Association for Behavior Analysis International

Rats were given a choice between two running wheels. Responding on one wheel resulted in the delivery of a smaller, immediate reward (Impulsive) while responding on the other wheel resulted in the delivery of larger but delayed reward (Self-control). A variable inter-trial-interval was used to control the session duration regardless of the distribution of choices made by the subject. The results demonstrate that, with the wheel-running response, the number of self-control choices are significantly greater than that predicted by chance. The present results are inconsistent with previous studies using rats using a lever pressing response. Results are discussed in terms of general behavior systems approach and lower discounted value when compared to more proximate types of reinforcement in the lab (i.e. food reinforcer).

Economic games are useful tools for decision making studies and aim to analyze how participants allocate resources. In a within-subjects design, two experiments were conducted involving repeated-trials over20 opportunities in which 636 male and female undergraduates made choices to distribute resources between themselves and an unseen, passive other, either optimally but non-competitively, equally but non-optimally, or least optimally but competitively. It should be clear from the point of view of maximizing money earned, participants should always choose the option that gives them more money. The questions were presented in a paper and pen format and the experimental sessions run in a room at a North American university. The results showed some gender difference; men consistently behaved more optimally (in a strictly economic sense) than women. Amount of money showed difference in the results, raising the amount of money drove participants to be more optimal. These games are important because they analyze the contingencies involved in decision making; characterize the choices, such as ideal, fair or competitive, and bring under scrutiny the exam of the possible effects of other variables on the distributions of choices to determine whether these are stable or influenced by such variables.

Impulsivity has been conceptualized as the allocation of choice responding to small, immediate reinforcement at the expense of concurrently available larger, delayed reinforcement. Much of the applied research in this area has consisted of evaluations of choice responding within concurrent operants paradigms in which choices are presented between small, immediate reinforcers (e.g., 10 s activity/delay for 10 s break) and large, delayed reinforcers (e.g., 60 s activity/delay for 30 s break). During baseline conditions, individuals often allocate choice responding to the small, immediate reinforcement option thereby demonstrating impulsivity (thus, setting the stage for interventions aimed at teaching self-control). One hypothesis regarding impulsive choice responding is that it may be the more economically favorable choice based on the unit price associated with the small immediate and large delayed options (e.g., 1:1 vs. 2:1 in the above example). In the current study, we used concurrent operant arrangements in which the ratio of response/reinforcement options was systematically manipulated to favor either impulsive or self-control responding; or to be equivalent for both options. Results suggested that responding was controlled by ratios such that choice responding was allocated to the most economically favorable option in each choice arrangement. Thus, impulsivity and self-control were demonstrated respectively depending on which option was most favorable in terms of unit price.

The value of rewards is reduced when delivery of the reward is delayed (delay discounting), when delivery is uncertain (probability discounting) or when substantial amounts of effort are required to ensure delivery (effort discounting). There is evidence to suggest that delay, probability, and effort discounting are related. However, no study has examined all three measures in animals using a within-subjects design. We examined the interrelationships between delay, probability, and effort discounting in 30 rats. All subjects were exposed to delayed, probabilistic, and effort-requiring rewards using the adjusting amount procedure. This task determines the amount of immediate sucrose equivalent to the discounted value of a 150 l sucrose reward following a specific delay, with a particular probability, or requiring a certain amount of effort. Rats randomly rotated through the three types of cost, completing a minimum of 30 sessions in each condition (delay, probability, and effort). Positive correlations were found between all tasks, though only correlations between probability and effort reached significance. Hyperbolic and exponential curves used to fit the data accounted for similar amounts of variance. The results suggest that the three types of discounting share common basis, but that the limited overlap implies differences in the mechanisms underlying each type of reward devaluation.

