — To obtain a curve of learning for direction. Work consisted in lolling a steel ball at a mark. 2, o, 2, o, 3, o, o, 2, I, 3, — I, o, o, — I, — 3, o, I, o, 2, I, — 1» 3, 3> o, 2, o, 5, 3, I, o, — I, — I, o, — I, — I, o, o, 2, o, I, I, I, I, — I, o, o, — I, o, o, o, 4, I, 4, o, o, o, — 2, 0, — 3, — 3, o, I, 2, 2, — I, o = 104. Apr. 30. 3, 2, o, — 3, — ^r, 2, I, o, o, i, — 2, 5, o, 2, 3, o, o, 1, — 4, — I, — I, — 2, o, — 2, — I, 2, o, I, 2, — I, — I, o, I, o, I, — I, 2, 2, o, o, o, I, o, o, 2, o, I, 2, o, — I, — I, 5, 2, o, — 3, 2, O, O, — I, O, O, O, — I, O, O, — 2, O, 0, — I, O, O, — 2, I, o, 2, 5, 2, I, O, O, O, O, —2, O, O, I, O, O, O, O, — I, — I, 0, O, I, I, 2, 2, 3, O = 104. May I. 2, o, 2, o, o, i, 2, i, — 2, — 2, i, o, i, 2, o, 3, o, o, 3, —3, —3, 2, —2, —3, o, I, —I, 2, —3, —I, I, o, o, I, —3, — i, 1, o, I, I, o, 2, I, 4, o, o, I, — 3, o, I, 4, o, o, o, o, — I, 0, o, o, o, o, o, I, 4, — I, — ^i, I, o, I, o, 7, o, o, — 3, 1, O, 2, — I, 5, O, I, O, I, O, I, O, 2, O, O, I, O, O, I, o, o, o, o, I, o = 105. 'May 2. o, o, — I, o, I, 5, I, 2, o, o, i, o, o, 2, o, i, o, — i, O, O, — I, O, O, O, O, O, I, I, O, O, O, 2, 2, I, O, O, — I, 2, O, O, I, 2, — I, O, 3, I, O, 2, — r, — I, — 2, O, O, I, — I, o, ■ — I, O, O, O, 2, I, — 2, 2, 2, O, O, I, O, 2, O, O, O, O, O, O, o, o, o, o, 0, 3, — I, — I, o, 3, o, — I, o, 2, o, 3, — 3, 4, 3, — I, O, I, O, I = 82. May 31. — 2, 4, o, 5, 3, I, I, — I, — 2, o, o, 2, i, o, — 2, o, o, ' — ^3, I, o, o, I, — I, — I, o, o, o, o, I, — I, o, 2, 3, I, 3, 2, o, 3, — I, O, O, O, 2, O, O, 2, O, O, — I, 3, 4, 3, O, O, 2, 3, I, O, I, O, O, O, 2, O, O, O, — 3, I, O, — 3, O, I, O, O, I, o, o, o, 3, I, — 3, I, o, I, 2, I, o, o, o, — 2, 2, 2, o, 2, 4, o, — 2 = 107. May 4. o, r, 2, 2, 2, 2, — i, — i, o, o, o, — i, 3, o, 3, 3, o, — i, o, — 2, o, o, 2, o, — 4, o, 2, o, 5, — 4, o, o, 2, o, o, — I, 2, I, o, — 2, o, — 2, O, — I, O, O, — 2, O, — I, O, — I, — I, 0, — I, O, o, o. o, 7, — 2, — 5, — 3, — 5, o, o, o, o, o, — i, o, o, — i, 2, — 2, o, — 2, 4, I, — I, 2, I, 2, I, 3, I, 2, O, O, O, O, — I, I, O, I, 2, 2, — 2, I, O = Il6. May s. o, i, 3, 5, o, o, o, o, 2, o, o, o, 2, o, i, 2, i, i, — i, 3, 2, o, I, o, o, I, o, o, I, o, r, — I, o, o, 2, o, o, 3, o, O, I, O, 2, O, O, O, O, — 2, O, O, O, O, — I, O, 5, 2, — I, 3, 2, O, O, O, O, O, — 2, I, 2, 2, O, O, O, I, 3, O, O, I, — I, I, I, O, 0, I, 0, O, I, O, I, O, — I, O, 4, — I, 2, O, I, O, O, — I, 2 = B>2. May 6. 3, 2, — 2, o, o, o, — 2, o, o, o, — i, 3, o, o, — 2, i, 3, o, o, I, o, I, o, o, 2, I, — 2, 2, I, o, I, 4, I, o, I, o, 1, 0, 0, 2, O, O, I, 2, O, 2, 2, I, 3, — I, — I, O, I, O, O, I. — 2, — 2, 3, I, I, 3, — I, 4, 2, O, 2, I, — 2, O, O, I, O, — I, 2, — 2, I, 2, O, O, — I, O, O, O, O, O, O, 2, I, — I, 2, — 2, O, 1, I, 3, o, o, I, —3 = los. 'May 7. i, i, i, o, o, — i, — 2, o, o, o, — 3, o, i, o, o, o, o, — I, o, — I, — I, 2, I, — I, — 3, o, — 2, o, o, o, o, 2, o, o, o, 2, O, O, — I, O, — 2, O, 2. O, O, — 2, O, O, O, — I, O, — I, — I, O, — I, O, O, — 2, O, 2, O, O, O, O, O, — I, —I, — I, 3, O, — I, O, O, O, O, 1, I, O, 2, — 2, O, I, O, O, I, O, — 2, O, — 2, — I, O, O, O, — I, O, — I, = 64. May 8. 2, o, i, 0, o, o, — i, — 2, 3, 0, 0, o, — i, 2, — i, 2, o, 3, O, I, 2, 3, — 2, I, I, O, O, — I, o, — 4, O, O, O, — I, — I, — I, — I, O, O, — I, — I, O, O, — 2, O, — I, O, — I, 2, 2, O, 2, — 2, 0, — I, 2, O, O, I, O, I, r, O, 2, O, O, I, O, — I, o, o, 46 WILLIAM HOWARD BATSON TABLE VIII {Continued) Subject — D. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. 3, o, — 2, — 2, o, o, o, o, — 2, — I, 4, o, o, 3, o, o, o, o, o, o, — I, —I, o, o, o, 3, —2, I, o = 86. iMay 9. 4, 2, — I, o, — I, — I, — I, — I, o, — 2, I, — I, o, — I, o, 2, 2, 0, 2, o, o, o, o, o, 2, o, o, 3, I, o, o, I, o, I, o, I, — 2, — I, — I, — I, 2, I, I, o, 0, I, — I, o, o, — I, 2, o, 2, 2, O, 4, O, O, O, — I, O, O, O, O, 2, O, 2, O, I, I, O, 4, O, O, 0, 2, O, 4, O, I, O, — I, O, O, O, O, I, — I, O, O, O, 3, I, o, 1, O, I, I, o, o = 81. May 10. 3, o, — i, 2, — i, o, o, 2, 2, 2, — i, 2, o, — 3, 2, o, O, — I, O, O, — I, O, O, O, O, O, 3, O, — I, O, O, — I, — 2, 2, O, O, 2, — I, I, — I, O, O, O, O, — I, 0, o, o, o, o, o, o, o, o, o, — I, O, O, O, — I, O, O, O, O, I, — I, I, I, O, O, I, O, 3, O, 2, O, O, O, O, I, O, O, O, O, O, O, O, 2, 3, 0, O, — 4, — I, 2, — I, o, o, o, — I, - 1 =66. May II. I, o, o, — i, o, o, o, 2, i, 2, i, — i, o, 0, 3, o, o, o, O, O, I, O, O, I, O, O, O, O, O, 2, 2, O, — I, I, O, — I, O, O, — 2, 2, 2, — I, O, O, — 2, O, O, — 2, O, O, O, I, O, O, 2, — I, O, 0, — I, — I, O, O, 2, O, O, O, I, I, I, O, O, O, I, o, o, o, o, 1, 2, 2, I, — 3, — 2, — I, O, — 2, 2, 6, O, O, I, 2, O, O, O, O, O, O, I = 64. May 12. o, I, o, i, o, o, 3, o, i, 2, o, o, 2, o, o, o, o, — i, — 2, O, O, I, O, O, I, O, O, O, O, O, O, I, O, O, O, O, O, — I, O, 2, O, — I, I, O, O, O, I, O, 0, O, O, O, — 2, 0, O, O, 0, O, O, O, O, I, — 2, — I, O, O, O, O, O, O, O, O, 2, O, O, O, O, O, 2, O, O, O, O, O, I, O, O, O, O, O, 0, O, — 2, O, O, O, O, O, 0, O i= 35. May 13. o, o, — i, o, o, o, 2, o, i, 3, 2, i, o, 2, — i, 0, — i, 0, I, O, O, O, O, O, O, O, O, O, O, O, O, O, O, O, 0, I, I, 1, I, o, O, O, O, O, O, O, O, O, I, I, I, I, O, O, O, I, 2, O, O, O, O, O, O, O, O, I, O, — I, — I, O, O, O, I, O, — I, I, O, O, O, 0, — I, I, O, O, — I, O, — I, O, O, I, 2, — I, — I, — I, O, — I, — I, I, I, o = 46. iMay 14. I, o, o, I, o, — i, — i, o, o, — 3, o, i, i, o, o, i, — i, O, O, O, O, 2, O, O, — I, I, O, 2, O, O, I, 2, O, O, I, I, O O, O, O, O, O, — 1, — I, O, 2, O, — I, I, O, I, I, — I, 3, O, I, I, 2, O, O, I, — I, O, O, O, O, O, O, O, I, I, 2, O, O, O, O, I, — I, — 3, o, o, o, I, o, o, I, o, o, o, o, 0, 5, o, o, o, i, 0, I, O, = 58. May 15. o, o, o, o, o, o, o, i, o, 2, i, 2, — i, — i, o, o, o, 1, I, I, I, I, — I, 2, I, — I, O, O, O, I, O, O, O, O, 0, O, I, I, O O, O, I, O, O, O, O, I, 2, O, O, O, O, 0, I, 3, I, 3, o O, I, I, 4, I, — I, — 2, O, O, — I, O, 0, O, O, I, 0, 2, 2, 2, 0, O, O, I, O, 2, O, O, — 2, — 2, O, I, O, 2, I, O, O, O, 0, O, — 2, —I, o = 65. May 16. I, o, o, o, i, — 2, 3, 4, 2, — 2, o, i, — 2, o, i, 2, 1, O, 4, I, O, O, O, 2, I, 2, O, O, — I, — 2, I, 3, O, 2, O, O, I, O, 2, O, O, I, O, O, O, O, O, O, I, O, I, I, — I, I, O, O, I, 1, 3, I, — I, — I, 2, — I, 0, O, I, I, I, I, I, O, I, — I, I, O, — I, 2, O, O, 2, O, 2, O, — I, O, O, O, O, — I, O, I, 3, 0, — I, O, 2, I, 3 = 89. May 17. o, o, i, 2, o, o, o, o, — i, o, i, o, 2, 3, o, i, o, o, o, O, O, I, O, O, O, O, O, O, I, O, O, O, I, I, I, O, O, O, O, I, o. O, I, O, O, O, O, O, O, O, O, I, — I, O, I, — 'I, O, I, 0, O, — I, O, O, O, 2, O, O, O, — I, — I, — I, O, O, O, O, I, o, o, o, o, o, o, o, I, o, o, o, o, o, I, I, — I, o, o, o, o, o, I, o = 35. May 18. 2, 2, o, o, o, — i, — i, o, o, o, o, o, o, o. o, i, i. I, 0, 0, 3, I, ), 0, o, o, o. 0, o, o, 0, , 0, 0, o. I, O, 0, 0, o, , o, o, I, I, O, 0, o, 0, >. — I. ■ — 2, o, 0, o, o. 0, ACQUISITION OF SKILL 47 TABLE VIII Subject — D. Experiment — To obtain a curve of learning for direction. Work consisted in rolling a steel ball at a mark. O, 3, O, O, 2, O, O, O, 2, I, O, O, O, — I, O, O, O, — I, O, O, 2, o, o, o, o, 3, o, o, o, o, o, o, I, o, I, I, — I, I, — I, o, O, O, O, I, O, O, O, O, O, I, O, — I, I, O, O, I, O, O, I, o, o, O, O, — I, O ,1, 2, 2, O, O, O, O, O, O, O, O, O, O, O, O, O, O, O = 4I. May 19. I, I, I, o, — I, o, o, — i, o, o, o, — 2, — i, 0, o, o, o, i, O, O, — I, I, 2, O, O, I, O, O, O, 0, O, — I, — I, O, O, O, O, 0, O, I, O, O, I, I, O, 2, 2, O, O, I, O, 0, I, O, O, I, O, O, O, 0, O, I, O, O, O, O, O, I, O, 2, O, O, I, O, O, I, O, — 2, I, 0, 1, o, 0, o, o, o, o, o, I, 0, o, o, — I, o, o, o, o, o, o = 38. May 21. — 2, — I, o, o, o, — i, — 2, o, o, i, o, — I, O, O, O, O, O, O, — I, I, 2, 2, O, O, O, O, O, — I, O, O, O, — I, I, O, O, O, — I, o, o, o, o, o, 1, O, O, O, O, O, O, O, O, O, O, I, O, O, I, o, o, O, O, O, I, I, — I, o, o, o, o, o, o, o, o, o, —I, I, o, o = 32. May 22. o, 2, o, 3, o, o, 0, i, o, i, o, — i, o, 2, O, I, O, O, I, O, O, O, O, O, O, O, I, o, o, 0, I, I, O, — I, O, I, — 2, O, 0, O, I, O, 4, o, o, o, 1, O, O, —2, O, O, O, I, O, O, O, O, O, I, o, O, O, 0, O, O, 0, 0, O, — I, — I, O, I, O, O, O, — I, O, 0, — I, o, o = 40 Curves XVI and XVII (Plate IV) are obtained by letting the total score for each day's practice represent divisions on the y-axis and practice periods represent divisions on the x-axis. Curves were also plotted representing the improvement meas- ured in terms of actual hits. The total number of hits made in each day's practice was taken as a basis for divisions on the y- axis and the number of practice periods as the basis for divisions on the X-axis. Curve XVIa (Plate VIII) shows the typical form. The two subjects were very different in their manner of throw- ing. Subject W., after a few days practice, threw the steel ball so that it curved from right to left. He used only enough force to enable the ball to reach the mark. He aimed at a mark on the board about eight inches from where he released the ball. After he had this plan perfected, it made no difference if a piece of card-board was set up between him and the block of wood in such a way as to shut off the view entirely. Subject D. threw directly at the mark. The ball rolled straight for the most part and with considerable force. Notwithstanding the difference in their method of practice, the curves show con- siderable similarity in the form of the improvement. Although 48 WILLIAM HOWARD BATSON the daily fluctuations are quite large, there is an almost constant downward course to each curve. The large fluctuations from the thirteenth to the twenty-first practice of Curve XVI repre- sent the period when the subject was attempting to curve the ball and to locate a place on the board near his hand through which he could throw the ball. It will be noted that at only one place are there as many as four consecutive points on the same side of the line passed from point seventy on the y-axis to point thirty-six on the x-axis of Curve XVI. With the exception of the period when the subject was attempting to find a new method, the curve follows the general direction of this line. The average deviation of the points from this line is eight. The upper heavy line shows the position where the plus and minus deviations are balanced. The sum of the deviations above this line is equal to the sum of the deviations below the line. The two dotted lines show the average deviation from the line. The course of Curve XVII is fairly well shown by a line drawn from 135 on the y-axis to a point 37 units above 35 on the X-axis. There is a period from the thirteenth to the twen- tieth practice when the score was rather high which was the result of the subject trying to spin the ball on its axis. The only explanation for the rise at the thirtieth and thirty-first practices was the bad weather which may have affected the sub- ject. Only in one case here do as many as four consecutive points remain on the same side of the line and three of these are within the average deviation limit, which is twelve in this case and is represented by the two dotted lines. An examination of Tables VII and VIII shows that the sul> jects were able to do almost equally well in each half of a prac- tice period. The scores that Subject W. made in the first half of his daily practices were as follows : 50, ii, 31, 20, 24, 24, 44, 31, 22, 19, 27, 19. 22. 2>7, 31, 19, 17, 25, 28, 30, 13, 9, 12, 24, 12, 16, 20, 6, II, I, 4, 4, 12, 8, 9, 5. Total 619. The scores for the last half of each practice were : 40, 30, 42, 24, 25, 23, 21, 21, 26, 27, 29, 24, 29. 