*Organizational Change *Program Evaluation Monitoring Organizational Behavior [3660]. Human. ACT-R associative strength cognitive architectures cognitive modeling cognitive skill acquisition Common Lisp computational linguistics DMAP dynamic task environment EPIC expertise extreme expertise GOMS GOMS/Analysis methods/Empirical studies/User models/Cognitive models/Methods for analysis or assessment/Prototyping/Protocol analysis/Theory in HCI interaction design interactive behavior leaps Measures of Semantic Relatedness natural language processing performance plateaus Rational analysis Semantic similarity measures soar Space Fortress Tetris Transfer tutorial VCR
2017 |
Gray, Wayne D; Lindstedt, John K Plateaus, Dips, and Leaps: Where to Look for Inventions and Discoveries during Skilled Performance Journal Article Cognitive Science, 41 (7), pp. 1838-1870, 2017. Abstract | Links | BibTeX | Tags: BreakOut, changepoint detection, digit span, dips, expertise, extreme expertise, leaps, performance, plateaus, Space Fortress @article{gray17csj-pdl, title = {Plateaus, Dips, and Leaps: Where to Look for Inventions and Discoveries during Skilled Performance}, author = {Wayne D. Gray and John K. Lindstedt}, url = {http://homepages.rpi.edu/~grayw/pubs/papers/2017/gray17csj-pdl.pdf}, doi = {10.1111/cogs.12412}, year = {2017}, date = {2017-10-18}, journal = {Cognitive Science}, volume = {41}, number = {7}, pages = {1838-1870}, abstract = {The framework of plateaus, dips, and leaps shines light on periods when individuals may be inventing new methods of skilled performance. We begin with a review of the role performance plateaus have played in (a) experimental psychology, (b) human--computer interaction, and (c) cognitive science. We then reanalyze two classic studies of individual performance to show plateaus and dips which resulted in performance leaps. For a third study, we show how the statistical methods of Changepoint Analysis plus a few simple heuristics may direct our focus to periods of performance change for individuals. For the researcher, dips become the marker of exploration where performance suffers as new methods are invented and tested. Leaps mark the implementation of a successful new method and an incremental jump above the path plotted by smooth and steady log--log performance increments. The methods developed during these dips and leaps are the key to surpassing one's teachers and acquiring extreme expertise.}, keywords = {BreakOut, changepoint detection, digit span, dips, expertise, extreme expertise, leaps, performance, plateaus, Space Fortress}, pubstate = {published}, tppubtype = {article} } The framework of plateaus, dips, and leaps shines light on periods when individuals may be inventing new methods of skilled performance. We begin with a review of the role performance plateaus have played in (a) experimental psychology, (b) human--computer interaction, and (c) cognitive science. We then reanalyze two classic studies of individual performance to show plateaus and dips which resulted in performance leaps. For a third study, we show how the statistical methods of Changepoint Analysis plus a few simple heuristics may direct our focus to periods of performance change for individuals. For the researcher, dips become the marker of exploration where performance suffers as new methods are invented and tested. Leaps mark the implementation of a successful new method and an incremental jump above the path plotted by smooth and steady log--log performance increments. The methods developed during these dips and leaps are the key to surpassing one's teachers and acquiring extreme expertise. |
Gigerenzer, Gerd ; Gray, Wayne D Topics in Cognitive Science, 9 (3), pp. 260–263, 2017. Abstract | Links | BibTeX | Tags: heuristics @article{Gigerenzer2017, title = {A Simple Heuristic Successfully Used by Humans, Animals, and Machines: The Story of the {RAF and Luftwaffe}, Hawks and Ducks, Dogs and Frisbees, Baseball Outfielders and {Sidewinder Missiles}---Oh My!}, author = {Gigerenzer, Gerd and Gray, Wayne D.}, url = {http://homepages.rpi.edu/~grayw/pubs/papers/2017/gigeGray17topiCS.pdf}, doi = {10.1111/tops.12269}, year = {2017}, date = {2017-04-15}, journal = {Topics in Cognitive Science}, volume = {9}, number = {3}, pages = {260--263}, abstract = {Please note that this paper is an introduction that we wrote to the next paper in the current (April 2017) issue of the journal Topics in Cognitive Science. Hamlin, R. P. (2017). ``The gaze heuristic:'' Biography of an adaptively rational decision process. Topics in Cognitive Science, 9(2):264--288. We do NOT actually cite that paper in our introduction as it immediately follows our introduction and the relationship is obvious in the online version of the journal. You can also find this Introduction as well as all other of my recent papers at: http://homepages.rpi.edu/~grayw/pubs/ Prof. Gigerenzer's works may be found at: https://www.mpib-berlin.mpg.de/de/publikationen/publikationen-im-volltext-in-auswahl Enjoy our Intro and don't forget to download and read the Hamlin's paper --- it is fascinating!}, keywords = {heuristics}, pubstate = {published}, tppubtype = {article} } Please note that this paper is an introduction that we wrote to the next paper in the current (April 2017) issue of the journal Topics in Cognitive Science. Hamlin, R. P. (2017). ``The gaze heuristic:'' Biography of an adaptively rational decision process. Topics in Cognitive Science, 9(2):264--288. We do NOT actually cite that paper in our introduction as it immediately follows our introduction and the relationship is obvious in the online version of the journal. You can also find this Introduction as well as all other of my recent papers at: http://homepages.rpi.edu/~grayw/pubs/ Prof. Gigerenzer's works may be found at: https://www.mpib-berlin.mpg.de/de/publikationen/publikationen-im-volltext-in-auswahl Enjoy our Intro and don't forget to download and read the Hamlin's paper --- it is fascinating! |
Gray, Wayne D Game-XP: Action Games as Experimental Paradigms for Cognitive Science Journal Article Topics in Cognitive Science, 9 (2), pp. 289–307, 2017. Abstract | Links | BibTeX | Tags: Action games, Chess, cognitive skill acquisition, Cohort analysis, Computer games, Expertise sampling, extreme expertise, Halo, Longitudinal studies, Skilled performance, Space Fortress, StarCraft, Tetris, Verbal protocol analysis, Video games @article{gray17topiCS.gxp, title = {Game-XP: Action Games as Experimental Paradigms for Cognitive Science}, author = {Gray, Wayne D.}, url = {http://homepages.rpi.edu/~grayw/pubs/papers/2017/gray17topics.gxp.pdf}, doi = {10.1111/tops.12260}, year = {2017}, date = {2017-04-15}, journal = {Topics in Cognitive Science}, volume = {9}, number = {2}, pages = {289--307}, abstract = {Why games? How could anyone consider action games an experimental paradigm for Cognitive Science? In 1973, as one of three strategies he proposed for advancing Cognitive Science, Allen Newell exhorted us to ``accept a single complex task and do all of it.'' More specifically, he told us that rather than taking an ``experimental psychology as usual approach,'' we should ``focus on a series of experimental and theoretical studies around a single complex task'' so as to demonstrate that our theories of human cognition were powerful enough to explain ``a genuine slab of human behavior'' with the studies fitting into a detailed theoretical picture. Action games represent the type of experimental paradigm that Newell was advocating and the current state of programming expertise and laboratory equipment, along with the emergence of Big Data and naturally occurring datasets, provide the technologies and data needed to realize his vision. Action games enable us to escape from our field's regrettable focus on novice performance to develop theories that account for the full range of expertise through a twin focus on expertise sampling (across individuals) and longitudinal studies (within individuals) of simple and complex tasks.}, keywords = {Action games, Chess, cognitive skill acquisition, Cohort analysis, Computer games, Expertise sampling, extreme expertise, Halo, Longitudinal studies, Skilled performance, Space Fortress, StarCraft, Tetris, Verbal protocol analysis, Video games}, pubstate = {published}, tppubtype = {article} } Why games? How could anyone consider action games an experimental paradigm for Cognitive Science? In 1973, as one of three strategies he proposed for advancing Cognitive Science, Allen Newell exhorted us to ``accept a single complex task and do all of it.'' More specifically, he told us that rather than taking an ``experimental psychology as usual approach,'' we should ``focus on a series of experimental and theoretical studies around a single complex task'' so as to demonstrate that our theories of human cognition were powerful enough to explain ``a genuine slab of human behavior'' with the studies fitting into a detailed theoretical picture. Action games represent the type of experimental paradigm that Newell was advocating and the current state of programming expertise and laboratory equipment, along with the emergence of Big Data and naturally occurring datasets, provide the technologies and data needed to realize his vision. Action games enable us to escape from our field's regrettable focus on novice performance to develop theories that account for the full range of expertise through a twin focus on expertise sampling (across individuals) and longitudinal studies (within individuals) of simple and complex tasks. |
Sibert, Catherine ; Gray, Wayne D; Lindstedt, John K Interrogating Feature Learning Models to Discover Insights Into the Development of Human Expertise in a Real-Time, Dynamic Decision-Making Task Journal Article Topics in Cognitive Science, 9 (2), pp. 374–394, 2017. Abstract | Links | BibTeX | Tags: Cognitive skill, cognitive skill acquisition, Cross-entropy reinforcement learning, expertise, Experts, Machine learning, Methods, Perceptual learning, Strategies, Tetris @article{sibert17topiCS.gxp, title = {Interrogating Feature Learning Models to Discover Insights Into the Development of Human Expertise in a Real-Time, Dynamic Decision-Making Task}, author = {Sibert, Catherine and Gray, Wayne D. and Lindstedt, John K.}, url = {http://homepages.rpi.edu/~grayw/pubs/papers/2017/sibert17topics.gxp.pdf}, doi = {10.1111/tops.12225}, year = {2017}, date = {2017-04-15}, journal = {Topics in Cognitive Science}, volume = {9}, number = {2}, pages = {374--394}, abstract = {Tetris provides a difficult, dynamic task environment within which some people are novices and others, after years of work and practice, become extreme experts. Here we study two core skills; namely, (a) choosing the goal or objective function that will maximize performance and (b) a feature-based analysis of the current game board to determine where to place the currently falling zoid (i.e., Tetris piece) so as to maximize the goal. In Study 1, we build cross-entropy reinforcement learning (CERL) models (Szita & Lorincz, 2006) to determine whether different goals result in different feature weights. Two of these optimization strategies quickly rise to performance plateaus, whereas two others continue toward higher but more jagged (i.e., variable) heights. In Study 2, we compare the zoid placement decisions made by our best CERL models with those made by 67 human players. Across 370,131 human game episodes, two CERL models picked the same zoid placements as our lowest scoring human for 43% of the placements and as our three best scoring experts for 65% of the placements. Our findings suggest that people focus on maximizing points, not number of lines cleared or number of levels reached. They also show that goal choice influences the choice of zoid placements for CERLs and suggest that the same is true of humans. Tetris has a repetitive task structure that makes Tetris more tractable and more like a traditional experimental psychology paradigm than many more complex games or tasks. Hence, although complex, Tetris is not overwhelmingly complex and presents a right-sized challenge to cognitive theories, especially those of integrated cognitive systems.