Violations of stochastic transitivity on concurrent chains: Implications for theories of choice

The concurrent-chains procedure has been used to measure how choice depends on various aspects of reinforcement, such as its delay and its magnitude. Navarick and Fantino (1972, 1974, 1975) have found that choice in this procedure can violate the condition of stochastic transitivity that is required if a unidimensional scale for reinforcements is to be possible. It is shown in this paper that two simple unidimensional models of choice on concurrent chains can produce violations of stochastic transitivity. It is argued that such violations may result from the complex contingencies of the concurrent-chains procedure.     

Behavioral economics of drug self-administration. II. A unit-price analysis of cigarette smoking.

In behavioral economics, the ratio of response requirement to reinforcer magnitude is referred to as unit price. Previous research with nonhuman subjects has demonstrated that (a) comparable amounts of food are consumed at the same unit price even though different response requirements and reinforcer magnitudes comprise that unit price and (b) increases in unit price decrease food consumption in a positively decelerating fashion. The present study assessed the generality of these findings to the cigarette smoking of 5 human volunteers. During approximately 18 3-hr sessions, various combinations of response requirement (fixed-ratio 200, 400, and 1,600) and reinforcer magnitude (1, 2, and 4 puffs per bout) were arranged. Consumption (i.e., the number of puffs) generally was comparable at the same unit price independent of the response requirement and reinforcer magnitude comprising that unit price. In addition, increasing unit price generally decreased consumption in a positively decelerating fashion. These results extend the generality of the unit-price analysis to human cigarette smoking. Moreover, these results further support the position that reinforcer magnitude and response requirement are functionally equivalent and interact to determine consumption. The concept of unit price, by integrating and summarizing the effects of those two operations, provides a more parsimonious explanation of the results than do separate evaluations of the effects of response requirement and reinforcer magnitude.     

OVERJUSTIFICATION EFFECTS IN TOKEN ECONOMIES

This study tested the relevance to clinical token economies of the overjustification hypothesis that tangible reward interferes with intrinsic interest in target behaviors and causes such behaviors to be less probable following a period of reinforcement than preceding such a period. The study was carried out in an ongoing token economy for chronic psychiatric patients. Alternated over an 8-week period were weeks of token and no-token reward for one of the program's target behaviors, toothbrushing. Two different amounts of token reward were employed in order to examine whether reward magnitude might influence the presence or extent of overjustification effects. Little evidence was found for the presence of overjustification effects in token economies. However, maintenance of toothbrushing was greater in no-token weeks following weeks of low amounts of token reward than in no-token weeks following weeks of higher amounts of reward. The importance of such complex functional relationships is discussed.     

Reinforcement magnitude and the inhibiting effect of reinforcement

In a two-key concurrent variable-interval schedule (using pigeons), if the reinforcement frequency for one response is held constant while that for the other is increased, the rate of response on the constant key decreases. The immediate reinforcement for key pecking can usually be conceptualized as the change from a condition in which the key light is on and the food hopper light is off to one in which the key light is off and the hopper light is on. The prechange condition is associated with a delay to food of one-half the average interreinforcement interval in effect during this condition. The postchange condition is associated with a delay to food of about .5 seconds. The programming of additional reinforcement results in a decrease in the delay to food associated with the prechange stimulus condition, and thus a decrease in the value of the improvement that results from the change. This would appear to be analogous to a decrease in the amount of reinforcement, and thus sufficient explanation for the decrease in the rate of the response.     

Temporal control of behavior and the power law

The performance of rats and pigeons under fixed-interval schedules was studied in two experiments. The duration of postreinforcement pause was a declining proportion of fixed-interval duration. For pigeons this was true both when the duration of the reinforcer was fixed and when it was increased in direct proportion to increases in fixed-interval duration; the longer reinforcer durations did, however, lengthen the postreinforcement pause at higher schedule values. A quantitative analysis of data from Experiments 1 and 2 and from other studies showed that fractional exponent power functions described the relationship between postreinforcement pause and fixed-interval value; similar functions have previously been observed in studies of temporal differentiation. It was concluded that power functions reflect a direct causal, rather than artifactual, relationship between performance and the temporal requirements of reinforcement schedules.     

Effects of food deprivation and reinforcement magnitude on conditioned suppression

In Experiment I, the responding of rats lever pressing on a variable-interval schedule for sucrose solution was partially suppressed by a variable duration conditioned stimulus followed by shock. When food deprivation was increased, response rates during and before the conditioned stimulus increased monotonically. Varying the concentration of sucrose across blocks of sessions or from session to session in a semi-random sequence had little effect on response rates either before or during the conditioned stimulus. With a fixed sequence of increasing concentrations across a five-session block, increased concentration produced much more rapid increases in response rates before than during the conditioned stimulus. In Experiment II, rats were presented with the same sequence of increasing concentrations across a five-session block. When tested at 80% body weight, response rates increased rapidly as concentration increased, but at 100%, body-weight rates increased only slightly. The effect of a change in body weight in Experiment II thus mimicked the effect of the conditioned stimulus in the latter part of Experiment I. These findings support the view that the effect of a pre-aversive conditioned stimulus is similar to that of a change in food deprivation, but unlike that of a change in reinforcement magnitude.     

音乐愉悦体验的神经机制

愉悦情绪体验是音乐活动中最普遍的心理现象。通过系统回顾相关的神经科学研究, 认为音乐愉悦体验与大脑奖赏系统的活动有关, 并涉及伏隔核与听觉皮层等其他脑区的交互。在这个过程中, 多巴胺的传递与音乐愉悦体验存在因果联系。基于预期视角, 奖赏预测误差理论和音乐信息理论模型可以解释音乐愉悦体验的产生机制。未来研究应进一步检验伏隔核及各皮层在音乐愉悦体验中的功能, 并整合不同的预期理论。     

非符号数量表征和符号分数表征的关系

个体学习符号分数的一个关键是能对其数值形成准确表征。现有研究假设符号分数表征的认知基础是人类自婴幼儿期就具有的非符号数量表征(如表征两个集合各自的数量, 或两个数量的比例)。其证据包括表征非符号数量(尤其是非符号数量比例关系)和表征符号分数在行为和大脑神经活动层面上都表现出相关性。然而要说明非符号数量表征是符号分数表征的认知基础, 还需更多研究表明两者在数量概念上的独特相关和因果联系, 并阐明符号分数表征形成的认知机制。     

Persistence during extinction: Are judgments of persistence affected by contingency information?

The partial reinforcement extinction effect (PREE), defined as increased behavioral persistence following intermittent reward, is considered an important outcome of instrumental learning contingencies, both inside and outside the laboratory. Since adults have a rich experience with situations in which desired outcomes depend on instrumental responding, we asked whether that experience affects judgments of persistence when relevant contingency information is manipulated. Subjects read simple scenarios with information about behaviors generated by high vs. low reward rate, and then judged the resultant persistence of these behaviors under no-reward conditions. Studies 1 and 2 found no evidence that persistence judgments were affected by contingency information in naive subjects. Studies 3 and 4 compared groups with and without explicit knowledge about behavioral psychology and thus tested possible effects of that knowledge for persistence more directly. Judgments in naive subjects were not reliably influenced by reward rate information, but subjects possessing expert knowledge demonstrated judgments that were reliably affected by contingency information. The results indicate that people do not generate generalized knowledge from normal experience with occasional vs. regular reward. Possible explanations and implications of these findings are discussed.