28, 26, 25, 15, 21, 28, 12, 17, 4, 6, 4, 2, 6, 16, 10, 2, 4, 5, 5, o, I, I, 2, 7. Total 608. ACQUISITION OF SKILL 49 The scores for the first half of each practice for Subject D. were: 33, 79, 63, 71, 43, 71, 49, 67, 53, 49, 58, 47, 58, 64, 38, 56, 55, 39, 5i, 34, 47, 43, 38, 32, 22, 21, 28, 25, 46, 18, 24, 24, 23, 22, Total 1496. The scores for the last half of each practice were : 71, 46, 59, 71, 47, 51, 47, 57, 49, 35, 49, 2>7, 57, 46, 41, 44, 5i, 6r, 44, 54, 30, 39, 38, 2%, 32, 13, 25, 30, 40, 43, 17, 17, 14, 9, 18. Total 1410. On July 17, 191 4 Subject W. was tested again in this work. The only difference was that the board this time was covered with heavy paper. He had had no practice since May 18, 191 3. This practice consisted of a hundred trials. The result was : First fifty. — i, 4, 5, 4, 7, o, o, o, 3, o, i, o, 6, 3, o, o, o, o, o, o, o, o, o, 2, o, o, o, o, o, o, 3, o, I, o, — 2, o, — 2, I, 2, o, o, o, 4, o, o, I, o, o, o, o. Total 52. Second fifty, o, o, o, o, o, 2, o, o, o, o, 5, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, o, 5, o, o, o, o, o, o, 4, o, o, o, o, o, o, o, o, o, o, o, o, o, o. Total 16. The subject explained his difficulty in the beginning by the fact that the ball would not curve on the paper as it had done on the board. The second fifty shows how fast he was able to adapt his method to the covered board and then his work was nearly as good as it had been at the end of the practice. In order to give an idea of the amount of skill that had Ix^en acquired by the above subject a comparison practice was given Subject Do. (Dockeray) on the same day as the above practice and the following score was made : 2, o, 6, o, 5, —2, —4, I, o, 5, 4, 2, o, 4, 3, I, I, 3, 5, 6, o, —5, —6, 7, I, —5, 6, 5, —I, I, —7, —2, o, 5, 6, 5, —I, I, —7, —2, o, 5, 6, 5, — r, I, —I, —2, I, 5- Total 154. This is a considerably higher score than either of the other two subjects made in their first practice but it indicates that the putting on of the paper had not made the work easier. Experiment II. On Force Four subjects took part in this experiment : Subject W. (Wang), Subject D. (Decamp), Subject C. (Cole), and Sub- ject Do. (Dockeray). The last two were also graduate students so WILLIAM HOWARD BATS ON in psychology. The apparatus for Subject W. was the same as in the preceding experiment, except the block of wood was not used and instead of the steel ball, two rubber balls that h'iN bo-y o n c o .5 »-• ■*-* >♦- " rt • o o *-■ ^ MH U. (U W (U >-'Xi U (U C 361 ■* rt O " 4-> 4) .S c£ n J3 3 0*0 "> o S ^ (N Hts2 UJ3 , • Ji^S -^ < ? « >> ,_, ^ 3 >5 u t/3W fc n i-i M N i-i ro PI 11 >0 >0 lO t^ •^ J-^ 1-1 ro w i-ii-i^i-iroiOroi-HM ""jvO ON fO N "-i ii >-< M >-i i-i ^^ i-H pt -t fo 1^ c^i lo 1^ oi vo fo <-< n w t«^ -f rovo r'j •I M ro <^ fO <^ i-i "l-^O hh >-< i-i ►H 01 1-1 ro Iv ■^■ f^l l^ fO ►-< fO 'O O 1-1 N 11 M 0\'0-+ii lO-^-tr^M <^-fii n ro '^) n "■) fO fO r^ C\ >»■' f^ *t fO (^l '-' 1-" M ►-. ro n M 1^1 "t O 0| M >0 -r r^O m « n n (V) ^ M ro -fO n ^1 -t ■^- if/^O "^1 "t '^J Cv) 01 (r> r\| M lO r^l Tf ro H- 00 -t " VO O) 01 n M 01 >0 i^vo r*5iir^iOM-t<^fO01iii •- i_i M n n \o -^O ■^- 01 Tt 1^ in >0 n n n Cv|(\|i_ii>iiirorO'+ rovD Tr-t-^-r^O)ro MO! ►-, (\| ►-, Ot 1 01 ^t l^ ro 1 01 01 -^-VO <^ lO 01 n COM noioiooioioo rrvo tx in 01 0< 01 OlOlOlcoN ii01-^01|^r^T}-Ttoi"i-OI01 i_iM 11 ii010irOc*5r0'^0|roinPO01'^-t0liioi01 mOI nii'iii'l'rOfOfO l^io infOioOlMMH m^h01>i>i 01 ii<^rooi-roioooi'*50lOlT)-ir5ro i >i>i 01 OlOl^Of^iOf^ iO01>ir«-;^-t nn-i -< ro oi M o O\00 1n\0 iOTfr •JH e o 3 rrt XI ^ 1-. CU x: •^ H in Q ■w u E x; H X WW f^-»tMMr)N0)TJ-J^TthHP)h-iM»i-)i-il-ifOMi-i fO C^) Mir) TfcOTTTj-iNrt-rflNPOi-iC^i-H m ^H M M ci i-H M (V) M --t ro 'O -^-OO ro 01 10 M "^ 1-1 p-i MM i-i 10 ,ri -^-vo 00 -O M M ■+ CM <^ i-H Tf in l^ >r) 10 Tt fO fO 0) •-< ?^ Mm 1-1 1-1 10 Tt m M \0 "tVO 01 ro "'t >-i i-i c^i M i-i ro 't M hi 01 ro>0 00 ■^- ro>0 "^ m 1-1 ►-, 0< M Tj- Ol TtOO IC; Tf 01 ro 01 ^ H-l 01 O) >-( w I-I r<^ 1-1 10 0) r<1VO 01 CO Tj- irj IT; ro 1-1 1-1 i-i i-> 1-1 0| 01 t 01 fH M c>t oi 00 -^^ 0| -f M 01 fO irj 1-1 i-i M ro 01 M -rf M 01 10 fO ro -t- 01 -t 01 -t 01 w 1-11-1 01 0< 11 1-1 01 -^ 01 w rf 10 10 PI 't 0< 01 01 0) 11 1-1 01 1-1 1-1 "I- ro^O oiPi-t-troNro01i-i01 rOM ■^• 1-1 00 0< 1-1 01 t^ 10 01 "t 01 -i I-. 01 1-1 w o» ^ 1-1 I-I PO 1-1 00 ^o "+ 01 CO 01 in 1^ 1-1 fo 01 1-1 po f^ M 1-1 01 0) 01 M 01 01 1-1 -1- -t ro PI -t VD 1-1 PO PO 1-1 01 01 1-1 in hi M w PI M M 01 01 01 h-i ro ro 01 01 <*5 rn co Oi "^f >-i 01 m pi 1^ POi-ii-ii-'PD M|^r^i-iroPii-HOiP| m '^ 00 o» 11 o Onco in,vo 10 -f po pi 1-1 o ■-< 01 po ■^- in>0 1 ^00 0\ O "-i pi ro '^ »o |_|IHI-ll-ll-l|-|^^ |_,l--,l-.l-|l-l|-l I I I I I I I I M I I I I 1 ^ ^ ^ ^ §8 O T3 lA c OJ E "^ ^ t/1 "o o ^ o CJ ^ x: tfl H G s T3 ^ u nj S n • -! o X 1-. -c n u o J pq < H H:S . «n V, !> O "\ <+; w 1-. >-, a. o w <« •- bfl4:i _G JJ "c7! c > e Ui 3 3 O "o o >, O ■4.1 "-(-I -i 1-1 HH ^ (N (^ .000 rovo 01 O) -rt 04 CO i-i MM ot 01 lo tn -^^ ^ CO "^ 01 1-1 M PO M ro 01 01 IT) ON^O l^ -i- 01 01 1-1 M 01 ooro-^Tj-roi-i ro>o>ni-i m ro 01 Tj-cooi\o Tj-ovo ^01 0001 01 M 1-1 1-1 CO ro M -^ 0\\0 CO CO ^ O) 01 01 01 1-1 01 1-1 M 01 CO C^ >o lO in i^ CO f^i 01 1-1 MMCOMI-l^wOll^^COl^ TfOlOl CO 01 l-IM M 01 CO 01 CO 01 01 -|- CO -^-VO 01011-iOIOlMCOOl Mi-i 1-1 1-1 01 01 CO 01 CO 01 CO'OmrocOTfCOi-iOlOlMM oico'-ioii-ioii-ico-^mt^cooii-'Tfi-iMi-i oiM 04h-( (^^MMl-l^-lO^C^llOT^01l-l■i-01lOCOCOl-lCO M 01 01 ooOImmOI '^cocoOlrOOli-iOliOCOl^ mmw CO 01 H-i CS ■* CO 11 coo CO 01 Tf 01 to 01 1-1 Mi-iMOli-ii-iVOcoOlt-icocOcocool'^cOOlOl OIMM 01 1-1 1-1 m01i-i i-il-lOlONt^OlMVOOllOTf M I-I 1-1 "t'O^COOl lOiO'^'+Ol 0) 01 01 1-1 OOl-l MOlOlOll-lOl lO-^Ol^COMlOM HHM„M CO •^•^ oicocoM oicooi invo I-I 01 01 M 01 1-1 i- m Tt CO 01 M O 0\X) t^vO I0^cooli-i0i-ir\|co-+ iO\0 1^00 O O " 01 co 'i" lO ^ 1^ s^ flj ^ •n 3 CTl i/j 6 )-i V lU u, ^ lU t« U. _C C O ^ B^ !fl 3 ^'o'rt O o -K ^ « 'o'd ^ u 1- o . rtx: horizontal on the bo column s ^ >< W pq t? s s < .S.s't: H x: •- _ H-^-S o rt .j= w t/> o L-. u <+-! o u M-4 a. !> c_c 3 EH o ^ rt % .5 . — 1 O 'rS J x: Jo i_, "'-' O (U >. o > -*-» 1+-1 H 1n ^c 1 >-. jr 1 t-n u 6 (U .E c« P o O 3 X t/3W cq cq i-H N fotni^irsr^t^cvi tJ-m m po tM 1-1 Tf in 0>^ ^ m m 0) M c^ii-i-^i-ioir^rorxoi-rti-i oiw c^ w CM mvo 'tvo »n CO lO r~^ w m >-i •^ trioo CO ^ on ■^ r-x ^ 11 tst w l-lC•^u^^^loc^<^^o^om^'5f5^^c^^^-^ ■-I M ro Tt-vO t^ Tf ir> rovo fO 0) to i-ii-i MTfNuOi-ivoo Ttvo fO N cq M I-H lo IT! rxoo t^ lo 1-1 0) N 01 M c^i 1-1 1-1 w lOOO C^ lO ro t^OO <*) N 1-1 cv) w riq i^ CO ro -t fn TtvO TfP>)rOTt-+i-iLo i-iw 1-1 /'Ml HHCOC^I fOC^rohxi-iON'^mvoi-t 1-1 1-1 ro i-< VO ■^00 00iO-<1-fOi-i Oli-in 11 1-1 N in r<^ lO -rfvo f*^ 1-" >o "^ f*^ 01 M C^ i-i Wm 01 Mi-ioit^ lO-^^^c^oifOTj-ro Mi-i i-i rOOli-ii-iVO'^rrM-TMxrOi-i m OlfTji-i CS '-if»5Tt(OIVsir)Tl-i-ivO ■^VO M 01 1-1 I-I roi-i >-< 1-^ ir>tN>rOT}-rff<0f<^Tti-i -"tM 01 HH w ^ ■H 1-1 01 01 m to rooo On to 01 1-1 01 1-1 M M _ „^(vr, u-)l-lf»2-^ VOmHTfOlOlOll-tOll-1 Tf M01N0101\0^1'^tr)oiroO|iOi-i i-i i-< 01>-i kh lO'TfrC'^l " O C'OO I^\0 mTrccoiwOi-Oin-^ ir,\0 t^OO 0\ O ►< 01 I I I I I i 1 I I TTT 56 WILLIAM HOWARD BATS ON on the y-axis to a point sixty eight units above twenty one on the X-axis leaves eleven points above and ten below with not more than four consecutive points on either side. The sum of the plus deviations is 144+ and the sum of the minus deviations is 131 +. The average deviation is 13.3. Curve XX does not yield to this method of treatment very well, partly for the reason mentioned in the preceding page. No line can be drawn through the curve in such a way as not to leave several consecutive points on one side. A line drawn as indicated from one hundred and twenty on the y-axis to forty units above twenty one on the x-axis shows its general direction about as well as any one line will do. This line leaves ten points above and ten below. The sum of the plus deviations is 1 1 1 + and the sum of the minus deviations is 1094-. The average deviation is II. The direction of Curve XXI is indicated by a line drawn from eighty three on the y-axis to a point forty eight units above twenty one on the x-axis. This line leaves ten points above and eleven below. There are not more than three consecutive points on either side. The sum of the plus deviations is 84-f- and the sum of the minus deviations is 88-|-. The average deviation is 8.6. If the slope of the line is taken to represent the improvement it is seen that Curve XX stands first with a slope of .80. Curves XIX, XXI, and XVIII have respectively .50, .35, and .20. For the purpose of comparison curves were drawn which rep- resented a range of -l-i to — i on the board. For instance in Table IX on the first trial the ball stopped eight times on +1, o, and — I. The sum of these three divisions was taken each trial for divisions on the y-axis and the number of trials for divisions on the X-axis. Curves XXa and XXIa show the general form of the curves obtained by this method. It would seem in this experiment that the person who is able to make a low score would, be the one who is able to perceive ac- curately small amounts of force. It would require the same kind of ability as distinguishing small dififerences in weights. Iti order to test this, three subjects (Subjects D., Do., and C.) ACQUISITION OF SKILL 57 were tested for their ability to distinguish small differences in the weights of bodies. Eight weights were used for the test. Hollow hard rubber cylinders were used for this purpose and loaded with shot and felt so that they weighed 75.001, 76.004 77.002, 78.001, 78.996, 79.998, 85.004, 90.005 grams respectively. The weights were numbered from one to eight. They were presented to the subjects in pairs. The subject was blindfolded and required to lift the cylinders two inches from the table or until they touched a rod running horizontally. The subject re- ported the last weight as heavier or lighter. The tables below show the result of the test : Weights in order i — 8 i — 7 i — 6 i — 5 r — 4 i — 3 i — 2 Subject C. — Right 10 9 10 10 10 9 6 — Wrong I I 4 Weights in order 8 — i 7 — i 6 — i 5 — i 4 — i 3 — i 2 — i Subject C. —Right 10 9 8 5 4 4 5 — Wrong I 2 5 6 6 5 This shows that the subject gave 109 right judgments and 31 wrong ones or he obtained a percentage of 77.8 correct. Weights in order 1--8 1—7 1—6 1—5 1—4 i— 3 1—2 Subject Do. —Right 10 9 5 10 7 8 7 —Wrong I 5 3 2 3 Weights in order 8—1 7—1 6—1 5—1 4—1 3—1 2—1 Subject Do. —Right 8888766 — Wrong 2 2 2 2 3 4 4 This shows that the subject gave 107 right judgments and 22 wrong ones or he obtained a percentage of 76.4 correct. Weights in order i — 8 i — 7 1—6 1—5 1—4 i — 3 i — 2 Subject D. —Right 10 8 7 9 9 9 6 —Wrong 2 3 I I I 4 Weights in order 8—1 7—1 6—1 5—1 4— r 3—1 2—1 Subject D. —Right 10 8 7 2 3 5 4 —Wrong 2 3 8 7 5 6 This sihows that the subject made 97 right judgments and 43 wrong ones or he obtained a percentage of 69. correct. If these figures are compared with those representing the slope of the lines on the curves, it is seen that the subject that showed the greatest improvement in the force test has the largest per- centage of correct judgments of weight and that the other two follow in order. S8 WILLIAM HOWARD BATSON Experiment III. On Time Subjects W., B., C, and D. took part in this experiment. The apparatus consisted of a circular piece of board about twelve inches in diameter and three quarters of an inch thick, placed in a vertical plane on an