THE DEVELOPMENT AND GENERALIZATION OF DELAYED IMITATION1

An individual's behavior can be identified as imitative if it temporally follows the behavior of another individual and if its topography is controlled by the demonstrated behavior [Baer, Peterson, and Sherman (Journal of the Experimental Analysis of Behavior, 1967, 10 , 405–416)]. This definition takes into account both temporal and topographical characteristics of the behavior in question. More recent research in the area of imitation has interpreted the temporal component of the above definition differentially by limiting imitation to those topographically similar responses occurring within 3, 5, or 10 sec after a model's demonstration. Yet, Gewirtz and Stingle (Psychological Review, 1968, 75 , 375–397) pointed out that much of the imitation seen in young children is not of this immediate nature, but instead occurs sometime after a model's response. They further suggest that this type of imitative behavior can be characterized as a response class and is susceptible to development and modification as a function of consequences delivered to subjects contingent on this type of delayed responding. Four retarded children, three initially imitative and one nonimitative, were individually trained to imitate a number of motor responses in an immediate and a delayed fashion. Immediate imitation was defined as a response similar to a model's demonstration occurring within 5 sec after the model's demonstration; delayed imitation was defined as a response similar to a model's demonstration occurring more than 5 sec, but not more than 25 sec, after the model's demonstration. A reversal (ABAB) design was employed to examine the experimental development of a generalized delayed imitative repertoire. Untrained probe responses were demonstrated to subjects systematically through the ongoing training. Generalized immediate and delayed imitation were observed in each subject; this generalization was restricted to the type of imitation currently undergoing training. This development of a generalized imitation repertoire was observed in each subject. That is, these subjects imitated some responses that had never been specifically trained. More importantly, a training package consisting of prompting, fading, and consequences for delayed imitation functioned to develop generalized delayed imitation. These data exemplify a special case of generalization that was a function of the most recent training history of immediate or delayed imitation. The reversal design demonstrated that imitations of nontrained models were either delayed or immediate, depending upon which form of imitation was currently receiving training. Therefore, for each form of imitation trained, delayed or immediate, a corresponding response class was demonstrated. These data relate to data reported by Garcia, Baer, and Firestone (Journal of Applied Behavior Analysis, 1971, 4 , 101–112). The association lies in the proposition that there are identifiable boundaries of generalized imitation and that these boundaries are functionally related to previous training histories.     

The development of generalized imitation within topographically determined boundaries

A multiple baseline technique was employed to examine the experimental development of an imitative repertoire within preselected topographical boundaries. Four severely retarded children, initially nonimitative, were individually trained to imitate a number of motor and vocal responses by shaping and fading procedures. Other untrained responses (probes) were demonstrated to the subjects systematically throughout the ongoing training. Training responses were divided into three topographical types: small motor, large motor, and short vocal responses. Probe responses were divided into four topographical types: small motor, large motor, short vocal, and long vocal responses. Following a multiple baseline format, sequential training of the first three types was begun at different temporal periods of the study; unreinforced imitative generalization was continually measured by the probes. Generalized imitation was observed in each subject (untrained responses were imitated even though unreinforced); but this generalization was restricted to the topographical type of imitation currently receiving training or having previously received training.     

COMPARISON OF EXTINCTION,DRO 0-SEC,AND DRO 6-SEC IN THE ELIMINATION OF IMITATIVE RESPONDING UNDER DISCRETE-TRIAL PARADIGMS1

Imitative response levels were observed for normal 5-yr-old children within a discrete-trial imitation paradigm. Following imitation training sessions, children were observed under conditions of contingent reinforcement for three experimental sessions. Following the contingent reinforcement phase, children were then observed under reversal conditions for four sessions. A 3 × 4 repeated measures design was employed for observing response decrements under conditions of extinction, DRO 6-sec, and DRO 0-sec. The DRO 0-sec group demonstrated a significantly greater decrement in imitative responding than did the DRO 6-sec and extinction group. The differences in effects between the two DRO conditions and results from previous studies in which DRO procedures have had little effect in reducing imitative responding, suggest an incomplete understanding of the ways in which DRO procedures may operate.     

