Association for Behavior Analysis International

Behavioral history effects are those of prior experience on subsequent current behavior. This study investigated history effects of reinforcement schedules correlated with high or low reinforcement rates on stimulus control in later contingencies using chicks. In the first phase, four chicks were exposed to two tandem variable-interval differential-reinforcement-of-low-rate schedules under different stimulus conditions. The values of DRL schedules were adjusted so that reinforcement rates in one stimulus condition were higher than those in the other, even though response rates in the two conditions were nearly identical. In the second phase, a fixed-interval schedule was introduced to each chick under the same stimulus conditions as those in the first phase. In the final phase, chicks' key-peck responses were extinguished under the same stimulus conditions as those in the previous phases. All chicks showed similar response rates in each phase. Figure 1 showed the typical results of two chicks. The effects of reinforcement rates under two stimulus conditions on later stimulus control were not clear. The chicks' key-peck response rates seem to be affected not by the prior reinforcement rates but by contingencies of reinforcement in each phase. Thus, the reinforcement rates may not be an important variable for behavioral history effects.

Recent studies have found that individuals with autism spectrum disorder (ASD) show robust impairments in behavioral flexibility. The current study aims to better understand how decreased 5-HT6 receptor activation may lead to an alleviation of the reversal learning impairments found in the BTBR mouse model of ASD. Control C57BL/6J and BTBR mice were tested on three separate behavioral measures including repetitive grooming, locomotor activity, and probabilistic reversal learning. We predicted that the 5-HT6 receptor antagonist BGC 20-761 would reduce repetitive grooming and alleviate the probabilistic reversal learning impairment found in these mice, while having no overall effect on locomotor activity. For the probabilistic reversal learning task, mice were tested on acquisition then reversal learning in a T-maze using an 80/20 probabilistic reinforcement procedure. The high dose of 2.5 mg/kg BGC 20-761 reduced repetitive grooming in BTBR mice compared to controls, but also decreased locomotor activity. The low dose of 0.25 mg/kg BGC 20-761 also reduced repetitive grooming but did not similarly reduce locomotor activity. BGC 20-761 treatment similarly improved probabilistic reversal learning in the BTBR mice. These findings highlight how 5-HT6 receptor down regulation may aid in attenuating lower order repetitive motor behaviors and higher order behavioral rigidity.

Cooperative responding has been investigated under continuous reinforcement (CRF or FR 1). The present study investigated the effects of fixed-ratio (FR) and variable-ratio (VR) schedules of reinforcement on the patterns of cooperative responding in seven pairs of rats. Experiment 1 arranged FR 1, FR 10, and VR 10 schedules for cooperation (order was counterbalanced). The contingency required the responding of both rats within a short time interval (< 500 ms). Cooperative responding was higher under intermittent schedules than under continuous reinforcement. FR 10 generated a break-and-run pattern, while VR 10 schedule generated a constant responding pattern. Experiment 2 compared responding under a cooperative intermittent schedule (FR 6), and under a yoked schedule programming the same reinforcement rate, without a requirement for joint responding of two rats for reinforcement. Coordinated behavior was maintained at higher rates under social intermittent contingencies than under yoked contingencies. Therefore, cooperative responding in rats was established and maintained under intermittent schedules of reinforcement, the patterns of responding were similar to those obtained for a single organism, and the removal of the contingency for coordinated responding did not maintain cooperation. These results extend previous findings on the conditions that foster coordinated behavior in non-human subjects. Keywords: cooperation, schedules of reinforcement, fixed-ratio, variable-ratio, water, rats

Most studies on food accumulation by rats involve two different response classes: earn and collect. However, in experiments done in our laboratory accumulated food has only been delayed from the earn response. By contrast the present study compared the effects of delayed reinforcement on both, earn and collect responses. For the earn condition three rats each were exposed to a chained FI 30 s FT t FR1 while for the collect condition three other rats were each exposed to a chained FI 30 s FR1 FT t. For both conditions t was either 0, 2, 4, 8, 16, or 32. Results showed that lengthening reinforcement delay in both conditions increased the rate of the earn responses although more markedly when food was delayed from the earn response than when it was delayed from the collect response. These results suggest that delaying reinforcement on the earn responses, as done in studies from our laboratory, yield similar functions than when reinforcement is delayed from the collect responses, a fact not hitherto known.

