mammw MB" Wfififlfifififi EH 5. DESCBEMINAflfi-H LEARNifiG TASK: l I A cow 15.31505; 0.? f mam m RETAEBATES } Thesis fez-9 {‘56 00mm 64? M. R MC in?“ STATE UEE'EVEESETY lante- Rafih Dickie “£9.70 LIBRARY Michigan State : _ . . j: Unlvcrsxcy ABSTRACT INHIBITION AND VERBALIZATION IN A DISCRIMINATION LEARNING TASK: A COMPARISON OF NORMALS AND RETARDATES BY Jane Rath Dickie Normals 5-6 years old, 9-12 year old normals, and, retardates with MA's from 5-6 and CA's of 9-12 were matched on WISC vocabulary scores and presented with a multidimen- _ sional discrimination task. Method of presentation (suc- cessive vs. simultaneous) and amount of time spent in view- ing the SA (maximum vs. minimum) were compared. All gs were taken to a trials—criterion of 10 trials correct in a row and a verbal criterion of correctly verbalizing the rule. The groups did not differ significantly in reaching the trials-criterion, but did differ significantly in reach- ing a verbal criterion. Almost without exception 9-l2 year old normals were able to verbalize the rule at the same time the trials-criterion was reached regardless of the Jane Rath Dickie method of presentation; the 5-6 year old normals were in- ferior to the 9-12 year olds in verbalizing the rule. A relationship between WISC vocabulary score (verbal ability) and ability to verbalize the rule at the same time the trials—criterion was reached was found. Also, for the younger normals, the ability to verbalize the rule was sub- stantially aided by the method of successive presentation. For the majority of retardates, ability to verbal— ize the rule was lacking, and for them, there was no rela- tionship between WISC vocabulary scores and verbalizing the rule. In addition, successive presentation did not help the retardate to verbalize the rule. The present study was interpreted as supporting White's (1965) proposition that normal children between the ages of 5 and 7 shift from a mode of responding (prob- lem solving) which is largely dependent on simple S—R asso- ciations to a secondary system based more on cognitive, de- cision making processes. In Luria's (1961) terms this would be a shift from the first to the second signal system, i.e., a shift to verbalization about one's own behavior. Inhibition of the primary mode of responding in favor of Jane Rath Dickie the second, language-based mode of responding is postulated to account for the present results. Concomitantly, these results support the notion of a basic inhibition deficit in the retardate (Denny, 1964), an inhibition deficit which prevents the emergence of the secondary, verbal mode of problem solving. INHIBITION AND VERBALIZATION IN A DISCRIMINATION LEARNING TASK: A COMPARISON OF NORMALS AND RETARDATES BY Jane Rath Dickie A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1970 ‘ To Larry, for his love and support through all my tirades, and for his encouragement to strive for distant goals. ii ACKN OWLE DGMENTS My sincere thanks to the members of my committee: To Dr. Hyman for the initial inspira— tion to critically evaluate the literature.... To Dr. Strommen for guidance in the literature and for an inspiring seminaire relating to this topic....And eSpecially to Dr. Denny for his con- cern and expertise in all matters relating to this work. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES. . . . . . . . . . . . . . . . . . vii INTRODUCTION . . . . . . . . . . . .'. . . . . . . 1 REVIEW OF THE LITERATURE . . . . 9 . . . . . . . . 5 Elicitation Theory . . . . . . . . . . . . . . 5 Inhibition . . . . . . . . . ; . . . . . . . . 10 Verbalization. . . . . . . . . . . . . . . . . 18 Integration of Inhibition and Verbalization. . 24 THE PRESENT STUDY. . . . . . . . . . . . . . . . . 34 Hypotheses . . . . . . . . . . . . . . . . . . 34 Procedure (Methodology). . . . . . . . . . . . 36 Subjects . . . . . . . . . . . . . . . . . 36 Apparatus. . . . . . . . . . . . . . . . . 37 Procedure. . . . . . . . . . . . . . . . . 37 Results. . . . . . . . . . . . . . . . . . . . 42 Ability to reach a trials-criterion. . . . 42 Ability to reach a verbal criterion. . . . 45 iv Table of Contents (cont.) Page Ability to verbalize within inhibition 7 1 conditions (method of presentation). . . 46 Nature of the incorrect verbalizations . . 51 WISC vocabulary scores and the ability to verbalize . . . . . . . . . . . . . . 53 DISCUSSION . . . . . . . . . . . . . . . . . . . q 56 IMPLICATIONS (for future research) . . . . . . . . 64 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . 65 LIST OF TABLES Table a Page 1. Results: Number of trials to reach a trials and a verbal criterion for all gs. . . . . 43 vi LIST OF FIGURES Figure Page 1. Normals and retardates: a comparison of number of trials to a trials-criterion and number of trials to a verbal cri- terion; each S is plotted under simul- taneous or successive presentation. . . . i 44 vii INTRODUCTION Elicitation theory(Denny, 1966; Denny and Adelman, 1955) emphasizes that learning occurs when a response is consistently elicited in contiguity with a stimulus. (This requires keeping competing responses to a minimum. From such an analysis of learning, it can be deduced that, to the extent competing responses intrude, learning will be retarded. Young children between 4-7 years of age and the mentally deficient seem particularly prone to competing re- sponse intrusion. In retardates especially, this has been termed an inhibition deficit (Denny, 1964). If the retard- ate or young child actually does have a greater difficulty inhibiting his responses to an SA, than older normals, then clearly his ability to learn a complex, multidimen- sional discrimination would be handicapped, especially if such a discrimination involves a process of eliminating responses to irrelevant cues as Zeaman and House (1963) have suggested. If an inhibition deficit exists in the retardate, then a situation in which competing responses were very likely to occur in the initial stages of learning would retard learning for the retardate more than for either a normal MA or CA control group. If inhibition as a process is developing in the normal child between the ages of 5-7 years, then learning should be slower for the younger MA control than for the CA control. Such findings have been indicated in several situations.' For example, when stimuli are presented simultaneously with position irrelevant, the very young subject and particularly the retardate is prone to respond perseveratively to position (Sidman and Stoddard, 1966; Spitz, 1966) or symmetry of response (Schusterman, 1964); the result in either case is slower learning for the young child and the retardate. Zeaman and House (1963) have discovered a hierarchy of "attended to" cue dimensions in discrimination learning by the retardate and the young child: 1) color-form, 2) form, 3) color. This suggests that when both color and form are relevant, the discrimina- tion is easiest because competing R's are minimal and that when form is relevant, there is less competition than when color is relevant. However, not all investigators (Zeaman and House, 1963) have found the retardate inferior in discrimination- problems when the criterion used has been the number of trials to a criterion of so many correct in a row (trials- criterion). In fact, retardates have even been found to. perform better than an MA control group (Klugh and Janssen, 1966). Even so, the retardate's performance does seem to break down when the ability to verbalize the rule (verbal criterion) is used. Stolurow and Pascal (1950) found low_ MA §s unable to verbalize the correct choice after a trials- criterion had been reached. Hermeline and O'Connor (1958) found that only 2 retardates out of 20 could verbalize an abstract concept, although the trials-criterion suggested. that the concept was known. An interesting finding by Klugh and Janssen (1966) is that even when retardates and normals were matched on vocabulary scores (WISC) the re-_ tardates' ability to verbalize the correct solution on a discrimination problem was significantly worse than_the normals. Not only has a significant difference in ability to verbalize between normals and retardates been found, as described above, but method of presentation seems to affect the retardates' ability to verbalize the correct solution on a multidimensional discrimination problem, with perform- ance on successive stimulus presentation being superior to simultaneous stimulus presentation. This suggests a pos- sible interaction between inhibition and verbalization, in which the retardate is particularly prone to intrusion of competing verbal R's. This interaction may be true to a lesser extent of a young normal population than of retard- ates of the same MA. This interpretation is further sup- ported by the observation that many retardates would verbal- ize an incorrect solution after a trials-criterion had been reached. Once false verbalization had occurred, the retard- ates would act on it making incorrect choices (Klugh and Janssen, 1966). After a trials-criterion was again reached, the retardate would again verbalize an incorrect solution. (This last observation has not been systematically investi- gated.) The aim of the present investigation was to deter— mine whether the ages of 5-7 are critical for the formation of an inhibition process (White, 1965) and if mentally retarded children with MA's of 5-7 and CA's of 10-14 are lacking in ability to inhibit incorrect responses as com- pared with MA and CA controls. The review of the literature deals with learning in retardates and normals (ages 5-7) within the elicita- tion framework, in an attempt to analyze just what the effects of inhibition and verbalization are and how they interact. Therefore, the theory, concepts of inhibition and verbalization, and an attempted