A computer model of context dependent perception in a very simple world

Lara-Dammer, F., Hofstadter, D. R., & Goldstone, R. L. (in press). A computer model of context dependent perception in a very simple world.  Journal of Experimental & Theoretical Artificial Intelligence.

We propose the foundations of a computer model of scientic discovery that takes into account certain psychological aspects of human observation of the world. To this end, we simulate two main components of such a system. The first is a dynamic microworld in which physical events take place, and the second is an observer that visually perceives entities and events in the microworld. For reason of space, this paper focuses only on the starting phase of discovery, which is the relatively simple visual inputs of objects and collisions.

Download PDF of article

The Multiple Interactive Levels of Cognition (MILCS) perspective on group cognition

Goldstone, R. L., & Theiner, G. (2017). The Multiple Interactive Levels of Cognition (MILCS) perspective on group cognition.  Philosophical Psychology, 1-35.  DOI: 10.1080/09515089.2017.1295635

We lay out a multiple, interacting levels of cognitive systems (MILCS) framework to account for the cognitive capacities of individuals and the groups to which they belong. The goal of MILCS is to explain the kinds of cognitive processes typically studied by cognitive scientists, such as perception, attention, memory, categorization, decision-making, problem solving, judgment, and flexible behavior. Two such systems are considered in some detail—lateral inhibition within a network for selecting the most attractive option from a candidate set and a diffusion process for accumulating evidence to reach a rapid and accurate decision. These system descriptions are aptly applied at multiple levels, including within and across people. These systems provide accounts that unify cognitive processes across multiple levels, can be expressed in a common vocabulary provided by network science, are inductively powerful yet appropriately constrained, and are applicable to a large number of superficially diverse cognitive systems. Given group identification processes, cognitively resourceful people will frequently form groups that effectively employ cognitive systems at higher levels than the individual. The impressive cognitive capacities of individual people do not eliminate the need to talk about group cognition. Instead, smart people can provide the interacting parts for smart groups

Download PDF of article

Discovering Psychological Principles by Mining Naturally Occurring Data Sets

Goldstone, R. L., & Lupyan, G. (2016).  Harvesting naturally occurring data to reveal principles of cognitionTopics in Cognitive Science, 8, 548-568.

The very expertise with which psychologists wield their tools for achieving laboratory control may have had the unwelcome effect of blinding psychologists to the possibilities of discovering principles of behavior without conducting experiments. When creatively interrogated, a diverse range of large, real-world data sets provides powerful diagnostic tools for revealing principles of human judgment, perception, categorization, decision-making, language use, inference, problem solving, and representation. Examples of these data sets include patterns of website links, dictionaries, logs of group interactions, collections of images and image tags, text corpora, history of financial transactions, trends in twitter tag usage and propagation, patents, consumer product sales, performance in high-stakes sporting events, dialect maps, and scientific citations. The goal of this issue is to present some exemplary case studies of mining naturally existing data sets to reveal important principles and phenomena in cognitive science, and to discuss some of the underlying issues involved with conducting traditional experiments, analyses of naturally occurring data, computational modeling, and the synthesis of all three methods.This article serves as the introduction to a TopiCS topic with the same name.  The rest of the downloadable papers in this Topic are:

Download a PDF of this paper

Berger, J. (2016). Does presentation order impact choice after delay? Topics in Cognitive Science.

Christiansen, M. H., & Monaghan, P. (2016). Division of labor in vocabulary structure: Insights from corpus analyses. Topics in Cognitive Science, 8, 670684. doi: 10.1111/tops.12205.

Griffiths, T. L., Abbott, J. T., & Hsu, A. S. (2016). Exploring human cognition using large image databases. Topics in Cognitive Science, 8, 569588. doi: 10.1111/tops.12209.

Heit, E., & Nicholson, S. P. (2016). Missing the party: Political categorization and reasoning in the absence of party label cues. Topics in Cognitive Science, 8, 697714. doi: 10.1111/tops.12206.

Koedinger, K. R., Yudelson, M. V., & Pavlik Jr., P. I. (2016). Testing theories of transfer using error rate learning curves. Topics in Cognitive Science, 8, 589609. doi: 10.1111/tops.12208.

Moat, H. S., Olivola, C. Y., Chater, N., & Preis, T. (2016). Searching choices: Quantifying decision making processes using search engine data. Topics in Cognitive Science, 8, 685696. doi: 10.1111/tops.12207.

Pope, D. G. (2016). Exploring psychology in the field: Steps and examples from the used-car market. Topics in Cognitive Science, 8, 660669. doi: 10.1111/tops.12210.

Vincent-Lamarre, P., Blondin Masse, A., Lopes, M., Lord, M., Marcotte, O., & Harnad, S. (2016). The latent structure of dictionaries. Topics in Cognitive Science, 8, 625659. doi: 10.1111/tops.12211.


