Search

We analyze the visual perception task that home plate umpires (N = 121) perform calling balls and strikes (N = 3,001,019) in baseball games, focusing on the topics of perceptual learning and bias in decision-making. In the context of perceptual learning, our results show that monitoring, training, and feedback improve skill over time. In addition, we document two other aspects of umpires' improvement that are revealing with respect to the nature of their perceptual expertise. First, we show that biases in umpires' decision-making persist even as their overall accuracy improves. This suggests that bias and accuracy are orthogonal and that reduction of bias in decision-making requires interventions aimed specifically at this goal. Second, we measure a distinct difference in the rate of skill improvement between older and younger umpires. Younger umpires improve more quickly, suggesting that the decision task umpires engage in becomes routinized over time.

We develop a conceptual framework for studying collective adaptation in complex socio-cognitive systems, driven by dynamic interactions of social integration strategies, social environments and problem structures. Going beyond searching for 'intelligent' collectives, we integrate research from different disciplines and outline modelling approaches that can be used to begin answering questions such as why collectives sometimes fail to reach seemingly obvious solutions, how they change their strategies and network structures in response to different problems and how we can anticipate and perhaps change future harmful societal trajectories. We discuss the importance of considering path dependence, lack of optimization and collective myopia to understand the sometimes counterintuitive outcomes of collective adaptation. We call for a transdisciplinary, quantitative and societally useful social science that can help us to understand our rapidly changing and ever more complex societies, avoid collective disasters and reach the full potential of our ability to organize in adaptive collectives.

How do people use information from others to solve complex problems? Prior work has addressed this question by placing people in social learning situations where the problems they were asked to solve required varying degrees of exploration. This past work uncovered important interactions between groups' "connectivity" and the problem's "complexity": the advantage of less connected networks over more connected networks increased as exploration was increasingly required for optimally solving the problem at hand. We propose the Social Interpolation Model (SIM), an agent-based model to explore the cognitive mechanisms that can underlie exploratory behavior in groups. Through results from simulation experiments, we conclude that "exploration" may not be a single cognitive property, but rather the emergent result of three distinct behavioral and cognitive mechanisms, namely: (1) breadth of generalization; (2) quality of prior expectation; and (3) relative valuation of self-obtained information. We formalize these mechanisms in the SIM, and explore their effects on group dynamics and success at solving different kinds of problems. Our main finding is that broad generalization and high quality of prior expectation facilitate successful search in environments where exploration is important, and hinder successful search in environments where exploitation alone is sufficient.

To identify the ways teachers and educational systems can improve learning, researchers need to make causal inferences. Analyses of existing datasets play an important role in detecting causal patterns, but conducting experiments also plays an indispensable role in this research. In this article, we advocate for experiments to be embedded in real educational contexts, allowing researchers to test whether interventions such as a learning activity, new technology, or advising strategy elicit reliable improvements in authentic student behaviours and educational outcomes. Embedded experiments, wherein theoretically relevant variables are systematically manipulated in real learning contexts, carry strong benefits for making causal inferences, particularly when allied with the data-rich resources of contemporary e-learning environments. Toward this goal, we offer a field guide to embedded experimentation, reviewing experimental design choices, addressing ethical concerns, discussing the importance of involving teachers, and reviewing how interventions can be deployed in a variety of contexts, at a range of scales. Causal inference is a critical component of a field that aims to improve student learning; including experimentation alongside analyses of existing data in learning analytics is the most compelling way to test causal claims.