}, keywords = {Cognitive skill, cognitive skill acquisition, Cross-entropy reinforcement learning, expertise, Experts, Machine learning, Methods, Perceptual learning, Strategies, Tetris}, pubstate = {published}, tppubtype = {article} } Tetris provides a difficult, dynamic task environment within which some people are novices and others, after years of work and practice, become extreme experts. Here we study two core skills; namely, (a) choosing the goal or objective function that will maximize performance and (b) a feature-based analysis of the current game board to determine where to place the currently falling zoid (i.e., Tetris piece) so as to maximize the goal. In Study 1, we build cross-entropy reinforcement learning (CERL) models (Szita & Lorincz, 2006) to determine whether different goals result in different feature weights. Two of these optimization strategies quickly rise to performance plateaus, whereas two others continue toward higher but more jagged (i.e., variable) heights. In Study 2, we compare the zoid placement decisions made by our best CERL models with those made by 67 human players. Across 370,131 human game episodes, two CERL models picked the same zoid placements as our lowest scoring human for 43% of the placements and as our three best scoring experts for 65% of the placements. Our findings suggest that people focus on maximizing points, not number of lines cleared or number of levels reached. They also show that goal choice influences the choice of zoid placements for CERLs and suggest that the same is true of humans. Tetris has a repetitive task structure that makes Tetris more tractable and more like a traditional experimental psychology paradigm than many more complex games or tasks. Hence, although complex, Tetris is not overwhelmingly complex and presents a right-sized challenge to cognitive theories, especially those of integrated cognitive systems. |
Gray, Wayne D Plateaus and Asymptotes: Spurious and Real Limits in Human Performance Journal Article Current Directions in Psychological Science, 26 (1), pp. 59-67, 2017. Abstract | Links | BibTeX | Tags: asymptotes, cognitive skill acquisition, expertise, memory, performance, plateaus, spurious limits @article{gray17cdps, title = {Plateaus and Asymptotes: Spurious and Real Limits in Human Performance}, author = {Gray, Wayne D.}, url = {http://homepages.rpi.edu/~grayw/pubs/papers/2017/gray17cdps.pdf}, doi = {10.1177/0963721416672904}, year = {2017}, date = {2017-02-15}, journal = {Current Directions in Psychological Science}, volume = {26}, number = {1}, pages = {59-67}, abstract = {One hundred twenty years ago, the emergent field of experimental psychology debated whether plateaus of performance during training were real or not. Sixty years ago, the battle was over whether learning asymptoted or not. Thirty years ago, the research community was seized with concerns over stable plateaus at suboptimal performance levels among experts. Applied researchers viewed this as a systems problem and referred to it as the paradox of the active user. Basic researchers diagnosed this as a training problem and embraced deliberate practice. The concepts of plateaus and asymptotes and the distinction between the two are important as the questions asked and the means of overcoming one or the other differ. These questions have meaning as we inquire about the nature of performance limits in skilled behavior and the distinction between brain capacity and brain efficiency. This article brings phenomena that are hiding in the open to the attention of the research community in the hope that delineating the distinction between plateaus and asymptotes will help clarify the distinction between real versus ``spurious limits'' and advance theoretical debates regarding learning and performance.}, keywords = {asymptotes, cognitive skill acquisition, expertise, memory, performance, plateaus, spurious limits}, pubstate = {published}, tppubtype = {article} } One hundred twenty years ago, the emergent field of experimental psychology debated whether plateaus of performance during training were real or not. Sixty years ago, the battle was over whether learning asymptoted or not. Thirty years ago, the research community was seized with concerns over stable plateaus at suboptimal performance levels among experts. Applied researchers viewed this as a systems problem and referred to it as the paradox of the active user. Basic researchers diagnosed this as a training problem and embraced deliberate practice. The concepts of plateaus and asymptotes and the distinction between the two are important as the questions asked and the means of overcoming one or the other differ. These questions have meaning as we inquire about the nature of performance limits in skilled behavior and the distinction between brain capacity and brain efficiency. This article brings phenomena that are hiding in the open to the attention of the research community in the hope that delineating the distinction between plateaus and asymptotes will help clarify the distinction between real versus ``spurious limits'' and advance theoretical debates regarding learning and performance. |
Veksler, Bella Z; Boyd, Rachel ; Myers, Christopher W; Gunzelmann, Glenn ; Neth, Hansjörg ; Gray, Wayne D Visual Working Memory Resources Are Best Characterized as Dynamic, Quantifiable Mnemonic Traces Journal Article Topics in Cognitive Science, 9 (1), pp. 1-19, 2017. Abstract | Links | BibTeX | Tags: ACT-R, Eye tracking, Resource allocation, visual search, Visual working memory @article{bella17topiCS, title = {Visual Working Memory Resources Are Best Characterized as Dynamic, Quantifiable Mnemonic Traces}, author = {Veksler, Bella Z. and Boyd, Rachel and Myers, Christopher W. and Gunzelmann, Glenn and Neth, Hansjörg and Gray, Wayne D.}, doi = {10.1111/tops.12248}, year = {2017}, date = {2017-01-15}, journal = {Topics in Cognitive Science}, volume = {9}, number = {1}, pages = {1-19}, abstract = {Visual working memory (VWM) is a construct hypothesized to store a small amount of accurate perceptual information that can be brought to bear on a task. Much research concerns the construct's capacity and the precision of the information stored. Two prominent theories of VWM representation have emerged: slot-based and continuous-resource mechanisms. Prior modeling work suggests that a continuous resource that varies over trials with variable capacity and a potential to make localization errors best accounts for the empirical data. Questions remain regarding the variability in VWM capacity and precision. Using a novel eye-tracking paradigm, we demonstrate that VWM facilitates search and exhibits effects of fixation frequency and recency, particularly for prior targets. Whereas slot-based memory models cannot account for the human data, a novel continuous-resource model does capture the behavioral and eye tracking data, and identifies the relevant resource as item activation.}, keywords = {ACT-R, Eye tracking, Resource allocation, visual search, Visual working memory}, pubstate = {published}, tppubtype = {article} } Visual working memory (VWM) is a construct hypothesized to store a small amount of accurate perceptual information that can be brought to bear on a task. Much research concerns the construct's capacity and the precision of the information stored. Two prominent theories of VWM representation have emerged: slot-based and continuous-resource mechanisms. Prior modeling work suggests that a continuous resource that varies over trials with variable capacity and a potential to make localization errors best accounts for the empirical data. Questions remain regarding the variability in VWM capacity and precision. Using a novel eye-tracking paradigm, we demonstrate that VWM facilitates search and exhibits effects of fixation frequency and recency, particularly for prior targets. Whereas slot-based memory models cannot account for the human data, a novel continuous-resource model does capture the behavioral and eye tracking data, and identifies the relevant resource as item activation. |
2016 |
Sangster, Matthew-Donald D; Mendonca, David J; Gray, Wayne D Big Data Meets Team Expertise in a Dynamic Task Environment Conference Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 60 (1), Sage , 2016. Abstract | BibTeX | Tags: dynamic task environment, League of Legends, LoL, team, team expertise @conference{sangster16hfes, title = {Big Data Meets Team Expertise in a Dynamic Task Environment}, author = {Sangster, Matthew-Donald D. and Mendonca, David J. and Gray, Wayne D.}, year = {2016}, date = {2016-09-21}, booktitle = {Proceedings of the Human Factors and Ergonomics Society Annual Meeting}, volume = {60}, number = {1}, pages = {158-162}, publisher = {Sage }, abstract = {Objective; This research employs large-scale data from a massively multiplayer online game to examine the links between the composition, processes and outcomes of teams operating in high tempo, data-rich environments. Background: Research on the performance of teams-- particularly over long time scales--is often expensive and time-consuming. But Big Data from competitive, team-based games can mitigate these costs. Methods: Data visualization techniques are used to explore team data harvested from publicly accessible sources for the online game League of Legends™, one of the most popular such games in the world. Results: The exploratory results suggest potentially complex relationships between team composition, processes and outcomes, and in particular how team composition and process may unfold over longer time spans. Conclusions: The results point to the potentially substantial benefits of large-scale studies of teamwork, and--in parallel--to the need for the development of tools, techniques and measures to bring Big Data to bear in teamwork studies. Application: This work demonstrates the feasibility of exploring online gaming data for new insights into team and individual performance.}, keywords = {dynamic task environment, League of Legends, LoL, team, team expertise}, pubstate = {published}, tppubtype = {conference} } Objective; This research employs large-scale data from a massively multiplayer online game to examine the links between the composition, processes and outcomes of teams operating in high tempo, data-rich environments. Background: Research on the performance of teams-- particularly over long time scales--is often expensive and time-consuming. But Big Data from competitive, team-based games can mitigate these costs. Methods: Data visualization techniques are used to explore team data harvested from publicly accessible sources for the online game League of Legends™, one of the most popular such games in the world. Results: The exploratory results suggest potentially complex relationships between team composition, processes and outcomes, and in particular how team composition and process may unfold over longer time spans. Conclusions: The results point to the potentially substantial benefits of large-scale studies of teamwork, and--in parallel--to the need for the development of tools, techniques and measures to bring Big Data to bear in teamwork studies. Application: This work demonstrates the feasibility of exploring online gaming data for new insights into team and individual performance. |