axis. Beneath the board was a pulley so that a motor could be attached and the board rotated. Two circular bands of cardboard were made by bending strips into the form of a hoop. These two bands were placed on the board, the one inside of the other. The inner band was one and one half inches in height and nine and five eights inches in diameter. The outer band was one and one half inches outside of the inner band and was one half inch high. The space between the two bands of cardboard was divided into twelve parts by tacking pieces of tin on the radii of the circular board. These pieces of tin were the same height as the outside band of cardboard. A trough about twenty inches long was supported so that the lower end rested one half inch above the little pockets on the circular board. The other end of the trough was six and one half inches higher. The trough was made by nailing two pieces of wood together at right angles and then it was lined with soft tin so that it was perfectly smooth. A mark was made in the trough eighteen inches from the lower end. On this mark the subject held a piece of wood that fit the bottom of the trough and behind this was placed an ordinary B.B. shot. When the subject lifted the piece of wood the shot rolled down the trough and fell into one of the pockets on the circular board. One of these pockets was marked by placing a piece of green paper in the bottom of it. The circular board was covered with a piece of cardboard from which a sector had been cut so that only a small part of the circular board was visible beneath the end of the trough. As the board rotated, the subject could see the pockets pass under the end of the trough and then disappear under the cardboard. The object was to release the shot so that they would roll down the trough in time to fall into the marked pocket. The circular board was rotated forty six times per ACQUISITION OF SKILL 59 minute or a pocket passed the end of the trough in one ninth of a second. The movement of the board made it impossible to distinguish any of the pockets except the one that had the green paper in it. The subject waited a short time after the marked pocket had passed the end of the trough and then released the shot. The aim was to improve his ability to judge this period of time. The subject could see the shot hit the pocket if they hit near the middle but if they hit near the side of the pocket he was not able to tell if they hit too early or too late. The pockets were marked zero, one, minus one etc. out in either direction. The score was kept by counting the number of shot in each pocket. These numbers were multiplied by the number of the pocket and then the sum of these products taken for the daily score. Subject W. commenced this work on May 13, 191 3 and con- tinued it until May 24, 191 3. The practice consisted of fifty trials daily. The result of his work is shown in Table XIII. Subjects B., C, and D. commenced work on June 30, 1914 and continued until July 19, 191 4. The practice was daily and consisted of fifty trials. The results are shown in Tables XIV, XV, and XVI. Curves XXII, XXIII (Plate IV), XXIV and XXV (Plate VI) show the form of the improvement. All the curves show rather a large drop during the first two or three practices. After that there is a slow but gradual improvement to the end of the practice. The first large drop in each case is the result of the subject's learning the apparatus. He would have a few very bad results at first which would not occur again after a few trials. Curve XXIII does not show this first early drop to so great an extent because the subject had set up the apparatus and knew by a few trials previous to the first test what to expect. The direction of these curves after the first two or three prac- tices is shown by the straight lines drawn through them. A line run from forty one on the y-axis to seventeen units above thirteen on the x-axis of Curv^e XXII shows that with the ex- ception of the first two practices the points alternate on either side of the line. The sum of the plus deviations is 10 and the 6a WILLIAM HOWARD BATSON sum of the minus deviations is 15. The average deviation is 2.5. On Curve XXIII, a line run from thirty five on the y-axis to fifteen units above twenty one on the x-axis leaves not more than two consecutive points on either side. Omitting the first practice, the sum of the plus deviations is 35 and the sum of the minus deviations is 36. The average deviation is 3.5. On Curve XXIV, a line run from forty three on the y-axis to twenty nine units above twenty one on the x-axis leaves no more than two consecutive points on either side of the line. The sum of the plus deviations is 29.4 and the sum of the minus devia- tions is 33. The average deviation is 3.2. On Curve XXV, a line run from thirty four on the y-axis to a point eighteen units above twenty one on the x-axis leaves no more than three consecutive points on the same side of the line. The sum of the plus deviations omitting the first one, is 41.8 and the sum of the minus deviations is 41.7. The average devia- tion is 4.1. The slope of these lines is .40, .20, .14 and .16 for the curv^es taken in order. When the actual number of hits, that is the number of shot that went into the marked pocket was taken to represent divisions on the y-axis and the trials to represent divisions on the x-axis. the curves take the form represented by Curves XXIVa and XX Va (Plate VII). TABLE XIII Subject W. Experiment — To obtain a curve for learning to judge time. The first horizontal column represents the number of practice days. The first vertical col- umn to the left shows the number of the pockets. The other vertical columns to the right show the distribution of the fifty shot in each prac- tice. The number at the foot of each vertical column shows the total score. May 13 14 15 16 17 18 19 20 21 22 23 24 27 5 I I 4 I 3 2 2 6 7 5 I I 6 6 13 14 12 17 18 23 I 10 10 13 10 2 4 4 I I 3 6 4 I 4 2 ■S I 6 I I 3 i.S 18 8 13 10 T3 8 22 18 25 28 26 zi 30 12 13 15 6 12 3 12 I I 2 r 78 59 38 31 34 29 ii 27 19 >-< CD ■T^00 CO I- i-i r^ O\00 o f^ 0) OvOO t^ 0\ W H o5 ^ ft) rt rr O •° " ^ HI w I o.S?« u, ^ ^ (U D O •£ P ° JJ S w (ij 1- -J 3-£ o "o o ^ '-Ex: 4; -Q a 6 o -M O rt .£ +^ c c>+-"~ u. oj ^ rt — ' o ^ S Jri O >-Q .Si Sj ^ ex P X LOW -^ 10 t^ M PO w M VO 0^ C^ ►-. u^OO On l^ w 10 ts>VO \0 -Tf W IT) t>.\0 >-< 0< ro\0 000 M tN. r^vo 00 M (N 00 f^ iJ-j IT) 01 i-i cq ^ 1^00 -^ I I I I I I ^ X w < -^ o O 1- (U "^ _c ,a-h <^ 6 is 3 O O ^ E E C/3 .— • * — I ■ti o o C u CJ JJ rt 3 nH o o "^^ u ™ tn O u U O. (U o (u E-^ •5 E . rt n *" (U 1-^ O >> > c^ 3 E 1) rt o o ■ o >* -Q u •-> t^ -^ JO a. 3 X C/2W oooo -* 0)M3 COOO M ►H r^ 0\ 1-1 M I I I I I ^ ACQUISITION OF SKILL 63 An examination of all these curves where a single element is involved shows that there is a large degree of similarity in them. All three of the experiments were such that it was hardly possi- ble to reach and maintain the limit of perfection. Except in the cases where the subjects made very high scores during the first two or three practices, the curves show a downward slope with considerable daily fluctuations. These daily fluctuations balance each other in almost every case within four or five days. It is likely that if the practice had been continued until the sub- jects had reached a point where they were no longer able to distinguish the selective element as force or time, the curves would have approached parallelism with the x-axis. In the ball tossing experiments the work was such that a' number of elements had to be improved together or the unit of measure did not indicate the advance. In the second set of ex- periments where the elements of direction, force and time are involved a relatively simple reaction was required. The next set of experiments are such that a number of elements are in- volved but in such a way that the attention may be focused on any one of them or distributed over the whole of them. THE EFFECT OF SEVERAL FACTORS WORKING IN SUCCESSION The ball-tossing experiments showed the nature of the curves of learning when several factors must be attended simultan- eously. The experiments on the judgment of direction, force, and time showed the form of the curves of improvement when the attention is centered on a single factor. It is the purpose of the following experiments to show what form the curves of learning take when several factors influence the attention but in such a way that they can be considered separately. Four subjects — M. (Martin), Fe. (Feemster), S. (Miss Scott), Cd. (Miss Coldwell) took part in the work. They were all undergraduates in college. The apparatus (See Fig. i) consisted of a disk ten inches in diameter cut from a pine board three quarters of an inch thick. This disk was placed in a horizontal plane on an axis. The axis ran through and was fastened beneath a table. Beneath the disk was a pulley by which a motor could be attached. On top of the diik, eight holes had been scooped out similar in form to the bottom of a spoon. These holes were three quarters of an inch in length, one half of an inch wide and one fourth of an inch deep. They were located three quarters of an inch from the circumference of the disk and extended across the ends of four diameters that divided the disk into eight equal parts. A piece of cardboard covered three fourths of the disk leaving one fourth in the form of a sector exposed. The top of the disk was three inches from the table. A frame was set up twenty-three inches from the center of the disk. This frame was perpendicular to the table and was in a plane parallel to the radius that divided the exposed sector of the disk into two equal parts. The frame was sixteen inches square and inclosed a hoop that just touched the sides. A piece of cloth was spread over this hoop and sewed to the hoop all fc ACQUISITION or SKILL 65 the way around the border. y\t the center a circular piece of wire was inserted which was four inches in diameter. A piece of cloth was sewed to this wire ring in such a way as to leave a pocket extending out on the side opposite to the disk. The center of this pocket was eight inches from the table and about five inches above the level of the top of the disk. The subject held in his right hand a pair of tongs nine inches long which were bent down at right angles one inch from the lower end. The seizing end of the tongs was one eighth of an inch wide and was ground perfectly smooth. The work of the subjects was to take these tongs and seize shot from the eight holes on the disk and throw them into the four inch pocket in the middle of the circular hoop twenty-four inches to the right. The disk was rotated eleven times per minute by means of a motor. This made the shot pass the point where they were to be seized at the rate of eighty-eight per minute. After the subject had struck at the shot during one revolution of the disk, he rested until it was refilled. The sub- ject was instructed to strike at every shot that appeared and to throw as many of those that he seized into the pocket as possible. In taking the data, four things were considered : First, the hitting the shot; second, the seizing the shot; third, the hittini: the large circle which included all of the area of the hoop which was sixteen inches in diameter; fourth, the hitting the pocket. The practice consisted of eighty strokes at shot, taken daily, except as noted in the tables. The practice of all