UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The relation of intelligence and the spread of effect Sutherland, Margaret Ruth 1949

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1949_A8 S8 R3.pdf [ 2.78MB ]
Metadata
JSON: 831-1.0107028.json
JSON-LD: 831-1.0107028-ld.json
RDF/XML (Pretty): 831-1.0107028-rdf.xml
RDF/JSON: 831-1.0107028-rdf.json
Turtle: 831-1.0107028-turtle.txt
N-Triples: 831-1.0107028-rdf-ntriples.txt
Original Record: 831-1.0107028-source.json
Full Text
831-1.0107028-fulltext.txt
Citation
831-1.0107028.ris

Full Text

a ^  ft T> The R e l a t i o n of I n t e l l i g e n c e and the Spread of E f f e c t by Margaret Ruth Sutherland A Thesis Submitted i n P a r t i a l F u l f i l m e n t of the Requirements f o r the Degree of MASTER OF ARTS i n the Department of EDUCATION The U n i v e r s i t y of B r i t i s h Columbia August, 1949 ABSTRACT The problem investigated was the relation of intelligence to the spread of effect. A n u l l hypothesis was set up, that the spread pattern obtained from more intelligent subjects would not d i f f e r from that found in the less intelligent. Data were collected from two groups of thirty subjects each, one composed of "bright*1 and the other of "dull** students* A l l were pupils in Grades V, VI, and VII in the same school. The material was of the conventional type (word-stimulus, number-response) used in many "effect" experiments, but the typical procedure of rewarding correct responses with the i i announcement "right" and punishing wrong responses with the announcement "wrong" was modified by the omission of the announcement "wrong" during the course of the experiment* Serial position effects were obviated by making successive presentations of the l i s t of stimulus words continuous, and by the length of the l i s t . Favored responses were determined with the help of two presentations free from reward at the beginning of the experiment, and were eliminated from a l l calculations in order to establish a neutral base-line, which was determined by computing the percentage of total repetitions throughout the five presentations during which rewards were given. Gradients were plotted for each group from the percentages of repetition of rewarded responses and of repetitions one, two, and three steps before and after rewarded responses. The results are such that the null hypothesis must be accepted, that i s , intelligence as measured by a standard test i s not a variable factor in determining the spread of effect. Of the group differences found, none i s s t a t i s t i c a l l y reliable. In so f a r as can be judged from a single experiment and within the parameters of that experiment, i t is conoluded that reward has equal effeot on bright and d u l l students in a se r i a l learning situation. The relatively low percentage levels of repetition as compared with those of previous studies is attributed to one of two factors, or possibly to a combination of both. In the f i r s t place, the method of assembling the data precluded favored responses from contributing to the gradients obtained, and so reduced the number of repetitions calculated. In' the second, the word-list was of a length commonly used with subjects at the college l e v e l . From these facts, two tentative conclusions are reaohed. One, that to the extent that favored responses contribute to gradient data, levels of repetition obtained in a number of previous studies and attributed to the effect of reward, are spuriously high, and the influence of reward has been exaggerated. Two, the length of the test (in this case, the word-list), i s a factor deter-mining the influence of reward. It i s thought that both of these conditions may be responsible, to a degree so f a r - 3 -undetermined, for the results obtained in this study. Similarly to the levels of repetition found in eaoh response category, the height of the established base-line i s conspicuously lower than any previously adopted, and i s noteworthy in i t s close approximation to the pure chance figure* It has been accounted f o r on the basis'of elimination of favored responses and ser i a l position effects, and would also have been affected by the factor of length of the word-list i f this were an experimental variable* A summary consideration of the slope of the gradients from these data compared with gradient curves from previous studies where punishment i n the form of the announcement "wrong,H as well as reward, was administered to the subjects, revealed no consistent trends and added nothing conclusive by way of evidence on the influenoe of punishment in a learning situation, other than to emphasize the apparently varying roles this type of "punishment" can play, and the inadvisability of generalizing fiom the evidence thus f a r available on i t s modus operandi. In addition, the fact that unrewarded responses, which were not punished, in seven out of twelve categories were repeated less frequently than consideration of the neutral base-lines would have led one to expeot, requires explanation. It has been hypothesized that reward, in emphasizing the correct response, acts as a distraction on neighboring connections and thereby reduces their rate of repetition below the chance level* - 4 -Suggestions were made f o r f u r t h e r r e s e a r c h i n t o the r e l a t i o n of i n t e l l i g e n c e and extent of spread; i n t o the technique f o r e s t a b l i s h i n g a b a s e - l i n e ; i n t o the f a c t o r of favored responses as unduly magnifying the e f f e c t of reward; and i n t o the f a l l i n g o f f of r e p e t i t i o n s of unrewarded though unpunished responses below the obtained chance l i n e . ( i ) Acknowledgements In presenting t h i s study, the writer wishes to make the following acknowledgements: to Dr. F. T. Tyler, now of the Unive r s i t y of C a l i f o r n i a at Berkeley, f o r his able introduction into the psychology of learning i n general and the spread phenomenon in p a r t i c u l a r , and f o r i n i t i a t i n g t h i s pieoe of research; to Mr. Harold F. George, mathematics s p e c i a l i s t i n the Mission, B. C. School D i s t r i c t , whose in t e r e s t and c r i t i c a l a ppraisal during the development of the problem were of inestimable value; to Mr. 0. J. Thomas, Inspector of Elementary Schools, and to the P r i n c i p a l , Mr. E. V. Oaspell, and the s t a f f of Renfrew Elementary School, Vancouver, f o r th e i r w i l l i n g cooperation i n providing subjects; to Dr. J . R. Mcintosh, of the Eduoation Department at the University of B r i t i s h Columbia, f o r his encouragement and thoughtful c r i t i c i s m throughout; and to Dr. Maxwell A . Cameron, Head of the Education Department, U n i v e r s i t y of B r i t i s h Columbia, whose great kindness and consideration have been n e v e r - f a i l i n g . ( i i ) Table of Contents Page Chapter I Introduction — The Problem — I t s Background and Development. 1 Chapter I I The Experiment -- Subjects, Material, Procedure, Calculations. 5 Chapter I I I Obtained Data — Analysis, . Findings, Interpretations. 13 Chapter TV Summary, Conclusions, Hypotheses. 