integration of the concepts of inhibition and verbaliéation will be dealt with in that order. REVIEW OF THE LITERATURE Elicitation Theory The framework in simple terms is a continguity theory that focuses on the importance of incentives and the removal of incentives as consistent elicitors (UCS) of the to-be-learned response. "Learning depends on eli- citation of a particular class of responses in close temporal contiguity with the stimuli in question" (Denny, 1966, p. 2), So, learning depends on the relative elimin- ation of competing responses (inhibition) in the presence of the cue stimuli. Note that reinforcement is not used in the Thorndikian sense; rather it lets the organism "know" what the response to be learned is. Thus it can include instructions for an adult, candy for a child, or food pel- lets for a rat. Reinforcement (incentives), then, solve the problem of how to communicate with a nonverbal organ- ism by ensuring the elicitation of the required response. That is, in the organism's approach to the incentive, the correct response is automatically made. Imitation, prompt- ing, knowledge of results, and instructions are all capable of eliciting the to-be—learned response. The greater the response involvement of the organ- _ism in both intensity and duration, the better the condi- tioning. That is, a long lasting or intense response has a greater likelihood of being conditioned. Presumably in the verbal organism, verbalization is a long-lasting re- sponse, thus leading to stronger conditioning. In order for a response to be elicited consistently, competing responses must be inhibited. Within the theory, inhibition is regarded as the result of response competi- tion. Such competing responses are conditioned to a stimu- lus situation which is similar to the one to which the original response is conditioned. Changes in the stimulus situation weaken the strength of the original response ten— dency through generalization decrement and may also produce competing responses; thus stimulus change can be viewed as a source of inhibition. The amount of inhibition generated by competing responses is a direct function of their incom- patibility with the original response. No unhooking of re- sponses or direct weakening of responses is postulated. In part, at least, the theory defines stimulus in terms of a particular class of organisms. A stimulus is a particular object or event which has been shown in the life history of the organism to elicit a response reliably in that organism. A response class is defined as a set of response occurrences for a particular stimulus situation for a particular class of organisms (i.e. approach, avoid- ance, aggression, etc.). Elicitation is the relationship between the S-R which holds whenever a class of responses is contingent on an immediately antecedent class of stimuli. At any moment in time the stimuli impinging on an organism are arranged in a hierarchy. These stimuli main- tain particular positions in the hierarchy depending on such variables as satiation, intensity of stimulation in- cluding magnitude of reward, deprivation, set producing instructions, etc. For stimulus generalization to occur, it is import- ant that the effective stimulus elements are common to all situations under consideration. When few, if any, of the ~effective stimulus elements are present when changing from one stimulus situation to another, then transfer will be poor, or the generalization gradient steep. Discrimina- tion training is one means of ensuring that essential ele- pments of the stimulus situation become the discriminative stimuli. Another means of stimulus control is through ini- tial saliency for such elements. For example, in the con- servation tasks the similarity from one task to another is often small. It is no wonder that transfer does not occur until the child is capable of carrying a set of verbal abstractions from one situation to the next. Otherwise, each time the controling or salient perceptual elements of the situation may actively compete against the desired S-R connection, namely conservation. Many experiments show that a change in responding occurs between 5 and 9 for the normal child. For the re- tarded child many of these changes do not occur. In eli- citation terms, what makes the child between ages 5-7 be- gin to respond consistently to new stimulus elements? What makes new stimuli salient? What makes the child re- spond consistently? Basically what I would argue is that (1) speech is gaining a regulatory function, thus provid- ing a long-lasting response necessary for certain learning situations. (2) Inhibition of other responses, competing responses, is facilitated by verbalizations which actively compete with the old responses. (3) Inhibition as a pro- cess is beginning, at this age to develop strongly. (4) LFor the retardate, although verbalization may be develop— irug, somehow it does not seem to get integrated into the pxnoblem solving process. Nor does the ability to inhibit in correct responses appear. 10 Inhibition Inhibition generally occurs in the process of dis- crimination training. During a discrimination task, par- ticular stimulus elements gain control, while others lose their saliency. The stimulus gains control of the response m which occurs consistently in its presence. Terrace (1966) attempted to discover the nature of inhibition in the stimulus discrimination task. He hypo- L thesized that the nature of inhibition is responsible for behavioral contrast (increased rate of responding to the D . A . S after many presentations of the S ) and peak shift (higher rates of responding to stimuli farther away from A . D the S even though the peak of responding was not the S .) He compared discrimination training in which inhibition was a fundamental part--trial and error 1earning—-and dis- crimination training where inhibition played no part-- errorless learning. Errorless learning was achieved by . . A a fading process. Responding to the S never occurred and so never needed inhibiting. Terrace's comparison between errorless learning and trial and error learning revealed the following ll findings: 1) After criterion had been reached the error group which relied on inhibiting incorrect responses dis- A played bursts of responses to the S ; the errorless group did not. 2) Learning was slower for the error group since errors perseverated. 3) Once errors diminished gs were p observed making competing emotional responses whenever f A ' . the S was presented. 4) Behav1oral contrast and peak shift were only observed in the error (inhibiting) group. From all these differences between error and error- less learning, Terrace suggested that inhibition is inde- pendent of excitory states, and that inhibition involves aversive conditioning to the SA; in errorless learning the SA is neutral. Most of Terrace's data were from studies done with pigeons; however, others working with children have ob- tained similar results; for example Sidman and Stoddard (1966) in their work with the mentally retarded (MR) began by asking the question, "How does one communicate with a child who understands neither word nor gesture?" They pointed out that often verbal instructions were not under- stood by the MR or the young child. They made the same 12 point that Vygotsky (1962) made in speaking about pseudo- concepts; i.e., although the child may utter the same sounds, the words they speak may mean entirely different things. Sidman and Stoddard's whole procedure of fading Fa involved the assumption that errors on the part of the child reflect a lack of understanding of what is wanted by the adult. An added difficulty occurs because once errors occurred in the young child they tended to perse- verate; in other words, once an incorrect response occurred, the child could not inhibit continued errors. Several steps were taken in the Sidman and Stoddard procedure to guarantee the need tofiinhibit was minimized for these children. First, they adapted the child to the situation. This was to prevent competing emotional R's in the stimulus situation. Then they slowly focused the child's attention on the important or relevant stimulus elements in the discrimination task. In their final in- terpretation of the results, Sidman and Stoddard pointed out the comparability of reinforcement for a child and verbal instructions for an adult. Both are elicitors of the desired response. . 