Category learning in the color-word contingency learning paradigm

In the typical color-word contingency learning paradigm, participants respond to the print color of words where each word is presented most often in one color. Learning is indicated by faster and more accurate responses when a word is presented in its usual color, relative to another color. To eliminate the possibility that this effect is driven exclusively by the familiarity of item-specific word-color pairings, we examine whether contingency learning effects can be observed also when colors are related to categories of words rather than to individual words. To this end, the reported experiments used three categories of words (animals, verbs, and professions) that were each predictive of one color. Importantly, each individual word was presented only once, thus eliminating individual color-word contingencies. Nevertheless, for the first time, a category-based contingency effect was observed, with faster and more accurate responses when a category item was presented in the color in which most of the other items of that category were presented. This finding helps to constrain episodic learning models and sets the stage for new research on category-based contingency learning.     

Contiguity and contingency in action-effect learning

According to the two-stage model of voluntary action, the ability to perform voluntary action is acquired in two sequential steps. Firstly, associations are acquired between representations of movements and of the effects that frequently follow them. Secondly, the anticipation or perception of an acquired action effect primes the movement that has been learnt to produce this effect; the acquired action-effect associations thus mediate the selection of actions that are most appropriate to achieve an intended action goal. If action-effect learning has an associative basis, it should be influenced by factors that are known to affect instrumental learning, such as the temporal contiguity and the probabilistic contingency of movement and effect. In two experiments, the contiguity or the contingency between key presses and subsequent tones was manipulated in various ways. As expected, both factors affected the acquisition of action-effect relations as assessed by the potency of action effects to prime the corresponding action in a later behavioral test. In particular, evidence of action-effect associations was obtained only if the effect of the action was delayed for no more than 1 s, if the effect appeared more often in the presence than in the absence of the action, or if action and effect were entirely uncorrelated but the effect appeared very often. These findings support the assumption that the control of voluntary actions is based on action-effect representations that are acquired by associative learning mechanisms.     

Generalization decrement in human contingency learning

An association between a cue and an outcome will generalize to a similar novel cue to some extent, but not completely. Learning theorists refer to the discrepancy between responding elicited by the original cue and the novel cue as a generalization decrement. Two experiments used a contingency learning task with human participants to compare the size of a generalization decrement between configurations of cues that were altered by adding or subtracting compositional elements. The results suggest that adding elements to a configuration can produce a generalization decrement, but removing elements produces a more robust generalization decrement. Furthermore, the generalization decrement caused by adding elements was not likely to be caused by competing orienting responses. The results are used to contrast Pearce's (1987, 1994) and Wagner's (2003) models of stimulus generalization.     

Generalization decrement in human contingency learning

An association between a cue and an outcome will generalize to a similar novel cue to some extent, but not completely. Learning theorists refer to the discrepancy between responding elicited by the original cue and the novel cue as a generalization decrement. Two experiments used a contingency learning task with human participants to compare the size of a generalization decrement between configurations of cues that were altered by adding or subtracting compositional elements. The results suggest that adding elements to a configuration can produce a generalization decrement, but removing elements produces a more robust generalization decrement. Furthermore, the generalization decrement caused by adding elements was not likely to be caused by competing orienting responses. The results are used to contrast Pearce's (1987, 1994) and Wagner's (2003) models of stimulus generalization.     

Abstraction and model evaluation in category learning

Thirty previously published data sets, from seminal category learning tasks, are reanalyzed using the varying abstraction model (VAM). Unlike a prototype-versus-exemplar analysis, which focuses on extreme levels of abstraction only, a VAM analysis also considers the possibility of partial abstraction. Whereas most data sets support no abstraction when only the extreme possibilities are considered, we show that evidence for abstraction can be provided using the broader view on abstraction provided by the VAM. The present results generalize earlier demonstrations of partial abstraction (Vanpaemel & Storms, 2008), in which only a small number of data sets was analyzed. Following the dominant modus operandi in category learning research, Vanpaemel and Storms evaluated the models on their best fit, a practice known to ignore the complexity of the models under consideration. In the present study, in contrast, model evaluation not only relies on the maximal likelihood, but also on the marginal likelihood, which is sensitive to model complexity. Finally, using a large recovery study, it is demonstrated that, across the 30 data sets, complexity differences between the models in the VAM family are small. This indicates that a (computationally challenging) complexity-sensitive model evaluation method is uncalled for, and that the use of a (computationally straightforward) complexity-insensitive model evaluation method is justified.     