The Structural Aspects of Schizophrenic Responses to Auditory Stimuli

The function of reinforcement as a performance versus a learning variable was examined with the use of a number of varying imitation training conditions. Forty white and 40 black Ss were used. The analysis consisted of five 2 (Testing Situation) × 2 (White versus Black children) × 8 (Training Conditions) factorials with the use of an analysis of variance technique. The results question the hypothesis that reinforcement is only a performance variable. Partial support is provided for the assumption that imitative behavior may be maintained through similarity of responding where imitation has been acquired through a strong history of reinforcement for imitation. Few racial differences were discovered.     

Homo imitans or homo provocans? Human imprinting model of neonatal imitation

Searching for the mechanism of neonatal imitation resulted in the discovery of neonatal initiative capacity, here called “provocation”. Newborns spontaneously produced previously imitated gestures while waiting for the experimenter’s response. A psychophysiological analysis revealed that imitation was accompanied by heart rate increase while gesture initiation was accompanied by heart rate deceleration, suggesting different underlying mechanisms. Results imply that infants are not only capable of responding to a model movement by imitating, but that they also have the capacity to provoke an imitative response, thus sustaining an interaction. These findings may constitute a laboratory demonstration of the first dialogue and, according to our hypothetical model, they represent how human imprinting begins.     

GENERALIZED IMITATION AND RESPONSE-CLASS FORMATION IN CHILDREN WITH AUTISM

An experimental analysis of imitation was conducted to examine the influence of response topography on generalization of imitation across three response types. Four children with autism were presented with both reinforced training trials and nonreinforced probe trials of models from vocal, toy-play, and pantomime response types. The probe trials were used to examine generalization within each response type. A multiple baseline design was used to analyze percentage of matching and nonmatching responses to models across response types. This study, the first to analyze imitative response classes in children with autism, showed that imitation generalized from reinforced training models to nonreinforced probe models within a response type, but it did not generalize across response types. Thus, functional response classes determined by topographical boundaries were exhibited within generalized imitation.     

Generalized imitation of facial models by children with autism

Imitation is an essential skill in the acquisition of language and communication skills. An initial phase in teaching young children with autism to engage in appropriate affective responding may be to teach the imitation of facial models. Using a multiple baseline across participants design, imitation training (consisting of modeling, prompting, differential reinforcement, and error correction) was introduced successively across 3 participants. Low and inconsistent rates of imitation of facial models were observed in baseline. All of the participants learned to imitate some of the facial models presented during imitation training, but only 2 of the 3 participants demonstrated generalized responding across stimuli.     

An analysis of generalized imitation

An experimenter presented English words to three intermediate-level children and reinforced accurate imitation of these words. The experimenter also presented novel Spanish words, but the imitation of these words was never experimentally reinforced. One subject quickly ceased performing non-reinforced imitative responses. The other two subjects were exposed to a series of conditions designed to facilitate discrimination. Upon observing the first subject for one session they immediately ceased imitating Spanish demonstrations. For all three subjects, when reinforcement was delivered for responses other than imitation, all imitative responses eventually ceased. When reinforcement was reintroduced for English imitations there was an immediate resumption of such responses to their previous 100% level. The occurrence of non-reinforced imitations in this and previous studies was discussed as being a function of one or combination of four variables: (1) similarity acquiring conditioned reinforcing properties, or (2) instructional, (3) coincidental, or (4) conditional stimulus generalization.     

The development of imitation by reinforcing behavioral similarity to a model

This research demonstrated some of the conditions under which retarded children can be taught to imitate the actions of adults. Before the experiment, the subjects were without spontaneous imitative behavior, either vocal or motor. Each subject was taught, with food as reinforcement, a series of responses identical to responses demonstrated by an experimenter; i.e., each response was reinforced only if it was identical to a prior demonstration by an experimenter. Initially, intensive shaping was required to establish matching responses by the subjects. In the course of acquiring a variety of such responses, the subjects' probability of immediate imitation of each new demonstration, before direct training, greatly increased. Later in the study, certain new imitations, even though perfect, were never reinforced; yet as long as some imitative responses were reinforced, all remained at high strength. This imitativeness was then used to establish initial verbal repertoires in two subjects.     

