Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

"Errorless" discrimination of reversed letters by children Groves, Muriel Kathleen 1972

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1972_A8 G76.pdf [ 4.59MB ]
Metadata
JSON: 831-1.0101591.json
JSON-LD: 831-1.0101591-ld.json
RDF/XML (Pretty): 831-1.0101591-rdf.xml
RDF/JSON: 831-1.0101591-rdf.json
Turtle: 831-1.0101591-turtle.txt
N-Triples: 831-1.0101591-rdf-ntriples.txt
Original Record: 831-1.0101591-source.json
Full Text
831-1.0101591-fulltext.txt
Citation
831-1.0101591.ris

Full Text

"ERRORLESS" DISCRIMINATION OF REVERSED LETTERS BY CHILDREN by MURIEL KATHLEEN GROVES B.A., University of B r i t i s h Columbia, 1963 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n the Department of Psychology We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October, 1972 In p.-resentlng th i s thesis^ in pa r t i a l f u l f i lment o f the requirements fo r an advanced degree at the Un iver s i t y of B r i t i s h Columbia, I agree that the L ib ra ry sha l l make i t f r e e l y ava i l ab le for reference and study. I f u r ther agree that permission for extensive copying of th i s thes i s fo r s cho l a r l y purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or pub l i c a t i on of th i s thes i s f o r f i nanc i a l gain sha l l not be allowed without my wr i t ten permiss ion. Department of Psychology The Un ivers i ty o f B r i t i s h Columbia Vancouver 8, Canada D a t e September 29. 1972. ABSTRACT "E r r o r l e s s " Discrimination of Reversed Letters by Children by Muriel K. Groves The purpose of t h i s project was to devise a method to teach preschool c h i l d r e n to discriminate between the l e t t e r s b and d quickly and with a minimum of errors. In the two p i l o t studies a simultaneous discrimination task was employed with one l e t t e r serving as Sf and one as S-. Children who did not reach c r i t e r i o n during a pretest matching-to-sample task involving the l e t t e r s b and d served as Ss. During t r a i n i n g t r i a l s the childre n were reinforced with candy a f t e r each correct response. In both studies control Ss received the f i n a l form of the l e t t e r s t i muli throughout the t r a i n i n g t r i a l s . For the progressive Ss i n the f i r s t experiment, the c i r c u l a r part of the negative l e t t e r was i n i t i a l l y absent and then was introduced gradually over t r i a l s . None of the progressive or control Ss acquired the b-d discrimination i n t h i s experiment. In the second study three-dimensional l e t t e r s were used. For the progressive Ss the upright negative l e t t e r i n i t i a l l y i was set so that the c i r c u l a r part of the l e t t e r was perpendicular to the front of the body of the S, who was seated facing the l e t t e r s . The c i r c u l a r part of the l e t t e r pointed toward the S. Then, over t r i a l s , the negative l e t t e r was rotated gradually around i t s v e r t i c a l axis u n t i l the c i r c u l a r part of the l e t t e r was p a r a l l e l to the body of the S. The c i r c u l a r part of the p o s i t i v e l e t t e r always was p a r a l l e l to the front of the S. Thus i n the f i n a l step of fading the c i r c u l a r parts of the p o s i t i v e and negative l e t t e r s were p a r a l l e l to the S and were facing i n an opposite l e f t - r i g h t d i r e c t i o n . In the second study one of the eight control Ss and four of the seven progressive Ss reached c r i t e r i o n . The majority of the progressive Ss who reached c r i t e r i o n made a number of errors. It was suggested that one possible reason f o r the lack of the success of the p i l o t studies was that the fading of the S- l e t t e r was not done i n a manner which was very "relevant" to the f i n a l l e f t - r i g h t discrimination. In other words, the steps i n the fading sequence did not, through a kind of successive approximations approach, ensure that the behaviour of the c h i l d would come gradually under the control of the d i r e c t i o n a l i t y of the l e t t e r s . In the p r i n c i p a l experiment an attempt was made to devise a fading sequence which would be more "relevant" to i i the f i n a l l e f t - r i g h t discrimination. This fading sequence was compared with a control procedure i n which the f i n a l forms of the l e t t e r s were used throughout the t r i a l s . Subjects who did not reach c r i t e r i o n on the baseline matching-to-sample t r i a l s with the l e t t e r s b-d were assigned to the progressive or control groups. The progressive Ss f i r s t received a series of matching-to-sample t r i a l s involving arrows pointing e i t h e r to the l e f t or r i g h t . Those who reached c r i t e r i o n on the arrow stimuli received a fading program involving a gradual progression from arrows to f i n a l letter-forms. The Ss who did not reach c r i t e r i o n on the arrow series received matching t r i a l s involving simultaneously moving arrows. The arrows were moved v e r t i c a l l y and then h o r i z o n t a l l y . The distance of the horizontal movement was reduced gradually u n t i l the arrows were stationary. The Ss then received the fading sequence from stationary arrows to f i n a l letter-forms. Eight of the nine progressive Ss reached c r i t e r i o n on the b-d discrimination. Seven of the progressive Ss who reached c r i t e r i o n responded c o r r e c t l y on more than 90 percent of the t r i a l s while the eighth S was correct on 77 percent of the t r i a l s . Only one of the nine control Ss reached c r i t e r i o n . A l l Ss reaching c r i t e r i o n on the b-d discrimination successfully transferred to a p-q i i i discrimination. The r e s u l t s of these experiments indicate that many preschool c h i l d r e n do not learn to discriminate between the reversed l e t t e r s b and d under a d i f f e r e n t i a l reinforcement procedure. However, the majority of such children learn to make t h i s discrimination with no or few errors when they receive a fading sequence which i s "relevant" to the f i n a l discrimination. i v TABLE OF CONTENTS Chapter Page I INTRODUCTION 1 1. Stimulus Control and the Left-Right Discrimination 1 2. Normative Studies on Left-Right Discrimination i n Children 4 3. Experimental Attempts to Teach Left-Right Discriminations 10 4. Purpose of Present Research 25 II PRELIMINARY STUDIES 26 1. Preliminary Experiment I 26 Method 26 Subjects 26 Materials and Apparatus 26 Design and Procedure 27 Results and Discussion 30 2. Preliminary Experiment II 31 Method 31 Subjects 32 Materials and Apparatus 32 Procedure 32 Results and Discussion 35 3. Summary 40 III MAIN EXPERIMENT 42 1. Introduction 42 2. Method 44 Subjects 44 v Chapter Page Materials and Apparatus 46 Procedure 48 3. Results 57 4. Discussion 65 General Summary and Comment 65 Reasons for the Success of the Program and Possible Modifications 70 Comparison of the Program with Other Training Methods 73 Theoretical Implications 77 5. Summary 80 REFERENCES 83 v i LIST OF TABLES Table Page 1. Age and Sex of Subjects i n the Di f f e r e n t Conditions 45 v i i LIST OF ILLUSTRATIONS Figure Page 1. P o s i t i v e stimulus d and steps 2, 3, 14, and 20 of the progressive modification of b 28 2. P o s i t i v e stimulus d and steps 1, 3, 7, 14, and 19 of the progressive r o t a t i o n of b 34 3. Summary and i l l u s t r a t i o n s of the phases i n the progression of the sti m u l i from thick stationary arrows to f i n a l letter-forms 47 4. Summary of experimental procedure 50 5. I l l u s t r a t i o n of how the arrows were moved 54 6. The percentage of t r i a l s on which Control and Progressive Ss responded c o r r e c t l y 59 7. Cumulative correct responses of Ss i n the Control condition, i n the Progressive i condition, and i n the Baseline-Only group 64 v i i i ACKNOWLEDGEMENTS I wish to thank my advisors, Dr. D. M. Wilkie and Dr. G. J . Johnson for t h e i r assistance and comments. I e s p e c i a l l y would l i k e to thank my chief advisor, Dr. D. M. Wilkie, for a l l h i s help and encouragement. I also would l i k e to thank my husband, Alan, f o r his support. A spe c i a l thanks to my parents, F l o r a and Arthur Croker, for t h e i r very great help and encourage-ment, and to my l i t t l e son John f o r being such a good boy and making i t possible f o r me to do t h i s project. I also would l i k e to thank Mrs. H. I. Jacobson for looking a f t e r John so well and Mrs. M. J . Fi n l a y for typing the manuscript with such care. ix CHAPTER I INTRODUCTION 1. Stimulus Control and the Left-Right Discrimination In h i s 1966 a r t i c l e Terrace defined stimulus control ^ as follows: Stimulus control r e f e r s to the extent to which the value of an antecedent stimulus determines the p r o b a b i l i t y of the occurrence of a conditioned response. It i s measured as a change i n response p r o b a b i l i t y that r e s u l t s from a change i n stimulus values. The greater the change i n response p r o b a b i l i t y the greater the degree of stimulus control with respect to the continuum being studied, (p. 271) As an example of stimulus c o n t r o l , suppose that a pigeon i s reinforced for pecking a key illuminated by monochromatic l i g h t . To determine i f pecking i s under the control of wavelength, the wavelength projected on the key i s varied and pecking during these various test wavelengths i s not reinforced. The pecking response i s said to be under the control of wavelength i f the p r o b a b i l i t y of ^The concept of stimulus control i s similar to the concepts of discrimination and generalization ( c f . , Terrace, 1966). However, the l a t t e r terms have connotations not implied by the term stimulus c o n t r o l . While both stimulus control and discrimination-generalization w i l l be used interchangeably i n the present paper, only the meaning of the former i s intended. 1 2 pecking covaries with wavelength - the steeper the gradient r e l a t i n g response p r o b a b i l i t i e s to stimulus values, the greater the degree of stimulus control. I f pecking and wavelength do not covary during the te s t , pecking i s not under the control of wavelength. Terrace (1966), a f t e r reviewing the l i t e r a t u r e , concluded that the establishment of control by a p a r t i c u l a r stimulus continuum requires a d i f f e r e n t i a l reinforcement procedure; that i s , one or more values along the continuum must be associated with reinforcement while one or more values must be correlated with non-reinforcement or a d i f f e r e n t schedule of reinforcement. As support f o r t h i s conclusion, Terrace c i t e d a number of studies (e.g., Jenkins and Harrison, 1960). In the study of Jenkins and Harrison one group of pigeons was reinforced for pecking i n the presence of a tone of a c e r t a i n frequency (S+- or posit i v e stimulus). A second group of birds was reinforced f o r pecking i n the presence of the same tone (S+) and was not reinforced f o r pecking during i t s absence (S- or negative stimulus). During a subsequent test f o r stimulus control, tones of d i f f e r e n t frequencies were presented. The Ss who received n o n - d i f f e r e n t i a l t r a i n i n g produced f l a t gradients, while the Ss who received d i f f e r e n t i a l t r a i n i n g produced gradients which peaked at the frequency correlated with 3 reinforcement. Thus stimulus control was established only when a d i f f e r e n t i a l reinforcement procedure was used. In most experiments on stimulus control the stimuli selected f o r study l i e on a definable continuum. In some cases, however, the stimuli do not appear to l i e along any e a s i l y definable continuum. For example, the stimuli of the present study, which d i f f e r only i n whether they face to the l e f t or to the r i g h t , do not seem to f a l l along a continuum of l e f t - r i g h t d i r e c t i o n a l i t y . There appear to be only two values on the l e f t - r i g h t dimension; i n r e l a t i o n to the organism, stimuli may face to the l e f t or to the r i g h t . In the case of such s t i m u l i , stimulus control cannot be assessed by measuring response p r o b a b i l i t i e s during several test values from the continuum as i s commonly done with stimuli f a l l i n g along definable continua. Instead stimulus control i s estimated with reference to the two values. For example, i n a l e f t - r i g h t discrimination task ( i e . , one stimulus i s pointing to the r i g h t ; the other, to the l e f t ) stimulus control can be assessed by comparing the pro b a b i l i t y of responding to the positive stimulus and the pr o b a b i l i t y of responding to the negative stimulus. I f the organism responds d i f f e r e n t i a l l y during presentation of the l e f t and r i g h t s t i m u l i , i t may be said that he i s under the control of the l e f t - r i g h t dimension. 4 Both psychologists and educators have been i n t e r -ested i n determining how children's behaviour comes under the control of the l e f t - r i g h t aspect of objects. The concern of educators with t h i s problem r e s u l t s from the observation that the l e f t - r i g h t aspect of printed l e t t e r s and words does not control the behaviour of many children who are learning to read. Such children may i n c o r r e c t l y i d e n t i f y the l e t t e r s b-d and p-q and may read f o r example, "bog" instead of "dog". In addition, these childr e n may not read across the page i n a consistent l e f t - r i g h t d i r e c t i o n . For example, "was" may be read as "saw" and "scared" may become "sacred". Some children p e r s i s t i n making these reversals longvafter the majority of children have learned to attend to the d i r e c t i o n a l component of l e t t e r s and words. Some educators (e.g., Teegarden, 1932; Wolfe, 1939) have suggested that the persistence of reversals i n reading i s one of the major symptoms of reading retardation and may be even the primary factor. 