Two mechanisms of human contingency learning

How do humans learn contingencies between events? Both pathway-strengthening and inference-based process models have been proposed to explain contingency learning. We propose that each of these processes is used in different conditions. Participants viewed displays that contained single or paired objects and learned which displays were usually followed by the appearance of a dot. Some participants predicted whether the dot would appear before seeing the outcome, whereas other participants were required to respond quickly if the dot appeared shortly after the display. In the prediction task, instructions guiding participants to infer which objects caused the dot to appear were necessary in order for contingencies associated with one object to influence participants' predictions about the object with which it had been paired. In the response task, contingencies associated with one object affected responses to its pair mate irrespective of whether or not participants were given causal instructions. Our results challenge single-mechanism accounts of contingency learning and suggest that the mechanisms underlying performance in the two tasks are distinct.     

Augmentation in contingency learning under time pressure

Recent research suggests that cue competition effects in human contingency learning, such as blocking, are due to higher‐order cognitive processes. Moreover, some experimental reports suggest that the effect opposite to blocking, augmentation, could occur in experimental preparations that preclude the intervention of reasoning mechanisms. In the present research, we tested this hypothesis by investigating cue interaction effects in an experimental task in which participants had to enter their responses under time pressure. The results show that under these conditions, augmentation, instead of blocking, is observed.     

The response-stimulus contingency and reinforcement learning as a context for considering two non-behavior-analytic views of contingency learning

This paper introduces a special section on the contingency. Bower and Watson were invited to present their views of contingency learning in human infants from outside the context of behavior analysis, and Cigales, Marr, and Lattal and Shahan provided commentaries that point out some of the more interesting and controversial aspects of those views from a behavior-analytic perspective. The debate turns on how to conceptualize the response-stimulus contingency of operant learning. The present paper introduces the contingency concept and contingency detection by subjects, as well as research practices in behavior analysis, in a context in which the dependency between infant responding and the presentation of environmental consequences may be disrupted through procedures in which ordinarily consequent events occur before the response or in its absence. These points can relate to and serve as an introduction to the Bower and Watson papers on infant contingency learning as well as to the three commentaries that follow.     

Contingency proportion systematically influences contingency learning

In the color-word contingency learning paradigm, each word appears more often in one color (high contingency) than in the other colors (low contingency). Shortly after beginning the task, color identification responses become faster on the high-contingency trials than on the low-contingency trials—the contingency learning effect. Across five groups, we varied the high-contingency proportion in 10% steps, from 80% to 40%. The size of the contingency learning effect was positively related to high-contingency proportion, with the effect disappearing when high contingency was reduced to 40%. At the two highest contingency proportions, the magnitude of the effect increased over trials, the pattern suggesting that there was an increasing cost for the low-contingency trials rather than an increasing benefit for the high-contingency trials. Overall, the results fit a modified version of Schmidt’s (2013, Acta Psychologica, 142, 119–126) parallel episodic processing account in which prior trial instances are routinely retrieved from memory and influence current trial performance.     

Overt attention and predictiveness in human contingency learning

Two experiments used eye-tracking procedures to investigate the relationship between attention and associative learning in human participants. These experiments found greater overt attention to cues experienced as predictive of the outcomes with which they were paired, than to cues experienced as nonpredictive. Moreover, this attentional bias persisted into a second training phase when all cues were equally predictive of the outcomes with which they were paired, and it was accompanied by a related bias in the rate of learning about these cues. These findings are consistent with the attentional model of associative learning proposed by Mackintosh (1975), but not with that proposed by Pearce and Hall (1980).     