13 A study by Touchette (1969) sheds further light on the problem immature subjects have in inhibiting incorrect responses. In his study Touchette compared two groups of matched mentally retarded children (IQ=27-45) on a discrim- ination learning task. Group 1 was given 370 trials or $1 until they learned the problem using the standard trial and error method. Only one Of the subjects from this group learned the discrimination. Group 2 was given a WT fading procedure taking 70 trials. During this time errors I were completely eliminated or significantly reduced. All of the subjects in this group learned the discrimination. In the second phase of the study group 1, the trial and error group, was switched to the fading procedure. All subjects learned the discrimination which originally they were unable to learn in 370 trials. However, when group 1 and 2 were compared on a final criterion run of 40 trials, group 1 had many more bursts of errors than did group 2, the group which had never gone through the trial and error procedure. This finding is similar to time finding of Ter- race with pigeons, i.e., the error group frequently made bursts of responses to the SA even after the discrimination was well barned. This finding can be interpreted in terms 14 of pro-active inhibition. The trial and error group had lots of practice responding the wrong way (to the SA). Once these responses were made, inhibiting them was much more difficult. So, as elicitation theory implies, the degree to which the to-be-learned response is consistently elicited is important. A study by Zaporozhets (1969) points to a similar assessment of inhibition for young children. This study focused on exploratory behavior rather than discrimination learning. Zaporozhets argued that appropriate orienting responses must occur for learning to take place. In this case learning involved the development of efficient sensory exploratory behaviors in yOung children. Previous studies by Zaporozhets had shown that young children (3-4 years) grasped objects rather than exploring the contours haptic- ally and visually. He argued that the traditional Montes- sori--trial and error--method for training sensory explor— ation is not satiSfactory. He had the children out out contours Of shapes. This forced their attention on to the relevant part of the stimulus--or forced the response of observing contours. In this case the stronger explor- atory response, i.e., responding to texture or size but 15 not contour, was inhibited by forcing the to-be-learned orienting response to contours. Zaporozhets' analysis of why the Montessori method did not work as well as his own method was not based on the difference between errorless and trial and error 1earn- I? ing, but in light of the fading studies and Terrace's an- alysis of inhibition it seems to be reasonable. One essential aspect Of errorless learning is fo- i} cusing attention on the relevant dimension. How does one make certain aspects of the stimulus situation salient for the child? One important variable would seem to be inhib- iting responses to extraneous distracting stimuli. Zeaman and House (1963) have analyzed the discrimination learning task in terms of selective attention. They have pointed out that during discrimination learning there is an early attentional phase during which time the child learns to attend to the relevant dimension and inhibit his attending to the irrelevant dimensions. By plotting backward learn- ing curves for each individual subject, they discovered that slow and fast learners and low and high mental age (MA) subjects are not differentiated by the rate at which they learn once responding above chance begins. Rather, 16 these subjects differed in the amount of time spent in the attentional phase. A key to determination of the length of the atten- tional phase for any subject would be the ability to in- hibit attention to irrelevant dimensions and extraneous stimuli in the environment. This variable has been called distractibility. The level of distractibility must be re- lated to the ability to inhibit responses. ‘Elicitation theory would predict that distractions would have the greatest effect during the learning of a task, (i.e., when consistent responding is most necessary) and a much lesser effect after the important stimuli in the situation had control over the behavior (i.e. during the performance of the task.) It would seem that the need to inhibit is greatest during the learning phase or what Zeaman and House called the attentional phase. Two studies by Sen and Clarke (1968) related these‘ variables. They studied some factors influencing distrac- tibility in the mentally retarded. Three findings from the first study relate to inhibition. 1) They found that low MA subjects (less than 7.5 years) were more easily distracted than older MA subjects (more than 7.5 years). 17 This finding suggests that younger subjects would have greater difficulty inhibiting responses to extraneous stimuli. 2) Higher MA subjects were only distracted by auditory stimuli which were similar to the relevant stim— uli. Lower MA subjects were distracted by both similar and dissimilar stimuli. This finding also makes sense in terms of ability to inhibit responding to stimuli which i are similar to those which are to elicit responding. 3) up -. -4“ . Duration of the task was also relevant in terms of how dis- tractible all subjects were. Presumably, fatigue would weaken one's ability to inhibit responding to distracting stimuli. In their second study, Sen°and Clarke showed that distractors had the greatest deleterious effect during the learning of a task rather than during the performance of the task after the learning had taken place. This follows from the prediction of elicitation theory and relates to the necessity of consistent responding during learning. So far, the discussion has outlined some of the be- havioral aspects of inhibition by comparing learning during which inhibition of a response was an integral part (trial 18 and error) and learning which was not dependent upon inhi- bition of a response (errorless learning). We have also seen some factors which influence the ability to inhibit (as in the distractibility study); these include mental age, fatigue, and how well the response in question has been learned. Later we will return to inhibition in an effort to explain some of the changes which occur during the transitional stage of 5-7 in normals, and whether such changes occur in the retardate. But first a discussion of studies on verbalization is appropriate. Verbalization Verbalization can be studied as either the indepen- dent or dependent variable. That is, one can look at the effect verbalization has on behavior, or on the effect be- havior, or other variables, have on verbalization. Most of the following discussion has to do with the effect ver- balization has on the behavior of the young child. ~That is, what is the regulatory function, if any, of verbaliza- tion. Some of the following studies deal with normal chil— dren, others with mental retardates. A“IF‘._II' 19 Denny (1964) stated that the retarded lack self— initiating sets which function to maintain continuous re- sponding. Luria (1961) pointed out the same sort of thing for young normals when he spoke about the development of self-regulation of the second-signal system (verbalization). Reese (1962) indicated that young subjects verbalize about a situation but their verbalizations are not always congru- ent with their non-verbal behavior. He termed this a media- tional deficiency. Kendler and Kendler (1962) were refer- ring to the same sort of discrepancy between verbal and other overt behaviOr when they spoke about vertical pro- cess--or several S—R levels occurring simultaneously. ". . . it was not uncommon for children to verbalize spon- taneously the correct solution while simultaneously making an incorrect choice. A few children did this for many con- secutive trials." (p. 9). All of these authors wonder what the role of verbalization is in the problem solving behavior of the young or low mental age child. This is a valid inquiry because many studies have shown that a young child's verbalization is not always congruent with his other behavior. Many of these studies have used discrimin- ation learning and reversal tasks. 20 Tracy Kendler (1964), while studying an optional reversal shift with kindergarden children, found that prompting S to verbalize helped both in the initial learn- ing task and in the reversal, even though the verbaliza- tions about the reversal were often incorrect. In her first experiment she compared a verbalizing group (those children who were instructed to say "white wins, black loses") with a control group (said nothing), on a discrim- !r (11.! ‘ ination learning task. Following the learning of task 1, all subjects were shifted to an optional reversal. They could either choose on the basis of color (previously the relevant dimension) or on the basis of form (previously irrelevant). That is, they could either make an intradi- mensional shift or an extradimensional shift. Following the learning of task 2, all subjects were given a test to discover which shift they had made. The paradigm is illu- strated below: Task 1 Task 2 Test + - + - + + 00 'IO ’05] C] II "white wins, intradimensional test to discover black loses." shift to black, which shift extradimensional was made. shift to square. 21 She found as expected that verbalizing facilitated the learning of task 1, and that more of the verbalizing group made the intradimensional shift. However, almost half of the verbalizing group continued to verbalize the original phrase, "white wins, black loses" on task 2, even though it was no longer correct, and their other overt behavior was correct. In her second experiment, the same procedure was followed, except that another group was added. Group V and VP were instructed to verbalize on task 1, however; only group VP was prompted to verbalize correctly on task 2. Group C was a control group, i.e., was not instructed to verbalize. She found that groups V and VP both learns faster than group C when shape, but not color, was rele- vant. (They all learned color very quickly, i.e. black and white was salient for this age group.) Verbalization did not affect the number of trials to learn the optional shift. Furthermore, there was no evidence that requiring S3 to verbalize correctly (VP) facilitated or interfered with choice learning as compared with those who continued to verbalize incorrectly. V and VP both had a greater proportion of intradimensional reversals than did group C. 22, And finally, the results from experiment 1 were replicated for group V. Over half (62%) continued to verbalize the incOrrect sentence during task 2, although their choice behavior was correct. Verbalizing an incorrect choice did not interfere with choosing correctly. Others who have obtained similar results have theorized about the discrepancy between the verbal and other overt behavior of the young child. Keeney, Cannizo, and Flavell (1967) after studying spontaneous verbaliza- tions during a recall task concluded that " . . . there could be a longer or shorter period in a subject when the component means (verbal and nonverbal) are well developed in themselves and yet tend not to be spontaneously re- cruited and integrated into the problem solving sequence." (p. 965) Piaget (1932) pointed out that before age 8 a child may say, "I'm not going to do that," and then proceed to do exactly what he said he wouldn't. Luria, in speaking about the regulatory function of speech pointed out that this regulation is slow to develop. Although the child can verbalize, his verbalizations do not regulate the behavior, even when the verbalizations of other people which are directed at the child do. 23 All of these results and speculations are based on a developmental factor. At some point the child inte- grates verbal and non-verbal behavior. A study by Milgram and Furth (1967) clarifies the development of the regula- tory function of speech which occurs between the ages of 5 and 9. They compared normal children, ages 6 and 9, to retarded children with mental ages of 6 and 9 on a concept development task. Three conditions for attainment ‘5‘. V.