2. Normative Studies on Left-Right Discrimination i n Children Several studies have shown that the behaviour of many preschool and kindergarten childr e n i s not under the control of the l e f t - r i g h t aspect of objects. Newhall (1937), for example, showed pictures including those of a chair, 5 boat, rabbit, and horse to three- to five-year-old children. The children were instructed to select from the tray before them a cut-out block figure i d e n t i c a l to the picture being presented. The figures were arranged on the tray so that they were rotated i n eithe r a l e f t - r i g h t or an up-down d i r e c t i o n from the sample. A f t e r the S i d e n t i f i e d the matching cut-out by handing i t to the E, the E then handed back the form rotated i n ei t h e r an up-down or l e f t - r i g h t d i r e c t i o n from the sample. The S was asked to place i t the "same" as the picture on the chart. While the childre n matched blocks to chart pictures almost without error, they did not always c o r r e c t l y place the objects handed back by the E, p a r t i c u l a r l y when the objects were reversed i n the l e f t - r i g h t d i r e c t i o n . When the blocks were rotated i n the up-down d i r e c t i o n , approximately 80 percent of the blocks were c o r r e c t l y placed. When the blocks were reversed i n the l e f t - r i g h t d i r e c t i o n , correct placements ranged from 17 percent to 27 percent over Ss. Gibson, Gibson, Pick, and Osser (1962) studied four- to eight-year-old c h i l d r e n . Their Ss were required to choose from twelve s t i m u l i , the stimulus which was the "same" as the sample. Among the comparison s t i m u l i were several rotated transformations of the sample stimulus including a 45-degree ro t a t i o n , a 90-degree r o t a t i o n , a 180-degree 6 rot a t i o n , and a l e f t - r i g h t r e v e r s a l . Approximately 45 percent of the l e f t - r i g h t reversals were selected by the four-year-olds; 15 percent by the six-year olds; and 5 percent by the eight-year-olds. In another study, Wechsler and Hagin (1964) evaluated the extent to which childre n i n the f i r s t grade and the t h i r d grade of school discriminated between objects having d i f f e r e n t orientations. A six-choice matching-to-sample task was used. The sti m u l i were asymmetrical figures roughly shaped l i k e lamb chops. The f i v e incorrect comparisons included a 90-degree rota t i o n , a l e f t - r i g h t r e v e r s a l , an up-down rev e r s a l , a 180-degree r e v e r s a l , and one of ei t h e r the combined 90-degree - l e f t - r i g h t reversal or the 90-degree - up-down rever s a l . I t was found that f i r s t grade £>s i n c o r r e c t l y selected the l e f t - r i g h t r e versal of the sample on 13 percent of the t r i a l s , while the third-grade £>s selected the l e f t - r i g h t reversal on 3 percent of the t r i a l s . Another study by Rudel and Teuber (1963) employed simultaneous discrimination tasks involving up-down rotations and l e f t - r i g h t reversals of a U-shaped figure. The Ss were four- to eight-year-old children. The Ss were to l d a f t e r every t r i a l whether t h e i r choice had been correct or incorrect. C r i t e r i o n for each task was 10 7 successive correct responses i n 10 t r i a l s . I t was found that nearly a l l the three- to eight-year-old c h i l d r e n reached c r i t e r i o n when they were required to discriminate between " J l " and "U". However, approximately 85 percent of the three-, four-, and five-year-olds did not learn to discriminate between the l e f t - r i g h t r eversals, and " • " . A f t e r schooling began there was a considerable improvement i n the a c q u i s i t i o n of the l e f t - r i g h t discrim-ination; approximately 33 percent of the six-year-olds and 85 percent of the eight-year-olds reached c r i t e r i o n on the task. Swanson and Benton (1955) attempted to determine the extent to which childre n of ages f i v e to nine years c o r r e c t l y i d e n t i f i e d the r i g h t - l e f t parts of t h e i r own body and the body of a picture of a front view of a male c h i l d . Such commands as "Show me your r i g h t hand." and "Show me the l e f t leg of the boy." were given. I t was found that some children were f a i r l y consistent i n l a b e l l i n g l e f t as r i g h t and vice versa. Thus, over Ss, the correct verbal l a b e l l i n g scores were le s s than the "discrimination" scores which included c r e d i t f o r consistent incorrect l a b e l l i n g . Consideration of the "discrimination" scores indicated that f o r childre n of a l l ages the i d e n t i f i c a t i o n of the body parts of the boy i n the picture was more 8 d i f f i c u l t than the i d e n t i f i c a t i o n of the parts of t h e i r own body. The percentage of correct responses f o r the s e l f -i d e n t i f i c a t i o n and the o t h e r - i d e n t i f i c a t i o n tasks were, respectively, 71.5 percent and 52.0 percent f o r the six-year-olds, 86.8 percent and 56.0 percent for the seven-year-olds, 89.7 percent and 69.0 percent f o r the eight-year-olds, and 93.6 percent and 80.8 percent f o r the nine-year-olds. In studies i n which alphabet-letters have been used as s t i m u l i , i t generally has been found that preschool children, and to a lesser extent school children, have d i f f i c u l t y i n discriminating between the l e t t e r - p a i r s b-d and p-q. In a study by Davidson (1935), one of the sample l e t t e r s d,e,p,q, or b was printed on the l e f t side of each sheet of paper. To the r i g h t of the sample l e t t e r were four rows of ten l e t t e r s , including f i v e i d e n t i c a l forms, f i v e rotated forms, several l e t t e r s s i m i l a r i n shape to the sample, and a number of other l e t t e r s . Subjects were required to cross out l e t t e r s which were the "same" as the sample. Davidson found that approximately 90 percent of the kindergarten childr e n i n c o r r e c t l y crossed out one or more of the reversed comparison l e t t e r s when the sample l e t t e r s were d,p,b, and q. A smaller, but sub s t a n t i a l l y large percentage (approximately 60 percent) of f i r s t - g r a d e 9 chi l d r e n made i d e n t i c a l errors. Popp (1964) reported s i m i l a r findings with f i v e -and six-year-old kindergarten children. In a two-choice matching-to-sample task, each l e t t e r of the alphabet served as.the sample with one of each of the other 25 l e t t e r s as the incorrect comparison. I t was found that the children made the most errors with the pairs b-d and p-q. Approxi-mately 35 percent of the children made incorrect matches when b or d were the samples and d and b were the comparison s t i m u l i . Approximately 38 percent of the children made sim i l a r errors when p or q were the samples and p and q were the comparison s t i m u l i . A number of studies have indicated that d i f f i c u l t i e s with these l e t t e r - p a i r s s t i l l occur, but to a les s e r extent, a f t e r several months of reading i n s t r u c t i o n ( H i l l , 1936) and even i n higher grades (Kennedy, 1954; Wilson and Flemming, 1938). For example, Kennedy used a matching-to-sample task with single l e t t e r samples such as p, d, and t and 12 comparison s t i m u l i including 3 correct ones and 3 of each of the inverted, reversed, and inverted-reversed rotations. I t was found that Ss i n the f i r s t grade of school i n c o r r e c t l y selected 30 percent of the reversed comparison s t i m u l i . At the end of grade two, the Ss chose 3.3 percent of these incorrect items. 10 3. Experimental Attempts to Teach Right-Left Discriminations The preceding studies show that the behaviour of young childre n often i s not under the control of the l e f t -r i g h t aspect of s t i m u l i . Since i t i s necessary that c h i l d r e n learn to discriminate between l e f t and r i g h t i n order to perform adequately such tasks as reading and following d i r e c t i o n s , considerable attention has been paid to devising methods to teach c h i l d r e n t h i s discrimination. The previously mentioned study of Rudel and Teuber (1963) c l e a r l y indicated that i n a simultaneous discrimination task, simply t e l l i n g the S a f t e r each t r i a l i f he was correct or incorrect does not ensure that the Si's behaviour w i l l come under the control of the l e f t - r i g h t aspect of the s t i m u l i . A number of other researchers have devised and tested somewhat more elaborate methods to teach t h i s discrimination. In a study by J e f f r e y (1958), two methods of teaching a l e f t - r i g h t discrimination to preschool childr e n were compared. The Ss were shown on each t r i a l a s t i c k figure with one of i t s arms at i t s side and the other arm pointing e i t h e r to the l e f t or to the r i g h t . For the f i r s t teaching method Ss were instructed to l a b e l the s t i c k figures according to which way the arm pointed. They were t o l d to c a l l the figure with the arm pointing " i n ' t h i s way, J i l l " , and the figure with the arm pointing " t h i s way, Jack". These instructions were repeated at i n t e r v a l s from 10 to 20 t r i a l s . Only 3 of 14 Ss reached the c r i t e r i o n of 10 successive correct l a b e l l i n g responses; they required 30 to 40 t r i a l s . A second group had 20 t r i a l s i d e n t i c a l to the t r i a l s received by the previous group throughout the task. The Ss not reaching c r i t e r i o n on t h i s l a b e l l i n g phase received t r i a l s on which they were required to push a button i n the d i r e c t i o n i n which the s t i c k figure was pointing. A f t e r a c r i t e r i o n of 10 successful correct responses the Ss then returned to the l a b e l l i n g task. Five Ss reached c r i t e r i o n on the i n i t i a l l a b e l l i n g phase. Seven of the eight Ss who did not reach c r i t e r i o n on the i n i t i a l l a b e l l i n g reached c r i t e r i o n on the l a b e l l i n g a f t e r they had received the button-pressing t r i a l s . The number of t r i a l s (including the i n i t i a l l a b e l l i n g and the button-pressing t r i a l s ) required by these seven S_s to reach c r i t e r i o n ranged from approximately 40 to 80 t r i a l s . Consideration of the data suggests that the Ss made a lower percentage of errors on the button-pressing t r i a l s than on the l a b e l l i n g t r i a l s which followed. J e f f r e y suggested that the superior performance of the second group resulted from the l a b e l l i n g task being preceded by a "simpler" task, i n t h i s case d i f f e r e n t i a l 12 pressing. According to J e f f r e y , success i n learning to respond d i f f e r e n t i a l l y to s t i m u l i d i f f e r i n g i n o r i e n t a t i o n depends upon the nature of the discriminatory response to be learned. Certain discriminatory responses such as d i f f e r e n t i a l pressing are easier to learn than are others such as d i f f e r e n t i a l l a b e l l i n g . Having the Ss i n i t i a l l y l e arn a "simpler" discriminatory response ( i n t h i s case d i f f e r e n t i a l pressing) f a c i l i t a t e s l a t e r learning of more " d i f f i c u l t " discriminatory responses such as l a b e l l i n g . J e f f r e y does not c l e a r l y explain the mechanisms of t h i s f a c i l i t a t i o n . However, i t has been suggested (e.g., Fellows, 1966) that button pressing f a c i l i t a t e d the l a t e r l a b e l l i n g by making the s t i m u l i more " d i s t i n c t i v e " . According to t h i s view, considerably d i f f e r e n t kinesthetic cues r e s u l t i n g from reaching to the l e f t or to the r i g h t were associated with each stimulus. Thus each stimulus became more " d i s t i n c t i v e " and the l a t e r attaching of labels to the stimuli was f a c i l i t a t e d . In the case of Ss who had the verbal l a b e l l i n g task throughout, t h i s task was not learned by many Ss because there was no i n i t i a l d i f f e r e n t i a l pressing which would make the s t i m u l i " d i s t i n c t i v e " enough for correct l a b e l l i n g . A l a t e r study by Hendrickson and Muehl (1962), however, casts doubt upon the i n t e r p r e t a t i o n that the 13 benefits of J e f f r e y ' s pressing method resulted from d i f f e r e n t k i n e s t h e t i c cues being associated with each stimulus. In t h i s study kindergarten childr e n served as Ss. There were two main experimental groups. The st i m u l i f o r the f i r s t group consisted of the l e t t e r s b and d, below which were printed arrows always pointing outward ( i e . , b, d ). In the case of the st i m u l i f o r the second group, h a l f of the l e t t e r s shown had the arrows below them pointing outward ( i . e . , ^ b ^ d^_^ ) while the other h a l f of the l e t t e r s had the arrows pointing inward ( i . e . , _ b ^ _>d<_)« I n the t r a i n i n g phase the l e t t e r s were presented successively. The Ss were required to indicate which arrow below the l e t t e r pointed the same way as the l e t t e r by pushing the r i g h t handle i f the correct arrow was on the r i g h t side and the l e f t handle i f the correct arrow was on the l e f t side. For the f i r s t group correct responding involved a consistent l e f t arm response to the l e t t e r d and a consistent r i g h t arm motor response to the l e t t e r b. For the second group the correct arm responses to the l e t t e r s were inconsistent; for example, the l e f t arm response was required on h a l f the presentations of d; the r i g h t arm response on the other h a l f . A f t e r 30 t r a i n i n g t r i a l s a paired associate transfer task was administered. The a n t i c i p a t i o n method was used. The c h i l d was required to le a r n which picture, a pumpkin 14 or a snowman, was associated with each l e t t e r . If the benefits of the handle pushing method, and thus presumably J e f f r e y ' s button pressing method, resulted from d i f f e r e n t kinesthetic cues being associated with each stimulus, i t would be expected that the f i r s t group would perform better on the transfer task than would the second group. Presumably, i n the case of the former group the complex of kin e s t h e t i c cues r e s u l t i n g from reaching to the l e f t would be associated with one stimulus while the complex of cues r e s u l t i n g from reaching to the r i g h t would be associated with the other stimulus. In the case of the l a t t e r group, the kinesthetic s t i m u l i produced by reaching both to the l e f t and to the r i g h t would be associated with both s t i m u l i . The r e s u l t s , however, indicated that there was no s i g n i f i c a n t difference between the performance of the groups. In both groups 12 of 16 Ss learned to select the correct picture ( i e . , pumpkin or snowman) f o r each l e t t e r . As might be expected, the f i r s t group made a s i g n i f i c a n t l y lower percentage of errors i n the tr a i n i n g tasks than d i d the second group. However, i n the transfer tasks there was no s i g n i f i c a n t difference between the groups i n the percentage of errors. Hendrickson and Muehl con-cluded that methods which require the Ss to indicate the d i r e c t i o n a l aspect of l e t t e r s and arrows by pushing a handle or button with the hand toward which the s t i m u l i are facing are e f f e c t i v e because they ensure that the Ss attend to the d i r e c t i o n a l aspect of the s t i m u l i . Methods which require the Ss to indicate the d i r e c t i o n a l aspect of the s t i m u l i i n other ways including inconsistent hand movements s i m i l a r l y ensure that the S_s attend to the d i r e c t i o n i n which the s t i m u l i are facing. It would seem that the experimental techniques of Hendrickson and Muehl and p a r t i c u l a r l y J e f f r e y were e f f e c t i v e i n ensuring that a high percentage of Ss learned the l e f t - r i g h t discrimination. However, i t i s possible that at l e a s t some of the Ss who reached c r i t e r i o n i n these studies did not perform without a number of err o r s . While i t i s not c l e a r from Jef f r e y ' s data how many errors were made by the various Ss who were given the button pressing task, consideration of the number of t r i a l s required by the Ss to reach c r i t e r i o n would suggest that i n the button-pressing stage and p a r t i c u l a r l y i n the l a b e l l i n g task which followed, at l e a s t some Ss may have made a number of err o r s . In the case of the study of Hendrickson and Muehl i t i s even more d i f f i c u l t to estimate from the data provided how many errors were made by the Ss who reached c r i t e r i o n . Results from a l l the Ss i n both groups indicated that i n the transfer t r i a l s , the group 16 who made the movement consistent with the d i r e c t i o n of the l e t t e r were correct on 83 percent of the t r i a l s , while the other group, who