Association for Behavior Analysis International

Transfer of stimulus control is an essential feature of many acquisition procedures, such as in errorless learning, fading, treatments for prompt dependence, and skill learning. Learning new behavioral skills often involves the transfer of stimulus control from discriminative stimuli present during early training to new endogenous cues that gain stimulus control with extended practice, often leading to skill autonomy. Four studies examined how behavioral skills may become autonomous by developing reliable practice cues when the originally predictive cues were modified or eliminated. As exteroceptive cues were degraded and eliminated over four successive phases using multiple schedules, pigeons solved the discrimination problem by transferring stimulus control to other available exteroceptive and endogenous cues. By measuring and comparing conditional discriminations in signaled vs. unsignaled components, we quantified the degree of stimulus control and measured changes in discriminative control across sessions. The final phase eliminated all discriminative stimuli. Pigeons achieved autonomy by ingeniously developing a new behavioral skill that could always yield reinforcement, but not via the transfer of stimulus control observed in Phases 1-3. Transfer of stimulus control to other available exteroceptive and endogenous cues is one process that may lead to skill autonomy, but it’s not the only one.

This study compared operant behavior on fixed ratio (FR) and random ratio (RR) food reinforcement schedules in enriched and non-enriched female Sprague-Dawley rats. The enriched rats were given post-session access to enrichment objects and other rats; non-enriched rats were kept in standard housing. Demand functions were generated for all subjects under both FR and RR schedules, with schedule order counterbalanced across subjects. Rats from the enriched group consistently responded at higher rates and at higher prices than did rats from the non-enriched group. Additionally, non-enriched rats showed the typical schedule difference, with higher levels of responding on RR than comparable FR schedules. Such differences were absent for enriched rats. For enriched rats, there was correlation between response rates on the FR schedule and social dominance in the enrichment setting. The results both join with other findings on the facilitative behavioral effects of enrichment, and add a much-needed level of quantitative rigor to the analysis of enrichment effects.

Obesity is one of the most prevalent behavioral disorders in the United States affecting approximately 40% of the adult population. Nearly 93.3 million individuals suffer from aversive clinical issues comorbid with obesity, including heart disease, diabetes, and certain types of cancer (Centers for Disease Control and Prevention, 2018). Currently, the role of motivating operations (MO), or the stimulus control that alters the value of a commodity, in obesity is unknown. Further, present technology used to explore genetic and behavioral interactions in obesity remains subpar. Fatty Zucker rats are widely used to model and test genetic obesity as they are Leptin-deficient; Leptin being a chemical mediator of energy balance and food intake suppressor. In this study, a reversal design was used to analyze the effects of deprivation on food motivated responding using 6 fatty and lean Zucker rats. Motivation was assessed via lever-pressing demand tasks following baseline (deprivation) and experimental condition (satiation). Data analysis suggests elevated food motivation in the fatty rats across many work requirements despite food satiation. This is significant as it addresses the role of MOs in genetic obesity and provides potential behavioral targets to address satiation and obesity.

Prior research has shown that responding can be maintained under concurrent food and social reinforcement in rats, but little is known about interactions between these reinforcers. In the present study, we approached the problem from a behavioral economic perspective, using demand-curve methods to analyze interactions between food and social reinforcement. Four rats were given repeated choices between food and 10-s of social access to a familiar rat on concurrent schedules. Social access was arranged by lifting a door to a restraint, within which the partner rat was held. The price of social access was held constant at fixed ratio (FR) 1 across all conditions, while the price of food was systematically increased from FR 1 to FR 64. Of interest was cross-price elasticity, or demand for social reinforcement as a function of changes in the price of food reinforcement. Food responding was maintained at lower to moderate prices but declined to low levels at higher prices. Social responding was relatively constant at the lower to moderate food prices but increased when food responding dropped at the higher FR food prices, suggesting a substitutable relationship. The methods show promise as a way to quantify interactions between qualitatively different reinforcers.