‘ m 1 of the part-whole concept were compared: E—-explanation, J--justification of choice (trial and error), and V-- verbal sentence frame, i.e., "A is part of B." Milgram and Furth found that normals showed a developmental in- crease in performance under the verbal conditions whereas the retardates did not. That is, at 6 years of age the} normals utilized the verbal aid much more than did the retardates of the same mental age, and the normals at 9 utilized the verbal aid more than any of the other groups. This study highlights 2 basic findings: that the regula- tory function of verbalization developes between 6 and 9 (possibly 5 is a better lower limit than age 6) and that retardates generally have a verbalization deficit; i.e., they do not use external verbal cues as well as do normals. 24 Other investigators have found similar results when comparing a trials—criterion with a verbal criterion in a discrimination learning task. That is, non—verbal behavior suggested that the low MA child knew the concept but he was unable to verbalize the rule (Stolurow and Pas- '71 cal, 1950; Hermeline and O'Connor, 1958; Klugh and Janssen, r 1966). The preceding discussion emphasizes 2 important points: 1) In problem solving behavior there are two 3‘ levels of operation--verbal and non-verbal behavior--and these are not necessarily congruent in the young or low MA child. 2) At some point in development these two modes become integrated and this may involve an inhibition pro- cess. This brings us to an attempt in theory to integrate verbal behavior and the ability to inhibit. Integration of Inhibition and Verbalization Two basic approaches to the interaction between verbalization and inhibition have been made. Luria (1961) Vygotsky (1962) and Bruner (1964) emphasize the effect of verbalization on inhibition. Denny (1964) and White (1965) 25 emphasize the effect of inhibition on verbalization. That is, one group holds that verbalization is more basic, the other that the ability to inhibit is more basic. Luria related verbalization and inhibition via Pavlov's second-signal system (verbalization). They are related in that the second-signal system serves to gener- alize, abstract, and systemize the stimuli which man reacts i to. The second-signal system further serves to inhibit Ill \ - A... 'l the organisms direct impulse reaction. So, verbal rules prevent impulse responses. The second—signal system can reorient a stimulus completely from an excititory to an inhibitory stimulus just by inserting "do not." However, in young children, the inhibitory function of verbaliza- tions, particularly their own verbalizations, is not com- plete. Excititory instructions can be followed quite well. But self-initiated inhibitory responses are not functioning. In Luria's (1961) classic technique, children were given a bulb, and instructed to squeeze it when a green (or red) light came on. This they did very well as long as they were older than about 2. (At this age they would often squeeze when the light was off.) However, when the children were instructed to squeeze for the green light 26 and not for the red light, they could not inhibit responses to the red light. They were told to say "stop, stop," when the red light came on. This made inhibiting even more difficult. The child while saying "stop, stop," would squeeze, squeeze! Luria interpreted this to mean that the young child's (age up to 5) behavior is condi- tioned to the impulse functiOn of speech, rather than to the meaning of the speech. Self-regulatory behavior or inhibition in Luria's framework is finally complete when speech is internalized, in the Vygotsky sense of internalized speech. This occurs -between 5-1/2 and 7 years. According to Luria, the regula- tory function is steadily transferred from the impulse side of speech to a system of associations based on the meaning of speech. During this time period speech shifts from external speech to internal speech. After this occurs the child can inhibit by means of his own verbalizations. So the meaning or significance of what is said regulates the behavior rather than the impulse creating diffuse ex- citation. Luria and Vygotsky emphasized the effect of verbal- ization on inhibition, and Bruner followed this reasoning 27 in tracing the cognitive growth of the child. Bruner traced 3 stages of cognitive growth: Enactive (like Piaget's sensori-motor), iconic (based on images of the environment), and finally symbolic. When the child en- counters information in the environment that conflicts with past information, he moves from one stage to the next to handle the inconsistency.' In the first and most of the second stages the child is tied to the happenings in the immediate environment. However, from the redun- dancy he acquires a model of the environment which per- mits him to go beyond the immediate information before him. Symbolic coding (language) is necessary for this -step. "To transcend the immediate perceptual, to get beyond what is vividly present to a more extended model of the environment, the child needs a system that permits him to deal with the nonpresent. . ." (p. 13, Bruner, 1964).A Implicit in this change from enactive to iconic and especially from iconic to symbolic, is the ability of the child to inhibit old modes of responding and implement the new method of responding to the consistencies in the environment. For example in a study by Frank (1966) young children, ages 4 and 5, were given the classic jar problems 28 for water conservation. However, in this technique the jars were partially concealed from the children until they had said that the water in both jars was the same. Then the sheet concealing the jars was removed. Following this they saw the whole jar during the transfer of water from one to the other; now, however, the five year olds were able to conserve, even in the face of the salient percep- tual cues. In this "say before see" method the child did not initially have to inhibit attending to the salient perceptual cues because they were hidden. Once the ver- bal response had been made, inhibiting attention to the visual cues was easier. Internal, verbal cues became the clue to the constancy of the situation. As shown, inhibi- tion is important within the Bruner framework too. Denny, with some good evidence, stated that prob- ably the inhibition deficit in the young is more basic than the ability to verbalize. Using data from classical conditioning studies Denny (1964) pointed out that the retarded (and this could apply to young children as well) have an inhibition deficit. This was manifest by increased resistance to extinction in classical conditioning. That is, these SS did not inhibit the conditioned response 29 once it was learned (Krasnogorski, 1913; Mateer, 1918; Razran, 1933; Luria, 1959). Also conditioning inhibition was very difficult, and once conditioned, disinhibition (or continued responding) occurred readily. However, these low MA children did condition an excititory or appetitive response rapidly. Marinesco and Kreindler (1933) have re- lated the inhibition deficit to poor attention or distractibility. This same hypothesis has prevailed from 1904 (Kuhlmann) until today (Zeaman, 1959) as a way to account for the retardate's difficulty in discrimination- learning (specifically, to explain the re- _ tardate's ignoring of relevant stimuli.) But the question which must be asked is this: Is the inability to attend caused by the in- ability to inhibit? May not the retardate's failure to inhibit the effects of extraneous stimuli account for his inability to attend to the particular task? Perhaps, the inhibi— tion deficit is the more basic (Berger, 1954). Luria, for one, lumps together distractibil- ity, impulsiveness, and inhibitory defect in his discussions. (Denny 1964, p. 4) -Denny went on to relate inhibition to verbalization. That the mentally retarded may condition poorly to verbal stimuli and may fail to inhibit because of lack of verbal control appears to strike at the heart of the men- tal retardation problem. Yet we are only groping in the dark until more research is done. Also, it would be unwise to imply that the inhibition deficit is necessarily a function of poor verbal control, for in 30 lower animals the juvenile organism exhibits an inhibition deficit as compared with the mature adult (Thompson, 1941; Panchenkova, 1956; Vince, 1959). The deficit is reflec- ted in slow extinction, poor differential learning, and perseverative behavior. The parallel with the mentally retarded would seem to be more than coincidental. Moreover, natal or prenatal anoxia in rats results most ya clearly in impairment in the inhibition ' sphere. (p. 6) Denny introduced comparative literature on rats and monkeys in which retardation had been experimentally I. '- imposed by anoxia at birth or by frontal leisions. Re- versal learning of a simple position response was impeded. Stereotypy of a position response in a discrimination task followed. Both of these findings reveal an inhibition deficit in that position is an early habit which must be inhibited for learning to occur, and inability to reverse in a reversal shift shows an inability to inhibit the first learned R. Pascal et a1. (1951) have found a relationship between MA and ability to delay a response. Delaying a response is, in effect, inhibiting that response. (Pascal and Stolurow (1952) found that this effect interacted with the relevant cue selected. That is, young children were more likely to succeed if form was relevant. 