University of Southern California Dissertations and Theses

The Use Of The Orienting Reflex To Test The Zeaman And House Theory Of Attention

(USC Thesis Other)

The Use Of The Orienting Reflex To Test The Zeaman And House Theory Of Attention

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 72-17,516 STREIFEL, John Arthur, 194<+- THE USE OF THE ORIENTING REFLEX TO TEST THE ZEAMAN AND HOUSE THEORY OF ATTENTION. University of Southern California, Ph.D., 1972 Psychology, experimental University Microfilms, A XEROX Company, Ann Arbor, Michigan THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. THE USE OP THE ORIENTING REFLEX TO TEST THE ZEAMAN AND HOUSE THEORY OF ATTENTION by John Arthur Streifel A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Education) January 1972 U I N I V t K S I I T U f S U U 1 r i L K I N ^ M L - i r V J I f i N I M THE GRADUATE SC H O O L U NIVERSITY PARK LOS ANGELES, C A LIFO R N IA 9 0 0 0 7 This dissertation, written by .......Johjn,Arth^^ .............. under the direction of h..L? . . . Dissertation C o m mittee, and a pproved by all its members, has been presented to and accepted by The G radu ate School, in partial fulfillment of require ments of the degree of D O C T O R O F P H I L O S O P H Y Dean Drt^....?ebruarY..197_2 DISSERTATION COMMITTEE - CT . I , / PLEASE NOTE: Some pages may have indistinct print. Filmed as received. University Microfilms, A Xerox Education Company TABLE OB CONTENTS Page LIST OP TABLES.................................. . V LIST OP FIGURES.................................. vi Chapter I. INTRODUCTION AND SURVEY OP THE LITERATURE . 1 PURPOSE OP THIS INVESTIGATION ............ 2 CRITIQUE OP THE ZEAMAN AND HOUSE THEORY OP . ATTENTION ORGANIZATION OP REMAINDER OP THIS CHAPTER . 4 THE ORIENTING REPLEX ...................... 5 History of the Orienting Reflex Description of the Orienting Reflex Changes Occurring During an OR The Relation of the OR to Attention and Consciousness The OR as a Measure of Attention An Example of Physiological Recording of the OR to Specific Stimuli Individual Differences in the OR and Learning Measurements of the OR in GSR Research An Example of Behavioral Indicators of the OR THE ZEAMAN AND HOUSE THEORY OP ATTENTION . . 19 The Backward Learning Curves Reanalysis of Data Via Backward Learning Curves Interpretation- of Backward Learning Curves The Zeaman and House's Model of Attention Research Prior to the Development of Attention Theory The Effects of Intelligence Relationships Between MA and Intelligence Beginning to the Attention Hypotheses Intelligence Versus Attention Investigation ii Chapter Page Stimulus Factors Affecting Attention Subjects Apparatus Dis criminanda Procedure for Assignment to Groups Experimental Procedure Measurement Operational Definition HYPOTHESES................................ 43 Hypotheses H1 through H. Hypothesis He through Hq Hypothesis Hq through H-p Hypothesis H^ through H^ DATA ANALYSIS.............................. 60 STATISTICAL PROCEDURES .................... 62 III. RESULTS.................................. 64 Hypothesis H1 Hypothesis Hp Hypothesis H, Hypothesis H? Hypothesis H? Hypothesis H? Hypothesis H« Hypothesis Hq Hypothesis Hq Hypothesis H^n Hypothesis hJ. Hypothesis H-p Hypothesis h J, Hypothesis H-? Hypothesis H-c Hypothesis H^ RECAPITULATION OP THE RESEARCH 32 II. METHOD 36 iii Chapter IY. DISCUSSION AND CONCLUSION Overview of Study Suiimlary of Purpose Summary of Findings GSR Comparisons of Past Learners and Yoked Controls Behavioral Comparisons of Past Learners and Yoked Controls GSR Comparisons of Past and Slow Learners Behavioral Comparisons of Past and Slow Learners Discussion of Results Summation of the Present Study APPSNDIX 113 BIBLIOGRAPHY 11? iv LIST OP TABLES Table Page 1. Summary of Analysis of Variance for Past Learners and Yoked Controls (GSR data). . . 67 2. Mean GSR Magnitudes of Past Learners and Yoked Controls to Stimulus Situations During the Different Levels of Reinforcement............................ 67 3. Summary of Analysis of Variance for Past Learners and Yoked Controls (behavioral data) ............... 71 4. Mean Percentage of Eye Orientations of Past Learners and Yoked Controls to Both Stimulus Situations Buring the Learning of a Discrimination...................... 71 5. Summary of Analysis of Variance for Past and Slow Learners (GSR data).............. 75 6. Mean GSR Magnitudes of Past and Slow Learners During the Learning of a Discrimination................... 75 7. Summary of Analysis of Variance for Past and Slow Learners (behavioral data) .... 83 8. Mean Percentages of Eye Orientations of Past and Slow Learners to Both Stimulus Situations During the Learning of a Discrimination .......................... 83 V LIST OF FIGURES Figure Page 1. Comparisons of Average GSR Magnitudes of Fast Learners and Yoked Controls to the Stimulus Situations During the Different levels of Reinforcement (level of reinforcement equals level of learning by fast learners)........ 68 2. Comparisons of Averaged Percentage of Eye Orientations of Fast Learners and Yoked Controls to the Stimulus Situations During the Different Levels of Reinforcement Equals Levels of Learning by Fast learners 72 3. Comparisons of Averaged GSR Magnitudes of Fast and Slow Learners to the Stimulus Situations at the Different Levels of Percentages of Correct Responding (levels of learning)............... 76 4. Comparisons of Averaged Percentages of Eye Orientations of Fast and Slow Learners to the Stimulus Situations During the Different Levels of Percentages of Correct Responding (levels of learning) ......... 84 5. Backward Learning Curves for all Fast and Slow Learners............. 83 6. Averaged Backward Learning Curves for Individual and Pooled Groups. . ........... 86 yi CHAPTER I INTRODUCTION AND SURVEY OP THE LITERATURE Since the 1950’s, research on the construct of attention is beginning to appear more frequently in the investigations by behavioristic psychologists, particularly iin the study of discrimination learning (Trabasso, 1966). As a perceptual process, attention is commonly studied indirectly; overt behavioral changes evidence its occurrence. Zeaman and House theorized that attention is the factor determining the speed of learning a two-choice discrimi nation by mentally retarded children. Learning to attend is shown by a sudden progressive improvement in correct responses. Slow learners require more trials before suddenly improving, but once they do, their rate of impro vement is the same as fast learners. Inferences about attentional differences are based on behavior and explanations of behavioral differences are based on attention. A method of measuring attentional differences, independent of consequent behavior, is necessary to clearly demonstrate the causality theorized by Zeaman and House. As a tool for analysis the OR (orienting reflex), to be discussed below, refers to a readiness or alerting response I to the onset of stimuli, novel^or significant. As a tool “ 2 for analysis, the OR can serve as an independent objective |measure of perceptual processes. An independent objective measure of perceptual processes is necessary if circularity I is to be avoided in the conceptualizing of the relationships :of the causal factors involved. As Maltzman and Raskin (1965, p 1) pointed out ”... concepts of perceptual processes typically suffer from a methodological short coming. Hypothetical perceptual processes or responses are | inferred from the same observed behavioral changes which they are assumed to influence." PURPOSE OP THIS INVESTIGATION The purpose of this investigation was to give clarity to the Zeaman and : House theory of the role of attention in mental retarded two-choice discrimination learning. Specifically, the purpose of this investigation was to determine the nature of the attentional differences between fast and slow learners that controls the speed of learning. CRITIQUE OP THE ZEAMAN AND HOUSE THEORY OP ATTENTION A critique of the Zeaman and House treatment of |attention shall now be presented to demonstrate the need | for clarification. In their theorizing about the parti cular role of attention, based on the shapes of the 3 backward learning curves, Zeaman and House (1963) postulated | that "the length of the initial flat chance level stages of the performance is controlled primarily by an attention process" (p 162). It would be helpful if Zeaman and House were more explicit about what they considered to be the attention process. Again, based on the appearance of the shapes of the backward learning curves, they postulated that the S did not sample relevant cues on every trial but "... may have to learn to attend to the relevant stimulus i dimension" (p 166). This indicates that the attention process they assumed to be controlling the learning rate is selective in character (as opposed to the position held by the continuity theories that attention is non-selective in character). However, this is an unsatisfactory characteri zation of the attention process involved. Since attention is the cornerstone of their theory, Zeaman and House should consider what differentiates the fast learners from the slow learners in terms of their attending to stimuli. Given that all learners attend to some degree and since rates of learning differ among Ss, there should be quantitative (i.e., levels) and/or qualitative (strategies for acquiring information) differences in the attending behavior of fast and slow learners. The study reported here investigated the quantitative attentional differences !in terms of the OH measured autonomically (GSR) and [qualitatively in terms of the OR measured behaviorally ■ ....... ' ... 4 (eye orienting to stimuli), I ORGANIZATION OF REMAINDER OF THIS CHAPTER The remainder of this chapter concerns itself with ;the following areas: the orienting reflex, the Maltzman land Raskin research, the Galvanic Skin Response and the 'Zeaman and House theories and research. Finally, the chapter closes by relating the essential aspects of the literature specifically to the conduct of the present study. The discussion of the OR is composed of a review of 'its history, a description of the phenomenon, the relation |of the OR to attention and consciousness and lastly, how it serves as a measure of attention. The review of the Maltzman and Raskin research gives an account of how the OR has been used to study awareness, a perceptual process. The review of the GSR research is to give acquaintance with its proper use as a measure of the OR. The presentation of the Zeaman and House research begins with a discussion of their theory, followed by a review of their investigations conducted prior to development of their theory of the role of attention. THE ORIENTING REFLEX The OR, a readiness response to novel or significant ■stimuli, is defined in terms of the behavioral and physiological changes concomitant with its occurrence. The more the OR is investigated, the more complex and conscious related the reaction becomes (Razran, 1961). History of the Orienting Reflex The OR has been known to exist since the turn of the century, but only since the 1950's has the phenomenon become popular as a subject for investigation (Razran, 1961). In 1902 Pavlov dropped the word "psychic" from his terminology and replaced it with the word "conditional". This change in terminology was, in part, due to Pavlov's growing disinterest with stimuli that precondition sensory experiences and stimuli that postcondition image experien ces. Another reason for the change was Pavlov's belief that to be truly objective one must avoid psychological formu lations as explanations of phenomenon. Pavlov was apparent ly equating psychology with its methodology which, at that time, was introspectionism. The avoidance of explanations derived by introspecting, which became his position in 1904, anticipated the major cornerstone of behaviorism which John Watson in 1913 articulated. Pavlov, in 1902, did give some recognition to the stimuli that precondition sensory experiences. These stimuli evoked in Ss (dogs) an observable, but many times, weak reaction. These reactions by his dogs were termed "orienting", the "what is it", the "investigatory" or the "attitudinal" reflex and the terms 6 interchangably by Pavlov (Razran, 1961). Pavlov noted that ithe rate of conditioning was facilitated if the OR occurred land was of some intermediate intensity, neither too weak nor too strong and as the CR began to appear, the OR began i ito disappear. | Pavlov missed an opportunity to make a significant : contribution to the studies of consciousness because as -Razran (1961) pointed out, these characteristics of the OR I1 '... might have been correlated with conscious experiences by saying, for instance, that the conditioned stimuli must arouse adequate but not overwhelming consciousness and that consciousness tends to disappear as habit develops” (p 109). After 1902 Pavlov was no longer interested in consciousness nor anything that smacked of introspectionism and consequently, research on the OR was relatively stagnant after its discovery. Since 1950, with changes in Russian psychology and political ideology, consciousness has become more popular as a subject for study and, staying within the Pavlovian conditioning paradigm, the OR has become the subject of a great deal of investigation (Razran, 1961). Description of the Orienting Reflex The OR involves a constellation of physiological changes which vary in magnitudes and appearances with the intensity and quality of the stimuli that evoke it and the level of alertness of the organism prior to the reception of stimuli (Lynn, 1966). The antecedent condition for the OR is the onset or i ! discovery of stimuli that are novel, novel in pattern, significant (Lynn, 1966) or absent (Badia and Defran, 1970). Changes in stimulation that are qualitative, intensive or temporal may also evoke the OR. The consequent; !of the OR is an increase in arousal, increase in sensitivity or perception (Lynn, 1966) and increased learning ability (Maltzman and Raskin, 1965). ! The OR is a nonspecific reaction which distinguishes from other reflexive behavior such as defensive, adapta- tional and specific reflexes which are evoked only by particular classes of stimuli (Maltzman and Raskin, 1965). Changes Occurring During an OR When the OR is initiated, sensory receptors increase in sensitivity and are directed towards the stimulus; other ongoing activity is momentarily arrested. Physiological and vegetative changes take place: respira tion and heart rate are delayed, the pupil dilates, skeletal muscle tones increase, the cortical alpharftythm is depressed, the GSR occurs and cephalic vasodilation and peripheral vasoconstriction take place (Lynn, 1966). I The Relation of the OR to Attention and Consciousness The OR in activity and function corresponds to the conditions present when attention occurs. The initial I ’ condition for the onset of attention is a stimulus change. |In a like manner, the initial condition for the OR is a I 8 I |stimulus change. The phenomenon of attention, paying i attention, etc., refers to conscious activities. The l physiological and behavioral measures of the OR correspond to attentional or conscious activities. Pillsburg, in 1908, reported that measures taken of people while "paying attention" indicated an increase in cerebral blood volume and a decrease in blood volume in the limbs. Soviet researchers adopted these responses as measures of the OR. Razran (1961) reported many Soviet psychologists have found positive correlations "between the scope and duration of elicited ORs and the scope and duration of conscious experience" (p 119). Lynn (1966) referred to the OR as "... a mechanism for paying attention to novel stimuli" (p 1). As a mechanism for attention, the,OR functions to facilitate the reception of stimuli by increasing the sensitivity of sense receptors and directing the organism to the source of stimulation. The OR does not manage stimuli as other reaction patterns (i.e., defensive, sexual and alimentary reaction patterns), it merely reacts to the presence of stimuli. The OR is "more preparatory than consumatory and preadaptive rather than adaptive" (Razran, 1961, p 114). These characteristics refer to a readiness or attentional response by an organism prior to the immediate encounter with the impinging stimulation. These character istics suggest a general controlling role of the OR and a I 9 r i |function like that of cognition (Razran, 1961). According to the model proposed by Sokolov (1960), Ithe OR represents conscious activity; the cortex initiates its excitation and inhibition. The cortex analyses |incoming stimuli for their novelty or significance and upon their detection activates the reticular formation which produces an OR. The persistence and strength of the OR correlates with the unfamiliarity and difficulty of the learning task. Grastyan (1961) reported that when the meaning of stimuli were unknown, the OR was persistent; the more difficult the discrimination to be made between stimuli, the stronger the OR. The OR as a Measure of Attention The OR served as a measure of attention to the extent that it corresponds to objective events occurring during the phenomenon. Jeffrey (1968) pointed out that the process of attention is made up of "an alerting or arousal component, a receptor orienting component and an internal cue selecting component" (p 325). The moment to moment changes in arousal are amenable to physiological recording and receptor orienting can be measured in terms of observable behavior. However, moment to moment changes in cue selection cannot be unambiguously measured physiologi cally when stimuli impinging on the organism are of the same modality and are configuerated (dimensions of the stimulus |occurring innately together, i.e., shape, color, size and position)• An Example of Physiological Recording of the OR to 10 Specific Stimuli The work of Hernandez-Peon, Scherrer and Jomvet (1956), recorded the occurrence of the OR to specific stimuli (or cues) in demonstrating the involvement of the reticular formation in blocking of non-significant stimuli !to the brain. In their study, evoked potentials of two Isensory modalities (sound, smell and vision) were recorded via recording electrodes. During the experiment evoked potentials from the cochlear nucleus of the ear were recorded from a cat. During the auditory stimulation, a imouse in a bottle was then placed in the cat's visual field, producing a blocking of the evoked potentials in the auditory sensory pathway. Y/hen the visual stimulus was removed, evoked potentials to the auditory stimulus returned to the previous level. The procedure was also carried out with the smell of fish which was; aghin intro duced while the cat was exposed to the auditory stimulus. iDuring olfactory stimulation, evoked potentials in auditory sensory pathways were again blocked and returned after ithe olfactory stimulus was removed. Individual Differences in the OR and learning The Maltzman and Raskin (1965) research to be presented demonstrated the OR can serve as an aid to imeasuring perceptual processes and that individual I 11 ! jdifferences in magnitudes of the OR exist which relate I idirectly to learning performance* I | Possibly, because of their political ideology, the |closest the Russians got to studying individual differences I in the OR was the report by Voronin and Sokolov (1960) which indicated that not all the different physiological changes involved during the OR occurred in everyone. Maltzman and Raskin (1965)» concerned with individual 'differences in the magnitudes of the OR, conducted studies which investigated the relationship between learning and awareness as a function of the OR. They made two important assumptions concerning individual differences in the OR. | The first assumption was that individuals will give different magnitudes of the OR to the same stimulus condi tions. The second assumption was that "the OR is related tc the discrimination of complex stimuli" (p 2). Raskin (1963) studied individual differences in the OR in relationship to differences in performance in semantic conditioning of the GSR. The results indicated that there are individual differences in the magnitude of the OR as measured by GSR recordings to the CS (the word plant was ipresented ten seconds before the onset of the UCS [110 db iwhite noise, 1 second}). A within S. control design was used and Ss were divided into four groups on the basis of level of orienting (high and low) and instructional conditions I (uninformed and partially informed). The high orienters jwere those whose GSR was below the median response level. iThe uninformed group received no information about when the UCS would occur. The partially informed group was told that a loud noise would follow certain words. Inspection iof the data indicated that the high orienters were "reliably superior" to low orienters in conditioned GSR magnitudes to both the GS and the neutral words, under both instructional conditions (Maltzman and Raskin, 1965). When the data were examined in terms of the instruc tional differences, the high orienters gave "reliably superior" GSRs to the CS under both instructional conditions, To insure that the GSR magnitudes were not in fact a measure of emotionally based drive rather than a measure of the OR, the Ss were given a short form of the MAS (Taylor Manifest Anxiety Scale) and the scores were correlated with the GSR magnitudes of the Ss at the conclusion of the experiment. "Magnitudes of the GSR taken as the measure of emotionally based drive were not reliably correlated" (Maltzman and Raskin, 1965, p 4). Also, the Ss were assigned on the basis of MAS scores to high and low anxiety groups (and were subdivided into the appropriate uninformed and partial ly informed groups) for a reanalysis of the data. A reliable relationship between MS scores and semantic conditioning was noted, but it was clearly different from the reliable relationship between the levels of GSR magnitudes and superiority of conditioning. That is, when |GSRs were plotted as a measure of the OR, the various groups in superiority of semantic conditioning ranked as ; follows: high orienter - partially informed, high orienter- ; i uninformed, low orienter - partially informed, and low i orienter - uninformed. When the GSR scores