The effect of incongruent instructions and consequences on imitation in retarded children

Three retarded boys served as subjects in a 13-phase experiment. In eight of these phases, the instructions administered by the experimenter before demonstrating a behavior and the consequences for imitative behavior were incongruent (the consequences were not those indicated by the instructions). Consequences rather than instructions controlled imitative behavior when (a) subjects were instructed not to imitate but received reinforcers if they imitated; (b) subjects were instructed to imitate but were differentially reinforced for other behavior; (c) subjects were instructed to imitate but were verbally reprimanded for imitation. Although subjects were highly imitative at the beginning of the study, when there was no reinforcement for imitation subjects gradually stopped imitating when instructed not to imitate. Instructions seemed to control imitative behavior when there was no reinforcement for imitation and subjects were instructed to imitate. These results indicated a need for further investigation of antecedent and consequent variables in imitation experiments and pointed out that certain techniques may be more efficient than others in eliminating well-established responses.     

Attention-getting vs. imitation effects on depressed mother-infant interactions

The effects of instructing mothers to “imitate” their infant versus “keep their infant's atten tion” were examined during mother-infant face-to-face interactions of 18 mothers reporting depressive symptoms as compared with 22 mothers who did not report such symptoms. Mothers were generally rated as showing more positive facial expressions and more game playing (particularly the depressed mothers) during the attention-getting versus the imitation sessions. The infants received more optimal physical ac tivity, and facial expression ratings during attention getting, and the infants of depressed mothers, in par ticular, showed more positive facial expressivity and more joy expressions. As might be expected for the imitation condition, mothers showed more imitative behavior, contingent responsivity, and silence during gaze aversion. Infants generally showed more disinterest and self-comfort behaviors, and the infants of depressed mothers, in particular, showed more anger expressions, fussiness, and squirming during the imitation condition. The data suggest that the attention-getting condition was the most effective “intervention” for eliciting positive behavior in the depressed mother-infant dyads.     

Behavioral similarity as a reinforcer for preschool children

Three experiments evaluated whether behavioral similarity provided by an adult could serve as a reinforcer for the modelling behavior of four preschoolers. In each experiment, sessions consisted of two kinds of trials: (1) experimenter-modelled trials, when the child's imitation of modelled motor responses was reinforced with praise and tokens, and (2) child-modelled trials when experimenter imitation of child-modelled responses was contingent upon the child's modelling one of three alternative responses: operation of a ball, horn, or clicker. Experiment I showed that the children consistently modelled whichever responses the experimenter imitated. Experiment II determined whether that performance was due to differences in the amount of experimenter behavior following imitated versus nonimitated child models or to experimenter imitation. Neither reducing nor increasing the amount of experimenter behavior following the children's nonimitated models altered their modelling of imitated responses. Experiment III evaluated whether experimenter imitation of child models was a reinforcer because the child's imitative responses were reinforced on experimenter-modelled trials. In Experiment III, the children's nonimitation of experimenter-models was reinforced with praise and tokens on a schedule of differential reinforcement of other behavior, yet they continued to model experimenter-imitated responses on child-modelled trials. These results indicate behavioral similarity was reinforcing, though no conditioning history through which it acquired that function was demonstrated.     

