Manish Vaidya

Concurrent performance in a three-alternative choice situation: Response allocation in a Rock/Paper/Scissors game ☆

Adult human subjects engaged in a simulated Rock/Paper/Scissors game against a computer opponent. The computer opponents responses were determined by programmed probabilities that differed across 10 blocks of 100 trials each. Response allocation in Experiment 1 was well described by a modified version of the generalized matching equation, with undermatching observed in all subjects. To assess the effects of instructions on response allocation, accurate probability-related information on how the computer was programmed to respond was provided to subjects in Experiment 2. Five of 6 subjects played the counter response of the computers dominant programmed response near-exclusively (e.g., subjects played paper almost exclusively if the probability of rock was high), resulting in minor overmatching, and higher reinforcement rates relative to Experiment 1. On the whole, the study shows that the generalized matching law provides a good description of complex human choice in a gaming context, and illustrates a promising set of laboratory methods and analytic techniques that capture important features of human choice outside the laboratory.

Trends in presentations at the annual conference of the Association for Behavior Analysis

The present report analyzes trends in attendance and presentations at the annual conference of the Association for Behavior Analysis (ABA). Numbers of registered attendees were plotted over time. The trends show that the number of registered attendees has grown considerably over the last three decades, with the largest proportion of the growth occurring during the last 10 years. This growth is shown to be correlated with the introduction of board certification in behavior analysis (BCBA and BCABA). In addition, conference programs from 1980 through 2007 were coded, and all presentations were categorized into one of four areas (application, basic research, conceptual, and verbal behavior) based on the primary designator codes chosen by the authors at the time of submission. An analysis of the total number of presentations in each category indicates that applied research presentations have always outnumbered the other three categories. The absolute number of presentations related to application has grown faster than presentations in other categories. However, correcting for population growth shows that the relative proportion of presentations in the four areas has remained fairly constant over the last 28 years.

Effects of Stimulus Discriminability on the Acquisition of Conditional Discriminations in Adult Humans

The current study investigated the effects of the discriminability of sample and comparison stimuli on the acquisition of conditional discriminations in adult humans. In an arbitrary matching-to-sample procedure, five university students were trained on four types of conditional discriminations between simple (one-element) and complex (two-element) stimuli. The matching-to-sample task involved simple-simple, simple-complex, complex-simple, and complex-complex conditional discriminations, where the first term designates the type of sample and the second term designates the type of comparison stimuli. The effects of the discriminability of sample stimuli on the acquisition of conditional discriminations did not differ greatly from those of the comparison stimuli. Nevertheless, the results are in general agreement with prior findings showing that, in a matching-to-sample procedure, the rate of acquisition of conditional discriminations is a function of the discriminability of the sample and comparison stimuli and that the former is a more important variable with respect to the rate of acquisition than the latter.

Effects of select and reject control on equivalence class formation and transfer of function.

The present study used a single-subject design to evaluate the effects of select or reject control on equivalence class formation and transfer of function. Adults were exposed to a matching-to-sample task with observing requirements (MTS-OR) in order to bias the establishment of sample/S+ (select) or sample/S- (reject) relations. In Experiment 1, four sets of baseline conditional relations were taught-two under reject control (A1B2C1, A2B1C2) and two under select control (D1E1F1, D2E2F2). Participants were tested for transitivity, symmetry, equivalence and reflexivity. They also learned a simple discrimination involving one of the stimuli from the equivalence classes and were tested for the transfer of the discriminative function. In general, participants performed with high accuracy on all equivalence-related probes as well as the transfer of function probes under select control. Under reject control, participants had high scores only on the symmetry test; transfer of function was attributed to stimuli programmed as S-. In Experiment 2, the equivalence class under reject control was expanded to four members (A1B2C1D2; A2B1C2D1). Participants had high scores only on symmetry and on transitivity and equivalence tests involving two nodes. Transfer of function was extended to the programmed S- added to each class. Results from both experiments suggest that select and reject controls might differently affect the formation of equivalence classes and the transfer of stimulus functions.

A neurobiology of learning beyond the declarative non-declarative distinction.

In neuroscience and psychology it has become common to distinguish between declarative and non-declarative forms of learning (e.g., Squire and Zola, 1996). The perspective (Squire, 2004, p. 173) that declarative learning “refers to the capacity for conscious recollection about facts and events and is the kind of memory that is impaired in amnesia and dependent on structures in the medial temporal lobe” developed when it was observed that patients with hippocampal damage were able to learn and improve on “procedural” tasks, such as hand-eye coordination, without being able to remember the specific learning episodes (e.g., Scoville and Milner, 1957; Squire, 1992). Such evidence led researchers (e.g., Moscovitch, 1995) to suggest that there are two separate learning systems: one involved in declarative, conscious, learning situated in the Medial Temporal Lobe (MTL) comprising the hippocampus, and another involved in non-declarative, implicit, skill/habit learning involving the basal ganglia. An interpretation of the roles of the hippocampus and the basal ganglia in learning independent from the declarative/non-declarative dichotomy may allow understanding the role of neural structures critical in learning without relying on phenomenological categories—difficult to examine from a scientific perspective—such as awareness and consciousness. The very nature of what consciousness is in fact still hotly debated (e.g., Morin, 2006) and grounding an entire taxonomy of learning and memory on consciousness-based criteria might complicate, instead of simplifying, scientific interpretations, and progress. An alternative perspective to the declarative—non-declarative distinction (e.g., Packard and McGaugh, 1992; McDonald and White, 1993) emphasizes the role of the hippocampus in learning stimulus-stimulus relations, and of the dorsal striatum in acquiring stimulus-response relations. Such perspective is consistent with a recent review (Henke, 2010) suggesting that neurobiological models of learning based on consciousness might not be adequate to describe the available experimental evidence.