29 Bibliography 33 Appendix 35 Figures 1. Gradients of E f f e c t Around Rewarded Responses 15 Tables 1. Composition of Experimental Groups 5 2. C l a s s i f i c a t i o n of S i g n i f i c a n t Responses 13 3. Percentages of Repetitions i n S i g n i f i c a n t Categories 14 4. Group Differences and t h e i r T-Ratios 16 5. Differences Between Obtained Base-lines and the Theoretical Chance Base-line 20 6. Comparison of Gradient Figures Obtained by T i l t o n and i n the Present Study 22 7. Gradient Figures from Five Studies Showing Var i a t i o n i n Gradient Slopes 25 8. Differences Between Gradient Figures and Obtained Base-lines 27 The Relation of Intelligence  and the Spread of Effect Chapter I Introduction — The Problem — Its Background and Development  Baokground The purpose of this study was to discover whether there i s any relation between the degree of measured i n t e l -ligence and the spread of effect as demonstrated f i r s t by E. L. Thorndike 1 and later by other investigators into the psychology of learning. Thorndike*s "The Fundamentals of Learning," which was published in 1932, contained among other revisions of his original theory of learning, a re-formulated law of effeot which minimized the direot weakening influence of punishment. In the following year he published experimental evidence to demonstrate that a reward strengthens not only the connection to which i t belongs, but also neighboring oonneotions both preceding and following the rewarded connection, such strengthening gradually diminishing as the distance from the rewarded connection increases. It 1. Thorndike, E. L. "A Proof of the Law of Effeot," Science. 1933, 77, PP. 173-175 "An Experimental Study of. Rewards," Teachers College, Columbia Univ., Contr. Educ, No. 580. New York: Bureau of Publications, Teachers College, Columbia Univ., 1933. - 2 -is this phenomenon to whioh Thorndike applied the term r _ 2 "spread (_of e f f e c t j , " and whioh has become the primary a r t i c l e of f a i t h in the Thorndike theory. 3 In 1947 Postman published an artiole in whioh he reviewed the development and the present position of the law of effect. The results of major significant studies were considered and questions not yet satisfactorily answered were put forward. Among these was one whioh arose in connection 4 with the findings of Muenzinger and Dove, that responses surrounding a rewarded response tend to be repeated while those around a punished response tend to be varied. Postman here makes this comment: "This analysis s t i l l leaves open the question as to what intervening mechanism the changes in va r i a b i l i t y should be ascribed." 5 Thorndike himself states that ". . . the extent of spread may vary with the kinds of learning and the learners." 2. Thorndike, E. L., passim 3. Postman, L.,"The History and Present Status of the Law of Effect, Psychol. B u l l . t 1947, 44, PP. 439-563. 4. Muenzinger, K. F., and Dove, 0. C., "Serial Learning: I. Gradients of Uniformity and Variability Produced by Suooess and Failure of Single Responses," J. Gen. Psychol.. 1937, 16, pp. 403-413. 5. Postman, op. o i t . , p. 518. 6. Thorndike, E. L., Oontr. Educ, No. 580, p. 54. I n t h i s regard, Wallach and Henle have demonstrated w i t h c o l l e g e students t h a t the "Thorndike e f f e c t " does not appear when mot i v a t i o n to l e a r n i s l a c k i n g . I t appears, then, t h a t the spread o f e f f e c t may be subject t o c e r t a i n v a r i a b l e f a c t o r s of which m o t i v a t i o n may be one. C o n s i d e r a t i o n of a mechanistic theory of l e a r n i n g such as Thorndike*s leads to the question of c o g n i t i v e behaviour i n r e l a t i o n to behaviour a u t o m a t i c a l l y determined by the formation of bonds or connections w i t h i n the organism. Whether we accept Sandiford's i n t e r p r e t a t i o n t h a t these connections " . . . presumably have t h e i r p h y s i c a l b a s i s i n the nervous system, where the connections between neuron 8 and neuron e x p l a i n l e a r n i n g " , o r whether w i t h Gates we regard a connection s i m p l y as ". . . a f u n c t i o n a l r e l a t i o n between a s i t u a t i o n and a response . . . . which i m p l i e s no Q n e u r o l o g i c a l c o r r e l a t e " 7 , the automatic and meohanioal element remains and gives r i s e to s p e c u l a t i o n concerning the p o s s i b l e operation of c o g n i t i v e f a o t o r s i n the l e a r n i n g process. This s p e c u l a t i o n , coupled p a r t i c u l a r l y w i t h Thorndike*s statement regarding v a r i a t i o n i n extent of spread, and the Wallaoh and Henle study on m o t i v a t i o n and reward, l e d to the f o r m u l a t i o n of the present problem — i s i n t e l l i g e n c e a v a r i a b l e f a c t o r i n r e l a t i o n to the spread of e f f e c t ? 7. Wallaoh, H. and Henle, Mary, "An Experimental A n a l y s i s of the Law of E f f e c t , J . Exp. Psychol..1941. 28, pp. 340 - 349. * "A" F u r t h e r Study of the Function of Reward,"J. Exp. Psychol.,1942, 30, pp. 147 - 160. 8. Sandiford, P., "Conneetionism: I t s O r i g i n and Major . Features," F o r t y - f i r s t Yearbook, Nat. Soc. Study E d u c , P a r t I I , Bloomington, P u b l i c School P u b l i s h i n g Co., 1942, p. 98. 9. Gates, A. I . , "Conneetionism: Present Concepts and I n t e r -p r e t a t i o n s , F o r t y - f i r s t Yearbook. Nat. Soc. Study Educ., P a r t I I , Bloomington, P u b l i c School P u b l i s h i n g Co., 1942, p. 145. - 4 -In reviewing the l i t e r a t u r e of the f i e l d , one i s s t r u c k r a t h e r f o r c i b l y by the f a c t t h a t i n the s t u d i e s so f a r reported, w i t h one exception where students from Grades 6, „ 10 7, and 8 were used, the subjects have been e i t h e r c o l l e g e students o r r a t s . This observation suggests the p r a c t i c a l need, p a r t i c u l a r l y from the viewpoint of general e d u c a t i o n a l a p p l i c a t i o n , of i n v e s t i g a t i o n i n the one instance of a l e s s s e l e c t , more r e p r e s e n t a t i v e group, and i n the other, of subjects more n e a r l y r e l a t e d i n kind to the t y p i c a l school p o p u l a t i o n . A c c o r d i n g l y , i t was decided to use i n t h i s i n v e s t i g a t i o n s u b j e c t s of a grade l e v e l as r e p r e s e n t a t i v e as p o s s i b l e of the average elementary school p u p i l . Since i n t e l l i g e n c e was to be the s i g n i f i c a n t v a r i a b l e , a comparative study was planned around two groups, one composed of p u p i l s who obtained h i g h scores on a standard i n t e l l i g e n c e t e s t , the o t h e r of p u p i l s whose scores on the same t e s t were r e l a t i v e l y low. The s p e c i f i c o b j e c t of the i n v e s t i g a t i o n was to determine whether the spread of e f f e c t p a t t e r n obtained from a group of subjects of high I.Q, i s s i m i l a r to that manifested by a group of subjects whose I.Q. fs are r e l a t i v e l y low. A n u l l hypothesis was set up, to the e f f e c t t h a t the spread patterns obtained would not d i f f e r s i g n i f i o a n t l y from each other. 