Destefano, Marc ; Gray, Wayne D Proceedings of the 38th Annual Conference of the Cognitive Science Society, Cognitive Science Society, Austin, TX, 2016. Abstract | Links | BibTeX | Tags: changepoint analysis, dips, expertise, leaps, method invention, performance, plateaus, SAX, skill acquisition, Space Fortress, strategy discovery @conference{marc16csc, title = {Where Should Researchers Look for Strategy Discoveries during the Acquisition of Complex Task Performance? The Case of Space Fortress}, author = {Destefano, Marc and Gray, Wayne D.}, editor = {Papafragou, A. and Grodner, D. and Mirman, D. and Trueswell, J. C.}, url = {http://homepages.rpi.edu/~grayw/pubs/papers/2016/marc16csc.pdf}, year = {2016}, date = {2016-08-05}, booktitle = {Proceedings of the 38th Annual Conference of the Cognitive Science Society}, publisher = {Cognitive Science Society}, address = {Austin, TX}, abstract = {In complex task domains, such as games, students may exceed their teachers. Such tasks afford diverse means to tradeoff one type of performance for another, combining task elements in novel ways to yield method variations and strategy discoveries that, if mastered, might produce large or small leaps in performance. For the researcher interested in the development of extreme expertise in the wild, the problem posed by such tasks is ``where to look'' to capture the explorations, trials, errors, and successes that eventually lead to the invention of superior performance. In this paper, we present several successful discoveries of methods for superior performance. For these discoveries we used Symbolic Aggregate Approximation as our method of identifying changepoints within score progressions in the venerable game of Space Fortress. By decomposing performance at these changepoints, we find previously unknown strategies that even the designers of the task had not anticipated.}, keywords = {changepoint analysis, dips, expertise, leaps, method invention, performance, plateaus, SAX, skill acquisition, Space Fortress, strategy discovery}, pubstate = {published}, tppubtype = {conference} } In complex task domains, such as games, students may exceed their teachers. Such tasks afford diverse means to tradeoff one type of performance for another, combining task elements in novel ways to yield method variations and strategy discoveries that, if mastered, might produce large or small leaps in performance. For the researcher interested in the development of extreme expertise in the wild, the problem posed by such tasks is ``where to look'' to capture the explorations, trials, errors, and successes that eventually lead to the invention of superior performance. In this paper, we present several successful discoveries of methods for superior performance. For these discoveries we used Symbolic Aggregate Approximation as our method of identifying changepoints within score progressions in the venerable game of Space Fortress. By decomposing performance at these changepoints, we find previously unknown strategies that even the designers of the task had not anticipated. |
Hope, Ryan M Cognitive Control of Saccadic Behavior in the Antisaccade Task: A model of Voluntary and Involuntary Eye Movements PhD Thesis Rensselaer Polytechnic Institute, 2016. Abstract | BibTeX | Tags: ABS-CRISP, CRISP, eye fixations, eye movements, saccades @phdthesis{hope16phdThesis, title = {Cognitive Control of Saccadic Behavior in the Antisaccade Task: A model of Voluntary and Involuntary Eye Movements}, author = {Hope, Ryan M.}, year = {2016}, date = {2016-04-11}, address = {Troy, NY}, school = {Rensselaer Polytechnic Institute}, abstract = {Performance detriments in the antisaccade task have been linked to numerous psychiatric and neurological disorders yet, there is no consensus as to how healthy individuals perform the task. Most computational models of the antisaccade task assume that cue onset automatically triggers programming of a prosaccade towards the cue and that successfully performing an antisaccade away from the cue requires top-down inhibition of the erroneous prosaccade before a correct antisaccade can be made. However, a recent body of research on oculomotor control suggests that humans have much less control over their eye movements. Eye trackers have revealed that the eyes are in constant motion, even when fixating, and that these fixational eye movements are possibly functional. The growing consensus is that saccades are initiated automatically by a rhythmic trigger from the brainstem. An important question now is how does a system based on automatic (involuntary) saccade timing still allow for top-down (voluntary) control, like that which is needed in the antisaccade task? In order to test this idea, a new model called ABS(Attention Biased Salience)-CRISP was created which builds upon the CRISP (Nuthmann, Smith, Engbert, & Henderson, 2010) model of saccade generation (which models the automatic saccade timer as a random walk process) by adding a spatial component that computes the saccade target location as the weighted sum of a bottom-up saliency map and a top-down attentional map. The CRISP and ABS-CRISP models were evaluated and compared to human performance in a mixed-block antisaccade task. The ABS-CRISP model was able to replicate individual distributions of saccade latencies that were indistinguishable from a majority of the subjects data. The results support the idea that the initiation of saccade timing is not tied to cognitive events that occur during fixations but instead, are triggered by a random timer. The results also support the idea that interindividual, intra-individual and inter-task differences in performance can be explained, in large, by changes in the bias between bottom-up and top-down information in the spatial component of saccade programming.}, keywords = {ABS-CRISP, CRISP, eye fixations, eye movements, saccades}, pubstate = {published}, tppubtype = {phdthesis} } Performance detriments in the antisaccade task have been linked to numerous psychiatric and neurological disorders yet, there is no consensus as to how healthy individuals perform the task. Most computational models of the antisaccade task assume that cue onset automatically triggers programming of a prosaccade towards the cue and that successfully performing an antisaccade away from the cue requires top-down inhibition of the erroneous prosaccade before a correct antisaccade can be made. However, a recent body of research on oculomotor control suggests that humans have much less control over their eye movements. Eye trackers have revealed that the eyes are in constant motion, even when fixating, and that these fixational eye movements are possibly functional. The growing consensus is that saccades are initiated automatically by a rhythmic trigger from the brainstem. An important question now is how does a system based on automatic (involuntary) saccade timing still allow for top-down (voluntary) control, like that which is needed in the antisaccade task? In order to test this idea, a new model called ABS(Attention Biased Salience)-CRISP was created which builds upon the CRISP (Nuthmann, Smith, Engbert, & Henderson, 2010) model of saccade generation (which models the automatic saccade timer as a random walk process) by adding a spatial component that computes the saccade target location as the weighted sum of a bottom-up saliency map and a top-down attentional map. The CRISP and ABS-CRISP models were evaluated and compared to human performance in a mixed-block antisaccade task. The ABS-CRISP model was able to replicate individual distributions of saccade latencies that were indistinguishable from a majority of the subjects data. The results support the idea that the initiation of saccade timing is not tied to cognitive events that occur during fixations but instead, are triggered by a random timer. The results also support the idea that interindividual, intra-individual and inter-task differences in performance can be explained, in large, by changes in the bias between bottom-up and top-down information in the spatial component of saccade programming. |
2015 |
, Proceedings of the 13th International Conference on Cognitive Modeling Proceeding Rijsuniversiteit Groningen Groningen, de Nederland, 2015. BibTeX | Tags: @proceedings{conf:iccm15, title = {Proceedings of the 13th International Conference on Cognitive Modeling}, author = { }, editor = {Niels A. Taatgen and Marieke K. van Vugt and Jeimer P. Borst and Katja Mehlhorn}, year = {2015}, date = {2015-04-01}, booktitle = {Proceedings of the 13th International Conference on Cognitive Modeling}, address = {Groningen, de Nederland}, organization = {Rijsuniversiteit Groningen}, keywords = {}, pubstate = {published}, tppubtype = {proceedings} } |
Gray, Wayne D Asymptotes, Plateaus, and Limits to Human Performance Conference Presentation to the IBM Cognitive Systems Institute, IBM Cognitive Systems Institute 2015. Abstract | Links | BibTeX | Tags: @conference{gray15ibm, title = {Asymptotes, Plateaus, and Limits to Human Performance}, author = { Wayne D. Gray}, doi = {10.13140/2.1.1603.1529}, year = {2015}, date = {2015-02-01}, booktitle = {Presentation to the IBM Cognitive Systems Institute}, organization = {IBM Cognitive Systems Institute}, abstract = {120 years ago the emergent field of experimental psychology became embroiled in debates as to whether plateaus in performance are real (or not) and if so whether they were due to periods in which league-stepping methods (originally defined as a hierarchy of habits that enabled experts to step leagues while novices were ``bustling over furlongs or inches'') were being acquired (or not). 20 years ago both the human-computer interaction and cognitive science communities were seized with concerns over performance plateaus (i.e., extended periods of stable suboptimal performance) from experts. I briefly review this history with the aim of drawing distinctions between performance asymptotes and performance plateaus, and argue that remediating one is the domain of design while remediating the other is the domain of training.}, keywords = {}, pubstate = {published}, tppubtype = {conference} } 120 years ago the emergent field of experimental psychology became embroiled in debates as to whether plateaus in performance are real (or not) and if so whether they were due to periods in which league-stepping methods (originally defined as a hierarchy of habits that enabled experts to step leagues while novices were ``bustling over furlongs or inches'') were being acquired (or not). 20 years ago both the human-computer interaction and cognitive science communities were seized with concerns over performance plateaus (i.e., extended periods of stable suboptimal performance) from experts. I briefly review this history with the aim of drawing distinctions between performance asymptotes and performance plateaus, and argue that remediating one is the domain of design while remediating the other is the domain of training. |
Hope, Ryan M; Gray, Wayne D A scanpath algorithm for dynamic regions of interest and complex task environments Incollection European Conference on Eye Movements, University of Vienna, Austria, 2015. BibTeX | Tags: @incollection{hope15ecem, title = {A scanpath algorithm for dynamic regions of interest and complex task environments}, author = { Ryan M. Hope and Wayne D. Gray}, year = {2015}, date = {2015-01-01}, booktitle = {European Conference on Eye Movements}, publisher = {University of Vienna}, address = {Austria}, keywords = {}, pubstate = {published}, tppubtype = {incollection} } |
Lindstedt, John K; Gray, Wayne D MetaT: Tetris as an Experimental Paradigm for Cognitive Skills Research Journal Article Behavior Research Methods, 2015. @article{lindstedt15brm, title = {MetaT: Tetris as an Experimental Paradigm for Cognitive Skills Research}, author = { John K. Lindstedt and Wayne D. Gray}, doi = {10.3758/s13428-014-0547-y}, year = {2015}, date = {2015-01-01}, journal = {Behavior Research Methods}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
Neth, Hansj"org; Sims, Chris R; Gray, Wayne D Rational Task Analysis: A Methodology to Benchmark Bounded Rationality Journal Article Minds and Machines, pp. 1-24, 2015, ISSN: 0924-6495. Links | BibTeX | Tags: Bounded rationality; Benchmarking; Optimality; Task environment; Rational analysis; Ecological rationality @article{neth15mm, title = {Rational Task Analysis: A Methodology to Benchmark Bounded Rationality}, author = { Hansj"org Neth and Chris R. Sims and Wayne D. Gray}, url = {http://dx.doi.org/10.1007/s11023-015-9368-8}, doi = {10.1007/s11023-015-9368-8}, issn = {0924-6495}, year = {2015}, date = {2015-01-01}, journal = {Minds and Machines}, pages = {1-24}, publisher = {Springer Netherlands}, keywords = {Bounded rationality; Benchmarking; Optimality; Task environment; Rational analysis; Ecological rationality}, pubstate = {published}, tppubtype = {article} } |
