Meet Dr. Patrick Cox

Patrick Cox

Dr. Patrick Cox

Advisor: Dr. Stephen Mitroff

Institution: George Washington University 

Bio:  Patrick H. Cox earned his PhD in Neuroscience from Georgetown University in 2017 with Maximiliam Riesenhuber before becoming a postdoctoral researcher in the Department of Psychological & Brain Sciences at The George Washington University with Drs. Steve Mitroff and Dwight Kravtiz. He is currently an Intelligence Community Postdoctoral Fellow at GWU sponsored by the National Geospatial Agency. His research program investigates how past experience and current context interact to shape perception using a combination of behavioral big data, traditional psychophysics, computational modeling, and neural measures (MRI/EEG).

Abstract:  The National Geospatial Agency (NGA) oversees the critical national security task of analyzing complex images to identify potential threats and/or issues of interest to the intelligence community. This is an incredibly challenging search task, as targets of interest can be rare, variable in features (e.g., varying sizes, shapes, colors), can be embedded in any number of background environments, and there can be more than one target at a time. Decades of research have shown these factors can systematically reduce the likelihood of successfully detecting targets of interest. Since success of NGA’s mission ultimately relies on humans conducting complex searches of geospatial intelligence (GEOINT), a direct way to improve overall operational success is to optimize the acquisition of image analysis expertise. Visual search expertise arises through the interaction of innate ability and experience, so to optimize the acquisition of expertise the current project proposes you must find individuals with the most innate ability for the task (selection) and expose them to the ideal set of experiences to develop to their full potential (optimized training). An intriguing hypothesis, explored in the current project utilizing a behavioral big-data approach, is that there might be an ideal set of parameters that can lead to optimal performance when analysts are exposed to them in early search experiences—that there is, for all intents and purposes, a “critical period” early in training wherein exposure to the right set of search statistics will be more likely to ultimately obtain their full potential as an expert GEOINT analyst.