David Eagleman - US grants

[unreadable] DESCRIPTION (provided by applicant): Understanding the timing of incoming neural signals is fundamental to moving, speaking and perception. To properly determine causality, animals must judge motor-sensory signals with the correct timing. However, this is not an easy task for brains to solve, because signals carried by different sensory modalities are processed at different speeds. The problem of sensory delays is compounded by factors ranging from changing lighting conditions to limb growth. This suggests that the brain continually recalibrates its motor- sensory timing judgments. Our preliminary studies seem to confirm this: we make the novel demonstration that an artificially injected delay between actions and effects leads to a recalibration of temporal order judgments. Under some circumstances this leads to an illusory reversal of action and effect. The goal of this proposal is to elucidate how the brain determines causality, and how it dynamically recalibrates that determination when the timing of feedback changes. We employ a battery of techniques including novel psycho physics, Virtual Reality and fMRI. We have four aims: (SA1) Demonstrate and quantify the conditions under which brains recalibrate motor-sensory timing. To this end, we have developed novel psychophysical experiments to analyze the effect of injected delays on temporal order judgments. (SA2) Determine the neural basis of causality timing judgments using fMRI. (SA3) Explore the neural basis of causality recalibration using fMRI and computational theory. (SA4) Root these findings in the context of timing in the brain more generally. Throughout the proposal, novel psychophysical findings are buttressed by brain imaging, and both sets of results steer our understanding of the neural mechanisms. Collectively, these experiments shed light on the dynamic temporal interactions of motor-sensory systems, leading to new insights about disorders that might be properly understood as disorders of temporal calibration. [unreadable] [unreadable] [unreadable]

Crime records are housed in hundreds of counties across the nation, each employing local laws and idiosyncratic information management systems. As a result, it is difficult to analyze detailed crime records across time and place. To address this challenge, the researcher proposes to build a database for integrating 23 million crime records in Houston, NY and Miami. The researcher proposes a workshop strategy to address three distinct aims: 1) Build an interdisciplinary community including statisticians, criminologists, legal scholars, behavioral scientists, neuroscientists, economists, sociologists and programmers, all of whom share an interest in the intersection of crime and their fields. 2) Work with the community to prototype a database of deep criminal records obtained via public access laws. 3) Work with the community to design advanced interface and visualization tools for the data, opening the data to the widest possible audience.

This project builds capacity for data-intensive research on national criminal records, guided at all stages by the stakeholder research communities. Previously, criminal research has relied on the FBI's Uniform Crime Reports, a tool with two weaknesses: no unique identifiers to identify re-offense rates, and a lack of detail that dilutes the ability to use 21st century tools and analysis techniques. The project breaks the information bottleneck by building a community of quantitative researchers - from a range of disciplines in the social, behavioral, and economic sciences - with a shared interest in next-generation questions at the intersection of crime and their fields. Through multiple workshops, the community establishes database ontologies and prototypes a transformative database of millions of criminal records spanning several decades (obtained via public access laws). The researchers also implement advanced interface and visualization tools to maximize the user audience for the database. This project provides an unprecedented level of detail about offenders, their crimes, and their interactions with the criminal justice system. By enabling an exploration of the relationships between external factors like legal policies and a decision to commit a crime, this project promises results across and beyond the contributing disciplines. Critically, the prototype database will include anonymized identifiers to enable exploration of desistance from crime, greatly advancing the study of recidivism (reoffenses). Such advances should in turn help efforts to identify high-risk offenders, allowing policy makers to base law enforcement decisions on direct, proven, open-source assessments of human behavior.

Scientists long have explored and measured a range of decision-making traits, such as the ability to inhibit impulses or to understand other people's emotions. Researchers have developed, validated, and refined a wide range of neurocognitive assessments, some of which have been used to try to make reasonable predictions of how dangerous prisoners will be in the future. This interdisciplinary research project will build on these previous efforts to test whether a newly developed tablet-based battery of tests can approve capabilities to improve prediction of future dangerousness by identifying and evaluating the relationships among several aspects of decision making and criminal recidivism. The assessment battery is designed to quantify an individual's decision-making profile, even if a user has limited reading proficiency and little to no previous experience with tablet computers. The project will explore a methodology than may be able to better tie details of individual decision making directly to past and future behavior. Although this project will focus on the topic of criminal recidivism, the software to be tested will provide an open-source, validated tool for correlating decision making with a variety of behavioral variables. Project findings will provide a deeper understanding of offenders' deficits, which may yield more rational, evidence-based policies aimed at the prevention and control of crime. The project also will provide education and training opportunities for graduate students and a post-doctoral researcher who are members of groups that are underrepresented in scientific research activities.

This project will focus on the core question of whether signatures of individual decision making can help to identify high-volume and high-risk criminal offenders. Past studies have employed a range of techniques, such as psychiatric assessment, behavioral inventories, and brain scanning, which have generally found that offenders have problems with anger, empathy, aggression, impulse-control, and self-management. The project advances efforts to directly measure those traits among offenders by accomplishing three specific aims: (1) finalize development of the battery, (2) validate it against standardized neurocognitive tests, and (3) quantify patterns of criminal recidivism by correlating decision-making metrics with future criminal activity of 300 released probationers. This project is supported through the NSF Interdisciplinary Behavioral and Social Sciences Research (IBSS) competition.