Some problematic human behavior occurs when a single choice produces consequences that include both reinforcing and aversive outcomes. The present experiment explored this type of choice situation with rats by investigating how the value of an immediate food reinforcer followed by a shock changes as a function of the delay to the shock. The rats chose between two food pellets delivered immediately and followed by delayed shock, and two food pellets delivered alone after a delay. Within each condition, the delay to food was adjusted based on the rats’ previous choices until both consequences were chosen equally often and the delay to food was stable. At this “indifference point,” the delayed food was equal in value to the immediate food followed by shock. Across conditions, the delay to shock was manipulated. Generally, the shock devalued the immediate food to the greatest extent when the delay to shock was short, and the effects of shock weakened as the delay was raised in a pattern resembling a hyperbola. The finding that effects of aversive events on choice are weakened by delay parallels research on temporal discounting of positive reinforcers.

We describe the implementation of five different measures of behavior: directionality, vigor, preference, variation and persistence to analyze and represent the effect of food delivery according to Fixed-Ratio and Fixed-Interval schedules of reinforcement. The subjects were 4 rats, divided in two groups. We used an enlarged experimental chamber of 92 cm x 92 cm with three levers and one food dispenser. Responses on any lever produced reinforcers according to the current schedule. In different phases, the location of levers (on the same or different walls) and schedule of reinforcement (Fixed Ratio-1 or Fixed Interval-30s) varied. We found that the five measures of behavior were sensitive to the current schedule of reinforcement and that changes in some of those dimensions did not correlate with changes in other measures of behavior. The importance of analyzing different dimensions of behavior and thinking of new forms of representing behavioral data will be discussed.

Sexual behavior and feeding have been described separately without specifying the interaction of both under the same contingencies. It has been affirmed that the analogous behavioral repertoires described in both behaviors could be to the manipulation of the same variables. The purpose is to analyze sexual and feeding behavior through the manipulation of reinforcement parameters: magnitude, delay and novel sexual stimulation. Four experiments were designed that are logically concatenated. In Experiment 1, the effect of a novel sexual stimulation on sexual behavior and feeding will be described. In Experiment 2, the effect of the deprivation of sexual activity will be analyzed according to the receptivity of the females. Experiment 3 will describe sexual behavior based on Lee's Body Mass Index and Experiment 4 will compare the copulatory efficacy of standard males versus a common male rat. The design of the experiments follows the Mixed Factorial Design, fully covering the sources of internal disability and obtaining external validity, show effect and interaction of the manipulated variables. The contribution consists in showing the generality of the principles of the Experimental Analysis of Behavior to the case of sexual behavior and its interaction with food, two basic behavioral repertoires common to all organisms.

The empirical evidence about the use of different reinforcers still does not clearly show its replaceability. In the present study the effect of four reinforcers on feeding behavior in rats is described throughout three independent experiments. In Experiment 1 rats were food deprived and rolled pellets were delivered. In Experiment 2 the rats were water deprived and 0.10ml of water was delivered. For both experiments a CRF on a lever was used to feed rats during one-hour of experimental sessions. In Experiment 3 rats were food-deprived and industrial pellets or tapioca was delivered according to a VI 10, 20 or 30 seconds on successive experimental phases. The temporal distribution for all the reinforcers used was similar to the typical patterns for food commonly used in experimental chambers. Response rate was higher for industrial pellets, followed by tapioca, rolled pellets and water, in this order. These findings suggest that it is feasible to replace the traditionally used industrial pellets by cheaper ones to generate evidence comparable with both, operant and motivation literature. It is analyzed the usefulness of cheaper reinforcers with different palatability.

Behavioral variability has been proposed as a dimension of operant behavior that is vulnerable to environmental manipulations. For example, it has been demonstrated that variability of 4-lever press sequences can be affected by dopamine drugs such as amphetamines and SKF-38393. In a previous experiment, we tested the effects of over-the-counter energy drinks on variability of rats sequential responding. Three rats were trained on a procedure similar to that of Neuringer (1991) and Pesek et al. (2011) where sequences of four level presses were reinforced if they were novel compared to the past eight response sequences, referred to as a lag 8 condition. While the OTC energy drink Red Bull did not appear to have any discernible effects on U (entropy – the variability of emitted sequences), a control component from the experiment was missing. In this experiment, we will directly replicate those of Pesek et al. (2011) using a multiple schedule which includes a control condition. In addition, we will provide doses of energy drinks determined by body surface rather than weight, as indicated by current literature.

