September 2019
Volume 19, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2019
Perceptual factors in mental maze solving
Author Affiliations & Notes
  • Dian Yu
    Computer Science and Artificial Intelligence Lab, MIT
  • Qianqian Wan
    Department of Psychology, Hong Kong University
  • Benjamin Balas
    Department of Psychology, North Dakota State University
  • Ruth Rosenholtz
    Computer Science and Artificial Intelligence Lab, MIT
    Brain & Cognitive Sciences, MIT
Journal of Vision September 2019, Vol.19, 68b. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Dian Yu, Qianqian Wan, Benjamin Balas, Ruth Rosenholtz; Perceptual factors in mental maze solving. Journal of Vision 2019;19(10):68b. doi:

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Mentally solving visual mazes can be difficult cognitively, e.g. when each intersection has many branches. But sometimes, due to perceptual factors, even tracing the one and only path can be difficult in a compact and complex maze. Perceptual aspects of maze solving have been largely overlooked in previous research. We hypothesize that the perceptual difficulty of mental maze solving is related to visual crowding, which adds ambiguity about the shape of the path. In the current study, participants solved mazes by judging whether a continuous path connected the start and end points of each maze. We altered the level of crowding for each maze by changing its visual features and examined how this influences maze-solving time. In Experiment 1, we varied wall thickness in each maze, a manipulation previously shown to affect crowding (Chang & Rosenholtz, 2016). As walls get thicker, adjacent paths become more separated (less crowded). Indeed, observers solved the mazes faster with increasing wall thickness (F(2,44) = 28.86, p< .001). In Experiment 2, we again varied wall thickness, and also added visual complexity to the mazes by replacing straight walls with wavy lines. Observers solved mazes with four different wall types: thin-straight, thick-straight, thin-wavy, thick-wavy. We replicated results from Experiment 1 that observers were faster solving mazes with thick compared to thin walls (F(1,23) = 35.50, p< .001). Observers were also consistently faster with straight compared to wavy walls for both thickness (F(1,23) = 16.13, p=.001). Together, we show mental maze solving is affected by features that influence visual crowding. Tracing paths in more crowded mazes may require more and smaller saccades due to a smaller “uncrowded window” (Pelli & Tillman, 2008), leading to slower performance. Crowding may also interfere with figure-ground segmentation – more crowded mazes make it harder to perceive the path as figure and recognize its shape.

Acknowledgement: NSF CRCNS:Neurocomputation in the Visual Periphery 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.