Essential value, a measure of reinforcement, is presented as the reciprocal of a value in a behavioral-economic Equation LogQ=LogQo+k(e^-aQoC -l) (Hursh and Silberberg, 2008). The a value is capable of representing the essential value of a reinforcer in a range of contexts for animals in closed settings (Foster et al., 2009; Hursh & Silberberg, 2008; Christensen et al., 2008a, 2008b). We employ this equation to explain human consumption behavior among substitutable food brands, applying a consumer choice model the Behavioral Perspective Model (BPM, Foxall, 1990, 2005). The BPM denotes that utilitarian reinforcement (UR, functional benefit) and informational reinforcement (IR, symbolic benefit) determine consumers' choices. Using the ACNielsen HomescanTM panel, covering 10,000+ UK households in Great Britain, 52 weeks from July 2004 for four food products--baked beans, yellow fat, biscuits and fruit juice, we found that (a) essential value varies across different brand groups within the same product, (b) the higher informational levels of brands the larger the essential value of these brands, (c) the higher utilitarian levels of brands the larger the essential value of these brands, (d) the essential value of brands varies inversely with the degree of openness of consumer settings, and (e) informational and utilitarian benefits are influential in predicting the quantity of consumption.

The current study examined changes in demand elasticity as a function of session duration. Six Long Evans rats lever pressed under an open economy for either one- or two- pellets according to a fixed-ratio (FR) schedule. Each session lasted 6 hr, and the ratio requirement for food increased across sessions. Another lever was always available that provided 8-s of access to water according to an FR-1 schedule. Demand curves (i.e., the number of food pellets obtained under each FR value) were plotted for each successive hour, providing an hour-by-hour index of demand elasticity. Previous research suggests that elasticity for commodities available in open economies remains relatively high; however, the results from the current study suggest that longer open-economy sessions may produce demand curves comparable to those obtained under closed economies. Elasticity during the first hour of the session tended to remain quite high; however, as the session progressed, elasticity of demand decreased, ultimately approximating inelasticity noted for 24-hour closed economy sessions (cf. Hursh, et al., 1988).

A common measure of impulsivity that has been reliably associated with substance use disorders (SUDs) is delay discounting (DD). DD describes how a reward loses value as a function of increasing delay to its receipt. Various theoretical models of DD exist. Area under the curve (AUC) has been espoused as a theoretically neutral measure of DD with additional statistical advantages. The current analysis compared a distribution of hypothetical AUC values with a DD parameter (k) from a common model of DD. The results suggest that while the AUC measure may be a theoretically neutral measure of DD, it appears to be influenced by procedural variables in important ways. Additionally, the AUC measure appears to be a less sensitive measure of DD when compared to k at relatively low and high values of k. The current manuscript is not a condemnation of the AUC measure in DD research, but it does suggest that researchers should be cautious when utilizing the AUC measure under certain circumstances.

The majority of research on discounting has examined either temporal discounting or probability discounting. Few studies have studied effort discounting. Most notably Sugiwaka and Okouchi (2004) looked at how often Japanese students would be willing to wash a bathtub for 100,000 yen. They found that although the discounting of effort was well-represented by Mazer’s (1987) hyperbolic curve, effort discounting was not correlated with delayed discounting of money. In the current study we examined how students discounted effort in terms of reading passages of text for extra credit points in their general psychology course. Students were asked to make hypothetical linear choices about earning up to 1, 5, or 10 extra credit points for reading varying amounts of time; in addition, students were asked to read out loud to earn points in real time according to a linear or logarithmic reward scale. Results indicated that logarithmic real-time effort discounting was correlated with hypothetical discounting, but linear real-time effort discounting was not. Further, delayed discounting of money was not correlated with either real-time or hypothetical effort discounting. These results suggest that students may not discount real-time points in the same manner in which they discount hypothetical points.

Research outside of behavior analysis has yielded findings confirming that there are boundary conditions to humans' preference for extensive numbers of choice options. This area of inquiry has been termed "choice overload," suggesting that too much choice may be undesirable and could result in negative decision outcomes. To date, no behavioral account of the choice overload phenomenon has been described in the decision making literature. The current project sought to explain choice overload within a discounting framework. Specifically, we sampled approximately 100 undergraduate students using the Maximization Scale (a self-report Likert-type scale from the social psychology literature that captures decision making tendencies). From this sample, we identified the top (Maximizers; individuals who seek the most options when deliberating on a decision) and bottom (Satisficers; individuals who settle on "just enough" options to make a decision) quartiles. Both groups were then presented with three forms of a discounting assessment using a computer-based paradigm; delay, probability, and progressively increasing numbers of choice alternatives (i.e., choice overload). No differences in rates of discounting (k) were found in any forms of discounting between groups, calling into question whether self-report scales of choice overload are predictive of the behavioral aspect of deciding amongst extensive options.