Sibert, Catherine; Gray, Wayne D; Lindstedt, John K Tetris: Exploring Human Performance via Cross Entropy Reinforcement Learning Models Inproceedings submitted for Proceedings of the 38th Annual Conference of the Cognitive Science Society, 2015. @inproceedings{sibert15csc, title = {Tetris: Exploring Human Performance via Cross Entropy Reinforcement Learning Models}, author = { Catherine Sibert and Wayne D. Gray and John K. Lindstedt}, year = {2015}, date = {2015-01-01}, booktitle = {submitted for Proceedings of the 38th Annual Conference of the Cognitive Science Society}, abstract = {What can a machine learning simulation tell us about human performance in a complex, real-time task such as Tetris? Although Tetris is often used as a research tool (Mayer, 2014), the strategies and methods used by Tetris players have seldom been the explicit focus of study. In Study 1, we use cross-entropy reinforcement learning (CERL) (Szita & Lorincz, 2006; Thiery & Scherrer, 2009) to explore (a) the utility of high-level strategies (goals or objective functions) for maximizing performance and (b) a variety of features and feature-weights (methods) for optimizing a low-level, 1-zoid optimization strategy. Two of these optimization strategies quickly rise to performance plateaus, whereas two others continued towards higher but more jagged (i.e., variable) plateaus. In Study 2, we compare the zoid (i.e., Tetris piece) placement decisions made by our best CERL models with those made by the full spectrum of novice-to-expert human Tetris players. Across 370,131 episodes collected from 67 human players, the ability of two CERL strategies to classify human zoid placements varied with player expertise from 43% for our lowest scoring novice to around 65% for our three highest scoring experts.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } What can a machine learning simulation tell us about human performance in a complex, real-time task such as Tetris? Although Tetris is often used as a research tool (Mayer, 2014), the strategies and methods used by Tetris players have seldom been the explicit focus of study. In Study 1, we use cross-entropy reinforcement learning (CERL) (Szita & Lorincz, 2006; Thiery & Scherrer, 2009) to explore (a) the utility of high-level strategies (goals or objective functions) for maximizing performance and (b) a variety of features and feature-weights (methods) for optimizing a low-level, 1-zoid optimization strategy. Two of these optimization strategies quickly rise to performance plateaus, whereas two others continued towards higher but more jagged (i.e., variable) plateaus. In Study 2, we compare the zoid (i.e., Tetris piece) placement decisions made by our best CERL models with those made by the full spectrum of novice-to-expert human Tetris players. Across 370,131 episodes collected from 67 human players, the ability of two CERL strategies to classify human zoid placements varied with player expertise from 43% for our lowest scoring novice to around 65% for our three highest scoring experts. |
2014 |
Gray, Wayne D What can Video Games Tell Us About Extreme Expertise? Conference Talk Presented at the Rice University Psychology Department Speakers Series, Rice University Houston, TX, 2014. @conference{gray14.Rice, title = {What can Video Games Tell Us About Extreme Expertise?}, author = { Wayne D. Gray}, year = {2014}, date = {2014-11-01}, booktitle = {Talk Presented at the Rice University Psychology Department Speakers Series}, address = {Houston, TX}, organization = {Rice University}, abstract = {We are interested in studying the acquisition and deployment of extreme expertise in tasks entailing the real-time interaction of a single human with a complex, dynamic decision environment. Our dilemma is the skills that we wish to generalize to (tasks such as helicopter piloting, laparoscopic surgery, and air traffic control) require hundreds if not thousands of hours to achieve expertise and the people who already possess such skills are very rare in the college population and too expensive to bring into our laboratory. Our solution to this dilemma is to study expert and novice video game players. In this talk I will provide a broad overview of three convergent lines of research on Tetristexttrademark players, with the primary focus being on our analyses of expert vs novice differences in eye data. We maintain that cognitive science has for far too long been fixated on isolating small components of individual cognition, that such an approach has the danger of overfitting our theories to our paradigms, and that the way out of this dilemma is to bring (a) powerful computational modeling, (b) machine learning techniques, and (c) human performance from extreme performers together to test and extend cognitive theory.}, keywords = {}, pubstate = {published}, tppubtype = {conference} } We are interested in studying the acquisition and deployment of extreme expertise in tasks entailing the real-time interaction of a single human with a complex, dynamic decision environment. Our dilemma is the skills that we wish to generalize to (tasks such as helicopter piloting, laparoscopic surgery, and air traffic control) require hundreds if not thousands of hours to achieve expertise and the people who already possess such skills are very rare in the college population and too expensive to bring into our laboratory. Our solution to this dilemma is to study expert and novice video game players. In this talk I will provide a broad overview of three convergent lines of research on Tetristexttrademark players, with the primary focus being on our analyses of expert vs novice differences in eye data. We maintain that cognitive science has for far too long been fixated on isolating small components of individual cognition, that such an approach has the danger of overfitting our theories to our paradigms, and that the way out of this dilemma is to bring (a) powerful computational modeling, (b) machine learning techniques, and (c) human performance from extreme performers together to test and extend cognitive theory. |
Gray, Wayne D; Hope, Ryan M; Lindstedt, John K; Destefano, Marc Plenary Presentation at the 12th Biannual Meeting of the German Cognitive Science Society, Universit"at T"ubingen T"ubingen, Germany, 2014. Abstract | Links | BibTeX | Tags: @conference{gray14.KogWis, title = {Elements of Extreme Expertise: Searching for Differences in Microstrategies Deployed by Experts and Novices}, author = { Wayne D. Gray and Ryan M. Hope and John K. Lindstedt and Marc Destefano}, doi = {10.13140/2.1.3809.1529}, year = {2014}, date = {2014-10-01}, booktitle = {Plenary Presentation at the 12th Biannual Meeting of the German Cognitive Science Society}, address = {T"ubingen, Germany}, organization = {Universit"at T"ubingen}, abstract = {We are studying the acquisition and deployment of extreme expertise during the real-time interaction of a single human with complex, dynamic decision environments. Our dilemma is that people who have the specific skills we wish to generalize to (such as helicopter piloting, laparoscopic surgery, and air traffic control) are very rare in the college population and too expensive to bring into our lab. Our solution has been to study expert and novice video game players. Our approach takes the position that Cognitive Science has been overly fixated on isolating small components of individual cognition. That approach runs the danger of emphoverfitting theories to paradigms. Our way out of this dilemma is to bring together (a) powerful computational models, (b) machine learning techniques, and (c) microanalysis techniques that integrate analyses of cognitive, perceptual, and action data collected from extreme performers to develop, test, and extend cognitive theory. Since our January 2013 start, we have built our experimental paradigm, collected naturalistic and laboratory data, published journal and conference papers, won Rensselaer Undergraduate research prizes, developed ``single-piece optimizers'' (SPOs, i.e., machine learning systems), compared machine performers to human performers, and begun analyzing eye and behavioral data from two 6hr human studies. Our tasks have been the games of Tetris and Space Fortress. Future plan include (a) using our SPOs to tutor piece-by-piece placement, (b) developing integrated cognitive models that account for cognition, action, and perception, and (c) continued exploration of the differences between good players and extreme experts in Tetris and Space Fortress. Games such as Tetris and Space Fortress are often dismissed as ``merely requiring reflex behavior.'' However, with an estimated total number of board configurations of $2^199$ (approx. 8 followed by 59 zeros), Tetris cannot be ``merely reflect behavior.'' Our preliminary analyses show complex goal hierarchies, dynamic ``two-piece'' plans that are updated after every episode, sophisticated use of subgoaling, and the gradual adaptation of strategies and plans as the speed of play increases. These are very sophisticated, human strategies, beyond our current capability to model, and are challenging topic for the study of the emphElements of Extreme Expertise.}, keywords = {}, pubstate = {published}, tppubtype = {conference} } We are studying the acquisition and deployment of extreme expertise during the real-time interaction of a single human with complex, dynamic decision environments. Our dilemma is that people who have the specific skills we wish to generalize to (such as helicopter piloting, laparoscopic surgery, and air traffic control) are very rare in the college population and too expensive to bring into our lab. Our solution has been to study expert and novice video game players. Our approach takes the position that Cognitive Science has been overly fixated on isolating small components of individual cognition. That approach runs the danger of emphoverfitting theories to paradigms. Our way out of this dilemma is to bring together (a) powerful computational models, (b) machine learning techniques, and (c) microanalysis techniques that integrate analyses of cognitive, perceptual, and action data collected from extreme performers to develop, test, and extend cognitive theory. Since our January 2013 start, we have built our experimental paradigm, collected naturalistic and laboratory data, published journal and conference papers, won Rensselaer Undergraduate research prizes, developed ``single-piece optimizers'' (SPOs, i.e., machine learning systems), compared machine performers to human performers, and begun analyzing eye and behavioral data from two 6hr human studies. Our tasks have been the games of Tetris and Space Fortress. Future plan include (a) using our SPOs to tutor piece-by-piece placement, (b) developing integrated cognitive models that account for cognition, action, and perception, and (c) continued exploration of the differences between good players and extreme experts in Tetris and Space Fortress. Games such as Tetris and Space Fortress are often dismissed as ``merely requiring reflex behavior.'' However, with an estimated total number of board configurations of $2^199$ (approx. 8 followed by 59 zeros), Tetris cannot be ``merely reflect behavior.'' Our preliminary analyses show complex goal hierarchies, dynamic ``two-piece'' plans that are updated after every episode, sophisticated use of subgoaling, and the gradual adaptation of strategies and plans as the speed of play increases. These are very sophisticated, human strategies, beyond our current capability to model, and are challenging topic for the study of the emphElements of Extreme Expertise. |