The author investigated interrelationship between attacker pigeons' extinction-induced attack (EIA) and their target pigeons' pecking towards the attackers by comparing inter-response times (IRTs) per opportunity of respective behavior. Three of seven pigeons were attackers and the remaining four were their targets. All attackers were paired with each of thFe targets. Each attacker was introduced into an operant chamber with one key and the target was introduced into a box adjacent to the operant chamber via transparent panel. When the attacker emitted EIA or the target pecked towards the attacker, this panel was operated. Due to this operation, attackers' EIA and the target's pecking were measured. Each attacker's key pecking was exposed to continuous reinforcement and extinction schedules during the respective pairing. The figure shows distributions of IRTs per opportunity in each pairing of an attacker (#2) and the four targets (#12, #24, #31, #82) and correlation coefficient r between their behaviors' IRTs per opportunity. The distributions of IRTs per opportunity of the attacker and the targets were very similar and all correlation coefficients were high. As other attackers' pairs showed comparable results, the author found that attackers' EIA and targets' pecking towards the attackers are related to each other.

fAn organism's behavior is shaped through interactions with the environment or an agent acting as a "shaper." Herrnstein (1964) demonstrated that a teacher-pigeon could shape a student-pigeon's behavior so that both could obtain food. This experiment replicated this demonstration using a pigeon as a teacher to a second pigeon, designated the student. Each pigeon faced a separate panel divided by a transparent wall. The panels included a food hopper for each pigeon, and a response key was located on the teacher's side of the wall. Food was delivered to both pigeons when two criteria were met: 1) the student was standing in a specific location designated by an orange rectangle on the floor and 2) the teacher pecked the key. When the orange rectangle moved around the chamber across sessions, the student tracked the location. The teacher showed some evidence of differentially responding when the student was closer to the designated location, although this was not as clear of an effect. Herrnstein, R. (1964). Will. Proceedings of the American Philosophical Society, 108(6), 455-458. Retrieved from http://www.jstor.org/stable/985862

An experimentally used pigeon 2748 was placed in a holding cage. Once in the cage 2748 would exhibit an escape behavior. The holding cage is commonly used in the lab and no prior learning of the escape behavior or shaping. The pigeon doesn't receive a food reinforcer upon escaping. The reinforcer in the experiment is the opportunity to escape. After conducting baselines on 2748, we received an average escape time. Next, we placed a mirror in front of 2748's holding cage, on average 2748 significantly escaped faster. This ruled out the possibility of social facilitation. We then conducted a baseline on the naïve pigeon. The pigeon did not exhibit the escape behavior. We plan to place twonaïve pigeons across from each other. We are looking to see if the one pigeon will imitate the other's behavior and escape. We do not expect that to occur since the pigeon has not previously escaped. Once all baselines are conducted we will place thenaïve pigeon across from 2748. We expect 2748 will continue to escape in the presence of thenaïve pigeon. After repeated exposure, thenaïve pigeon will imitate 2748 and escape, in the presence of the 2748 and on his own.

Studies suggest that social enrichment speeds up many learning processes, including habituation to novel and aversive stimuli. However, the effect of social enrichment on habituation to positive reinforcers has not been examined. The current investigation compared within-session responses of socially enriched (SE; n = 4) and socially isolated (SI; n = 4) male Wistar rats on (1) a VI-7.5 s schedule of reinforcement and (2) an open-field apparatus. Repeated measures analyses of variance were used to analyze proportions of operant responses and open-field activity made in successive 2-minute bins. Both SE [F(14, 42) = 22.96, p < .001, ?2p = .88] and SI [F(14, 42) = 19.40, p < .001, ?2p = .87] subjects exhibited significant within-session decreases in operant responses rates. However, while SE subjects exhibited decreased locomotion in the open field [F(6, 18) = 6.23, p = .001, ?2p = .68], SI subjects did not [F(6, 18) = 1.89, p = .14, ?2p = .39]. Further research will examine brain and adrenal gland weights relative to body weight in the two groups. Results from this investigation will aid in interpreting future data collected from socially housed subjects, which may better generalize to real world applications.