31 Although Denny was directing his thoughts toward an inhibition deficit in the mentally retarded many of his findings are applicable to young normals as well. White's (1965) theorizing on inhibition followed the line of Denny's. Although verbalizing may (and prob- ably does) interact with inhibition, the inhibition defi— cit seems more basic to the understanding of a change in behavior from ages 5-7 in the normal. He pointed out that inhibition had its first sizable influence during this age range and suggested that there are two levels of functioning: that which is based on more primitive S-R connections and a second level which is dependent upon in- -hibition of early modes of responding in favor of a more cognitive or decision based type of responding. He stated: Perhaps the 5-7 period is a time when some maturational development, combining perhaps with influences in the modal environment, inhibits a broad spectrum of first level functioning in favor of a new, higher level of function (White, 1965, p. 213). A final decision as to which aspect (inhibition or verbalization) in problem solving behavior is more basic and the exact nature of the change which occurs awaits direct research. 32 IA study by Klugh and Janssen (1966) provides one opportunity for elucidating the complex relationship be- tween age (CA and MA), ability to inhibit both non-verbal and verbal responses and ability to verbalize the correct responses. They presented adult institutionalized retard- ates (mean CA : 44 years, mean MA = 6-4) and a normal MA control (matched also on WISC vocabulary scores) with a discrimination learning rask. Method of presentation was varied. That is, simultaneous (2 stimuli at a time) and successive (one stimulus at a time) methods of presenta- tion were compared on a trials-criterion and on ability to verbalize the correct cue (which was always square). On the trials-criterion, the retardates did slightly better than the normals. The retardate's performance did not differ for the 2 methods of presentation; however, the normals did much better on the successive presentation than the simultaneous presentation. Of those able to reach criterion verbalizing the correct one was much easier for the normals than for the retardates. The normals were able to verbalize equally well under both conditions. However, the retardates verbalized best under the successive presentation with 67% able to 33 verbalize as compared with only 14% able to verbalize under the simultaneous presentation. The major finding then, was that successive presentation was simpler for both normals and retardates with MA's = 6.2. For the normals the effect was on the trials-criterion and for the retardates on a verbal-criterion. Other authors (Erickson and Lippsitt, 1960; Hor- witz and Armentrout, 1965; and Lipsitt, 1961) have found that simultaneous was easier than successive. However, the subjects for all of these studies were 4th graders which would make them from 9-10 years old. Why this difference in ability to solve the dis— -crimination depending on method of presentation? Why an interaction between method of presentation and ability to verbalize? A possible explanation lies in the develop- ing, but incomplete ability to inhibit responding to the SA, and in an interaction between verbalization and inhi- bition. In the simultaneous condition gs must inhibit responding verbally to the cues of the incorrect stimulus. In the successive condition there are fewer distracting or irrelevant cues to interfere with correct responding. THE PRESENT STUDY Hypotheses The present study was designed to answer several questions arising from the theories and research described above. What do Ss verbaliZe? Do they verbalize the incor- rect R due to an inhibition deficit? If the learning prob- lem maximizes this deficit will retardates perform signif- iczntly worse than normals? The three main effects inves- tigated were: 1) minimizing vs. maximizing the need to ‘inhibit competing responses as reflected in performance on simultaneous or successive presentation, 2) verbal vs. trials criterion, and 3) the interaction between inhibi- tion and verbalization. The present hypotheses are (1) when the need to inhibit (competing responses) is minimal (successive pre- sentation) the retardate will perform better than when the need to inhibit is maximized. When inhibition is minimized the retardate and his normal MA control will perform the same. Normals will be superior to retardates 34 35 when the need to inhibit is maximized (simultaneous). However, the normal MA control should also show difficulty since the ability to inhibit responding is just developing from ages 5-7. Therefore, the normals should show a de- velopmental increase in ability to solve the discrimina- tion when inhibition is maximized. (2) When a trials- criterion is used to measure perfOrmance the retardate should perform as well as his normal MA control. (3) There will be an interaction between the retardates' and young normals' ability to reach a verbal criterion and the amount of inhibition necessary to perform a task. The need for increased inhibition will have a much greater effect on the retardate than the older normals (possibly a facilitory effect on normals) and inhibition will have a greater effect on a verbal criterion than on a trials— criterion. (4) The particular type of competing responses which interrupt correct verbalization will depend on the type of inhibition necessary. (In the conditions succes- sive and simultaneous inhibition minimized, competing R's from the irrelevant dimensions of the SD; in successive and simultaneous inhibition maximized competing R's from the cues of the_SA.) (5) For both retardates and young 36 normals the salient cues (i.e. form) should interfere most in both number of incorrect verbalizations and in which one dimension is the easiest. (6) There should be a cor- relation between verbal ability (WISC) and ability to verbalize the rule. Procedure (Methodology) ‘ Subjects.--A group of 36 normal children (from the Holt School System) with a mean CA of 5.89 (range 5.4 -6.5) and a mean WISC vocabulary score of 14.8 (range 5-23) were randomly assigned to 4 different inhibition conditions ‘and within each condition, three different "relevant" cue groups. Each S was matched on WISC vocabulary score to a group of 36 retardates (from the Jackson School for Trainable Retardates) with a mean CA of 12.82 (range = ‘8.4 - 17.4) and a mean MA of 6.31 (range = 4.9 - 8.5) and mean WISC vocabulary score of 13.28 (range = 7 - 22). The retarded group was correspondingly assigned to 4 inhibi- . tion and three "relevant" cue groups. Finally a normal CA control group of 36 SS (from the Holt School System) was assigned to the different conditions. An equal num- ber of males and females was in each group. 37 Apparatus.--A modified WGTA was used to present the stimuli to each subject. This device consisted of the normal WGTA with a small slotted box placed to the side into which SS placed the correct stimulus. Between the box and the display tray a vertical board prevented §s from viewing the remaining stimulus during the simultan- eous condition.. The SS were unable to View the stimuli once they had been placed into the box. The stimuli con- sisted of plastic cut-outs varying in size (2" or 4"), shape (square or round), and color (blue or red), so that there was a total of 8 different stimuli. Procedure.--Each S was given 2 pre-tests; the WISC vocabulary subtest and a test to make sure all of the con- cept instances could be independently verbalized, i.e., blue, red, circle, square, big, small. Any SS unable to verbalize the concept instances were eliminated immediately. Then each S was given the discrimination problem by one of 4 methods (inhibition groups), varying in type of presenta— tion and the degree to which viewing the incorrect stimu- ~lus (SA) was maximized. Viewing the SA was intended to maximize the need to inhibit incorrect responding. The 4 conditions were: 1) Successive presentation with 38 inhibition minimized, 2) Successive presentation with inhibition maximized, 3) Simultaneous presentation with inhibition minimized and 4) Simultaneous presentation with inhibition maximized. (These distinctions will be explained later in the paper.) The gs were then given the following instructions: "I want to see how quickly you can solve a puzzle. This puzzle has to do with little pieces like these." (Two of the stimuli were presented differing in shape, size, and color, these were counterbalanced within groups.) "You see these pieces are not alike, in fact they are different in several ways. For instance, they are different sizes. Which one is bigger (smaller)? . . . They are different shapes. Which one is round (square)? . . . They are dif- ferent colors. Which one is blue (red)? . . . Now I am going to show you some more pieces like these. I'll put them out here one at a time" (if successive; two at a time if simultaneous). "Then I want you to tell me whether you think that piece is the kind that goes in this box."