of the Ss j jreassigned on the basis of MAS scores were plotted they ranked in superiority of semantic conditioning as follows: jhigh anxious - partially informed, low anxious - uninformed, low anxious - partially informed, and high anxious - uninformed. When GSR is interpreted to represent emotionally based drive, one set of results is obtained. When GSR is ! I interpreted .to represent the OR, another set of results is obtained. Since GSR magnitudes were not "reliably corre lated" with MAS scores, the alternative conclusion of GSR representing the OR was accepted. Individual differences in the OR, which are directly related to learning performance, have been demonstrated. The question then becomes, what is the involvement of the OR in learning? The characteristics of the OR suggested to Maltzman and Raskin that it served to facilitate awareness and thus, learning. Regarding the role of awareness in learning, Barber (1963) described three positions that may be taken. Briefly, they are: (1) awareness is a byproduct of learning occurring after the acquisition of S-R associations, (2) |awareness occurs between stimulus and response and is responsible for the S-R association made and (3) awareness land learning interact. Maltzman and Raskin offered a fourth, position with supporting evidence that "awareness is a function of the OR" (1965, p 5). In the next study, awareness as a function of the OR was investigated during a semantic conditioning and generalization experiment. In the experiment loud white Inoise (UCS) was to condition vasoconstriction in the fingers to the word light (CS). During the semantic conditioning, filler words and the generalization test words (dark, heavy, lamp and soft) were also presented. At the conclusion of the experiment, all the words were rated as pleasant, neutral or unpleasant. After rating the words, an inquiry was made concerning words rated as unpleasant. In the semantic conditioning aspect of the experiment, almost all Ss indicated awareness by being able to identify the contingency: "The word light was followed by a loud sound". About half of the Ss rated one or more of the generalization words as being unpleasant. The word lamp was most often rated as unpleasant. The typical reason given was that the word lamp made the j 3 think of light. In the data analysis, these Ss were categorized as verbalizers. Non-verbalizers were those who could not state the relevant contingencies between the GS word light and one of the generalization ! words. Comparisons were then made of the vasoconstriction responses by the verbalizers and non-verbalizers. The 15 jvasoconstriction responses were plotted during both phases J of the experiment. Vasoconstriction was first taken as a ;measure of the OR to the first UCS following a seried of |adaptation words and later as a measure of semantic condi tioning and generalization. The graphs indicated that the verbalizers gave "reliably larger" vasomotor responses (ORs) than the non-verbalizers. However, vasoconstriction responses, as a measure of semantic conditioning and generalization, were not significantly greater in magnitude in the verbalizers over the non-verbalizers. The variabili ty of the vasomotor response was suggested as the reason for the failure of the verbalizers to show a significant ; difference in semantic conditioning. In a "replication" study, GSR and vasomotor responses were simultaneously conditioned to the same words used previously. The verbalizers now showed reliably superior semantic generalizations as measured by the GSR index. The vasomotor conditioning did ishow the same trend as in the previous experiment, but again, failed to show a reliable difference in semantic generalization between verbalizers and non-verbalizers (Maltzman and Raskin, 1965). A graphic representation of the data in the initial conditioning phase of the experiment, in which vasocon striction was the measure of the OR, the verbalizers j(later identified) showed "superior conditioning ... before I there was any contingency between the CS and generalization 16 words for them to verbalize" (Maltzman and Raskin, 1965, p 6)* Maltzman and Raskin concluded that the superior conditioning of the respondents and the ability to j |discriminate and verbalize the relevant contingencies were ;both a function of the magnitude of the OR within limits. Raskin, in 1963, obtained supporting evidence of the relationship between the OR and the ability to verbalize the generalization contingency. Brotsky (1964) failed to demonstrate a relationship between the magnitude of the OR and the ability to verbalize the generalized test contingen cies. Using highly discriminable training and generali zation test words, the conditional stimulus (word) was too ireadily discriminable for the college students who were able to verbalize the contingencies while showing varying magnitudes of the OR. This study pointed out that the complexity of the task and the method of measuring verbali zation of the contingencies are important variables in demonstrating the facilitation of high OR magnitudes in awareness and learning. Measurements of the OR in GSR research Studies of the OR have generally been in terms of changes in one or more of autonomic components; "BEG, GSR (or electrodermal changes), heart rate, digital pulse, volume and pressure, and breathing rate" (Uno and Grings, 1965, P 312). Of these various ways of measuring the OR, Ithe GSR component has been selected for this study. Voronin and Sokolov (1960) 8howed that the GSR was |the most reliable autonomic component to manifest an OR to ;the first presentation of a novel stimulus. Badia and iDefran (1970), heeding the suggestion by Stewart, Stern, ;Winokur and Fredman (1961) of the need for an analysis of GSR conditioning in terms of the OR, conducted two experi ments to examine the involvement of the OR in GR acquisition and in a second study, UCR diminution. The two major hypotheses of Badia and Defran were "that orienting responses result from classical conditioning test proce dures of both CS and UGS omission and that under certain conditions, OR, GR and UCR are inextricable" (p 171). In the first study, tone (CS) and light (UCS) were paired together at interstimulus intervals of .5 seconds for one group and five seconds for another. This was the typical GSR conditioning experiment, except for one difference. Both stimuli were of moderate intensity (50 db tone, 6 V/att light) which were capable of eliciting GSRs (CRs) but not conditioning (GRs). Fifteen training trials were given which produced a continuous decline in GSR magnitudes to both tone and light which indicated that habituation was taking place. During the test trials when light was omitted, GSR magnitudes increased significantly during both the .5 and five-second intervals. The second experiment and discussion were not relevant to this study land will not be presented (Badia and Defran, 1970). i Grings and Carlin (1966) instrumentally modified the GSR by making shock contingent upon the response to light (punishment group) and shock contingent upon not responding |to the light. The GSRs of the avoidance group were in«* creased in frequency and decreased in magnitude. The GSRs of the punishment group decreased in frequency, and decreased in magnitude more than the GSRs of the avoidance i group. Yoked control groups were used for the avoidance and I punishment groups to control for the effect of ISs' respon- sitivity and the number of reinforcements of the GSR. Schnidman (1970) instrumentally conditioned electrodermal ORs by making reinforcement contingent upon at least a + .25 mv skin potential change to a triangle stimulus during the five-second presentation. The reinforcements used were presentations of picture slides of landscapes, seascapes, animals or numbers. Bach picture slide presentation repre sented a ten cent bonus for participating in the experiment. Yoked control Ss were matched with experimental iSs on the basis of responsitivity to ten habituation trials. The iyoked control members received reinforcement on the same numbered trials as their counterparts but produced significantly fewer GSRs. An Example of Behavioral Indicators of the OR The study by Dodd and Lewis (1969), which examined the magnitude of the OR in children as a function of changes in color and contour, used behavioral measures rather than 19 |measures of autonomic components in measuring the OR. The i |indices used were fixation time (time spent looking at the stimulus) and frequencies of smiling, pointing and surprize. J The results indicated that increases in complexity (addition of color to achromatic pictures) produced | i significantly greater ORs than in the reverse operation, while increases in intensity (straight colored lines changed |to curved lines) produced greater ORs but not significantly different from ORs produced by changing curved lines to straight lines. THE ZEAEAN AND HOUSE THEORY OF ATTENTION On the basis of reanalyzed data from experiments conducted in 1960 and 1962, Zeaman and House (1963) formulated a theory about the role of attention in mentally retarded two-choice discrimination learning. The mechanism for reanalysis of the data was the backward learning curve. The backward learning curve was: also the major vehicle for presenting their theory. The backward learning curve will be discussed prior to the presentation of the Zeaman and House theory. [ The Backward learning Curves The backward learning curves are learning curves shifted to the right until the terminating point on a graph I for all ogival shape lines are the same. The slopes of the | 20 |cui*ves are then easily compared for rate of improvements. 'The number of trials required before the progressive rate of I improvement takes places by different Ss and are also easily compared. The equality in the rate of improvement by fast and slow learners and the variance in the number of trials ! required before the rate of improvement took place led Zeaman and House (1963) to the formulation of their theory. Reanalysis of Data Via Backward Learning Curves Zeaman and House constructed backward learning curves (a graphic technique developed by Hayes in 1953) from the data of two color-form discrimination experiments conducted in 1960 and 1962. In the 1962 study, the data were organized in homogeneous groups according to the number of days required to reach criterion of 20 correct responses in 25 trials. In the 1960 study, the data were organized into homogeneous groups according to MA and IQ (2-4 years with mean IQ of 30 and 4-6 years with mean IQ of 41) and whether they were learners or nonlearners. Inspection and statistical analysis of the backward learning curves revealed, in summary, that learning curves of the fast and slow learners differed only in the initial portion when the performance was at the chance level. The length of the line, in this portion of the graph, was proportionally longer for the slower learners, later, as |the line became ogival in shape, indicating a sudden i 21 I progressive improvement, the curves for the fast and slow ] |learners became statistically the same, indicating equal rates of improvement in performance. This observation led Zeaman and House to theorize that the "difference between fast and slow learners is not so much the rate at which improvement takes place, once it starts, but rather the inumber of trials for learning to start” (1963, p 163). Interpretation of Backward learning Curves Zeaman and House interpreted these learning curves not as indicating the usual gradual improvements with occasional decrements in which refinements in the responses were constantly being made, but rather as indicating that two phases were involved. The first phase indicating that a. chance level of responding in which there were no improve ments occurring until the second phase in which ”... instru- mental discriminative learning...” was taking place (1963, p 162). The initial chance level of the performance portion of the curve (first phase) became the prime interest of Zeaman and House which they believed to be "controlled primarily by an attention process” (p 162). This all-or-none characteristic, either learning takes place at a rate which is equal for both fast and slow learners or not at all, led Zeaman and House (1963) to perceive that a discontinuity was theorized to be "produced by two continuous processes." The first process was the !"••• learning to attend to the relevant dimension..." and | 22 the second, process was "... approaching the correct cue of j ithat dimension.*1 This two-process idea provided for the "possibility that the relevant cues were not attended to on every trial because the S may first have to learn to attend to the relevant stimulus dimension" (p 166). The Zeaman and House’s Model of Attention The belief in the existence of the "learning to iattend" process, based on the apparent discontinuity in the learning curves, sets Zeaman and House's conceptualizations apart from the quantitative discrimination theories of Burke and Estes (1957), Bush and Mosteller (1951), Restle (1955) and Atkinson (1958) which all assumed that the S_ samples relevant stimuli on every trial. Since the learning to attend process is lacking in their theories, their predictions of the learning curves are without abrupt transition (1963, p 166), Wyckoff's (1952) observing response model best accomodates the Zeaman and House's two-process idea which they further modified and extended. Generally, what Zeaman and House (1963) believed to be taking place during any j discrimination trial is as follows: at the beginning of every trial, a variety of stimulus dimensions, both relevant and irrelevant, are present. Only a portion of the available dimensions were effective as cues since the S's attention was assumed to be limited. Eor simplicity, Zeaman jand House postulated for their first submodel that only one 23 dimension was to be attended to at a time. For a more detailed description of what Zeaman and House believed to be taking place when a discrimination was ; being made, imagine "a somewhat wall-eyed £ 3 ... viewing a simultaneous presentation of two stimulus objects differing in color and form." At the time of choice, the S was either attending to the relevant dimension S (form in this case) or the irrelevant dimension S (color or position). If the S then made "the relevant observing response 0^, ... the specific values of cues of the form dimension ... were seen; if the S made the irrelevant observing response, the specie ! fic values of cues of the other dimensions were seen." Following 0.j, an instrumental response was me.de to one of the cues of the relevant dimension, "approaching the positive cues R-j, or approaching the negative cues R^; after Og, the £ 3 approached one of the cues of the irrele vant dimension and received no reward" (p 167). Zeaman and House admitted that in the previous description of what was taking place, none of the responses was observable. The only observable responses in the system were the overt picking up of the correct object Rc\ The probabilities of these overt responses were predicted by ! :their theory, which is a mathematical model of discriminal tion learning. i Research Prior to the Development of Attention Theory In this section issues confronted by Zeaman and 24 House prior to the formulation of their theory are dis cussed, The first set of studies reviewed were concerned with the general question: what type of information is used !to learn a discrimination? House, Orlando and Zeaman (1957) studied how the mentally defective child makes use of positive and negative cues in making visual discriminations. Positive cues were those which were involved in making a correct discriminatior. i and receiving a reward. The question that House, at, al., were trying to answer was whether the tendency to approach positive cues was the same for negative cues. Of the two existing views of the role of negative cues in making a discrimination (the negative cues were or were not used), they indirectly concluded that the negative cues were not used. The suggestion that "imbeciles make relatively little, or perhaps no use of the negative cue in a two- choice discrimination" prompted House and Zeaman (1958) to make a study of the effects of reward and non-reward in the discrimination learning of mentally deficient children, Harlow made a similar study with monkeys in which a forced positive or a forced negative response (one stimulus present with or without a reward) was made in the first trial. The responses on subsequent two-choice discriminations were the basis for determining the effect of reward and non-reward |in discrimination learning. ' 25 The same procedure was used with the mentally iretarded children. The results indicated that there was a i ^stronger (more correct) performance following an initial jnon-reinforced trial. After a comprehensive examination of the data, this was explained on the basis of a tendency to approach novel stimuli rather than to the possible greater associative value of a negative trial. As to the role of negative cues, House and Zeaman stated that after trial two there was no evidence that an unrewarded trial contributed iany thing to the learning process. In fact, the occurrence of an incorrect response was followed by a decrease, rather than an increase in the probability of a correct response. If the negative stimulus does, in fact, play no part in making a choice, one might ask if the £ 3 actually was making a discrimination or was the £ only learning a discriminative stimulus as one does during instrumental conditioning. The Effects of Intelligence House and Zeaman (1958) made comparisons between normal and mentally defective children on the basis of MA to measure the effect of intelligence on ability to learn a discrimination. The two major findirtgs of this experiment were: (1) with increases in MA, normals and mental defect ives made fewer errors in visual discriminations and (2) normal children of the same MA but of higher intelligence ilearned significantly faster than mental defectives, particularly on the more difficult discriminations. I 26 j | This study cannot be considered as conclusive ! evidence that intelligence is related to discrimination learning ability as evidenced by the comparisons made. As House and Zeaman (1958) pointed out, besides intelligence, normals and mental defectives differ from one another in many other ways such as physical health and enriched environment. To explain this possible confounding of the variables in the experiment, House and Zeaman believed that "such differences, however, may not be sufficiently Independent of IQ to allow them to be controlled and yet leave the required variation in IQ. And more importantly, it is not established that such differences are relevant Ivariables for learning ability when IQ is controlled" (p 414). The first statement is an admission that physical and environmental differences do contribute to IQ. These contributions affect intelligence to the extent that the test is measuring responses which are relevant to and predi cated on these different variables. That is to say, an intelligence test samples a portion of the person's mental ability, knowledge and perceptual capabilities which are partly, if not entirely, acquired from learning as a consequence of experience in a given type of environment. 