Difficulties with "self-control" affect us all. Yet, because a large repertoire of other prosocial behaviors discounts the occasional lapse in judgment, the overall effect of impulsive choice-making is not devastating to the lives of many. For certain populations, however, deficient repertoires coupled with cognitive impairments prove to amplify the detrimental effects of occasional impulsivity. Individuals with disabilities may choose to engage in aberrant behaviors instead of concurrently available adaptive behaviors (Dixon & Holcomb, 2000; Myerson & Hale, 1984; Neef, Mace, & Shade, 1993; Neef, Marckel, Ferreri, Bicard, Endo, Aman, Miller, et al., 2005; Neef, Shade & Miller, 1994). The consistent theme is the inability to delay gratification (Mischel, 1972), that is, forego small, immediate reinforcers in favor of large, delayed reinforcers (Rachlin & Green, 1972). Methods for increasing self-control are many, but those that focus on gradually increasing preferences for large, yet, delayed reinforcers are well-supported in the basic (Ainslie, 1974; Fantino, 1966; Mazur & Logue, 1978; Navarick & Fantino, 1976; Rachlin, & Green, 1972) and applied literature (Dixon et al., 1998; Dixon & Holcomb; Dixon & Tibbetts, in press; Schweitzer & Sulzer-Azaroff, 1988). However, questions remain about the generalization of such treatment effects on overall impulse-control of individuals with disabilities. Therefore, the current study aimed to extend previous self-control research by providing explicit training to individuals with dual diagnoses and demonstrating the possibility for generalized self-control.

When students are making decisions regarding studying, there is often a cost-benefit analysis regarding how much they would be willing to study to earn a certain grade. Study-time-to-grade distributions most likely follow a standard learning curve, which is roughly logarithmic in shape, such that even a little studying will result in substantially increased grades, whereas it takes a great deal of studying to receive the highest possible grade. The current study examined this discounting of effort by students enrolled in a general psychology course. Students were asked to make decisions about how much studying they would engage in to receive a corresponding grade when the points are roughly logarithmically distributed and the total study time to receive the maximum grade was varied (1, 2, 3, 4, 7, 10, and 15 hrs). These results were then compared to their discounting of hypothetical extra credit points and delayed discounting of money ($100 or $1000 delayed from 1 day to 25 years). Results indicated that students discounted exam grades significantly less (as measured by area under the curve) than either extra credit or delayed money.

Attention-deficit hyperactivity disorder (ADHD) is a heterogeneous disorder with disruption in cognitive control and reward processing, including preferences for smaller, immediate rewards (SS) over greater, larger rewards (LL). We examined: (a) the relationship between delay discounting, academic achievement, IQ, working memory (WM) and ADHD ratings in 26 adolescents (14.6 years) and (b) the relationship between delay discounting and the index measure, k of self-control, and fMRI BOLD signal acquired during a cognitive control flanker task in 19 adolescents (14.6 years). Participants chose between SS and LL rewards with one choice randomly selected for actual reimbursement. The findings suggest greater self-control (lower k) is associated with better visual-spatial, math, attentional and WM functioning and lower ADHD ratings in adolescents. Brain activation during the performance of the cognitive control-flanker task for those with greater impulsivity (higher k), is associated with significant activation in motor, visual processing and response to reward (lenticulate) regions. Engaging these regions, rather than those associated with self-control and deliberation, (e.g.,prefrontal cortexor posterior parietal cortex) suggests more impulsive individuals are less likely to use top-down regions during cognitive tasks as well. Similarly, these regions are often significantly more active in individuals with ADHD, rather than higher order regions (dorsolateral prefrontal cortex).

