Wayne at work.
Wayne at work

I am a Professor of Cognitive Science at Rensselaer Polytechnic Institute. Rensselaer is America’s oldest technological research university dating its founding to 1824. At Rensselaer I founded the CogWorks Laboratory (CWL), which has a productive history for research as well as of training PhDs. Today our research focuses on detailed studies of longitudinal changes in individual human performance -- especially performance in dynamic, real-time tasks -- tasks in which even hesitating requires a decision to hesitate. These types of tasks require us to focus on the mind’s eye and the mind’s hand (that is, the interaction of perception, action, and cognition) within dynamic, externally-paced, task environments.

In the 90’s, we modeled the interactions of single humans in complex interactions with various interactive systems as well as with other humans (e.g., human-human-computer interaction, see Gray & John, 1993). Later our work focused on identifying the common “interactive routines” that undergird much of daily interactive behavior (Gray & Boehm-Davis, 2000). In the early 2000’s the lab focused on the paradox of the active user -- namely “why” people who used the same software systems day-in and day-out performed suboptimally. Wai-Tat Fu and I found an answer to much of that (Fu & Gray, 2004) and my research group went on to develop the Soft Constraints Hypothesis (Gray & Fu, 2004; Gray, Sims, Fu, & Schoelles, 2006), which speaks, more generally, to the non-deliberate modality choices made by human cognition when alternative methods exist of performing the same task.

Until lately, most of ours as most of others’ work in human interactive behavior, implicitly assumed that advances in human cognitive science could be obtained by studying simple situations that could quickly be mastered by minimum training of individual humans who had no prior experience in the tasks they were being trained to do. Stated so baldly, that seems wrong. Today my CWL has invested in collecting massive amounts of data from individuals in the lab and harvesting massive amounts of data from individuals and teams from the web and elsewhere. Over the past 5-years, we have (a) trained people, in our lab, for 31 hours on a novel video game (Space Fortress), (b) “sampled expertise” by collecting data from people with vastly different amounts of prior experience playing the game “Tetris” (sampling individuals in our lab and at regional and international tournaments), and (c) accessing public APIs to download data from millions of the people who have played one to several thousands of hours of the game “League of Legends.” These studies have led us to formulate the “Plateau, Dips, & Leaps” framework (Gray & Lindstedt, 2017) that focuses on identifying -- for individual humans -- periods in which no progress is being made (plateaus), periods in which people discover or invent new methods (sometimes signaled by “dips” in performance), and periods in which new methods are implemented and performance soars (“leaps”) over what would have been expected by those slow and gradual forces postulated by the log-log law of learning. Our most recent publications focus on behavioral differences between humans and machine models in the play of Tetris (Sibert & Gray, 2018) and a principle component analysis (PCA) of an hour of data from each of 240 student players across a wide-span of human expertise. The PCA analysis revealed three components that, when applied to a multiple regression analysis across levels of Tetris play revealed interpretable differences between novice, average, and expert level student players (see Lindstedt & Gray, 2019, for a discussion of these points).

Professor Gray earned his Ph.D. from U. C. Berkeley in 1979. His first position was with the U. S. Army Research Institute where he worked on tactical team training (at the Monterey Field Unit) and later on the application of artificial intelligence (AI) technology to training for air-defense systems (HAWK) (at ARI-HQ Alexandria, VA). He spent a post-doctoral year with Prof. John R. Anderson's lab at Carnegie Mellon University before joining the AI Laboratory of NYNEX' Science & Technology Division. At NYNEX he applied cognitive task analysis and cognitive modeling to the design and evaluation of interfaces for large, commercial telecommunications systems. His academic career began at Fordham University and then moved to George Mason University. He joined the Cognitive Science Department at Rensselaer Polytechnic Institute in 2002.

Gray is a Fellow of the Cognitive Science Society, the Human Factors & Ergonomics Society (HFES), and the American Psychological Association (APA). In 2008, APA awarded him the Franklin V. Taylor Award for Outstanding Contributions in the Field of Applied Experimental & Engineering Psychology. He is a past Chair of the Cognitive Science Society and the founding Chair of the Human Performance Modeling technical group of HFES. At present he is the Executive Editor for the Cognitive Science Society’s first new journal in 30 years, Topics in Cognitive Science (topiCS). In 2012, he was elected a Fellow by the Alexander von Humboldt Foundation and spent his sabbatical in research at the Max Planck Institute Center for Adaptive Behavior and Cognition (ABC) in Berlin. Most recently, he received an IBM Faculty Award from IBM's Cognitive Systems Institute.


Images and Design © Robyn Gray 2008