the subjects was in the afternoon between twelve-thirty and one-thirty o'clock. Subject M. commenced this work on Jan. 20, 191 4 and con- tinued it until June 5, 1914. He was given a certain position in which to stand, that is on a certain side of the table, and was told to strike at every shot as it appeared with the aim of seizing it and throwing it into the pocket. The importance of striking • every shot was emphasized and at the end of eight strokes he was required to make a stroke at the hole in which he had struck at the first shot. This was done so that he could not take extra time to throw the last shot. The result of his work is shown in Table XVII. Curve XXVI 66 WILLIAM HOWARD BAT SON (Plate V) is plotted from these figures. The spaces on the y- axis represent the number of shot. The spaces on the x-axis show the number of practice periods. The part at the top shows the number of shot hit ; the next cne below it, the number of shot seized ; the third one, the number of shot that hit the big circle ; and the fourth one, the number of shot that were thrown in the pocket. As nearly all of the shot that were seized were thrown so that they hit the big circle, the second and third parts run along almost together. The subject's attention during the first two or three practices was almost entirely upon the disk. He was not able to take his eyes off the board long enough to look at the pocket and then return them in time to make a stroke at the next shot. He then tried to remember the general direction of the pocket. About the third practice, he felt that he was getting a better idea of this direction. He soon found that he had trouble handling the tongs. He would hit the shot but they would roll away from the end of the tongs. This trouble represents a period from the third up to about the fifteenth practice. The wide fluctuations on the sixteenth, seventeenth, eighteenth and nineteenth practices are the result of trouble with the motor. The belt slipped at first so that the disk rotated about one revolution per minute slower. After it was tightened the disk was unsteady in its mo- tion so that the subject was not able to get hold of the shot. It will be noted that very little progress had been made in the second and third parts during the time that the subject was having trouble with the tongs. The fourth part showed some slight improvement which was due -to the subject's gaining a better idea of the general direction of the pocket. The first part shows that the subject had trouble hitting the shot up to the fifty-first practice. About that time he com- menced to watch the disk that carried the shot make one full revolution before he started to strike. He noted the exact po- sition of every shot and, as he said, had a mental picture of just how it looked. It may be seen that parts two and three shortly after this, rise to a higher level. They had remained almost on the same level from the nineteenth to the sixtieth practice with ACQUISITION OF SKILL ^7 the exception of the marked drop between the thirty-fifth and the forty-second practices. Here the subject developed a habit of bringing the shot back with the return stroke of the tongs. This bothered him so that he was not able to grasp the next shot. He discovered the trouble to be due to having changed very slightly the angle of the tongs when releasing the shot. Instead of having the end that contained the shot point toward the pocket, he had been holding them so that the end pointed down toward the table. When he opened the tongs under this condition the quick back motion did not give the shot time to drop but carried it back. During the time from the fifteenth to the fifty-seventh practice the fourth part shows practically no improvement. He had been trying to seize and throw all the shot without trying, directly, to improve the manner of locating the pocket. It may be noted that shortly after the first part had been perfected, the second and third also rose to a higher level. His concentrating his attention upon the position of the shot enabled him to seize them oftener. This may in part account for the rise in the fourth part from the fifty-ninth to the sixty-third practice. But the subject dur- ing this time discovered another change. He at first noted that standing with his feet in a certain position gave him a certain balance that he did not have otherwise. He later found that when he stood with his feet in this position and stretched out his arm straight towards the pocket, the end of the tongs rested about one and one half inches from the center of the pocket. This was the position in which he could do the best work and after he had discovered this way of locating it he always stretched out his arm and took the proper position before starting. There is not much improvement in any of the parts after this. The subject was very desirous of getting a perfect record even up to the end of the practice and thought he could do so if he had had a few more times to practice. He succeeded a few times in making as many as six rounds of the disk or seventy-two strokes without missing. This chance to see any variation in the progress gave an added incentive for greater effort. At the end of the practice, the subject was asked to write out the fact^ s, as he remembered them, that helped him in the work. 68 WILLIAM HOWARD BATSON He gave the following outline : "Some factors that were important in this experiment : I — ^Concentration of mind on the work before beginning. a. Must have a clear conception of the apparatus in mind. b. Must think of the actual work to be done. c. Must think of the distance, direction, and position of the pocket as related to the revolving disk. 2 — Getting the proper position. a. Stand with the feet separted about two feet. b. Stretch the hand that holds the tongs toward the pocket until the tips of the tongs are about one and one-half inches from the center of the pocket. Then look down the arm to see that the same relative position is gained each time. After this, move the hand back and forth as if throwing the shot to insure absolute freedom of the arm. Just before beginning, fix the eyes upon the revolving disk, and at the same time look out of the corner of the eye at the pocket. After the body is in a proper position, let the disk make one complete revolution. During this time look steadily at the shot to see that they are in their proper position. Not only the motion of the disk and the position of the pockets must be related but the time rate of the arm movement must be clearly in mind. 3 — In order to do the best work, must have a feeling that the arm is absolutely under control. 4 — iMust have confidence; yet overconfidence sometimes produces a low score. 5 — Must have a strong purpose in mind, in order to do best work." The increase in the subject's power to concentrate his atten- tion upon the work was very noticeable. After he had practiced several times, the work appeared to take hold of him. There was a set to his whole body and a certain rigidity that was en- tirely absent at first. This grew into a habit so that when the subject watched one of the other subjects perform he was ap- parently under this strain. A very high degree of accuracy with great speed was required but since each trial lasted but slightly over five seconds, the fatigue element did not enter. ,* n ACQUISITION OF SKILL 69 TABLE XVII Subject M. Experiment — Throwing shot. The first column shows the date of practice. The second column shows the number of shot hit; the third, the number of shot seized; the fourth, the number of shot in the big circle; and the fifth, the number of shot in the pocket. This outline is true of each of the three divisions on the page. Jan. 20 56 40 27 12 Mch. 8 Apr. 24 21 79 58 46 16 9 78 60 50 27 25 80 78 77 74 22 80 61 51 21 10 80 67 56 41 26 23 80 59 52 26 II 80 72 68 57 27 80 78 73 69 24 7Q 55 49 19 12 80 67 63 43 28 80 73 73 67 25 13 79 66 60 28 29 So 77 77 70 26 80 70 60 31 14 79 67 59 38 30 80 78 78 69 27 80 59 53 21 15 May 1 80 73 72 58 28 74 55 51 17 16 80 72 70 35 2 80 77 75 66 29 79 60 52 17 17 79 68 60 34 3 30 80 60 57 21 18 80 68 66 45 4 80 76 76 70 31 80 59 52 24 19 80 72 69 52 5 80 74 71 67 Feb. 1 20 77 65 64 41 6 80 77 74 60 2 80 59 53 32 21 80 71 68 49 7 80 71 71 60 3 77 51 43 24 22 8 80 73 74 51 4 79 61 51 36 23 80 66 66 26 9 80 74 74 61 5 79 61 so 29 24 80 71 69 33 10 6 80 65 59 47 25 80 72 66 38 II 80 75 75 67 7 80 74 68 49 26 80 66 63 44 12 80 78 75 69 8 27 13 80 79 75 73 9 79 58 48 18 28 80 74 69 49 14 80 73 73 67 10 79 66 59 26 29 15 80 74 74 67 II 80 71 63 47 30 80 66 65 29 16 80 75 73 68 12 80 68 63 31 31 17 13 75 65 59 44 Apr. i 80 71 67 57 18 80 77 7^ 74 14 79 75 68 56 2 80 73 7i 5o 19 80 77 76 71 15 3 80 73 71 54 20 80 72 72 67 16 79 74 70 28 4 80 72 66 56 21 80 80 78 73 17 80 74 68 45 5 22 80 78 77 74 18 78 71 67 43 6 80 73 68 53 23 80 77 77 72 19 79 68 65 42 7 80 75 73 68 24 20 8 80 78 78 68 25 80 76 76 70 21 9 80 74 74 70 26 80 78 77 67 22 10 80 77 75 63 27 80 78 78 75 23 80 74 65 54 II 80 78 74 64 28 80 80 78 72 24 78 65 61 45 12 29 80 73 73 68 25 80 69 61 42 13 80 76 72 59 30 80 77 76 70 26 80 72 70 62 14 80 75 75 65 31 27 80 70 67 58 15 80 76 75 65 June 1 80 76 75 67 28 79 69 64 50 16 80 79 73 61 2 80 75 75 68 Mch. 1 17 80 80 79 71 3 80 76 75 65 2 78 67 60 35 18 80 77 77 72 4 80 78 77 75 3 80 71 62 54 19 5 80 78 77 73 4 79 65 53 47 20 80 75 75 67 S 79 62 50 36 21 80 78 77 62 6 80 63 40 32 22 79 75 75 7i 7 78 62 48 37 23 80 78 77 68 Subject Fe. commenced work on Jan. 20 and continued until June 4, 1 91 4. The practice was daily except as noted in Table 70 WILLIAM HOWARD BAT SON XVIII. The conditions under which he worked and the instruc- tions were the same as those for Subject M. The resuh of his work is shown in Table XVIII and the form of his improve- ment by Curve XXVII (Plate VI). This Subject did not analyse his manner of procedure as closely as did Subject M. He at first gave his attention to the seizing of the shot but after a few trials he grasped the situation as a whole. By examining the curves, it is seen that there is considerable difference between his work and that of Subject M. His first part was perfected somewhat more quickly. He failed to make a perfect score only four times after the thirty-second practice. Subject M. had trouble with this part of the work up to the fifty-first practice. The second and third parts rise rapidly at first and then gradu- ally until about the fiftieth practice. After that they remain al- most on a level fluctuating between seventy-six and eighty. The fourth part makes practically no progress until the sixth practice ; after that it rises rapidly until the ninth practice. The large fluctuations between the fourteenth and seventeenth practices were due to the belt slipping on the motor. If this trouble had not occurred, it is likely the rise would have followed a more gradual course up to the thirty-sixth or thirty-ninth practice. After that there is a period of practically no progress up to the sixty-first practice, or during twenty-three or four days. At that point, he found that he could give a longer swing to his arm and thus hit the pocket more often. This brought the fourth part to the final level which lasted to the end of the practice. Except this last case, the subject did not discover any factor that had helped him in the improvement. ACQUISITION OF SKILL 71 TABLE XVIII Subject Fe. Experiment — Throwing shot. The first column shows the date of practice. The second column shows the number of shot hit; the third, the number of shot seized; the fourth, the number of shot in the big circle; and the fifth, the number of shot in the pocket. This outline is true of each of the three divisions on the page. Jan. 20 76 58 30 oMch. 7 80 72 67 SpApr. 