'House and Zeaman, in final support of their position, gave an anecdotal report of frequent observations by "teachers | of the mentally retarded who had long suspected that MA ' 27 scores of intelligence tests were overestimating the jlearning ability of their students" (1958, p 415). This applies to discrimination learning only in a general way and is without empirical verification. However, the information may be helpful (in terms of the difficulty of the information selected for teaching and the expectations of success) to those involved in teaching the mentally retarded. j i Relationships between MA and Intelligence In 1960, House and Zeaman conducted another study along the same theme. The relationship between learning and MA and intelligence was examined. This time only mental [defectives of low MAs were used as Ss. The improvements of the 1960 study over the 1958 study were the increased controls for individual and environmental differences. That is, widely different populations of Ss were not being compared in a specific situation in which differences might be due to the richness of different environments. House and Zeaman found that "MA and IQ are independently related to learning”(1960, p 57). They interpreted this in two ways. The first interpretation of the data was that intelligence, at lower levels, is related to visual discrimination learning. This interpretation |was analogue but more specific than the conclusion in the j1958 study. | ' 28 Beginning to the Attention Hypotheses The second interpretation by Zeaman and House represented a step forward from their earlier postulation (1959) which was that learning curves indicated two proc*: esses involved and attention might be the first process which precedes discrimination learning. The second inter pretation made in 1960 was that "intelligence is related not ito learning in this task, but rather to attention" (House iand Zeaman, 1960, p 57). If perceptual acuity can be thought of as being involved in attention, the criticism made of the 1958 study takes on greater validity. Children, in a stimulating environment, learn by experiencing the differences between stimuli and thus, develop relatively more acute perceptions than a person in an institutionalized setting where the stimulus situation is more constant. Perceptual acuity is necessary and increases the likelihood that a child will discern the relevant stimulus dimensions sooner, if at all. The rationale for postulating the role of attention rather than intelligence in discrimination learning begins with considering learning as representing a modification or improvement of a behavior. In the 1960 study, learning was measured in terms of progressive reduction of errors to criterion. This measure of learning was used by Zeaman and ;House because results by other researchers suggested exami nation of errors to criterion may serve to better reveal the* 29 jprocesses involved. Many theorists (none was named) performed such an analysis of visual discrimination learning: i land discovered two component processes were involved; "one j :of which required that a S_ direct and maintain attention to the relevant cues of the situation before and during learning” (House and Zeaman, 1960, p 57). Intelligence Versus Attention Investigation Zeaman and House (1959) and House and Zeaman (1960) ! made such an analysis discovering that ”the main source of individual differences in total errors comes from the duration of chance performance before any learning is observed.” This is an interpretation of discrimination learning curves (S shaped) which characteristically show a flat line, representing chance success, and eventually a rapid rise until the criterion is reached. The data of this experiment do not differentially support the intelligence or attention interpretation but merely indicated that there is reason to believe that the individual difference construct is involved in discrimination learning and that "it is correlated with MA and IQ" (House and Zeaman, 1960, p 57). The problem was how to separate the effects of attention from the effects of intelligence to demonstrate clearly that . i the construct of attention is the primary variable in the rate of learning a discrimination. Zeaman and House (1965) were able to answer this |problem by utilizing the backward learning curve. Evidence 30 for the attention interpretation came from the observation ! (of data from the 1960 study) that the apparent rate of learning by Ss having Ms of 4-6 and 2-4 years were the same iafter the lower MA group learned to attend to the relevant |stimulus dimension. More strikingly, there were Ss in both MA groups who failed to learn the discrimination indicating that intelligence, in relation to attention, played a secondary role. Stimulus Factors Affecting Attention Information about the effect of stimulus factors presumably came from the results of the 1962 study by House and Zeaman which was reorganized for the 1963 presentation of their theory. In this study, S>s of different MA levels made color-object-form discriminations. Only the results of the Ss of Ms of 4-6 years were reported. The Ss learned to discriminate between multidimensional stimulus objects, colors and forms. Examples of multidimensionally different stimuli (junk) were an aluminum pot cover versus a green plastic soap dish (Zeaman and House, 1963). In the order of easiest to more difficult to learn to discriminate were multidimensional stimulus objects (with Ss experienced at discrimination learning) and objects differing in color alone (with Ss experienced at discrimina tion learning). Zeaman and House pointed out that "... the difficulty of the object-form problem may be underestimated" jbecause the Ss used had previous experience in learning ' 31 to discriminate (1963, p 164). In a study to be reported, Shepp and Zeaman (1966) [discovered that discriminations were more quickly made if ; there were wide differences in cue properties. Discrimi- i jnation could also be taught on the basis of position (right or left) of the stimulus objects. Zeaman and House, based on an experiment conducted in 1959, reported that position [discriminations were learned in an average of three trials. i The learning of a discrimination can be facilitated by introducing a novel stimulus as a substitute for a previously used stimulus. Zeaman, Hduse and Orlando (1958) discovered that the introduction of a novel stimulus [produced a "sudden and unexpected learning of a discrimi nation" which facilitated the learning of very difficult discriminations (Zeaman and House, 1963, p 199). Another procedure for catching the S/s attention is to reverse "the roles of the positive and hegative cues." This means making the reward now contingent upon responding :to the stimulus object previously not associated with a reward. This kept the probability of observing the once.: rewarded stimulus high (if the S was close to reaching the criterion). However, the probability of responding to it became increasingly lower, while the probability of observing and responding to the previously unrewarded stimulus object continued to increase (Zeaman and House, 11963, P 199). 32 j RECAPITULATION OP THE RESEARCH ! j Zeaman and House (1963) reanalyzed data from experiments conducted in 1960 and 1962, using backward i ^learning curves. Using this method of organizing the data provided new comparisons between fast and slow learners. i The comparisons revealed that fast and slow learners differ ionly in the number of trials required before learning began; i once learning began, the rate of improvement in reaching the criterion was the same. Zeaman and House postulated that the determinant of when learning began to occur was atten tion. Attention was thus theorized as that which differen tiates the fast from the slow learners. Zeaman and Hotise used attention to explain the differences in learning performance which was the basis for inferring the effect of attention. This is a circularity, a typical methodological shortcoming of studies involving |perceptual processes (Maltzman and Raskin, 1965). An independent objective measure of attention, a way of measuring attention other than consequent learning perfor mance is necessary to gain a clear assessment of the effect ! of attention. Another shortcoming in the Zeaman and House theory, perhaps forgivable, is the lack of information about :how fast and slow learners differ in their attention. ! The OR has been used as a tool for analysis of 33 perceptual processes (awareness) by Maltzman and Raskin (1965). The OR, a non-specific reaction of cortical origin to stimuli novel or significant resulting in a readiness or I preparatory response, corresponds, in part, to the phenome non of attention. The initial condition for the occurrence of atten tion is a stimulus change. A stimulus change is also a ;necessary condition for the occurrence of the OR, "as a mechanism for paying attention to novel stimuli" (Lynn, 1966, p 1). The OR facilitates the reception of stimuli by sensitizing the sensory receptors and directing the organism to the source of stimuli. Consciousness refers to attention jand awareness? the OR, in activity, corresponds to conscious activity, to the learning of stimuli and making discrimi nations between stimuli. Maltzman and Raskin (1965) studied individual differences in the OR and its effect on learning and ^awareness. As a tool for analysis of perceptual processes, ;high magnitudes of the OR corresponded with S_ awareness which facilitated learning of semantic conditioning and generalization. Of the available autonomic responses for measuring the OR, the GSR was selected for this study. The GSR is also conceived of as a measure of emotionally based drive. In the Raskin (1963) study of individual differences in the i OR and its effect on semantic conditioning, a short form of 34 the MAS was given to Ss to check the GSR as a measure of orienting rather than emotionality. MAS scores showed no reliable correlation with GSRs, indicating in this instance, |that the measure was of orienting. Such a procedure of i i testing for emotionality with mentally retarded Ss is not possible. Raskin used a 110 db white noise to elicit an OR. This study used a 6 Watt light and colored wooden figures ! to elicit an OR. These stimuli should not induce stress or ! emotionality, yet are still capable of producing an OR (Badia and Defran, 1970). Badia and Defran (1970) pointed out that ORs, CRs, and UCRs can be mistaken for one another; stimuli of weak intensity (50 db sound, 6 Watt light) occurring together do not lead to the production of CRs, which is desired in this study. Grings and Carlin (1966) demonstrated that magnitude and frequency of GSRs can be instrumentally conditioned. To demonstrate the contingencies, Grings and Carlin used yoked control groups as a control for Ss responsitivity and the number of reinforcements of the GSR. Sehnidman (1970) has ; shown that eleetrodermal orienting responses can be instrumentally conditioned. In the study reported here, Ss were reinforced with M & M candy for correct discriminations. The involvement of reinforcement, as shown in the above studies, can affect the magnitude of the OR and its metric the GSR. A methodologi- leal question arises: are the Ss learning faster because 35 they are orienting higher or are they orienting higher because they are learning faster? If the first part of the question is true, the OR is not being affected by the reinforcer, nor is its metric the GSR. If the second part |of the question is true, the OR is affected as is the GSR. |The use of yoked controls, as used by Grings and Carlin I i(1966) will provide answers to the question posed. CHAPTER II METHOD Subjects All of the Ss used in this experiment were resident patients at Pacific State Hospital, a California Institution for the mentally retarded. The Ss had MAs ranging from 4-6 years, CAs of 11.6 to 12.2 years and IQs of 35 to 45* The ,Ss had no severe visual, auditory nor neurological deficiencies, did not show frequent behayioral problems as autism, withdrawal, aggression, hyperacticity nor psychosis, and were not on drug therapies. Prom the available total population of 20 Sa having the above stated characteristics, 16 Ss were used in the experiment. During the course of the experiment, one S became progressively uncooperative and unmanagable and had to be released from the experiment along with his yoked control member. Two of the Sa failed to learn the discrimination. Their fixated basis for responding was a right or left position. These Ss were dismissed two sessions (40 trials) after the other experimental Ss had reached criterion. In a pilot study, in which procedures were tested, three Ss who had received the same procedure as Ss in the experiment proper were then reintroduced into 36 .................................................. 37 the experiment for the data analysis. A new median was computed to determine fast and slow learner classification. The three pilot Ss qualified as slow learners. i { Apparatus I i A two-choice discrimination apparatus was used, {functionally resembling the Wisconsin General Test apparatus, but with some modifications. The modifications i were a replacement of a see-through mirror by a wooden louver partition which allowed the 35 (experimentor) to i observe the S, without being seen. A warning light (6 Volt) was positioned in the middle of the apparatus, 44 inches above the food tray. A 3/4 inch hole was drilled in the center of the apparatus with the opening on the EJ's side, one inch below the light. The hole was drilled on about a 45° downward slant such that the S_ could not see what was taking place on the other side of the apparatus. Four push button switches were mounted to the foodtray. Two switches were mounted on the E's side of the tray, one controlling the onset of the warning light and a light on the E’s side of the apparatus (to signal the onset-offset of the warning light). The other switch controlled the offset of the lights when the E desiredj otherwise a Hunter i ; Timer controlled the offset of the lights. Two of the jswitches were on the S.'s side of the tray, one positioned one inch in front of each of the two stimulus objects. The experimental room was 81 x 10' in dimension, 38 adequately lighted, relatively silent (except for the I constant sound of the air conditioning unit) and had no ! iwindows. The GSR recording unit and the Hunter Timer were j |hidden behind a partition from the S3. Discriminanda | The stimulus objects used were a red square :(2i" x 2i") and a green square of the same dimensions, cut from i inch masonite and mounted on 4” x 4" gray squares. i |A type R.S. Dynograph recorder in conjunction with a type I9892A Skin Resistance Coupler was used as an event indicator which marked the occurrence of the warning light onset and ithe button pushes by the S and the end of three-seconds' |time intervals timed by a Hunter Timer. Procedure for Assignment to Groups Prior to the experiment, Ss responsitivity, in terms of averaged magnitude of GSR to a 6 Yolt light, was ! measured for matching the yoked S pairs. A S was brought | individually to the experimental room and allowed sufficient time to habituate to the environment such that a baseline j level of resistance could b® ascertained. Then measuring i i of the S's responsitivity commenced. The £ 3 was told to sit very still and watch for the light to come on. The light i |was presented six times randomly at 30, 40 and 50-second intervals for three seconds duration. GSRs beginning one second to five seconds after the onset of the light were jtaken as the GSR to the light. The maximum change in " ■ ........ ' ' ......“... 39 resistance reached after about ten seconds of the light onset was the magnitude of the response. The magnitudes of the GSRs were averaged and matched pairs of Ss were formed. Assignment of a member in each yoked pair to the experi mental group was determined by a coin flip. I The Ss assigned to the experimental group were run | ifirst in an experimental day and were immediately followed |by their control counterparts. The control members of the jyoked pair received reinforcement on the same numbered trial as their experimental counterparts, regardless whether the control S's discrimination was correct or not. At the completion of the experiment, the experi mental Ss were divided into fast and slow learner groups depending on whether the number of experimental sessions required to reach criterion were below (fast learners) or above (slow learners) the median number of experimental I isessions required for all experimental Ss. The GSRs and the ;behavioral indices of orienting of the Ss were then analyzed. Experimental Procedure During the experiment, the Ss were brought into the | experimental room one at a time. Upon arrival into the room, the E seated the i S and commenced to attach the lelectrodes to the fingertips of the first and third finger |on the left hand. The S_ was then encouraged to relax and i |sit quietly. The E then explained to the S that he was going to play a guessing game and if he guessed right, he ] Iwould receive a candy# The food tray, containing the two stimulus figures, was then pushed out. While pointing to lone of the stimulus figures the E said: "See this one? If I |you pick this one, point at it and then push this button, I like this." The E then pushed the button and said: "Now you try it." The S commenced to follow the instructions and was assisted by the E when help was indicated. The i Iprocedure was repeated for the other stimulus figure. The !E then said: "Do you see how we play this game? Now, remember, if you pick the correct one, you will get candy. The candy will come out of here." The E pointed to the hole in the apparatus. The E then attached the electrodes and told the S_ to sit still and went behind the apparatus to prepare for the trial. The stimulus figures were placed in their right-left positions of the correct stimulus figure. To avoid the learning of a position, the correct stimulus figure was not placed in the same right or left position during three successive trials. The red square was the correct stimulus figure for half the experimental Ss and the green square was i correct for the other half. After sufficient time (average of ten minutes) was allowed for the S. to habituate to the environment such that a baseline level of resistance could be ascertained, E [activated the three-second warning light. E, six seconds 41 after the warning light onset, pushed out the tray- containing the two stimulus figures for S. to make his choice. i Sixty seconds were allowed for S to make his choice; non choosing was recorded as an incorrect choice. After £ 3 made |a choice, the tray was withdrawn. Reinforcement was ! delivered three seconds after the button push if the choice !was correct. This concluded one trial of the experimental isession which was composed of 20 trials. The next trial i ; began on an average of 30 seconds later. For the next trial, the stimulus figures were repositioned and the baseline resistance was ascertained. The criterion used for learning the two-choice |discrimination was 90$ or more correct responding to the presentations during an experimental session. Upon reaching the criterion, the £ and his control counterpart were excused from the experiment. Measurement During each trial, GSRs occurring to the warning light and to the discriminanda at the moment of choice were recorded. Also, behavioral evidence of the OR (eye fixa tions) to the two stimulus events was recorded. The GSR occurring from one to five seconds after the warning light onset was recorded as the response to the iwarning light. The maximum change in conductance occurring within three seconds of the onset of the GSR was the jmagnitude of the response. The GSR given to the warning 42 light was interpreted as the OR to the light and represented ithe S's level of attention. The GSR occurring from one to Ifour seconds after one of the two figures of the discrimi- j !nanda was selected, via pointing to the figure and then jpushing the appropriate button by the S, was interpreted as Ithe level of attention during the process of cue selection iin which the meaning of the stimuli were learned. The imaximum change in conductance occurring three seconds after the onset of the GSR was the magnitude of the response. The behavioral measurements of the OR representing the occurrence of attending to stimuli during each trial were as follows: Measurements of orienting to the task 1. Eye orientation. The occurrence of eye orientation to the warning light during onset. Eye fixation had to be greater than one second. 2. Bye orientation. The occurrence of eye orientation to both of the stimulus figures during selection of one of the figures. Eye fixation time to each of the figures had to be greater than .5 seconds. An interrater reliability coefficient obtained during the first session for six of the Ss was: r = .96. In the experiment, the ]>s were rated during each trial and a score of 1 or 0 was given for each of the stimuli depending upon whether the S, oriented or not. l operational Definition Attention was evidenced by the OR which is autono- Imically recorded in terms of the GSR and behaviorally in I terms of eye orientations to the warning lifeht and stimulus 43 figures of the discriminanda. ! i HYPOTHESES The sixteen research hypotheses are presented each with its experimental hypothesis. Research hypotheses I through are concerned with the GSRs of the experimental iSs and yoked control Ss. Research hypotheses through Hg lare concerned with the frequencies of eye fixations to ^stimuli by the experimental Ss and yoked controls. Research hypotheses Hg through are concerned with quantitative and qualitative attentional differences between fast ilearners and slow learners. Hypotheses H. ) through This group of research hypotheses are concerned with GSR differences between fast learners and their yoked control partners. Hypothesis H^ i The first hypothesis H^ is concerned with the causal relationship between levels of orienting and the speed of learning. Research hypothesis H^ is as follows: fast learners learn faster because their GSR magnitudes are. higher (rather than GSR magnitudes being higher because they ;learn faster). The direction of this' causality should be exhibited in accordance with the Maltzman and Raskin (1963) I Iresearch on the relationship of the OR (measured as GSR) to I awareness and learning. ___ ____________________________ ” 44 The proposition of this hypothesis is necessary by |the presence of reinforcements which might affect the jmagnitudes of the GSR. The timing of the reinforcement (three-second delay) was planned to reinforce a correct jchoice but not to affect the GSR activity at the moment of choice. A yoked control design was used to control for the ieffects of the reinforcer. The logic of the yoked control jwas that by noncontingently reinforcing choices of yoked jcontrol members, learning would be prevented. The differences in the GSRs in favor of the yoked control members oyer the experimentals would indicate reinforcements effects on the GSR. i The two experimental hypotheses involved in testing research hypothesis are H^s fast learners (X^) and yoked control members (X2) will not differ in tonic arousal time during the first twenty trials. This experimental hypothesis was formulated to test the adequacy of matching Ss. Tonic arousal is an index of responsitivity. Statistical Hypotheses: H1a1: = ^2 H1a2: X1 > X2 < .05 j The second experimental hypothesis H^: fast ilearners (X^) will give higher magnitudes of GSR than will yoked controls (X2) to the warning light and during the jmaking of a discrimination at the 50%, 65%, 75% and 90%+ llevels of reinforcement (reinforcement contingent upon 45 correct responding by the fast learners). Statistical Hypotheses: H1b1 H1b2 H1b3 I1 = 12 I1 > X2 < .05 X1 < X2 < .05 jHypothesis H2 The second research hypothesis H2 is concerned with jthe differential effect of the stimuli in the two stimulus |situations, warning light and discriminanda, to elicit GSRs from the fast learners and yoked controls. Based on the report by Grastyan (1961), the warning light is lesser of a i learning task than learning to discriminate the stimuli of ithe discriminanda and should elicit lower GSRs. The second research hypothesis H2&: the stimuli of the discriminanda (ILj) shall elicit higher GSR magnitudes from fast learners and yoked controls than will the warning light stimulus situation (X2). Statistical Hypotheses: H2a1: X1 = x2 H2a2: I > X2 < .05 H2a3:V x 2 < .05 Hypothesis H^ The third research hypothesis H^ is concerned with the habituation process. Habituation is the progressive extinction of the OR as stimuli become familiar or learned. Since fast learners are learning but yoked control members 46 are not, the point of comparisons between the two groups is the level of reinforcement. The level of learning and the level of reinforcement for the fast learners are the same (i.e., at the 75% level of reinforcement the fast jlearner is correctly choosing 75% of the presentations of ithe discriminanda). Yoked control members received reinforcement on the same corresponding trials as the fast ( learners. Thus, their level of reinforcement equals that of ;the fast