- (For simultaneous: "Then I want you to tell me which One you think goes in this box.") "I'll tell you if you're right or wrong.. Then you put only the right kind in the 39 box." Correction technique was used. "Let's see how quickly you can learn to tell me which kind goes in the box. Do you understand?" Before each discrimination E asked, "Is this the kind that goes in the box?" or "Which one is the kind that goes in the box?" the emphasis was placed on the Eifl§.0f object which was correct. The par- ticular dimension and cue for each group was counterbal- é anced and allowed for tests between dimensions. In the simultaneous presentations S was shown a 4 sequence of pairs always consisting of a square and a circle, one blue, the other red, one large, the other small. The eight possible pairs were presented randomly with the restriction that no pair was duplicated until all pairs had appeared and no pair was presented on con- secutive trials. The SS were given as much time as needed to respond, although an attempt was made to present a pair every 10 seconds. For the successive presentations the same restrictions held except that the stimuli were pre— sented every 5 seconds to equalize the number of stimuli per unit time with the simultaneous groups. Within the simultaneous presentation there were two conditions of inhibition. These were accomplished by 40 the following procedure: after S had selected the correct stimulus he placed it through the slotted top of the box, out of View. For the inhibition maximizing condition, S then requested S to push the display tray back to S. This U 0 O A procedure forced S to View the incorrect stimulus (S ) last on each trial. This procedure maximized viewing of A . . . . . . the S and was intended to maXimize the need to inhibit responding to this stimulus. For the inhibition minimized condition, S selected the correct stimulus and placed it . . . . A in the box while S retracted the display tray With the S . Because of the positioning of the box,,Ss were unable to . A . . D , . View the 8 while plaCing the S into the box. This con- . . . . . . . A . . . dition minimized VieWing the S and thus minimized the need to inhibit responding to this stimulus. Within the successive presentation the two condi- tions of inhibition differed only in the ninth and tenth trial of a ten trial correct criterion. For the inhibi- .. .. .. A .. tion maXimized condition, Vlerng the S was maXimized, i.e., following 8 consecutive correct trials, 2 trials 0 A I DID. on presenting the S were given. For the inhibition mini- . . . . . A . . . mized condition v1ew1ng the S was minimized, so that following 8 consecutive correct trials the last two trials 41 in the ten trial criterion consisted of presenting the SD. (This procedural difference was expected only to have an effect on the ability to verbalize the correct cue.) The trials-correct criterion was 10 consecutive correct responses. At this time S asked S, "Tell me, which is the ggly kind that goes in the box." If S was correct the session stopped. If S was incorrect, the response was noted and S continued to trials-correct cri- terion again. This procedure was repeated up to 96 trials; Ss unable to reach a trials-criterion were continued to 96 trials and asked which kind always went into the box. A record of all verbalizations made by S during the session was kept. RESULTS The results are given in Table 1. As can be seen from the table and Figure l, most.subjects either solved the discrimination within 18 trials or else were unable to solve the problem by 96 trials. Since the minimum num- ber of trials to reach criterion was 10 and the maximum allowed was 96, there was a considerable floor and ceil- ing effect. Because of this, the originally planned analy- sis of variance for the data could not be performed. (The assumption of homoscedastisity was not even approximated.) ‘However, the data lent themselves well to a Chi square analysis. Ability to reach a trials criterion.--As predicted, when the criterion was 10 trials correct in a row, there was no difference between retardates and MA and CA con- trols. This result is shown below. Actually, all groups did so well that an analysis based on the trials—criterion could reveal very little. 42 43 Table 1. Results: Number of trials to reach a trials and a verbal criterion for all Ss. RELEVANT CONDITIONS DIMENSIONS Simultaneous Simultaneous Successive Successive Maximized Maximized Minimized Minimized trials verbal trials verbal trials verbal trials verbal RETARDATES . 87 87 63 96 26 96 77 91 form 39 96 12 12 15 81 18 96 12 12 10 10 12 12 15 15 47 96 34 34 37 37 14 14 color 10 10 96 96 96 96 13 13 13 96 33 33 96 96 17 58 14 96 11 96 18 38 36 96 size 14 96 11 96 15 96 19 96 10 96 12 96 51 96 10 10 -NORMALS 5-6 11 ll 54 96' 10 10 ll 11 form 11 11 10 10 15 15 10 10 12 12 12 12 ‘ 12 12 12 12 96 96 16 16 13 13 16 16 color 96 96 21 21 63 96 30 30 96 96 44 96 18 18 10 10 12 12 14 96 g l4 l4 l4 14 size 11 11 12 96 12 96 16 16 18 96 ll 11 18 18 10 10 NORMALS 9-12 ~ 10 10 11 26 14 14 l4 14 form 12 12 12 12 ll 11 14 14 12 12 ll 11 10 10 10 10 35 35 96 96 13 13 12 12 color 13 13 13 13 12 12 l4 14 13 13 12 12 12 12 16 16 12 12 12 12 12 12 ll 21 size lI ll 10 10 15 15 15 15 16 16 ll 11 14 14 12 12 100 ‘ ELVIS (5—6) & g 90 ‘ ' = successive ":1 o = simultaneous 3 80 - .H 3.. :1: 7o - 3 +33 60 - - o a; o -50 ‘ o .p . m o .3 1+0 . ’2 4’ 30 _ ‘H O o s. g 20 ‘ fig." 0 a a 2: 1O .. 6:93” g 1 L 1 1 1 1 1 l 1 l 10 20 30 1+0 50 60 70 80 90 100 100.” Number of trials to a-verbal criterion Retardates 9. 8 90- .5 ' = successive ° .3 80" 0 = simultaneous ‘1 o 2’. 7°- 31 k 60- o .p U o .9 5°‘ o ‘3 uo~ o O E oo' '2: ’°‘ . g 20. . 5 r... ' . 2 1m 3* ' ' 1 1 1 1 I 1 J l 1 4 1O 0 0 1H) 2) 60 0 80 0 100 Numger 013 trials 1.05 a verb . criterign Figure 1. Nomals and retardates: a comparison of number of trials to a trials-criterion and number of trials to a verbal criterion; each S is plotted under simultaneous or successive presentation. 45 Retardates Normals (5-6) Normals (9-12) pass 33 33 35 fail 3 3 1 Ability to reach a verbal criterion.--Comparing groups on a criterion of being able to verbalize the rule (verbal criterion) we see Significant differences between all groups, (X2 = 22.43, p < .01). The results are shown below.. Also, the retardates differed in this regard from 5-6 year old normals: 72% of the 5—6 group passed as com- pared with 47% of the retarded group (X2 = 3.7, p < .05, .one-tailed). Also, the 5-6 year old normals differed from Retardates Normals (5—6) Normals (9—12) pass 17 26 36 fail 19 10 ’ l the 9-12 year olds: 72% of the 5-6 group passed as com- . 2 pared Wlth 97% of the 9-12 year olds (X = 6.85, p < .01). The results supported the anticipated interaction between population (normals vs. retardates) and ability to 46 verbalize the correct response at the same time the trials- correct criterion was reached. Figure l for the 5-6 norm- als shows the ability to verbalize after reaching the trials correct criterion. All Ss on the diagonal reached the trials and verbal criterion at the same time. All Ss off the diagonal reached a verbal criterion later than the trials criterion or never reached a verbal criterion at all (although the trials criterion had been met). This relatiOnship is shown below. Retardates Normals (5-6) solving both together 16 29 solving at different times 20 7 X = 8.5 p < .01 or not at all Ability to verbalize within inhibition conditions (method of presentation).--A comparison between groups on the inhibition conditions with respect to the verbal cri- terion follows. It was predicted that difficulty in solv- ing the discrimination increases as the need to inhibit incorrect responses is maximized (being able to see the SA). It was predicted that the performance under these conditions orders itself from high to low, in the following 47 way: Simultaneous maximized, Simultaneous minimized, Suc- cessive maximized, Successive minimized. Although the differences were not significant over-all, the number of failures on the verbal criterion within each group (condi- tion) suggested the predicted order of difficulty. The results are shown below (the numbers in the table indi- cate the number of failures within each group). Simult. Simult. Success. Success Max. Min. Max. Min. Retardates 6 5 . 5 3 Normals (5-6) 4 4 2 0 Normals (9-12) 0 1 0 0 Total * 10 ' 10 7 3 The difference between the methods of presentation (successive vs. simultaneous) seemed to produce more of an effect than maximum-minimum treatment. As predicted all of the 9-12 year.old normals verbalized the rule, with one exception. The following analysis deals only with the 5-6 year old normals and the retardates. For both groups considered together, the differ— ence between simultaneous and successive presentation was 48 . . . 2 . Significant at the .05 level (X = 3.7, one-tailed). Suc- cessive presentation was superiOr to simultaneous for these SS; more of those in successive presentation verbal- ized successfully. The results are shown below. Simultaneous Successive pass 17 26 fail 19 10 X2 = 3.7 p < .05 (one-tailed) Considering the retardates alone, we see that the difference between methods of presentation was not signif- icant, though it was in the right direction. The results are Shown below. Simultaneous Successive pass 7 10 fail ' 11 ' 8 X2 not Sig. For the young normal children, the difference be- tween methods of presentation was significant, again with successive presentation superior to simultaneous. The re- sults are shown below. 