10. Rook, R. T., J r . , "The Influence Upoa Learning of the  q u a n t i t a t i v e V a r i a t i o n ~of A f t e r - E f f e c t s , " Teachers Coll e g e , Columbia Univ., Contr. E d u c , No. 650, New.York: Bureau of P u b l i c a t i o n s , Teachers Coll e g e , Columbia Univ. , 1935. - 5 -Chapter II The Experiment — Subjects. Material, Procedure, Calculations Subjects Data were collected from two groups, each composed of thirty pupils from Grades V, 71, and 711', at Renfrew Elementary School in 7anoouver, B. C. The school is located in what might be called an average residential d i s t r i c t . Rele-vant details concerning each group are presented in Table I. For convenience, the group with the higher I.Q. rating w i l l be identified as the high group, and the group with the lower I.Q. rating as the low group. Table I COMPOSITION OF EXPERIMENTAL GROUPS Grade Plabement M Sex I. ft. Age (in years) V 71 f 711 M I F Range jMean Range Mean High Group 30 11 j 8 1 11I12 18 |20-1541127.91 10.4-13.4 11.96 Low Group 30 15 8 I 7118 |12 f 73-94 I 87.31 10.8-14.41 12.71 I.Q,. ratings were obtained by means of the National Intelligence Test administered by personnel from the Testing Bureau of the city school system. The groups were differentiated by 40.6 § In previous studies reviewed N has varied from twenty to . seventy-five. - 6 -mean I.Q,. points, and by 26 points separating the lowest rating i n the high group from the highest i n the low group. The differences apparent i n chronological age were considered not large enough to a f f e c t m a t e r i a l l y the obtained data. Material The material used was of the conventional type employed i n many experiments designed to t e s t the spread of eff e c t phenomenon, that i s , a l i s t of word-stimuli, each word requiring a response i n the form of a number from within a designated range. In t h i s case, the word l i s t consisted of 40 simple one-syllable verbs, chosen on the basis of about equal f a m i l i a r i t y to a l l the subjects (Appendix). The range of numbers from which responses were to be made was one to s i x . 1  Procedure In the t y p i c a l method of procedure, the word-stimuli are presented i n f i x e d order and at a spec i f i e d rate of speed to each subjeot. Rewards i n the form of the announcement "ri g h t " , and punishments i n the form of the announcement "wrong" are administered by the experimenter according to a 1 In previous studies of s i m i l a r nature which have been examined the length of the word-series has varied from twenty-five to eighty words, and the range of responses from one to f i v e to one to ten. 7 -key or pl a n by which r i g h t and wrong responses are pre-determined. The l i s t i s presented to each subject i n a number of continuous successive t r i a l s and a record i s kept, by the examiner, of each response to each stimulus word (Ap-pendix). This method was f o l l o w e d except f o r one m o d i f i c a t i o n . For the purpose of t h i s study i t was decided t h a t the e f f e c t of the announcement "wrong" was not of p a r t i c u l a r moment. Consequently t<he procedure was modified i n t h i s r espect, and o n l y reward i n the form o f the announcement " r i g h t " was used i n the p r e s e n t a t i o n of the t e s t m a t e r i a l . The t o t a l number o f t r i a l s f o r each subjeot was seven. During the f i r s t two of these, no rewards were given. Much d i s c u s s i o n of the evidenoe supporting the law and the spread of e f f e c t has centred around the determination of a ba s e - l i n e of chance expectation from which to evaluate the a c t u a l percentages of r e p e t i t i o n s obtained. I t i s apparent that i n the s o r t of " f r e e a s s o c i a t i o n " technique employed i n experiments of t h i s k i n d , some favored word-number combinations are extremely l i k e l y to occur and recu r . T i l t o n has stated the problem c l e a r l y when he says t h a t what i s needed " . . . i s the r e p e t i t i o n which o b t a i n s before the 2 •rewards* and •punishments* to be studied are a p p l i e d . " From t h i s premise he p l o t t e d a curve of r e p e t i t i o n s f o r 360 observations a t each s e r i a l p o i n t i n h i s experimental m a t e r i a l , based on the responses o f su b j e c t s before they had been informed as to r i g h t and wrong responses. The curve proved throughout most of i t s l e n g t h to be s t r a i g h t and 2 . T i l t o n , J . W., "Gradients of E f f e c t , " J . Genet. Psyohol.. 1 9 4 5 , 6 6 , p. 8 . • h o r i z o n t a l i n nature and was adopted as the zero o r b a s e - l i n e f o r h i s experimental data. I n c a l c u l a t i n g e f f e c t g r a d i e n t s , a c o r r e c t i o n was made f o r s e r i a l p o s i t i o n e f f e c t s which ap-peared i n the b a s e - l i n e data, by o m i t t i n g from the gradient data items l o c a t e d a t the top and bottom of the two pages of l e a r n i n g m a t e r i a l , and whioh showed an extremely high percentage of r e p e t i t i o n . At the same time T i l t o n found t h a t f o r such data a s i m i l a r b a s e - l i n e was a r r i v e d a t simply by computing the percentage of t o t a l r e p e t i t i o n s , r i g h t and wrong, during the course of the experiment. He p o i n t s out that i n h i s m a t e r i a l the number of r i g h t s was about equal to the number of wrongs, and t h a t where t h i s i s aot the case, t h e r e s u l t s f o r r i g h t s •and wrongs would have to be averaged w i t h equal weights before a s i m i l a r curve could be obtained by t h i s method. 3 Martens estimated a b a s e - l i n e by using a c o n t r o l group. The experimental m a t e r i a l was presented to the c o n t r o l group i n two successive t r i a l s without reward o r punishment, and the percentage of r e p e t i t i o n s which r e s u l t e d was used as the chance b a s e - l i n e . The number of s u b j e c t s i n the Martens study has not been reported, so t h a t some doubt a r i s e s as to the v a l i d i t y of such a technique unless the numbers are very l a r g e . For small numbers i t would seem p r e f e r a b l e to have experimental groups serve as t h e i r own c o n t r o l s , 3. Martens, D., "Spread of E f f e c t i n V e r b a l S e r i a l Learning," (Abstract) Amer.-Psychologist. 1946, 1, 448-449. - 9 -with cognizance taken of the fact that there may be strengthening of a response due simply to its occurrence. In the present study, the Tllton and Martens procedures were combined and modified to obtain a neutral base-line. The experimental groups served as their own controls in the determination of favored responses. The f i r s t two trials of the word series were run through without reward and repetitions were noted as probable favored responses. Only on rare oooasions were such responses rewarded throughout the remaining five t r ials . With six possible responses to eaoh stimulus, and seven presentations of the word-list, two occurrences of the same number-response to a single word-stimulus oould be regarded as due to ohanoe. More than two occurrences without the influence of reward could be considered therefore as favored responses. Thorndike has remarked that the spread appears to stop at some point between the third and sixth step positions removed.^ With this in mind, i t had been decided that gradients plotted to the third step would serve the purpose of the present study. When the problem of determining favored responses arose, the third step was again selected as the point beyond which reward is unlikely to show marked effeot. Favored responses, then, were identified as those responses which occurred three times or more before reward oould be considered to have influenced them, and were omitted entirely 4 . Thorndike, E. L . , Contr. Educ, No. 5 8 0 , p. 5 4 . 