Choices between immediate and delayed consequences have been studied through mathematical models. Several studies have investigated the delay discounting phenomenon with gains. Few have studied this phenomenon with losses. In previous research, data from 36 undergraduate students (aged 18 to 28, both sexes) in a series of hypothetical choices between immediate and delayed monetary losses (fixed delayed value of R$1,000.00 and a variable immediate value from R$1.00 to R$1,000.00) suggested that models found in the literature failed to describe loss delay discounting. An alternative model has been proposed in which the aversive value of the delay is modulated by two parameters added to the Hyperbolic-Exponent model. Although this model provided a better description of the obtained data, it is less parsimonious. This poster presents new analysis of the same data comparing the initial results obtained based on goodness of fit (R) criterion and conclusions based on Akaike Information Criterion (AIC). Pairwise comparisons revealed that the complex model proposed still can be considered more adequate that the Hyperbolic-Exponent model, but similar to Green's model. Implications of the use of both criteria are discussed.

Sunk cost effects (also known as'throwing good money after bad') bedevil decision makers making choices about risky ventures such as research and development (R&D) projects. A series of studies explored the effect of construct measurement and decision structure on the sunk cost effect. A single binary choice yielded a sunk cost effect; a single continuous measure did not. Repeated measurements revealed a weak sunk cost effect that fades over time. Measurement of the sunk cost effect changes whether or not it exists, which is explained by the economics of uncertainty.

The present paper expands upon the use of foraging theory as a general model for human decision making, and presents a replication of a previous study of price and delay equivalence (Smith & Hantula, 2003). Participants shopped for musiccompact discs (CDs) in a simulated Internet mall in a successive-choice procedure lasting five sessions. Five virtual music stores featuring incremental increases in mean CD prices from $9.99 to $17.99 were presented. Preference was measured as the proportion of total purchases allocated to each store. Consistent with previous research, a hyperbolic decay function provided the best fit to both aggregate and individual preference data. The results add to the evidence of price-delay equivalence and bolster existing evidence of the generality of hyperbolic discounting in human decision making.

In one variant of a matching-to-sample procedure a delay (or, "retention interval") is introduced between the offset of a sample stimulus (following the completion of an observing-response requirement) and the presentation of an array of comparison stimuli. In general, the accuracy of the performance decays as a function of increases in the retention-interval value. Another measure that appears to change in a direct relation with the duration of the retention interval is the latency to begin the sample-observing response requirement. We hypothesized that increases in the programmed retention interval decrease the value of the conditioned reinforcer (the onset of the comparison-array) via discounting processes. In an effort to test this interpretation, we arranged a series of conditions in which a varying number of 0-second dealyed matching-to-sample (DMTS) trials were randomly inserted in a block of 10-s DMTS trials allowing us to manipulate the average duration of the retention interval at a session-wide level. The data from one subject show a systematic decrease in sample latencies as a function of a decrease in the average retention interval value. These data suggest that behavioral economic processes (such as discounting) play a role in organizing in a wide variety of contexts.

The Sharing Game was created to study non-normative decision making using a non-zero sums economic game (Kennelly & Fantino, 2007). Earlier studies using the Sharing Game have yielded various gender differences, of those including males behaving more optimally and self-interested, and females behaving more equitably. This original study used monetary resources, meaning 'the more the better' to participants. In this study, 'time on a boring task' was allocated between participants, thus now meaning 'the more the worse' to participants. Participants chose one of five options, each one related to a social value orientation, which distributed some amount of minutes on the boring task between themselves and another participant. A Chi-Square Test was run to compare overall choice distributions between the two and proved the two to be different (Female X (4, N = 125) = 128.16, p < .001; Male X (4, N = 55) = 35.17, p < .01). An independent samples t-test revealed that males were more likely to choose the altruistic option (t(56.22) = 2.17, p = .04), whereas females were, once again, more likely to choose the egalitarian option (t(120.51) = -2.38, p = .02). A parametric analysis will be run in future studies to model choice behaviors among both genders.

Previous work in our lab trained rats on a dual-duration, peak-interval procedure in which one modal stimulus (e.g. tone) signaled reinforcement after 4-s, and another modal stimulus (e.g. light) signaled reinforcement after 12-s. When tested on non-reinforced compound-stimulus trials, rats showed a scalar peak of responding at a time in-between the trained durations when the tone signaled the short duration and the light signaled the long duration, but not vice versa. Based on research showing a modality effect for clock-speed (i.e., tones drive an internal clock faster than lights) that causes auditory stimuli to appear longer than visual stimuli (Wearden et al., 1998), the present experiment investigated modality effects on clock-speed as a potential explanation for the stimulus modality effects observed in temporal stimulus compounding. In an adjusting delay procedure, rats chose between a standard duration (1-s) providing one food pellet and an adjusting duration providing two. The adjusting delay increased by 25% with two successive adjusting choices during free choice trials. Results showed no effect of stimulus modality on the rate of delay discounting, though there was a trend towards steeper discounting when the standard and adjusting durations were accompanied by stimuli of different modalities rather than identical modalities.