made the inconsistent movement, were correct on 76 percent of the t r i a l s . The standard deviation f o r both groups was approximately 19 percent. Consideration of these means and standard deviations would suggest that i t i s possible that some of the Ss who reached c r i t e r i o n may have made a number of errors. A number of studies have attempted to devise techniques which, i n addition to enabling a high percentage of Ss to learn the l e f t - r i g h t discrimination, would ensure that the Ss performed with a minimum of errors. In these studies the differences between the stimuli were i n i t i a l l y exaggerated and then gradually reduced u n t i l the f i n a l forms were reached. The methods of these studies are based on the techniques of Lawrence (1952) and Terrace (1963a, 1963b). Lawrence showed that the most e f f i c i e n t way of estab l i s h i n g a discrimination was to devote the early t r a i n i n g t r i a l s to "easier" discriminations on the same continuum of s t i m u l i as the f i n a l discrimination. Rats which received simultaneous discrimination t r a i n i n g i n four stages, from a dark gray-light gray discrimination to the f i n a l discrimination between two s l i g h t l y d i f f e r e n t shades of gray, were able to achieve c r i t e r i o n on the f i n a l discrimination i n fewer t r i a l s ( t r i a l s on the e a r l i e r discriminations were counted) than r a t s who received t h e i r practice solely on the f i n a l discrimination. Lawrence's transfer on a continuum method has yielded s i m i l a r l y successful r e s u l t s i n a v a r i e t y of studies. For example, Baker and Osgood (1954) found that the most e f f i c i e n t way to e s t a b l i s h a p i t c h discrimination was to approach the task i n steps of increasing d i f f i c u l t y . S i m i l a r l y House and Zeaman (1960) showed that discrimination t r a i n i n g on three dimensional s t i m u l i p r i o r to t r a i n i n g on two dimensional patterns produced f a s t e r learning of the two dimensional discrimination than did training on the two dimensional patterns alone. Terrace showed that a d i f f i c u l t successive d i s -crimination could be acquired with no or few errors i f the difference between the p o s i t i v e and negative s t i m u l i was i n i t i a l l y large and was reduced very gradually. Terrace further observed that the reduction i n errors produced by a very gradual fading sequence eliminated the occurrence of a number of possibly undesirable e f f e c t s . In one experiment (Terrace, 1963a) pigeons learned a d i f f i c u l t d iscrimination without making errors ( i e . , responses during the non-reinforced negative stimuli) when the negative stimulus was introduced just a f t e r responding during the p o s i t i v e stimulus was conditioned. In t h i s study the negative stimulus, i n addition to d i f f e r i n g i n wavelength, was dimmer and shorter i n duration than the posit i v e stimulus. The brightness and duration of the negative stimulus was progressively increased u n t i l the brightness and duration of the po s i t i v e stimulus were reached. The Ss who received the fading sequence a f t e r they were exposed to the posit i v e stimulus for a period of time d i d not exhibit " e r r o r l e s s " performance. S i m i l a r l y , the Ss who received the f i n a l value of the sti m u l i throughout t r a i n i n g performed with e r r o r s . Terrace observed that some of the Ss who performed with errors showed bursts of incorrect responding even a f t e r a stable near perfect rate of responding had been reached. The Ss who learned e s s e n t i a l l y without errors did not exhibit these disruptions i n performance. In another experiment Terrace (1963b) demonstrated that the d i f f i c u l t d i s crimination of the orient a t i o n of two white l i n e s could be learned by pigeons without errors when the backgrounds of the two d i f f e r e n t orientations were of d i f f e r e n t colours which gradually faded to black. The control Ss who received the f i n a l d iscrimination throughout the t r i a l s made many errors. In both the preceding studies i t was observed that pigeons who had learned the discrim-i n a t i o n with errors often exhibited "emotional" behaviour 19 such as wing flapping i n the presence of the negative stimulus. Following " e r r o r l e s s " learning no such emotional responses were observed. Terrace (1966) further noted that when the t r a i n i n g technique permits the Ss to make errors, i t i s possible that responding might come under the c o n t r o l of s t i m u l i other than those which are experimentally manipulated. Under c e r t a i n conditions permanent f a u l t y performance might r e s u l t due to an intermittent reinforcement e f f e c t . This control of responding by factors beyond the E's control would not occur under an e f f e c t i v e progressive procedure. Terrace's fading method has been applied i n a number of teaching programs for children to ensure that discrimina-tions not only are learned, but are learned with a minimum of errors. For example, Sidman and Stoddard (1967) devised a teaching program involving a number of discriminations between c i r c l e s and e l l i p s e s . Over t r i a l s the r a t i o of the v e r t i c a l and horizontal axis of the e l l i p s e s was made to approach one. Using t h i s program retarded Ss learned to discriminate between a c i r c l e and an e l l i p s e which were very s i m i l a r . In another study (Moore and Goldiamond, 1964) involving a matchlng-to-sample task, c h i l d r e n were taught to select from three compar-ison t r i a n g l e s , the t r i a n g l e having the same o r i e n t a t i o n as the sample t r i a n g l e . The incorrect comparison s t i m u l i , which varied s l i g h t l y i n o r i e n t a t i o n from the sample, i n i t i a l l y 20 were of low l i g h t i n t e n s i t y . The l i g h t i n t e n s i t y of the incorrect matches was gradually increased to that of the correct match. While fading methods have been e f f e c t i v e i n teaching a number of d i f f i c u l t discriminations, attempts to produce a fading sequence which would ensure that children would acquire the l e f t - r i g h t discrimination have not been p a r t i c u l a r l y successful. In one of these attempts, Karraker (1968) investigated both the progressive-constant variable and the e a r l y - l a t e introduction of the negative stimulus. Both the progressive and constant groups received two sessions, each involving two phases. The f i r s t phase of each session was i d e n t i c a l for the constant and progressive groups. A set of seven s l i d e s was shown at each session. There was a d i f f e r e n t set f o r each session and only one l e t t e r was shown i n each set. The f i r s t s l i d e of the set showed eithe r an orange b or green d accompanied by a drawing of a bat and b a l l i n the case of b or a dog i n the case of d. An arrow pointed either to the drawing or to the l e t t e r . In the case of the l e t t e r b when the arrow pointed to the drawing, Ss were to say "bat and b a l l " , and when i t pointed to the l e t t e r , "b, as i n bat and b a l l " . Over the series of s l i d e s the l e t t e r which was being shown i n the series and i t s associated 21 drawing decreased i n size and faded to black. By s l i d e seven the drawing had been faded out, and there remained a black, lower-case l e t t e r . The second phase of each session involved successive presentations of the l e t t e r s b and d. In one session one l e t t e r was defined as the p o s i t i v e stimulus; i n the: next session the other l e t t e r was p o s i t i v e . The negative l e t t e r was introduced a f t e r e i t h e r 1 t r i a l or 23 t r i a l s of the p o s i t i v e l e t t e r . In the constant groups, black lower-case l e t t e r s were presented throughout the t r i a l s . In the progressive groups the positive l e t t e r was i n the black lower-case form. The negative l e t t e r was faded i n along four dimensions ( s i z e , duration of exposure, colour, presence of drawings). I t i n i t i a l l y was larger than the p o s i t i v e l e t t e r and had a shorter duration of exposure. Depending upon whether the l e t t e r was a "b" or "d" i t was coloured orange or green and was accompanied by drawings of e i t h e r a bat and b a l l or a dog. A f t e r both phases had been completed i n both sessions, there was a c r i t e r i o n session of 14 simultaneous b-d discriminations followed by 62 randomly alternated presentations of the black lower-case l e t t e r s . In the 14 s l i d e s of the simultaneous phase, an arrow pointed to the p o s i t i v e stimulus, the one the c h i l d had to i d e n t i f y . The 22 l e t t e r s d i f f e r e d i n colour, with the d coloured green and the b, red. In addition, the posit i v e l e t t e r also was larger than the negative one. The po s i t i v e stimulus gradually was reduced i n size while the colour of both st i m u l i faded to black. In the successive phase of the c r i t e r i o n session the Ss were required to name the l e t t e r s which were i n black lower-case form. Consideration of the mean errors on c r i t e r i o n t r i a l s of the four groups showed that the time variable was s i g n i f i c a n t , whereas the progressive-constant variable was not s i g n i f i c a n t . The number of Ss emitting fewer than 10 percent errors was for each group: Early-Progressive 13/16, Early-Constant 11/16, Late-Progressive 6/16, Late-Constant 4/16. Due to the design of the study i t i s not possible to estimate whether many Ss who learned the discrimination under the progressive procedure might have required fewer t r i a l s to acquire the discrimination under a simple d i f f e r e n t i a l reinforcement procedure. This p o s s i b i l i t y i s great i n that the Ss ranged i n age from 5.5 years to 6.6 years with a mean age of 5.9 years. According to the previously mentioned study of Rudel and Teuber (1963) the percentage of Ss who learned quickly to make a l e f t - r i g h t d iscrimination increased greatly from ages f i v e to six. Thus i t i s possible that a number of 23 Ss would have acquired the discrimination a f t e r very few t r i a l s . It would seem that a better test f o r a fading procedure would employ Ss i n both control and fading groups who d i d not discriminate between reversed forms during some kind of baseline procedure. Schutz ( c i t e d by Karraker, 1968) was unsuccessful i n teaching the b-d discrimination with both l i g h t and colour fading, but was somewhat more successful with s t r u c t u r a l fading when the negative l e t t e r i n i t i a l l y was f i l l e d i n (k) and gradually was modified to achieve the normal form (b). Bijou (1963) also reported an experiment involving s t r u c t u r a l fading. Both normal and retarded c h i l d r e n were used as Ss. In a matching-to-sample task involving f i v e choices, a deformed version of the mirror-image of the sample nonsense shape was introduced as one of the choices and was a l t e r e d gradually over t r i a l s to achieve a f i n a l mirror-image form. For example the incorrect "mirror" comparison stimulus f o r the sample stimuluspwas progress-i v e l y modified from 3 to the f i n a l reversed form ^ by gradually moving the lower bar upward. Although Bijou's study often i s c i t e d as an example of successful teaching of the l e f t - r i g h t discrimination, the program was not e f f e c t i v e f o r a l l Ss; Ss varied considerably i n the extent of t h e i r progression through the program. In f a c t , i t i s not too clea r from the study what percentage of the Ss were successful. While a short pretest involving no reinforcement was given, there was no control group or any form of reinforced baseline procedure. These omissions question any claim f o r the success of the fading method, as there i s no way of estimating how many of the Ss would have learned the l e f t - r i g h t discrimination under a simple d i f f e r e n t i a l reinforcement procedure - perhaps i n fewer t r i a l s than were required with the progressive procedure. This p o s s i b i l i t y i s great i n that many of the normal Ss were f i v e or six years of age and came from a medium-high socioeconomic l e v e l . In addition, many of the retarded Ss had mental ages of greater than six years. Furthermore, a matching-to-sample task may be easier than the simultaneous or successive discrimination tasks which were used i n the previously mentioned fading experiments. In summary, the previously mentioned t r a i n i n g methods have not been e f f e c t i v e i n ensuring both that a high percentage of Ss learn the l e f t - r i g h t discrimination and that the performance of these Ss i s e s s e n t i a l l y " e r r o r l e s s " . Rudel and Teuber (1963) showed that simply t e l l i n g the S a f t e r each t r i a l i f he was correct or incorrect was i n e f f e c t i v e i n teaching the majority of 25 preschool c h i l d r e n to make the discrimination. In the case of the methods involving arm movement ( i e . , J e f f r e y , 1958; Hendrickson and Muehl, 1962), i t was found that these techniques, p a r t i c u l a r l y the method of J e f f r e y , taught a high percentage of Ss to make the l e f t - r i g h t discrimination. However, as mentioned previously, i t would seem possible that the performance of at l e a s t some of these Ss was not without a number of errors. With regard to the fading techniques of B i j o u (1963), Karraker (1968) and Schutz (1964), none of these methods appears to have been e f f e c t i v e i n teaching the l e f t - r i g h t discrimination to a high percent-age of Ss. It i s impossible to determine from the data presented by B i j o u (1963) and Karraker (1968) whether Ss who learned the discrimination made very few errors. It would seem possible, however, that at least some of the Ss made a number of e r r o r s . 