learners. As higher levels of learning are reached I by the fast learners, the meaning of the stimuli are better I learned, more familiar and thus, less capable of eliciting GSRs. logically, habituation should occur in accordance iwith the report by Grastyan (1961) that the OR was persistent when the meaning of stimuli were unknown. Learning the meaning of stimuli would then accompany extinction of the OR and habituation. In opposition of this logic is that habituation will not occur or be greatly impeded because the stimuli involved have acquired signal characteristics. The involvement of reinforcement, while not instrumentally affecting the OR, may be causing the stimuli to develop particular signifi cance to the S* Lynn (1966), in a summation of habituation j istudies, reported that when stimuli are or become signifi cant (i.e., baby cries to mother, hearing one's name, watch outl) habituation to these stimuli is "very greatly prolonged or never occurs" (p 30). The presence of rein- 47 forcement relates to prolonging the occurrence of the OR without directly modifying its magnitude. J I Research hypothesis is: as levels of reinforce- jment increase, the stimuli involved will become and remain significant to the S, leading to a prolongation of the I occurrence of the OR, impeding habituation. If reinforce ment does not modify the magnitude of the OR but causes |stimuli to become significant by its presence, GSRs given by fast learners and yoked controls should not increase as levels of reinforcement increase. Experimental hypothesis H^a: magnitudes of GSRs by Ifast learners and yoked controls to the warning light and j idiscriminanda will not increase as levels of reinforcement increase (50%, 65%, 75% and 90%+). Statistical Hypotheses: H3ar X1=X2 = X3 = X4 H5a2: X1 < X2 < X3 < X4 < .05 H3a3: r, > X2 > X3 > X4 < .05 Hypothesis H4 Research hypothesis H4 is concerned with interaction effects that may exist between combinations of levels of stimulus situations (warning light and discriminanda) and/or ilevels of reinforcement (50%, 65%, 75% and 90%+). ! Experimental hypothesis H4&: an interaction between |combination; of main effects of levels of Ss (a^) and levels !of stimulus situations (b.) will produce an effect greater i _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ J _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I 48 than the sum of these two main effects. Statistical Hypotheses: H4a1: ab13 = *13 - <ai + bj + *13k> H4a2: abi3 > *13 - (ai + b5 + ^ ’ <05 i I | Experimental hypothesis H^: an interaction between combinations of main effects of levels of Ss (a.) I ~ 1 and levels of reinforcement (ck) shall produce an effect greater than the sum of these two main effects. | Statistical Hypotheses: H4b1: acik = fik “ <ai + ck + ^ijk) H4b2: acik > Yik ” (ai + ck + < ‘°5 Experimental hypothesis H^c: an interaction between combinations of main effects of levels of stimulus situa tions (b^) and levels of reinforcements (ck) shall produce an effect greater than the sum of these two main effects. Statistical Hypotheses: H4c1! bcjk = *jk ~ + ck + ^ijk) H4c2: bcjk > ^jk " ^bj * ck + ^ijk^ < *°5 |An interaction between two main effects in combination (acik» a^i^» or acjk^ and a "kkird effect (b^, ck or a^) |shall produce an effect greater than the sums of the three i main effects a^, b^ and ck). ’ Hypotheses H,. through Ho j This group of research hypotheses are concerned with l ithe percentage of eye orientation differences between the 49 fast learners and their yoked control partners. Hypothesis Research hypothesis H^, having concerns very similar to hypothesis H«j, is as follows: fast learners are learning ifaster because they give a higher percentage of eye orientations rather than giving a higher percentage of eye orientations because they are learning faster. Experimental hypothesis H^a: fast learners (X1) will give higher per<- icentages of eye orientations than will yoked controls (1L,) to the warning light and to both stimulus figures of the discriminanda during the 50$, 65$, 75$ and 90$+ levels of reinf ore ement• Statistical Hypotheses: H5a1 H5a2 H5a3 X l = x2 > X2 < .05 X., < X2 < .05 Hypothesis Hg Research hypothesis Hg, essentially the same as research hypothesis H2, is formulated in terms of the behavioral data: the stimuli in the two stimulus situations will have a differential effect in eliciting eye orienta tions from both the fast learners and yoked controls. The experimental hypothesis related to this research hypothesis is as follows. Hg&: the stimuli of the discriminanda (X^) [will elicit higher percentages of eye orientations (to both of the stimulus figures) from fast learners and yoked 50 controls than will the warning light stimulus situation joy. Statistical Hypotheses: ! Hypothesis Hy j Research hypothesis Hy is generated from the same -concerns as research hypothesis H^. The only differences ! between the H^ and Hy research hypotheses is how the ! behavioral data are being tested. Research hypothesis H7: j ' jas levels of reinforcement increase, the stimuli will become significant and eye orientations to the warning light and tc ■both of the stimulus figures of the discriminanda will increase in percentage. Experimental hypothesis Hy&: percentage of eye |orientations by the fast learners and yoked controls to the jwarning light and to both stimulus figures will not increase jas levels of reinforcement increase (50% * X^t 65% = X2, 75% = X3 and 90%+ = 14). Statistical Hypotheses: H7a1! *1 = = x2= x3 = X4 H7a2! *1 < ' 12 * X3 4 x4 > .05 H7a3: X-, ? > X2 > > X4 < .05 Hypothesis Hq Research hypothesis Hg, like research hypothesis H^, 51 but the data to be tested is behavioral, is concerned with [possible interaction effects that exist between combinations j |of levels of Ss (fast learners and yoked controls) and jlevels of stimulus situations (warning light and discrimi- i Inanda) and/or levels of reinforcement (50%, 65%, 75% and j j 90%+)• The experimental hypothesis related to research i jhypothesis Hg is Hg&: an interaction between combinations i |of main effects of levels of Ss (a^) and levels of stimulus ^situations (b.) will produce an effect greater than the sums J of these two main effects. Statistical Hypotheses: W abij = *ij - <ai + + j W abij > - <ai + bj + < -05 Experimental hypothesis Hgb: an interaction between combinations of main effects of levels of Ss (a^) and levels of reinforcement (c^.) will produce an effect greater than the stun of these two main effects. Statistical Hypotheses: H8b1! acik = Xik “ ^ai + ck + ^ijk^ H8b2: acik > Xik ” (ai + ck + Xijk^ < *05 Experimental hypothesis Hgc: an interaction betweer icombinations of main effects of levels of stimulus situa tions (b..) and levels of reinforcement (c^) will produce an |effect greater than the sum of these two main effects. Statistical Hypotheses: j H8c1: bcjk = xjk ” + ck + Xijk^ 52 H8c2: bcjk > ^jk “ (bj + ck + ^ijk^ < *05 Experimental hypothesis Hg^: an interaction between two main effects in combination (acik, abij» or bc^) and a third main effect (b^, c^ or a^) will produce an effect greater than the sume of the three main effects a^, b.. and c^. Hypothesis Hg through H12 These hypotheses are concerned with quantitative and ! qualitative attentional differences between fast and slow ilearners measured in terms of GSR. jHypothesis Hg I Research hypothesis Hg is concerned with quantita^ tive differences of the fast and slow learners in orienting to stimuli and is as follows: fast learners will give larger GSRs than slow learners during the learning of a discrimination. The related experimental hypothesis Hga is: jfast learners ) will give larger magnitudes of GSRs than I slow learners (X2) to the warning light and during the making of a discrimination at the 50$, 65$, 75$ and 90$+ ilevels of learning. Statistical Hypotheses: H9a1 H9a2 H9a3 x1 = x2 > X2 < .05 X1 < T2 < .05 53 Hypothesis H^Q Research hypothesis H^0 is concerned with the qualitative nature of attention of both fast and slow learners during the learning of a discrimination* The test jis to determine the differential effect of the stimuli in j ithe two stimulus situations to elicit GSRs from the Ss. The |stimuli in the two situations (warning light and discrimi- inanda) are of different instrumental worth for acquiring | ! reinforcement. The warning light was not directly related ;to solving the problem to receive reinforcement. Some consideration of the stimulus figures of the : discriminanda was necessary if a correct choice was to be jmade and reinforcement acquired. The stimuli of the Idiscriminanda should be of greater significance and thus, oriented to in greater magnitude. Differential responding on the basis of significance of stimuli indicates the qualitative nature of attentional differences to stimuli. Research hypothesis H^0 is then: the stimuli in the two stimulus situations will have a differential effect in eliciting GSRs from both the fast and slow learners. Experimental hypothesis H-joa1 stimuli of the discrimi- nanda (X^) will elicit higher GSR magnitudes from fast and jslow learners than will the warning light stimulus situation (X2). Statistical Hypotheses: jHypothesis The eleventh research hypothesis is concerned with Ithe habituation process as it relates to quantitative |characteristics of the attentional behavior of the fast and I Islow learners. For both the fast and slow learners, the jmagnitudes of the GSRs given to the stimuli should be the highest at the 50% level of learning and gradually become less in magnitude as the meanings of stimuli are learned. This would be a demonstration of the habituation phenomenon jwhich occurs during repeated exposure to stimuli. The i jincreases in the rate of reinforcement, that accompanies reaching higher levels of learning, may make the stimuli significant to which the 8 continues to orient to and thus, fails to habituate. | Research hypothesis is: as levels of learning increase, the stimuli involved become significant to the S, Heading to prolongation of the occurrence of the OR, impeding habituation. Experimental hypothesis H11&: the magnitudes of the jGSRs given to stimuli by the S will not differ at the 50% (X^), 65% (^2 75% (X3), 90%+ (X4) levels of learning. Statistical Hypotheses: 55 H11a3s X1 < X2 ^ V V - 05 Hypothesis H-jg Research hypothesis iss interaction effects exist between combinations of levels of Ss (fast learners, jslow learners) and levels of stimulus situations (warning ilight, discriminanda) and/or levels of learning (50%, 65$, 15% and 90%+). Experimental hypothesis H12a: an interaction between combinations of main effects - levels of Ss (a.^) and levels of stimulus situations (bi) will produce an effect J greater than the sums of the two main effects. Statistical Hypotheses: H12a1! abij = " <ai + bj + \jk> H12a2: abij > - ^ai + bj + ^ijk^ -05 Experimental hypothesis H12t): an interaction between combinations of main effects - levels of Ss (a^) and levels of learning (ck) will produce an effect greater : than the sum of the two main effects. Statistical Hypotheses: H12b1: acik = xik “ ^ai + ck + Xijk^ H12b2: acik > xik " ^ai + ck + Xijk^4 *°5 Experimental hypothesis H12cl an interaction jbetween combinations of main effects - levels of stimulus jsituations (b^) and levels of learning (c-^) will produce Ian effect greater than the sum of the two main effects. 56 Statistical Hypotheses: H12c 1s bcjk = ^jk “ (bj + ck + ^ijk^ H12c2! bc;jk > ^jk " ^bj + ck + ^ijk^ *°5 Experimental hypothesis a*i interaction between two main effects in combination (acik, ab^.. or bCjk) and a third main effect (b^, ck or a.^ will produce an effect greater than the sums of the three main effects b^ and ck» Hypotheses g through H-)g Hypotheses concerned with quantitative and qualita tive attentional differences between fast and slow learners measured in terms of percentage of eye orientations will now be presented. Hypothesis H^ Research hypothesis H^, concerned with qualitative differences of fast and slow learners, is as follows: fast learners will give higher percentages of eye orientations during the learning of a discrimination than will slow learners. Experimental hypothesis H ^ &: fast learners (X^) will give higher percentages of eye orientations than will slow learners (X2) "to the warning light and to both of the stimulus figures of the discriminanda during the 50%, 65%, 75% and 90%+ levels of learning. 57 Statistical Hypotheses: H13a1 H13a2 H13a3 x1 = x2 > X2 < .05 X1 < I2 < .05 Hypothesis H^ Research hypothesis H14 is based on the same rationale that underlies research hypothesis H^q. The stimulus situations differ in their instrumental value for |learning the discrimination and acquiring reinforcement. ^Research hypothesis H^: the stimuli in the two stimulus situations will have a differential effect in eliciting eye orientations from both the fast and slow learners. | Experimental hypothesis H ^ a: the stimuli of the discriminanda (X^) will elicit higher percentages of eye orientations (to both stimulus figures) from fast and slow learners than will the warning light stimulus situation :< x2). Statistical Hypotheses: H14a1! *1 = x2 H14a2! x1 > x2 H14a3: V t2 Hypothesis H^ I Research hypothesis H^ is concerned with the habituation process as it relates to the qualitative | characteristics of the attentional behavior of the fast and I jslow learners. For both the fast and slow learner?, the___ 58 percentage of eye orientations given to the stimuli should be the highest at the 50% level of learning and gradually becoming less as the meaning of stimuli becomes learned. The warning light stimulus situation, having little instru mental value, should elicit eye orientations less frequent ly as higher levels of learning are reached. Also, as jhigher levels of learning are reached, eye orientations to |both stimulus figures should occur only if the S 3 should jhappen to orient first to the incorrect stimulus figure. (The presence of the reinforcer may make the stimuli take on significance, thus, holding the Ss' attention to the (stimuli involved. The developing significance of the stimuli, expecially the stimuli of discriminanda, may thwart the habituation phenomenon. Eye orientations during the iwarning light stimulus situation and to the two stimulus figures of the discriminanda may continue even though it is not necessary for making a correct choice. | Research hypothesis H ^ f stated as a null hypothesis i iis: the frequencies of eye orientations given to stimuli |by Ss, will not differ at the 50% (X-j), 65% (S2)» 75% (X^) land 90%+ (X^).levels of learning. Statistical Hypotheses: H15a1: X1 = y2 = x3 *4 H15a2: x1 > X2 > X3 > X4 < .05 H15a3: x1 < X2 < x3 < I4 < .05 ' .59 Hypothesis Research hypothesis H^: interaction effects exist between combinations of levels of Ss (fast learners, slow learners) and levels of stimulus situations (warning light, discriminanda) and/or levels of learning (50%, 65%, 75% and 90%+). Experimental hypothesis H^g&: an interaction between combinations of main effects - levels of Ss (a^) and levels of stimulus situations (b^) will produce an effect greater than the sums of the two main effects. Statistical Hypotheses: H16a1! abij = xij " (ai + b3 + | H16a2: *>14 > \i - <ai + »4 + *t4k> '< -°5 Experimental hypothesis an interaction between combinations of main effects - levels of Ss (a^) and levels of learning (c^) will produce an effect greater than the sum of the two main effects. Statistical Hypotheses: ! j H16b1: acik = Xik “ ^ai + ck + Xijk^ H16b2: acik Xik “ (ai + ck + \jk^ < ,05 Experimental hypothesis H^gc: an interaction between combinationof main effects - levels of stimulus i Isituations (b.) and levels of learning (ct.) will produce an j |effect greater than the sum of the two main effects. I Statistical Hypotheses: _______________ H 16c1;_bc;jk = Xjk_~^o__+ck + Xijk^______ 60 | H16c2! boJk> V " <bj + °k + < *05 | Experimental hypothesis H-jg^x an interaction |between two main effects in combination (acik, ab^ or |bc^k) and a third main effect (b^, ck or ai) will produce ar |effect greater than the sums of the three main effects a., |bj and ck. | BATA ANALYSIS I To accomodate the variance among Ss in each group (fast and slow learners) in the number of sessions required |to reach criterion, the data were organized into four levels !of learning; 50$ correct responding (48$ to 52$), 65$ (62$ to 68$), 75$ (72$ to 78$) and 90$+ (90$ to 100$). The reason for selecting the first and fourth levels of learning is obvious. The second and third levels selected because the percentages of correct responding during pre-criterion i sessions clustered around these values. In the organization of the data, the percentages of correct responding by the Ss during the sessions were inspected to find natural breaks in the learning curves. I In cases where the level of correct responding by a S had |not changed from one session to the next or when the average i of the level of correct responding for the two sessions was ;a value that fell within the ranges set for the four levels of learning, both sessions were used to represent the """" 61 the particular levels of learning in question. This | afforded the opportunity to use all the sessions and trials |of each S. In the case of three of the fast learners in which i the criterion was reached within one, two or three sessions, j ; groups of trials (four or more) were used; in some |instances, the groups of trials overlapped one another by jone or more trials. Where it was possible to use an entire jsession of a S to represent a particular level of learning, the opportunity was taken. After the trials to be used for each of the levels of learning were identified, the GSR values and behavioral lvalues were then averaged for the trials. The behavioral values, eye orientations, were recorded as frequencies to each of the events. Percentages of eye orientations, the ratio of the number of eye orientations to the total number ;of times stimuli were made available during the set of ! trials representing a given lfivel of learning and were then calculated. These percentage values for each level of i learning, to each of the stimulus situations (warning light iand discriminanda), were then averaged by groups. Values for the yoked control members of the fast learners were averaged over the exact numbered trials and sessions. The ! i |GSR values and behavioral values did change from session to |session, but the usual orderly rate of habituation did not |take place, so the instances of using two sessions and 62 groups of trials to represent levels of learning provided good estimates of central tendency. STATISTICAL PROCEDURES | The statistical test used to partially evaluate hypothesis was a t-Test for differences between means of ithe experimental hypothesis H^&. The means for each fast i ! learner and his yoked control member were the correlated observations. Experimental hypotheses of research hypotheses H1, iHgj and were tested simultaneously in a 2 x 2 x 4 factorial design. The A factor was Ss which had two levels: 1 fast learners and yoked control members. The B factor was stimulus situations which had two levels: warning light and discriminanda on which repeated measures were taken. The C factor was levels of reinforcement of 50%, 65%, 75% and 90%+ on which repeated measures were taken. All factors iwere fixed since assignment of Ss to the fast learner group was not random and levels of B represented q effective levels of Q and C represented r effective levels of R. The experimental hypotheses of research hypotheses H H y and Hg were also simultaneously tested in a 2 x 2 x 4 factorial design. The factors A, B and C were the isame as in the above three-way analysis of variance. The lonly difference* was that the behavioral data (eye orienta- jtions) rather than GSR data were used. The experimental hypotheses Hg, H10, h- j- j and H^2 were evaluated as before. The same factorial design was used with changes being factor A changed from level of Ss to speed of learning, having the two levels: fast and slow learning, and factor C changed from levels of reinfor cement to levels of learning, having the four levels of 50% 65%, 75% and 90%+. | CHAPTER III ! RESULTS j The presentation of results of experimental hypo theses shall be preceeded by citations of relevant tables |and figures. Families of research hypotheses and their lexperimental hypotheses have been presented separately. The results of the experimental hypotheses ipresented below belong to research hypotheses through which were concerned with GSR differences between fast learners and their yoked controls. Table 1 is a summary of analysis of variance of the GSRs for fast learners and yoked controls. The experimental hypotheses of research hypotheses H1 through were tested simultaneously in a 2 x 2 x 4 analysis of variance. Table ' . 2 gives the mean GSR magnitudes of the fast learners and yoked controls to the two stimulus situations during the different levels of reinforcement. Figure 1 is a graphic comparison of the averaged GSRs of the fast learners and yoked controls to the warning light and discriminanda during ithe four levels of reinforcement. Hypothesis The t-Iest used to test the hypothesis H^a that fast learners (X^) and yoked control members (X2) will not differ i in tonic arousal during the first experimental session L _______ 64______________________________ r "....................................... 65 i jresulted in a t = .041, d.f. 12 > .05. Thus H^a^s = X2 |was supported. The reason for testing this hypothesis iwas to check how well the fast learners and yoked control imembers were matched on responsitivity. Differences in the jGSRs of the fast learners and yoked control members could have been due to differences in the tonic arousal levels of the Ss. This would be particularly true, were the differences extremely large between the groups (Schell, personal communication). i Experimental hypothesis tested the causality between speed of learning and magnitude of GSRs. Table T indicates no significant difference occurred for factor A (fast learner, yoked control). Thus, statistical hypothesis X^ = X2 was supported. Figure 1, on appearances, indicates the yoked controls to have somewhat larger averaged GSRs during the making of a discrimination than fast learners. Hypothesis Hp Experimental hypothesis H2a tested the differential effects of the stimulus situations to elicit GSRs from the fast learners and yoked controls. Table 1 indicates no significant differences occurred for factor B (warning light, discriminanda). The statistical hypothesis H2a^: X^ = X2 was supported. Figure 1, on appearances, indicates the yoked controls to have differentially responded to the two I jstimulus situations. In data form, Table supplements 66 this observation. The non-differential responding by the fast learners to the stimulus situations, as indicated in Figure 1, is probably responsible for the non-significant result. [Hypothesis i Experimental hypothesis tested the differences lin habituation by the fast learners and yoked controls to the two stimulus situations during the different levels of reinforcement. Table . 