22 80 79 yy 58 21 80 69 47 I 8 23 80 77 73 70 22 79 70 41 2 9 24 80 78 78 74 23 80 67 40 o 10 80 75 75 61 25 80 80 80 69 24 80 65 58 3 II 80 75 73 64 26 25 12 80 75 71 56 27 26 13 80 77 74 62 28 80 80 80 70 27 78 61 62 12 14 80 77 74 61 29 80 76 76 72 28 80 61 57 16 15 30 9o 77 77 7Z 2f) 79 68 60 28 16 'May i 80 80 78 7Z 30 79 59 53 31 17 80 77 75 64 2 31 78 68 64 26 18 80 78 77 65 3 Feb. 1 19 79 69 69 58 4 2 20 80 79 76 68 5 80 79 78 57 3 78 64 58 20 21 80 79 76 68 6 80 77 76 66 4 79 70 62 29 22 7 80 77 76 67 5 79 73 70 25 23 8 80 78 78 67 6 75 70 56 33 24 80 78 76 62 9 80 79 79 72 7 17 75 64 53 25 79 78 77 67 10 8 26 80 80 79 66 II 9 27 79 72 72 65 12 80 80 79 74 10 79 73 68 35 28 80 79 76 69 13 80 77 73 67 II 80 74 64 56 29 14 80 80 80 74 12 80 74 71 45 30 15 80 80 80 72 13 80 73 64 45 31 80 75 74 61 16 80 80 80 65 14 79 70 62 46 Apr. 1 80 77 77 68 17 15 2 18 16 3 19 80 78 78 70 17 80 71 66 46 4 20 80 79 79 69 18 79 71 69 39 5 21 80 80 80 77 19 80 75 71 43 6 22 So 77 77 6g 20 78 70 68 36 7 80 77 76 70 23 80 79 79 69 21 80 71 67 36 8 24 22 '9 80 77 76 83 25 23 10 80 79 79 67 26 80 79 77 70 24 80 73 69 35 II 80 77 77 62 27 80 80 79 72 25 79 77 74 50 12 28 80 80 80 75 26 80 72 71 55 13 29 80 77 76 72 27 80 77 75 50 14 80 76 76 61 30 80 78 75 64 28 80 77 76 54 15 80 79 77 52 31 Mch. 1 16 79 78 77 52 June i 2 17 80 78 77 60 2 80 77 76 73 3 79 77 74 5i 18 80 79 79 69 3 80 79 78 76 4 19 4 80 76 76 71 5 80 74 73 56 20 6 80 77 72 54 21 80 77 75 65 Subject S. commenced work on Apr. 2, and continued until May 29, 191 4. She worked under the same conditions as the other two subjects and was given the same instructions. The 72 WILLIAM HOWARD BAT SON results of her work are given in Table XIX. Curve XXVIII (Plate VII) shoves the nature of her improvement. During the first two practices she said that the motion of the disk made her dizzy. After that she experienced no more trouble in this respect. She mastered the manipulation of the tongs more quickly than did the other two subjects. The quickness and accuracy of the hand movement enabled her, after about the seventh practice, to devote all her attention to the throwing of the shot. This made her problem nearly the same as when only one factor was to be observed. An examination of the curve that represents her progress shows that after the seventh practice the first three parts show little improvement. With a few exceptions, the second and third parts fluctuate between seventy-six and eighty. The fourth shows an almost constant rise with the usual daily fluctuations. TABLE XIX Subject S. Experiment — Throwing shot. The first column shows the date of practice. The second cokimn shows the number of shot hit; the third, the number of shot seized; the fourth, the number of shot in the big circle; and the fifth, the number of shot in the pocket. This outline is true of each of the three divisions on the page. Apr. 2 71 55 54 laApr. 23 80 77 77 35May 13 80 78 7^, 51 3 7:i 65 59 12 24 80 75 75 45 I4 80 75 74 62 4 79 69 64 23 25 80 78 78 47 15 80 79 79 59 5 26 16 80 77 76 42 6 7Z 69 57 17 27 80 75 74 46 17 7 79 72> 61 22 28 80 79 77 60 18 80 77 77 52 8 78 72 68 25 29 80 79 79 65 19 80 79 78 59 9 80 75 72 23 30 80 79 76 55 20 80 77 76 67 10 80 76 76 35May i 80 77 64 48 21 80 78 78 59 II 80 80 77 28 2 80 77 77 42 22 80 76 76 54 12 3 23 80 79 79 63 13 80 75 71 44 4 80 76 73 44 24 14 80 75 75 39 5 80 80 78 51 25 80 79 79 63 15 80 73 72 37 6 80 77 75 57 26 80 78 78 60 16 80 80 78 37 7 79 77 77 S6 27 80 79 79 63 17 80 80 77 34 8 80 77 77 48 28 80 77 77 69 18 80 74 73 32 9 80 78 78 61 29 80 78 78 72 19 10 30 80 78 77 67 20 79 71 70 27 II 80 78 76 60 31 21 80 78 77 40 12 June 1 80 78 78 72 22 80 77 77 37 13 ACQUISITION OF SKILL 73 Subject Cd. was selected primarily because she had had a large amount of practice on the piano. She commenced work on April 2, and continued until May 29, 19 14. The conditions under which she worked and the instructions given were the same as those for the other subjects. The result of her practice is shown in Table XX and the form of improvement is shown in Curve XXIX (Plate VIII). She did the work with a sort of passive indifference. There was no sign of the effort which was so characteristic of the other subjects. She was able to improve very rapidly. In six days the first three parts had reached a level that was not improved afterwards. In twelve practices the fourth part had reached a level that on an average was not sur- passed during the rest of the practice. From the twelfth to the forty-third practice there was little improvement. This seems to indicate that the work was partly done when she started. The point that was hardest for the other subjects, that of the position of the pocket, caused her no particular trouble. After just a few practices she would move her nand exactly to the same point and in the same way. TABLE XX Subject Cd. Experiment — Throwing shot. The first column shows the date of practice. The second column shows the number of shot hit; the third, the number of shot seized; the fourth, the number of shot in the big circle; and the fifth, the number of shot in the pocket. This outline is true of each of the three divisions on the page. Apr. 2 64 61 30 ooApr. 22 79 72 7i 44'May 11 3 72 64 60 15 23 80 74 72> 55 12 79 79 76 59 4 78 70 69 21 24 79 70 67 45 13 80 7i 7Z 52 5 25 80 76 72 65 14 80 74 75 61 6 74 72 68 13 26 15 ^ 77 77 (>7 7 77 72 71 30 27 80 72 70 55 16 80 74 74 62 8 80 75 72 21 28 80 74 73 61 17 9 79 71 70 26 29 79 72, 71 54 18 10 80 78 76 38 30 79 68 67 42 19 80 74 7^ 54 II 80 79 79 35May i 80 76 75 58 20 80 75 74 46 12 2 76 68 67 53 21 79 71 71 54 13 79 71 71 30 3 22 14 78 75 74 44 4 80 70 69 55 23 80 76 76 61 15 80 78 74 57 5 79 73 69 51 24 80 73 72 64 16 79 71 69 48 6 80 76 75 45 25 17 80 79 76 62 7 78 69 69 47 26 18 80 76 72 62 8 80 72, 72, 53 27 19 9 80 72 71 55 28 79 71 71 S6 20 10 29 80 77 74 64 21 74 WILLIAM HOWARD BAT SON By comparing these four curves, it may be seen to what extent each of the four factors influenced the work and what the effect is on the form of the curves. It will be seen that in all the curves except XXVIII the problem of hitting the shot was not mastered until several weeks practice. This period was extended longest in the case of Curve XXVI. In Curve XXIX, this element caused trouble until the end of the practice. The second and third parts of Curve XXVI show three rises and three plateau periods. The first rise from the first to the third practice represents the period when the subject was getting acquainted with the rate of rotation of the disk. From the third to the fifteenth practice, his attention was on the general direc- tion of the pocket and on the tongs. At the sixteenth practice there was a marked improvement in his ability to handle the tongs. From the sixteenth to the fiftieth practice the attention was centered on seizing and throwing the shot while the hitting of the pocket was ignored. At the fifty-first practice, the subject commenced to watch one full rotation of the disk before starting to strike and he continued to attend to this phase of the work until the fifty-seventh practice. Here he improves his seizing on the basis of better striking and also notices a better way of standing. Curve XXVII shows no such steps in the second and third parts. There is a general rise up to about the fiftieth practice and after that little improvement. The subject's manner of at- tending gave him no factor for separate consideration. The second and third parts of Curve XXVIII show about the same form as the same parts of Curve XXVII. The first factor hav- ing been mastered after about six practices there was no element to cause trouble and no particular improvement in the plan of the work. In Curve XXIX the second and third parts rose rapidly for four or five practices, but did not succeed in reaching so high a level for this part of the work as that reached by the other subjects. Curve XXVI shows in the fourth part a very pronounced plateau type of curve. At first, there is a slight rise partly due to the subject's getting acquainted with the apparatus and partly ACQUISITION OF SKILL 75 the result of his improving the method then being used. After about sixteen practices, a level was reached in this part of the curve that remains practically constant until the fifty-eighth prac- tice. What was going on at this time may be partly accounted for by noticing the other parts. The rise to the next level in- volved two factors at least. One was this mastering the first element and the second followed as a consequence from this. The subject being somewhat free discovered the proper way to gtt the most suitable standing position. He had noted a feeling of ease when in the proper position even before he had analysed a method to make sure of it every trial. The fourth part of Curve XXVII shows no improvement for four practices. During this time, the subject had given all his attention to the other factors. Then by attending the whole problem he reached a level at about the forty-first practice which seemed to have been the limit for the method he was using. After he had discovered a new method, that of extending his arm farther, he brought this part of the curve to the next level. The fourth part of Curve XXVIII shows a constant rise for the reason that the subject was attending to this part from the first. No especial changes in method were noted. The fourth part of Curve XXIX shows a rapid rise for a few practices which seems to have been the result of adapting an ability already ex- isting to this new situation. After this process had become per- fect no further advance was made. No new method was devised and no special effort was put forth. There was a drop from the twenty-first to the thirty-second practice which appeared to be the result of indifference. GENERAL DISCUSSION The first aim of this study was to get further information in regard to the long period fluctuations or plateaus in the curve of learning. It is seen from these experiments that only in the last division of the work are there plateaus of long length. Neither in the ball tossing experiments nor in the work done by Swift, although the work extended over a considerable period of time in some cases, does there appear a long arrest between two rapid 76 WILLIAM HOWARD BATSON progress periods. The middle division of this work where only one element was involved did not in any case extend over a period of time of sufficient length to show a long plateau. In the last division of the work, however, in one case in particular the plateaus were very pronounced if the part representing the shot thrown in the pocket is considered. No one, so far as the writer is aware, has contended that the long period plateaus are a necessary part of all learning curves regardless of the type of learning. Bryan and Harter, however, in their work were convinced that these long periods of slow progress were necessary in the curves for receiving the tele- graphic language. Book found the long period fluctuation in two of the curves he obtained for learning typewriting but he did not think they were a