, The 36th Annual Meeting of the Cognitive Science Society Inproceedings Bello, Paul; Guarini, Marcello; McShane, Marjorie; Scassellati, Brian (Ed.): Proceedings of the 36th Annual Conference of the Cognitive Science Society, Cognitive Science Society, Austin, TX, 2014. BibTeX | Tags: @inproceedings{conf:cogsci14, title = {The 36th Annual Meeting of the Cognitive Science Society}, author = { }, editor = {Paul Bello and Marcello Guarini and Marjorie McShane and Brian Scassellati}, year = {2014}, date = {2014-07-01}, booktitle = {Proceedings of the 36th Annual Conference of the Cognitive Science Society}, publisher = {Cognitive Science Society}, address = {Austin, TX}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Hope, Ryan M; Schoelles, Michael J; Gray, Wayne D Simplifying the interaction between cognitive models and task environments with the JSON Network Interface Journal Article Behavior Research Methods, 46 , pp. 1007-1012, 2014. Abstract | Links | BibTeX | Tags: Cognitive architecture; ACT-R; EPIC; IPC; TCP; JSON; Common Lisp; Python @article{hope14brm, title = {Simplifying the interaction between cognitive models and task environments with the JSON Network Interface}, author = { Ryan M. Hope and Michael J. Schoelles and Wayne D. Gray}, url = {http://dx.doi.org/10.3758/s13428-013-0425-z}, doi = {10.3758/s13428-013-0425-z}, year = {2014}, date = {2014-01-01}, journal = {Behavior Research Methods}, volume = {46}, pages = {1007-1012}, publisher = {Springer US}, abstract = {Process models of cognition, written in architectures such as ACT-R and EPIC, should be able to interact with the same software with which human subjects interact. By eliminating the need to simulate the experiment, this approach would simplify the modeler's effort, while ensuring that all steps required of the human are also required by the model. In practice, the difficulties of allowing one software system to interact with another present a significant barrier to any modeler who is not also skilled at this type of programming. The barrier increases if the programming language used by the modeling software differs from that used by the experimental software. The JSON Network Interface simplifies this problem for ACT-R modelers, and potentially, modelers using other systems.}, keywords = {Cognitive architecture; ACT-R; EPIC; IPC; TCP; JSON; Common Lisp; Python}, pubstate = {published}, tppubtype = {article} } Process models of cognition, written in architectures such as ACT-R and EPIC, should be able to interact with the same software with which human subjects interact. By eliminating the need to simulate the experiment, this approach would simplify the modeler's effort, while ensuring that all steps required of the human are also required by the model. In practice, the difficulties of allowing one software system to interact with another present a significant barrier to any modeler who is not also skilled at this type of programming. The barrier increases if the programming language used by the modeling software differs from that used by the experimental software. The JSON Network Interface simplifies this problem for ACT-R modelers, and potentially, modelers using other systems. |
Gittelson, Logan; Lindstedt, John K; Sibert, Catherine; Gray, Wayne D Comparing Reinforcement Learning in Humans and Artificial Intelligence Through Tetris Inproceedings 2014. BibTeX | Tags: @inproceedings{gittelson14csc.poster, title = {Comparing Reinforcement Learning in Humans and Artificial Intelligence Through Tetris}, author = { Logan Gittelson and John K. Lindstedt and Catherine Sibert and Wayne D. Gray}, year = {2014}, date = {2014-01-01}, crossref = {conf:cogsci14}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Gray, Wayne D Elements of Extreme Expertise in Cognitive Skill: Studies of Video Gamers Conference Talk Presented at the Union College Psychology Department Speakers Series and Honors Colloquium, Union College Schenectady, NY, 2014. BibTeX | Tags: @conference{gray14.UnionCollege, title = {Elements of Extreme Expertise in Cognitive Skill: Studies of Video Gamers}, author = { Wayne D. Gray}, year = {2014}, date = {2014-01-01}, booktitle = {Talk Presented at the Union College Psychology Department Speakers Series and Honors Colloquium}, address = {Schenectady, NY}, organization = {Union College}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Gray, Wayne D; Hills, Thomas Does Cognition Deteriorate with Age or is it Enhanced by Experience? Journal Article Topics in Cognitive Science, 6 (1), pp. 2-4, 2014. Links | BibTeX | Tags: Cognitive aging, Cognitive functioning, Domain knowledge, Longitudinal change, Normal aging, Rational analysis @article{gray14topiCS, title = {Does Cognition Deteriorate with Age or is it Enhanced by Experience?}, author = { Wayne D. Gray and Thomas Hills}, url = {http://dx.doi.org/10.1111/tops.12080}, doi = {10.1111/tops.12080}, year = {2014}, date = {2014-01-01}, journal = {Topics in Cognitive Science}, volume = {6}, number = {1}, pages = {2-4}, keywords = {Cognitive aging, Cognitive functioning, Domain knowledge, Longitudinal change, Normal aging, Rational analysis}, pubstate = {published}, tppubtype = {article} } |
Gray, Wayne D Introduction to Volume 6, Issue 2 of topiCS Journal Article Topics in Cognitive Science, 6 (2), pp. 197–197, 2014, ISSN: 1756-8765. @article{gray14topiCS62, title = {Introduction to Volume 6, Issue 2 of topiCS}, author = { Wayne D. Gray}, url = {http://dx.doi.org/10.1111/tops.12091}, doi = {10.1111/tops.12091}, issn = {1756-8765}, year = {2014}, date = {2014-01-01}, journal = {Topics in Cognitive Science}, volume = {6}, number = {2}, pages = {197--197}, keywords = {}, pubstate = {published}, tppubtype = {article} } |
2013 |
Gray, Wayne D Human Interactions in Abstract Visual Spaces Conference Dagstuhl Seminar on Interaction with Information for Visual Reasoning, Schloss Dagstuhl -- Lebniz-Zentrum f"ur Informatik Germany, 2013. @conference{gray13Dagstuhl, title = {Human Interactions in Abstract Visual Spaces}, author = { Wayne D. Gray}, year = {2013}, date = {2013-08-01}, booktitle = {Dagstuhl Seminar on Interaction with Information for Visual Reasoning}, address = {Germany}, organization = {Schloss Dagstuhl -- Lebniz-Zentrum f"ur Informatik}, abstract = {Human behavior is interactive behavior. Behavior emerges from the interaction of bounded cognition with the natural or designed task environment and task goals. EXTENDED ABSTRACT: Human behavior is interactive behavior. Behavior emerges from the interaction of bounded cognition with the natural or designed task environment and task goals. Topics covered included: (a) Interactive routines, (b) The eye-hand span, (c) Analogies for memory, (d) Local, not global optimization, (e) Modeling the whole human (not just the convenient bits!), (f) Modeling pre-attentive and attentive visual processes, (g) Tools for the statistical analysis of visual saliency and similarity, (h) Tools for analyzing eye data, and (i) Possibilities for more indepth talks including emphThe Cognitive Science of Natural Interaction and emphElements of Extreme Expertise.}, keywords = {}, pubstate = {published}, tppubtype = {conference} } Human behavior is interactive behavior. Behavior emerges from the interaction of bounded cognition with the natural or designed task environment and task goals. EXTENDED ABSTRACT: Human behavior is interactive behavior. Behavior emerges from the interaction of bounded cognition with the natural or designed task environment and task goals. Topics covered included: (a) Interactive routines, (b) The eye-hand span, (c) Analogies for memory, (d) Local, not global optimization, (e) Modeling the whole human (not just the convenient bits!), (f) Modeling pre-attentive and attentive visual processes, (g) Tools for the statistical analysis of visual saliency and similarity, (h) Tools for analyzing eye data, and (i) Possibilities for more indepth talks including emphThe Cognitive Science of Natural Interaction and emphElements of Extreme Expertise. |
, Proceedings of the 12th International Conference on Cognitive Modeling Proceeding Carleton University Ottawa, Canada, 2013. BibTeX | Tags: @proceedings{conf:iccm13, title = {Proceedings of the 12th International Conference on Cognitive Modeling}, author = { }, editor = {Robert West and Terry Stewart}, year = {2013}, date = {2013-07-01}, booktitle = {Proceedings of the 12th International Conference on Cognitive Modeling}, address = {Ottawa, Canada}, organization = {Carleton University}, keywords = {}, pubstate = {published}, tppubtype = {proceedings} } |
Neth, Hansj"org; Sims, Chris R; Gray, Wayne D Rational Task Analysis: Eine Methode zur Erforschung Adaptiver Kognition Conference Albert-Ludwigs-Universit"at Freiburg, Germany 2013. BibTeX | Tags: @conference{neth13talk, title = {Rational Task Analysis: Eine Methode zur Erforschung Adaptiver Kognition}, author = { Hansj"org Neth and Chris R. Sims and Wayne D. Gray}, year = {2013}, date = {2013-07-01}, organization = {Albert-Ludwigs-Universit"at Freiburg, Germany}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
, The 35th Annual Meeting of the Cognitive Science Society Inproceedings Knauff, Markus; Pauen, Michael; Sebanz, Natalie; Wachsmuth, Ipke (Ed.): Proceedings of the 35th Annual Meeting of the Cognitive Science Society, Cognitive Science Society, Austin, TX, 2013. BibTeX | Tags: @inproceedings{conf:cogsci13, title = {The 35th Annual Meeting of the Cognitive Science Society}, author = { }, editor = {Markus Knauff and Michael Pauen and Natalie Sebanz and Ipke Wachsmuth}, year = {2013}, date = {2013-07-01}, booktitle = {Proceedings of the 35th Annual Meeting of the Cognitive Science Society}, publisher = {Cognitive Science Society}, address = {Austin, TX}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Gray, Wayne D Elements of Extreme Expertise in Cognitive Skill Conference Talk Presented at the Center for Cognitive Science, Albert-Ludwigs-Universit"at Freiburg, Germany 2013. BibTeX | Tags: @conference{gray13.Freiburg, title = {Elements of Extreme Expertise in Cognitive Skill}, author = { Wayne D. Gray}, year = {2013}, date = {2013-07-01}, booktitle = {Talk Presented at the Center for Cognitive Science}, organization = {Albert-Ludwigs-Universit"at Freiburg, Germany}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
, The European ACT-R Workshop Inproceedings Taatgen, Niels; van Rijn, Hedderik; Borst., Jelmer (Ed.): The European ACT-R Workshop, Rijsuniversiteit Groningen 2013. BibTeX | Tags: @inproceedings{conf:europeanActr:13, title = {The European ACT-R Workshop}, author = { }, editor = {Niels Taatgen and Hedderik van Rijn and Jelmer Borst.}, year = {2013}, date = {2013-04-01}, booktitle = {The European ACT-R Workshop}, organization = {Rijsuniversiteit Groningen}, series = {Groningen, The Netherlands}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } |
Gray, Wayne D Elements of Extreme Expertise in Cognitive Skill Conference Talk Presented at the Naval Submarine Medical Research Laboratory, Groton, CT, 2013. BibTeX | Tags: @conference{gray13.Groton, title = {Elements of Extreme Expertise in Cognitive Skill}, author = { Wayne D. Gray}, year = {2013}, date = {2013-04-01}, booktitle = {Talk Presented at the Naval Submarine Medical Research Laboratory}, address = {Groton, CT}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Lindstedt, John K; Gray, Wayne D Extreme expertise: Exploring expert behavior in Tetris Inproceedings pp. 912-917, 2013. @inproceedings{lindstedt13csc, title = {Extreme expertise: Exploring expert behavior in Tetris}, author = { John K. Lindstedt and Wayne D. Gray}, year = {2013}, date = {2013-01-01}, pages = {912-917}, crossref = {conf:cogsci13}, abstract = {Expertise is easy to identify in retrospect. It is the most expert player who wins the meet and the most proficient team that wins the playoffs. However, sometimes during play we see a masterful move that clearly separates one player from the competition. Our goal, in this work, is to identify the masterful moves or elements of expertise that predict the continuum of performance in the game of Tetris. As a first step we have collected data from a wide variety of Tetris Tournament players and used it to derive metrics of global, local, and immediate interactions. Here we present statistical models of these data and report the initial success of these models at predicting level of expertise.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Expertise is easy to identify in retrospect. It is the most expert player who wins the meet and the most proficient team that wins the playoffs. However, sometimes during play we see a masterful move that clearly separates one player from the competition. Our goal, in this work, is to identify the masterful moves or elements of expertise that predict the continuum of performance in the game of Tetris. As a first step we have collected data from a wide variety of Tetris Tournament players and used it to derive metrics of global, local, and immediate interactions. Here we present statistical models of these data and report the initial success of these models at predicting level of expertise. |