10 -from a l l c a l c u l a t i o n s . When a l l the data had been c o l l e c t e d , the percentage 5 of t o t a l r e p e t i t i o n s of r i g h t s and wrongs combined, exoluding the favored responses, was c a l c u l a t e d f o r each group and used as the n e u t r a l b a s e - l i n e . T i l t o n 1 s p r o v i s o concerning equal average we i g h t i n g where the number of r i g h t s I s not about equal to the number of wrongs was not considered to hold i n t h i s case where o n l y the e f f e c t of reward i s under o b s e r v a t i o n . However, i t should be recognized here t h a t the b a s e - l i n e so obtained may i t s e l f have been a f f e c t e d by the i n f l u e n c e of reward, w i t h the r e s u l t that the l i n e obtained i s a o t u a l l y higher than pure chance would have d i c t a t e d . This p o i n t w i l l be discussed more f u l l y In the f o l l o w i n g chapter. The s u b j e c t was seated about f o u r f e e t from the examiner*s desk and i n a p o s i t i o n p a r a l l e l to the desk. He was given the f o l l o w i n g d i r e c t i o n s : What you are going to do f o r me now has nothing to do w i t h your school work and won't make any d i f f e r e n c e to your grades or marks. I'm doing an experiment w i t h boys and g i r l s of your age and Mr. 0. £the school p r i n o i p a l j says t h a t you w i l l be a good one to help me. This i s what we are to do: I have a l i s t of words which I'm going to read to you, 5. I t w i l l be r e o a l l e d t h a t the e f f e o t of punishment i s not being i n v e s t i g a t e d , and throughout the experiment no announcements of "wrong" were made. For convenience, responses not c a l l e d r i g h t by v i r t u e of reward are termed wrong, but no announcement of "wrong" accompanied them. - 1 1 -one a f t e r the other. Each word has a number from one to s i x belonging to i t , and you are to guess t h a t number. We'll p r a c t i s e a b i t f i r s t , I reading a word and you g i v i n g a number from one to s i x , and then when you have the i d e a , I ' l l s t a r t t e l l i n g you when you have guessed the r i g h t number. I ' l l simply say " r i g h t " and w e ' l l go s t r a i g h t on to the next word. Remember, any number from one to s i x . Always say the f i r s t number t h a t comes i n t o your head and don't t r y to f o l l o w a p a t t e r n of numbers l i k e saying them i n a c e r t a i n order. Ready? The words were read a t a rate of one -about every 2% to 3 seconds, and the p r e s e n t a t i o n was made i n as automatic and stereotyped a manner as p o s s i b l e . Rewards were assigned a r b i t r a r i l y and o n l y r a r e l y to what seemed to be a favored response. I t had been Intended o r i g i n a l l y to keep rewards separated by a t l e a s t seven o r e i g h t unrewarded responses i n an e f f o r t to o b v i a t e i n t e r f e r e n c e of backward and forward g r a d i e n t s . With subjects as young as these, and as eager to guess the r i g h t number, i t was found • that a g r e a t e r number of rewards was necessary to maintain i n t e r e s t and r e l i e v e the boredom and apparent f a t i g u e induced by the monotony of the procedure. During the t h i r d t r i a l , which marked the beginning of the announcement of " r i g h t , " rewards were 'separated v a r y i n g l y by s i x , seven, and e i g h t unrewarded responses. Thereafter, r e p e t i t i o n s of responses from the immediately preceding t r i a l were rewarded; and a d d i t i o n a l rewards were given as necessary to insure s i x o r seven rewards per t r i a l . There was thus some "crowding" of - 12 -r i g h t responses, but since the experiment was not designed w i t h the purpose of determining to what extent the backward gradient i s a f u n c t i o n of a preceding forward one, complete c o n t r o l of t h i s f a c t o r was not deemed imperative to the t e s t i n g of the hypothesis which had been set up. C a l c u l a t i o n of Responses R e p e t i t i o n s were i d e n t i f i e d as those responses whioh d u p l i c a t e d responses made i n the Immediately preceding presentation-of the s e r i e s , always exc l u d i n g favored responses. -Three step p o s i t i o n s before and a f t e r a rewarded response were thought to be adequate f o r the determination of spread p a t t e r n s , and r e p e t i t i o n s beyond these were not c a l c u l a t e d i n the g r a d i e n t s . Where crowding of r i g h t responses was evident, a r e p e t i t i o n of a wrong response was o r e d i t e d to the nearest rewarded response which i t preceded or f o l l o w e d . R e p e t i t i o n s i n an e x a c t l y midway p o s i t i o n between two rewarded responses were omitted from the c a l c u l a t i o n of g r a d i e n t s . To the extent that the backward gradient may be a f u n c t i o n of a l a r g e r forward one, the backward g r a d i e n t s obtained from t h i s data w i l l be spurious, but the o r i g i n a l purpose of the study i s not thereby a f f e c t e d . - 13 -Chapter I I I Obtained Data — Analys i s , Findings, Interpretations Tabulation of Data The record of each subject was analyzed and responses f o r the f i v e t r i a l s during whioh rewards were given were c l a s s i f i e d as i l l u s t r a t e d In Table I I , which contains composite figures f o r each group. Table I I CLASSIFICATION OF SIGNIFICANT RESPONSES Number of Responses Categories of Responses High Group Low Group Total possible j 6000 6000 X _ Total calculated 7^ 5725 1 5764 Y Total r e p e t i t i o n s 850 | 921 A mm Total rewarded responses 783 835 B Repetitions of rewarded responses 162 201 C Repetitions one step before 87 113 D two steps before 103 96 E mm three steps before 54 48 F one step a f t e r 103 127 G two steps a f t e r 104 95 H mm three steps a f t e r 56 5^ # i . e . omitting favored responses. - 14 -F o l l o w i n g t h i s b a s i c t a b u l a t i o n , percentages of r e p e t i t i o n i n each oategory to be examined were computed to ob t a i n comparable f i g u r e s . These are reported i n Table I I I , F i g u r e 1 i s a graphic d e s c r i p t i o n of the same m a t e r i a l , and presents f o r both groups the grad i e n t s of e f f e o t preceding and f o l l o w i n g rewarded responses. Table I I I PERCENTAGES OF REPETITIONS IN SIGNIFICANT CATEGORIES CATEGORIES OF REPETITIONS Percentages of R e p e t i t i o n s High Group Low Group y T o t a l r e p e t i t i o n s : T o t a l ji v 2 - - responses' 14.85 15.98 E/A - Three steps before 6.89 5.75 D/A - Two steps before 13.15 11.49 C/A - One .step-before 11.11 13.53 B/A - REWARDED RESP0NSE3 20.69 24.07 F/A - One step a f t e r 13.15 15.21 G/A - Two steps a f t e r 13.28 11.38 H/A - Three steps a f t e r 7.15 6.47 # These r a t i o s represent the chance b a s e - l i n e s used. - 1 5 -5 o I ' ' 1 » } i E/A D / A 0/A B / A F / A G/A H / A F i g u r e 1. Gradients of E f f e c t Around Rewarded Responses A n a l y s i s and F i n d i n g s The percentage of r e p e t i t i o n s of rewarded responses f o r both groups i s s i g n i f i c a n t l y b e t t e r than chance. Using the Y/X r a t i o as the chance b a s e - l i n e , t - r a t i o s of 3 . 