AN INVESTIGATION OF THE EFFECTS OF TOKEN REINFORCEMENT ON DIVERGENT VERBAL RESPONDING1

The effect of token reinforcement on three classes of divergent verbal responding to verbal stimulus items from three measures of the Wallach-Kogan Creativity test was examined. The subjects consisted of two “gifted”, two “average”, and two “learning disabled” children from a public-school setting. The design utilized both an intra-subject and inter-subject multiple-baseline design with a reversal design added. Tokens were dispensed contingent upon the number of appropriate verbal responses to each stimulus item, i.e., a continuous schedule of reinforcement was used—one token per appropriate response. Interobserver agreement on “appropiateness” was 993%. The tokens could be used to “purchase” items listed in a “menu” format (e.g., ball and jacks—100 tokens). Consistently large effects of token reinforcement were observed, without generalization of effect across the three classes of behaviors, or conditions. These results support the works of Goetz and Baer (Journal of Applied Behavior Analysis, 1973, 6 , 209–217), Goetz and Salmonson (Behavior Analysis and Education, G. Semb, (Ed.), University of Kansas, 1972, 53–61), and Maloney and Hopkins (Journal of Applied Behavior Analysis, 1973, 6 , 425–434). Implications lie in continued experimental analysis of behavior approach to the concept called “creativity”. Additionally, implications are made in classroom application of assessing and intervening in the area of divergent responding for children with deficits in those areas.     

THE TRAINING AND GENERALIZATION OF A CONVERSATIONAL SPEECH FORM IN NONVERBAL RETARDATES1

Differential reinforcement and imitation were used with two retarded children to train three sequential verbal responses associated with the display of a picture and questions related to that picture. Each response consisted of a three-word chain in sentence form; combined with verbal responses from the experimenter, this trained sequence formed a short conversational unit. Three experimenters measured the use of each sentence in settings different than the one in which training took place, and with pictures different than those used during training. Two types of generalization sessions were used: (1) General sessions, during which 10 pictures never used during training were displayed to the subject with reinforcement delivered on a noncontingent basis, and (2) Intermixed sessions, during which 10 pictures never used during training were displayed to the subject, but a picture having received training was also displayed, and correct responses to this picture were reinforced on a variable schedule. Both subjects learned the sentences being trained. However, little generalization was evident from this training when all experimenters conducted General probe sessions. Generalization occurred with one experimenter only after that experimenter conducted Intermixed probe sessions. Generalization to a third experimenter was then observed (i.e., after the first two experimenters had conducted Intermixed probe sessions) without the use of Intermixed probe sessions by this third experimenter.     

PRODUCING GENERATIVE SENTENCE USAGE BY IMITATION AND REINFORCEMENT PROCEDURES1

Three retarded subjects and two developmentally normal toddlers were trained using imitation and reinforcement procedures to use correct sentences. The experimental task was to use sentences with correct subject-verb agreement to describe pictures that were presented to the subjects. Two classes of sentences were taught: those involving a plural subject that required the use of the verb “are” (for example, “the boys are running”) and those involving a singular subject that required the use of the verb “is” (for example, “the boy is running”). The basic design of the study involved multiple baselines for each class of sentences. Four of the subjects began to produce novel, untrained sentences of a particular type to generalization probe pictures when that particular class of sentence was currently being trained. Thus, the imitation and reinforcement procedures appeared to be functional in producing generative sentence usage for both types of sentences. One subject produced correct sentences to both singular and plural probe pictures when only “is” sentences had been taught. A reversal procedure and retraining phase indicated that for this subject, imitation and reinforcement procedures for training one class of sentence behavior seemed functional in producing generative responses of the other class of sentences.     

PARENTAL TRAINING OF THE PLURAL MORPHEME IN NORMAL TODDLERS1

Experimental studies with young normal children can provide useful strategy for the functional analysis of language. Research by Whitehurst (Journal of Experimental Child Psychology, 1972, 13 , 502–515) exemplifies this approach, in which 2-yr-old children were exposed to a training procedure that involved imitation and differential reinforcement for a two-word productive sequence of the adjective-noun form. Results indicated that these young children could be trained to produce rudimentary novel utterances that were grammatically appropriate. More recently, Hursh and Sherman (Journal of Experimental Child Psychology, 1973, 15 , 328–339) reported the existence of a functional relationship between parental modelling and reinforcement of vocalizations, and increased instances of these vocalizations in young (15- to 20-month-old) children. These demonstrations relate to a summary of findings with linguistically deviant populations discussed by Sherman (Advances in Child Development, 1971). The present study attempts to extend these “generative” investigations to normal children and to provide useful information for those interested in teaching speech forms to them. Modelling and differential reinforcement were used by three mothers to establish the use of the plural morpheme in the speech of their 19- to 25-month-old children. During training trials, verbal praise was presented contingent on correct labelling of singular and plural items, while correct labelling was modelled if the child responded incorrectly. Children learned to label specifically trained sets of singular and plural items and also exhibited correct labelling when asked to label never-trained singular and plural items. After establishing correct usage, the same training procedures were used to train reversed labelling (plural responses to single items and singular responses to plural items). This produced a corresponding reversal of responding (both trained and untrained items) by each child for the plural items but not for singular items. Correct singular and plural labelling was recovered by returning to the initial procedures.     