Cognitive capacities such as learning, reasoning, anddecision-making are often studied in tasks where a single par-ticipant acts in isolation. Yet humans don’t learn, reason, andmake decisions in a vacuum. Human cognition is distinc-tively social: Much of what we do influences—and is influ-enced by—other people.The goal of this workshop is to bring together diverse per-spectives on the interplay between human cognition and thedynamic, social environments we inhabit. The workshop isorganized around three key themes. Theme 1 lays out thecognitive tools that equip individuals to thrive in social en-vironments, including specialized mechanisms for teachingand learning from others. Theme 2 examines how the socialenvironment is itself shaped by the dynamic interactions be-tween multiple individuals, producing emergent behaviors atthe level of the collective. Finally, Theme 3 explores howhuman cognition responds to the demands of particular so-cial environments, including how cultural variability in socialcognition might emerge across development.Collectively, the research showcased in this workshopenriches this year’s conversation on “How to Develop aMind: Learning in Humans, Animals, and Machines”by highlighting the social and cultural context of learn-ing and development. In addition, our speakers representa broad cross-section of the conference, spanning multi-ple disciplines (computer science, anthropology, psychol-ogy), perspectives (computational, ecological, developmen-tal), and career stages (from research assistants to full pro-fessors). Below, we describe each theme and presentercontributions in detail. To take part in the workshop,visit cognitioncollectivesandculture.github.io forthe current schedule.

One of the major ways that people engage in adaptive problemsolving is by copying the solutions of others. Most of the workon this field has focused on three questions: when to copy, whoto copy from, and what to copy. However, how much to copyhas been relatively less explored. In the current research, weare interested in the consequences for a group when its mem-bers engage in social learning strategies with different tenden-cies to copy entire or partial solutions and different complex-ities of search problems. We also consider different networktopologies that affect the solutions visible to each member.Using a computational model of collective problem solving,we demonstrate that strategies where social learning involvespartial copying outperform strategies where individuals copyentire solutions. We analyze the exploration/exploitation dy-namics of these social learning strategies under the differentconditions.

We explore different ways in which the human visual system can adapt for perceiving and categorizing the environment. There are various accounts of supervised (categorical) and unsupervised perceptual learning, and different perspectives on the functional relationship between perception and categorization. We suggest that common experimental designs are insufficient to differentiate between hypothesized perceptual learning mechanisms and reveal their possible interplay. We propose a relatively underutilized way of studying potential categorical effects on perception, and we test the predictions of different perceptual learning models using a two-dimensional, interleaved categorization-plus-reconstruction task. We find evidence that the human visual system adapts its encodings to the feature structure of the environment, uses categorical expectations for robust reconstruction, allocates encoding resources with respect to categorization utility, and adapts to prevent miscategorizations.

The division of labor phenomenon has been observed with re-spect to both manual and cognitive labor, but there is no clearunderstanding of the intra- and inter-individual mechanismsthat allow for its emergence, especially when there are multipledivisions possible and communication is limited. Situationsfitting this description include individuals in a group splittinga geographical region for resource harvesting without explicitnegotiation, or a couple tacitly negotiating the hour of the dayfor each to shower so that there is sufficient hot water. We stud-ied this phenomenon by means of an iterative two-person gamewhere multiple divisions are possible, but no explicit commu-nication is allowed. Our results suggest that there are a lim-ited number of biases toward divisions of labor, which serveas attractors in the dynamics of dyadic coordination. How-ever, unlike Schelling’s focal points, these biases do not attractplayers’ attention at the onset of the interaction, but are onlyrevealed and consolidated by the in-game dynamics of dyadicinteraction.

Effective problem solving requires both exploration and ex-ploitation. We analyze data from a group problem-solving taskto gain insight into how people use information from past expe-riences and from others to achieve explore-exploit trade-offs incomplex environments. The behavior we observe is consistentwith the use of simple, reinforcement-based heuristics. Partic-ipants increase exploration immediately after experiencing alow payoff, and decrease exploration immediately after expe-riencing a high or improved payoff. We suggest that whetheran outcome is perceived as “high” or “low” is a dynamic func-tion of the outcome information available to participants. Thedegree to which the distribution of observed information re-flects the true range of possible outcomes plays an importantrole in determining whether or not this heuristic is adaptive ina given environment.