8 and 5.2 were found f o r the high and low groups r e s p e c t i v e l y . 1 1. The formulae, '** *Pg > < ^ , and t • were used throughout i n determining the s i g n i f i c a n c e of d i f f e r e n c e s . - 16 -The s i g n i f i c a n t values of t a t the 5% and 1% l e v e l s are 1.96 and 2.576. This f i n d i n g s u b s t a n t i a t e s the strengthening e f f e c t of a reward on the stimulus-response connection to which i t belongs. The t - t e s t was ap p l i e d to the d i f f e r e n c e s between the groups i n eaoh response category and none of these i s s t a t i s t i c a l l y s i g n i f i c a n t . The l a r g e s t group d i f f e r e n c e , 3.38$ between percentages of r e p e t i t i o n of rewarded responses, has a t - r a t i o of 1.63. D i f f e r e n c e s and t h e i r t - r a t i o s f o r each category are presented i n Table IV. Table IV GROUP DIFFERENCES AND THEIR T-RATIOS .CATEGORIES DIFFERENCES T-RATIOS SIGNIFICANT T VALUES 5% l e v e l V/o l e v e l T/X 1.13 1.69 1.96 2.576 E/A 1.14 # 1.94 1.96 2.576 D/A 1.66 # 1.01 1.96 2.576 C/A 2.42 1.48 1.96 2.576 B/A 3.38 1.63 1.96 2.576 F/A 2.06 1.19 1.96 2.576 G/A 1.90^ 1.16 1.96 2.576 H/A .68 # .54 1.96 2.576 § These d i f f e r e n c e s are i n f a v o r of the high group. - 17 -Interpretations The similarity, both apparent and s t a t i s t i c a l , of the effect gradients of both groups supports the hypothesis that the spread patterns obtained from a group of bright subjects and from a group of d u l l subjects would not d i f f e r significantly from eaoh other, and leads to the conclusion that, within the parameters of this study, intelligence as measured by a standard test is not a variable factor in relation to the spread of effeot. The larger over-all percentage of repetitions in the low group, though not reliable, suggests that there.may be a greater tendency towards stereotypy among d u l l students than there i s among the bright. This observation seems to be in acoord with the reoognized use of d r i l l as a teaching method for d u l l students. However, extreme individual differences apparent in spread patterns and in repetitions of rewarded responses among subjects of both groups make impossible any broad generalizations along this line, and point up Thorndike*s acknowledgement of the need for "an enormous body of data" in connection with the possible 2 * variation i n spread. At this point, the data might be viewed in the light of an assumption at variance with the original hypothesis. It might be postulated thai; the more intelligent subjects would tend to a greater extent than the d u l l to vary their 2. Thorndike, Contr. Educ, No. 580, p. 54. - 1 8 -responses, particularly those responses which had previously not "been rewarded. The lower over-all percentage of repetition among the bright lends some support to such a thesis, but again qualifications as to s t a t i s t i c a l r e l i a b i l i t y and individual differences must be applied, and no conclusions oan be drawn. In addition, differences at the second and third steps both before and after a rewarded response are in favor, though not reliably, of the bright subjects. The slightly steeper nature of the gradients for the low group suggests that the effect of reward might have a wider spread among bright subjects. Further investigation into the questions here raised might prove of value. Supplementary Considerations Base-lines The matter of the determination of a neutral base-line deserves some additional comment. To the writer's knowledge, Stephens,-^ Martens, and T i l ton are the only investigators who have systematically attacked the problem. Although Stephens' assumption, viz., that the base-line i s indicated by the percentage of "weak untreated connections" persisting^ i s basioally similar to that adopted here., the nature of his investigation and the precise technique he employed to secure a base-line make comparison with the present study impractical. Likewise, sinoe no 3. Stephens, J. M., "Further Notes on Punishment and Reward," J. Genet. Psychol.. 1934, 44, PP. 464-472. 4. Ibid,, p. 469» - 19 -f i g u r e s are reported f o r Martens* experiment, the d i s c u s s i o n w i l l oentre around T i l ton's work. I t i s g e n e r a l l y accepted t h a t , i n l e a r n i n g m a t e r i a l of the kind employed here, the p r o b a b i l i t y of occurrence of a s i n g l e response i s not l/6 (where the denominator equals 1 the number of p o s s i b l e responses per s t i m u l u s ) , but 1/6 p l u s the f a c t o r of .'favoritism p l u s the f a c t o r of s e r i a l p o s i t i o n . Favored responses w i l l so vary among i n d i v i d u a l s t h a t , unless N approaches i n f i n i t y , to c o r r e c t o n l y f o r group p r e d i s p o s i t i o n s determined by s e r i a l p o s i t i o n as d i d T i l t o n , o r _ t o use a c o n t r o l group as d i d Martens, w i l l not produoe an accurate estimate of the tru e b a s e - l i n e p o s i t i o n . I n T i l t o n ' s study, the number of responses per stimulus was f o u r . Oould we accept pure chance as the determinant, the b a s e - l i n e f o r h i s data would f a l l a t the 25$ l e v e l . As a r e s u l t of the procedure he used to secure a b a s e - l i n e , i t f a l l s a t the 38% l e v e l . I n the l i g h t of what has been remarked concerning the p r o b a b i l i t y of occur-rence of a s i n g l e response, 13% of the area whioh hiSrbase-line represents must be a t t r i b u t e d to some f a c t o r o r f a c t o r s o t her than ohanoe. S e r i a l p o s i t i o n e f f e c t s as manifested by the group were corrected f o r , but i n d i v i d u a l p r e d i s p o s i t i o n s were not taken i n t o account. What appears needful i n the matter of b a s e - l i n e s and i n the c o l l e c t i o n of g r a d i e n t data i s the e l i m i n a t i o n of both s e r i a l p o s i t i o n e f f e c t s and favored responses. At t h i s 5. Stephens obtained 36% as the b a s e - l i n e f o r h i s data. Op. c i t . , p. 469. - 20 -juncture, attention is invited to the technique employed in this study to establish a neutral base-line, and i t s results. The material was so arranged and presented that s e r i a l position effeots would be negligible. Favored responses were deter-mined and eliminated from all.-calculations. When the collection of data was completed, the percentage of total repetitions, exoluding favored responses was computed to obtain the neutral base-line. For the high group, i t f a l l s at 14.85$; for the low group, at 15.98$. Pure ohanoe, with six possible responses as in this case, would yield l/6t or 16.66$. The similarity between the obtained base-lines and the chance figure is not a l i t t l e striking. Differences between the obtained figures and the pure chance figure, with their t-ratios, are shown in Table V. While the figures for the high group and for Table V DIFFERENCES BETWEEN OBTAINED BASE-LINES AND THE THEORETICAL CHANCE BASE-LINE CHANCE -OBTAINED DIF-FERENCE T-RATIO SIGNIFICAI S# level IT T-VALUES 1% level HIGH GROUP 16.66 • 14.85 1.81 2.701 1.96 2.576 LOW GROUP 16.66 15.98 .68 1.000 1.96- 2.576 GROUPS COMBINED 16.66 15.41 1.25 2.604 1,96 2.576 the high and low groups combined d i f f e r significantly from the pure chanoe percentage, the faot remains that the procedure employed has resulted in base-lines which more nearly approximate the theoretical ohanoe line than any previously reported. - 2 1 -As observed i n the preceding chapter, i n v o l v e d i n t h i s prooedure are the f a c t o r s of reward and occurrence o p e r a t i n g to increase the percentage of r e p e t i t i o n which represents the b a s e - l i n e . The i d e a l s i t u a t i o n has been remarked by Thorndike when he proposes f o r a b a s e - l i n e . . . "the fo of r e p e t i t i o n s there would have been i f rewards., punishments, and occurrences had had no strengthening o r weakening e f f e o t upon any connections."