1 indicates no significant differences to exist for factor G (levels of reinforcement). Statisti cal hypothesis H^-jS = X2 = Xj = X^ was supported. Figure 1 indicates no habituation taking place. Fast ! learners showed some change in GSR magnitudes over levels of reinforcement; yoked controls showed rather consistent responding. Table 2. indicates the GSRs of the fast learners fluctuated more to the discriminanda than to the warning light. Hypothesis Experimental hypotheses H^a, H^, H^c and tested for the possible two and three-way interaction effects. For all experimental hypotheses, their statistical hypotheses stated as null hypotheses were supported; no interaction jeffects of combinations AB, AC, CB and ABC were large enough |to deny their null hypotheses. TABLE 1.— Summary of analysis of variance for fast learners! and yoked controls (GSR data) | Source of Variation SS df MS P Between Ss .8684 7 .1241 A .0256 1 .0256 S w Groups .8428 6 .1405 Within Ss 1.7780 56 .0317 B .2631 1 .2631 AB .0912 1 .0912 B x S w Groups .4391 6 .0772 C .0698 3 .0233 AC .0628 3 .0209 C x S w Groups .6425 18 .0357 BC .0396 3 .0132 ABC .0055 3 .0018 BC x S w Groups .1644 18 .0091 TABLE 2.— Mean GSR-magnitudes of fast learners and yoked controls to stimulus situations during the different levelsj of reinforcement Warning Light Past Learners Yoked Controls 50# 65# 75# 90#+* .24327 .23429 .32639 .27552 .13469 .16204 .16472 .15610 Discriminanda Past Learners Yoked Controls 50# 65# 75# 90#+ .34936 .21877 .38000 .39213 .46266 .31677 .34345 .35963 ♦Percentage of correct responding equal percentage of . trials Ss were reinforced (levels of reinforcement); for statistical purposes Ss are compared on levels of rein forcement. 68 FIGURE 1.— Comparisons of averaged GSR magnitudes of fast learners and yoked controls to the stimulus situations at:the different levels of reinforcement (level of reinforcement equals level of learning by fast learners) 0 - ---- 20 .00 ___ levels of Reinforcement 75% 65% Key: Fast learners Choice *---a Yoked Controls Choice * > • ---® Fast learners light a---* Yoked Controls light •---4 The results of the experimental hypotheses presented] below belong to the family of research hypotheses through Ho which were concerned with percentage of eye orientation ! ® differences between fast learners and their yoked controls. Table 3 is a summary of analysis of variance of the percent age of eye orientations for the fast learners and their yoked controls. The experimental hypotheses of research hypotheses through Hq were tested simultaneously in a 2x2x4 analysis of variance. Table 4 gives the mean percentage of eye orientations of the fast learners and j yoked controls to the two stimulus situations during the i different levels of reinforcement j Figure 2 is a graphic comparison of the averaged j percentage of eye orientations of the fast learners and j yoked controls to the warning light and discriminanda during the four levels of reinforcement. i hypothesis Experimental hypothesis H^a tested the causality between speed of learning and percentage of eye orienta tions. Table 3 indicated no significant differences occurred for factor A (fast learners, yoked controls). Statistical hypothesis H^a1: X.| = X2 was supported. Figure 2, on appearances, indicates fast learners are giving higher averaged percentage of eye orientations than jthe yoked controls. Table 4 supplements this observation. * " 70 Hypothesis Experimental hypothesis Hga tested the differential effect of the stimulus situations to elicit eye orientations of the Ss.at the different levels of reinforcement. Table 3 indicates no significant differences occurred for factor B (warning light, discriminanda). Statistical hypothesis H6a1: ^1 = ^"2 was suPP°r‘ fced* Figure 2, on appearances, indicates fast learners are differentially responding more to the discriminanda than to the warning light; yoked con trol members are responding more to'the warning light than to the discriminanda after the 75% level of reinforcement was reached. Hypothesis H^ Experimental hypothesis H^a tested the habituation differences of the fast learners and yoked controls during the different levels of reinforcement. -Table 3 indicates no significant differences occurred for factor C (level of reinforcement). Statistical hypothesis H^a1: = Xg = X3 = X4 was supported. Figure 2 indicates no habituation taking place. Fast learners showed fairly consistent ‘ responding while yoked controls fluctuated in the responding i Hypothesis H£ Experimental hypotheses Hga, Hg^, Hgc and Hgd tested for the possible two and three-way interaction effects. All statistical hypotheses stated as null hypotheses were supported. Ho interaction effects of combinations AB, AC, 71 TABLE 3.— Summary of analysis of variance for fast learners j and yoked controls (behavioral data) Source of Variation SS df MS P Between Ss .9908 7 .1415 A .4472 1 .4472 S w Groups .5436 6 .0906 Within Ss 2.5375 56 .0453 B .00002 1 .00002 AB .1691 1 .1691 B x S w Groups 1.4342 6 .2390 G .1282 3 .0427 AC .0875 3 .0291 C x S w Groups .3431 18 .1900 BC .1148 3 .0382 ABC .0757 3 .0252 BC x S w Groups .1849 18 .1020 - - - 1 - TABLE 4.— Mean percentage of eye orientations of fast learners and yoked controls to both stimulus situations during the learning of a discrimination Warning Light 50% 65% 75% 90%+* Past Learners .85 .90 .84 .93 Yoked Controls .87 .83 .66 .90 Discriminanda 50% 65% 75% 90%+ Past Learners .98 1.00 .95 1.00 Yoked Controls .81 .85 .66 .53 '♦Percentage of correct responding equals percentage of 1 trials Ss were reinforced (levels of reinforcement); for ; statistical purposes Ss were composed on levels of rein- ! forcement. 72 FIGURE 2.— Comparisons of averaged percentage of eye orientations of fast learners and yoked controls to the stimulus situations at the different levels of reinforcement (equals levels of learning by fast learners)________ 1.00 r T & f 5 P- a 3 ? P - P £ 3 L £ I r F & O 5 ( 0 80 ft- 60 .00__ levels of Reinforcement 50% Key: Fast learners Choice *----- - a Yoked Controls Choke©— — — © Fast learners light a----- Yoked Controls light*----- -• ................ 73.| CS and ABC were large enough to deny their null hypotheses, j The results of the experimental hypotheses presented below are associated with research hypotheses Hg through which were concerned with GSR differences between fast jand slow learners. Table 5 is a summary of analysis of ! [variance of the GSRs for fast and slow learners. The experi mental hypotheses of research hypotheses Hg through were tested simultaneously in a 2 x 2 x 4 analysis of variance. i Table 6 gives the mean GSR magnitudes of the fast and slow learners to the two stimulus situations during the different levels of learning. Figure 3 is a graphic comparison of thej averaged GSRs of the fast and slow learners to the warning light and discriminanda during the four levels of learning. [ Hypothesis Hg ! ^ ! Experimental hypothesis Hga tested the quantitative differences in the magnitude of GSRs by fast and slow learners to the stimulus situations during the learning of [a discrimination. Table 5 indicates no significant differences to occur for factor A (fast learners, slow learners). Statistical hypothesis Hga1: = Xr> was supported. Figure 3, on appearances indicates the fast [learners to generally be giving a higher averaged GSR to Ithe stimulus situations than the slow learners but the difference was not great enough to be significant. Table 6 [supports this observation. | 74 Hypothesis H1Q Experimental hypothesis H^Qa tested the qualitative differences in magnitude of GSRs of fast and slow learners jto the stimulus situations during the learning of a I discrimination. Table 5 indicates that a significant i I difference for Factor B occurred. Statistical hypothesis H10a1: = ^2 failed ‘ to be supported; the alternative hypothesis H-jQa2: X-j > X2 > .05 d.f. 6,1 was supported. Table 6 indicates that both fast and slow learners responded with higher mean GSRs to the discriminanda than to the warning lights. Figure 3 indicates differential responding to the stimulus situations by the fast learners at the 50$ j and 90$+ levels of learning and markedly at the 50$ level j by the slow learners. Hypothesis - j Experimental hypothesis H ^ a tested the habituation process of the fast and slow learners during the different ;levels of learning. Table 5 indicates no significant differences occurred for factor G (levels of learning). Statistical hypothesis H.^^: X1 = X2 = X^ = X^ was supported. Figure 3» on appearances, indicates the slow |learners habituating to both stimulus situations while the fast learners were continuing to give high averaged GSR magnitudes. |Hypothesis H^2 | Experimental hypotheses H^2a. H12b, H^2c and H^2d ! 75 TABLE 5.— Summary of analysis of variance for fast and slow learners (GSR data) Source of Variation SS df MS P Between Ss 1.0685 7 .1526 A ... ... .2590 1 .2590 S w Groups .8085 6 .1349 Within Ss 1.4986 56 .0268 B .1115 1 .1115 6.00* AB .0124 1 .0124 B x S w Groups .1115 6 .0186 G .0306 5 .0102 AC .2049 3 .0680 C x S w Groups .7248 18 .0403 BC .0550 3 .0183 ABC .0368 3 .0123 BC x S w Groups .2121 18 .0118 *P = 6.00 d.f. 6,1 .05 TABLE J6♦▼-Mean GSR magnitudes of fast and slow learners during the learning of a discrimination Warning Light 50$ ; 65$ 75$ 90$+ Past Learners .24327 .23429 .32639 .27552 Slow Learners .05139 .22366 .14455 .08097 Discriminanda 50$ 65$ 75$ U3 o + Past Learners .34946 .21877 .38000 .39213 Slow Learners .31010 .33345 .19827 .10956 76 FIGURE 3*— Comparisons of averaged GSR magnitudes of fast and slow learners to the stimulus situations at the different levels of percentages of correct responding (levels of learning) 70 ’ ■ v / y o'n 20 levels of Learning 50% 15% Keys Fast Learners Choice a------ a Slow Learners Choicer----- o Fast Learners Light a----- * Slow Learners Light •----- • ! ' ' 77.| itested for the possible two and three-way interaction jeffects. For all experimental hypotheses, the statistical hypotheses, stated as null hypotheses, were supported; no | interaction effects of combinations AB, AC, CB and ABC were large enough to deny their null hypotheses. i The results of the experimental hypotheses presented below are associated with research hypotheses H-j^ through TT 16 which are concerned with percentage of eye orientation differences between fast and slow learners. Table 7 is a i summary of analysis of variance of the percentage of eye ! orientations for fast and slow learners. The experimental j hypotheses of research hypotheses through were j | tested simultaneously in a 2 x 2 x 4 analysis of variance. Table 8 gives the mean percentage of eye orientations of the! I fast and slow learners to the two stimulus situations during the different levels of learning. Figure 4 is a graphic comparison of the averaged percentage of eye orientations of the fast and slow learners to the warning light and discriminanda during the four levels of learning. Hypothesis Experimental hypothesis H ^ a tested the qualitative ; difference between fast and slow learners in mean percentage of eye orientations to the stimulus situations during the learning of a discrimination. Table 8 indicates a |significant difference occurred for factor A (fast learners and slow learners). Statistical hypothesis H ^ al: = X2 ' 78 i j was not supported. The alternative hypothesis H ^ a2s | > X2 < .05, I * = 87.60 d.f. 6,1 < .01 was supported. Table 8 indicates the significant differences that occurred between fast and slow learners was due to the differences jin eye orientations to the discriminanda. j Experimental hypothesis H ^ a tested the qualitative differences between fast and slow learners in mean percentage of eye orientations to the stimulus situations during the learning of a discrimination. Table 8 indicates a significant difference occurred for factor A (fast ! learners and slow learners). Statistical hypothesis : X1 = X2 was not supported. The alternative hypothesis H 1 3 a 2 : X1 > X2 < .05, E = 87.60 d.f. 6,1 < .01 was j supported. Table 8 indicates the significant difference i that occurred between fast and slow learners was due to the differences in eye orientations to the discriminanda. Hypothesis i Experimental hypothesis H ^ a tested the qualitative differences in mean percentage of eye orientations of both Ifast and slow learners to the different stimulus situations during the learning of a discrimination. Table 7 indicates jno significant difference occurred for factor B (warning light, discriminanda). Statistical hypothesis H ^ a^; j _ _ _ X^ = X2 was supported. Eigure 4, on appearances, indicates that both fast and slow learners were differentially j responding to the stimulus situations. However, whereas thej fast learners were responding more to the discriminanda than |the warning light for the slow learners, the results are just the opposite^. i Hypothesis Experimental hypothesis H^, having qualitative concerns, tested the effect of different levels of learning j on the habituation process, a progressive reduction in the ; j mean percentage of eye orientations of the Ss during the two: stimulus situations. Table 7 indicates that a significant j difference occurred for factor C (levels of learning). I ! Statistical hypothesis s = X2 = X^ = X^ was not | supported. Significant differences occurred for levels of j learning: F = 27.92, d.f. 6,1 < .01 but not as anticipated by either H ^ a2 or H ^ a^. Eigure 4 graphically and Table 8 numerically, indicate the patterns of responding for both |the fast and slow learners were: increasing percentage of eye orientations going from the 50% to the 65% levels, decreasing from 65% to 75% and finally, increasing, going from the 75% level to the 90%+ level of learning. The means for both fast and slow learners were: X^ = 85%, X2 = 91%, X3 = 85%, and X^ = 91% eye orientations to the available stimuli during appropriate trials. i ‘ Hypothesis j Experimental hypotheses Hi6a» Hl6b» H16c ' tested the 80 possible two-way interaction effects. The results indicated the effects of each of the combinations of main effects AB, [AC and BC were not enough to exceed chance. Statistical j hypotheses H ^ a1, H1gb1 and stated as null hypotheses, were supported. Experimental hypothesis H-jg^ tested the J Ithree-way interaction effect. Table 7 indicates an ABC interaction to exist: P = 26.86, d.f. 18,3 < .01. The significant three-way interaction was then investigated to determine which two factor interaction effects were i differing in magnitude at the various levels of a third j factor: AC (B), AB (C), or BC (A). Tests for an inter- J action between a particular factor and two-factor inter- ! I actions revealed that the three-way interaction was due to the C factor (at 50% level of learning) interacting with thej i AB interaction (P = 8.81, d.f. 1,24 < .01). Thus, the j I three-way interaction is AB (G). | Comparisons were then made of A at b^c^ and B at - L J a.c. to discover which levels of the A and B factors were 1 J contributing to the interaction. Using the formula: tAbi°d %/*?' ~ (Kirk> 1968) d y 2n. ms w. all the results were ^Ab^c^ = .003 d.f. 1,48 > .05 and ^At^c-j = |4.5 d.f. 1,48 < .01. Por the comparisons of B at a.c., the ifollowing formula was used: j ^ B a . c . - — *•*» ~ ^ 2- ! 1 3 "v/ssbx s.w.g. +S5bc X -s.w.g. (Kirk, 1968) i A. f . b. X s.w.g. + d .f.b. X ■ s- w -§* 81 The results were: = 2.06 d.f. 1,24 < .025» and ’ tBa2e1 = 2.14 d.f. 1,24 < .025. This is interpreted to mean that the frequencies of eye orienting produced by the combination of fast and slow learners to the discriminanda at the 50% level of learning is greater than that produced by the unit additive effects of the three factors. Figure 5» Figure 5 represents the backward learning curves of the experimental Ss (fast and slow learners); percentage learned is plotted over sessions. Point N in the figure is the total number of sessions required for any S to reach criterion; or numerically, it was the last session; i.e., for a particular £ five sessions were required to reach criterion, session five was the last session and N = 5. Reading from right to left, N-1 is the next to last session for all Ss and so on until the first session was reached which is numerically one less than N, i.e., if for a particular S five sessions were required, the first session would be at N-4. The curves for the individual Ss showed more homogenuity in the rate of learning after the level of learning around 65% correct was reached. When the curves for the experimental Ss were averaged by groups (and by Ss) at the points N, R-1, N-2, ... N-5, the curves indicated little differences in the rate of learning after the 65% level of learning was reached. 82 Figure 6 Figure 6 is an averaging of all the Ss together at corresponding points of N, N-1, ... N-5 for illustrative purposes to show the deviation from the "grand" mean by both groups. The deviations by the groups appear small and the general observation by Zeaman and House regarding the ^ equality in the rate of improving is supported. 83 TABLE 7.— Summary of analysis of variance learners (behavioral data) for fast and slow Source of Variation SS df MS F Between Ss .1404 7 .0200 A .1314 1 .1314 .87.60* S w Groups .0090 6 .0015 Within Ss .4853 56 .0087 B .0064 1 .0064 AB .0927 1 .0927 B x S w Groups .1064 6 .0177 C .0401 3 .0134 .27.92*-* AC .0066 3 .0022 C x S w Groups .0865 18 .0048 BC .0140 3 .0047 ABC .0412 3 .0137 .26.86*** BC x S w Groups .0914 18 .0051 *F = 87.60 d.f. 6,1 .01 **F = 27.92 d.f. 18,3 .01 ***F = 26.86 d.f. 18,3 .01 TABLE 8.— Mean percentages of eye orientations of fast and slow learners to both stimulus situations during the learning of a discrimination Warning Light Fast Learners Slow Learners 50% 65% 75% 90%+ .85 .90 .84 .93 .87 .95 .85 .83 Dis criminanda Fast Learners Slow Learners , 50% 65% 75% 90%+ *98 1.00 .95 1.00 .73 .82 .81 .88 84 FIGURE 4.— Comparisons of averaged percentages of eye orientations of fast and slow learners to the stimulus situations at the different levels of percentages of correct ___________responding (levels of learning) 1.00 80 o cr> 60 00 Levels of Learning 65# 75# Key: Fast Learners Choice a- ----a Slow Learners Choice o----o Fast Learners Light *----* Slow Learners Light •---- • 85 FIGURE 5.— Averaged backward learning curves for fast and slow learners 100# 90# 80 # 70# 60# N-4 N-1 Sessions Key: S 11 a----- a S 51 *----- -* 21 Ar------a “ 81 a-------n 41 © - — •— ° 1 11* 86 FIGURE 6,— Averaged backward learning curves for individual and pooled groups 30% 10% 60% Sessions N-4 N-3 N-1 N-2 Kfey: Fast Learners o------ o Slow Learners A------ A Slow Learners and Fast Learners a--- CHAPTER IV i I DISCUSSION AND CONCLUSION | This chapter overviews the study and its findings, then discusses these findings in terms of the involvement of attention in learning. Overview of Study This section summarizes the essential aspects of thej literature that gave purpose to this study} summation of purpose and of findings follows. I ! Zeaman and House (1963) reorganized learning curves ; to test whether attention or intelligence was the primary determinant in speed of learning a two-choice discrimination. The backward learning curves revealed their fast and slow learners differing only in the number of trials required before the chance level of responding (50% correct) was surpassed. After the chance level of responding correctly was surpassed, the rate of improvement or learning by their ;fast and slow learners was the same. Zeaman and House thus concluded that attention was the primary determinant in |speed of learning. The effect of attention was inferred on the basis of |the behavior which attention was assumed to influence. Their methodology generated circularity in the explanations 87 88 of the results. As pointed out by Maltzman and Raskin (1965), studies of perceptual processes require objective j independent measures to avoid this typical methodological jshortcoming. Maltzman and Raskin used the OR to study awareness during semantic conditioning and generalizations. The OR is a readiness response by an organism to the onset of novel and/or significant stimuli. The initiation or inhibition of the OR is controlled by the cortex (Sokolov, 1960). The intensity and/or duration of the OR is j correlated with the unfamiliarity of the stimulus and the I j difficulty of the discrimination to be made between stimuli j (Grastyan, 1961). These characteristics of the OR make it j the mechanism for paying attention (Lynn, 1966). j Following the lead of Maltzman and Raskin (1965), the OR was used in this study as an objective independent measure of attention to avoid circularity in the methodology and to extend knowledge of attentional differences in the mentally retarded. i Summary of Purpose This study sought to contribute some clarity to the : Zeaman and House theory of the role of attention in mental ,retarded two-choice discrimination learning. | Summary of Findings i During the course of learning the discrimination, reinforcement was delivered for correct choices. To check i ifor the possible modifying effects of reinforcement on the 89 ORpGSR and eye fixations of the yoked control Ss were compared with their matched fast learner members. The j findings are summarized first in terms of fast learners and yoked control comparisons, then in terms of fast and slow | |learner comparisons. The results are as follows: 1) Tests of the hypotheses concerning GSR comparisons revealed the following: (a) fast learners and their yoked controls did not differ as groups in their GSRs, (b) these two groups did not differentially respond to the two different stimulus situations (warning light, discrimi- i nanda), (c) GSRs did not change with different levels of j reinforcement, and (d) the factors: subjects, stimuli and I j reinforcement were not operating in combination or inter- j acting. ! 