Intracytoplasmic sperm injection (ICSI) is an in vitro fertilization (IVF) method that involves the direct delivery of spermatozoon into the ooplasm of an egg. Since its introduction in 1992, ICSI has grown in popularity and is now the most commonly used IVF method worldwide. Despite this, some concerns have arisen regarding the procedure, namely that it circumvents natural sperm selection at the level of conception and could therefore lead to adverse postnatal outcomes. These concerns have led researchers to begin investigating ICSI outcomes in both humans and mouse models. This poster describes the first stage of an interdisciplinary collaboration in which we investigated operant learning in ICSI mice. First generation ICSI mice (n = 26) and naturally-conceived controls (n = 28) were compared across a battery of operant conditioning procedures: nose poke acquisition, a discrimination task, a do-not-match-to-position (DNMTP) recognition memory task, and retention assessments to measure the extent to which DNMTP performance was maintained across longer periods of time. The main finding was differences between ICSI and same-sex controls in nose poke acquisition and the discrimination task performance. We consider the implications of these initial findings for further behavioral research with ICSI mice.

The availability heuristic (reliance on recent events or information for decision making) is a systematic bias in human cognition. Few studies have been conducted to determine if the availability heuristic impacts risky decisions or if animals utilize such heuristics. This experiment sought to determine if recent events influenced risk sensitivity in rats. Sprague Dawley rats were randomly assigned to a Win and Loss Group. Both groups chose between a constant option that delivered 2 100% sugar pellets following a 3 s delay (p = 1.0)and a variable option that delivered 2 pellets 100% sugar pellets after a 1 s or 5 s delay (p = .5). In the Win Group (n = 10) subjects experienced a run of good luck or short delays prior to choices while the Loss Group (n = 10) experience a run of bad luck or long delays prior to choice. No statistically significant difference in choice was observed between groups (see Figure 1). These findings indicate that recent events or the availability heuristic do not drive variable or risky decision making. Rather, data is best described by the sequential choice model, which predicts that animals make choices in a chronological manner as they encounter different choice options.

This research has the purpose to investigate the influence of the use, chronic and acute, of caffeine in insight learning process on albino Wistar rats. The research was divided on six stages: 1) Discriminative training to signalize reinforcer availability; 2) an open field test; 3) a pre-test with each animal in the experimental cage; Then the animals will be divided into three groups, a control group (GC), an acute-use caffeine group (GCA) and a chronic-use caffeine group (GCC), each one with four animals; 4) consists on an independently training of dig and climbs repertoire; 5) individuals of each group were displayed to a puzzle that needed previous learned repertoire to be solved; 6) they were placed on the open field teste for the second time. Results shown that in the training stage all group had similar results, although GCC showed a better result on both repertoires then GCA. On GCC group all subjects were able to solve the problem, but in the GC and GCA only two animals solved the problem. Overall the acute-use of caffeine may produce a better performance on problem solving situations and it also can affect some process of learning.

Operant responding may habituate (decreased responsiveness over time) or sensitize (increased responsiveness over time) to repeatedly-presented reinforcers (McSweeney & Murphy, 2014). These within-session changes have been studied as they relate to broader behaviors such as addiction and food intake. 16 male Wistar rats were classified as either sign-trackers (n = 8) or goal-trackers (n = 8) in a Pavlovian conditioning task and then trained on a variable interval 7.5-s schedule of reinforcement. Groups did not differ in habituation, sensitization, or response rates when food restricted. Free-feeding decreased responding similarly in the two groups and did not affect habituation or sensitization. Additionally, because of previously documented increases in sensitization caused by aversive stimuli, 6 subjects were tested the day of a magnitude 7.1 earthquake. Within-session patterns of responding differed significantly between the day of the earthquake and the previous day F(29, 145) = 2.23, p = .001, with larger increases and smaller decreases observed after the earthquake. Taken together the current results are consistent with previous findings that (1) within-session decreases in operant responding are attributable to habituation rather than satiety, and (2) aversive stimulation can increase responsiveness to unrelated stimuli, including positive reinforcers.

