Cognitive Science, Volume 46, Issue 5, May 2022.

Unlike behaviorism, cognitive psychology relies on mental concepts to explain behavior. Yet mental processes are not directly observable and multiple explanations are possible, which poses a challenge for finding a useful framework. In this article, I distinguish three new frameworks for explanations that emerged after the cognitive revolution. The first is called tools‐to‐theories: Psychologists' new tools for data analysis, such as computers and statistics, are turned into theories of mind. The second proposes as‐if theories: Expected utility theory and Bayesian statistics (...) are turned into theories of mind, describing an optimal solution of a problem but not its psychological process. The third studies the adaptive toolbox (formal models of heuristics) that describes mental processes in situations of uncertainty where an optimal solution is unknown. Depending on which framework researchers choose, they will model behavior in either situations of risk or of uncertainty, and construct models of cognitive processes or not. The frameworks also determine what questions are asked and what kind of data are generated. What all three frameworks have in common, however, is a clear preference for formal models rather than explanations by general dichotomies or mere verbal concepts. The frameworks have considerable potential to inform each other and to generate points of integration. (shrink)

Mental chronometry is a classical paradigm in cognitive psychology that uses response time and accuracy data in perceptual-motor tasks to elucidate the architecture and mechanisms of the underlying cognitive processes of human decisions. The redundant signals paradigm investigates the response behavior in Experimental tasks, where an integration of signals is required for a successful performance. The common finding is that responses are speeded for the redundant signals condition compared to single signals conditions. On a mean level, this redundant signals effect (...) can be accounted for by several cognitive architectures, exhibiting considerable model mimicry. Jeff Miller formalized the maximum speed-up explainable by separate activations or race models in form of a distributional bound – the race model inequality. Whenever data violates this bound, it excludes race models as a viable account for the redundant signals effect. The common alternative is a coactivation account, where the signals integrate at some stage in the processing. Coactivation models have mostly been inferred on and rarely explicated though. Where coactivation is explicitly modeled, it is assumed to have a decisional locus. However, in the literature there are indications that coactivation might have at least a partial locus in the nondecisional or motor stage. There are no studies that have tried to compare the fit of these coactivation variants to empirical data to test different effect generating loci. Ever since its formulation, the race model inequality has been used as a test to infer the cognitive architecture for observers’ performance in redundant signals Experiments. Subsequent theoretical and empirical analyses of this RMI test revealed several challenges. On the one hand, it is considered to be a conservative test, as it compares data to the maximum speed-up possible by a race model account. Moreover, simulation studies could show that the base time component can further reduce the power of the test, as violations are filtered out when this component has a high variance. On the other hand, another simulation study revealed that the common practice of RMI test can introduce an estimation bias, that effectively facilitates violations and increases the type I error of the test. Also, as the RMI bound is usually tested at multiple points of the same data, an inflation of type I errors can reach a substantial amount. Due to the lack of overlap in scope and the usage of atheoretic, descriptive reaction time models, the degree to which these results can be generalized is limited. State-of-the-art models of decision making provide a means to overcome these limitations and implement both race and coactivation models in order to perform large scale simulation studies. By applying a state-of-the-art model of decision making to the investigation of the redundant signals effect, the present study addresses research questions at different levels. On a conceptual level, it raises the question, at what stage coactivation occurs – at a decisional, a nondecisional or a combined decisional and nondecisional processing stage and to what extend? To that end, two bimodal detection tasks have been conducted. As the reaction time data exhibits violations of the RMI at multiple time points, it provides the basis for a comparative fitting analysis of coactivation model variants, representing different loci of the effect. On a test theoretic level, the present study integrates and extends the scopes of previous studies within a coherent simulation framework. The effect of experimental and statistical parameters on the performance of the RMI test is analyzed via Monte Carlo simulations. Specifically, the simulations treated the following questions: what is the power of the RMI test, is there an estimation bias for coactivated data as well and if so, in what direction, what is the effect of a highly varying base time component on the estimation bias, type I errors and power rates, and are the results of previous simulation studies replicable, when current models of decision making are used for the reaction time generation. For this purpose, the Ratcliff diffusion model was used to implement race models with controllable amount of correlation and coactivation models with varying integration strength, and independently specifying the base time component. The results of the fitting suggest that for the two bimodal detection tasks, coactivation has a shared decisional and nondecisional locus. For the focused attention experiment the decisional part prevails, whereas in the divided attention task the motor component is dominating the redundant signals effect. The simulation study could reaffirm the conservativeness of the RMI test as latent coactivation is frequently missed. An estimation bias was found also for coactivated data however, both biases become negligible once more than 10 samples per condition are taken to estimate the respective distribution functions. A highly varying base time component reduces both the type I errors and the power of the test, while not affecting the estimation biases. The outcome of the present study has theoretical and practical implications for the investigations of decisions in a multisignal context. Theoretically, it contributes to the locus question of coactivation and offers evidence for a combined decisional and nondecisional coactivation account. On a practical level, the modular simulation approach developed in the present study enables researchers to further investigate the RMI test within a coherent and theoretically grounded framework. It effectively provides a means to optimally set up the RMI test and thus helps to solidify and substantiate its outcomes. On a conceptual level the present study advocates the application of current formal models of decision making to the mental chronometry paradigm and develops future research questions in the field of the redundant signals paradigm. (shrink)