How does cooperation arise in an evolutionary context? We ap-proach this problem using a collective search paradigm whereinteractions are dynamic and there is competition for rewards.Using evolutionary simulations, we find that the unconditionalsharing of information can be an evolutionary advantageousstrategy without the need for conditional strategies or explicitreciprocation. Shared information acts as a recruitment sig-nal and facilitates the formation of a self-organized group.Thus, the improved search efficiency of the collective bestowsbyproduct benefits onto the original sharer. A key mecha-nism is a visibility radius, where individuals have uncondi-tional access to information about neighbors within a lim-ited distance. Our results show that for a variety of initialconditions—including populations initially devoid of prosocialindividuals—and across both static and dynamic fitness land-scapes, we find strong selection pressure to evolve uncondi-tional sharing.

Across three experiments featuring naturalistic concepts (psychology concepts) and naïve learners, we extend previous research showing an effect of the sequence of study on learning outcomes, by demonstrating that the sequence of examples during study changes the representation the learner creates of the study materials. We compared participants' performance in test tasks requiring different representations and evaluated which sequence yields better learning in which type of tests. We found that interleaved study, in which examples from different concepts are mixed, leads to the creation of relatively interrelated concepts that are represented by contrast to each other and based on discriminating properties. Conversely, blocked study, in which several examples of the same concept are presented together, leads to the creation of relatively isolated concepts that are represented in terms of their central and characteristic properties. These results argue for the integrated investigation of the benefits of different sequences of study as depending on the characteristics of the study and testing situation.

Information sharing in competitive environments may seemcounterintuitive, yet it is widely observed in humans and otheranimals. For instance, the open-source software movement hasled to new and valuable technologies being released publiclyto facilitate broader collaboration and further innovation. Whatdrives this behavior and under which conditions can it be ben-eficial for an individual? Using simulations in both static anddynamic environments, we show that sharing information canlead to individual benefits through the mechanisms of pseudo-reciprocity, whereby shared information leads to by-productbenefits for an individual without the need for explicit recipro-cation. Crucially, imitation with a certain level of innovation isrequired to avoid a tragedy of the commons, while the mecha-nism of a local visibility radius allows for the coordination ofself-organizing collectives of agents. When these two mecha-nisms are present, we find robust evidence for the benefits ofsharing—even when others do not reciprocate.

Despite its omnipresence in this information-ladensociety, statistics is hard. The present study explored theapplicability of a grounded cognition approach to learningbasic statistical concepts. Participants in 2 experimentsinteracted with perceptually rich computer simulationsdesigned to foster understanding of the relations betweenfundamental statistical concepts and to promote the abilityto reason with statistics. During training, participants wereasked to estimate the probability of two samples comingfrom the same population, with sample size, variability, anddifference between means independently manipulated. Theamount of learning during training was measured by thedifference between participants’ confidence judgments andthose of an Ideal Observer. The amount of transfer wasassessed by the increase in accuracy from a pretest to aposttest. Learning and transfer were observed when tailoredguidance was given along with the perceptually salientproperties. Implications of our quantitative measures ofhuman sensitivity to statistical concepts were discussed.

We investigated whether, and in what, ways people use visualstructures to evaluate mathematical expressions. We alsoexplored the relationship between strategy use and othercommon measures in mathematics education. Participantsorganized long sum/products when visual structure wasavailable in algebraic expressions. Two experiments showed asimilar pattern: One group of participants primarily calculatedfrom left to right, or combined identical numbers together. Asecond group calculated adjacent pairs. A third group tendedto group terms which either produced easy sums (e.g., 6+4),or participated in a global structure. These different strategieswere associated with different levels of success on the task,and, in Experiment 2, with differential math anxiety andmathematical skill. Specifically, problem solvers with lowermath anxiety and higher math ability tend to group by chunksand easy calculation. These results identify an important rolefor the perception of coherent structure and patternidentification in mathematical reasoning.