^ Rewards and punish-ments, i t i s t r u e , can be c o n t r o l l e d , but how data oan be c o l l e c t e d without occurrences and the e f f e c t s thereof i s somewhat p r o b l e m a t i c a l . T i l t o n approaches the desideratum i n h i s observation t h a t ". . . i n c e r t a i n cases . . . . c o l l e c t i o n of data f o r the b a s e - l i n e would need to p a r a l l e l the c o l l e c t i o n 7 f o r the g r a d i e n t s from t r i a l s as w e l l as from p o s i t i o n s . " I n the absence of any c o n c l u s i v e evidence r e g a r d i n g v a l i d b a s e - l i n e s , and i n view of those obtained here, the technique used i n t h i s study appears to m e r i t f u r t h e r experimentation. Q u a n t i t a t i v e A n a l y s i s Consequent to the c o n s i d e r a t i o n of n e u t r a l b a s e - l i n e s i s the question of the q u a n t i t a t i v e aspects of the g r a d i e n t s obtained i n t h i s study. I n general i t may be s t a t e d t h a t the percentages of r e p e t i t i o n obtained here are lower than those u s u a l l y reported from s i m i l a r experiments. S p e c i f i c comparison i s again l i m i t e d to the work of T i l t o n because of the c o r r e c t i v e 6 . Thorndike, op o i t . , p. 5 0 . 7 . T i l t o n , J . W ., "Gradients of E f f e o t , " J . Genet. P s y c h o l . . 1 9 4 5 , 6 6 , p. 9 ( f o o t n o t e ) . - 22 -f a c t o r s used i n assembling the da t a . Gradient f i g u r e s from T i l t o n and from an average of the two groups of t h i s study are presented i n Table VI. As T i l t o n has reported o n l y to two step p o s i t i o n s , the t h i r d step has been omitted. Table VI COMPARISON OF GRADIENT FIGURES OBTAINED BY . TILTON AND IN - THE PRESENT STUDY - . Two Steps Before One Step Before Rewarded Responses One Step A f t e r Two Steps A f t e r TILTON 29.3 30.2 41,35 # 35.1 32.4 PRESENT STUDY 12.32 12.32 22.38 14.07 12.33 PRESENT STUDY AS $ OF TILTON .42 .41 .54 .40 .38 With the exception of the rewarded responses, the percentage of r e p e t i t i o n a t each p o s i t i o n i n t h i s study i s l e s s than h a l f of tha t reported by T i l t o n . How are these c o n s i s t e n t d i f f e r e n c e s i n degree to be acoounted f o r ? I n the mind of the w r i t e r there are two p o s s i b l e i n t e r p r e t a t i o n s . F i r s t , they may be a m a n i f e s t a t i o n of the v a r i a t i o n i n extent of spread which Thorndike suggested might occur r e l a t i v e to the kind o f l e a r n i n g and the l e a r n e r s . I f so, i t w i l l be noted t h a t not o n l y the spread has v a r i e d , but so has the r e p e t i t i o n of rewarded responses themselves'. I t w i l l be r e o a l l e d t h a t the mean age of the su b j e c t s was approximately 12 years, whereas i n previous s t u d i e s subjeots have been o o l l e g e students. The l e a r n i n g m a t e r i a l here oon-# T i l t o n reported h i s backward gradient from one composite rewarded response f i g u r e (40.9$), and h i s forward g r a d i e n t from another (41.8$). I b i d . , p. 12. These were averaged to o b t a i n a s i n g l e f i g u r e f o r ease of comparison. - 23 -s i s t e d of a l i s t of f o r t y words, a l e n g t h commonly used w i t h c o l l e g e students. I s i t p o s s i b l e that the a c t i o n of reward on l e a r n i n g m a t e r i a l of t h i s nature i s i n some way r e l a t e d to the l e n g t h of the s e r i e s and the age or m a t u r a t i o n a l l e v e l of the subjeots? I s the strengthening e f f e o t of reward d i s s i p a t e d i n c e r t a i n s u b j e c t s by the i n t e r f e r e n c e of i n t e r v e n i n g stimulus-response connections beyond a c e r t a i n number? I s the time between successive p r e s e n t a t i o n s of a s i n g l e stimulus a f a c t o r i n determining the i n f l u e n c e of i t s reward? These questions would seem to be of v i t a l importance, both t h e o r e t i c a l l y 8 and p r a c t i c a l l y , and present problems f o r f u r t h e r study. The second i n t e r p r e t a t i o n d e r i v e s from the d i s c u s s i o n of l e v e l s of chance expectation and the se c u r i n g of a v a l i d b a s e - l i n e . To the w r i t e r ' s knowledge, no previous attempt has been made, w i t h the p o s s i b l e exception of the Martens study, to c o n t r o l as f u l l y as p o s s i b l e the f a c t o r of favored responses i n r e l a t i o n to an o p e r a t i o n a l l y determined b a s e - l i n e and the c o l l e c t i o n of gr a d i e n t data. The data of t h i s JLnvestigation do not lend themselves to examination i n order to d i s c o v e r how the g r a d i e n t s would have appeared had favored responses been 8. Stephens has already presented evidence which suggests that the l e n g t h of the t e s t may indeed a f f e c t the number of r e p e t i t i o n s , and has concluded t h a t n. . . the longer the t e s t the l e s s l i k e l i h o o d of p e r s i s t e n c e Cof weak untreated connections}," J . Genet. Ps y c h o l . . 1934, 44, p. 470. -- 24 -rewarded and had not been eliminated from a l l calculations. However, It i s submitted that the lower percentages-of repetition obtained from these data may be attributed to the elimination of favored responses with a consequent reduction throughout of the levels of repetition and an apparent mini-mizing of the influence of reward. This interpretation too requires further investigation. Role of Punishment Out of the preceding discussion there arises the interesting question as to what effect the omission of punish-ment in the form of the announcement "wrong" may have had on the repetition of unrewarded responses. Ho s t a t i s t i c a l analyses have been made of the gradients obtained here as compared to those reported in other similar studies, but the steepness of these gradients favors the tentative suggestion that the announcement "wrong" in learning material of this kind has a questionable effect on the repetition of wrong 9 responses. Gradients obtained by Muenzinger and Dove are 10 similar in slope; Tilton*s are markedly more gradual; 11 -Zirkle has reported one similar set and one markedly steeper. 9. Muenzinger, K. F., and Dove, 0. 0., "Serial Learning: I. Gradients of Uniformity and Variability-Produced by Success and Failure of Single Responses," J. Gen. Psychol.. 1937, 16, p. 10. Tilton, op. o i t . , p. 12. 