2) Tests of the hypotheses concerning eye fixation comparisons between fast learners and their yoked controls revealed the following: (a) fast learners and their yoked controls did not differ as groups in frequencies of eye fixations, (b) these two groups did not differentially respond to the two stimulus situations (warning light, discriminanda), (c) frequencies of eye fixations did not ; change with different levels of reinforcement, and (d) the Ifactors: subjects, stimuli and reinforcement were not operating in combination or interacting. These results indicate that GSRs and eye fixations |as of the OR were not modified by reinforcement, that the 1 ’" ' 90] stimulus situations were equally significant for both fast j learners and their yoked controls and further, that habituation in GSR and eye fixations by both groups did not occur to either of the stimulus situations. The results also showed that the factors were not operating in combi nation, thus possible latent effects of the factors finding i expression as combinations did not exist. 3) Tests of the hypotheses concerning GSR compa- i risons revealed the following: (a) fast and slow learners ! i did not differ as groups in their GSRs, (b) these groups did differentially respond to the two stimulus situations, (c) GSRs did not change with different levels of learning, and j (d) the factors: subjects, stimuli and reinforcement were not operating in combination or interacting. ! 4) Tests of the hypotheses concerning frequencies of eye fixations revealed that a three-way interaction occurred between the factors. Further examination of the three-way interaction revealed that at the 50% level of learning, both levels of Ss and the discriminanda level of stimuli were contributing to the interaction effects. The results of the comparisons of the fast and slow jlearners in terms of the GSRs and eye fixations indicate ’ that fast and slow learners in terms of their GSRs do not Idiffer quantitatively in their attention during the learning: |of a discrimination. The differential GSR responding by fast and slow learners to the two stimulus situations does not reveal attentional differences per se but does suggest that the outstanding characteristic of mental retardate 'attention is its qualitative nature. Habituation of GSR and eye fixations of fast and slow learners did not occur to jeither of the stimulus situations. The results of the frequencies of eye fixation measures indicate that qualitative differences exist between fast and slow learners. I Results of each hypothesis tested in this investi gation shall be presented to be followed by a discussion of results and a review of the literature dealing with the ! involvement of attention in learning. j j GSR Comparisons of Fast Learners and.Yoked Controls Fast learners and yoked controls did not differ in their GSR magnitudes to the stimulus situations in all points of comparison. The test of the causality between magnitude of the GSR and speed of learning was inconclusive. Comparisons of GSR magnitudes of fast and slow learners (later to be dis cussed) revealed no differences occurring, suggesting the test of the hypothesis as stated was not appropriate. The lumderlying rationale for testing the causality was.to [determine the effect of reinforcement on the magnitudes of j I the GSR. The no-differences result between the fast learners and yoked controls demonstrated that reinforcement ;was not affecting the magnitude of the GSR. The 92 non-contingent reinforcement of the yoked controls for their choices did not modify their behavior from involvement to non-involvement. Statistically, the stimulus situations had i the same GSR eliciting potentials for the fast learners and yoked controls. The GSR, eliciting potentials of the stimu lus situations, was dependent upon the significance of the stimuli to the Ss. The significance of the stimuli was the same for the fast learners and their yoked controls and of a rather high degree as evidenced.by the lack of habituation j I i during the different levels of reinforcement. The lack of any two-way interaction effects indi cated that the factors were not operating in combination to I produce greater GSR magnitudes than that accounted for by i the additive effects of the factors. Importantly, the j effect of reinforcement previously shown to be benign, did not become manifest in the two-way interactions in which reinforcement was one of the factors involved. The non- occurrence of the three-way interaction revealed, as above, that factors involved were not operating in a multiplicative way to produce results greater than that accounted for by the additive contributions of the factors. Behavioral Comparisons of Past Learners and Yoked Controls Fast learners and yoked controls did not differ in their averaged percentage of eye fixations in any points of ! comparison. Hypotheses involving the behavioral data were janalogus to those for the GSR data and the discussion and 93 ; I conclusion of the analysis of the behavioral data shall be somewhat abbreviated. The causality between speed of learning and percent age of eye orientations was not demonstrated as stated in the hypotheses. However, reinforcement, an important issue |in the causality tested, was shown to have no effect on eye orientation. This conclusion is underscored by the lack of a two-way interaction involving levels of reinforcement as one of the factors. Statistically, the stimulus situations were shown to be equally significant to both fast learners j and yoked controls to the extent that habituation to the stimuli did not occur over the different levels of reinforce-- ment. The lack of any two-way interactions, as in the GSR data, indicated the factors were not operating in combina tion to produce higher average percentages of eye orienta- , tions than that accounted for by the additive effects of the factors. The non-occurrence of the three-way interaction indicated the three factors were not operating in combina tion to produce effects greater than that, accounted for by the additive contributions of the factors. GSR Comparisons of Past and Slow Learners In the tests to determine the quantitative- qualitative nature of attentional differences, quantitative idifferences in GSR or behavior were not detected. The one significant finding indicated differential responding to ! 94 1 the two stimuli by both fast and slow learners suggesting the outstanding characteristic of mental retardates, attention, is its qualitative nature. The comparisons of the GSR magnitudes of the fast land slow learners to test the hypothesized quantitative [difference, based on the Maltzman and Raskin (1963) finding that levels of orienting are related to learning perfor mance, revealed no differences occurring. The magnitude of the GSR, as an index of the OR, in itself did not prove to j be a causal factor in the speed of learning. The test to i ascertain whether the qualitative aspect was a relevant I issue in the study of mental retardate attention produced positive results. Past and slow learners differentially i responded to the two stimulus situations. The stimuli of j | the discriminanda, having greater significance or informa- j tional value, were responded to with higher GSR magnitudes. Over the different levels of learning, fast and slow learners did not habituate to the stimulus situations. The slow learners did show a trend towards habituation to both stimulus situations. The fast learners were just the oppo- ; i site, giving progressively higher GSR magnitudes to the jstimuli of the discriminanda. The two-way interaction effects which were to detect latent reinforcement effects, and/or provide further infor- Imation about the qualitative aspects in terms of attentional i j differences between fast and slow learners, did not occur. 95 | The lack of any of the two-way interaction effects indicates that none of the hypothesized two-factor combinations was operating to produce greater GSR magnitudes than that which Would be predicted by their additive contributions. The pLack of the three-way interaction effect means, again, the jthree-factors involved are not operating in a multiplicative way to produce GSR magnitudes greater than that predicted by their additive effect. Behavioral Comparisons of Fast and Slow Learners | The results of the analysis of the behavioral data amplifies the possible qualitative differences involved in i the GSR data. The significant three-way interaction indi cated fast learners to be qualitatively superior in j orienting differentially to the stimulus situations at the i 50% level of learning. At the 50% level of learning, the ; fast learners gave a higher percentage of eye orientations to the stimulus figures of the discriminanda than to the warning light. At the 50% level of learning, the slow learners gave a higher percentage of eye orientations to the warning light than to the stimulus figures of the discrimi nanda. i Discussion of Results In overview of this section to be discussed are: !(a) the non-habituation of the Ss which is explained in j terms of stimuli becoming significant and acquiring signal lvalue, (b) a commentary of the Zeaman and House theory in terms of the role of attention in the learning process, (c) research related to attentional differences, (d) developmental studies of attention, (e) the Russian studies of perceptual development demonstrating qualitative differ- jces in orienting exploratory movements, (f) the results of Russian studies in relation to the present study in terms of development modes of information gathering, (g) Russian Psychology versus Western Psychology on the developmental nature of perception, and (h) training of selective ! i attention as a partial settlement of the developmental j i nature of perception. The chapter concludes with a summa- j tion of the present study. Non-Habituation of Ss The non-habituation by all Sis involved in this study is an interesting result explainable in terms of stimuli becoming significant and acquiring signal value. Stimuli take on signal value when they develop a particular significance for the observer (Sokolov, 1963). i The developing significance of stimuli does not mean that jmotivational, emotional changes are taking place concomi-*- tantly. The OR functions to enhance the reception of stimuli selected by the sensory analyzer in the cortex. The OR does not function as an energizer; the defensive ! reaction to strong, noxious stimuli performs this function (Maltzman and Raskin, 1965). In this study the reinforcers, delivered three seconds after a correct choice, were involved in the acquisition of signal value of the stimuli. The effect of j ! the reinforcer did not modity the magnitude of the OR but did influence its non-habituation. Sokolov (1963) and Lynn (1966) reported that when stimuli acquire signal value, habituation of the OR to these stimuli can require hundreds of trials. Commentary of the Zeaman and House Theory Zeaman and House (1963) stated that in the attending; behavior of the mentally retarded there is the "possibility I that the relevant cues are not attended to on every trial because the £ > may first have to learn to attend to the i j relevant stimulus dimension" (p 166). On the basis of thesej data, one could conclude that even before the learning to j i attend to the relevant stimulus dimension, the ; S must first discover which of the classes of stimuli are of instrumental value for solving the discrimination. Next, the S_ must learn how to acquire related informational cues (attending : to both of the stimulus figures of the discriminanda) before stimulus dimensions can be explored to discover which of the dimensions are relevant. In clarification of the Zeamar. and House (1963) theory of differences of attention as that which distinguishes fast and slow learners, fast and slow i learners differ qualitatively in their attention to stimuli. Research Related to Attentional Differences ! i As previously mentioned, direct investigations of ■ ■ ■ . . . 98 I qualitative and quantitative differences in attention (with j the exception of the awareness studies by Maltzman and Raskin) were not found. Extrapolations of research with normal children will now be presented which lend support to jthe results of this study. i The work of Irzhanskaia and Felderbaum (1967) demonstrated that prior sensitization of premature infants ■(1^ to 2h months old) to a stimulus (mint odor) fostered the development of future conditioned reflexes to the stimulus. j The stimulus of mint odor, which the infants had become sensitized to, had thus, become a very significant stimulus to the infant and thereby, increased the probability that iti would be selectively attended to (oriented toward). All thej i infants, two weeks prior to the conditioning, were exposed j i i to the mint odor via feeding them milk through nipples j stored in mint. For one half of the 33 infants, mint odor on the nipples (CS) was paired with a stream of air (UCS). On the first trials, mint and air were presented simulta- i neously and later air was delayed three or four seconds. After mint and air presentations, anise and air were presented during the same session; the other infants re ceived anise and air during each session. Irzhanskaia and !l'elderbcaum concluded that ”... conditioned reflex formation I also depends on the child's capacity to differentiate among stimuli, not only in terms of their physical qualities, but |also in terms of the degree to which they are significant for that organism" (p 249). The OR, although weakly, j preceded the formation of the conditioned reflex. In terms of this study, the classes of stimuli had idifferential significance for both the fast and slow jlearners. Because of the different modes of attending :(selective attention to one or both of the stimulus objects of the discriminanda), the fast learners were able to differentiate among the stimuli with greater rapidity and to recognize sooner the differential significance of the j j stimulus objects. The Maltzman and Raskin (1965) studies indicated (in terms of the OR which was to represent aware- j j ness) that for college sophomores quantitative differences j i were related to better performance in semantic conditioning.; Qualitative differences in awareness were not investigated but it is suspected, by this writer, that given the Ss had reached their peak in cognitive development, J their ability in developing strategies to acquire relevant information for problem solving is probably not different. This is not to say that they make equal use of the infor mation once acquired, but only, that they know where to loot: for it. Developmental Studies of Attention Apparently, quantitative-qualitative differences in ; orienting (whether orienting represents attention or i awareness or some other perceptual process as it affects j ! performance) are of a developmental nature with experience 100 facilitating the learning of differences between stimuli. For the very young, differences in attending appear to be predominantly qualitative. As cognitive development proceeds to maturity, differences in attending would appear to be predominantly qualitative. As cognitive development proceeds to maturity, differences in attending would appear to be progressively more quantitative. The work of Kagan (1970) gives support to this | contention. An important assumption that Kagan was opera- I I ting under is that "with age, the child becomes more selec- j tively attentive and better able to differentiate the rele- j vant signals from background noise" (p 827). The work j reported by Kagan of studies in infant attention was in j terms of preferences of the young to attend to certain j i classes and/or patterns of stimulation. The studies repor- j ted the development of selective attention in the infant. During the first months of life, the infant's attention was chaotic, captured by "stimuli that move, have many discrete elements or possess contour contrast" (p 827). During this time of orienting to the environmental stimuli, relative priorities of stimuli, in terms of their significance to the infant, began to develop. These most significant stimuli I occurred together from the basis for schemata. Jeffrey's serial habituation hypothesis was an ! account of how the orientations by the infant to related | I stimuli gave rise to schemata. A schema, according to Kagan (1970) "is a representation of experience that pre- j serves the temporal and spatial relations of the original jevent, without being necessarily isomorphic with that event” l(p 827). Kagan pointed out that a central assumption |surrounding early schemata formation states that the first schema represents invariant stimulus patterns that are part of a larger context characterized by high rates of change (movement, contour, contrast and acoustic shifts). The invariant arrangements, eyes, nose and mouth of the human | i face (and because of the significance of these stimuli, particularly a mother's face which represents nurturance, comfort, protection, etc.) would make the human face one of | the earliest schemata. Experimental observations of young I infants suggest this is so. j With the development of schemata somewhere during j | the second month of life, the attentional preference | switches to those stimuli that are moderately discrepant with the schemata the child now posesses for that class of stimuli. Moderately discrepant stimuli will "elicit longer orientations than do either minimally discrepant (that is, familiar) events or novel events that bear no relation to the schemata. The relation between attention and magnitude iof discrepancy is assumed to be curvilinear (an inverted U). Although an orientation reflex can be produced by any change I in quality or intensity of stimulation, duration of sus tained attention is constrained by the degree of discrepancy 102 | i between the event and the relevant schemata" (Kagan, 1970, | I p 828). | Russian Studies of Perceptual Development A. V. Zaporozhets (1965) used the orienting explo ratory movement to study perceptual development in the pre school child. Orienting exploratory movements occurred when perception, as a process, goes beyond the direction of the sensory analyzer, during an OR, to employment of another sense modality (tactual in this case). Other sense modali- j ties were employed as the organism's activity investigated j the characteristics or properties of stimuli. Prom the presentation by Zaporozhets of the Russian research, in ; ontogenic changes in the orienting exploratory movement, it appears that the essential characteristics of this movement is the purposive exploration or investigation of the source of stimuli following the onset of the OR (D. B. El'Konin, 1969). Berlyne (1960), from an S-R point of view, developed a concept of exploratory behavior which was based on the orienting response. With a major interest in the processes that subserve selective attention, he postulated that exploratory responses served to "help one stimulus to win the contest for attention by raising its intensity and weakening or eliminating its most formidable rivals" (p 78). This is a synpnjjraious description of the function of the OR. However, Berlyne designated other functions to the 103 i concept of exploratory movements. Besides its stimulus function, the principal function of exploratory responses is "to afford access to environmental information that was not previously available" via searching or exploring the environ ment to create the presence and/or clarity of particular ! stimuli (Berlyne, 1960, p 79)* This discussion of the orienting exploratory move ment and exploratory response is to whow, conceptually, that orienting reactions, orienting exploratory movements and ; exploratory responses, are not really different (all maximize the reception of stimuli) and all can be used to ! [ represent the process of attention. ! A. V. Zaporozhets (1965) reported the results of j j Russian developmental studies of perception in preschool i i children. The work of U. P. Zenchenko and A. G. Ruzskaya j i studied developmental differences in the haptic sense (active touch, object unseen) and visual sense (object explored visually but not touched) in preschool children. In the first experiment, movie film was used to record the haptic orienting exploratory movement. In one of the experimental groups, the children explored an object tactually and were later required to identify the object visually. The objects explored were of two shapes, one heart |shaped and one bread shaped. Both had smooth, flat surfaces; |The descriptions of the tactual movements of the different : 104 age groups, according to Zaporozhets (1965), were as j follows. The three year olds exhibited movements that "were more like catching than like touching" (p 85). The children appeared to be playing with the object rather than exploring it. His grasp of the object was palmar with the fingers pressing against the surface of the figure; the palm remained motionless. The four to five year olds also used a palmar grasp of the object with the thumb and four i fingers pressing against the figure. The four year olds j differed from the three year olds in that the four year oldsj soon began "to acquaint themselves with the objects more j activily by using the palms and the surfaces of the fingers" (P 85). The grasp of the object was generally onehanded and the fingertips, though they touched the object, were believed to be "absolutely passive in the tactile process" (p 85). The children that were five to six years old grasped the object with two hands, moving the hands towards or away from one another simultaneously. Bottom surfaces of the fingers and finger tips were employed to examine carefully specific features of the figure (concave portions, ! corners, projections, etc.). The entire outline of the figure was not traced. The six year olds had the most refined responses, using both hands, touching the figure Iwith their fingertips and systematically tracing the entire [ |outline of the figure "as if the children were reproducing r io5H i the form of the figure with their tactile movements by- modeling its form" (p 85). The results have been reported here at length not only because they were, interesting in themselves, but also to show the definite breaks in the modes of haptic orienting exploratory movements. Differences between the age groups are qualitative in nature with each succeeding age group showing a progressive refinement. Prom the data, however, one is not able to determine whether qualitative differences I I exist within groups nor is one able to determine whether j quantitative differences exist between groups and/or within | groups. The descriptions of the tactile explorations offered by Piaget and Inhelder, although sketchy and obser- J ved in a less controlled situation, are generally in agree- j ment with the work of Zenchenko and Ruzskaya (Zaporozhets, 1965). In their study of developmental differences in visual orienting exploratory movements, movie film was used to record eye movements. The children were given 20 seconds to explore the object and were required to later identify ;the object. The same two forms were presented individually. Again, the data showed changes in orienting exploratory movements occurring in the development of the child. A description of the eye movements characteristic of |each age group was as follows: the three to four year olds |showed little visual exploration of the objects. Eye ; 106 i ; 1 fixation times were longer than for the older groups. Eye i fixations were generally at the center of the figure with little exploration of the outlines. The children were many jtimes distracted by the camera lens (located in the center jof the figure), and were wrong half of the time during jrecognition. The children also mixed up figures that were quite different in form. For the four to five year olds, eye fixations were still within the figure but the number of eye fixations i increased greatly in number while concomitantly, fixation | i times were greatly reduced. The general tracings of eye fixations indicated possibly the child was "orienting to the size and length of the figure. Although we did not see movements tracing the outline of the object, ... we did I find some groups of fixing points that were close to each other and related to the most specific features of the object" (Zaporozhets, 1965, p 87). This age group had a Ihigher percentage of correct recognition than did the three year olds. The beginning of tracing the outline of the figures started with the five to six year olds; only part of the outline was traced, many of the fixations were still within the figure. The number of eye fixations of the five year olds was the same as the four year olds; however, the five I year olds had a higher percentage of later correct recogni tions. The six year olds made complete tracings of the ! ....... 