TWO-TO-ONE versus ONE-TO-ONE STUDENT-TEACHER RATIOS IN THE OPERANT VERBAL TRAINING OF RETARDED CHILDREN1

The operant training of two retarded children simultaneously on a picture-naming task was investigated as an alternative to the more commonly reported one-to-one student-teacher ratio. In Experiment I, two conditions were compared in which the children received primary reinforcement on a fixed-ratio schedule for responding correctly on prompt and probe trials in a standardized picture-naming procedure. During the “Group Condition”, the experimenter alternated from one child to the other after each primary reinforcement, after each incorrect response, after each response omission, and after each 10-sec period in which a child did not “attend” (by making a trial-initiating response) when it was his or her turn to be worked with. During the “Individual Condition”, the experimenter worked with only one child, and presented trials whenever the child made attending responses. Experiment I demonstrated that the Group Condition was more efficient than the Individual Condition in terms of total correct responses and total pictures learned per unit of training time. Incidental learning was also found in that the children learned some of each others' pictures as well as their own, thus indicating a further advantage of the larger student-teacher ratio. In Experiment II, an attempt was made to equate the two conditions, except for the presence of two children in the Group Condition, by ignoring the child in the Individual Condition for brief periods equal to those that occurred in the Group Condition when the experimenter presented training trials to the other child. The results demonstrated that the greater efficiency of the Group Condition was not due to the manner in which training time was allocated to the two members of a group. It also replicated the finding that the children learned some of each others' words in the Group Condition.     

Generalized imitative affection: Relationship to prior kinds of imitation training

Two experiments utilizing first- and second-grade Canadian children showed that generalized imitative physical affection was most facilitated by prior imitative physical contact training (as opposed to verbal contact training) when the object of the affection was either a toy teddy bear, Experiment I, or an adult human, Experiment II.Additional findings from Experiment I showed that generalized imitative physical aggression was equally facilitated by imitative physical contact training and that punishment, as well as extinction operations applied to training imitations, resulted in suppression of all generalized imitations with no differential effect of punishment on affection or aggression being noted.The lack of any persisting imitations in a control group in Experiment II, which received noncontingent reinforcement but instructional prompting for training imitations, suggested that instructional control of imitation responses was initially weak and that the contingency between reinforcement and training imitations was critical for continued occurrence of training imitations and any occurrence of generalized physical affection imitations.     

The effects of differential reinforcement on generalized imitation

The present study examines the transfer of imitative learning to other nonimitative performance conditions and compares imitative and nonimitative performance under contingencies of differential reinforcement for S⁰ behavior, differential reinforcement for nonimitative behavior, and extinction. Many authors have suggested that a child's continued imitative performance of rewarded S^D and unrewarded S^Δ behavior is a function of subtle social cues or experimental demand present in most generalized imitation procedures. The two experiments presented here support that conclusion but also provide evidence that conclusions drawn from such generalized imitation studies were generally accurate. Even though a child's trial-by-trial imitative performance appeared to be a function of procedural artifacts, the child's later performance in the role of a model indicated that a functionally interdependent generalized response class of imitative behavior had been learned while the child imitated. As such, these experiments generally supported Baer's secondary reinforcement hypothesis for imitative performance and suggest that future research employ nonimitative tasks such as reversed imitation as a measure of imitative learning.