11. Zirkle, G. A., "Success and Failure in Serial Learning. I. The Thorndike Effect," J. Exp. Psychol., 1946, 36, p. 232. - 25 -For convenience of comparison, gradient f i g u r e s f o r these s e v e r a l s t u d i e s are presented i n Table VTI. Table V I I GRADIENT FIGURES FROM FIVE STUDIES SHOWING VARIATION IN GRADIENT SLOPES PRESENT^ STUDY MUENZINGER AND DOVE TILTON ZIRKLE (1) ZIRKLE (2) Three Steps Before 6.31 18.3 . 13.8 13.6 Two Steps Before 12.32 20.9 29.3 14.0 11.8 One Step Before 12.32 22.1 30.2 16.1 11.6 Rewarded Responses 22.38 51.4 41.35 42.5 26.1 One Step A f t e r 14.07 22.9 35.1 21.1 19.5 Two Steps A f t e r 12.33 19.5 32.4 14.9 16.5 Three \ Steps A f t e r | 6.74 17.7 13.6 12.5 # Composite f i g u r e s from g r a d i e n t s of two groups combined. These instances of v a r i a b i l i t y i n obtained g r a d i e n t s seem to i n d i c a t e that the r o l e of punishment a t l a r g e and i n general i n r e l a t i o n to a l e a r n i n g s i t u a t i o n i s d i f f i c u l t i f not r impossible t o determine on the b a s i s of experimental evidence discovered through the medium of the announcement "wrong." I t i s r e a l i z e d t h a t t h i s i s not a new ob s e r v a t i o n , but the co n d i t i o n s of the present experiment w i t h respeot to the omission of s p e c i f i c i n f o r m a t i o n regarding wrong responses seemed to the w r i t e r to r e q u i r e some comment. The r e s u l t s of - 26 -t h i s study when viewed i n r e l a t i o n to other experiments serve to point up Postman's statement that "The oomplex ways i n whioh the effects of 'wrong' vary with the parameters of the experimental s i t u a t i o n highlight the need f o r caution i n 12 generalizing about the effects o f punishment." A further question a r i s e s in t h i s connection. In spite of the f a c t that no punishment was given to wrong responses throughout this experiment, r e p e t i t i o n of unrewarded responses i n the second and third steps before and a f t e r reward i n the low group, and i n the f i r s t and t h i r d steps before and t h i r d steps a f t e r reward i n the high group, are r e l i a b l y lower than would have been expected having regard f o r the neutral base-line. Table VIII shows the difference between r e p e t i t i o n s at each step p o s i t i o n and the obtained l e v e l of chance expectation, together with t - r a t i o s . S i g n i f i c a n t values of t i n each case are 1.96 at the 5$ l e v e l and 2.576 at the 1% l e v e l . 12. Postman, L., "The History and Present Status of the Law of E f f e c t , " Psychol. B u l l . . 1947, 44, p. 504. - 27 -Table V I I I DIFFERENCES BETWEEN GRADIENT FIGURES AND OBTAINED BASE-LINES . HIGH LOW GROUP GROUP Three steps before - d l f f . 7.96 10.23 - t 7.80 10.88 Two steps before - d i f f . 1.70 4.49 - t -1.32 3.74 One step before - d i f f . 3.74 2.45 - t 3.09 1.91 One step a f t e r - d i f f . 1.70 .77 - t 1.32 .57 Two steps a f t e r - d i f f . 1.57 4.60 - t 1.21 3.83 Three steps a f t e r - d i f f . 7.70 9.51 - t 7.47 9.70 E x p l a n a t i o n of t h i s seeming anomaly i s d i f f i o u l t . I n experiments where punishment was administered, such a drop i n r e p e t i t i o n s oould be a t t r i b u t e d to i t s e f f e c t . I n t h i s case the oause must be sought elsewhere. The hypothesis i s o f f e r e d t h a t w h i l e a reward strengthens the stimulus-response connection to which i t belongs, i t a c t s i n the nature of a d i s t r a c t i o n on neighboring connections and thereby reduces t h e i r r a t e of r e p e t i t i o n below what could be expected i f there were no d i s t r a c t i o n o r i n t e r f e r e n c e i n t h e i r p r o x i m i t y . This i n t e r p r e t a t i o n i s a t vari a n c e w i t h the Jenkins and - 28 -13 Sheffield hypothesis that distraction is l i k e l y to bring about an increase in the level of repetition of errors remote from reward because under distraction subjects w i l l be less l i k e l y to remember and therefore vary wrong responses in subsequent t r i a l s . However, accepting for the moment a oonneotlonist theory of learning, the distraction hypothesis is tenable on the grounds that the word-stimuli were presented too quickly during the present study to permit any rehearsal of right or wrong responses. Whatever the interpretation, i t seems that further study needs to be made of the effects of both reward and punishment on the learning process. 13. Jenkins, W. 0., and Sheffield, F. D., "Rehearsal and Guessing Habits as Sources of the 'spread of effect'," J. Exp. Psychol., 1946, 36, pp. 316-330. - . - 2 9 -Chapter 17 Summary, Conclusions, Hypotheses Summary The problem i n v e s t i g a t e d was the r e l a t i o n of i n t e l l i g e n c e t o the spread of e f f e c t . A n u l l hypothesis was set up, that the spread p a t t e r n obtained from more i n t e l l i g e n t s u b j e c t s would not d i f f e r from t h a t found i n the l e s s i n t e l l i g e n t . Data were c o l l e c t e d from two groups of t h i r t y s u b j e c t s eaoh, one odmposed of " b r i g h t " and the other of " d u l l " students. A l l were p u p i l s i n Grades 7, 71, and 711 i n the same school. The m a t e r i a l was of the conventional type (word-stimulus, number-response) used i n many"effeot" experiments, but the t y p i o a l prooedure of rewarding c o r r e c t responses w i t h the announcement " r i g h t " and punishing wrong responses w i t h the announcement "wrong" was modified by the omission of the announcement "wrong" d u r i n g the course of the experiment. S e r i a l p o s i t i o n e f f e c t s were obviated by making successive presentations of the l i s t of stimulus words continuous, and by the l e n g t h of the l i s t . Favored responses were determined w i t h the help of two pre s e n t a t i o n s f r e e from reward a t the beginning of the experiment, and were e l i m i n a t e d from a l l c a l c u l a t i o n s i n order t o e s t a b l i s h a n e u t r a l base-l i n e , which was determined by computing the percentage of t o t a l r e p e t i t i o n s throughout the f i v e p r e s e n t a t i o n s 'during whioh rewards were giv e n . Gradients were p l o t t e d f o r each group from the percentages of r e p e t i t i o n of rewarded responses and of r e p e t i t i o n s one, two, and three steps before - 3 0 -and a f t e r rewarded responses. Conclusions The r e s u l t s are such t h a t the n u l l hypothesis must be accepted, that i s , i n t e l l i g e n c e as measured by a standard t e s t i s not a v a r i a b l e f a c t o r i n determining the spread of e f f e c t . Of the group d i f f e r e n c e s found, none i s s t a t i s t i c a l l y r e l i a b l e . I n so f a r as can be judged from a s i n g l e experiment and w i t h i n the parameters of t h a t experiment, i t i s concluded that reward has equal e f f e c t on b r i g h t and d u l l students i n a s e r i a l l e a r n i n g s i t u a t i o n . Hypotheses The r e l a t i v e l y low percentage l e v e l s of r e p e t i t i o n as compared w i t h those of previous s t u d i e s i s a t t r i b u t e d to one of two f a c t o r s , or p o s s i b l y to a combination of both. I n the f i r s t p l a c e , the method of assembling the data precluded favored responses from c o n t r i b u t i n g to the g r a d i e n t s obtained, and so reduced the number of r e p e t i t i o n s c a l c u l a t e d . I n the