Previous research has shown that the sequence in whichconcepts are studied changes how well they are learned. In aseries of experiments featuring naturalistic concepts(psychology concepts) and naïve learners, we extend previousresearch by showing that the sequence of study changes therepresentation the learner creates of the study materials.Interleaved study leads to the creation of relatively interrelatedconcepts that are represented by contrast to each other andbased on discriminating properties. Blocked study, instead,leads to the creation of relatively isolated concepts that arerepresented in terms of their central and characteristicproperties. The relative benefits of these representationsdepend on whether the test of conceptual knowledge requirescontrastive or characteristic information. These results arguefor the integrated investigation of the benefits of differentsequences of study as depending on the characteristics of thestudy and testing situation as a whole.

Recent results demonstrate that inducing an abstractrepresentation of target analogs at retrieval time aids access toanalogous situations with mismatching surface features (i.e.,the late abstraction principle). A limitation of currentimplementations of this principle is that they either require theexternal provision of target-specific information or demandvery high intellectual engagement. Experiment 1 demonstratedthat constructing an idealized situation model of a targetproblem increases the rate of correct solutions compared toconstructing either concrete simulations or no simulations.Experiment 2 confirmed that these results were based on anadvantage for accessing the base analog, and not merely on anadvantage of idealized simulations for understanding the targetproblem in its own terms. This target idealization strategy hasbroader applicability than prior interventions based on the lateabstraction principle, because it can be achieved by a greaterproportion of participants and without the need to receivetarget-specific information.

Formal mathematics is a paragon of abstractness. It thus seems natural to assume that the mathematical expert should rely more on symbolic or conceptual processes, and less on perception and action. We argue instead that mathematical proficiency relies on perceptual systems that have been retrained to implement mathematical skills. Specifically, we investigated whether the visual system-in particular, object-based attention-is retrained so that parsing algebraic expressions and evaluating algebraic validity are accomplished by visual processing. Object-based attention occurs when the visual system organizes the world into discrete objects, which then guide the deployment of attention. One classic signature of object-based attention is better perceptual discrimination within, rather than between, visual objects. The current study reports that object-based attention occurs not only for simple shapes but also for symbolic mathematical elements within algebraic expressions-but only among individuals who have mastered the hierarchical syntax of algebra. Moreover, among these individuals, increased object-based attention within algebraic expressions is associated with a better ability to evaluate algebraic validity. These results suggest that, in mastering the rules of algebra, people retrain their visual system to represent and evaluate abstract mathematical structure. We thus argue that algebraic expertise involves the regimentation and reuse of evolutionarily ancient perceptual processes. Our findings implicate the visual system as central to learning and reasoning in mathematics, leading us to favor educational approaches to mathematics and related STEM fields that encourage students to adapt, not abandon, their use of perception.

Categorical perception (CP) effects manifest as faster or more accurate discrimination between objects that come from different categories compared with objects that come from the same category, controlling for the physical differences between the objects. The most popular explanations of CP effects have relied on perceptual warping causing stimuli near a category boundary to appear more similar to stimuli within their own category and/or less similar to stimuli from other categories. Hanley and Roberson (2011), on the basis of a pattern not previously noticed in CP experiments, proposed an explanation of CP effects that relies not on perceptual warping, but instead on inconsistent usage of category labels. Experiments 1 and 2 in this article show a pattern opposite the one Hanley and Roberson pointed out. Experiment 3, using the same stimuli but with different choice statistics (i.e., different probabilities of each face being the target), obtains the same pattern as the one Hanley and Roberson showed. Simulations show that both category label and perceptual models are able to reproduce the patterns of results from both experiments, provided they include information about the choice statistics. This suggests 2 conclusions. First, the results described by Hanley and Roberson should not be taken as evidence in favor of a category label model. Second, given that participants did not receive feedback on their choices, there must be some mechanism by which participants monitor their own choices and adapt to the choice statistics present in the experiment.