Previous research has suggested that the presence of global uncertainty in choice between delayed outcomes can reduce an individual's sensitivity to delay. However, this result has only been shown using hypothetical rewards, and the question of whether the same effect would be observed in an experience-based paradigm remains unanswered. To this end, a first-person shooter video game was adapted to present participants with choices between outcomes that were both delayed and uncertain. Delayed outcomes took the form of shots from a weapon that 'charged' to increase its power in the time between firings. Uncertainty was incorporated by allowing for weapon misfires. Participants were not given exact figures for delay or probability but instead had to learn them from experience. Results were analyzed using linear mixed-effects modeling. Participants showed less discounting when impatience was more disadvantageous, and lower outcome probabilities produced greater discounting overall, but these two effects did not interact.

The economic game developed in this experiment simulated investments situations. The model employed in this study used multiple trials and a forced choice paradigm between two alternatives: the safe alternative and the investment alternative (risky). The safe alternative offered the possibility to maintain (and not to invest) the amount of hypothetical money received, situation in which the participant could not profit or lose money; and the risky alternative provided the opportunity to invest money, opting for a situation in which the participant could have a profit or loss. The risky alternatives combined several probabilities of profits and losses. Each block of questions was compounded by combinations between a specific probability of losses (10%, for instance) with different probabilities of profits. The experiment was carried out in a room at the local university and the questions were presented to 44 adults of both genders in a paper and pencil format. Preliminary results showed that in some situations the possibility of gaining nothing (safe alternative) became more aversive than the possibility of losing something (risky alternative).

An outcome will decrease in value as the delay to its receipt increases. This devaluation, or discounting, is very well established in the human literature with animal models now emerging as appropriate avenues of investigation. Cross-species (e.g., human and pigeon) similarities and differences in discounting have been revealed but are not yet fully understood. For a more thorough understanding of the factors influencing behavior, researchers may need to consider the biological relevance of the required response in addition to the economic context in which choices are made. In the present study, pigeons completed a delay discounting task where the nature of the response varied across conditions. One condition required a key-peck (biologically relevant response) and another condition required a treadle-press (non-biologically relevant response). The pigeons tended to discount much more steeply when key-pecking than when treadle-pressing. An important follow-up manipulation indicates that the differences in the rate of discounting are not due to the effort differential between the two responses, suggesting that the form of the required response may have a meaningful impact (e.g., more impulsivity is demonstrated with a biologically relevant response) and that a fuller understanding of discounting may require an integration of biological considerations.

The reinforcing effectiveness of some reinforcers may depend partly on their continuity through time (Hackenberg & Pietras, 2000). This experiment assessed whether continuous access, mediated through token delivery, produced results similar to those produced by distributed access (i.e., immediate but interrupted access), when their 'cost' (i.e. the number of responses required to produce reinforcement) was increased while the cost of a concurrently available alternative remained low and constant. The participant was a 15-year-old boy admitted to an inpatient facility for the assessment and treatment of severe problem behavior. Separate evaluations for food and non-food items were conducted. Results suggested that as response requirements for toys increased, accumulated access supported more responding than when access to the same toys was distributed throughout the session. This result was not observed when tokens exchangeable for food were evaluated. That is, distributed delivery of the food item supported more responding than its accumulated delivery.