Index of Supplemental Videos for “An Integrated Computational Model of Perception and Scientific Discovery in a Very Simple World, Aiming at Psychological Realism”

Below is an index of supplemental videos for the manuscript:

Lara-Dammer, F., Hofstadter, D. R., & Goldstone, R. L. (under review).An Integrated Computational Model of Perception and Scientific Discovery in a Very Simple World, Aiming at Psychological Realism

Overview V1

Plausible Approach V1

Plausible Approach V2

Object Identification V1

Object Identification V2

Object Identification V3

Object Identification V4

Object Identification V5

Same Angle V1

Same Angle V2

Ambiguous Event V1

Ambiguous Event V2

Direction Parameter V1

Direction Parameter V2

Direction Parameter V3

Direction Parameter V4

Direction Parameter V5

Stability V1

Stability V2

First Discovery 1

First Discovery 2

First Discovery 3

First Discovery 4

Pressure 1

Pressure 2

Pressure 3

Pressure 4

Kinetic Energy 1

Kinetic Energy 2

Kinetic Energy 3

Kinetic Energy 4

Free Space 1

Free Space 2

Free Space in a Circle (Tricycle)

Boyle’s Law Sophisticated A (Tricycle)

Boyle’s Law Sophisticated B (Tricycle)

Boyle’s Law Sophisticated C (Tricycle)

Non-ideal Gas A (non-success, Tricycle)

Non-ideal Gas B (Tricycle)

Non-ideal Gas C (Tricycle)

Understanding Noise (Tricycle)

Failed Discovery (Tricycle)

Thinking in Groups A (Tricycle)

Thinking in Groups B (Tricycle)

Effects of variation and prior knowledge on abstract concept learning

Braithwaite, D. W., & Goldstone, R. L., (2015). Effects of variation and prior knowledge on abstract concept learning. Cognition and Instruction, 33, 226-256.

Learning abstract concepts through concrete examples may promote learning at the cost of inhibiting transfer. The present study investigated one approach to solving this problem: systematically varying superficial features of the examples. Participants learned to solve problems involving a mathematical concept by studying either superficially similar or varied examples. In Experiment 1, less knowledgeable participants learned better from similar examples,while more knowledgeable participants learned better from varied examples. In Experiment 2, prior to learning how to solve the problems, some participants received a pretraining aimed at increasing attention to the structural relations underlying the target concept. These participants, like the more knowledgeable participants in Experiment 1, learned better from varied examples. Thus, the utility of varied examples depends on prior knowledge and, in particular, ability to attend to relevant structure. Increasing this ability can prepare learners to learn more effectively from varied examples.

Download a PDF of the paper


Integration of social information by human groups

Granovskiy, B., Gold, J. M., Sumpter, D., & Goldstone, R. L. (2015). Integration of social information by human groups. Topics in Cognitive Science, 7, 469-493.

We consider a situation in which individuals search for accurate decisions without direct feedback on their accuracy, but with information about the decisions made by peers in their group. The “wisdom of crowds” hypothesis states that the average judgment of many individuals can give a good estimate of, for example, the outcomes of sporting events and the answers to trivia questions. Two conditions for the application of wisdom of crowds are that estimates should be independent and unbiased. Here, we study how individuals integrate social information when answering trivia questions with answers that range between 0% and 100% (e.g., “What percentage of Americans are left-handed?”). We find that, consistent with the wisdom of crowds hypothesis, average performance improves with group size. However, individuals show a consistent bias to produce estimates that are insufficiently extreme. We find that social information provides significant, albeit small, improvement to group performance. Outliers with answers far from the correct answer move toward the position of the group mean. Given that these outliers also tend to be nearer to 50% than do the answers of other group members, this move creates group polarization away from 50%. By looking at individual performance over different questions we find that some people are more likely to be affected by social influence than others. There is also evidence that people differ in their competence in answering questions, but lack of competence is not significantly correlated with willingness to change guesses. We develop a mathematical model based on these results that postulates a cognitive process in which people first decide whether to take into account peer guesses, and if so, to move in the direction of these guesses. The size of the move is proportional to the distance between their own guess and the average guess of the group. This model closely approximates the distribution of guess movements and shows how outlying incorrect opinions can be systematically removed from a group resulting, in some situations, in improved group performance. However, improvement is only predicted for cases in which the initial guesses of individuals in the group are biased.

Download PDF of paper

Memory constraints affect statistical learning; statistical learning affects memory constraints

de Leeuw, J. R., & Goldstone, R. L. (2015).  Memory constraints affect statistical learning; statistical learning affects memory constraints.  Proceedings of the Thirty-Seventh Annual Conference of the Cognitive Science Society.  (pp. 530-535).  Pasadena, CA: Cognitive Science Society.

We present evidence that successful chunk formation during a statistical learning task depends on how well the perceiver is able to parse the information that is presented between successive presentations of the to-be-learned chunk. First, we show that learners acquire a chunk better when the surrounding information is also chunk-able in a visual statistical learning task. We tested three process models of chunk formation, TRACX, PARSER, and MDLChunker, on our two different experimental conditions, and found that only PARSER and MDLChunker matched the observed result. These two models share the common principle of a memory capacity that is expanded as a result of learning. Though implemented in very different ways, both models effectively remember more individual items (the atomic components of a sequence) as additional chunks are formed. The ability to remember more information directly impacts learning in the models, suggesting that there is a positive-feedback loop in chunk learning.

Download a PDF of the paper here