Sims, Chris R; Neth, Hansj"org; Jacobs, Robert A; Gray, Wayne D Melioration as Rational Choice: Sequential Decision Making in Uncertain Environments Journal Article Psychological Review, 120 (1), pp. 139-154, 2013. Abstract | BibTeX | Tags: Bayesian modeling, Melioration, Rational analysis, sequential decision making @article{sims13psycRvw, title = {Melioration as Rational Choice: Sequential Decision Making in Uncertain Environments}, author = { Chris R. Sims and Hansj"org Neth and Robert A. Jacobs and Wayne D. Gray}, year = {2013}, date = {2013-01-01}, journal = {Psychological Review}, volume = {120}, number = {1}, pages = {139-154}, abstract = {Melioration -- defined as choosing a lesser, local gain over a greater longer term gain -- is a behavioral tendency that people and pigeons share. As such, the empirical occurrence of meliorating behavior has frequently been interpreted as evidence that the mechanisms of human choice violate the norms of economic rationality. In some environments, the relationship between actions and outcomes is known. In this case, the rationality of choice behavior can be evaluated in terms of how successfully it maximizes utility given knowledge of the environmental contingencies. In most complex environments, however, the relationship between actions and future outcomes is uncertain and must be learned from experience. When the difficulty of this learning challenge is taken into account, it is not evident that melioration represents suboptimal choice behavior. In the present article, we examine human performance in a sequential decision-making experiment that is known to induce meliorating behavior. In keeping with previous results using this paradigm, we find that the majority of participants in the experiment fail to adopt the optimal decision strategy and instead demonstrate a significant bias toward melioration. To explore the origins of this behavior, we develop a rational analysis (Anderson, 1990) of the learning problem facing individuals in uncertain decision environments. Our analysis demonstrates that an unbiased learner would adopt melioration as the optimal response strategy for maximizing long-term gain. We suggest that many documented cases of melioration can be reinterpreted not as irrational choice but rather as globally optimal choice under uncertainty.}, keywords = {Bayesian modeling, Melioration, Rational analysis, sequential decision making}, pubstate = {published}, tppubtype = {article} } Melioration -- defined as choosing a lesser, local gain over a greater longer term gain -- is a behavioral tendency that people and pigeons share. As such, the empirical occurrence of meliorating behavior has frequently been interpreted as evidence that the mechanisms of human choice violate the norms of economic rationality. In some environments, the relationship between actions and outcomes is known. In this case, the rationality of choice behavior can be evaluated in terms of how successfully it maximizes utility given knowledge of the environmental contingencies. In most complex environments, however, the relationship between actions and future outcomes is uncertain and must be learned from experience. When the difficulty of this learning challenge is taken into account, it is not evident that melioration represents suboptimal choice behavior. In the present article, we examine human performance in a sequential decision-making experiment that is known to induce meliorating behavior. In keeping with previous results using this paradigm, we find that the majority of participants in the experiment fail to adopt the optimal decision strategy and instead demonstrate a significant bias toward melioration. To explore the origins of this behavior, we develop a rational analysis (Anderson, 1990) of the learning problem facing individuals in uncertain decision environments. Our analysis demonstrates that an unbiased learner would adopt melioration as the optimal response strategy for maximizing long-term gain. We suggest that many documented cases of melioration can be reinterpreted not as irrational choice but rather as globally optimal choice under uncertainty. |
Veksler, Vladislav Daniel; Gray, Wayne D; Schoelles, Michael J Goal-proximity decision making Journal Article Cognitive Science, 37 (4), pp. 605-774, 2013. @article{vdv13cogSci, title = {Goal-proximity decision making}, author = { Vladislav Daniel Veksler and Wayne D. Gray and Michael J. Schoelles}, year = {2013}, date = {2013-01-01}, journal = {Cognitive Science}, volume = {37}, number = {4}, pages = {605-774}, abstract = {Reinforcement learning (RL) models of decision-making cannot account for human decisions in the absence of prior reward or punishment. We propose a mechanism for choosing among available options based on goal-option association strengths, where association strengths between objects represent previously experienced object proximity. The proposed mechanism, Goal-Proximity Decision-making (GPD), is implemented within the ACT-R cognitive framework. GPD is found to be more efficient than RL in three maze-navigation simulations. GPD advantages over RL seem to grow as task difficulty is increased. An experiment is presented where participants are asked to make choices in the absence of prior reward. GPD captures human performance in this experiment better than RL.}, keywords = {}, pubstate = {published}, tppubtype = {article} } Reinforcement learning (RL) models of decision-making cannot account for human decisions in the absence of prior reward or punishment. We propose a mechanism for choosing among available options based on goal-option association strengths, where association strengths between objects represent previously experienced object proximity. The proposed mechanism, Goal-Proximity Decision-making (GPD), is implemented within the ACT-R cognitive framework. GPD is found to be more efficient than RL in three maze-navigation simulations. GPD advantages over RL seem to grow as task difficulty is increased. An experiment is presented where participants are asked to make choices in the absence of prior reward. GPD captures human performance in this experiment better than RL. |
Gray, Wayne D; Schoelles, Michael J Elements of Extreme Expertise in Cognitive Skill Technical Report Rensselaer Polytechnic Institute (ONR-N00014-13-1-0252, RPI-A12477), 2013. @techreport{gray13onr, title = {Elements of Extreme Expertise in Cognitive Skill}, author = { Wayne D. Gray and Michael J. Schoelles}, year = {2013}, date = {2013-01-01}, number = {ONR-N00014-13-1-0252, RPI-A12477}, institution = {Rensselaer Polytechnic Institute}, abstract = {We propose to study the cognitive elements that lead to extreme expertise in skilled performance. The tasks we examine will require the real-time interaction of a single human with a complex, dynamic decision environment. Initial performance will be poor, but over the course of practice, which will span hours and days, some people will achieve mastery and others will not. Successful mastery will require the discovery and timely deployment of strategies and microprocedures, which will require real-time problem solving as well as the skillful interweaving of cognitive, perceptual, and motor operations. Our data collection and analyses will examine cognitive elements at the millisecond level to understand changes at the level of minutes, hours, and days. Beyond the task of collecting, analyzing, and tabulating human data in these demanding tasks, we strive for predictive validity; by which we mean, the development of computational cognitive models that not only perform the task, but which show the same developmental trajectory across practice sessions as humans. Our goal is to develop cognitive theories and modeling formalisms powerful enough to understand the acquisition of expertise in tasks given to college students that point the way to developing powerful training programs for tasks necessary for eet operations.}, keywords = {}, pubstate = {published}, tppubtype = {techreport} } We propose to study the cognitive elements that lead to extreme expertise in skilled performance. The tasks we examine will require the real-time interaction of a single human with a complex, dynamic decision environment. Initial performance will be poor, but over the course of practice, which will span hours and days, some people will achieve mastery and others will not. Successful mastery will require the discovery and timely deployment of strategies and microprocedures, which will require real-time problem solving as well as the skillful interweaving of cognitive, perceptual, and motor operations. Our data collection and analyses will examine cognitive elements at the millisecond level to understand changes at the level of minutes, hours, and days. Beyond the task of collecting, analyzing, and tabulating human data in these demanding tasks, we strive for predictive validity; by which we mean, the development of computational cognitive models that not only perform the task, but which show the same developmental trajectory across practice sessions as humans. Our goal is to develop cognitive theories and modeling formalisms powerful enough to understand the acquisition of expertise in tasks given to college students that point the way to developing powerful training programs for tasks necessary for eet operations. |
Hope, Ryan M; Gray, Wayne D Eye movement optimization in visual search Inproceedings pp. 609-614, 2013. Abstract | BibTeX | Tags: eye movements, microstrategy, optimization, return saccades, visual search @inproceedings{hope13csc, title = {Eye movement optimization in visual search}, author = { Ryan M. Hope and Wayne D. Gray}, year = {2013}, date = {2013-01-01}, pages = {609-614}, crossref = {conf:cogsci14}, abstract = {In the present study we investigated whether eye movements in visual search are optimized to reduce time on task. Subjects task was to find a target object in a large field of objects that differed based on shape, color, size and numeric label. The target specification was manipulated, directly influencing the average number of fixations it took subjects to find the target object. Although a microstrategy that allowed for parallel saccade programming and information processing was found to be more ecient in terms of time, a serial microstrategy where saccade programming always follows information processing was found to be the more prevalent microstrategy.}, keywords = {eye movements, microstrategy, optimization, return saccades, visual search}, pubstate = {published}, tppubtype = {inproceedings} } In the present study we investigated whether eye movements in visual search are optimized to reduce time on task. Subjects task was to find a target object in a large field of objects that differed based on shape, color, size and numeric label. The target specification was manipulated, directly influencing the average number of fixations it took subjects to find the target object. Although a microstrategy that allowed for parallel saccade programming and information processing was found to be more ecient in terms of time, a serial microstrategy where saccade programming always follows information processing was found to be the more prevalent microstrategy. |
2012 |