The temporal discounting literature demonstrates that the loss in value for delayed monetary outcomes is best described using a hyperbolic discounting equation. Hyperbolic shaped functions not only describe the shape of discounting for monetary outcomes, but also for a variety of other tangibles outcomes. The 4hyperbolic shape of these functions has important theoretical implications. Despite this broad list, there is one class of outcomes that has not been addressed: delayed social interactions. In the current study 93 undergraduate psychology students completed a standard binary choice temporal discounting procedure. For each choice, participants chose between an immediate face-to-face social interaction lasting between 0 and 100 minutes or a delayed (between 6 hours and 1 year), longer social interaction (100 minutes). Results suggest that the same equations (exponential, hyperbolic, and hyperboloid) that describe temporal discounting for non-social commodities also describe the relationship between delay and the perceived value of delayed social interactions and that, like non-social outcomes, the hyperbolic shaped models are superior to the exponential model. The similarity in function suggests that a similar decision process is used for evaluating both tangible and non-tangible commodities.

A behavioral economic analysis was conducted on genetically engineered mice lacking dopamine D2 receptors (DA D2Rs) to assess the role of these receptors in food reinforcement. The demand curve (consumption vs. FR value) generated from responding under various fixed-ratio schedules decreased more steeply in DA D2R knockout (KO) than in heterozygous (HET) or wild-type (WT) mice indicating a role of the DA D2R in food reinforcement. Pre-feeding decreased, whereas extinction increased overall response rates as a proportion of baseline similarly in all mice. The absence of an effect of genotype indicates a lack of involvement of DA D2Rs in satiation and extinction. Both the DA D2R antagonist, (-)-eticlopride, and its (+)-enantiomer dose-dependently decreased responding in all genotypes with (-)-eticlopride ~30-fold more potent than (+)-eticlopride in all but KO mice. The enantiomers were equipotent in KO mice, and similar in potency to (+)-eticlopride in WT and HET mice. Differences between (-)-eticlopride effects and extinction indicate that, despite previous suggestions, pharmacological DA D2R blockade does not mimic extinction. The maintenance of responding in KO mice indicates that the DA D2R is not necessary for reinforcement. However, the economic analysis indicates that the DA D2R contributes substantially to the effectiveness of food reinforcement.

A first-person shooter video game was adapted for the study of impulsive choice to identify state variables that can affect the trade-off between smaller sooner and larger later outcomes. College students chose when to fire a weapon that increased in damage potential over a 10 s interval, an escalating interest situation. Across two experiments, participants demonstrated sensitivity to the nature of the mathematical function that defined the relationship between waiting and damage potential. In Experiment 1, people tended to wait longer when doing so allowed them to eliminate targets in a more timely fashion. In Experiment 2, people tended to wait longer to increase the likelihood of a constant magnitude of damage than to increase the magnitude of a certain amount of damage that was matched for the same expected value. The two experiments demonstrated sensitivity to the way in which an outcome improves when the outcome is continuously available. The results also demonstrate that this new video game task is useful for generating sensitivity to delay to reinforcement over time scales that are typically used in nonhuman animal studies.

For almost 4 decades, intertemporal choice tasks in animals have produced consistent, highly-reproducible behavior. These results are thought to imply that many species are highly impatient, with hyperbolic discounting curves that result in rewards losing half their value when delayed by only seconds. However, the link between intertemporal choice tasks and impulsivity as an ethological and psychological construct has gone largely unquestioned. Here, we review evidence from our own work, as well as from studies of foraging and behavioral economics, suggesting that standard intertemporal choice paradigms may lack robustness as measures of intrinsic impulsivity, or even of discounted value. We conjecture that animals' behavior may be better understood as maximizing rates of reward subject to cognitive constraints, and that the structure of intertemporal choice tasks may primarily reflect myopia in the algorithm by which animals estimate these rates, thereby inflating measures of discounting. We offer alternative measures of discounting, and specific tests for these biases. Finally, we will discuss the implications of these ideas for neuroscience and for addiction, as well as for human studies of temporal discounting.

Cocaine-dependent individuals (N = 38) participated in a study designed to investigate the potential role of delay discounting in the high rates of sexual HIV risk behavior observed in cocaine dependence. Participants completed a traditional hypothetical money delay-discounting task in addition to a novel Sexual Discounting Task assessing choices between immediate unprotected sex and delayed sex with a condom across four hypothetical partners (most likely to have a sexually transmitted disease (STD), least likely to have an STD, most want to have sex with, least want to have sex with). Delays to sex with a condom ranged from 1 hour to 3 months. The reported likelihood of using a condom decreased as a function of delay across all conditions, and a hyperbolic equation well described this function. Hypothetical money discounting and sexual discounting were generally correlated. Sexual discounting was significantly greater when making responses for partners judged least (compared to most) likely to have an STD, as well as for partners with whom participants indicated they most (compared to least) wanted to have sex. Results suggest that sexual HIV risk behavior and its prevention may be understood in terms of delay discounting.