107 j figure, as if reproducing or modeling its form. "At the same! i time, we could observe the movements across the figure. Probably, these movements also solved an important ^orienting problem, measuring the area of the figure” i (Zaporozhets, 1965, p 88). The number of eye fixations were larger in number and shorter in duration; the six year olds were 100% correct in later identifications of the..figures explored. Results of Russian Studies in Relation to the Present Study The qualitative differences between age groups that j characterized the development in the mode of orienting in haptic exploratory movements showed a similar development j i in the visual sense. In both situations, the three year j olds1 mode of exploring the objects were inappropriate in terms of making the necessary information available to gain sufficient acquaintance with the objects to be able to later recognize the objects with some degree of accuracy. Over successive age groups, the mode of exploring the objects progressed to the point where, in the case of the six year olds, attempts were made to gain a sensorial representation of the object by actively acquainting themselves with the entire shape and size of the objects. i Based on the differential amounts of eye orienta tions to stimuli by fast and slow learners at the 50% level J of learning, it has been postulated that mentally retarded children, prior to learning the relevant stimulus dimension, 1 108 j must first learn the class of stimuli having instrumental ! yalue and then, how to best acquire the relevant informa tion. This suggests teaching techniques generated from an |emphasis on process. In problem solving situations, such his learning a discrimination, performance is highly ! dependent on perceptual ability (i.e., selective attention). The improvement of perceptual abilities should, hopefully, lead to faster learning. Russian Psybhology Versus Western Psychology | Underlying the Russian educational system for j children is the position that sensory processes become gradually more complicated. As a result, "perceptive images, appearing at different ontogenetic stages, become more and more orthoscopic; that is, reflex the environments more fully and adequately" (Zaporozhets, 1965, p 82). This position, in opposition to the Gestalt theorists that "a newly born child possesses the basic features of perception in ready made forms" (p 82) places |perceptual abilities under developmental control. One difficulty with the Russian position is that development in cognition, as studied by Piaget, aids the child in giving a |more adequate or correct rendition of his world. Thus, |whether the causality of the development to more orthoscopic iperceptions is due to development in cognition, giving | ! perception only the appearance of being developmental, or j |due to neurological reasons, is not determined. r ' ~ ' 109 The Russian psychologists also did not study individual differences, but rather, group differences which | imay be adding to the developmental appearances of percep- i I tion. Educators can either conduct studies to determine the extent of a developmental influence of perception or stay with the position of the Gestalt theorists and study how perceptual processes such as selective attention might be improved. In hand with such studies would be how to instill cognitive strategies for acquiring information and the relation of one to another. Training of Selective Attention Gaines (1970) studied whether selective attention | to classes of stimuli (color and form) was a function of stage of development or set( predisposition! to respond predictably as a function of training). Gaines subsumed the term preference (responding to a particular stimulus dimension) under selective attention as a process that occurs as a function of preference, but, in addition, |suggested that the stimulus controls learning or produces a reliable response over time and conditions. The interest ing feature of Gaines' study to education and this investi gation is that "several studies showed that children's i consistent color and form choice does influence concept ^attainment, discriminations and reversal learning" (Gaines, i 11970, p 980). I The results of Gaines' investigations were that children of ages 4.6 to 7.6 years of age, ranging in IQ of 98-164, could learn to attend consistently and selectively ;to the dimension that was not their initial dominant preference. Summation of the Present Study ! In summation, Zeaman and House (1963) have theorized the role of attention to be the determining factor in the speed of learning a two-choice discrimination by mentally retarded children. Their methodological shortcoming of not having an objective independent measure of attention has | limited them to inferences about attention based on differences in learning performance and then explaining ! differences, in learning speed on the basis of attention. ! In this study, the employment of the OR provided a j ] means for an objective measure of attention that was inde pendent of learning performance. Thus, circularity was avoided in the methodology and the causality of attention ;in the learning was more clearly apprehended. The results of the OR comparisons (GSR and eye fixations) between fast learners and their yoked control members revealed no OR differences existing between the Igroups, to the stimulus situations, over the different i levels of reinforcement. The conclusion was that reinforce ments were not modifying the magnitude of the GSR nor the frequencies of eye fixations. The results of the OR comparisons, in terms of GSR 111] magnitudes, between fast and slow learners, revealed no j differences existing between these groups. However, both groups did differentially respond to the stimulus situations], ! ■suggesting the outstanding characteristics of mental retard- j dte attention is its qualitative nature. No differences in GSR magnitudes occurred during the different levels of learning, indicating non-habituation occurring to the sti^ muli involved in the study. It is concluded on the basis of these results that j fast and slow learners do not differ quantitatively in ! their attention during the learning of a two-choice discri- ! mination. j The results of the OR comparisons, in terms of j frequency of eye fixations, revealed a three-way interaction i between the factors occurred. The interaction indicated ! I i fast learners to be qualitatively superior in orienting differentially to the stimulus situations at the 50% level of learning. Thus, fast and slow learners differ only qualitatively in their attention to stimuli. Past learners, in giving higher percentages of eye orientations (particularly during early trials) indicated a more efficient mode of attending in that they were ibringing themselves into contact with necessary information for a correct choice. This supports the position of quali tative differences to be existing between fast and slow I Ilearners in their attention to stimuli during the learning n ' ' ..~ ... 112 |of a two-choice discrimination. ! i ! In a search for confirmation in the literature, no studies of quantitative and/or qualitative differences in attention of the mentally retarded (nor normal children) were found. Extrapolations of results from attention studies of infants and normal children indicate more efficient modes of encountering stimuli, suggesting attentional differences among the young to be primarily qualitative. Alternative conceptualizations of perception (Gestaltists and Russian) have been presented to educators j as issues for further investigation. The Gaines (1970) study indicated that selective attention as a preference for stimuli is modifiable and should generate much enthusiam: among educators concerned with improving teaching tech niques for the mentally Retarded and preschool child. APPENDIX Mean GSRs of Past Learners and Yoked Controls LIGHT Past Learners .25683 .42271 .29355 .00000 .20094 .37926 .14612 .21085 .15071 .97746 .10508 .27233 .22506 .57794 .05027 .24884 Yoked Controls .40754 .14820 .05664 .12640 .14149 .31495 .08767 .10408 .28058 .14818 .13638 .09374 .16385 .26557 .11748 .07751 DISCRIMINANDA Past Learners .19428 .61958 .30123 .28275 .00000 .144441 .37222 .35847 .00000 .88968 .35815 .27219 .41730 .55951 .12800 .46372 Yoked Controls .36383 .27574 .42576 .78533 .20657 .18925 .35245 .51881 .22502 .11458 .48947 .54475 .11335 .31524 .45282 .55712 Mean GSRs of Past and Slow Learners LIGHT Past Learners .25683 .42271 .29355 .00000 .20094 .37926 .14612 .21085 .15071 .97746 .10508 .27233 .22506 .57794 .05027 .24884 Slow Learners .03990 .07905 .08660 .00000 .41864 .15540- .13240 .17820 .01521 .24810 .16550 .11340 .01839 .15340 .06360 .08850 DISCRIMINANDA .19428 .00000 .00000 .41730 Past Learners .61958 _ .14441 .88968 .55951 113 50% 65% 75% 90%+ Fast learners Slow learners 114 30123 .37222 .35815 .12800 28275 .35847 .27219 .46372 35211 .39462 .12531 .11664 05890 .39310 .36000 .13380 59560 .44420 .15780 .10490 23380 .10190 .15000 .08290 Mean Percentage of Eye Orientations of Fast Learners and Yoked Controls LIGHT Fast Learners Yoked Controls Fast Learners Yoked Controls Fast Learners Slbw Learfcers 50% 65% 75% 90%+ .75 1.00 .75 .85 .92 .83 .91 .94 .85 .83 .88 .92 .90 .95 .83 1.00 .75 .66 .75 1.00 .85 .66 .44 .70 1.00 1.00 .65 .95 .90 1.00 .83 .95 DISCRIMINANDA 1.00 1.00 1.00 1.00 1.00 1.00 .85 1.00 1.00 1.00 1.00 1.00 .95 1.00 .95 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 .56 .15 ».25 .40 .08 .00 . * ■ ■ Bye Orientations of Fast and Slow Learners LIGHT .75 ' ■ 1.00 .75 .85 .92 .83 .91 .95 .85 .83 .88 .92 .90 .95 .83 1.00 .80 .90 .75 .75 .85 .85 .80 .95 .92 .95 1.00 .95 .92 1.00 .85 .70 115 DISCRIMINANDA Past Learners 50% 65% 75% 90%* 1.00 1.00 1.00 1.00 1.00 1.00 .85 1.00 1.00 1.00 1.00 1.00 .95 1.00 .95 1.00 .90 .80 .95 .95 .77 .85 .70 .85 .57 .80 .85 .85 .70 .85 .75 .90 T-Test of Differences Between Means of Past Learners and Yoked Controls Ionic Arousal During Session 1 Session 1 Pair 1 Pair 2 Pair 3 Pair 4 Controls 1.45303 1.20868 1.36213 1.13464 nED2 - (Ed)2 = 4(.173603) - (.19499)2 = .694412 - .038025 - .656387 S? = Ld = .656387 = .054698 d ri(n'- 'T 7 .. if. S?/n - .055 = .117 a — 3- .049 - 0 = .41 7— Exp erimentals 1.16980 1.46077 1.18958 1.14334 Ed = Ed* d .28323 -.25209 .17255 -.00870 .19499 .173603 .049 GSR Calibration Ohms A Ohms Sensitivity 50mv/cm 20mv/cm 50K . ' k . ' 51K 1000 4mm 9mm 75K 76K 1000 4mm 9mm 100K 101K 1000 3.5mm 9mm 116 Ohms AOhms Sensitivity 50mv/cm 20mv/cm 150K 151K 1000 3mm 8mm 200K 201K 1000 3mm 7mm 21 OK 211K 1000 2.9mm 7mm 225K 226K 1000 2.5mm 7mm 250K 251K 1000 2.5mm 6.5mm 300K 301K 1000 2mm 6mm 350K 351K 1000 2mm 5mm 400K 401K 1000 2mm 5mm 450K 451K 1000 2mm 4.5mm 500K 501K 1000 1,9mm 4mm Correlation Coefficient for Interrater Reliability. Sums of Frequency of Eye Orientations to Warning Light and Both Stimulus Figures During Twenty Trials r = NSXY - (EX)(EY) y [nEX2 - (EX)2] [NEY2 - (EY)23 r = 6(5272) - (177U175) ~]/ [6(5359)-31329] [6(5199)-30625], r = .96 X Y 24 25 28 28 35 34 27 25 37 35 26 28 117 Learning Record of Subjects 1 a correct choice, 0 = incorrect choice during a trial S11 Correct Color = Red Session 1 Session 2 10110 01110 00001 11111 10001 01000 10010 10011) : / ■ I ---------------------- 50%--------------------- » Session 3 Session 4 h i m 01000 00011 111111 [10111 11011 01111 01 m l I---------------------£C.at----------------- 1 *------------------- 7 5 ^ --------------------1 ■65$" Session 5 jl0101 11111 11111 11111 -90$+ S21 Correct Color = Green Ssssxon 1 ScsBion 2 101010 10100 01101 011111 101111 10111 01110 10100I ( --------- r - n a f----------- * T---------- CEO-/ 1 50$- Session 3 01011 00111 11110 01111 75$----- 65$' Session 4 111111 11111 11011 111011 1 ______ 90$+___l S41 Correct Color = Red Session 1 Session 2 110011 01110 oidoli 111|(11 01111 11011 11101 11011 I Ena*----:— Jl—--------- 50$- .. Session 3 | 0|1111 11111 11111 11111 75$' 90$+ S51 Correct Color = Green Session 1 Session 2 f1011(1 0 < L— j5C$- 2 2 P 11 1 65$— * 75$ 111f1 11111 11111 11111 11111 11111 -------- 90$+------------- S61 Correct Color = Red Session 1 1010111 30 1 75$ 1111 11111 11111 90$+--- < — * S81 Correct Color = Red Session 1 Session 2 [01110 01011 01111 00101 10101 11010 10010 011001 I ---------------- 50%- 1 Session 5 Session 4 , 101101 11-101 10010 10111 01111 11011 11010 01111 I ---------- 65% 1 1 --------- 75%----------1 Session 5 , Session 01111 10111 11111 11111 --------- 90%+-------- S91 Correct Color = Red Session 1 Session 2 |00110 00101 11101 10101 11010 10100 11011 10010 %----------------- . Session 3 Session 4 11011 00110 10110 01111 10110 01101 01001 01111 _________________________ (ktkaL__________________________ Session 5 Session 6 |l 1011 11011 01110 011101 101111 01111 11111 11111 75%---------1 I --------- 90%+ S111 Correct Color = Green Session " I |oi101 10111 01000 11110 10110 01000 11010 11010 I ---------------------- 50%---------------- -— — Ssssxon 3 (11010 10101 11011 100111 I ------- 65%----- 1 Sbssion 4 11011 11011 01111 01111 -------- 75%---------- i Session 5 11111 11111 11111 11111 ------ 90%+------— BIBLIOGRAPHY 1. Atkinson, R. C., A Markov Model of Discrimination Learning, Ps.vchometrika, 1958, 23* pp. 309-322 2. Badia, D., Defran, R. H., Orienting Responses and GSR Conditioning: A Dilemma, Psychol. Rev., 1970, 77, (3), PP. 171-181 i 3. Berlyne, D. E., Conflict, Arousal and Curiosity, Hew York: McGraw-Hill, i960 4. Brotsky, S. J., The Classical Conditioning of the GSR to Verbal Concepts, Unpublished Doctoral Dissertation, UCLA, 1964 5. Burke, C. J., Estes, W. K., A Component Model for Stimulus Variables in Discrimination Learning, Psychometrika, 1957, 22, pp. 133-145 6. Bush, R. R., Mosteller, R. A., A Model for Stimulus Generalization and Discrimination, Psychol. Rev., ! 1951, 58, pp. 413-423 j 7. Dodd, C., Lewis, M., The Magnitude of the Orienting | Response in Children as a Function of Changes in Color and Contour, J. Exp. Child Psychol., 1969, 8, pp. 296-305 8. Farber, I. B., The Things People Say to Themselves, Am. Psychol., 1963, 18, pp. 185-197 9. Gaines, R., Children’s Selective Attention to Stimuli: Stage or Set?, Child Development, 1970, 41, pp. 979- 991 110. Grastyan, B., The Significance of the Earliest Manifestations of Conditioning in the Mechanism of Learning, A. Fessard, R. W. Gerard and J. Konorski (Eds), Brain Mechanism and Learning, Springfield: C. C. Thomas, 1961 11. Grings, W. W., Carlin, S., Instrumental Modifications of Autonomic Behavior, Psychol. Record, 1966, 16, pp. 153-159 12. Grings, W. W. and Lockhart, R. A., Dameron, L. E., ! Conditioning Autonomic Responses of Mentally Subnormal Individuals, 1962, 76, (39 Whole No, 558) 13. Grings, W. W., and Shell, A. M., Magnitude of | Electrodermal Response to a Standard Stimulus as a i Function of Intensity and Proximity of a Prior I Stimulus, J. Comp. & Physio. Psychol., 1969, 67, (1), i . pp. 77-82 119_____ __________ _________ 120 14. Hernandez-Peon, R., Scherrer, H., Jouvet, H., Modification of Electric Activity in Cochlear Nucleus During "Attention1 1 in Unanesthetized Cats, Science, 1956, 123, pp. 331-332 15. House, B. J., Oddity Performance in Retardates, I., Acquisition and Transfer, 1964, 35, PP» 635-643(a) 16. House, B. J., Oddity Performance in Retardates, II.» Acquisition Discrimination Functions from Oddity and Verbal Methods, 19£>4, 35, pp. 645-651 (b) 17. House, B. J^, Effects of Discrimination Problems on Learning and Retention in Retardates, J. Exp. Child Psychol., 1968, 6, pp. 571-584 18. House, B. J., Orlando, R., and Zeaman, D., Role of Position and Negative Cues in the Discrimination Learning of Mental Defectives, Percept. Mot. Skills, 1957, 7, pp. 73-79 19. House, B. J., Zeaman, D., A Comparison of Discrimi nation Learning in Normal and Mentally Defective Children, Child Development, !958f 29, pp. 411-416 (a) 20. House, B. J., Zeaman, D., Position Discrimination and Reversals in Low-Grade Retardates, J. Comp. Physio. Psychol., 1959, 53, pp. 564-565 21. House, B. J., Zeaman, D., Reward and Non-Reward in the Discrimination Learning of Imbeciles, J. Comp. Physio. Psychol., 1958, 51, pp. 614-618 (b") 22. House, B. J., Zeaman, D., Transfer af a Discrimination from Objects to Patterns, J. Exp. Psychol., 1960, 59, pp. 298-302 (a) 23. House, B. J., Zeaman, D., Visual Discrimination learning and Intelligence in Defectives of Low Mental Ages, Amer. J. Ment. Def., 1960, 65, pp. 51*58 (b) 24. Irzhanskaya, K. N., Pelderbaum, R. A., Effects of Stimulus Sensitization on Ease of Condition, in Yvonne Brackbill and George G. Thompson (Eds), Behavior in Infancy and Early Childhood; A Book of Readings, New York; The Free Press, 1967, pp. 246-249 . 25. Jeffrey, W. E., The Orienting Reflex and Attention in Cognitive Development, Psychol. Rev., 1968, 75, (4), pp. 323-334 26. Kagan, J., Attention and Psychological Change in the Young Child, Science, 1970, 170, pp. 826-832 123 27. Kirk, R. E., Experimental Design: Procedures for the Behavioral Sciences. Belmont: Brooks/Cole, 1968 28. Lynn, R., Attention. Arousal and the Orientation Reaction, New Yorks Pergamon Press, 1966 29. Maltzman, I., Raskin, D. C., Effects of Individual Differences in the Orienting Reflex on Conditioning and Complex Processes, J. Exp. Res, in Pers., 1965, (1), pp. 1-16 30. Piaget, J., Inhelder, B., The Child's Conception of Space, Langdon and Lunzer (trans), London: Routledge and Kegan Paul, 1963 31. Raskin, D. C., Some Factors Influencing Semantic Conditioning and Generalization of Autonomic Responses, Unpublished Doctoral Dissertation, UCLA, 1963 32. Razran, G., The Observable Unconscious and the Inferable Conscious in Current Soviet Psychology: Interoceptive Conditioning, Semantic Conditioning and the Orienting Reflex, Psychol..Rev., 1961, 68, (2), pp. 81-147 33. Restle, F., A IDheory of Discrimination learning, Psychol., 1955, 62, pp. 11-20 34. Schnidman, S. R., Instrumental Conditioning of Orienting Responses Using Positive Reinforcement, J. Exp. Psychol., 1970, 83, (3), pp. 491-494 35. Shepp, B., Zeaman, D., Discrimination Learning of Size and Brightness by Retardates, J. Comp. Physio. Psychol., 1966, 61, pp. 55-59 36. Stewart, M. A., Stern, J. A., Y/inokur, G., Fredman, S., Degree of Conditioning of the GSR as a Function of the Period of Delay, Psychol. Rev., 1961, 68, pp. 60-67 37. Trabasso, T., Attention in Learning Theory and Research, 1966, New York: John Wiley & Sons, Inc. 38. Uno, T., Grings, V /. Y/., Autonomic Components of Orienting Behavior, Psychophysiology, 1965, 1 (4), pp. 311-321 39. Voronin, L. G., Sokolov, E. N., Cortical Mechanisms of the Orienting Reflex and its Relations to the Conditioned Relfex, H. II. Jasper, G. D. Smirnov (Eds), The Moscow Colloquium on Electroencephalo graphy of higher Nervous Systems, Montreal: EEG J ournal, 1960 40. Wolfensberger, W., O'Connor, N., Stimulus Intensity and Duration Effects on EEG and GSR Responses of Normals and Retardates, Am. J. Ment. Def., 1965, 70, pp. 21-57 41. Wyckoff, L. B., The Role of Observing Responses in Discrimination Learning, Psychol. Rev., 1952, 59, pp. 431-442 42. Zaporozhets, A. V., The Development of Perception in Preschool Children, Monographs of the Society for Research in Child Development, Paul Mussen (Ed), 1965, pp. 82-101 43. Zeaman, D., House, B. J., Discrimination Learning in Retardates, Train. Sch. Bull., 1959, 56, pp. 62-67 44. Zeaman, D., House, B. J., The Role of Attention in Retardate Discrimination Learning in N. R. Ellis (Ed)i Handbook of Ment. Def., New York; McGraw-Hill, 1963, pp. m - m

