Indexador

The reinforcers produced by one response reduce the rate of other, concurrently reinforced responses. An analysis of the logical and empirical implications of the relation indicates that one reinforcer must have this effect on responses maintained by other reinforcers even when all reinforcers are produced by the same class of responses. A quantitative expression of the relation leads to a formulation, mathematically equivalent to Herrnstein's (1970), in which the rate of a reinforced response is a joint function of (1) an excitatory effect of the reinforcers produced by that class of responses, and (2) an inhibitory effect of the total reinforcers produced both by that class and by other classes of responses.

A token system was used to attempt to increase the accuracy with which special education students answered questions about reading assignments. In the token system, students recorded their own data, received toy money for accurately completing assignments, and were allowed to spend their toy money at the end of the week for educational activities. The accuracy with which students answered questions was higher when the token system was in effect than when it was not. When student teachers were used to manage the token system and when the self‐recording feature of the system was removed, only slight changes in the accuracy of the student performance were obtained.

Three pigeons pecked for food in an experiment in which reinforcements were arranged for responses terminating sequences of interresponse times. Each reinforced interresponse time belonged to a class extending either from 1.0 to 2.0 sec (class A) or from 3.0 to 4.5 sec (class B). Reinforcements were arranged by a single variable‐interval schedule and a random device that assigned each reinforcement to one of four sequences of two successive interresponse times: AA, AB, BA, or BB. Throughout the experiment, half of the reinforcements were delivered for interresponse times in class A and half for those in class B. Over conditions, the interresponse time preceding a reinforced interresponse time always, half of the time, or never, belonged to class A. The duration of the interresponse time preceding a reinforced one had a pronounced effect on response patterning. It also had a pronounced effect on the overall response probability, which was highest, intermediate, and lowest, when the interresponse time preceding a reinforced interresponse time always, half of the time, or never, belonged to class A, respectively. In no case were successive interresponse times independent, so that overall response probability was not representative of momentary response probabilities.

Two pigeons performed a three‐key matching‐to‐sample task. The comparison (side key) stimuli were either solid colors or white lines. The sample (center key) stimuli were either compounds (white lines on colored grounds) or elements (white lines on black grounds on some trials, and solid colors on other trials). Sample stimuli were presented for nine sample stimulus durations ranging between 0.04 and 5.00 sec. Within each daily session, both compound and element samples were presented at each sample duration in a random sequence. Compound samples controlled matching responses less effectively than did element samples at all sample stimulus durations.

Shock intensities (1 to 4 mA) and shock durations (0.3 to 0.75 sec) were concurrently varied over a range commonly used in free‐operant avoidance studies using a lever‐press response. Response rates were a positive linear function of the log of the product of intensity times duration. Shock rates were a negative linear function of that log. The increase in response rates was primarily due to a selective increase in the conditional probability of making responses with long interresponse times. The disproportionality of receiving shocks early in the session (warm‐up) was also a linear function of the log of the intensity‐duration product, with increasing disproportionality as the value of the intensity‐duration product was increased. Thus, with all measures of the avoidance performance, shock intensity and shock duration combine in a multiplicative fashion to determine the avoidance performance.

Five pigeons were exposed to multiple and concurrent variable‐interval, variable‐interval reinforcement schedules in which reinforcement availability in one component was never signalled. During certain phases of the experiment, reinforcement availability in the other component was signalled. Behavioral contrast was observed in seven of eight instances when reinforcement availability in the multiple schedules was signalled. Under the concurrent schedules in which reinforcement availability was signalled, the subjects did not always allocate more time to (prefer) the component containing non‐signalled reinforcement, as would be predicted by an account of behavioral contrast holding that contrast results from the introduction of a less‐preferred condition in one component of a multiple schedule.

The blockbuilding behavior of three preschool girls was analyzed in terms of the forms manifest in any completed block construction, and found to contain few different forms in baseline sessions. Social reinforcement, given contingent on the production of any form not previously constructed within the current session (i.e., every first appearance of any form within a session was reinforced but no subsequent appearances of that form within that session were), increased the number of different forms built per sessions. Social reinforcement, given for all second and later appearances within the session, decreased the number of different forms built per session. Furthermore, it was found that new forms (forms never seen before in the child's total prior sequence of blockbuilding sessions) emerged at higher rates during periods of reinforcement of different forms (first appearances) than during periods of baseline or reinforcement of same forms (second and later appearances).

The mother of a 3‐yr‐old girl who showed “autistic‐like” behavior was cued via an FM wireless microphone systematically to approve and disapprove of her child's behavior. After baselines were taken on two categories of problem behavior (Pre‐Academic and Social Behavior), the social contingencies were applied successively to each category. The Pre‐Academic task was quickly established in the child when the mother applied these social contingencies. This result was replicated with requests for social interaction. In the final phase, cueing was withdrawn from both situations and the mother was able to maintain the child's improved behavior. An analysis of the mother's behavior suggested that her increased use of social punishment for inappropriate behavior was the key factor in the child's increasing responsiveness. Follow‐up seven months later indicated that the improvements maintained.

The effectiveness of a token economy system in producing improvement in the academic performance of children with Down's syndrome was tested. One group of seven children received token reinforcement for correct responses and showed significant improvement both in arithmetic and language. A second matched group of six children received only verbal praise for correct responses to the same instructional materials and failed to improve in arithmetic but showed significant gains in language. Re‐test scores one year later revealed that the Token Group maintained its gains in both subjects whereas the language performance of the No‐Token Group showed a significant decline.

We have treated 20 autistic children with behavior therapy. At intake, most of the children were severely disturbed, having symptoms indicating an extremely poor prognosis. The children were treated in separate groups, and some were treated more than once, allowing for within‐ and between‐subject replications of treatment effects. We have employed reliable measures of generalization across situations and behaviors as well as across time (follow‐up). The findings can be summarized as follows: (1) Inappropriate behaviors (self‐stimulation and echolalia) decreased during treatment, and appropriate behaviors (appropriate speech, appropriate play, and social non‐verbal behaviors) increased. (2) Spontaneous social interactions and the spontaneous use of language occurred about eight months into treatment for some of the children. (3) IQs and social quotients reflected improvement during treatment. (4) There were no exceptions to the improvement, however, some of the children improved more than others. (5) Follow‐up measures recorded 1 to 4 yr after treatment showed that large differences between groups of children depended upon the post‐treatment environment (those groups whose parents were trained to carry out behavior therapy continued to improve, while children who were institutionalized regressed). (6) A brief reinstatement of behavior therapy could temporarily re‐establish some of the original therapeutic gains made by the children who were subsequently institutionalized.