Is Stimulus-Stimulus Pairing Sufficient to Produce Conditional Discriminations?

The purpose of the present study was to investigate the sufficiency of stimulus-stimulus pairing in establishing conditional discriminations in adult humans. The task involved learning 32 different conditional discriminations. During training, a correct pair of sample and comparison stimuli was presented after the subjects made a sample-observing response but before (prompt) or after (feedback) the subjects made a comparison-selection response. During testing, the same conditional relations were presented without prompting or feedback to assess the development of conditional relations among the stimuli. Despite the fact that accuracy on prompt trials was quite high from the beginning of training, accuracy on prompt trials dropped substantially during the test conditions for three out of four subjects. By contrast, accuracy on the feedback trials was well maintained under test conditions. These results suggest that, in this preparation, the contemporaneous presentation of experimenter-designated correct sample and comparison stimuli is not sufficient and that some other factors may have played an important role in the establishment of conditional discriminations.

Conditional discriminations, symmetry, and semantic priming.

Psychologists interested in the study of symbolic behavior have found that people are faster at reporting that two words are related to one another than they are in reporting that two words are not related - an effect called semantic priming. This phenomenon has largely been documented in the context of natural languages using real words as stimuli. The current study asked whether laboratory-generated stimulus-stimulus relations established between arbitrary geometrical shapes would also show the semantic priming effect. Participants learned six conditional relations using a one-to-many training structure (A1-B1, A1-C1, A1-D1, A2-B2, A2-C2, A2-D2) and demonstrated, via accurate performance on tests of derived symmetry, that the trained stimulus functions had become reversible. In a lexical decision task, subjects also demonstrated a priming effect as they displayed faster reaction times to target stimuli when the prime and target came from the same trained or derived conditional relations, compared to the condition in which the prime and target came from different trained or derived conditional relations. These data suggest that laboratory-generated equivalence relations may serve as useful analogues of symbolic behavior. However, the fact that conditional relations training and symmetry alone were sufficient to produce the effect suggests that semantic priming like effects may be the byproduct of simpler stimulus-stimulus relations.

A Comparison of Prompting Versus Feedback in the Development of Conditional Discriminations in Adult Humans

The purposes of the current study were to compare the relative effectiveness of procedures emphasizing prompting versus feedback in the development of conditional discriminations, and to investigate whether prompting and feedback interact with stimulus discriminability on their effects on the development of conditional discriminations. The task involved learning 32 different conditional discriminations. For half of these conditional discriminations, an experimenter-designated correct pair of sample and comparison stimuli was presented immediately after the observing response but before the choice phase of the trial (prompting). For the other half of the conditional discriminations, an experimenter-designated correct pair of sample and comparison stimuli was presented after a participant selected a comparison stimulus (feedback). In addition, effects of discriminability of sample and comparison stimuli were investigated by presenting combinations of simple (one-element) and complex (three-element) sample and comparison stimuli. Test trials, run without programmed consequences, presented the 32 conditional discriminations trials without prompting or feedback. All participants learned the conditional discriminations through non-differential prompting and feedback, but the feedback condition produced slightly more accurate performance on the test trials. Stimulus discriminability affected the development of conditional discriminations, but the effects of prompting and feedback on the development of conditional discriminations were not altered by the discriminability of sample and comparison stimuli and vice versa.

Effects of stimulus meaningfulness on the acquisition of conditional relations and the emergence of symmetrical relations in adult humans

ABSTRACT The goal of the present study was to investigate the effects of stimulus meaningfulness of sample and comparison stimuli on the acquisition of conditional relations and the emergence of symmetrical relations. Eight college students were trained on two types of baseline conditional discrimination tasks involving meaningful (English words) and non-meaningful (Swahili words) stimuli. One task involved English samples and Swahili comparisons (English–Swahili), whereas the other task involved Swahili samples and English comparisons (Swahili–English). Test trials, without programmed consequences, presented the symmetry probe of the baseline relations trained. That is, the English–Swahili baseline relations presented on the training trials were presented as the Swahili–English symmetrical relations on the test trials. Similarly, the Swahili–English baseline relations on the training trials were presented as the English–Swahili symmetrical relations on the test trials. The results show that the English–Swahili baseline relations were acquired faster than the Swahili–English baseline relations. The results also show that, relative to the acquisition of the baseline relations, the emergence of symmetrical relations was facilitated and delayed, respectively, on the English–Swahili and Swahili–English test trials. The present results suggest that stimulus meaningfulness of untrained successive discrimination of samples and simultaneous discrimination of comparisons on test trials affects the emergence of symmetrical relations, with the former playing a more important role than the latter.

Effects of heterogeneous controlling relations on tests of transitivity and equivalence: An exploratory study

ABSTRACT The establishment of sample/S+ (select) or sample/S− (reject) controlling relations affects equivalence-class formation. The present experiment aimed to evaluate transitivity and equivalence test results when, during training, some of the baseline relations among stimuli presented during tests were established under reject control and others were established under select control. Five participants had learned in a previous study to respond to conditional discriminations AB, BC, and CD under reject control and to DE and EF relations under select control. In the present investigation, such baseline-trained relations were reviewed and, after that, (1) one-node CE-transitivity and EC-equivalence tests, (2) two-node BE-transitivity and two-node EB-equivalence tests, and (3) four-node AF-transitivity and four-node FA-equivalence tests were carried out. Results suggest that during transitivity and equivalence tests involving one stimulus set related under reject and another under select control (CE and EC tests), participants scored high. Increasing the number of nodes involved during tests, however, resulted in variability. The possibility that the commonly observed failures of equivalence class formation may be the result of heterogeneous controlling relations among the stimuli comprising the set or putative equivalence class is discussed.