Social norms regulate and constrain individual behaviors to establish shared expectations within a social group. People with psychiatric diseases often exhibit impaired social norm perception and experience difficulties in interpersonal interactions. Although studies have suggested that social norm perception is pliable, the mechanism by which norm perception changes and the underlying neural substrates are not clear. We took a quantitative approach to manipulating fairness norms using the Ultimatum Game. We trained the players to adjust to a new norm by repeatedly exposing them to unfair, fair or hyper-fair conditions. We found that initial exposure to unfair or hyper-fair conditions affected behavior in the fair condition, and that those subjects who played the unfair condition first had greater activation in the ventromedial prefrontal cortex (VMPFC) and the nucleus accumbens. Additionally, training led to a reduction in activity in the bilateral putamen and the anterior insula, suggesting habituation to the new norm. We also found significant correlations between the activity in the bilateral anterior insula and deviations from a new norm.

Delay discounting describes how the value of an outcome decreases as its receipt is delayed over time. Such values are well described by a hyperbolic discounting model. A large number of studies have been conducted with both human and nonhuman subjects; however, our understanding of discounting in non-human primates is limited. One limiting factor is that quantifying delay discounting is time consuming, requiring multiple conditions run for several sessions each. The current study assessed discounting of sucrose pellets by rhesus macaques, utilizing two procedural manipulations aimed at decreasing time to stability in each condition. A within-session adjusting delay procedure was utilized, in which subjects chose between 1 pellet delivered immediately versus a larger reinforcer delivered after an adjusting delay, with the larger amount altered across conditions. In one version,five subjects began each condition with a low adjusting delay (10s). Remaining subjects began each condition at the delay predicted based on the hyperbolic model with parameter k = 0.05. Resulting data were well described by the hyperbolic model, and subjects showed moderate decreases in time to stability using predicted equivalent values as starting delays. Also discussed are the minimum number of different conditions required to achieve stability in derived k values.

Food choice directly impacts caloric intake, which has been shown to be a contributing factor to obesity. Previous research has suggested operant choice arrangements as an objective evaluator of the relative reinforcing value of food versus alternative reinforcers. The current study assessed differences in frequency of responding for food and non-food items by presenting obese and non-obese college students with concurrent VR schedules of reinforcement via a computer-based program. Participants (N = 16) were first given a practice session in which they experienced each schedule; they were then instructed to respond to either schedule throughout the experimental session. Activity reinforcers were available on a VR2 schedule throughout the study and food reinforcers were available on increasing VR schedules, requiring more response effort to procure food reinforcers as the session progressed. General trends across all participants suggest more responding was allocated to the denser (activity) schedule of reinforcement. Consistent with previous research, obese participants appeared to defend consumption of snack points as the VR schedule for snacks increased before switching preference to activity. Contrary to previously published research, non-obese participants earned more food points overall than their obese counterparts. Current efforts to examine previous methodologies and findings via a component analysis are underway.

The present study investigated the roles of work requirement (cost) and reward amount (benefit) in determining response allocation across three concurrently available schedules of reinforcement. The economic concept of unit price suggests that the ratio of costs to benefits determines consumption of a given commodity. We examined the utility of unit price to account for college students response allocation using a computer game that contained three distinct response options. Sessions began with each option having the same unit price but with different work requirements and reinforcer magnitudes. Unit Price was escalated within a session by increasing the work requirement or by decreasing the reinforcer magnitude for a response option following each schedule completion. That is, in some conditions the pay off for a schedule (i.e., amount of money or duration of the video) remained constant while the response requirement increased, while in other conditions reward amount decreased while response requirement remains constant. We compared response allocation in this arrangement using two qualitatively different reinforcers (points exchangeable for money and video). Results are discussed in terms of deviations from optimal responding and implications for the utility of unit price when using non-monetary reinforcers.