Inferences about unobserved causes in human contingency learning

Estimates of the causal efficacy of an event need to take into account the possible presence and influence of other unobserved causes that might have contributed to the occurrence of the effect. Current theoretical approaches deal differently with this problem. Associative theories assume that at least one unobserved cause is always present. In contrast, causal Bayes net theories (including Power PC theory) hypothesize that unobserved causes may be present or absent. These theories generally assume independence of different causes of the same event, which greatly simplifies modelling learning and inference. In two experiments participants were requested to learn about the causal relation between a single cause and an effect by observing their co-occurrence (Experiment 1) or by actively intervening in the cause (Experiment 2). Participants' assumptions about the presence of an unobserved cause were assessed either after each learning trial or at the end of the learning phase. The results show an interesting dissociation. Whereas there was a tendency to assume interdependence of the causes in the online judgements during learning, the final judgements tended to be more in the direction of an independence assumption. Possible explanations and implications of these findings are discussed.     

Aging and the detection of contingency in causal learning

Young and older participants' ability to detect negative, random, and positive response-outcome contingencies was evaluated using both contingency estimation and response rate adaptation tasks. Age differences in contingency estimation were consistently greater for negative than positive contingencies, and these differences, though still present, were smaller when response rate adaptation was used as the measure of contingency learning. Detecting causal contingency apparently becomes more difficult with age, especially when an oven numerical estimate of contingency must be provided and when the relationship between a causal event and an outcome is negative. A model that incorporates features of both associative and rule-based approaches to contingency learning (e.g., P. C. Price & J. F. Yates, 1995; D. R. Shanks, 1995) provides the best explanation for this pattern of findings.     

Striatal activity in concept learning

Striatal learning systems have been implicated in learning relationships between visual stimuli and outcomes. In the present study, the activity of the striatum during visual concept learning in humans was examined by using functional magnetic resonance imaging (fMRI). Participants performed three concept-learning tasks and a baseline task. The participants were trained to criterion before fMRI scanning on two tasks, verbal and implicit. In the verbal task, classification could be performed on the basis of a simple verbal rule, but in the implicit task, there was no simple verbal rule. The novel-implicit learning task, in which an implicit structure was used, was not encountered by the participants before scanning. Across all three concept-learning tasks, there was significant activation in the striatum, in comparison with the baseline task. The striatum was recruited similarly in classification when the participants had different levels of expertise (novel-implicit vs. verbal and implicit) and were able to verbalize their learning to different degrees (verbal vs. implicit and novel-implicit). There was left lateral occipital activation when learning was implicit (implicit and novel-implicit), but not when learning was easily verbalized (verbal).     

Transformation rules in concept learning

Two major classes of models have been proposed to explain concept learning: strength models and distance models (Hayes-Roth & Hayes-Roth, 1977). The present study demonstrates that subjects abstract transformation rules as defined by the Franks and Bransford 11971) distance model. Transformation rules characterize how the patterns of a concept differ from each other. Transformation rules are inconsistent with strength models, which assume that subjects abstract component features and not relational information characterizing the differences among patterns. Whether a strength model or a distance model is more appropriate in other instances of concept learning is probably a function of task demands, stimulus characteristics, and subject characteristics.     

Secondary task difficulty modulates forward blocking in human contingency learning

The influence of a secondary task on forward blocking of human contingency ratings was examined. A smaller blocking effect was found when participants performed a highly demanding secondary task than when they performed a less demanding secondary task. The modulatory effect of secondary task difficulty was significant only when the secondary task was administered during both the learning and the test phase of the contingency judgement task. The results suggest that forward blocking in human contingency learning cannot be fully accounted for by associative processes. Instead, forward blocking seems to depend at least partially on deliberate deductive reasoning processes.     

Generalization gradients for acquisition and extinction in human contingency learning

Two experiments investigated the perceptual generalization of acquisition and extinction in human contingency learning. In Experiment 1, the degree of perceptual similarity between the acquisition stimulus and the generalization stimulus was manipulated over five groups. This successfully generated a generalization gradient of acquisition. In the subsequent phase, the response to the generalization stimulus was extinguished in each group. Finally, the acquisition stimulus was presented again. The response recovered differently over groups, thereby establishing the generalization gradient of extinction. In Experiment 2, the acquisition stimulus itself was extinguished before the set of generalization stimuli was tested between groups. One group evidenced a response recovery at test, which suggests that the gradient of acquisition is somewhat broader than the gradient of extinction.