This article discusses how sequential sampling models can be integrated in a cognitive architecture. The new theory Retrieval by Accumulating Evidence in an Architecture (RACE/A) combines the level of detail typically provided by sequential sampling models with the level of task complexity typically provided by cognitive architectures. We will use RACE/A to model data from two variants of a picture–word interference task in a psychological refractory period design. These models will demonstrate how RACE/A enables interactions between sequential sampling and long-term (...) declarative learning, and between sequential sampling and task control. In a traditional sequential sampling model, the onset of the process within the task is unclear, as is the number of sampling processes. RACE/A provides a theoretical basis for estimating the onset of sequential sampling processes during task execution and allows for easy modeling of multiple sequential sampling processes within a task. (shrink)

Previous studies have shown that multiple reference frames are available and compete for selection during the use of spatial terms such as “above.” However, the mechanisms that underlie the selection process are poorly understood. In the current paper we present two experiments and a comparison of three computational models of selection to shed further light on the nature of reference frame selection. The three models are drawn from different areas of human cognition, and we assess whether they may be applied (...) to a reference frame selection by examining their ability to account for both existing and new empirical data comprising acceptance rates, response times, and response time distributions. These three models are the competitive shunting model, the leaky competing accumulator model, and a lexical selection model. Model simulations show that only the LCA model satisfactorily accounts for the empirical observations. The key properties of this model that seem to drive its success are its bounded linear activation function, its number and type of processing stages, and its use of decay. Uncovering these critical properties has important implications for our understanding not only of spatial term use, in particular, but also of conflict and selection in human cognition more generally. (shrink)

Recent years have witnessed an enormous increase in behavioral and neuroimaging studies of numerical cognition. Particular interest has been devoted toward unraveling properties of the representational medium on which numbers are thought to be represented. We have argued that a correct inference concerning these properties requires distinguishing between different input modalities and different decision/output structures. To back up this claim, we have trained computational models with either symbolic or nonsymbolic input and with different task requirements, and showed that this allowed (...) for an integration of the existing data in a consistent manner. In later studies, predictions from the models were derived and tested with behavioral and neuroimaging methods. Here we present an integrative review of this work. (shrink)

According to the dual-system theories, the decisions in an ultimatum game are governed by the automatic System 1 and the controlled System 2. The former drives the preference for fairness, whereas the latter drives the self-interest motive. However, the association between the contributions of the two systems in UG and the cognitive process needs more direct evidence. In the present study, we used the process dissociation procedure to estimate the contributions of the two systems and recorded participants eye movements to (...) examine the cognitive processes underlying UG decisions. Results showed that the estimated contributions of the two systems are uncorrelated and that they demonstrate a dissociated pattern of associations with third variables, such as reaction time and mean fixation duration. Furthermore, the relative time advantage and the transitions between the two payoffs can predict the final UG decisions. Our findings provide evidence for the independent contributions of preference for fairness and self-interest maximizing inclinations to UG and shed light on the underlying processes. (shrink)

In recent years quantum probability models have been used to explain many aspects of human decision making, and as such quantum models have been considered a viable alternative to Bayesian models based on classical probability. One criticism that is often leveled at both kinds of models is that they lack a clear interpretation in terms of psychological mechanisms. In this paper we discuss the mechanistic underpinnings of a quantum walk model of human decision making and response time. The quantum walk (...) model is compared to standard sequential sampling models, and the architectural assumptions of both are considered. In particular, we show that the quantum model has a natural interpretation in terms of a cognitive architecture that is both massively parallel and involves both co-operative and competitive interactions between units. Additionally, we introduce a family of models that includes aspects of the classical and quantum walk models. (shrink)