The foundational research in operant behavior exclusively employed electro-mechanical relay equipment to arrange contingencies of reinforcement. Variable interval schedules were programmed using a punched-tape reader that arranged a fixed series of repeating inter-reinforcer intervals. This repetition possibly undermines the construction of prediction-proof reinforcement schedules; however, it is unknown whether this feature affected behavior. The present study was conducted to determine whether a repeated loop of inter-reinforcer intervals would exert control over key pecking in pigeons. In the initial four phases of the experiment, pigeons pecked on either a VI 180-s or a VI 15-s schedule of either 15 or 5 total inter-reinforcer intervals generated using the Catania and Reynolds (1968) algorithm. A list of intervals were repeated each session. Analyses were conducted to discover evidence of behavior's sensitivity to the repetitive nature of the variable interval schedule. No such signatures were discovered, suggesting that the repetitive nature of the inter-reinforcer intervals of variable interval schedules used in the classic pigeon studies likely did not influence their outcomes. The components of the current condition, which mimic that of a two-component mixed fixed interval, could help to establish whether or not temporal learning can be accomplished with the current subjects and apparatus.

The effects of delay of reinforcement on resistance to change and resurgence were assessed in an experiment with four rats. In the Training phase, left-lever pressing was maintained on a two-component multiple schedule. In the immediate component, the schedule was a tandem variable-interval (VI) differential reinforcement of low rates (DRL); in the delay component, the schedule was a tandem VI fixed time (FT). Conditions were programmed so that both reinforcement and response rates, variables known to affect resurgence, were similar between components in this phase. In the Elimination phase, in both components, left-lever pressing was extinguished and right-lever pressing was maintained under equal VI schedules. In the Test phase, right-lever pressing also was extinguished in both components. The resistance to change of left-lever pressing was similar between components and, in the Test phase, resurgence did not occur in each component, for each rat. These are results of an ongoing experiment in which both delay of reinforcement and Training-phase response rates are being manipulated. The results obtained so far indicate the importance of considering response rates in the Training phase in analyses of the effects of delay of reinforcement on both resistance to change and response recurrence.

The present experiment investigated the effect of reinforcement delay on resistance to extinction. Four rats were exposed to a two-component multiple schedule. In the immediate component, a tandem VI DRL schedule was in effect; in the delay component, a tandem VI FT 3-s schedule was in effect. In baseline, reinforcement rates were similar between components; for two rats, response rates were similar between the components and, for the other two, response rates higher in the immediate component than in the delayed component. For three of four rats, resistance to extinction was greater in the delay than in the immediate component, regardless of the difference in baseline-response rate between components. However, the difference in resistance to extinction between components was greater when response rates differed between components in baseline. These results indicate that reinforcement delay affects resistance to change and that this relation can be modulated by the baseline rate of responding. In addition, the present results fail to replicate those of previous studies with pigeons, in which greater resistance to change in the immediate than in the delay component has been reported.

Numerous studies have demonstrated the utility of functional communication training (FCT) in the treatment of challenging behavior. Ringdahl et al. (2009) evaluated the role of proficiency across mand topographies with regard to the effectiveness of FCT. Ringdahl et al. found that FCT was more effective when higher proficiency mand topographies were targeted during FCT in the treatment of challenging behavior. The purpose of the current study was to examine mand proficiency on resurgence of challenging behavior within FCT. We conducted a mand proficiency assessment (MPA) with children who engaged in challenging behavior. Based on the results of the MPA, FCT was implemented with high and low-proficiency mand topographies; subsequently, we tested for resurgence (i.e extinction of mands and challenging behavior) across the two mand topographies. Results showed similar patterns of resurgence of challenging behavior in both high and low proficiency mand topographies; however greater persistence in high proficiency mands was observed during extinction phases relative to low proficiency mands.