4. Purpose of Present Research The purpose of the present research project was to develop a method which would ensure that a large percentage of preschool c h i l d r e n acquire quickly and with a minimum of errors a discrimination involving the l e t t e r s b and d. Two preliminary studies are reported i n Chapter I I . The p r i n c i p a l experiment of t h i s project i s reported i n Chapter I I I . CHAPTER II PRELIMINARY STUDIES 1. Preliminary Experiment I This study employed s t r u c t u r a l fading of the negative l e t t e r (S-) i n a simultaneous discrimination task. The c i r c u l a r part of the negative l e t t e r , e i t h e r b or d of the pair b-d, i n i t i a l l y was absent and gradually was introduced. Two extensions from the stem were increased gradually u n t i l they met to form the c i r c u l a r part of the l e t t e r . I t was thought that by having the c i r c u l a r part of the l e t t e r grow from the stem outward, t h i s fading sequence might ensure that the c h i l d ' s behaviour would come under the control of the d i r e c t i o n a l aspect of the l e t t e r . Method Subjects The Ss were 6 childre n from a Vancouver day care centre. The c h i l d r e n ranged i n age from four years, four months to six years, one month with a median age of f i v e years, one month. Materials and Apparatus Cardboard cards with the sample l e t t e r b or d 26 printed above the comparison l e t t e r s b and d were used during the pretest. During the simultaneous discrimination phase the cards used showed both l e t t e r s for the constant group and one l e t t e r and a stage of fading of the second l e t t e r f o r the progressive group. A l l the l e t t e r s or near letter-shapes were printed i n black ink on white cards. The l e t t e r s were 5.0 inches i n height and a maximum of approximately 2.5 inches i n width. Figure 1 shows examples of the st i m u l i used f o r the progressive group. The l e t t e r d i s shown as the po s i t i v e stimulus ( S t ) . Steps 2, 3, 14, and 20 of the progressive modification of b (S-) are shown. The p o s i t i v e l e t t e r was alone i n step one; then i n step two the stem of the negative l e t t e r also was shown. Over the next 19 steps the c i r c u l a r part of the negative l e t t e r was introduced gradually from the stem outward. The two extensions from the stem were increased gradually so that t h e i r combined lengths were 0.35, 0.40, 0.50, 0.60, 0.65, 0.70, 0.75, 0.80, 0.83, 0.86, 0.88, 0.90, 0.92, 0.94, 0.96, 0.98, 0.99, 0.995 and 1.00 times the circumference of the c i r c u l a r part of the l e t t e r . Design and Procedure There was a pretest of eight non-reinforced matching-to-sample t r i a l s . Children not achieving four 28 S T E P S d 14 d 20 S + S -Figure 1. P o s i t i v e stimulus d and steps 2, 3, 14 and 20 of the progressive modification of b. 29 successive correct matches during the pretest were assigned randomly to the control and experimental groups. For h a l f of the Ss i n each group b was the posit i v e stimulus (S+-). For the other h a l f of the Ss, d was the pos i t i v e stimulus. The Ss then received simultaneous discrimination t r i a l s . Subjects were instructed that one of the l e t t e r s would be correct throughout the t r i a l s and they should indicate t h e i r choice by pointing on every t r i a l . The S was t o l d when he was correct or incorrect ( i e . , " r i g h t " or "wrong") and when correct he received a candy ( i e . , a continuous reinforcement schedule) which was put i n a bag for his l a t e r consumption. In the progressive group the Ss received a backup procedure when an error was made. Other than for step 1 of the sequence, the S received the fading step appropriate to the preceding step, n-1. This backup procedure continued through n-2, n-3, etc. u n t i l the S was correct or reached the f i r s t step. Then he received the step succeeding the one upon which he was correct. The stim u l i were presented so that the l e f t - r i g h t p o s i t i o n of the positive stimulus followed a Gellermann seri e s . This randomization was modified i n the progressive group when an error was made and there was a backup to the preceding step. Each stimulus presentation was shown u n t i l the S made a choice. The i n t e r t r i a l i n t e r v a l was approximately 3 seconds. T r i a l s continued u n t i l the Ss reached a c r i t e r i o n of six successive correct responses on the f i n a l forms of the l e t t e r s or u n t i l they had received 50 t r a i n i n g t r i a l s . Results and Discussion None of the Ss i n eithe r group reached c r i t e r i o n . The control Ss performed at no better than chance l e v e l . For two of the three progressive Ss, responding was e s s e n t i a l l y without errors u n t i l the gap i n the negative stimulus was closed; then consistent correct responding ceased. Even a f t e r there were a number of backup t r i a l s to the next to l a s t step followed by progressions to the f i n a l value, consistent correct responding on the f i n a l values was not achieved. The t h i r d progressive S responded at somewhat better than chance l e v e l on the f i r s t 11 steps of fading. Then he began to make more errors as the gap i n the negative l e t t e r was reduced i n size and he never reached the f i n a l form of the negative stimulus. It would seem that the behaviour of the progressive Ss remained under the control of the open-closed aspect of the c i r c u l a r part of the l e t t e r u n t i l the gap was made very small or was closed altogether. Then, i n the absence of the open-closed feature of the stimulus, correct responding broke down. Similar breakdowns i n correct responding toward the end or at the end of the fading sequence was seen i n the previously mentioned studies of B i j o u (1963), Karraker (1968), and Schutz (1964). In the case of the present study, one possible reason why the behaviour of the chi l d r e n never came under the control of the l e f t - r i g h t dimension of the s t i m u l i i s that the gap i n the negative l e t t e r might have been considerably more " d i s t i n c t i v e " than the l e f t - r i g h t aspect of the l e t t e r s even toward the end of the fading sequence. According to Piaget and Inhelder (1956) the open-closed aspect of sti m u l i i s one of the f i r s t topological features to which the young c h i l d attends. Presumably, then, t h i s aspect would be very " d i s t i n c t i v e " . 2. Preliminary Experiment II Since the fading method of the f i r s t experiment proved unsuccessful, another method which involved ro t a t i n g the S- l e t t e r around the v e r t i c a l axis was t r i e d . I t was thought that by having the negative stimulus gradually turn toward the l e f t or r i g h t , t h i s fading sequence might ensure that the children would attend to the d i r e c t i o n i n which the stimulus was turning and thus t h e i r behaviour would come under the con t r o l of the d i r e c t i o n a l aspect of the l e t t e r s . Method 32 Subjects The Ss were 15 c h i l d r e n selected from a Vancouver day care centre. The children, ranging i n age from three years, six months to six years, seven months, had a median age of f i v e years, two months. Materials and Apparatus The s t i m u l i were the l e t t e r s b, d, p, and q, which measured approximately 7.5 inches i n height and 2.0 inches i n width. The c i r c u l a r part of each l e t t e r was made of thick black cardboard which was inserted i n s l i t s i n narrow black poles, approximately 0.3 inch i n diameter, which formed the stems of the l e t t e r s . The poles were inserted i n thick sheets of wood. Each sheet held a p a i r of l e t t e r s , e i t h e r b-d or p-q. Procedure The general procedure of the previous p i l o t study was followed. There was a pretest of eight non-reinforced matching-to-sample t r i a l s . Children not achieving four successive correct matches during the pretest were assigned randomly to the control and experimental groups. For h a l f the Ss i n each group b was the S+ l e t t e r . For the other h a l f of the Ss d was the S+ l e t t e r . Subjects were instructed that one of the l e t t e r s would be correct throughout the t r i a l s and they should indicate t h e i r choice by pointing appropriately on every t r i a l . The S was t o l d when he was correct or incorrect ( i e . , " r i g h t " or "wrong") and when correct he received a candy which was put i n a bag f o r h i s l a t e r consumption. Figure 2 shows how the S- l e t t e r was rotated over t r i a l s f o r the progressive group. Steps 1, 3, 7, 14, and 19 of the progressive r o t a t i o n of b are shown. In step one the upright form of the negative stimulus was set so that the c i r c u l a r part of the l e t t e r was perpendicular to the front of the body of the S, who was seated facing the l e t t e r s . The c i r c u l a r part of the l e t t e r pointed toward the S. The negative l e t t e r was rotated gradually around i t s v e r t i c a l axis u n t i l the c i r c u l a r part of the l e t t e r was p a r a l l e l to the body of the S. The steps i n the r o t a t i o n from the perpendicular p o s i t i o n were 10, 20, 30, 40, 50, 60, 65, 67%, 70, 72%, 75, 77%, 80, 82%, 85, 87, 88% and 90 degrees. The c i r c u l a r part of the positive l e t t e r always was p a r a l l e l to the front of the S. The constant group received the l e t t e r s i n the f i n a l p o s i t i o n throughout the t r i a l s ( i e . , the c i r c u l a r parts of the po s i t i v e and negative l e t t e r s were p a r a l l e l to the S and were facing i n an opposite l e f t - r i g h t d i r e c t i o n ) . The Ss i n the progressive group received a backup procedure when an error was made. Other than f o r step 1 of the sequence, the S received the fading 34 Figure 2. P o s i t i v e stimulus d and, steps 1, 3, 7, 14, and 19 of the progressive r o t a t i o n of b. 35 step appropriate to the preceding step, n-1. This backup procedure continued through n-2, n-3, etc. u n t i l the S was correct or reached the f i r s t step. Then he received the step succeeding the one upon which he was correct. In both groups the st i m u l i were presented so that the l e f t - r i g h t p o sition of the positive stimulus followed a Gellermann series. This randomization was modified i n the progressive group when an error was made and there was a backup to the preceding step. Each stimulus presentation was shown u n t i l the S made a choice. The i n t e r t r i a l i n t e r v a l was approximately 5 seconds. T r i a l s continued u n t i l the Ss reached a c r i t e r i o n of six successive correct responses on the f i n a l forms of the l e t t e r s or u n t i l they had received 50 t r a i n i n g t r i a l s . A f t e r the c r i t e r i o n or the cutoff number of t r i a l s was reached, Ss received 10 simultaneous discrimination t r i a l s with the l e t t e r s p and q. C r i t e r i o n on t h i s transfer task was six successive correct responses. Results and Discussion One of the eight control Ss and four of the seven progressive Ss reached c r i t e r i o n . A c h i square test indicated that the difference between the groups i n the number of Ss who reached c r i t e r i o n was not s i g n i f i c a n t (p<.10). The Ss reaching c r i t e r i o n showed positive transfer 36 when tested on p and q. The progressive method provided e s s e n t i a l l y " e r r o r l e s s " learning with only one of the four progressive Ss who reached c r i t e r i o n . Three progressive Ss who reached c r i t e r i o n made a number of errors. Two of these Ss made errors at the f i n a l step when the negative l e t t e r was i n i t s f i n a l p o s i t i o n . Several backups to the previous step followed by progressions to the f i n a l step were required before consistent correct responding on the f i n a l stage was achieved. The t h i r d S made errors at various steps throughout the program. Two of the three progressive Ss who did not reach c r i t e r i o n did not reach the f i n a l phase of fading. Due to numerous errors throughout the early part of the fading sequence, one of these Ss never went beyond the fourth step of fading. The second S responded c o r r e c t l y up to the l a s t t h i r d of the program. Then correct responding broke down and the S never reached the f i n a l steps of the fading sequence. The t h i r d S reached the f i n a l step of fading but d i d not reach c r i t e r i o n even a f t e r a number of backups and progressions involving the f i n a l and next to l a s t steps of the program. While the fading procedure of t h i s study was more e f f e c t i v e than that of the f i r s t , the percentage of Ss (approximately 43 percent) who did not reach c r i t e r i o n was s t i l l high. I t should be noted that, as i n the f i r s t study, the behaviour of some Ss appeared to be under the control of the l e t t e r s up to the l a s t part of the program or the f i n a l step of fading. Then successive correct responding broke down and i n some cases could not be regained even a f t e r a number of backups and progressions. According to Terrace (1966) such breakdowns i n responding are observed when there has been an abrupt reduction of reinforcement density i n the p o s i t i v e stimulus, when there have been large steps i n the progressive change of the negative stimulus, p a r t i c u l a r l y on the f i n a l steps, or when the dimension along which the values of the stimuli were changed was "inappropriate". Terrace i s i n no way precise as to what i s meant by an "appropriate" or "inappropriate" dimension. As an example of fading along an "inappropriate" dimension Terrace c i t e s a study involving t r a i n i n g a group of pigeons to make a h o r i z o n t a l - v e r t i c a l l i n e discrimination. In t h i s study the i n t e n s i t y of S- was increased gradually to that of S+. Whereas the previously mentioned fading from the discrimination of colour to the orientation discrimination was an e f f e c t i v e procedure, the fading of l i g h t i n t e n s i t y was not. Discrimination performance was perfect u n t i l the l a s t few increases i n i n t e n s i t y , at which point consistent correct responding broke down. In the case of both preliminary studies there was no reduction 38 i n reinforcement density and the difference between the f i n a l steps of fading was very small. Thus i t would appear that the f a i l u r e of these fading sequences was due to the faded dimensions being "inappropriate". I t i s possible to consider that an "appropriate" fading dimension might be one that i s "relevant" to the f i n a l discrimination. For example, i n the case of the experiments of Lawrence (1952) and Baker and Osgood (1954), the dimension which was progressively modified contained the values involved i n the f i n a l discrimination. For example, i n the l a t t e r study the f i n a l p i t c h discrimination was established a f t e r e a r l i e r t r a i n i n g t r i a l s on "easier" p i t c h discriminations. In cases such as t h i s , the faded dimension would seem to be very "relevant" to the f i n a l discrimination. Another example of "relevance" might occur when the faded dimension and the dimension containing the f i n a l discriminative s t i m u l i are frequently associated; that i s , s t i m u l i f a l l i n g along one of the dimensions often w i l l have values along the other dimension. For example,, a note of a c e r t a i n p i t c h also w i l l have a c e r t a i n loudness. Thus, the dimension of sound frequency might be "relevant" to the dimension of sound i n t e n s i t y , and fading along one of the dimensions might f a c i l i t a t e learning a discrimination involving the other. An example of t h i s type of "relevance" might be seen i n the fading sequence of Terrace (1963a). As may be r e c a l l e d from Chapter I, part of the fading i n thi s experiment involved the gradual modification of the brightness of the negative stimulus u n t i l the stimuli d i f f e r e d only i n wavelength. As s t i m u l i of d i f f e r e n t wavelengths have values on both the wavelength and brightness continua, discrimination learning between d i f f e r e n t wavelengths might have been greatly f a c i l i t a t e d by the gradual modification of brightness. It i s obvious that there must be a v a r i e t y of complex ways i n which one dimension might be "relevant" to another dimension. In addition, the degree of "relevance" between two dimensions would seem to vary greatly. Except i n the case when the fading i s done along the dimension of the f i n a l discrimination, there would appear, at t h i s point, to be no a p r i o r i method of determining whether and to what extent the faded dimension i s "relevant" to the f i n a l discrimination. Since there seems to be no definable continuum of l e f t - r i g h t d i r e c t i o n -a l i t y , the devising of a fading procedure which would be "relevant" to the l e f t - r i g h t l e t t e r discrimination would seem to depend upon a " t r i a l and er r o r " approach. The experiment reported i n Chapter III i s another attempt to devise a fading procedure which would be 40 e f f e c t i v e i n e s t a b l i s h i n g a r i g h t - l e f t l e t t e r discrimination. 