We explore how context change and item variation duringnatural category learning influence memory and generalizationto new examples. Participants studied either images of the samebird or varied birds from each of several categories. Theseimages could be presented in a constant background color ordifferent background colors. During test, birds were presentedin only one of the studied background colors. Performance attest depended on the context overlap between study and test,with better performance when there was minimal contextchange during study. Also, contrary to previous findings, wefound that learners generalized better when items wererepeated during study and remembered old items better whenitems were varied during study. When there is a moderatedegree of context change, there is no benefit of repetition orvariation for either novel or old items. These results indicatethat context change and item variation have complementaryeffects on learning.

The purpose of this paper is to explore how and when verbal labels facilitate relational reasoning and transfer. We review the research and theory behind two ways words might direct attention to relational information: (1) words generically invite people to compare and thus highlight relations (the Generic Tokens [GT] hypothesis), and/or (2) words carry semantic cues to common structure (the Cues to Specific Meaning [CSM] hypothesis). Four experiments examined whether learning Signal Detection Theory (SDT) with relational words fostered better transfer than learning without relational words in easily alignable and less alignable situations (testing the GT hypothesis) as well as when the relational words matched and mismatched the semantics of the learning situation (testing the CSM hypothesis). The results of the experiments found support for the GT hypothesis because the presence of relational labels produced better transfer when two situations were alignable. Although the CSM hypothesis does not explain how words facilitate transfer, we found that mismatches between words and their labeled referents can produce a situation where words hinder relational learning.

Study sequence can have a profound impact on learning.Previous research has often shown advantages for interleavedover blocked study, though the reverse has also been found.Learners typically prefer blocking even in situations forwhich interleaving is superior. The present study investigatedlearner regulation of study sequence, and its effects onlearning in an ecologically valid context – university studentsusing an online tutorial relevant to an exam that countedtoward their course grades. The majority of participantsblocked study by problem category, and this tendency waspositively associated with subsequent exam performance. Theresults suggest that preference for blocked study may beadaptive under some circumstances, and highlight theimportance of identifying task environments under whichdifferent study sequences are most effective.

We present evidence that successful chunk formation during astatistical learning task depends on how well the perceiver isable to parse the information that is presented betweensuccessive presentations of the to-be-learned chunk. First, weshow that learners acquire a chunk better when thesurrounding information is also chunk-able in a visualstatistical learning task. We tested three process models ofchunk formation, TRACX, PARSER, and MDLChunker, onour two different experimental conditions, and found that onlyPARSER and MDLChunker matched the observed result.These two models share the common principle of a memorycapacity that is expanded as a result of learning. Thoughimplemented in very different ways, both models effectivelyremember more individual items (the atomic components of asequence) as additional chunks are formed. The ability toremember more information directly impacts learning in themodels, suggesting that there is a positive-feedback loop inchunk learning.

Three experiments explore differences between blocked andinterleaved study with and without item repetition. In the firstexperiment we find that when items are repeated duringstudy, blocked study results in higher test performance thaninterleaved study. In the second experiment we find that whenthere is no item repetition, interleaved and blocked studyresult in equivalent performance during the test phase. In thethird experiment we find that when the study is passive andincludes no item repetition, interleaved study results in highertest performance. We propose that learners create associationsbetween items of the same category during blocked study anditem repetition strengthens these associations. Interleavedstudy leads to weaker associations between items of the samecategory and therefore results in worse performance duringtest when there are item repetitions

Category learning is an essential cognitive mechanism formaking sense of the world. Many existing computational categorylearning models focus on categories that can be representedas feature vectors, and yet a substantial part of the categorieswe encounter have members with inner structure andinner relationships. We present a novel computational modelthat perceives and learns structured concepts from physicalscenes. The perception and learning processes happen simultaneouslyand interact with each other. We apply the modelto a set of physical categorization tasks and promote specifictypes of comparisons by manipulating presentation order ofexamples. We find that these manipulations affect the algorithmsimilarly to human participants that worked on the sametask. Both benefit from juxtaposing examples of different categories– especially ones that are similar to each other. Whenjuxtaposing examples from the same category they do better ifthe examples are dissimilar to each other.

