• 1/21 Amy Needham - Duke University (Host: Robert Goldstone)
• 1/28 Tom Busey - with John Vanderkolk, Indiana State Police
• 2/4 Katy Börner
• 2/11 Mark Blair
• 2/18 Tony Chemero - University of Connecticut (Host: Andy Clark)
• 2/25 Shane Mueller
• 3/3 Heather Wild
• 3/10 Josh Goldberg
• 3/17 No meeting (Spring Break)
• 3/24 John Beggs
• 3/31 Eldad Yehiam
• 4/7 Amy Criss
• 4/14 Justin Kantner
• 4/21 Leslie Blaha
• 4/28 Mark Mon-Williams - University of Aberdeen (Host: Geoff Bingham)
• 5/5 No meeting (Finals Week)

Abstracts

1/28: Tom Busey with John Vanderkolk, Indiana State Police
Configural Processing in Expert Fingerprint Examiners: Evidence from Behavioral and EEG Measures

Studies conducted with Fingerprint examiners represent a superb opportunity to examine the results of extreme perceptual learning. These experts receive extensive training and have learned to extract the relevant signal from noise conditions that are highly variable. In this talk we examine how these conditions affect the information processing strategies adopted by fingerprint examiners. The results of a behavioral study suggest that experts can maintain images in memory for longer intervals, have greater tolerance to noise, and appear to adopt configural processing strategies when images are embedded in noise. We provide converging evidence for this conclusion with an EEG (brainwave) experiment that looks for the signatures of configural processing in the voltage data. At the end of the talk I hope to have time to discuss a future experiment that is designed to reveal the features that are used by examiners when matching prints.

2/4: Katy Börner
The Influence of Topics, Aging, and Recursive Linking on the Co-Evolution of Author and Paper Networks

There has been a long history of research into the structure and evolution of mankind’s scientific endeavor. However, recent progress in applying the tools of science to understand science itself has been unprecedented because only recently has there been access to high-volume and high-quality data sets of scientific output (e.g., publications, patents, grants), as well as computers and algorithms capable of handling this enormous stream of data. The talk will review major work on models that aim to capture and recreate the structure and dynamics of scientific evolution. We then introduce a general process model that simultaneously grows co-author and paper-citation networks. The statistical and dynamic properties of the networks generated by this model are validated against a 20-year data set of articles published in the Proceedings of the National Academy of Science. Systematic deviations from a power law distribution of citations to papers are well fit by a model that incorporates a partitioning of au thors and papers into topics, a bias for authors to cite recent papers, and a tendency for authors to cite papers cited by papers that they have read. In this TARL model (for Topics, Aging, and Recursive Linking), the number of topics is linearly related to the clustering coefficient of the simulated paper citation network.

2/11: Mark Blair
Accuracy and Simplicity: Opposing Goals in Adopting a Categorization Cue Set

The acquisition of expertise is a process in which increasing sensitivity to useful regularities in the environment leads to decreasing error. Categories that are overlapping when represented in one dimensionality may be separate in a higher-dimensional space. Three experiments are reported in which participants were shown an additional cue after learning to use 2 imperfect cues. The results reveal that participants can integrate new information into their categorization cue set. Wide individual differences are discovered, however, with many participants favoring simpler, but less accurate cue sets. Only some participants demonstrated the ability to discard information previously used when new, more accurate information was introduced. Current theories of selective attention that posit rapid shifts of learned attention show promise in accounting for these data.

2/25: Shane Mueller
The Case of the Disappearing Recency Effect and Other Mysteries of Strategy in Immediate Serial Recall

Across experiments using the immediate serial recall task, the recency effect ranges from being large (with the last item being recalled much better than the previous items) to small, disappearing entirely in some cases. This "disappearing recency effect" appears to stem from the adoption of different recall strategies in response to properties of the experiment and the instructions. I will show evidence I have found for this in a series of experiments in which the strategies participants use are manipulated through payoff and instructions. Along the way, I will also show a number of other ways in which participants' strategies affected performance in immediate serial recall. Finally, I will present a model using the EPIC computational architecture that allows these strategies to be represented explicitly, and demonstrate that by dealing with these strategies directly, we can also achieve a better understanding of the underlying structural limitations of verbal working memory.