3. Summary This chapter reported two preliminary attempts to develop a method which would teach preschool c h i l d r e n to discriminate between the l e t t e r s b and d quickly and with a minimum of errors. In both studies there was fading of the S- l e t t e r i n a simultaneous discrimination task. Children who did not reach c r i t e r i o n on a pretest matching-to-sample task involving the l e t t e r s b and d served as Ss. During t r a i n i n g t r i a l s the children were reinforced with candy a f t e r each correct response. In both studies control Ss received the f i n a l form of the l e t t e r s t i m u l i throughout the t r i a l s . For the progressive Ss i n the f i r s t study, the c i r c u l a r part of the negative l e t t e r was i n i t i a l l y absent and then gradually was introduced over t r i a l s . None of the progressive or control Ss reached c r i t e r i o n on the b-d discriminations i n this experiment. In the second study three dimensional l e t t e r s were used. For the progressive Ss the upright form of the negative l e t t e r i n i t i a l l y was set so that the c i r c u l a r part of the l e t t e r was perpendicular to the front of the body of the S, who was seated facing the l e t t e r s . The c i r c u l a r part of the l e t t e r pointed toward the S. The negative l e t t e r was rotated gradually around i t s v e r t i c a l axis u n t i l the c i r c u l a r part of the l e t t e r was p a r a l l e l to the body of the S. The c i r c u l a r part of the pos i t i v e l e t t e r always was p a r a l l e l to the front of the S. S l i g h t l y more than h a l f the progressive Ss reached c r i t e r i o n on the b-d discrimination i n t h i s experiment. However, the majority of the progressive Ss who reached c r i t e r i o n made a number of errors. Reasons as to why a fading program might not be very e f f e c t i v e was discussed. I t was concluded that the fading i n both these studies was not done i n a manner that was "relevant" or "appropriate" to the f i n a l discrimination. CHAPTER III MAIN EXPERIMENT 1. Introduction This experiment involved another attempt to develop a fading method which would be e f f e c t i v e i n establishing a l e f t - r i g h t l e t t e r discrimination. As previously mentioned, there would appear, i n many cases, to be no a p r i o r i method of determining whether and to what extent a fading sequence i s "relevant" to the f i n a l discrimination. However, a recent study by Underwood (1971) suggested a method of fading which i n t u i t i v e l y would seem to be p a r t i c u l a r l y "relevant" to the l e f t - r i g h t discrimination. In Underwood1s study, Ss were required on each t r i a l to indicate which of the f i v e s t i m u l i shown was not the "same" as the r e s t . There were four p r i n c i p a l phases i n t h i s "oddity" task. In the f i r s t phase the st i m u l i were arrows extending to the l e f t or r i g h t from the mid-point of v e r t i c a l l i n e s the ovals were replaced by c i r c l e s . The c i r c l e s , however, were s t i l l i n the middle of the v e r t i c a l l i n e . In the f i n a l " c r i t e r i o n " phase there were four sets of "oddity" In the second phase the arrows were In the t h i r d phase 42 43 problems, one set for each of the four combinations b-d, p-q, b-q, d-p. In t h i s phase the c i r c u l a r part of the l e t t e r s was i n the usual p o s i t i o n at the bottom of the stalk . To reach c r i t e r i o n Ss had to achieve 6 successive correct responses or 9 correct responses i n 10 t r i a l s i n each of the four sets. Although this method did not i n any way provide " e r r o r l e s s " learning (for example, one S took 395 t r i a l s to reach c r i t e r i o n ) , 8 out of 10 Ss reached c r i t e r i o n . This study can be c r i t i c i z e d i n that although a pretest was used, a reinforced baseline l e v e l of performance was not established for the Ss and there was no control group. As the Ss were f i v e and six years of age, the l i k e l i h o o d that many of them could have learned the discrim-i n a t i o n with a usual reinforcement procedure might have been high, p a r t i c u l a r l y as there were many t r i a l s spread over a number of sessions. The present study incorporated arrow discriminations into a fading program. A two-choice matching-to-sample task was used. Subjects who d i d not reach c r i t e r i o n on baseline t r i a l s involving the b-d sti m u l i were assigned to the control and experimental groups. The control Ss received the b-d sti m u l i throughout the t r a i n i n g t r i a l s . The progressive Ss f i r s t received matching-to-sample t r i a l s involving arrows pointing e i t h e r to the l e f t or to the r i g h t . Subjects reaching c r i t e r i o n on the arrow stimuli received a fading program involving a gradual progression from arrows to f i n a l letter-forms. The Ss who did not reach c r i t e r i o n on the arrow discriminations received matching-to-sample t r i a l s with simultaneously moving arrows. The arrows i n t i t i a l l y were moved v e r t i c a l l y and l a t e r h o r i z o n t a l l y . The distance of the horizontal movement of the arrows gradually was reduced u n t i l the arrows were stationary. The Ss then received the previously mentioned fading sequence from stationary arrows to f i n a l letter-forms. 2. Method Subjects The Ss were 25 c h i l d r e n from Vancouver and Burnaby day care centres. Shown i n Table 1 are the age and sex of each of the Ss. The Ss are l i s t e d i n 3 groups, Baseline Only, Control, and Progressive. These groupings are the r e s u l t of the experimental procedure. The ages of the Baseline-Only Ss ranged from four years, four months to six years, four months with a median age of f i v e years, seven months. The ages of the nine Control and nine Progressive Ss ranged,,respectively, from four years, 45 TABLE 1 Age and Sex of Subjects i n the D i f f e r e n t Conditions Subject Number Sex Age Years Months Baseline Only Control Progressive 7 Female 4 4 2 Male 4 4 1 Male 5 4 3 Male 5 7 5 Female 5 9 6 Female 5 11 4 Male 6 4 9 Male 4 3 11 Male 4 5 14 Female 4 5 13 Male 4 8 12 Male 4 ,10 8 Female 5 0 10 Female 5 5 15 Male 5 10 16 Male 5 10 23 Female 4 3 20 Male 4 7 17 Female 4 9 19 Female 4 9 21 Male 4 9 22 Male 4 9 18 Male 5 9 25 Female 5 9 24 Male 6 0 three months to f i v e years, ten months and from four years, three months to s i x years, zero months. The median ages of the Control and Progressive groups were four years, ten months and four years, nine months. Materials and Apparatus The l e t t e r s b-d or p-q, arrows pointing to the l e f t and r i g h t , and the s t i m u l i for the fading sequence were printed i n matching-to-sample form on white cardboard cards. For the b-d, p-q, and arrow stimuli there were cards f o r each of the four combinations of the two possible samples and the two possible positions of the correct match on the card (either on the l e f t or on the r i g h t ) . The cardboard cards measured approximately 12.0 inches by 16.0 inches. Figure 3 shows the various phases i n the fading sequence from thick stationary arrows to the f i n a l l e t t e r forms. The l e t t e r s and numbers i n brackets indicate, respectively, the abbreviated name of each phase and the number of steps i n each phase. In the width phase (Wid) the width of the arrows was reduced over 6 steps. Then, i n the separation phase (Sep) of 5 steps, the v e r t i c a l separation between the shafts of the arrows was increased. The stalk of the arrows then was introduced gradually i n the'/stalk phase (Sta), which involved 10 steps. In the next phase, the head phase (Hea) of 7 steps, the head of the arrows was gradually 47 r V,'i(K-.i\. reduced 3J. • Width' reduced tlu:oir?n 0 . 6 3 " , 0 . 4 4 " , 0.31 " , 0.19",0 . 09 " , to 0 . 0 2 " . Vertical distance between shafts i s 5 . 5 0 " . t o Separation, of arrows increased, LSfS>-tSX- Vertical distance between shafts increased through 5 . 75 " ,6.SO",7 .00 " , 7 . 5 0 " to 7 - 7 5 " . 16 Stajlk introduced (5ta , 1 0 ) . •Stalk length increased through , 0 . 2 5 " , 0 . 4 4 " , 0 . 8 1 " , 1 . 5 0 " , 2 . 1 9 " , 2 . S 3 " , 3 . 5 3 " , 4 . 2 5 " , 5 - 7 5 " to 7 . 0 0 " . Of stalk . lcnrrth,' 2 5 3 i s below shaft. to 'Circles forncd (Cir.15) . Perpendicular1 distance frosu ends of extensions to stalk increased through 0.13", 0.30",0.63",0.75",0.83", 1.00",1.13",1.25",1.33", 1.44",1.50",1.56",1.63", 1 . 6 9 " to 1.75". to -> Shaft reduced through ~2. 2 2 . 0 0 " , 1 . 7 5 " , 1 . 5 0 " , 1 . 2 5 " , • 1 .CO",0 .75 " ,0.63",0 .50 " 0 . 3 8 " ,0,31",0.25 " ,0.19", 0.13",0.06" t o ' 0 . 0 0 " . " 5", -) 6 to neA?Lj".2!L!?ilfLcL (Hca.7Jj.. Arrow head gradually rounded u n t i l i t Coras 0 . 4 2 " of circuufer-enee of c i r c l e of 2.00" radius. Centre of c i r c l e i s 1.50" frou stalk. ( t ) ( to d D a Figure 3. Summary and i l l u s t r a t i o n of the phases i n the progression of the sti m u l i from thick stationary arrows to f i n a l letter-forms. The l e t t e r s and numbers i n the brackets indicate, respectively, the abbreviation f o r each phase and the number of steps i n each phase. 48 rounded. Then i n the shaft phase (Sha) of 16 steps the length of the shaft of the arrow was reduced u n t i l the shaft disappeared. In the f i n a l phase, the c i r c l e s phase (Cir) involving 15 steps, the c i r c u l a r part of the l e t t e r s was completed by increasing gradually the length of two extensions from the stem outward u n t i l the extensions met. The moving arrows were arranged by having arrow forms move along s l i t s i n a 48.0 by 12.0 inch plywood board. An upper s l i t i n the board 2.5 inches from the top measured 44.0 inches by 0.6 inch. Four and one-half inches below the s l i t were two s l i t s each 22.0 inches by 0.5 inch. The lower s l i t s were separated h o r i z o n t a l l y by 2.0 inches. Arrows made of thick black cardboard were inserted into the s l i t s so that the head of the arrow was i n front of the s l i t while the shaft of the arrow showed through the s l i t . The arrows measured 4.0 inches i n length. Each side of the head of the arrow was 2.7 inches long. The angular separation of the head of the arrow was approximately 86 degrees. The width of both the head of the arrow and the st a l k which showed through the s l i t was 0.6 inches. I t was arranged so that the E could move three arrows simultaneously from behind the plywood. Procedure 49 Figure 4 provides a summary of the procedure which i s outlined below. Baseline t r i a l s and general procedure. Subjects were brought i n d i v i d u a l l y into a private room i n the day care centre and seated i n a chair so that the E was facing the c h i l d across a table. Each c h i l d had baseline t r i a l s which involved matching the normal p-q forms for the f i r s t 5 t r i a l s and the f i n a l b-d forms f o r the remaining t r i a l s . When the f i r s t card was shown, the c h i l d was asked to "point to the picture which was or d i d the same thing or was l i k e the picture on the top". The c h i l d was t o l d that the f i r s t picture he pointed to would be considered h i s f i n a l choice. He was instructed that he would be t o l d when he was correct or incorrect ( i e . , " r i g h t " or "wrong") and every once i n a while, when he was correct, he would receive a candy which would be put i n a bag f o r h i s l a t e r consumption. I f the c h i l d chose the correct l e t t e r on the f i r s t t r i a l the E said " r i g h t " and put a candy i n the ch i l d ' s bag. If the c h i l d was in c o r r e c t another card i d e n t i c a l to the f i r s t was shown. The E said, "Try the other side". A f t e r the f i r s t successful t r i a l the next card of the series was shown and the E urged the c h i l d to " t r y to be r i g h t every time". The E then made no further explanatory comments unless the c h i l d said he d i d not 50 BASELINE ( A l l Ss) - Reinforced matching-to-sample with p and q - 5 t r i a l s . - Reinforced raatching-to-sample with b and d. T r i a l s u n t i l c r i t e r i o n of 10 successive correct responses or cutoff of 25 t r i a l s , ( i f 3 correct on t r i a l 25, more t r i a l s u n t i l error made or c r i t e r i o n reached). Ss 1 to 7, who )-reached c r i t e r i o n , received transfer t r i a l s . Remaining Ss randomly assigned to either Control or Experimental Condition. Control Condition (Ss 8 to 16) - Reinforced matching-to-sample with b and d. T r i a l s u n t i l c r i t e r i o n of 10 successive correct responses or cutoff of 150 t r i a l s ( i f 5 correct on t r i a l 150, more t r i a l s u n t i l error made or c r i t e r i o n reached). "Experimental "Condition "CSs 17 to 25) - Reinforced matching-to-sample with stationary arrows. T r i a l s u n t i l c r i -t e r i o n of 8 successive correct responses or cutoff of 10 t r i a l s (If S correct on t r i a l 10, more t r i a l s u n t i l error made or c r i t e r i o n reached). Ss 17 to 20, v/ho did not reach arrow c r i t e r -i o n , received moving arrow program and then fading series f rom -t h i c k stationary arrows (^ «#) to b-d. After moving arrow program/ probe t r i a l s with b-d approximately every 15 t r i a l s (generally at end of a phase). I f S i n -correct on probe, more t r i a l s with b and d u n t i l S reached c r i t e r i o n of 10 successive correct responses or made an error and fading resumed. Ss 21 to 25, who reached arrow c r i t e r -ion, received fading series from t h i n stationary arrows (<~ ~^ ) to b-d. Probe t r i a l s with b-d approximately every 15 t r i a l s (gen-e r a l l y at end of a phase). If S in c o r r e c t on probe, fading r e -sumed. I f S correct on probe, more t r i a l s with b and d u n t i l S reached c r i t e r i o n of 10 successive correct responses or made an error and fading resumed. TRANSFER ( A l l Ss) <= — -Reinforced matching-to-saraple with p and q. T r i a l s u n t i l c r i t e r i o n of 10 successive correct responses or cut o f f of 10 t r i a l s ( I f S correct on t r i a l 10, more t r i a l s u n t i l error made or c r i t e r i o n reached). Figure 4. Summary of experimental procedure 51 understand the task or exhibited such behaviour as pointing to the sample stimulus or both comparison s t i m u l i when the card was presented. I f the c h i l d behaved i n t h i s manner, the preceding in s t r u c t i o n s were repeated. Baseline t r i a l s continued u n t i l the Ss reached a c r i t e r i o n of 10 consecutive correct matches on the b-d stim u l i or u n t i l they received 30 baseline t r i a l s . I f an S was correct on t r i a l 30, he received further t r i a l s u n t i l he made an error or reached c r i t e r i o n . The Ss who reached c r i t e r i o n on the baseline t r i a l s were eliminated as i t was considered that t h e i r behaviour had come under the control of the l e f t - r i g h t aspect of the l e t t e r s p r i o r to the experiment or during the baseline t r i a l s . The Ss who did not reach c r i t e r i o n were assigned randomly to the experi- . . mental and control groups. The assignment to one of the groups occurred immediately upon the S's completion of the baseline