second, the w o r d - l i s t was of a l e n g t h commonly used w i t h subjects a t the c o l l e g e l e v e l . From these f a c t s , two t e n t a t i v e conclusions are reached. One, t h a t to the extent t h a t favored responses c o n t r i b u t e to g r a d i e n t data, l e v e l s of r e p e t i t i o n obtained i n a number of previous s t u d i e s and a t t r i b u t e d to the e f f e c t of reward, are s p u r i o u s l y high, and the i n f l u e n c e of reward has been exaggerated. Two, the l e n g t h of the t e s t ( i n t h i s case, the w o r d - l i s t ) , i s a f a c t o r deter-mining the i n f l u e n c e of reward. I t i s thought t h a t both of these c o n d i t i o n s may be r e s p o n s i b l e , to a degree so f a r - 31 -undetermined, f o r the r e s u l t s obtained i n t h i s study. S i m i l a r l y to the l e v e l s of r e p e t i t i o n found i n each response category, the height of the e s t a b l i s h e d b a s e - l i n e i s conspicuously lower than any p r e v i o u s l y adopted, and i s noteworthy i n i t s olose approximation to the pure chance f i g u r e . I t has been accounted f o r on the b a s i s of e l i m i n a t i o n of favored responses and s e r i a l p o s i t i o n e f f e o t s , and would a l s o have been a f f e c t e d by the f a c t o r of l e n g t h of the w o r d - l i s t i f t h i s were an experimental v a r i a b l e . A summary c o n s i d e r a t i o n of the slope of the gradients from these data compared w i t h g r a d i e n t curves from previous s t u d i e s where punishment i n the form of the announcement "wrong," as w e l l as reward, was administered to the s u b j e c t s , revealed no c o n s i s t e n t trends and added nothing c o n c l u s i v e by way of evidence on the i n f l u e n c e of punishment i n a l e a r n i n g s i t u a t i o n , other than to emphasize the apparently v a r y i n g r o l e s t h i s type of "punishment" can p l a y , and the i n a d v i s a b i l l t y of g e n e r a l i z i n g from the evidence thus f a r a v a i l a b l e on i t s modus operandi. In a d d i t i o n , the f a c t t h a t unrewarded responses, whioh were not punished, i n seven out of twelve c a t e g o r i e s were repeated l e s s f r e q u e n t l y than c o n s i d e r a t i o n of the n e u t r a l b a s e - l i n e s would have ledoae to expeot, r e q u i r e s explanation. I t has been hypothesized t h a t reward, i n emphasizing the oorreot response, aots as a d i s t r a c t i o n on neighboring connections and thereby reduces t h e i r r a t e of r e p e t i t i o n below the chance l e v e l . - 32 -Emergent Questions A c r i t i c a l review of the r e s u l t s of t h i s study d i s o l o s e s c e r t a i n areas i n which f u r t h e r researoh appears warranted: 1. The l a s t word has not been spoken i n regard to the i n f l u e n c e of i n t e l l i g e n c e on the spread of e f f e o t . I t i s true that no s t a t i s t i c a l l y s i g n i f i c a n t d i f f e r e n c e s between the b r i g h t and the d u l l were found, but the higher o v e r - a l l percentage of r e p e t i t i o n and the s l i g h t l y steeper g r a d i e n t s obtained from the d u l l group suggest that f u r t h e r s t u d i e s be o a r r i e d out where i n t e l l i g e n c e i s the experimental v a r i a b l e and where the spread i s measured to t h a t step p o s i t i o n , before and a f t e r a rewarded response, a t which the e f f e c t of reward i s no l o n g e r d i s c e r n i b l e . 2. The approximation of the obtained b a s e - l i n e to the l e v e l of pure chance, and the technique by which i t was e s t a b l i s h e d , r e q u i r e a d d i t i o n a l i n v e s t i g a t i o n . 3. The extent to which favored responses have accounted f o r r e p e t i t i o n s , w i t h the p o s s i b l e consequence of an exaggeration of the e f f e c t of reward, c a l l s f o r systematic i n q u i r y . 4. The f a l l i n g o f f of r e p e t i t i o n s of unrewarded responses to l e v e l s below the n e u t r a l b a s e - l i n e cannot i n t h i s case be a t t r i b u t e d to the e f f e o t of punishment, and i n d i c a t e s the need f o r f u r t h e r research i n t o the e f f e c t s of both reward and punishment on the l e a r n i n g p rocess. - 33 -B i b l i o g r a p h y Gates, A. I . "Connectionism: Present concepts and i n t e r -p r e t a t i o n s . " I n F o r t y - f i r s t Yearbook Nat. Soc. Study E d u c , Pt. I I . . Blooming ton: P u b l i c School P u b l i s h i n g Co., 1942. Jenkins, W. 0 . , & S h e f f i e l d , F. D. "Rehearsal and guessing h a b i t s as sources of the 'spread of e f f e c t ' . " J . Exp. Psych o l . , 1946, 36, pp. 316-330. - -Martens, D. "Spread of e f f e c t In v e r b a l s e r i a l l e a r n i n g . " (Abstract) Amer. P s y c h o l o g i s t . 1946, 1, pp. 448-449.-Muenzinger, K. F., & Dove, C. C. " S e r i a l l e a r n i n g : I . Gradients of u n i f o r m i t y and v a r i a b i l i t y - p r o d u c e d by success and f a i l u r e of s i n g l e responses." J . Gen. Psychol.. 1937, 16, pp. 403-413. - • Postman, L. "The h i s t o r y and present s t a t u s of the law of e f f e c t . ? . P s y c h o l . B u l l . , 1947, 44, PP. 489 - 563. Rock, R. T., J r . "The i n f l u e n c e upon l e a r n i n g of the q u a n t i - t a t i v e v a r i a t i o n of a f t e r - e f f e c t s . " Teach. C o l l . . Columbia Univ., Contr. E d u c , No. 650. New York: Bureau of P u b l i c a t i o n s , Teachers College, Columbia Univ., 1935. , • *J Sandiford, P. "Connectionism: I t s o r i g i n and major f e a t u r e s . " I n F o r t y - f i r s t Yearbook,Nat. Soc. Study E d u c , P t . I I . Bloomington: P u b l i c School P u b l i s h i n g Co., 1942. Stephens, J . M. "Further notes on punishment and reward." J . Genet. Psy c h o l . , 1934, 44, PP. 464-472. Thorndike, E. L. "The fundamentals of l e a r n i n g . " New York: Teachers Coll e g e , Columbia Univ., 1932. -' Thorndike, E. L. "A proof of the law o f e f f e c t . " Science. 1933, 77, PP. 173-175. -Thorndike, E. L. "An experimental study o f rewards." Teachers C o l l e g e , Columbia Univ., Contr. E d u c ,-No. 580. New York: Bureau of P u b l i c a t i o n s , Teachers College, Columbia Univ., 1933. T i l t o n , J . W. "Gradients of e f f e o t . " J . GeneWPsychol., 1945, 66, pp. 3-19. - 34 -• Wallaoh, H., & Henle, Mary. "An experimental analysis of the law of.effect." J. Exp. Psychol., 1941, 28, pp.340-349. Wallach, H. & Henle, Mary. "A further study of the function of reward." J. Exp. Psyohol., 1942, 30, pp.147-160. Zirkle, G. A. "Success and fail u r e in s e r i a l learning. I v The Thorndike effect." J. Exp. Psychol., 1946, 36, pp. 230-236. ~ Zirkle, G. A. "Success and failure in serial learning. II. Isolation and the Thorndike effect." J. Exp. Psychol., 1946,-36, pp. 302-315. - •• - 35 -Appendix WORD-LIST AND RECORD SHEET CA Grade .......... SUBJECT. IQ, Sex T e s t TRIALS 3 4 l o s e swim r u n c r y dance f a l T throw l a u g h see c l i m b s I F trow Lraw l o o k move d r i v e skT reac walk b u i l d  cough  sleep  caton  sneeze  smile  p u l l  burn  shine  b i t e  f i n d fi v e H e —  count  f r e e z e  break d r i n k 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0107028/manifest

Comment

Related Items