3/3: Heather Wild
Electrophysiology and Psychophysics: Exploring Sensitivity and Bias in Face Perception

This work examines how cognitive processing related to sensitivity and bias in face perception is reflected in the EEG (electroencephalogram) signal. Behavioral indices and models of these aspects of processing are well-elucidated in psychophysics, but the link to EEG measures is not established. In this work, methodology from psychophysics is extended to ERP studies of face perception where bias and sensitivity effects are manifested in interesting ways. This methodology will potentially further understanding of the mechanisms underlying bias and sensitivity, specifically in the domain of face perception. Important applications also include testing of model assumptions based on inferences about internal distributions. Preliminary data illustrating this methodology will be presented and discussed.

3/10: Joshua Goldberg
Dynamical Field Theory Predicts a Developmental Reversal in an A-Not-B-Like Task

Dynamical Field Theory has made many predictions that go well beyond where conceptual models of Piaget's A-not-B task would lead. I will start by explaining how our Dynamical Field Model works. Then I will discuss a new variant on the A-not-B task. This new task pits against each other a cue and a distractor, occurring at different times. This lets us observe time-courses of competitive dynamics predicted by the Field Model that are not present in the basic A-not-B task.

Large numbers of simulations across a matrix of experimental conditions show regions of "parameter space" that have qualitatively different dynamics. Focusing on two small pieces of parameter space, we find an interesting prediction of a developmental reversal of the effectiveness of a distractor. (One type of distractor that works for young infants no longer works when they're old; another that does not affect young infants does work for the old ones.)

3/24: John Beggs
Information Storage and Avalanches in a Network of Cultured Cortical Neurons

The physiology of the cerebral cortex has been the subject of intense interest for many decades. This interest is perhaps driven by the desire to understand the mechanisms of higher cognitive functions that are dependent upon the cortex. While tremendous effort has been directed toward understanding cortical function at the cellular and large systems levels, much less experimental work has been done to uncover principles of neural interaction at the level of local cortical networks that comprise hundreds of neurons. We hypothesized that cortical networks were organized to accomplish at least three general tasks: Storing information, transmitting information, and maintaining network stability.

To investigate these issues, we cultured slices of rat cortex on 60 channel microelectrode arrays. Bursts of spontaneous electrical activity were recorded continuously after the cultures matured, and data were analyzed off line. Using this system, we found that spontaneous activity preferentially visited some network states more than would be expected by chance. These states were temporally precise and stable over a period of 10 hours, suggesting that they could be used by the network to store information. In addition, the number of electrodes activated in a burst of spontaneous activity was found to obey the same statistical rules that describe avalanches. Simulations of these avalanche dynamics showed that they were nearly optimal for information transmission and actually maintained network stability.

Thus, this work suggests that small cortical networks self-organize into a critical state that allows information storage, optimizes information transmission, and preserves stability. The principles governing cortical self-organization seem to be very general and share similarities with several non-biological phenomena like avalanches, earthquakes, nuclear fission and the spread of forest fires. Understanding emergent properties in these systems may therefore help us to understand emergent properties of cortical networks, and ultimately, the building blocks of cognition.

3/31: Eldad Yehiam
On the examination of basic assumptions embedded in learning models

The present study examines basic assumptions made by learning models for predicting behavior in decisions based on experience. In such decisions, the probabilities and payoffs are initially unknown and are learned from repeated choice with payoff feedback.

Different learning models can be conceived to predict players' behavior in these tasks. We systematically test the basic components embedded in these models. One component relates to the discounting of old information as new information is gathered. Another component relates to the choice rule for selecting alternatives. The result is a factorial-like design evaluating 8 (learning) x 3 (choice) models, which allows decomposing the unique contribution of each component of a model.

Two methods are employed to evaluate the success of learning models in approximating players' choices. The method of predicting one step ahead is compared to the simulation of an entire game path. The results indicate that the two methods lead to converging results with respect to the predictive power of basic assumptions.

4/7: Amy Criss, Krystal Klein, & Richard M. Shiffrin
What is the nature of temporal encoding? A look at judgments of recency

Models of memory have been very successful at explaining performance in certain memory tasks such as free recall or recognition. Here we explore a slightly different task where participants are presented with two studied words and are asked to judge which word occurred more recently in the studied list. The literature on these 'judgments of recency' is sparse and unconnected, and existing model proposals are vague. As such, we discuss three exploratory studies in light of a contextual hypothesis of recency judgments, showing that context does not seem to change much during a single study list, but can be forced to change under certain circumstances. Contributions of other theories to explaining the phenomenon of recency judgments will also be discussed.