t r i a l s . Then, the S immediately received further t r a i n i n g t r i a l s appropriate to the p a r t i c u l a r condition to which he had been assigned. The baseline and tra i n i n g t r i a l s had a number of features i n common. A fixed r a t i o (1:3) reinforcement schedule was used throughout. The st i m u l i were presented so that the l e f t - r i g h t p o s i t i o n of the correct match followed a Gellermann series. Whether the correct matches 52 were facing inward or outward was randomized. Each stimulus presentation was shown u n t i l the S made a choice. With a l l the sti m u l i presented on cards, the i n t e r t r i a l i n t e r v a l was approximately 3 seconds. During the i n t e r t r i a l i n t e r v a l the E put the candy i n the bag and noted the correctness of the response. Control procedure. Reinforced matching-to-sample t r i a l s with the b-d sti m u l i continued for the Control Ss u n t i l the S.s reached a c r i t e r i o n of 10 consecutive correct responses or u n t i l they received, i n addition to the baseline t r i a l s , 150 t r i a l s . I f the S was correct on t r i a l 150, t r i a l s were continued u n t i l the S reached c r i t e r i o n or made an error. Progressive procedure. In the Progressive group the Ss were given 10 reinforced t r i a l s of arrow cards following the baseline t r i a l s . The Ss who reached a c r i t e r i o n of 8 successive correct matches on the arrow cards began the fading series at the Sep phase. The Ss who did not reach t h i s c r i t e r i o n started the fading series with moving arrows. The arrows f i r s t were moved v e r t i c a l l y and then, h o r i z o n t a l l y . Figure 5 shows how the movement was accomplished. The two upper diagrams show fo r the f i r s t step of the movement fading, the beginning positions of the arrows (shaded) before the movement began, and the 53 end p o s i t i o n (unshaded) a f t e r the movement. The v e r t i c a l movements were always of the maximum length while the horizontal movements were decreased gradually over six steps u n t i l the arrows were stationary i n the centre of the board. The steps by which movement was reduced were 1.00, 0.56, 0.28, 0.11, 0.02, and 0.00 of the maximum movement of 18.0 inches. Questioning of Ss i n p i l o t studies indicated that c h i l d r e n sometimes considered that the correct comparison stimulus was the one closer to the sample. As can be seen i n the two upper diagrams i n Figure 5, the correct compar-ison i s usually c l o s e r to the sample during most of the movement. To control for the fa l s e hypothesis that the comparison closer to the sample was the correct one, the r e l a t i v e positions of the stimuli were modified on some t r i a l s i n the manner indicated i n the lower two diagrams of Figure 5. As can be seen i n these diagrams the correct comparison stimulus during much or a l l of the movement was farther away from the sample than was the correct comparison. This v a r i a t i o n i n arrow positioning was used every t h i r d or fourth t r i a l of the v e r t i c a l movements and every t h i r d or fourth t r i a l on the f i r s t step of the horizontal movements and on the 0.28 step of the horizontal movements. With the moving arrows the Ss were instructed not 54 Figure 5. I l l u s t r a t i o n of how the arrows were moved. Parts A and B show fo r the f i r s t step the beginning positions of the arrows (shaded) before the movements began and the end positions (unshaded) a f t e r the movements. Parts C and D show fo r the f i r s t step the modifications of the beginning positions (shaded) and the end positions (unshaded) of the arrows. Such var i a t i o n s i n the positions of the arrows were shown at int e r v a l s to negate any possible f a l s e hypothesis that the comparison stimulus c l o s e r to the sample stimulus for most of a l l the run was the correct match. 55 to indicate t h e i r choice of the moving comparison stimuli u n t i l the E said "now" which occurred a f t e r most of the movement had occurred. In the case of the moving arrows the i n t e r t r i a l i n t e r v a l was approximately 7 seconds. The Ss had to achieve 6 successive correct choices when the arrows were moving v e r t i c a l l y before they received the horizontal movement t r i a l s . During the horizontal movement t r i a l s , the Ss were required to make 4 successive correct choices on the longest runs before the fading began. A f t e r the moving fades were completed, the Ss were required to make 4 successive correct choices on the stationary arrows. Then the Ss started the fading sequence beginning with the Wid stage of the cards. The Ss were required to achieve 4 successive correct choices on the f i r s t step of the Wid stage of the cards before they began to progress through the sequence. I f an S requiring the moving arrows did not reach the Wid stage i n approximately one-half hour, t r i a l s were recommenced i n approximately two hours. A l l Progressive Ss received a backup procedure following errors. Other than f o r step 1 of the sequence, the S received the fading step appropriate to the preceding step, n-1. This backup procedure continued through h-2, n-3, etc. u n t i l the S was correct or u n t i l the f i r s t step 56 was reached. Then he received the step succeeding the one upon which he was correct. The previously mentioned randomization over t r i a l s of the l e f t - r i g h t p o s i t i o n of the correct match and the d i r e c t i o n i n which the comparison st i m u l i faced ( i e . , inward or outward) was modified when an error was made; i n the backup procedure the S received the same stimulus presentation for each step as he had received i n i t i a l l y . For a l l the Progressive Ss t r i a l s continued u n t i l the Ss achieved 10 successive correct t r i a l s on the f i n a l b-d s t i m u l i or u n t i l they received, i n addition to the baseline t r i a l s , 150 t r i a l s . If the S was correct on t r i a l 150, t r i a l s were continued u n t i l the S reached c r i t e r i o n or made an error. The Progressive Ss were "probed" with the f i n a l value of b-d approximately every 15 t r i a l s . The non-reinforced probe t r i a l s were made to determine whether the Ss required the complete fading sequence i n order to discriminate between the l e t t e r s . As the Ss d i f f e r e d with regard to the number of t r i a l s required i n the various phases and stages, where the probes occurred v a r i e d for each S. In general, however, probes were made only at the end of a phase of fading. The probes were not done during the moving arrow t r i a l s . I f the S was correct on the f i r s t probe, more probes immediately were given u n t i l the S made 10 successive correct matches or made an error. I f an error was made, the S received the stimulus presentation which had been given immediately preceding the probe t r i a l s ; then the fading proceeded. Transfer task. A f t e r a c r i t e r i o n number of correct discriminations or after the S had reached a cutoff number of t r i a l s , a matching-to-sample transfer task involving the l e t t e r s p and q was given. There were 10 reinforced (1:3) t r i a l s with a c r i t e r i o n of 10 successive correct responses. If the S was correct on t r i a l 10, the number of transfer t r i a l s was extended u n t i l the S made an error or reached c r i t e r i o n . 3. Results C r i t e r i o n was reached by 7 of the 25 Ss during the baseline t r i a l s . Eight of the nine Progressive Ss and one of the Control Ss reached c r i t e r i o n on the b-d tra i n i n g t r i a l s . A c h i square tested indicated that the difference between the groups i n the number of Ss who reached c r i t e r i o n was s i g n i f i c a n t (p<.01). A l l Ss who reached c r i t e r i o n on the b-d stimuli reached c r i t e r i o n on the p-q transfer t r i a l s . As the number of t r i a l s completed by the Ss varied, i n order to permit comparison between the Control and Progressive Ss, the number of correct responses times 100 divided by the number of t r i a l s ( i e . , the percentage of t r i a l s correct) was computed for each S. Figure 6 shows these percentage scores f o r the Control and Progressive Ss i n both the baseline and t r a i n i n g t r i a l s . The height of the striped bar indicates the baseline score while the height of the darkened bar indicates the t r a i n i n g score. The height of the u n f i l l e d bar above the darkened area indicates f o r the Progressive Ss the score computed by omitting the r e s u l t s of the stationary arrow series which immediately followed the baseline t r i a l s . This arrow series served to determine whether or not the Progressive Ss required the moving arrows. As can be seen, the median baseline score was 50 percent (S 13) i n the Control group and 52 percent (S 20) i n the Progressive group. The scores ranged from 37.5 percent (S 11) to 53 percent (S 12) i n the case of the Control Ss and from 30 percent (S 17) to 60 percent (S 25) i n the case of the Progressive group. In the t r a i n i n g stage f o r the Control group the scores ranged from 25.3 percent (S 8) to 100 percent (S 16) with a median of 50.9 percent (S 12). For the Progressive group, the scores f o r the t r a i n i n g trials,.when a l l t r i a l s were included, ranged 59 100| 90-LU 00 z o D_ (J) LU CC LU o cc 80-70-60-BE 5 ( > ^ o o k UJ 40-p 8 3 0 - " 2 0 -10-^ T - B A S E L I N E T R I A L S - A R R O W T R I A L S E X C L U D E D " A R R O W T R I A L S I N C L U D E D kits 1 A k l I % 'A HI 1 I M 1 J % I • B kl m I It I k# •S'J' I 4 I I / f -"I rt i f jay y%1 i k< 1 k l k l i S8- Sg S10 S11 Sl2 S 1 3 S14 Sl5 S 1 6 C O N T R O L C O N D I T I O N S17 Si8 S19 S20 S2I S22 S23 S24 S25 P R O G R E S S I V E C O N D I T I O N Figure 6. The percentage of t r i a l s on which Control and Progressive Ss responded c o r r e c t l y . For each Progressive S two percentages are shown. One percentage was calculated by including the arrow t r i a l s which immediately followed baseline t r i a l s i n the Progressive condition. The other percentage was calculated by excluding the arrow t r i a l s . <7 from 42 percent (S 17) to 95 percent (S 24) with a median of 87 percent (S 20). Thus only S 17, the one Progressive S who did not reach c r i t e r i o n , responded at e s s e n t i a l l y chance l e v e l . The other eight Progressive Ss (Ss 18 to 25) had scores greater than or equal to 77 percent; four of these Ss (Ss 22 to 25) had scores equal to or greater than 90 percent. When the stationary arrow t r i a l s were omitted the scores improved for a l l Ss. These scores ranged from 42.5 percent (S 17) to 100 percent (S 24, S 25) with a median of 91.4 percent (S 20). Thus seven of the Progressive Ss (Ss 18, 19, 20, 22, 23, 24, 25) had scores of greater than 90 percent while the others, S 17 and S 21, had scores of 42.5 percent and 77.9 percent respectively. Figure 7 (p. 64) shows the cumulative number of correct responses of the Baseline Only Ss, those Ss who reached c r i t e r i o n on the baseline t r i a l s , the Progressive Ss, and the Control Ss. The short v e r t i c a l l i n e s below the graphs indicate the beginning of the baseline, t r a i n i n g , and transfer stages. In the case of the Progressive group, the beginning of each phase of fading also i s indicated. The symbols under the ordinate r e f e r to the stage or phase and the s t i m u l i used. A key explains the various symbols. For the Progressive Ss, the short v e r t i c a l l i n e s 61 above the graphs indicate the probe t r i a l s . A t y p i c a l performance of a Control S i s shown i n the record of S 13. As can be seen he had 15 correct responses i n 30 baseline t r i a l s and 80 correct responses i n 152 t r a i n i n g t r i a l s . Like a l l Ss who did not reach c r i t e r i o n on the b-d t r i a l s he did not reach c r i t e r i o n on the p-q transfer task. In the case of S_ 16, the one Control S who reached c r i t e r i o n , he had 15 correct responses i n the 30 baseline t r i a l s . However, on t r i a l 1 of the t r a i n i n g session h i s successful run began. Four of the experimental Ss (Ss 17 to 20) required the moving arrows. The record of S 17 indicates the performance of the one experimental S who d i d not reach c r i t e r i o n . In the f i r s t t r a i n i n g session of 20 t r i a l s on the Ver M phase, she was correct on 14 t r i a l s . In the next session, involving 10 t r i a l s on the Ver M phase and 10 t r i a l s on the Hor M phase, she was correct on, respectively, 3 and 4 t r i a l s . She was the only S who required two sessions and who never learned the moving arrow discrimination. The record of S 18 shows the performance of one of the Ss who received the moving arrows and reached c r i t e r i o n . He had 16 correct responses i n 30 baseline 62 t r i a l s . On the Ver M phase he made two errors before he made the run of six successive correct responses. One other S (S 20) experienced a si m i l a r minor d i f f i c u l t y at the beginning of the Ver M phase. T y p i c a l of a l l the Ss who received the moving arrows and reached c r i t e r i o n , S 18 had a l i t t l e d i f f i c u l t y at the f i n a l step of the Hor M phase when movement had ceased. He required backups to the next to l a s t step two d i f f e r e n t times before he could achieve four successive correct responses on the stationary arrows on the board. He then made no addi t i o n a l errors. He was the only S requiring the moving arrows who did not require the complete program. Three other Ss (Ss 21, 22, 25) did not require the complete program. Subject 25 and S 18 reached c r i t e r i o n on the f i r s t probe which occurred during or a f t e r the Sep phase. Two other S_s (Ss 21, 22) reached c r i t e r i o n on the second probes a f t e r , r e s p e ctively, the Hea and Sha phases. The record of S 23 shows the performance of an S who did not require the moving arrows. She and S 24 were the two Ss not requiring the moving arrows who required the complete fading program. She had 14 correct responses i n 30 baseline t r i a l s . Her only d i f f i c u l t y i n the t r a i n i n g stage was i n the Hea phase, where she made two errors i n both step 4 and step 5 of the phase. Two other Ss (S 19, S 22) made one error on t h i s phase, both at step 4. As can be seen on the record of S 21, t h i s S had d i f f i c u l t y i n the Arr-Sep phases. He i n i t i a l l y reached c r i t e r i o n on the A r r t r i a l s . However, as indicated by the number of errors i n the Arr-Sep phases he had d i f f i c u l t y i n commencing the fading. A backup procedure to the A r r phase was required three d i f f e r e n t times. On two of the three backups the S made one or more errors on the A r r phase before progressing to the Sep phase, where errors then were made a f t e r the f i r s t two backups. Only a f t e r the t h i r d backup was the S able to progress successfully through the Sep phase. An analysis of the various phases of fading was done by determining the number of Ss who were incorrect on greater than 10 percent of the t r i a l s on that phase. Due to the Ver M, Wid, Sep, and Hea phases having fewer than 10 steps, i t was possible f o r an S to make greater than 10 percent errors on these phases by making only one mistake. However, except f o r the two previously mentioned Ss who each made one error on the Hea phase, a l l the Ss having greater than 10 percent errors on a p a r t i c u l a r phase made more than one error on that phase. Only S 17, the S who did not reach c r i t e r i o n , made more than 5 errors on any stage. In the Ver M and Hor M phases three Ss KEY Bas Baseline p q ;d b Cr i Criterion Run b d Train Training (Control) b d Arr Arrows Stationary V e r M Vertical Movement HorM Horiz. Movement * •* W i d Width Reduced Sep Separ. of Arrows Sta Stalk introduced Hea Head Rounded Sha Shaft Reduced IJ i\-\3 Cl Cir Circles Formed t> cf -b d Tran Transfer p q PROGRESSIVE Ss S 1 7 - S 2 5 CD S3 S4 S5 S6 S7 Figure 7. Cumulative correct responses of Ss i n the Control condition, i n the Progressive condition, aricl i n the Baseline-Only group, the group that reached c r i t e r i o n on baseline t r i a l s . The short v e r t i c a l l i n e s below the graphs indicate the beginning of the stages of the t r i a l s . In the case of the Progressive condition, these l i n e s also indicate the beginning of the phases of fading. The symbols under the graphs r e f e r to the various stages and phases. These symbols are explained i n the key, which shows a summary and i l l u s t r a t i o n of each stage and phase. Except for.the Sep phase, the key provides examples of only the comparison s t i m u l i . 