necessary part of the development. The last division of this work shows that there may be or may not be plateaus of long length in the same type of work. One way of deciding the question of the necessity of plateaus for a particular type of learning would be to try a large number of individuals and see if they uniformly had plateaus in their improvement. Another way and a more satisfactory plan would be to attempt to analyse out the factors and see just what is in- volved in the formation of the arrest period. It has been seen (see Introduction to this study) that Bryan and Harter attributed the plateaus to the time it takes to form diflferent-order habits : Swift, although he found only short plateaus, believes for the most part that they can be explained on the basis of the associa- tions being made automatic; and Book thought they were the result of lapse of attention or misdirected attention. In none of these publications was an attempt made to so arrange the work that an opportunity would be given to see just how the separate factors acted during a plateau period and what effect the dif- erent reactions to them by different subjects had upon the work as a whole. The last division of the work reported in this study was planned with the intention of having a number of elements in- volved in such a way that the subject could attend to all of them at once or attend to one or more of them separately from the ACQUISITION OF SKILL 77 rest. It may be objected that the factors that were measured in this work must follow one another in a certain order and that the subject could not modify this or in other words change the order of the factors. This is true,— the subject could not throw or hit the pocket until he had struck or seized the shot— but this did not prevent him from giving practically all his attention to the first factor or distributing it over the work as a whole. What occurred in the subjects tested for this work may be seen by examining the different divisions of the curves obtained for the four subjects used in this experiment. If the last or fourth part (the part obtained by using the shot that hit the pocket as a unit of measure) of the curves for each of these subjects be observed, it will be seen that one of the sub- jects (Subject M.) had two rises and two marked plateaus. Subject Fe. showed a short arrest period at first, a rapid rise and then another period of arrest. Subject S. showed no sign of an arrest period in this part of her work but an almost con- stant rise from one end of the curve to the other. Subject Cd. at first showed a rapid rise and then a long period of delay which was not broken at the end of the practice. If the hitting of the pocket had been the only goal and the unit of measure had considered nothing else, the explanation of the curve would have had to depend upon introspections. Here, however, the other parts of the curve give an objective basis for explanation. Subject M. who had two pronounced plateaus at- tacked the work in parts ; he attended to the first or basal factor until it was under a high degree of control. After that he took up the other factors. It is true the improvement was due, in part, to improvement in method but this depended upon the fact that his attention was free to attack the other part of the prob- lem. In this case the subject followed a plan parallel to what Bryan and Harter explained when they spoke of the lower order habits being completed or automatized first. There was no evi- dence in this subject of a lapse and renewal of attention as Book pointed out in his work on typewriting. It may be that the work was such that the subject could see that he was making progress all the time ; that is, he could see that he was hitting the shot or 78 WILLIAM HOWARD BAT SON seizing them better and this kept up his interest. However, the fact stands that it was not lapse of attention that caused the long arrest in the fourth part of the curve. Subject Fe. had a short period of arrest in the fourth part of his curve, while there was a relatively large improvement in the other parts. This indicated that he was attending to those first factors. Later all the parts of the curve make gains together. He was able to attend during this time to the process as a whole. When he reached a level near the limit of his ability to improve, all the divisions of the curve sloped off more nearly parallel to the X-axis. Subject S. had no plateau in the fourth division of her work, because, as can be seen from the other three parts of her curve the first three factors presented her no trouble. As has been explained, the quickness of her hand movement and the accuracy with which she used the tongs enabled her to master these factors in a short time. In her work, it was simply a case of giving her attention to a single factor and the curve of the type of the simple association curve (see discussion on the form of the curves) was the result. Subject Cd. had much the same form of curve in the fourth part but the rise was more rapid. The reason for that has been explained already. The practice on the piano was a training that only needed special adaptation. This took place rapidly. After that she improved the work in a short time to a point near the limit of her ability. From this study and from the other experiments done in this field, the writer feels justified in drawing the following conclu- sions in regard to the plateau. There is no evidence to show that they ever occur in learning processes where there is only a single association involved. They may or may not occur in a complex learning process. If the factors involved are of such a nature that they must be improved together or if the subject is able to attend them as a whole, there will be no plateau. If, however, the nature of the work is such that the factors must be attended in succession or the subject gives his attention to the separate factors, as such, there will be plateaus. ACQUISITION OF SKILL 79 Ball-tossing may be taken as an illustration of where the factors involved must be improved together. At least three of these factors have been mentioned and studied to some extent in this work. If these factors are taken, it is seen that it is prac- tically impossible for the subject to separate them. One of them counts for nothing unless the others go along with it. It might be conceived that the subject could practice throwing the ball a certain height or in a certain direction, yet in this case, he would either have to watch it or use some other means to get a selective factor, that is, to show when an improvement is made and this would give him a criterion for the time element. Again if he improves his ability to throw in a certain direction and does not improve his ability to shift his hand to the position where the ball will fall, there will be no improvement in the unit of measure. In the work on throwing the shot, the separate factors could be improved in succession. The subject could practice with his attention fixed on the seizing of the shot until he had acquired the ability to hit and seize practically every one in order, or he could distribute his attention on the separate factors about equal- ly. The aim was to have the subjects learn each factor in order but some of the subjects were more completely able to do this than others. In such cases as this last division of work where the shot throwing is the problem, the attention will be given to a part or the whole of the work according to all the conditions that govern the attention of the individual at the particular time of the practice. The Daily Fluctuations The daily fluctuations appeared in all the work that was done. In fact there seem to be no cases of learning where they do not appear. They indicate very clearly the variation in the physical and mental condition of the subject from day to day. At times the subjects were able to tell in advance that they would make a good score but in this work, just as Swift found, their thinking that they would make a good score was no proof that they would. It was noted that Subject F. in his tossing the pieces of hard rubber was set back by his confidence that he would make a high score after the first relatively high score. 8o WILLIAM HOWARD BAT SON Whether these fluctuations were greater at the beginning or at the end of the practice depended upon the nature of the work and the unit of measure. If there was a definite set Hmit for the unit of measure as in the case of the direction, force, and time curves, or in the shot throwing curves, the later variations were likely to be less. But on the contrary if the unit of measure had no limit other than fatigue or some physical or mental factor as in the case of the ball-tossing, the later fluctuations were rela- tively large. This work differs from a large part of the experimental work that has been done in this field in that the material from day to day was kept exactly the same. In work like typewriting, short- hand, and telegraphy where different material is used on suc- ceeding days, the fluctuation may be partly at least accounted for by the fact that the subject matter was more difficult on some days than on others. The fluctuations are useful for indicating the relative amount of ability shown daily by subjects doing the same kind of work but it must be remembered that the nature of the unit of measure will largely determine the apparent amount of fluctuation each day. It may also be noted that because the fluctuations are larger toward the end of the curve, one should not conclude that the daily ability varies more at that time of the practice. The time and direction curves may be taken to illustrate the first point. Because the average deviation of the time curves was less than that of the direction curves, it does not follow that the subjects varied less daily from a certain line of improvement in judging time than they did in judging direction. If more pockets had been made on the disk, the variation for the time test would have been greater though the actual daily ability would have been the same. Again if the ball tossing curves, where the number of catches form the unit of measure, are noticed, it will be seen that the latter part of the curves shows wide daily fluctuations. If it were true that catching ten balls meant the same whether they are the first ten or the ten between ninety and a hundred then these fluctuations near the end of the curve would mean a greater ACQUISITION or SKILL 8i daily variation in ability, but no one is likely to contend that this is true. That the fluctuations may be seen to be due to the nature of the unit of measure, attention need only be called to the forms of Curves Ila and IVa. The latter part of these curves shows little variation and gives a directly opposite result from what the first method of plotting the ball tossing curves seems to indicate. The General Form of the Curves For the purposes of discussion, the curves of learning may con- veniently be divided into four classes. When the number of catches in a given number of trials (Curves I to V), the recipro- cals of the number of errors per given group (Curves IX and X) as five hundred or one thousand, or the average number of catches in thousand groups (Curves XI to XV) are used as the basis for the unit of measure the curves all show the same general form. They are all concave to the y-axis. Some of them do not show this peculiarity to