In one previous study on the recurrence of observing responses, both primary and secondary reinforcers were removed during the resurgence test. The current procedure addressed this limitation by only withholding the conditioned reinforcer during the resurgence test. Responses on the central key were reinforced according to a mixed VI EXT schedule. During the training phase, the target and alternative responses were pecking on side keys reinforced by presenting the availability of reinforcement on a central key. In the alternative reinforcement phase, responses on one of the side keys no longer produced the color change on the center key. When the alternative response was no longer reinforced, there was an increase in responding on the target key. By the end of the resurgence test, responding on the center key increased. Thus, this experiment demonstrated the resurgence of responses maintained by secondary reinforcement without removing primary reinforcers.

The present study examined the effects of reinforcement probability of target and alternative responses on resurgence in humans. In Phase 1, a target response was reinforced. In Phase 2, the target response was placed on extinction and an alternative response was reinforced. In Phase 3, all responses did not produce reinforcers. Across all phases, two control responses were recorded but had no programmed consequences. The probabilities of reinforcement for the target and alternative responses in Phases 1 and 2 were varied among 0.2, 0.4, and 0.8 across conditions. All possible combinations of these probabilities yielded nine conditions and each participant was exposed to one of nine conditions. Although both the target and control responses increased in Phase 3 relative to Phase 2, resurgence of the target response exceeded the control response in eight of nine conditions. This result suggests the possibility that resurgence of the target response can be distinguished from extinction-induced variability. With respect to the relation between probability of reinforcement and resurgence, the reinforcement frequency for the target response had no effect on the magnitude of resurgence. On the other hand, the magnitude of resurgence was greater when the probability of reinforcement for the alternative response was high.

We investigated the impact of response effort on variability. Participants worked on a computer task moving a light around a five by five matrix. Participants moved a light from the upper left to the lower right position. The "A" key shifted the light down. Across components, the "S," "G," and "L" key moved the light rightward. Each time the light reached the target position, a point was earned. During baseline, any sequence of 4 presses on each key produced a point. Participants were instructed to earn as many points possible during each component. Once baseline responding was established, a lag five schedule was implemented to evaluate the effects of key distance on sensitivity to differential reinforcement of variability. The results showed that increasing the distance between keys reduced the number of novel sequences emitted and reduced the number of key switches per sequence. The lag schedule increased the number of different sequences emitted during the session, but those effects were not altered by key distance. The results indicate that response effort, here defined by key distance, may reduce spontaneous variability in response topography. Baseline variability, however, does not seem to limit the effects lag schedules to engender further variability.

Several studies have shown that reinforcement schedules systematically affect the dimensions of responses not directly reinforced under the schedules. For example, the spatial locations of pigeons' responses have been investigated under a fixed-interval (FI) schedule. The results show that the response location moves closer to that of the most recent reinforced response over time, and that the variability of the response location increases as an FI requirement. The present study compares the effects of ratio and interval schedules on pigeons' response locations. Pigeons are exposed to multiple fixed-ratio yoked FI and multiple variable-ratio yoked variable-interval schedules. A circular response area (22.5 cm in diameter) is used to ensure that pecking responses are effective over a wide range. The results indicate that the response locations vary between the schedules in both cases. In addition, the mean deviation, defined as the mean distance between the median of the coordinate and the location of each response, is higher for interval schedules than it is for ratio schedules. Therefore, the results show that interval schedules produce greater variability than ratio schedules do in terms of response location.

Zebrafish are used as an animal model in biomedical studies including vision research. However, behavioral contrast sensitivity, a basic measure of visual function, of zebrafish has not been measured. In the present experiment, zebrafish were trained to discriminate between a sinusoidal grating pattern and plain gray having equivalent luminance. The visual stimuli were presented by an organic light-emitting diode display that has high contrast ratio. The experimental chamber had one observing gate and two entrance gates through which subjects could see the discriminative stimuli. Trials were initiated and discriminative stimuli were presented when a fish passed through an observing gate. Fish were reinforced by an automated food delivery system only when it entered the gate where sinusoidal grating pattern was presented. If fish entered the gate where plain gray was presented, a timeout was implemented. Contrast of sinusoidal grating pattern was decreased until fish cannot detect the difference between a sinusoidal grating pattern and plain gray having equivalent luminance. Zebrafish showed high-level performance in preliminary discrimination between the sinusoidal grating pattern and plain gray having lower luminance. In discrimination between sinusoidal grating pattern and plain gray having equivalent luminance, the accuracy was not so robust but approached statistical significance.