65 (Ss 17, 18, 20) made errors on more than 10 percent of the t r i a l s . In the Hea phase three Ss (Ss 19, 22, 23) made greater than 10 percent errors. In each of the other phases only one S made errors on more than 10 percent of the t r i a l s . 4. Discussion General Summary and Comment The r e s u l t s c l e a r l y demonstrated the effectiveness of the fading method. Although the Control and Progressive Ss showed e s s e n t i a l l y the same baseline performance, the number of Progressive Ss who reached c r i t e r i o n greatly exceeded the number of Control Ss who reached c r i t e r i o n . Furthermore, the great majority of the Progressive Ss who reached c r i t e r i o n made very few errors. When the i n i t i a l 10 stationary arrow t r i a l s were excluded from the computation, the great majority of the experimental Ss who reached c r i t e r i o n were correct on more than 90 percent of the t r i a l s . When the computation included the i n i t i a l arrow series, the percentage correct scores of a l l Ss were reduced, but only those of the Ss requiring the moving arrows went below 90 percent. In the case of the Ss who required the moving arrows, i t would seem that the score r e s u l t i n g from excluding the arrow t r i a l s would be the most appropriate indicator of how well Ss who require the moving arrows perform with regard to learning without errors. Presumably, the stationary arrow t r i a l s served mainly to indicate that these Ss required the moving arrows and were not e f f e c t i v e i n ensuring that the behaviour of these Ss came under the control of l e f t - r i g h t d i r e c t i o n a l i t y . However, the Ss who d i d not require the moving arrows acquired a l e f t - r i g h t discrimination during the stationary arrow t r i a l s . Thus i t would seem that for these S_s the percentage correct score r e s u l t i n g from including the stationary arrow t r i a l s would be appropriate. Thus, consideration of the percentage correct score including the stationary arrow t r i a l s f o r Ss not requiring the moving arrows and excluding these t r i a l s f o r Ss requiring the moving arrows shows that a l l but one of the Progressive Ss who reached c r i t e r i o n made fewer than 10 percent e r r o r s . There appears to be no generally accepted cut-off point where performance i s no longer considered to be " e r r o r l e s s " . However, i t would seem that Ss who respond c o r r e c t l y on more than 90 percent of the t r i a l s would be considered to have exhibited " e r r o r l e s s " learning. Hence, with regard to the reduction of errors, the fading technique of the present study would seem to compare favourably with the previously mentioned successful fading studies of Terrace (1963a, 1963b), Sidman and Stoddard (1967), and Moore and Goldiamond (1964). It i s important to note, however, that two Progressive Ss made a number of errors; one of these Ss was incorrect on 20 percent of the t r i a l s , while the other performed at e s s e n t i a l l y chance l e v e l and never proceeded past the f i r s t step of the moving arrow phase. There i s a s l i g h t p o s s i b i l i t y that the r e l a t i v e l y poor performance of the former S may have resulted from h i s reaching c r i t e r i o n on the arrow t r i a l s by chance. Thus he may have been " i n c o r r e c t l y " assigned to the group which received only stationary s t i m u l i . When errors on the f i r s t step of fading required backups several times to the arrow s t i m u l i , t h i s S experienced d i f f i c u l t y with the arrows. With regard to the Progressive S who did not progress past the v e r t i c a l movement discrimination, i t i s possible that her lack of success was due, at l e a s t i n part to her exh i b i t i n g behaviour which would seem incom-patible with the behaviour required for successful performance of the task. She was considered by her day care centre teacher to be above average i n i n t e l l i g e n c e and willingness to cooperate. However, a f t e r making several errors i n the t e s t i n g s i t u a t i o n she repeatedly stood up and sat down. She often looked about the room 68 and asked several times when she might leave the room. On many t r i a l s she d i d not look at the arrows as they moved. Consideration of the findings of Rudel and Teuber (1963) indicate that the occurrence of children who do not learn e a s i l y an up-down discrimination would be infrequent i n a large sample of children. As may be r e c a l l e d from Chapter I, i t was found that fewer than seven percent of preschool childr e n from a low socioeconomic background did not learn an up-down discrimination within 50 t r i a l s . I t i s possible that the percentage of children who do not learn to discriminate between arrows moving i n an opposite up-down d i r e c t i o n might be considerably lower than the percentage found by Rudel and Teuber. Presumably a discrimination involving moving .arrows would be easier than one involving stationary U-shapes. Furthermore, although there i s no evidence, i t seems possible that the matching-to-sample task employed i n the present study would be easier than the simultaneous discrimination task employed by Rudel and Teuber. As mentioned e a r l i e r , i t was found that only 7 of the 25 Ss achieved c r i t e r i o n on the baseline t r i a l s under a simple d i f f e r e n t i a l reinforcement procedure and only 1 of 9 Control Ss was successful with a possible 150 more such t r i a l s . These findings are i n agreement with the previously c i t e d normative studies (e.g., Davidson, 1935; Gibson, et a l . , 1962; Newhall, 1937), which indicated that the behaviour of the majority of preschool c h i l d r e n was not under the control of the l e f t - r i g h t aspect of s t i m u l i . Furthermore, the f a i l u r e of t h i s procedure to teach many children the discrimination i n a matching-to-sample task p a r a l l e l s Rudel and Teuber's (1963) s i m i l a r lack of success with young children i n a simultaneous discrimination task. However, the f a c t that approximately one-third of the c h i l d r e n were successful under the reinforcement procedure supports the c r i t i c i s m of the studies of B i j o u (1963) and Underwood (1971) concerning t h e i r lack of baseline t r i a l s and/or control groups. Only with some form of control procedure can the effectiveness of a fading method be determined. The findings of the present study suggest that a control group might be omitted i f there were a s u f f i c i e n t number of baseline t r i a l s . Consideration of the performance of the Baseline-Only and Control Ss indicate that Ss who reach c r i t e r i o n do so within a l i m i t e d number of t r i a l s ; any lengthy extension of t r i a l s does not seem to produce an increase i n the number of Ss who reach c r i t e r i o n . While a l l the Baseline-Only Ss began t h e i r f i r s t t r i a l of the c r i t e r i o n run by t r i a l 28, one 70 Control S d i d not begin the c r i t e r i o n run u n t i l the f i r s t t r i a l a f t e r the baseline t r i a l s . I t would seem that a baseline procedure of approximately 40 t r i a l s would be s u f f i c i e n t to eliminate most, i f not a l l , of the Ss who would learn the l e f t - r i g h t l e t t e r discrimination under a d i f f e r e n t i a l reinforcement procedure i n a matching-to-sample task. It i s i n t e r e s t i n g to note that any Ss who were successful i n the b-d discrimination, whether i n the Baseline-Only group, or the Control and Progressive groups, successfully transferred to the p-q discrimination task. As might have been predicted, once the c h i l d has come under control of the d i r e c t i o n a l component of the f i r s t p a i r , the second p a i r , which are very s i m i l a r to the f i r s t , should pose no problem. I t would be i n t e r e s t i n g to see whether the c h i l d , a f t e r acquiring such l e t t e r discriminations, would be able to transfer to d i r e c t i o n a l discriminations between two and three dimensional figures which are quite d i s s i m i l a r to the arrow and l e t t e r forms. Reasons for the Success of the Program and Possible  Modifications The effectiveness of the present fading method with regard to both the number of Ss who learned the l e f t -r i g h t l e t t e r discrimination and the very few errors made 71 by these Ss may have resulted from the fading being done i n small steps along what would seem to be a very "relevant" continuum, or perhaps more appropriately stated, a very relevant sequence. Thus t h i s program was a successful combination of the techniques of Lawrence (1952) and Terrace (1963a, 1963b) i n that the fading was along presumably the "relevant continuum" and was done very gradually. It would seem that the use of arrows was an important factor contributing to the success of t h i s fading sequence. Results indicate that the d i r e c t i o n a l component of arrows i s considerably more " d i s t i n c t i v e " than the d i r e c t i o n a l aspect of l e t t e r s . Many Ss who did not reach c r i t e r i o n on the baseline t r i a l s learned e a s i l y to discriminate between the stationary arrows. Furthermore, even when Ss c o r r e c t l y matched the arrow stimuli they d i d not necessarily reach c r i t e r i o n on the f i r s t probe t r i a l s . Even Ss who v e r b a l l y had mentioned the d i r e c t i o n a l difference between the arrows did not necessarily respond d i f f e r e n t i a l l y to the l e t t e r s on the f i r s t probes. The use of movement with the arrows for Ss who i n i t i a l l y did not discriminate between the stationary arrows presumably increased the " d i s t i n c t i v e n e s s " of the d i r e c t i o n a l component of the arrows. Due to innate factors of the S and/or h i s 72 previous learning, movement would seem to be a very " d i s t i n c t i v e " aspect of s t i m u l i . Furthermore, by having the arrows move v e r t i c a l l y rather than h o r i z o n t a l l y on the f i r s t stage of the fading, t h i s program possibly ensured that the S's behaviour would come more quickly under the control of the orient a t i o n of the s t i m u l i . As previously mentioned i n Chapter I, Rudel and Teuber (1963) found that very few preschool childr e n learned to discriminate between l e f t - r i g h t reversals within 50 t r i a l s ; however, the great majority of these c h i l d r e n learned e a s i l y an up-down discrimination. Presumably the use of movement with arrow stimuli would f a c i l i t a t e t h i s up-down discrimination learning, and once t h i s discrimination was learned the S would transfer successfully to the horizontal movement discrimination and on through the program. It i s an i n t e r e s t i n g conjecture that the f i r s t part of the program from moving arrows to stationary arrows may p a r a l l e l i n some abbreviated form the stages that a c h i l d normally would pass through i n h i s preschool years. > Due to both the reinforcement contingencies of everyday l i v i n g and the p o s s i b i l i t y that movement may be innately " d i s t i n c t i v e " to humans, the c h i l d who has not been taught formally the various discriminations s t i l l 73 may pass through i n order at l e a s t three of four stages of discrimination learning. These stages involve learning to discriminate between (a) objects moving i n an opposite up-down d i r e c t i o n , (b) stationary objects facing i n an opposite up-down d i r e c t i o n , (c) objects moving i n an opposite l e f t - r i g h t d i r e c t i o n , and (d) stationary objects facing i n an opposite l e f t - r i g h t d i r e c t i o n . Possible modifications of the fading sequence probably would include shortening some phases and lengthening others. Elimination of steps seems p a r t i c u l a r l y f e a s i b l e i n the case of the Sha •••!)) and C i r ( [) -b) phases, each of which involved very gradual fading over 15 or more steps. As a l l Ss requiring the moving arrows made errors when the movement ceased, an improved program might increase the number of steps so that the t r a n s i t i o n from movement to no movement would be very gradual. Furthermore, consideration of the number and locus of errors i n the Hea phase ( ••• r-) ) would suggest that more steps be added, p a r t i c u l a r l y at l e v e l 4. Comparison of the Program with Other Training Methods Although i t i s d i f f i c u l t to compare across studies the success of various fading methods i n teaching the l e f t - r i g h t l e t t e r discrimination, i t would seem that the effectiveness.of the present method, as r e f l e c t e d i n the 74 number of successful Ss and the percentage of correct t r i a l s , iv-as greater than the success of the fading methods of the p i l o t studies and the studies of Bi j o u (1963), Karraker (1968), and Schutz (1964). In considering the effectiveness of the present fading method i h r e l a t i o n to methods-other than fading, i t i s important to question why the Ss instead of receiving a fading method simply should not be to l d v e r b a l l y about the d i r e c t i o n a l aspects of the l e t t e r s . There are at leas t three important responses to t h i s question. F i r s t , i t i s possible that due to a v a r i e t y of reasons the c h i l d ' s behaviour might not be under the control of verbal i n s t r u c t i o n s . For example the c h i l d of foreign parentage may not speak or understand the language of i n s t r u c t i o n . In addition, p a r t i c u l a r l y i n teaching situations involving large classes, the c h i l d often w i l l be attending to st i m u l i other than the teacher's i n s t r u c t i o n s . In these types of si t u a t i o n the fading sequence adapted to teaching machines, would seem to be very useful. Secondly, even i f chil d r e n do acquire the b-d discrimination with verbal i n s t r u c t i o n s , they may make more errors i n reaching c r i t e r i o n than do chi l d r e n who receive a fading method. While i t i s not cl e a r how many errors were made by the three Ss i n J e f f r e y ' s (1958) 75 study who reached c r i t e r i o n a f t e r verbal i n s t r u c t i o n s , consideration of the number of t r i a l s to reach c r i