so great an extent as others but this can be explained on the ground that the subject had had previous practice or similar practice which modified the form of the early part of the curve. Another class is where the slope is relatively constant up or down, accordingly as the unit of measure is such that it increases or decreases, until a point is reached near the limit of the sul)- ject's skill. This class is illustrated by the experiments on di- rection, force and time. A third class is where the curve rises or falls rapidly at first and then gradually approaches parallelism with the x-axis. This class is illustrated by the second and third divisions of some of the shot throwing curves and those for telegraphic sending in the study of Bryan and Harter; also by the ball-tossing experi- ments when the unit of measure is the number of errors made in catching a definite number, as five hundred or a thousand, or the percent the number of catches is of the number of tosses. The fourth class is the long level plateau class where there is a relatively rapid rise and then a period of delay and then an- other period of rapid rise. This is illustrated by the fourth part 82 WILLIAM HOWARD BATSON of the curve of Subject M. obtained for shot tossing, and also by some of the typewriting curves of Book and the telegraphic re- ceiving curves by Bryan and Harter. The curves for ball tossing when the unit of measure is the number of catches made in a certain number of trials, show a long period when comparatively little progress is made and then a sudden rise at the end. This is true regardless of whether the subject attempts to use a particular method, or to develop some special part of the work. It is true also regardless of the distri- bution of practice. When the subject had a practice period of approximately five hundred catches instead of ten trials, the curve obtained when ten trials was made the basis of the score, was the same in form (Curve VIII). There are a number of factors that must be considered in try- ing to interpret the significance of this type of curve. The long period of apparent slow progress at the beginning of the ball tossing curves where a number of trials is made the unit for the divisions on the x-axis and the sum of the catches is the basis for the divisions on the y-axis may be explained by the fact that the long period represents relatively little practice. When the subject is able to make less than one hundred catches during the ten trials he gets much less practice than when he can make a thousand or more catches in the same number of trials. This holds true for Curve VIII, when the practice period was five hundred but the divisions on the x-axis were based on groups of ten trials as well as for the first five ball tossing curves. The rapid rise at the end of all the curves of tlij| type is due to the peculiar nature of the unit of measure; that is, in the ball tossing curves where the number of trials is used as the basis for divisions on the x-axis (Curves I to V and also VIII) ; the curves where the reciprocals of the number of errors per given group as five hundred or one thousand (Curves IX and X) are used; and the curves where the averages per thousand (Curves XI to XV) are made the basis for the divisions on the y-axis. In the case of Curves I to V and VIII, ten trial groups are made the basis for divisions on the x-axis but a trial here means the number of catches between two errors or misses. It follows ACQUISITION OF SKILL 33 that the significance of a trial in the earHer and later part of the work in any experiment varies greatly. The implication is that ten catches represent the same amount of skill whether they are made on the first day's practice or on the last day's practice or whether it is the first ten caught or the ten between ninety and one hundred, and that to catch one hundred balls in one trial represents a degree of skill ten times as great as to catch ten balls in one trial. In all the other experiments of- this work, the unit of measure would be comparable to a single toss, that is the toss would con- stitute a trial. The degree of skill would be represented by the proportion of tosses that are successful. If the subject catches ten balls the first time before he misses, he has ten successes out of eleven trials or ninety-one percent would represent his degree of skill. When he catches one hundred balls without missing, he has ninety-nine percent for his degree of success. Instead of his success or skill being ten times as great in the second case as in the first as it would be by the first method, it is really only eight per cent greater. In the curves where the reciprocals of the number of errors is made the basis for the divisions on the y-axis the result is the same as where the ten trial basis is used. The assumption is that to catch one hundred balls in one trial represents a degree of skill ten times as great as to catch ten balls in one trial. The efifect can be seen readily by noting that the vertical distances (distances above the x-axis) vary as a constant divided by a decreasing variable. j^ When the data used in plotting these first curves, that is the first five for ball tossing are used so as to make the actual amount of practice the basis upon which to count gain in skill a very different form of curve is obtained (see curve Ila and IVa). In these cases the basis was one thousand catches which was a definite amount of practice and the success was measured by counting what percent this was of the total number of tosses necessary to make the thousand catches. The implication is that a miss is no practice, at least it gives no increase in skill. This method gives practically the same form of curves as taking the 84 WILLIAM HOWARD BAT SON number of catches made in each one thousand tosses as the basis for divisions on the y-axis. Instead of measuring the absohite gain this method measures the relative gain. With this method of measuring gain, there is no rapid rise at the end of the curves. Thorndike using Swift's data points out that when the average number of tosses is used as the basis for the divisions on the y-axis and the thousand groups are the basis for the divisions on the X-axis, a curve of the form of type three is the result.^^ The data obtained in this work do not give that form of curve when treated in that way. The second class of curves is where the slope remains up or down depending on whether the unit of measure is such that the values representing the divisicns on the y-axis increase or decrease with the increase of skill until a point of development is reached near the limit of its possibilities. This may be de- termined by the particular function having been perfected as in the case of direction if the subject could have hit the mark every time or by his approaching a place where he could no longer get a selective factor, that is he passes into a realm beyond the threshold of sensory discrimination in that particular field. For instance, if the experiment on force is taken as an example it may be conceived that the subject will be able to hit within a certain minimum distance of the mark every throw but after he reaches a plane where the difference required to put the ball within that space and beyond it falls below a certain fraction of the weight of the ball that he is throwing, he would be able no longer to control the ball so as to reduce the score. At this point the curve would slope off so as to become practically parallel with the x-axis. Sometimes in the case of a single factor acting, the ap- paratus may at first cause some confusion so that the first few scores may be abnormally high. This makes it appear that the curve rises or falls rapidly at first. The curves obtained for the judgment of time are examples of this. All the simple association experiments in this work as well as those in other experiments that have been examined show a con- stant and gradual improvement with the daily fluctuations bal- -■f Thorndike, E. L. "Educational Psychology," Vol. II, p. 122. ACQUISITION OF SKILL 8s ancing each other, as a rule, in such a way as to approach a straight line until near the limit of the development. The third type may be the result of a number of elements gradually becoming perfected together and as the elements ap- proach their limit the improvement grows less. Where the work is complex, and the unit of measure is such that it shows the full value of the improvement as fast as it is made, and the subject attends the process as a whole, the curve rises or falls rapidly at first and as the chance for improvement in the separate ele- ments becomes less the rise becomes less rapid. The curves for the sending of the telegraphic language illustrate this form as do also some of the curves obtained by Book in the typewriting experiments, and Curves Ila. and IVa. for ball tossing. The fourth class of curves is that of the long plateau form which has already been discussed under the head of long time fluctuations. In addition to what has already been said, it is interesting to note that Starch's work on tracing a six-pointed star {op. cit., p. 12) shows the effect of two elements working. He plotted a curve for speed and one for error. The error curve was of the third type and the time curve was in a measure like the fourth type. He does not so state but it is probable that the subject was giving his attention largely to the error side of the work while the plateau period lasted in the time or speed curve. The writer believes there is no typical curve for all types of learning. Where the work is so simple that the attention cannot be distributed as in the case of rolling a steel ball at a mark or the work is of such a nature that the attention must be given in a certain way, the curves are likely to be of similar form for different subjects. Where the work is complex and the attention can be distributed on different parts of the process, then the curve for different individuals will vary, though the work is the same. The Influence of Objective and Subjective Factors This subject has been discussed in detail by Swift and Book. Nothing more was noted in this work than that which empha- sizes what they have already said. The objective factors such as 86 WILLIAM HOWARD BAT SON light, temperature, nature of the work, and apparatus, were all very important and any change showed its effect especially in the early part of the practice. The same was true with the subjective factors. However these were not always so easily analysed out. All of these factors showed themselves more influential in the early part of the work when the attention was actively involved than later when the process was more automatic. The experience of expert telegraphers, pianists, and others where there is a large element of motor control involved, demonstrate this beyond doubt. This does not assert, however, that where speed is in- volved distraction of attention does not affect it. The Warming up Process A part of this work showed a warming up and a part did not. The work in ball-tossing shows clearly that after the first trial there were better results. Book also found this warming up ten- dency to a very marked degree in his work. In the experiments on the judgment of direction there was little evidence of warm- ing up. As the work was such that fatigue would not enter to any great extent, a division of the practice periods into two equal