Journal of the experimental analysis of behavior最新文献

My collaboration with Peter Urcuioli started with research on delayed matching to sample. Initially we asked, what do pigeons remember during the delay in delayed matching to sample: a retrospective coding of the sample or a prospective coding of the comparison-related response? This led us to examine the basis of the differential outcomes effect. Why are samples associated with differential outcomes learned faster and remembered better than samples associated with common outcomes? This research helped us discover a procedure that resulted in functional stimulus equivalence: Samples associated with the same comparison are commonly associated. This research led Peter to develop his creative model of pigeon equivalence class formation. His model predicts the conditions under which pigeons satisfy the three components of what is known as Sidman equivalence: reflexivity, symmetry, and transitivity, phenomena that are difficult to demonstrate in pigeons. Importantly, his model predicts the opposite of reflexivity (anti-reflexivity) and symmetry (anti-symmetry). Research confirming Peter's model laid to rest the belief that the emergent relations defining Sidman equivalence can be satisfied only by an organism capable of using language. In his long career, Peter Urcuioli has made an important and long-lasting contribution to the field of learning and comparative cognition.

Dogs show numerousness, which is the ability to identify the larger of two stimuli, most often the number of treats on a plate. However, dogs seem to use mechanisms other than counting to make this discrimination. This study builds on existing research by controlling for (a) olfaction, (b) the surface area of the stimuli, and (c) delivery of a single reinforcer contingent on correct choices in the trained task. Nine dogs were trained to select a dish with 5 dots/treats in a sealed container when presented with comparison stimuli of 1, 4, and 9 dots/treats. The dogs showed numerousness in discriminating between dishes with 1 versus 5 dots, with consistent performance of more than 80% correct. Two dogs could discriminate 4 versus 5 dots, and three dogs learned the 9- versus 5-dot discrimination in which there is a conflict between selecting the larger option and selecting the 5 dots to gain reinforcement in the task. Knowledge of numerical competency, particularly training dogs to select the nonreinforced choice, can offer strategies to facilitate cognitive enrichment and learning in animals or offer enhancement of the capabilities of working dogs where the concept of number might be advantageous in providing additional skills.

A recent survey of the evidence on associative symmetry in humans revealed that nearly all the demonstrations either unintentionally trained backward stimulus pairings and/or had a temporal overlap between the stimuli being trained. We consider these criticisms and improve on our own method of “associative networks.” In this method, participants learn multiple stimulus pairings via arbitrary matching-to-sample tasks in which the stimuli are concurrently presented as sample and comparison stimuli. In Experiment 1, human participants learned a bidirectional network (in which symmetry was synergistic) and a unidirectional network (in which symmetry was antagonistic) or two unidirectional networks (removing explicit reinforcement of backward stimulus pairings). In Experiment 2, participants learned two unidirectional networks; however, we removed the temporal overlap between sample and comparison stimuli by imposing a 1-s delay between them. Both experiments showed robust evidence of symmetry, suggesting that the expression of symmetry in humans survives the most common confounds in published research.

In real-world settings, stimulus and outcome associations often depend on situational factors, such as Pavlovian occasion setters (OSs), which disambiguate whether a conditional stimulus (CS) will predict an outcome (unconditional stimulus; US). Whereas previous studies show that OSs are often lower in salience than CSs, no study has examined how low-salience OSs affect learning. In two conditioning experiments, we investigated this from the premise that inconsistently reinforced CSs prompt searching for additional stimuli (OSs) that indicate whether the CS will be followed by the US. Occasion setting learning was assessed using extinction rate—as partial reinforcement slows extinction relative to continuous reinforcement—and self-reported latent learning of stimuli. We hypothesized that a high-salience OS would result in faster extinction rates and occasion setting learning, whereas a low-salience OS would result in slower extinction rates and CS partial reinforcement learning. The results of Experiment 1 were mixed; there was no effect of OS salience on extinction rate, but the results for latent learning supported the hypothesis. We conducted Experiment 2 to specifically test extinction rate, and the results supported our hypothesis. The findings suggest that if a salient OS is found, occasion setting is learned; otherwise, CS partial reinforcement is learned.

This experiment was an investigation how reinforcers for one response in the presence of one stimulus may generalize to other dimensionally related stimuli. Four pigeons were trained on a three-alternative concurrent variable-interval schedule in which, after an initial condition, extinction was arranged for one alternative. In Part 1, we varied the reinforcer rate on a dimensionally distant alternative while keeping the reinforcer rate on the dimensionally closer stimulus constant. In Part 2, the reinforcer rate for the distant alternative was kept constant and that for the closer alternative was varied. Increasing the reinforcer rate for the closer alternative increased responding on the extinction alternative, but increasing the reinforcer rate on the distant alternative decreased extinction response rates. This result is predicted by the generalization across dimensions model. This model also helps to explain the results from previously reported choice research that involves multiple alternatives, and particularly why Luce's indifference principle is sometimes supported and sometimes not.