49 Simultaneous Successive pass 10 16 fail 8 2 X = 3.46 p < .05 (one-tailed) Several comparisons between groups (normals (5-6), normals (9-12), and retardates) were made to determine whether differences in ability to solve the problem de- pended upon method of presentation. As predicted, this was the case for the normals, i.e., when the need for. Sglgeinitiated inhibition was minimized (successive pre- sentation) the young normals showed no particular deficit in ability to solve the discrimination; however, when the need to inhibit was maximized (simultaneous presentation) they did show a deficit. The comparisons are shown below. For the successive presentation: Normals (5-6) Normals (9-12) pass . 16 18 2 . fail 2 0 X not Sig. 50 For the simultaneous presentationz' Normals (5-6) Normals (9—12) pass 10 17 fail 8 l x = 5.33 p < .02 It was predictedthat retardates would perform approximately the same as the 5-6 year olds on the verbal criterion when the need to inhibit was minimized (succes— sive presentation), and would not perform as well as the normals when the need to inhibit was maximized (simultan- ious presentation). However, the results showed that -whereas the young normals were better in the successive presentation, the retardates were about equally poor in both conditions. These results are summarized below. For the successive presentation: Normals (5-6) Retardates pass ' 16 10 fail 2 8 x = 3.46 p < .10 51 For the simultaneous presentationz. Normals (5-6) Retardates pass 10 7 fail 8 11 X2 not sig. Nature of the incorrect verbalizations.--The na- ture of the incorrect verbalizations was investigated. Again as predicted, significantly more retardates made incorrect verbalizations than did their MA control group (normals 5-6). The results are shown below. Normals (5-6) Retardates Number of SS making 'incorrect verbalizations 9 - 23 . 2 Number of SS making ‘ 27 -13 X = 9.5 p < .01 correct verbalizations The young normal gave more incorrect verbaliza- tions than did the 9-12 year olds (9 vs. 3); however, the difference was not significant. The hypothesis that the source of incorrect ver- balizations would depend upon the last stimulus seen was not substantiated. However, the hypothesis that if certain salient dimensions were relevant, the discrimination prob- lem would be more easily solved than if the salient dimen- sions were irrelevant was substantiated for both normals (5-6) and retardates considered together. combining both groups are shown below. form color size pass 19 8 10 5 16 .14 fail The results Looking at the retardates and young normals separ- ately, we note the following minor differences: For the normals: form color size pass 11 4 8 fail l 8 4 For the retardates: form color size pass 8 4 2 fail 4 8 10 x =6.55p< .05 X = 4.4, p < .2, not sig. _53 For both groups, form was the easiest dimension. However, groups differed in performance on color and size. For the normals size followed form in difficulty and color was the hardest. For the retardates, form was followed by color and size was the most difficult (the differences among the three dimensions were not significant for the retardates). WISC vocabulary scores and the ability to verbalize.-- Several comparisons were made to determine if there were a relationship between CA, MA, or WISC vocabulary scores and the ability to verbalize the correct solution within groups. Only one of these relationships reached the .05 level of significance; however, the trends present are worth review- ing. The relationship between WISC vocabulary scores and ability to verbalize the correct solution was signifi- cant for the young normals. However, for the retardates there was no relationship at all. These results are shown 2 . . ‘ below. The X is based on a split at the mean WISC score. 54 For the normals: WISC < 14.8 WISC > 14.8 pass 8 18 . 2 fail 7 3 X = 3.00 p < .05 (one-tailed) For the retardates: WISC < 13.28 WISC > 13.28 pass 8 9 . - 2 . fail 10 9 X = not Sig. Further evidence for the lack of relationship be- tween verbal ability and ability to verbalize in the re- tardates is revealed by the following MA comparison. MA i 6-4 MA > 6-4 pass 10 6 . 2 . fail . 11 9 X = not Sig. The relationship of CA with verbalizing the cor— rect criterion was not significant for either the young normals or the retardates. Although in the case of the . normals this may be due to the very restricted range of ages. The results are shown below. For the normals: CA 5 5—10 CA > 5-10 pass 13 13 . 2 . - fail 8 2 X = 1.51,-not Sig. _ For the retardates: : CA > 11 pass 3 l4 . 2 . fail 8 11 X = 1.51, not Sig. DISCUSSION White (1965) has marshalled considerable evidence in support of the proposition that normal children from 5-7 shift from a mode of responding (problem solving) which is largely dependent on S-R associations to a secon- dary system based much more on cognitive, decision making processes. Luria (1961) refered to this shift in terms of a change from the primary to the secondary signal sys- tem. In both of these theories, inhibition of the primary ’ mode of responding in favor of the second, language based ‘mode of responding is implicit. The present data support these ideas and the basic notion of an inhibition deficit in the retardate (Denny, 1964). Definite differences among retardates, 5-6 year old normals, and 9-12 year old normals appeared in their .ability to apply verbal behavior to a problem solving task. Since retardates and normals (5-6) were matched on WISC vocabulary scores and all SS had passed the pretest re- quiring independent verbalization of the concept instances, 56 57 the obtained differences suggest what Kendler and Kendler (1962) and Reese (1962) have called a mediational deficit. That is, the retardate's and the 5-6 year old's inability to verbalize the correct solution was not due to the fact that the words were not in the child's repertoire, but that the child was unable to incorporate verbalization into the problem solving task. The fact that retardates, normals (5-6), and normals (9-12) were comparable on the trialsécriterion, suggests a dichotomy between these 2 modes of responding (i.e. verbal and non-verbal problem solving). Further support for this comes from the fact that only 2 of all the children verbalizing incorrectly followed their incorrect verbalizations; all the others continued to make correct choices while verbalizing in— correctly. That is, their non-verbal behavior and their verbal behavior were not integrated. However, inability to verbalize the correct solu- tion or inability to incorporate verbalization into prob- lem solving behavior is only part of the story. The fact that method of presentation also affected ability to ver— balize suggests that ability to inhibit incorrect verbal responses may interact with ability to incorporate _.. L91:- fifiufl' lr-er— 58 verbalization. This follows from the differences between successive and simultaneous presentation, which will be discussed later. The 9-12 year old normals almost without exception were able to verbalize the correct solution at the same time the trials correct criterion had been reached. Oper- ationally, this can be defined as incorporating verbal be- havior into the problem solving task. -For the 5-6 year olds, a relationship existed be- tween verbal ability (WISC) and ability to apply verbal behavior to the problem solving task (X2 = 3.00, p < .05). That is, SS with greater verbal ability were more likely to verbalize the correct solution at the same time they reached the trials correct criterion; This relationship was further clarified by a difference in ability to ver— balize the correct solution within the successive or simul— taneous presentation. The 5-6 year olds performed as well as the 9—12 year olds when presentation was successive but much worse when presentation was simultaneous. That is, they were able to incorporate verbal behavior when pre- sentation was successive but much less likely to do so when the presentation was simultaneous. 59 A look at the differences between methods of pre- sentation makes this difference easier to interpret. In the successive presentation SS had to respond with yes or no to one stimulus at a time, providing SS with 2 differ- ent conditions for verbalizing correctly. First of all, fewer irrelevant cues were present on any one trial to distract Ss as compared with the Simultaneous presenta- tion. Secondly, the successive presentation specified the negative information thus making it easier to inhibit verbal R's to these cues. In the simultaneous presenta- tion the emphasis was entirely on the correct stimulus. The negative information was more of an interference than -an aid. That is, specifying the incorrect cues explicitly brought them under the verbal contrOl of these young Ss, thus making it easier to inhibit responding to them. How- ever, no such conditions existed in the Simultaneous pre- sentation, where emphasis was on the correct stimulus only. For the retardate the method of responding differed considerably. Although the retardates demonstrated equal verbal ability (WISC and pretest) as did the 5-6 year old normals, they were not as successful in incorporating their verbal ability into the problem solving task. Unlike the 60 young normals, verbal ability (as measured by the WISC) had no relationship to ability to verbalize the correct choice (X2 < 1). Furthermore, even though the incorrect cues were explicitly pointed out in the successive presen- tation, they were unable to bring these under verbal con- trol. That is, the retardate seems to lack the ability to bring verbal behavior into the problem solving task. . Further support for the interaction of verbaliza- tion and inhibition came from the type of incorrect re- sponses made. It was implicit in the inhibition hypothe- sis that correct responding was more difficult due to the _intrusion of incorrect R's. The nature of these intruding R's did not seem to depend upon the particular cues most recently seen. It seems that SS verbalized particular cues or dimensions repeatedly regardless of whether that one had recently been seen, i.e., they perseverated. This was one of the criteria used by Denny (1964) for estab- lishing an inhibition deficit. Furthermore, the dimen- sions verbalized most frequently were those which Zeaman and House (1963) determined to be most salient for retard- ates and young children, i.e., form first and then color. 