parts ought to show better results for the last half if there was any noticeable warming up. These figures do not show enough advantage to the last half of the work to be of any weight. Wells found in another experiment where a single association had to be formed that there was evidence of warming up to a consider- able extent. In a complex work, the warming up may take the form of renewed associations but in a simple work this is not necessary. It would seem that the difference between the work that Wells had his subject do, that of tapping at maximum speed on a telegraph key, and the work in learning direction consisted in this, that in the first case special preparation was necessary while in the other the ordinary movements of the day would be all that is necessary to keep the arm in proper condition. The Effect of Short and Long Rest Periods An examination of the ball-tossing experiments shows that there is no regularity in what happens after a short rest period ACQUISITION OF SKILL 87 of a day or two. Sometimes the score is higher and sometimes lower. This was also true to a large extent in the shot throwing experiments. In the other experiments no opportunity was given to test the effect of short rest periods as the work was con- tinuous from the beginning to the end. The shot throwing experiments gave an opportunity to test the effect of a short rest period as the work was discontinued on Sunday each week. In the case of Subject Fe. both Sunday and Monday were missed each week. The work of Subjects M. and Fe. extended over nineteen Sundays. Subject M. threw more shot in the pocket on Saturday than on Monday eleven times out of the nineteen. Subject Fe., who had missed two days' practice threw more shot in the pocket on Saturday than on Tuesday nine times out of nineteen. Subject S. threw more shot in the pocket on Saturday than on Monday three times out of eight and Sub- ject Cd., six times out of eight. So far as this part of the com- parison goes it is seen that two of the subjects did better after the rest period and two did better before it. If, instead of counting only the number of shot put in the pocket, the whole work is considered, that is groups of figures in the four columns of the tables for the work immediately be- fore and after the rest periods, it is seen that for Subject M. thirty-one scores were better on Saturday than on Monday, twenty-four were better on Monday than on Saturday and the others are equal. Subject Fe. has twenty-eight better on Satur- day and twenty-two better on Tuesday. Subject S. has twelve better on Saturday and eleven better on Monday. Subject Cd. has twenty-one better on Saturday and eight better on Monday. The advantage seems to be slightly in favor of the practice pre- ceding the rest period. For the long rest period, the results show that where there was not a positive gain, a condition existed that made a rapid gain possible when practice was resumed anew. After a rest of more than six months, subject D. was able to make nearly as high a score the second practice period as he had made the last day in the regular period. Almost a year after this, he was able to make a score nearly twice as high as any he had ever made 88 WILLIAM HOWARD BATSON before. This was in the first test made. Here was a positive gain with less than a half hour's practice having been taken dur- ing the intervening period. Subject B., after more than a year's rest, made a score on the first day that the practice was resumed with his right hand almost as high as he had ever made before. Two days later he was able to make a higher score than he had ever made before. The left hand had apparently not retained the same amount of skill as the right hand. However, the second retention test showed a marked improvement over the first. The score was as high as any he had ever made in the regular practice except the last three or four where the usual final spurt had occurred. Subject W., after a rest period of nearly five months, made a score during his second practice higher than he had ever made during his regular practice. A little more than a year after this retention test, he made a score higher than any he had ever made in the regular practice and almost equal to the last score that he made in his first retention test. Subject F. was the only subject that did not show results in the retention test practically equal to or superior to the best scores in the regular practice. The reasons for this have been explained in the discussion of his work. Although the work was not carried very far, one interesting point to note is the rapidity with which the improvement takes place in the repeated retention tests. Subject D. in his second retention test which was taken two days after his first shows an improvement of nearly twenty-five per cent over the first one. Subject B., made a gain of almost thirty per cent in the second test over the first with his right hand. With his left hand there was a gain of forty-three per cent in the second test over the first. Subject W., in his second test, showed a gain of nearly ninety per cent over the first. Subject F. was the only subject that did not show a gain in the second retention test over the first. In his case there was a loss of about twenty per cent. It is interesting to note that this rapid regaining of an acquired ability is closely parallel to the rapid improvement in an instinctive response that has been delayed for a short time. Shepard and ACQUISITION OF SKILL 89 Breed found in their study of chicks^® that if the pecking re- sponse of the chicks is delayed for a few days the accuracy of the pecking increased much more rapidly than it did when the instinct took its normal course; in other words the chicks that had been delayed soon gained an ability equal to those that had normally developed. The only place in this study that a test was made for the effect of a long rest period, except the ball tossing experiments, was for direction with Subject W. His first test showed that he had apparently lost much of his former skill. The reason for that has been already given in the discussion of that experiment. His second test taken immediately afterwards showed that he had regained practically all the skill he had acquired by his previous training. The data obtained in these tests only confirmed what has been pointed out by Swift, Book, Wells, and others. Summary I- — The plateaus of learning depend upon the factors involved in the process to be learned and upon the distribution of atten- tion. They are not found in simple types of learning and may or may not be found in complex types. 2. — The daily fluctuations are common to all types of learning and depend upon the objective and subjective factors involved. They may be deceptive because of the influence of the nature of the unit of measure upon them but they are useful for the purpose of comparing the daily variability of two or more subjects that are at the same point of development in doing the same type of work. 3.— 'Both the factors involved and the attention of the subject influence the general form of the curves of learning. Where the work is simple, that is requires a simple sensori-motor associa- tion, there will be a typical curve common to different subjects. This is also true where the work is more complex and the factors must, from the nature of the work, be attended in a certain order. " Shepard, John F. and Breed, F. S., "Maturation and Use in the Develop- ment of an Instinct." Jour, of Animal Behavior. July-August, 1913, vol 3 no. 4, pp. 274-285. 90 WILLIAM HOWARD BATSON 4. — Objective and subjective factors have a very great influ- ence on the learning, especially in the early part of the vvrork. 5. — The w^arming up process is the necessary accompaniment of all work where the ordinary daily activity does not provide the special preparation required for performing the work effectively. 6. — The results do not show that a short rest period has any marked effect on the work one way or the other. 7. — After a long rest period the subject is found to be in a condition to improve very rapidly. In some cases the results show that they have actually gained power during the rest period. BIBLIOGRAPHY Bair, J. H. — "Development of Voluntary Control." Psy. Rev., vol. 8, pp. 474-510- "The Practice Curve." Mon. Sup. No. 19 Psy. Rev. Book, W. F. — "The Psychology of Skill : with Special Reference to Its Acquisition in Typewriting." Univ. of Montana Pub- lications in Psychology. Bui. No. 53. Bryan, W. L. and Harter, N. — "Studies in the Psychology and Physiology of the Telegraphic Language." Psy. Rev., vol. 4, pp. 27-53, and vol. 6, pp. 345-375- Johnson, W. S. — "Researches in Practice and Habit." Studies from the Yale Psychological Laboratory, vol. 6, pp. 51-103. Partridge, G. E.^ — "Experiments upon the Control of the Reflex Wink." Amer. Jour, of Psy., vol. 11, pp. 244-250. Scripture, E. W., Smith, T. L., and Brown, E. M. — "On the Education of Muscular Control and Power." Studies from the Yale Psychological Laboratory, vol. 2, pp. 114-119. Shepard, John F. and Breed, F. S.— "Maturation and Use in the Development of an Instinct." Jour, of Animal Behavior, July- August, 191 3, vol. 3, no. 4, pp. 274-283. Starch, D.— "A Demonstration of the Trial and Error Method of Learning." Psy. Bui., vol. 7, pp. 20-23, Swift, E. J. — "Studies in the Psychology and Physiology of Learning." Amer. Jour, of Psy., vol. 14, pp. 201-251. "The Acquisition of Skill in Typewriting; A Contribution to the Psychology of Learning." Psy. Bui., vol I, pp. 295- 305- "Memory of a Complex Skillful Act." Amer. Jour, of Psy. vol. 16, pp. 131-133- "Beginning a Language : A Contribution to the Psychology of Learning." Studies of Philosophy and Psychology by Former Students of Charles Edward Garman, pp. 297-313. "Memory of Skillful Movements." Psy. Bui, vol. 3, pp. 185-187. "Relearning a skillful Act: An Experimental Study in Neuro-Muscular Memory." Psy. Bui., vol. 7, pp. 17-19. "Learning to Telegraph." Psy. Bui., vol. 7, pp. 149-153. Swift, E. J. and Schuyler, W.— "The Learning Process." Psy. Bui., vol. 4, pp. 307-310. Thorndike, E. L.— "Educational Psychology." Vol. 11. Wells, F. L.— Normal Performance in the Tapping Test." Amer. Jour, of Psy., vol. 19, pp. 437-483. "The Relation of Practice to Individual Differences." Amer. Jour, of Psy., vol. 23, pp. 75-88. Whitley, M. T.— "An Empirical Study of Certain Tests for In- dividual Differences." Archives of Psychology, No. 19. CURVES AND PLATES Curves I to V (inclusive) Plate I Curves I la and IVa Plate I Curves Illb and IVb Plate I Curves VI and VII Plate II Curves VIII to X (inclusive) Plate III Curves XI to XIV (inclusive) Plate II Curve XV Plate III Curves XVI and XVII Plate IV Curve XVIa Plate VIII Curves XVIII and XIX Plate V Curves XX and XXI Plate VI Curves XXa and XXIa Plate VI Curves XXII and XXIII Plate IV Curves XXIV and XXV Plate VI Curves XXIVa and XXVa Plate VII Curve XXVI Plate V Curve XXVII Plate VI Curve XXVIII Plate VII Curve XXIX Plate VIII Bcse h moid Plate J. YJ^SndT}d_E_/\/Xl Plate]. /O /^ 10 ' is 30 35 HO 45 50 55 (^o 05 ^ /Q _ 75 ^0 t lA rv^ W L i Xri/g Z m m Cd i ry e lyh C "el- u r\ -m [a'i± 2IL. /5 10 \ 5 10 d IL TM.2^^ V H5 50 55 (>0 05 70 75 S'O ; // , o2/ j/ V/ 5/ ^/ 71 $1 Phtem PUtel. PlateW io i5 100 m no PXdite^. (>JLmL H5 ir 50 P/at- P bo ^6 ^o 75 100 m no 55 &0 ^5 70 75 $0 r5 90 PJditem ^1 -^i'^-riMj^'L PlateYW: 5 /O 15 XO 15 30 35 HO