How the subjective value of an outcome changes as a function of time, probability, or effort has been an active area of psychological and economic research for decades. The exact functional form of how a commodity is discounted has been debated, and there have been numerous forms proposed. One of the challenges when trying to determine the functional form of discounting data is how models are compared, what modeling methods are used, how many data points are used, and what comparison metrics were used. Thus, we sought to replicate and extend previous research comparing discounting model selection methods by simulating discounting data from five functional forms: the Mazur hyperbolic model (Mazur, 1987), Rachlin hyperboloid (Rachlin, 2006), Myerson–Green hyperboloid (Myerson & Green, 1995), Samuelson exponential model (Samuelson, 1937), and beta-delta model (Laibson, 1997). With each of these models we manipulated the number (i.e., density) of data points, used two forms of modeling, and assessed the degree to which each model generalizes to data it has not used in the fitting process. Model comparisons were conducted using the Akaike information criterion (AIC), Bayesian information criterion (BIC), and leave-one-out cross validation (LOOCV). In general, AIC, BIC, and LOOCV selected the correct model, whereas the Rachlin model had the lowest error across folds of LOOCV when relying on multilevel modeling.

Resurgence refers to the recurrence of a previously reduced or eliminated operant following the worsening of reinforcement conditions for an alternative behavior. The contextual account of resurgence posits that the discriminative properties of reinforcers modulate this relapse. Evidence supporting the contextual account of resurgence stems from translational laboratory investigations with nonhuman subjects. The purpose of the current study was to further translate this work using a human-operant preparation with a young boy with autism and intellectual disability across three experiments. Experiment 1 was conducted according to a conventional three-phase resurgence preparation. In Experiment 2, the reinforcer (O2) that maintained the alternative response (R2) in Phase 2 was delivered noncontingently in Phase 3. Finally, in Experiment 3 the reinforcer (O1) that maintained the target response (R1) in Phase 1 was delivered noncontingently in Phase 3. Although both O1 and O2 mitigated resurgence relative to extinction alone, the magnitude of resurgence was slightly higher with O2  deliveries. Our results are generally consistent with the contextual account of resurgence. The influence of a differential-reinforcement-of-other-behavior contingency in Phase 3 and other procedural aspects of this study are discussed, as are applied implications for reducing relapse in behavioral treatments.

Our behavioral units include stimulus classes and response classes. Peter Urcuioli's differential-outcomes research implies they should extend to the third term of the three-term contingency. Classes of consequences come in several varieties (e.g., conditional reinforcers, tokens), but our vocabulary does not coherently organize them. They are differentiated not only by physical properties such as type, location, and duration but also by the schedule contingencies in which they participate. We consider units ranging from the physical and chemical sciences to those based on the particular history of life on earth. The latter include biology, sociology, linguistics, and our own behavior analysis. Scientific units are typically nested (e.g., atoms within molecules, cells within organs, organisms within species). Comparing our units with those from other taxonomies raises questions about their emergence and evolution and their shared properties across levels of nesting (e.g., species within genus, subclasses within higher order operants, phonemes within words). Emergence necessarily occurs when higher order units have functions that are not shared with their lower order constituents. These nested and multileveled behavior classes challenge single-level views, such as metaphorical accounts of behavior as a totality contained within a pie, with slices corresponding to behavior classes matched to their outcomes.

We developed five novel quantitative models of punishment based on the generalized matching law (GML). Two of the new models were based on Deluty's additive theory of punishment, two were based on de Villiers's subtractive theory of punishment, and the last was based on the concatenated GML (cGML). Using information criteria, we compared the descriptive accuracies of these models against each other and against the GML. To obtain a data set that fairly compared these complex models, we exposed 30 human participants to 36 concurrent random-interval random-interval reinforcement schedules via a recently developed rapid-acquisition operant procedure (procedure for rapidly establishing steady-state behavior). This experimental design allowed us to fit the models to 30 data sets ranging from 22 to 36 data points each, comparing the models' descriptive accuracy using Akaike information criteria, corrected for small samples (AICc). The punishment model based on the cGML had the lowest AICc value of the set, with an Akaike weight of 0.99. Thus, this cGML-based punishment model is presumed to be the best contemporary quantitative model of punishment. We discuss the theoretical strengths and weaknesses of these models and future directions of GML-based punishment model development.

Resurgence is defined as an increase in a previously reinforced and reduced target response when conditions worsen for a more recently reinforced alternative response. The present experiment evaluated the effects of target- and alternative-reinforcer rate on resurgence in humans. We arranged combinations of contingent high- and low-rate target and high- and low-rate alternative reinforcement across four groups. When extinguishing all reinforcement, resurgence occurred only in groups experiencing a high rate of alternative reinforcement, with resurgence being higher in the group that experienced high rates of both target and alternative reinforcement. Our findings join those of others that indicate that rates of alternative reinforcement contribute more to resurgence than target-reinforcer rates. A quantitative model of resurgence, resurgence as choice in context (RaC²), tended to underpredict target responding and overpredict alternative responding in Phase 3. Including a misallocation parameter to account for a proportion of reinforcers misallocated between responses provided a better account of the findings.