61 Also, those S3 for whom form was relevant were signifi- cantly more successful in verbalizing the discrimination (X2 = 6.4 p < .05) than was any other group. Retardates were the most affected by intruding R's. Significantly more retardates gave incorrect ver- balizations than did the 5-6 year olds (X2 = 9.5, p < .01), again supporting an inhibition deficit. More 5-6 year olds gave incorrect responses than did the 9-12 year olds (9 vs. 3) although this difference was not signifi- cant.~ The largest portion of incorrect verbalizations was form, followed by color and very few of size. In summary, the young normal child from 5-7 is just beginning to inhibit S-R modes of responding and to incorporate verbal behavior into the problem solving— task. Therefore, when the incorrect cues are explicitly pointed out, as in the successive method of presentation,' the young child is able to incorporate these into his. verbal system. However, when the need to inhibit re- sponses is not explicit, as in the simultaneous method of presentation, the young child is unable to incorporate these into the verbal system. Correct verbalization is further dependent upon the saliency of the relevant cues. 62 If the cues are salient, then there appears to be less need to specify the incorrect cues. If the irrelevant cues are most salient, then there is a greater need to explicitly specify these. By the time the child is 9-12 years old, verbal behavior dominates in the problem solv- ing task and both methods of presentation are handled r“ equally well. The picture differs for the retardate (MA 5-7). It makes no difference how advanced his verbal ability . :3 becomes, he is unable to inhibit earlier modes of respond- ing and incorporate verbal behavior into the problem solving task. Even when the need to inhibit is explicit, verbalizing correctly is not facilitated. That is, the retardate has a low IQ deficit and does not show increased ,Verbal problem solving behavior as does his normal MA match, even though verbal ability may increase. The results of this study support the findings of Milgram and Furth (1967) who found that retardates as com- pared with normals showed no developmental increase in a concept formation task as a function of verbal aid. Nor- mals did show an increased ability to incorporate the ver- bal aid into the problem solving task. 63 These results also support the proposal that the retarded and young normals lack self-initiating sets (Denny 1964). The regulatory function of speech as in inhibiter is just beginning to develop in the normal 5-6 year old child (Luria, 1961). When inhibition of responding is externally initiated or assisted (as in the explicit spe- cification of negative information in the successive method of presentation) than the young normal is able to verbalize the concept. However, if the desired verbal behavior is left to be self-initiated (as in the simultan- eous presentation) the young normal is often unable to inhibit incorrect or simple S-R problem solving behavior in favor of verbal problem solving behavior. The retardate is frequently unable to inhibit ‘under either method of presentation and lacks self-initiated tsets. Further, the retardate often fails to employ verbal problem solving behavior even when this behavior is prompted by explicitly specifying the negative or SA information (as in the successive presentation). IMPLICATIONS (for future research) For normal children ages 5-6, specifying the in- correct stimulus may be enough to facilitate learning of the correct verbal response. However, if an attempt is not made to bring these incorrect responses (particularly if they are dominant) under the child's verbal control, then learning may proceed very slowly and be character- ized by perseveration. For the retardate, it appears that specifying the ‘incorrect stimulus is less likely to be enough. A better approach would be t£> ensure the saliency of the correct stimulus, perhaps by a process such as fading. 64 BIBLIOGRAPHY 3T“: BIBLIOGRAPHY Berger, A. Inhibition of the eyelid reflex in three etio- logic groups of mentally retarded boys as compared with normals. Training sch. bu11., 1954, 51. Bruner, J. The course of cognitive growth. Am. psycholo- gist, 1964. Denny, M. R. A theoretical analysis and its application to training the mentally retarded. In Ellis, N. R. (Ed.) International Review of research in mental retardation, Vol. 2. New York: Academic Press Inc., 1966. Denny, M. R. Research in learning and performance. In Stevens, H. A. and Heber, R. (Eds.) Mental Retarda- tion: a review of research. Chicago: University of Chicago Press, 1964. ‘Denny, M. R. and Adelman, H. M. Elicitation theory: I. An analysis of two typical learning situations. Psychological rev., 1955, 62, 290-296. Erickson, M. T. and Lipsitt, L. P. Effects of delayed reward in simultaneous and successive discrimina- tion learning in children. J. comp. physiol. psychol., 1960, 53, 256-260. Frank, F. Perception and language in conservation. In Bruner, J. S. Studies in cognitive growth, New York: Wiley, 1966. Hermeline, and O'Connor. Verbalizing concepts. J. ment. defic. res., 1958, 2, 21-27. Horwitz, F. D. and Armentrout, J. Discrimination learning, manifest anxiety, and the effects of reinforcement. Child deve1., 1965, 36, 731-747. 66 67 Keeney, T. J., Cannizo, S. R., and Flavell, J. H. Spon- taneous and induced verbal rehersal in a recall task. Child devel., 1967, 38, 953-966. Kendler, H. H. and Kendler,T. S. Vertical and horizontal processes in problem solving. Psych. rev., 1962, 69, 1-16. Kendler, T. S. Verbalization and optional reversal shifts among kindergarten children. J. of verb. learn. r1 and verb. behav., 1964, 3, 428-437. ‘ Klugh, H. E. and Janssen, Raylyn. Discrimination learning by retardates and normals: method of presentation and verbalization. Am. J. of ment. def., 1966, 70, 903-906. Kuhlmann, F. Experimental studies in mental deficiency. Amer. j. psychol., 1904, 15, 391-446. Lipsitt, L. P. Simultaneous and successive discrimination learning in children. Child devel., 1961, 32, 337-347. Luria, A. R. Experimental study of the higher nervous activity of the abnormal child. J. ment. defic. res., 1959’ 3' 1-22. Marinesco, G. and Kreindler, A. Des reflexes condition- nels. J._psychologie, 1933, 30, 855-886. Mateer, F. Child behavior. Boston: The Gorham Press, 1918. Milgram, N. and Furth, H. G. Factors affecting concep- tual control in normals and retarded children. Child devel., 1967, 38, 531- . Panchenkova, E. F. The ontogenetic development of condi- tioning in the white rat. Zh. vssh. nerv. deiat., 1956, 6, 312-318. 68 Pascal, G. R., Stolurow, L. M., Zaborenko, R. N., and Chambers, K. S. The delayed reaction in mental defectives. Amer. j. ment. defic., 1951, 56, 152-160. Piaget, J. Play, dreams, and imitation in childhood. New York: Norton, 1951. Razran, G. H. S. Conditioned responses in children-- a behavioral and quantitative critical review F of experimental studies. Arch. psychol., 1933, g No. 148. E Reese, H. W. Verbal mediation as a function of age level. _<..mu—.,v Sen, A. and Clarke, A. M. Some factors affecting distract- Q ability in the mental retardate. Am. J. of ment. defic. , 1968, 73, 50-60. Sen, A. and Clarke, A. M. The effect of distraction during and after learning a serial recall task. Am. J. of ment. defic., 1968, 73, 46-49. Schusterman, R. J. Strategies of normal and mentally re- tarded children under conditions of uncertain out- come. Am. j. of ment. defic., 1964, 69, 66-75. Sidman, M. and Stoddard, L. T. Programming perception and learning for the mentally retarded child. In Ellis, N. R. (Ed.) International review of research in mental retardation, Vol. 2. New York: Academic Press, 1966. Spitz, H. H. The role of input organization in the learn- ing and memory Of retardation. In Ellis N. (Ed.) International review of research in mental retarda- tion, Vol. 2. New York: Academic Press, 1966. Stolurow, L. M. and Pascal, G. R. Double alternation be- havior in mental defectives. Amer. psychol., 1950, 5' 273—2740 69 Terrace, H. S. Stimulus control. In Honig, W. (Ed.) Operant behavior: areas of research and applica- tion. New York: Appleton-Century-Crofts, 1966. Thompson, W. R. The ability of children of different grade levels to verbalize on sorting tests. 1. Touchette, P. The effects of graduated stimulus change on the acquisition of a simple discrimination in severely retarded boys. J. of exper. and applied behavior., 1969, 39-52. Vince, M. A. "String-pulling" in birds. II. Differences related to size in greenfinches, chaffinches, and canaries. Anim. behav., 1958, 6, 53-59. Vygotsky, L. S. Thought and language. Cambridge, MIT Press, 1962. White, S: H. Evidence for a hierarchical arrangement of learning processes. In Lipsitt, L. and Spiker, C. (Eds.) Advances in child development and behavior, Vol. 2. New York: Academic Press, 1965. Zaporozhets, A. V. Some psychological problems of sensory training in early childhood and the preschool period. In Cole, M. and Maltzmann, I. (Eds.) A handbook of contemporary_Soviet psychology. New York: Basic Books, 1969. Zeaman, D. Discrimination learning in retardates. Train- Zeaman, D. and House, B. The role of attention in retard- ates discrimination learning. In Ellis, N. (Ed.) Handbook of mental deficiency. New York: Mcgraw- Hill, 1963. m1|ififlflllfilflMlfiliilfifilfiilfliflljifllififs