t e r i o n would suggest that they may have made a number of errors. As mentioned i n Chapter I, possibly undesirable behaviour may r e s u l t from learning with err o r s . According to Terrace (1966) i t has been r e g u l a r l y observed that Ss who have learned with errors e x h i b i t emotional responses i n the presence of the negative stimulus. Furthermore, even a f t e r a stable near perfect rate of responding has been reached a f t e r learning with e r r o r s , there may be reoccur-rences of incorrect responding (Terrace, 1963a). In the case of Ss who have learned with no or few err o r s , these emotional responses and disruptions i n performance have not been observed. F i n a l l y i t i s possible that verbal instructions would not be e f f e c t i v e i n teaching the discrimination to many children, even i f they are under the control of the ins t r u c t i o n s . In Je f f r e y ' s (1958) study i t was found that the majority of the control group, who received instructions including gestures concerning the d i r e c t i o n a l i t y of the l e t t e r s every 10 or 20 t r i a l s , d i d not learn to i d e n t i f y c o r r e c t l y the l e t t e r s . By allowing errors, verbal instructions may permit responding to come under the control of st i m u l i other than thoser/which are experimentally 76 manipulated. As mentioned previously, under c e r t a i n conditions permanent f a u l t y performance may r e s u l t due to an intermittent reinforcement e f f e c t (Terrace, 1966). It i s in t e r e s t i n g to compare the effectiveness of the present method with the techniques involving arm movement employed by J e f f r e y (1958) andi Hendrickson and Muehl (1962). Other than the method of the present study, the techniques of J e f f r e y and Hendrickson and Muehl appear to be the only ones which have been shown experimentally to teach a high percentage of Ss to make the l e f t - r i g h t discrimination. In order to make the comparison between the arm-moving technique and the fading method of the present study i t would seem appropriate to compare only Je f f r e y ' s method with the fading technique; i n the case of the experiment of Hendrickson and Muehl, the e f f e c t s of arm movement were confounded by the presence of arrows below the l e t t e r s . The comparison between the method of J e f f r e y and that of the present study i s d i f f i c u l t i n that the studies varied with respect to a number of va r i a b l e s . For example, the c r i t e r i o n tasks were quite d i f f e r e n t . The Ss i n Jeff r e y ' s study were required to name the l e t t e r s "Jack" and " J i l l " , while the Ss of the present study did a matching-to-sample task. In the present study only Ss who had f a i l e d to achieve c r i t e r i o n on the baseline t r i a l s were employed, while i n J e f f r e y ' s study there was no baseline elimination procedure. Even with t h i s d i f f i c u l t y i n making a comparison i t might be safe to say that the success rate of the present study, 8 successful Ss i n 9 Ss, was roughly equivalent to J e f f r e y ' s rate of 13 successful Ss i n 14 Ss. Furthermore, as mentioned i n Chapter I, consideration of the number of t r i a l s required to reach c r i t e r i o n by the Ss i n J e f f r e y ' s study who had the arm moving stage would suggest that the performance of at l e a s t some of these Ss was not without a number of errors. Hence, even a f t e r c r i t e r i o n was reached, t h e i r performance could have been subject to reoccurrences of in c o r r e c t responding similar to the bursts of incorrect responding found by Terrace (1963a). In the case of the present method, these disruptions of correct responding presumably would not occur with the majority of Progressive Ss, who performed with minimal errors. Theoretical Implications It i s important to consider the extent to which the findings of the present study are i n agreement with Orton's theory concerning the r e l a t i o n s h i p of cerebral dominance and perception. Orton's attempt to explain i n 78 terms of incomplete cerebral dominance why many children's behaviour i s not c o n t r o l l e d by the l e f t - r i g h t aspect of st i m u l i has been by f a r the most important t h e o r e t i c a l influence on research on reversals i n reading. Orton (1937) postulated that cerebral dominance must be attained before the c h i l d i s able to discriminate between objects facing i n an opposite l e f t - r i g h t d i r e c t i o n . Before t h i s dominance i s achieved the? stimuli are registered and interpreted by both hemispheres and thus apparently cannot be " c o r r e c t l y processed" by the brain. A number of studies have been conducted to determine whether there i s a r e l a t i o n s h i p between incomplete cerebral dominance and reading d i s a b i l i t y . T y p i c a l l y , incomplete cerebral dominance has been infer r e d from mixed l a t e r a l preference; that i s , one side of the body i s not dominant i n a l l a c t i v i t i e s . An example of mixed l a t e r a l preference occurs when a person who writes with h i s r i g h t hand kicks with h i s l e f t foot or sights with h i s l e f t eye. Generally i n these studies, standardized tests of reading accuracy, speed, and/or comprehension have been used to measure reading d i s a b i l i t y . A number of c r i t i c i s m s of these studies can be made. It seems obvious that poor performance on the tests used i n these studies has a very complex etiology; thus a poor score does not necessarily r e s u l t from the f a c t that the S's behaviour i s not under the control of the l e f t - r i g h t aspect of l e t t e r s and words. Furthermore, incomplete cerebral dominance probably cannot be i n f e r r e d r e l i a b l y from mixed l a t e r a l preference. At least for some a c t i v i t i e s , environmental influences such as peer pressure or the design of equipment probably determine to a great extent which side the c h i l d uses. Thus i t i s not su r p r i s i n g that the r e s u l t s of these studies are c o n f l i c t i n g . While ce r t a i n studies (e.g., Harris 1957, Orton 1937) have indicated a r e l a t i o n s h i p between reading d i s a b i l i t y and mixed l a t e r a l preference, others have not (e.g., Balow, 1963; Belmont and B i r c h , 1965). With regard to the present study i t would seem clea r , however, that incomplete cerebral dominance was not responsible f o r the poor performance of at least most of the Ss who did not reach c r i t e r i o n . It i s d i f f i c u l t to conceive that the Progressive Ss' problems with incomplete cerebral dominance as r e f l e c t e d by poor performance on the baseline t r i a l s could be "cured" by the fading sequence i n such a short time. It i s quite possible as suggested by Money (1966) that the poor performance of childr e n on l e f t - r i g h t d iscrimination tasks may be due to the f a c t that preschool ch i l d r e n are not reinforced for responding d i f f e r e n t i a l l y to differences i n orientation, p a r t i c u l a r l y to l e f t - r i g h t o r i e n t a t i o n . Possibly the reason that young ch i l d r e n learn the up-down discrimination before the l e f t - r i g h t d iscrimination (Rudel and Teuber, 1963) i s that up-down discriminations are reinforced e a r l i e r . A c h i l d who turns an ink bottle or a haul-truck f u l l of d i r t upside down over a rug may be punished and c e r t a i n l y w i l l not be reinforced; yet the c h i l d who rotates the same object 180 degrees about the v e r t i c a l axis w i l l be neither reinforced nor punished. Generally c h i l d r e n are not required to make the l e f t - r i g h t discrimination u n t i l they begin to read; then'.after years of being i r r e l e v a n t , l e f t - r i g h t d i r e c t i o n a l i t y becomes a c r i t i c a l dimension to which the c h i l d must learn to attend. This learning probably i s f a i r l y d i f f i c u l t i n that the c h i l d ' s previous experience of three or four years must be revised. 5. Summary This study compared a fading sequence, which at least i n t u i t i v e l y seemed to be "relevant" to the f i n a l r i g h t - l e f t discrimination of the l e t t e r s b-d, to a control procedure In which the f i n a l forms of the l e t t e r s were presented throughout the t r i a l s . The Ss who did not reach c r i t e r i o n on the baseline matching-to-sample t r i a l s with the l e t t e r s b-d were assigned to the Progressive and Control 81 groups. A l l Progressive Ss received a series of matching-to-sample t r i a l s involving arrows pointing e i t h e r to the l e f t or r i g h t . Those who reached c r i t e r i o n on the arrow sti m u l i received a fading program involving a gradual progression from arrows to f i n a l letter-forms. The Ss who did not reach c r i t e r i o n on the arrow series received matching t r i a l s involving simultaneously moving arrows. The arrows were moved v e r t i c a l l y and then h o r i z o n t a l l y . The distance of the h o r i z o n t a l run was reduced gradually u n t i l the arrows were stationary. Then, the Ss received the fading sequence from stationary arrows to f i n a l l e t t e r -forms. Eight of the nine Progressive S_s reached c r i t e r i o n on the b-d discrimination. Seven of the Progressive Ss who reached c r i t e r i o n responded c o r r e c t l y on more than 90 percent of the t r i a l s while the eighth S was correct on 77 percent of the t r i a l s . Only one of the nine Control Ss reached c r i t e r i o n . A l l Ss reaching c r i t e r i o n on the b-d discrimination transferred to the p-q discrimination. The discussion of the r e s u l t s of the study included reference to the following points: 1. The effectiveness of the present fading sequence as compared to both other fading sequences, including those involving l e f t - r i g h t l e t t e r discriminations, and methods 82 other than fading. 2. Possible reasons for the "deviant" performances of the Progressive S_ who did not reach c r i t e r i o n and the Progressive S who made a f a i r number of errors while learning the discrimination. 3. Reasons as to the effectiveness of the program and any possible modifications. 4. Theoretical implications of the r e s u l t s . REFERENCES Baker, R.A. and Osgood, S.W. Discrimination transfer along a pitch continuum. Journal of Experimental Psychology, 1954,48,241-246. Balow, I.H. L a t e r a l dominance c h a r a c t e r i s t i c s and reading achievement i n the f i r s t grade. Journal of Psychology, 1963,55,323-328. Belmont, I. and B i r c h , H.G. L a t e r a l dominance, l a t e r a l awareness and reading d i s a b i l i t y . C h i l d Development, 1965,36,57-71. Bijou, S.W. Studies i n the experimental development of l e f t - r i g h t concepts i n retarded children using fading techniques. International Review of Research i n Mental  Retardation, 1963,3,66-95. Davidson, H. A study of the confusing l e t t e r s b,d,p, and q. Journal of Genetic Psychology, 1935,47,458-468. Fellow, B.J. The discrimination process and development. Oxford: Pergamon Press, 1966. Gibson, E., Gibson, J . , Pick, A., and Osser, H.A. Develop-mental study of the discrimination of l e t t e r - l i k e forms. Journal of Comparative and P h y s i o l o g i c a l  Psychology, 1962,55,897-906. H a r r i s , A.J. L a t e r a l dominance, d i r e c t i o n a l confusion and reading d i s a b i l i t y . Journal of Psychology, 1957,44, 283-294. Hendrickson, L.N. and Muehl, S. The e f f e c t of attention and motor response pretraining on learning to discriminate b and d i n kindergarten c h i l d r e n . Journal of Educational Psychology, 1962,53,236-241. H i l l , M.B. A study of the process of word discrimination i n i n d i v i d u a l s beginning to read. Journal of  Educational Research, 1936,29,487-500. 83 84 House, B.J. and Zeaman, D. Transfer of a discrimination from obiects to patterns. Journal of Experimental  Psychology, 1960,59,298-302. J e f f r e y , W.E. Variables i n e a r l y discrimination learning: Motor responses i n t r a i n i n g a l e f t - r i g h t discrimination. C h i l d Development, 1958,29,269-275. Jenkins, H.M. and Harrison, R.H. E f f e c t of discrimination t r a i n i n g on auditory generalization. Journal of  Experimental Psychology, 1960.59-:246-253. Karraker, R.J. Teaching beginning readers to d i s t i n g u i s h between s i m i l a r l e t t e r s of the alphabet. United States O f f i c e of Education Project No. 6-1955-24, 1968. Kennedy, H. Reversals, reversals, reversals! Journal of  Experimental Education, 1954,23,161-170. Lawrence, D.H. The transfer of a discrimination, along a continuum. Journal of Comparative and P h y s i o l o g i c a l  Psychology, 1952,45,511-516. Money, J . On learning and not learning to read. In J . Money (Ed.) The disabled reader. Baltimore: John Hopkins Press, 1966. Pp. 21-40. Moore, R. and Goldiamond, I. E r r o r l e s s establishment of v i s u a l discrimination using fading procedures. Journal of the Experimental Analysis of Behaviour, 1964,7,269-272. Newhall, S.M. I d e n t i f i c a t i o n by young childre n of d i f f e r e n t l y oriented v i s u a l forms. C h i l d Development, 1937,8,105-111. Orton, S.T. Reading, writing and speech problems i n  children. New York: Norton, 1937. Piaget, J . and Inhelder, B. The c h i l d * s conception of  space. New York: Humanities Press, 1956. Popp, H.M. V i s u a l discrimination of alphabet l e t t e r s . The Reading Teacher, 1964,17,221-226. 85 Rudel, R. and Teuber, H. Discrimination of d i r e c t i o n of l i n e i n chi l d r e n . Journal of Comparative and  Phys i o l o g i c a l Psychology, 1963,56,892-898. Schutz, R.E. A c q u i s i t i o n of alphabet-letter discrimination with and without errors. Paper read at the Annual Meeting of the American Psychological Association, Los Angeles, C a l i f o r n i a , 1964. Sidman, M. & Stoddard, L.T. The effectiveness of fading i n programming a simultaneous form discrimination f o r retarded c h i l d r e n . Journal of the Experimental  Analysis of Behaviour. 1967,10,3-15. Swanson, R. and Benton, A.L. Some aspects of the genetic development of r i g h t - l e f t discrimination. C h i l d  Development, 1955,29,123-133. Teegarden, L. C l i n i c a l i d e n t i f i c a t i o n of the prospective non-reader. C h i l d Development, 1932,3,346-358. Terrace, H.S. Stimulus c o n t r o l . In W.K. Honig (Ed.) Operant behaviour: areas of re search and ap p l i c a t i o n . New York: Appleton-Century Cr o f t s , 1966. Pp. 271-344. Terrace, H.S. Discrimination learning with and without errors. Journal of the Experimental Analysis of  Behaviour, 1963a,6,l-27. Terrace, H.S. Err o r l e s s transfer of a discrimination across 2 continua. Journal of the Experimental  Analysis of Behaviour, 1963b,6,223-232. Underwood, B.J. A program f o r grapheme discrimination f o r preschool children. Unpublished doctoral d i s s e r t a t i o n , U n i v e r s i t y of Arizona, 1971. Wechsler, D. & Hagin, R. The problem of a x i a l r o t a t i o n i n reading d i s a b i l i t y . Perceptual and Motor S k i l l s , 1964,19,319-326. Wilson, F.T. & Flemming, C.W. Reversals i n reading and wri t i n g made by pupils i n kindergarten and primary grades. Journal of Genetic Psychology, 1938,53,3-31. Wolfe, L. An experimental study of reversals i n readi American Journal of Psychology, 1939,52,533-561. 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0101591/manifest

Comment

Related Items