Mental Model Management: The effects of punishment and reinforcement re-using and re-configuring strategies within game spaces

• Author / Creator

  Zhang, Yajing

• The development, storage and deployment of mental models are keystone cognitive processes central to successful operation in everyday life. We investigate the effects of punishment and reinforcement on people’s ability to acquire, reuse, and reconfigure mental models. Across three experiments, 218 participants completed a competitive, binary-outcome dice game against a computerized opponent, where the goal was to defeat the opponent who played two different exploitable strategies. All participants played four blocks of game, and each block consists of Pre, During, and Post phases. In the Pre phase, a strategy was acquired. In the During phase, a fixed win rate manipulation was used to create conditions where the strategy learned in the Pre phase was either punished or reinforced. In the Post phase, to maximize wins participants had to either relearn the old strategy (where Pre and Post strategies were the same) or learn a new strategy (where Pre and Post strategies were different). The three experiments varied in their During phase design. Participants experienced a mild punishment (50%) for Pre strategy in Experiment 1, a severe punishment (25%) and a light reinforcement (75%) in Experiment 2, and an equal degree of punishment (44%) and reinforcement (88%) relative to the baseline Pre performance of 66% win rate in Experiment 3. Participants’ proportions of win trials and optimal behaviours were analyzed in response to the reinforcement or punishment of old and new mental model strategies. Data revealed: (1) It is easier to relearn an old strategy relative to learning a new strategy; (2) The benefit of relearning old knowledge is weakened following a strong but not mild punishment; (3) Reinforcement of a learned strategy strengthens relearning old information but sabotages learning new information; and (4) Behaviours following a win are generally a stronger predictor in future performance than behaviours following a loss. These results indicate that people have a tendency to stay at their previous strategies unless they are punished harshly. Additionally, wins produce more reliable and flexible behavior relative to losses, indicating the focus of future research should be on how individuals recover from loss rather than maintain success.

• Subjects / Keywords

  • Decision-making
  • Simple games
  • Mental model acquisition
  • Mental model update
  • Mental model reacquisition
  • Reinforcement
  • Punishment

• Graduation date

  Spring 2024

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-e81g-4c87

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.