The sequence of study influences how we learn. Previous research has identified different sequences as potentially beneficial for learning in different contexts and with different materials. Here we investigate the mechanisms involved in inductive category learning that give rise to these sequencing effects. Across 3 experiments we show evidence that the sequence of study changes what information learners attend to during learning, what is encoded from the materials studied and, consequently, what is remembered from study. Interleaved study (alternating between presentation of 2 categories) leads to an attentional focus on properties that differ between successive items, leading to relatively better encoding and memory for item properties that discriminate between categories. Conversely, when learners study each category in a separate block (blocked study), learners encode relatively more strongly the characteristic features of the items, which may be the result of a strong attentional focus on sequential similarities. These results provide support for the sequential attention theory proposing that inductive category learning takes place through a process of sequential comparisons between the current and previous items. Different sequences of items change how attention is deployed depending on this basic process. Which sequence results in better or worse learning depends on the match between what is encoded and what is required at test.

Humans routinely form groups to achieve goals that no individual can accomplish alone. Group coordination often brings to mind synchrony and alignment, where all individuals do the same thing (e.g., driving on the right side of the road, marching in lockstep, or playing musical instruments on a regular beat). Yet, effective coordination also typically involves differentiation, where specialized roles emerge for different members (e.g., prep stations in a kitchen or positions on an athletic team). Role specialization poses a challenge for computational models of group coordination, which have largely focused on achieving synchrony. Here, we present the CARMI framework, which characterizes role specialization processes in terms of five core features that we hope will help guide future model development: Communication, Adaptation to feedback, Repulsion, Multi‐level planning, and Intention modeling. Although there are many paths to role formation, we suggest that roles emerge when each agent in a group dynamically allocates their behavior toward a shared goal to complement what they expect others to do. In other words, coordination concerns beliefs (who will do what) rather than simple actions. We describe three related experimental paradigms—"Group Binary Search," "Battles of the Exes," and "Find the Unicorn"—that we have used to study differentiation processes in the lab, each emphasizing different aspects of the CARMI framework. Group coordination often brings to mind group members synchronizing their behaviors, as with soldiers marching in lockstep or audience members spontaneously clapping in unison. However, effective coordination often crucially involves a division of labor in which different members come to adopt different roles. The emergence of specialized roles in the three collective behavior paradigms we review can be understood in terms of five interactive processes: communication, adaptation to feedback, repulsion among member strategies, multi‐level planning, and intention modeling. [ABSTRACT FROM AUTHOR]

Many life-influencing social networks are characterized by considerable informational isolation. People within a community are far more likely to share beliefs than people who are part of different communities. The spread of useful information across communities is impeded by echo chambers (far greater connectivity within than between communities) and filter bubbles (more influence of beliefs by connected neighbors within than between communities). We apply the tools of network analysis to organize our understanding of the spread of beliefs across modularized communities and to predict the effect of individual and group parameters on the dynamics and distribution of beliefs. In our Spread of Beliefs in Modularized Communities (SBMC) framework, a stochastic block model generates social networks with variable degrees of modularity, beliefs have different observable utilities, individuals change their beliefs on the basis of summed or average evidence (or intermediate decision rules), and parameterized stochasticity introduces randomness into decisions. SBMC simulations show surprising patterns; for example, increasing out-group connectivity does not always improve group performance, adding randomness to decisions can promote performance, and decision rules that sum rather than average evidence can improve group performance, as measured by the average utility of beliefs that the agents adopt. Overall, the results suggest that intermediate degrees of belief exploration are beneficial for the spread of useful beliefs in a community, and so parameters that pull in opposite directions on an explore–exploit continuum are usefully paired. [ABSTRACT FROM AUTHOR]