Creating agents that proficiently interact with people is critical for many applications. Towards creating these agents, models are needed that effectively predict people's decisions in a variety of problems. To date, two approaches have been suggested to generally describe people's decision behavior. One approach creates a-priori predictions about people's behavior, either based on theoretical rational behavior or based on psychological models, including bounded rationality. A second type of approach focuses on creating models based exclusively on observations of people's behavior. At (...) the forefront of these types of methods are various machine learning algorithms. This paper explores how these two approaches can be compared and combined in different types of domains. In relatively simple domains, both psychological models and machine learning yield clear prediction models with nearly identical results. In more complex domains, the exact action predicted by psychological models is not even clear, and machine learning models are even less accurate. Nonetheless, we present a novel approach of creating hybrid methods that incorporate features from psychological models in conjunction with machine learning in order to create significantly improved models for predicting people's decisions. To demonstrate these claims, we present an overview of previous and new results, taken from representative domains ranging from a relatively simple optimization problem and complex domains such as negotiation and coordination without communication. (shrink)

Individuals with high anxiety show bias for threatening information, but it is unclear whether this bias affects memory. Recognition memory studies have shown biases for recognising and rejecting threatening items in anxiety, prompting the need to identify moderating factors of this effect. This study focuses on the role of semantic similarity: the use of many semantically related threatening words could increase familiarity for those items and obscure anxiety-related differences in memory. To test this, two recognition memory experiments varied the proportion (...) of threatening words in lists to manipulate the semantic-similarity effects. When similarity effects were reduced, participants with high trait anxiety were biased to respond “new” to threatening words, whereas when similarity effects were strong there was no effect of anxiety on memory bias. Analysis of the data with the drift diffusion model showed that the bias was due to differences in processing of the threatening stimuli rather than a simple response bias. These data suggest that the semantic similarity of the threatening words significantly affects the presence or absence of anxiety-related threat bias in recognition memory. The results indicate that trait anxiety is associated with a bias to decide that threatening stimuli were not previously studied, but only when semantic-similarity effects are controlled. Implications for theories of anxiety and future studies in this domain are discussed. (shrink)

Diffusion models are widely-used and successful accounts of the time course of two-choice decision making. Most diffusion models assume constant boundaries, which are the threshold levels of evidence that must be sampled from a stimulus to reach a decision. We summarize theoretical results from statistics that relate distributions of decisions and response times to diffusion models with time-varying boundaries. We then develop a computational method for finding time-varying boundaries from empirical data, and apply our new method to two problems. The (...) first problem involves finding the time-varying boundaries that make diffusion models equivalent to the alternative sequential sampling class of accumulator models. The second problem involves finding the time-varying boundaries, at the individual level, that best fit empirical data for perceptual stimuli that provide equal evidence for both decision alternatives. We discuss the theoretical and modeling implications of using time-varying boundaries in diffusion models, as well as the limitations and potential of our approach to their inference. (shrink)

The information age can be dated to the work of Norbert Wiener and Claude Shannon in the 1940s. Their work on cybernetics and information theory, and many subsequent developments, had a profound influence on reshaping the field of psychology from what it was prior to the 1950s. Contemporaneously, advances also occurred in experimental design and inferential statistical testing stemming from the work of Ronald Fisher, Jerzy Neyman, and Egon Pearson. These interdisciplinary advances from outside of psychology provided the conceptual and (...) methodological tools for what is often called the cognitive revolution but is more accurately described as the information-processing revolution. Cybernetics set the stage with the idea that everything ranging from neurophysiological mechanisms to societal activities can be modeled as structured control systems with feedforward and feedback loops. Information theory offered a way to quantify entropy and information, and promoted theorizing in terms of information flow. Statistical theory provided means for making scientific inferences from the results of controlled experiments and for conceptualizing human decision making. With those three pillars, a cognitive psychology adapted to the information age evolved. The growth of technology in the information age has resulted in human lives being increasingly interweaved with the cyber environment, making cognitive psychology an essential part of interdisciplinary research on such interweaving. Continued engagement in interdisciplinary research at the forefront of technology development provides a chance for psychologists not only to refine their theories but also to play a major role in the advent of a new age of science. (shrink)

A framing bias shows risk aversion in problems framed as “gains” and risk seeking in problems framed as “losses,” even when these are objectively equivalent and probabilities and outcomes values are explicitly provided. We test this framing bias in situations where decision makers rely on their own experience, sampling the problem's options and seeing the outcomes before making a choice. In Experiment 1, we replicate the framing bias in description-based decisions and find risk indifference in gains and losses in experience-based (...) decisions. Predictions of an Instance-Based Learning model suggest that objective probabilities as well as the number of samples taken are factors that contribute to the lack of framing effect. We test these two factors in Experiment 2 and find no framing effect when a few samples are taken but when large samples are taken, the framing effect appears regardless of the objective probability values. Implications of behavioral results and cognitive modeling are discussed. (shrink)

In many learning or inference tasks human behavior approximates that of a Bayesian ideal observer, suggesting that, at some level, cognition can be described as Bayesian inference. However, a number of findings have highlighted an intriguing mismatch between human behavior and standard assumptions about optimality: People often appear to make decisions based on just one or a few samples from the appropriate posterior probability distribution, rather than using the full distribution. Although sampling-based approximations are a common way to implement Bayesian (...) inference, the very limited numbers of samples often used by humans seem insufficient to approximate the required probability distributions very accurately. Here, we consider this discrepancy in the broader framework of statistical decision theory, and ask: If people are making decisions based on samples—but as samples are costly—how many samples should people use to optimize their total expected or worst-case reward over a large number of decisions? We find that under reasonable assumptions about the time costs of sampling, making many quick but locally suboptimal decisions based on very few samples may be the globally optimal strategy over long periods. These results help to reconcile a large body of work showing sampling-based or probability matching behavior with the hypothesis that human cognition can be understood in Bayesian terms, and they suggest promising future directions for studies of resource-constrained cognition. (shrink)

Preadolescence is an important period for the consolidation of certain arithmetic facts, and the development of problem-solving strategies. Obese subjects seem to have poorer academic performance in math than their normal-weight peers, suggesting a negative effect of obesity on math skills in critical developmental periods. To test this hypothesis, event-related potentials were collected during a delayed-verification math task using simple addition and subtraction problems in obese [above 95th body mass index percentile] and non-obese preteens with different levels of math skill; (...) thirty-one with low math skills and thirty-one with high math skills. No significant differences between weight groups were observed in task accuracy regardless of their mathematical skill level. For ERPs, electrophysiological differences were found only in the subtraction condition; participants with obesity showed an electrophysiologic pattern associated with a reduced ability to allocate attention resources regardless of their math skill level, these differences were characterized by longer P300 latency than their normal-weight peers. Moreover, the participants with obesity with high math skills displayed hypoactivity in left superior parietal lobule compared with their normal-weight peers. Additionally, obese preteens with low math skills displayed smaller arithmetic N400 amplitude than non-obese participants, reflecting difficulties in retrieving visual, semantic, and lexical information about numbers. We conclude that participants with obesity are less able than their normal-weight peers to deploy their attention regardless of their behavioral performance, which seems to have a greater effect on obese participants with low math skills because they also show problems in the retrieval of solutions from working memory, resulting in a delay in the development of mathematical skills. (shrink)

The speed‐accuracy trade‐off (SAT) effect refers to the behavioral trade‐off between fast yet error‐prone respones and accurate but slow responses. Multiple theories on the cognitive mechanisms behind SAT exist. One theory assumes that SAT is a consequence of strategically adjusting the amount of evidence required for overt behaviors, such as perceptual choices. Another theory hypothesizes that SAT is the consequence of the mixture of multiple categorically different cognitive processes. In this paper, these theories are disambiguated by assessing whether the fixed‐point (...) property of mixture distributions holds, in both simulations and data. I conclude that, at least for perceptual decision making, there is no evidence for a mixture of different cognitive processes to trade off accuracy of responding for speed. (shrink)

The speed-accuracy trade-off effect refers to the behavioral trade-off between fast yet error-prone respones and accurate but slow responses. Multiple theories on the cognitive mechanisms behind SAT exist. One theory assumes that SAT is a consequence of strategically adjusting the amount of evidence required for overt behaviors, such as perceptual choices. Another theory hypothesizes that SAT is the consequence of the mixture of multiple categorically different cognitive processes. In this paper, these theories are disambiguated by assessing whether the fixed-point property (...) of mixture distributions holds, in both simulations and data. I conclude that, at least for perceptual decision making, there is no evidence for a mixture of different cognitive processes to trade off accuracy of responding for speed. (shrink)

This article demonstrates the potential of using hierarchical Bayesian methods to relate models and data in the cognitive sciences. This is done using a worked example that considers an existing model of category representation, the Varying Abstraction Model (VAM), which attempts to infer the representations people use from their behavior in category learning tasks. The VAM allows for a wide variety of category representations to be inferred, but this article shows how a hierarchical Bayesian analysis can provide a unifying explanation (...) of the representational possibilities using 2 parameters. One parameter controls the emphasis on abstraction in category representations, and the other controls the emphasis on similarity. Using 30 previously published data sets, this work shows how inferences about these parameters, and about the category representations they generate, can be used to evaluate data in terms of the ongoing exemplar versus prototype and similarity versus rules debates in the literature. Using this concrete example, this article emphasizes the advantages of hierarchical Bayesian models in converting model selection problems to parameter estimation problems, and providing one way of specifying theoretically based priors for competing models. (shrink)