Gray, Wayne D; Schoelles, Michael J; Hope, Ryan M; Lindstedt, John K What Games Can Tell Us About the Acquisition and Transfer of Cognitive Skills Conference ONR Review, Office of Naval Research Arlington, VA, 2012. BibTeX | Tags: @conference{gray12onr.Arlington, title = {What Games Can Tell Us About the Acquisition and Transfer of Cognitive Skills}, author = { Wayne D. Gray and Michael J. Schoelles and Ryan M. Hope and John K. Lindstedt}, year = {2012}, date = {2012-10-01}, booktitle = {ONR Review}, address = {Arlington, VA}, organization = {Office of Naval Research}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Gray, Wayne D; Schoelles, Michael J Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload Conference Talk Presented at Arizona State University on 2012-09Sep-18, Office of Naval Research Arlington, VA, 2012. BibTeX | Tags: @conference{gray12onr.Phoenix, title = {Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload}, author = { Wayne D. Gray and Michael J. Schoelles}, year = {2012}, date = {2012-09-01}, booktitle = {Talk Presented at Arizona State University on 2012-09Sep-18}, address = {Arlington, VA}, organization = {Office of Naval Research}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Gray, Wayne D; Ralph, Jason Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload Conference Talk Presented at Technical University of Berlin on 2012-06Jun-13, Technical University of Berlin Berlin, Germany, 2012. BibTeX | Tags: @conference{gray12.TU, title = {Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload}, author = { Wayne D. Gray and Jason Ralph}, year = {2012}, date = {2012-06-01}, booktitle = {Talk Presented at Technical University of Berlin on 2012-06Jun-13}, address = {Berlin, Germany}, organization = {Technical University of Berlin}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Gray, Wayne D; Ralph, Jason Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload Conference Talk Presented at Central European University, Budapest on 2012-05May-07, Central European University Budapest, Hungary, 2012. BibTeX | Tags: @conference{gray12.CEU, title = {Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload}, author = { Wayne D. Gray and Jason Ralph}, year = {2012}, date = {2012-05-01}, booktitle = {Talk Presented at Central European University, Budapest on 2012-05May-07}, address = {Budapest, Hungary}, organization = {Central European University}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Schoelles, Michael J; Gray, Wayne D SimPilot: An exploration of modeling a highly interactive task with delayed feedback in a multitasking environment Inproceedings 12th International Conference on Cognitive Modeling, Berlin, Germany, 2012. Abstract | BibTeX | Tags: delayed feedback, Multitasking; interactive behavior, threaded cognition; cognitive control; task switching @inproceedings{schoelles12iccm, title = {SimPilot: An exploration of modeling a highly interactive task with delayed feedback in a multitasking environment}, author = { Michael J. Schoelles and Wayne D. Gray}, year = {2012}, date = {2012-04-01}, booktitle = {12th International Conference on Cognitive Modeling}, address = {Berlin, Germany}, abstract = {Taxiing an airplane at a major airport requires the pilot to interact the world outside the cockpit, the instrumentations within the cockpit, and the co-pilot. As many actions require time to pass before their outcome can be evaluated, the pilot must have an approximate sense of how much delay should occur before the outcome of an action can be evaluated. Finally, taxiing is a paradigmatic example of multitasking. These three ingredients (a) a high level of interaction with dynamic task environments, (b) a sense of time, and (c) multitasking, present challenges for theories of cognition and the building of process models of taxiing. We describe a model, SimPilot, its initial validation, and its implications for cognitive theory.}, keywords = {delayed feedback, Multitasking; interactive behavior, threaded cognition; cognitive control; task switching}, pubstate = {published}, tppubtype = {inproceedings} } Taxiing an airplane at a major airport requires the pilot to interact the world outside the cockpit, the instrumentations within the cockpit, and the co-pilot. As many actions require time to pass before their outcome can be evaluated, the pilot must have an approximate sense of how much delay should occur before the outcome of an action can be evaluated. Finally, taxiing is a paradigmatic example of multitasking. These three ingredients (a) a high level of interaction with dynamic task environments, (b) a sense of time, and (c) multitasking, present challenges for theories of cognition and the building of process models of taxiing. We describe a model, SimPilot, its initial validation, and its implications for cognitive theory. |
Gray, Wayne D; Kaber, David; Gil, Guk-Ho; Kim, Sang-Hwan; Cao, Shi; Liu, Yili; Gonzalez, Cleotilde; Gunzelmann, Glenn; Gluck, Kevin Symposium on Human Performance Modeling Inproceedings 12th International Conference on Cognitive Modeling, Berlin, Germany, 2012. Abstract | BibTeX | Tags: human factors, human performance modeling @inproceedings{gray12iccm, title = {Symposium on Human Performance Modeling}, author = { Wayne D. Gray and David Kaber and Guk-Ho Gil and Sang-Hwan Kim and Shi Cao and Yili Liu and Cleotilde Gonzalez and Glenn Gunzelmann and Kevin Gluck}, year = {2012}, date = {2012-04-01}, booktitle = {12th International Conference on Cognitive Modeling}, address = {Berlin, Germany}, abstract = {This symposium is co-sponsored by the Human Performance Modeling Technical Group (HPM-TG) of the Human Factors & Ergonomics Society. Three Research Talks and a Panel Discussion were presented. Each talk used a different style of cognitive modeling and addressed a different problem of interest to the human factors community. For the Panel Discussion, three additional members of the HPM-TG joined with our three speakers in a round table discussion of the similarities and differences between cognitive modeling in applied versus basic science.}, keywords = {human factors, human performance modeling}, pubstate = {published}, tppubtype = {inproceedings} } This symposium is co-sponsored by the Human Performance Modeling Technical Group (HPM-TG) of the Human Factors & Ergonomics Society. Three Research Talks and a Panel Discussion were presented. Each talk used a different style of cognitive modeling and addressed a different problem of interest to the human factors community. For the Panel Discussion, three additional members of the HPM-TG joined with our three speakers in a round table discussion of the similarities and differences between cognitive modeling in applied versus basic science. |
Gray, Wayne D; Ralph, Jason Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload Conference Talk Presented at Max-Planck-Institut f"ur Bildungsforschung on 2012-03Mar-27, Max-Planck-Institut f"ur Bildungsforschung Berlin, Germany, 2012. BibTeX | Tags: @conference{gray12.MPIB, title = {Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload}, author = { Wayne D. Gray and Jason Ralph}, year = {2012}, date = {2012-03-01}, booktitle = {Talk Presented at Max-Planck-Institut f"ur Bildungsforschung on 2012-03Mar-27}, address = {Berlin, Germany}, organization = {Max-Planck-Institut f"ur Bildungsforschung}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Gray, Wayne D; Ralph, Jason Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload Conference Talk Presented at University College London on 2012-03Mar-07, University College London London, England, 2012. BibTeX | Tags: @conference{gray12.UCL, title = {Contributions of Body-Bound & Brain-Bound Resources, and Control Processes to Cognitive Workload}, author = { Wayne D. Gray and Jason Ralph}, year = {2012}, date = {2012-03-01}, booktitle = {Talk Presented at University College London on 2012-03Mar-07}, address = {London, England}, organization = {University College London}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Gray, Wayne D; Lindstedt, John K; Hope, Ryan M; Destefano, Marc An update on the whys and wherefores of Space Fortress 5 and the mysteries of Tetris Conference Talk Presented at University College London on 2012-03Mar-06, University College London London, England, 2012. BibTeX | Tags: @conference{gray12a.UCL, title = {An update on the whys and wherefores of Space Fortress 5 and the mysteries of Tetris}, author = { Wayne D. Gray and John K. Lindstedt and Ryan M. Hope and Marc Destefano}, year = {2012}, date = {2012-03-01}, booktitle = {Talk Presented at University College London on 2012-03Mar-06}, address = {London, England}, organization = {University College London}, keywords = {}, pubstate = {published}, tppubtype = {conference} } |
Janssen, Christian P; Gray, Wayne D When, what, and how much to reward in reinforcement learning based models of cognition Journal Article Cognitive Science, 36 (2), pp. 333-358, 2012. Abstract | Links | BibTeX | Tags: Reinforcement learning; Choice; Strategy selection; Adaptive behavior; Expected utility; Expected value; Cognitive architecture; Skill acquisition and learning @article{janssen12csj, title = {When, what, and how much to reward in reinforcement learning based models of cognition}, author = { Christian P. Janssen and Wayne D. Gray}, doi = {10.1111/j.1551-6709.2011.01222.x}, year = {2012}, date = {2012-01-01}, journal = {Cognitive Science}, volume = {36}, number = {2}, pages = {333-358}, abstract = {Reinforcement learning approaches to cognitive modeling represent task acquisition as learning to choose the sequence of steps that accomplishes the task while maximizing a reward. However, an apparently unrecognized problem for modelers is choosing when, what, and how much to reward; that is, when (the moment: end of trial, subtask, or some other interval of task performance), what (the objective function: e.g., performance time or performance accuracy), and how much (the magnitude: with binary, categorical, or continuous values). In this article, we explore the problem space of these three parameters in the context of a task whose completion entails some combination of 36 state--action pairs, where all intermediate states (i.e., after the initial state and prior to the end state) represent progressive but partial completion of the task. Different choices produce profoundly different learning paths and outcomes, with the strongest effect for moment. Unfortunately, there is little discussion in the literature of the effect of such choices. This absence is disappointing, as the choice of when, what, and how much needs to be made by a modeler for every learning model.}, keywords = {Reinforcement learning; Choice; Strategy selection; Adaptive behavior; Expected utility; Expected value; Cognitive architecture; Skill acquisition and learning}, pubstate = {published}, tppubtype = {article} } Reinforcement learning approaches to cognitive modeling represent task acquisition as learning to choose the sequence of steps that accomplishes the task while maximizing a reward. However, an apparently unrecognized problem for modelers is choosing when, what, and how much to reward; that is, when (the moment: end of trial, subtask, or some other interval of task performance), what (the objective function: e.g., performance time or performance accuracy), and how much (the magnitude: with binary, categorical, or continuous values). In this article, we explore the problem space of these three parameters in the context of a task whose completion entails some combination of 36 state--action pairs, where all intermediate states (i.e., after the initial state and prior to the end state) represent progressive but partial completion of the task. Different choices produce profoundly different learning paths and outcomes, with the strongest effect for moment. Unfortunately, there is little discussion in the literature of the effect of such choices. This absence is disappointing, as the choice of when, what, and how much needs to be made by a modeler for every learning model. |