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A Cross-Cultural Study Of The Use Of Art Forms

PDF

The Conditioning Of A Discriminative Stimulus Measured As An Orienting Reaction In Profoundly Retarded Blind Children

PDF

A Study Exploring Two Different Approaches To Encounter-Groups: The Combination Of Verbal Encounter And Designed Nonverbal Activity Versus Emphasis Upon Verbal Activity Only

PDF

Determining Policies And Procedures In Opening New Elementary Schools

PDF

An Analysis Of The Philosophical Beliefs Implicit Within General Semantics And Their Relevance For Educational Theory

PDF

Delay Of Feedback And The Acquisition And Retention Of Verbal Material Inthe Classroom

PDF

Teaching As A Dramatic Performance: An Inquiry Into The Application Of The Stanislavski Method To The Classroom

PDF

A Comparative Analysis Of The Concepts Of Choice And Responsibility In The Educational Theories Of Theodore Brameld And Carl Rogers

PDF

An Evaluation Of Instruction By Student-Led Discussion In The College Classroom

PDF

A Study Of Philosophy Curriculum Dissonance In Selected Colleges And Universities In California

PDF

Capillary Pulse Pressure As A Means Of Discriminating Flaws And Involvement Between Cognitive And Affective Test Questions

PDF

A Semantic Differential Investigation Of Critical Factors Related To Achievement And Underachievement Of High School Students

PDF

An Investigation Of The Effect Of An Experimental Curriculum On The Knowledge And Attitudes Of High School Students Toward Consumer Education

PDF

Performance Of Retardates On Piagetian Tasks As A Function Of Ethnicity

PDF

Attitudes Toward Religion Of Saudi Arabian Students In The United States

PDF

Field-Independence And Typologies Of Delinquency

PDF

Signal Processing Time And Aging: Age Differences In Backward Dichoptic Masking

PDF

The Treatment Of Mexican-American History In Contemporary American High School Textbooks

PDF

An Individualized Approach: An Evaluation Of Cognitive And Affective Learning In Seventh-Grade And Eighth-Grade Mathematics Classes

PDF

A study of personality factors affecting cultural adjustment of American teachers overseas

Asset Metadata

Creator Streifel, John Arthur (author)

Core Title The Use Of The Orienting Reflex To Test The Zeaman And House Theory Of Attention

Degree Doctor of Philosophy

Degree Program Education

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Psychology, experimental

Format dissertations (aat)

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Meyers, Charles Edward (committee chair), Galbraith, Gary C. (committee member), O'Neill, William S. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-480413

Unique identifier UC11362219

Identifier 7217516.pdf (filename),usctheses-c18-480413 (legacy record id)

Legacy Identifier 7217516.pdf

Dmrecord 480413

Document Type Dissertation

Format dissertations (aat)

Rights Streifel, John Arthur

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA