Brian Glass

Biography (short): 


From Brian's personal web page

I am generally interested in human decision making. Specifically, I am interested in certain aspects of decision making at the levels of the individual as perceiver, social or motivational influences on the individual, and the individual interacting with agents in dynamic environments.

  • Cognitive consequences of video games: In recent years, video games have developed into rich, dynamic, and fast-paced virtual environments. Prior work reveals that those who play video games demonstrate enhanced cognitive and perceptual performance in a range of tasks. This is true for both chronic video game players (VGP) and non-video game players (NVGP) who are trained on video games (Dye, Green, & Bavelier, 2009). 

    Most previous research uses a subtype of video games known as action video games. Action video games feature a first-person perspective, and the participant navigates a three-dimensional world (e.g., UnrealDoomsee example here). The fast paced game play in action video games typically consists of engaging other agents in the virtual world (i.e., killing them with projectile weapons). Here, we are interested in an alternative subtype of video games known as real-time strategy games (RTS; e.g., StarCraftAge of Empiressee example here). Game play in RTS games is from a third-person perspective with the participant constructing and controlling an army represented by several agents on a large map. The participant commands this army in battle against an enemy army. Despite many concurrent game situations unfolding in real-time, only a portion of the entire game world (i.e., game map) and a portion of the agents can be seen on the monitor at a given time. Participant must track and switch between many evolving states in order to be successful. We employ a variety of novel methods to capture which game features are associated with various cognitive enhancements. Thus, RTS games represent a unique research opportunity due to the possibility that attentional mechanisms could be altered and perhaps enhanced by RTS training. Additionally, the long term effects of RTS gaming is not known.

  • Individual as perceptual classifier: the role of feedback, intradimensional shifts and extradimensional shifts in category learning. Multiple systems have been implicated in human category learning, and model based analyses help to determine strategies that people may adopt in different types of tasks.
  • Motivational influences on the individual: the role of regulatory focus on set shifting tasks that require cognitive flexibility and signal detection tasks that require shifts in response bias and perceptual sensitivity.
  • Individual interacting with agents
    1. By developing models that can drive simulated opponents, single participants perform resource acquisition tasks in controlled environments. These models are used in combination with a two-dimensional foraging task (see Goldstone and Ashpole 2005) to examine individual level behavior such as the tradeoff between exploration and exploitation.
    2. Humans can teach as well as learn from agents, and I am interested in psychological properties as well as properties of computer agents that drive learning in human-agent interactive scenarios.


research associate (postdoc)
decision making