John, Bonnie E; Patton, Evan W; Gray, Wayne D; Morrison, Donald F Tools for Predicting the Duration and Variability of Skilled Performance without Skilled Performers Journal Article Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 56 (1), pp. 985-989, 2012. Abstract | Links | BibTeX | Tags: {CogTool, SANLab} @article{bej12hfes, title = {Tools for Predicting the Duration and Variability of Skilled Performance without Skilled Performers}, author = { Bonnie E. John and Evan W. Patton and Wayne D. Gray and Donald F. Morrison}, doi = {10.1177/1071181312561206}, year = {2012}, date = {2012-01-01}, booktitle = {56th Annual Conference of the Human Factors & Ergonomics Society}, journal = {Proceedings of the Human Factors and Ergonomics Society Annual Meeting}, volume = {56}, number = {1}, pages = {985-989}, publisher = {HFES}, address = {Santa Monica, CA}, abstract = {Many devices are designed to allow skilled users to complete routine tasks quickly, often within a specified amount of time. Predictive human performance modeling has long been able to predict the mean time to accomplish a task, making it possible to compare device designs before building them. However, estimates of the variability of performance are also important, especially in real-time, safety-critical tasks. Until recently, the human factors community lacked tools to predict the variability of skilled performance. In this paper, we describe a combination of theory-based tools (CogTool and SANLab) that address this critical gap and that can easily be used by human factors practitioners or system designers. We describe these tools, their integration, and provide a concrete example of their use in the context of entering the landing speed into the Boeing 777 Flight Management Computer (FMC) using the Control and Display Unit (CDU).}, keywords = {{CogTool, SANLab}}, pubstate = {published}, tppubtype = {article} } Many devices are designed to allow skilled users to complete routine tasks quickly, often within a specified amount of time. Predictive human performance modeling has long been able to predict the mean time to accomplish a task, making it possible to compare device designs before building them. However, estimates of the variability of performance are also important, especially in real-time, safety-critical tasks. Until recently, the human factors community lacked tools to predict the variability of skilled performance. In this paper, we describe a combination of theory-based tools (CogTool and SANLab) that address this critical gap and that can easily be used by human factors practitioners or system designers. We describe these tools, their integration, and provide a concrete example of their use in the context of entering the landing speed into the Boeing 777 Flight Management Computer (FMC) using the Control and Display Unit (CDU). |
Patton, Evan W; Gray, Wayne D; John, Bonnie E Automated CPM-GOMS Modeling from Human Data Journal Article Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 56 (1), pp. 1005-1009, 2012. Abstract | Links | BibTeX | Tags: {CogTool, Log Analyzer}, SANLab @article{patton12hfes, title = {Automated CPM-GOMS Modeling from Human Data}, author = { Evan W. Patton and Wayne D. Gray and Bonnie E. John}, doi = {10.1177/1071181312561210}, year = {2012}, date = {2012-01-01}, journal = {Proceedings of the Human Factors and Ergonomics Society Annual Meeting}, volume = {56}, number = {1}, pages = {1005-1009}, publisher = {HFES}, address = {Santa Monica, CA}, abstract = {Our work with the Argus Prime (Schoelles & Gray, 2001) simulated task environment has uncovered a variety of strategies that subjects use, at least sometimes, during target acquisition. However, it is difficult to determine how well subjects implement these strategies and, if implemented, how much these strategies contribute to overall performance. Recently, we have adopted Byrne and Kirlik's (2002) cognitive-ecological approach to determine what strategies work best in different task environments. In the work reported here, we took one computational cognitive model and, holding all else constant, swapped in and out alternative strategies for target acquisition. We then ran each of these simulated human users ten times through each of four interface conditions.}, keywords = {{CogTool, Log Analyzer}, SANLab}, pubstate = {published}, tppubtype = {article} } Our work with the Argus Prime (Schoelles & Gray, 2001) simulated task environment has uncovered a variety of strategies that subjects use, at least sometimes, during target acquisition. However, it is difficult to determine how well subjects implement these strategies and, if implemented, how much these strategies contribute to overall performance. Recently, we have adopted Byrne and Kirlik's (2002) cognitive-ecological approach to determine what strategies work best in different task environments. In the work reported here, we took one computational cognitive model and, holding all else constant, swapped in and out alternative strategies for target acquisition. We then ran each of these simulated human users ten times through each of four interface conditions. |
Ralph, Jason; Gray, Wayne D; Schoelles, Michael J Cognitive workload and the motor component of visual attention Incollection Miyake, N; Peebles, D; Cooper, R P (Ed.): Proceedings of the 34th Annual Conference of the Cognitive Science Society, pp. 899-904, CSS, Austin, TX, 2012. @incollection{ralph12csc, title = {Cognitive workload and the motor component of visual attention}, author = { Jason Ralph and Wayne D. Gray and Michael J. Schoelles}, editor = {N. Miyake and D. Peebles and R. P. Cooper}, year = {2012}, date = {2012-01-01}, booktitle = {Proceedings of the 34th Annual Conference of the Cognitive Science Society}, pages = {899-904}, publisher = {CSS}, address = {Austin, TX}, abstract = {Cognitive workload effects behavior like squeezing a balloon. If you squeeze at one place, it pops out at another, and it is hard to predict where it's going to pop out. Understanding workload requires understanding the control of cognition at the 1/3 to 3s-time span during which cognitive, perceptual, and motor operations become bound together into interactive routines. Interactive routines constitute unit tasks (3 to 30 s), and unit tasks constitute subtasks (30s to 3min). To reduce cognitive workload and overload, the Functional Resource Hypothesis maintains that an optimal allocation of interactive routines to task performance would be based on the functional resource of time not modality. Some of the implications of this hypothesis are investigated in an empirical study that varied memory load as well as the demands on the eyes, visual attention, auditory cognition, and motor operations. A microanalysis of the data revealed tradeoffs between groups in their pattern of resource allocation that were compatible with the Functional Resource Hypothesis and led to surprising behavioral effects.}, keywords = {}, pubstate = {published}, tppubtype = {incollection} } Cognitive workload effects behavior like squeezing a balloon. If you squeeze at one place, it pops out at another, and it is hard to predict where it's going to pop out. Understanding workload requires understanding the control of cognition at the 1/3 to 3s-time span during which cognitive, perceptual, and motor operations become bound together into interactive routines. Interactive routines constitute unit tasks (3 to 30 s), and unit tasks constitute subtasks (30s to 3min). To reduce cognitive workload and overload, the Functional Resource Hypothesis maintains that an optimal allocation of interactive routines to task performance would be based on the functional resource of time not modality. Some of the implications of this hypothesis are investigated in an empirical study that varied memory load as well as the demands on the eyes, visual attention, auditory cognition, and motor operations. A microanalysis of the data revealed tradeoffs between groups in their pattern of resource allocation that were compatible with the Functional Resource Hypothesis and led to surprising behavioral effects. |
Sheridan, Thomas; Feary, Michael S; Gray, Wayne D; Hancock, Peter; Pew, Richard; Salvucci, Dario Tools for Predicting the Duration and Variability of Skilled Performance without Skilled Performers Inproceedings 56th Annual Conference of the Human Factors & Ergonomics Society, HFES, Santa Monica, CA, 2012. @inproceedings{sheridan12hfes.panel, title = {Tools for Predicting the Duration and Variability of Skilled Performance without Skilled Performers}, author = { Thomas Sheridan and Michael S. Feary and Wayne D. Gray and Peter Hancock and Richard Pew and Dario Salvucci}, year = {2012}, date = {2012-01-01}, booktitle = {56th Annual Conference of the Human Factors & Ergonomics Society}, publisher = {HFES}, address = {Santa Monica, CA}, abstract = {Our work with the Argus Prime (Schoelles & Gray, 2001) simulated task environment has uncovered a variety of strategies that subjects use, at least sometimes, during target acquisition. However, it is difficult to determine how well subjects implement these strategies and, if implemented, how much these strategies contribute to overall performance. Recently, we have adopted Byrne and Kirlik's (2002) cognitive-ecological approach to determine what strategies work best in different task environments. In the work reported here, we took one computational cognitive model and, holding all else constant, swapped in and out alternative strategies for target acquisition. We then ran each of these simulated human users ten times through each of four interface conditions.}, keywords = {}, pubstate = {published}, tppubtype = {inproceedings} } Our work with the Argus Prime (Schoelles & Gray, 2001) simulated task environment has uncovered a variety of strategies that subjects use, at least sometimes, during target acquisition. However, it is difficult to determine how well subjects implement these strategies and, if implemented, how much these strategies contribute to overall performance. Recently, we have adopted Byrne and Kirlik's (2002) cognitive-ecological approach to determine what strategies work best in different task environments. In the work reported here, we took one computational cognitive model and, holding all else constant, swapped in and out alternative strategies for target acquisition. We then ran each of these simulated human users ten times through each of four interface conditions. |
Wang, Ziheng; Hope, Ryan M; Wang, Zuoguan; Ji, Qiang; Gray, Wayne D Cross-subject workload classification with a Hierarchical Bayes Model Journal Article NeuroImage, 59 (1), pp. 64-69, 2012. Abstract | BibTeX | Tags: Articial Neural Network, EEG, Hierarchical Bayes Model, Workload Classication @article{wang12neuroImage, title = {Cross-subject workload classification with a Hierarchical Bayes Model}, author = { Ziheng Wang and Ryan M. Hope and Zuoguan Wang and Qiang Ji and Wayne D. Gray}, year = {2012}, date = {2012-01-01}, journal = {NeuroImage}, volume = {59}, number = {1}, pages = {64-69}, abstract = {Most of the current EEG-based workload classiers are subject-specific; that is, a new classier is built and trained for each human subject. In this paper we introduce a cross-subject workload classier based on a hierarchical Bayes Model. The cross-subject classier is trained and tested with data from a group of subjects. In our work, it was trained and tested on EEG data collected from 8 subjects as they performed the Multi-Attribute Task Battery across three levels of difficulty. The accuracy of this cross-subject classier is stable across the three levels of workload and comparable to a benchmark subject-specific neural network classier.}, keywords = {Articial Neural Network, EEG, Hierarchical Bayes Model, Workload Classication}, pubstate = {published}, tppubtype = {article} } Most of the current EEG-based workload classiers are subject-specific; that is, a new classier is built and trained for each human subject. In this paper we introduce a cross-subject workload classier based on a hierarchical Bayes Model. The cross-subject classier is trained and tested with data from a group of subjects. In our work, it was trained and tested on EEG data collected from 8 subjects as they performed the Multi-Attribute Task Battery across three levels of difficulty. The accuracy of this cross-subject classier is stable across the three levels of workload and comparable to a benchmark subject-specific neural network classier. |