Indexador

Either a partial blackout, or the blackout plus a “feeder flash”, occurred in lieu of reinforcement on two procedures that produced opposite patterns of responding after reinforcement. Response rate was elevated after reinforcement omission on the procedure that produced a “pause‐and‐respond” pattern following reinforcement, but depressed after reinforcement omission on the procedure that produced a “respond‐and‐pause” pattern. The effect of blackout plus feeder flash was generally intermediate between the effects of blackout and the effects of reinforcement. These results are consistent with an interpretation of reinforcement omission effects in terms of the discriminative temporal control exerted by reinforcement and stimuli similar to it.

Use of concurrent variable‐interval performances confirms that more‐complex visual patterns have greater reinforcement value for human subjects than less‐complex patterns. The findings tally with verbal evaluative ratings and with results of previous experiments using a discrete‐trial binary‐choice technique.

Lever‐pressing responses were maintained in the squirrel monkey when the only consequence of responding was the delivery of a response‐produced electric shock, or alternatively, a brief visual stimulus that was occasionally followed by an electric shock. When shock was produced by the first response occurring after 8 min (8‐min fixed‐interval schedule), a period of no responding at the beginning of the interval was followed by a gradual increase in response rate during the interval. Similar rates and patterns of responding were maintained when a 1‐sec visual stimulus was produced by the first response occurring after 8 min and shock delivery followed the brief stimulus. Subsequently, patterns of positively accelerated responding were engendered during individual fixed‐interval components when the first response occurring after 4 min produced a 1‐sec visual stimulus and shock delivery followed the second, and later the fourth, presentation of the 1‐sec stimulus. When the duration of the brief stimulus was varied over a 100‐fold range from 0.1 to 10.0 sec (1) mean response rates decreased monotonically as stimulus duration increased, and (2) patterns of positively accelerated responding were least variable and response rates during the initial part of each 4‐min interval were lowest at a stimulus duration of 1 sec.

A fixed‐interval schedule of reinforcement was modified by dividing each interval into 4‐sec trial periods. No more than one response could occur during each trial because the operandum was inactivated for the remainder of any trial in which a response occurred. For example, under a 28‐sec schedule, no more than seven responses could be emitted between reinforcements. Probabilities of responding by pigeons under six values of this discrete‐trial fixed‐interval schedule were best described by a two‐state model: responding was either absent or infrequent immediately after reinforcement; then, at some variable time after reinforcement, there was an abrupt transition to a high and constant probability of responding on each trial. Performances under the discrete‐trial procedure were less affected by uncontrolled sources of variance than performances under equivalent free‐operant fixed‐interval schedules.

Pigeons and rats were exposed to multiple schedules with different schedules of electric shock superimposed on identical schedules of food reinforcement during each of two components. During one component, (adjusting‐intensity) the intensity of electric shock depended on responding. Each response increased the intensity while intensity decreased between responses. During the other component (constant‐intensity) the intensity was fixed at the value at which it had been adjusted at the end of the immediately preceeding adjusting‐intensity component. In one experiment, shock was continuous during both components. In another experiment, instead of continuous shock, a brief pulse was delivered immediately after each response. During the adjusting‐intensity component of both experiments, pigeons and rats responded at a rate just sufficient to keep the shock constant (critical rate). During the constant‐intensity component, responding depended on whether shock was delivered continuously or in pulses. When shock was continuous, response rate during the constant‐intensity component was higher than the critical rate. When shock was pulsed, response rate during the constant‐intensity component was equal to the critical rate.

Control of avoidance response patterning by time‐correlated stimuli was studied in rhesus monkeys. At several shock‐shock = response‐shock intervals, the intervals were divided into 8, 4, or 2 time segments by correlating, respectively, 1, 2, or 4 discrete lights of an eight‐light display with each successive segment. A further condition examined response patterning when the avoidance interval was not segmented: all eight lights of the display were lit throughout the interval. Reversal of the order in which the lights were lit in sequence was also examined. Generally, increasing the number of lights lit at one time (decreasing the number of signalled time segments) increased response rates and shifted interresponse time distributions to the left. When the lights were lit one at a time, signalling eight discrete time segments of the interval, response rates were consistently low.

After training three albino rats to bar press during a multiple fixed‐interval variable‐interval schedule, the response‐reinforcer dependency was simultaneously removed from both components, converting the schedule to multiple fixed‐time variable‐time. Response rates were reduced in both components under these conditions but the fixed‐time schedule maintained relatively higher response rates with each rat. After reinstating the response‐reinforcer dependency in both components, responding was conventionally extinguished by rendering the pellet dispenser inoperative. Responding rapidly decreased to near‐zero levels. Differences in fixed and variable‐time schedules in sustaining behavior are discussed in terms of differences in response rates at the time of reinforcer delivery. Similarities and differences between conventional extinction and schedules delivering response‐independent reinforcers are also discussed.

Resumo não informado pela fonte indexada.

Resumo não informado pela fonte indexada.

Sixteen pupils in a psychiatric hospital were assigned to two tutorial reading classes and balanced on six pupil characteristics and teacher preferences for the children. The effects of reward and cost procedures in a token program were assessed using both within‐ and between‐subject comparisons in the following phases: (1) Baseline; (2) Token I, teacher evaluated and reinforced children for appropriate behavior; (3) Withdrawal of Tokens; (4) Token II, same as Token I; (5) Token III, same as Token I and II, but switched order of class meeting time; and (6) Self‐Evaluation, students rated their own behavior and received prizes based on their rating, rather than the teacher's rating. The token program was markedly successful in reducing disruptive behavior and in increasing reading skills in both the Reward and Cost Classes, but there were no significant differences in the effects of the reward versus the cost procedure. While cost may be seen as a punishment procedure, there were no adverse side effects observed in the Cost Class at any time when the token program was in effect. The order of the classes was unrelated to the level of disruptive behavior or academic progress. The Self‐Evaluative Phase, in which the students rated their own behavior, was included as an alternative to the abrupt withdrawal of tokens. In this phase, disruptive behavior remained at the previous low level.