This study compared the effects of (+/-)3,4-methylenedioxymethamphetamine, d-amphetamine, and cocaine on performance of rats in a delayed matching-to-sample procedure using a variety of indices of performance to determine the mechanism by which working memory task impairments arise. All 3 drugs produced an overall delay-independent decrease in accuracy rather than a delay-dependent increase in the rate of forgetting. This impairment arose as a result of current-trial choice responses being progressively more affected by responses made in the immediately preceding trial as drug dose increased. Therefore, all 3 drugs produced qualitatively similar disruptions in memory task performance best characterized as an impairment arising from proactive sources of interference.

The authors devised a nose-poke task with asymmetric position-reward mapping to distinguish between effects of bias and sensitivity in reaction times of rats. In all trials, the rats had to poke their noses into the hole to the left or to the right of center, corresponding to the side at which 4 lights were illuminated, while ignoring distracters on the other side. Reaction times were faster for large-reward trials than for small-reward trials. In large-reward trials, there was no influence of the number of distracters, whereas in small-reward trials, distracters produced an increase in reaction time. Analysis of reaction-time distributions according to a linear model of decision making suggests that most of the systematic variability was due to a reward-oriented bias.

Three dual-task experiments were conducted to investigate the relation between immediate, "on-line" judgments about visual features and delayed, "off-line" judgments. One hypothesis ("selective interference") predicted that dual-task performance would be challenged specifically within a visual dimension, as both tasks compete for the same resources. Another hypothesis ("cost of switching") made the opposite prediction. In Experiment 1, participants performed either color or shape discriminations in the on-line and off-line visual tasks, with systematic variation of feature similarity between the on-line and off-line features. In Experiment 2, participants performed either color or shape discriminations in the off-line task and color discriminations in the on-line task, with no overlap between the on-line and off-line features. In Experiment 3, participants performed color discriminations in both the on-line and off-line tasks, with partially overlapping stimulus sets. Altogether, the data from the three experiments provided evidence in favor of the hypothesis of cost of switching. Stimulus-stimulus compatibility effects between features in the off-line task and those in the on-line task further underscored the perceptual nature of the crosstalk.

Previous research has established that dopamine signals are crucial in orienting behavior to reward. Less is known, however, about the psychopharmacology of task performance under small-reward conditions as compared to large-reward conditions. The current study examined the effects of the noncompetitive N-methyl-D-aspartate (NMDA)-receptor antagonist dizocilpine (MK-801) on reaction time (RT) in a nose-poke task with rats completing an asymmetric reward schedule. In all trials, the rats were required to poke their nose in either the left or the right peripheral hole immediately adjacent to the centre hole when the corresponding light was illuminated. Depending on the stimulus-reward mapping, however, one position was associated with a large reward, while the alternative position was associated with a small reward. Correct performance was required in every trial; if the rat did not make a correct response within 20 s, the trial was aborted, and the same stimulus was presented again on the next trial. In this way, the rat was forced to perform the same visuo-spatial discrimination task under different reward conditions. Reaction times (ms) were faster for large-reward trials than for small-reward trials, replicating previous findings. At a dosage of MK-801 (0.04 mg/kg), there was no significant influence of on RT in large-reward trials. In contrast, the same dosage of MK-801 in small-reward trials produced a decrease in RT as compared to the control condition, implying an improvement of performance. Below 0.04 mg/kg of MK-801, a steady decrease of RT in small-trials was seen as a function of dosage. Above 0.04 mg/ kg of MK-801, the majority of rats failed to perform the task at all, whereas the rats that did manage to perform the criterion of 80 correct trials in a session showed no difference in RT between largeand small-reward trials. These data indicate that the systemic administration of a relatively small dosage of MK-801 facilitates performance when reward is small. It is suggested that the facilitation may be due to the reinforcement of mechanisms that work in opposition to response bias. As a corollary, the study provides a useful paradigm to study the voluntary control of unavoidable action. FindingsResponses are faster and more accurate with a large reward than with a small reward [1]. It is thought that predictive signals of dopamine neurons are crucial in orienting behavior to reward [2]. For instance, dopamine input to dorsal striatum may induce a response bias, with a high a priori likelihood to choose a response associated with a large reward [3,4]. However, many situations impose a mandatory requirement in favor of a less-desirable option. Recent evidence showed that thalamic neurons in