Delay discounting is sometimes discussed as a trait variable – a pre-existing characteristic people bring to a situation. A trait approach suggests an association between discounting across outcomes characteristics (e.g., amount, commodity, and gain/loss). A trait approach would predict related, but not necessarily equivalent, discounting across characteristics. In contrast, a specificity approach suggests people do not possess an underlying discounting trait and there should be little-to-no relationship between discounting across outcome characteristics. In this study, we used traditional quantitative analyses and machine learning to describe data obtained from 23 cocaine- and 24 never-using individuals. Participants completed 16 total discounting tasks spanning: gains and losses; money, health, and cocaine; and, amounts of $10, $100, and $1000. When traditional models incorporate independent parameters for amount, commodity, gain/loss, and discount rate, we found that a single discount parameter described discounting well only across amounts, but not across the other domains (commodity, gain/loss). In contrast, machine learning suggested a single discounting parameter could describe discounting well across all domains.

In probability discounting research, comparing real and hypothetical outcomes has mixed results. Furthermore, it seems necessary a test that asserts the understanding of the probability concept before the discounting tasks. The aim of this study was to compare the probability discounting (i.e., choose for probabilistic bigger rewards vs. certain smaller rewards) with real and hypothetical outcomes. Thirty-two teenagers (12-15 years old) participated. First, a condition for training the probability and certainty concept was conducted. The probability discounting tasks followed an adjusting amount procedure. Four trials were used in five levels of probability: 10%, 25%, 50%, 75% and 90%. The standard amount was 16. The real outcomes were programmed in a video-game. In one option the smaller certain reward was delivered (i.e., medals and trophies). In the other option, a tombola was presented, and the reward was delivered or not depend on the probability programmed. The hypothetical probabilities and rewards were programmed in a traditional probability discounting task. The order presentation of the probabilities and each task were randomized. In both tasks, higher adjustment was observed with hyperboloid (R2> .80) than hyperbolic model. No differences were found between tasks. Video-games could facilitate the identification of risk-taking in a target population.

It has been found that people with obesity tend to choose smaller-immediate amounts of money instead of greater-delayed ones, which reflects impulsive instead of self-controlled behavior according to the delay-discounting procedure. In previous studies conducted in the laboratory of the authors, the relation between the participants’ weight and their delay-discounting rates for different kinds of consumable rewards was shown. Globally, teenagers with higher body mass index (BMI) showed more self-controlled behavior, measured with delay-discounting rates, for soda and fast food than for water and healthy food, respectively. However, different results were found with body fat percentage (BFP). The purpose of the present study was to evaluate the generality of the delay-discounting findings of our previous studies, with adults and with both health indices, BMI and BFP. Sixty-one adults between 18 and 50 years old were exposed to a delay-discounting task, which tested five rewards that varied in its calorie level: water, soda, healthy food, fast food, and money as a control. The amounts of the rewards were varied according to an adjusting-amount procedure. The results were like those of our previous studies; that is, adults with higher weight showed more self-controlled behavior, measured with delay-discounting rates, for soda and fast food than for water, healthy food, and even money. These results were similar for both, BMI and BFP and contribute to the study of self-control as a state behavior, which changes according to features of both rewards and participants.

Progressive ratio (PR) schedules are often used to identify the value of a reinforcer, which is often defined by the point at which an organism stops responding. This point, the breakpoint, however, can be impacted by procedural considerations such as the starting response requirement, the algorithm used for increasing the response requirement, the rate of reinforcement, the criteria for identifying the breakpoint, and the context in which the PR schedule is arranged. The current presentation uses PR schedules to assesses the effectiveness of two putative reinforcers for completing a ratio requirement under varying contexts. Manipulations such as the response required to access the reinforcer and the use of a progress bar are focused on primarily and are assessed through measuring the total responses emitted and the total ratios completed per reinforcer. Results suggest that contextual variables such as the progress bar and response type impacts measurement of the reinforcer value.