August 2023
Volume 23, Issue 9
Open Access
Vision Sciences Society Annual Meeting Abstract  |   August 2023
Inferential tracking reveals context is more informative than faces in judgments of trustworthiness.
Author Affiliations
  • Yifan Fang
    University of California, Berkeley
  • Jefferson Ortega
    University of California, Berkeley
  • Necdet Gürkan
    Stevens Institute of Technology
  • Jordan W. Suchow
    Stevens Institute of Technology
  • David Whitney
    University of California, Berkeley
Journal of Vision August 2023, Vol.23, 5119. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Yifan Fang, Jefferson Ortega, Necdet Gürkan, Jordan W. Suchow, David Whitney; Inferential tracking reveals context is more informative than faces in judgments of trustworthiness.. Journal of Vision 2023;23(9):5119.

      Download citation file:

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

  • Supplements

Perception of trustworthiness significantly impacts economic (Rezlescu et al., 2012) and juridical decisions (Blair et al., 2004). Consequently, the misperception of trustworthiness can lead to devastating outcomes. Contextual threats can bias observer perception of the trustworthiness of face stimuli (Mattavelli et al., 2021), but studies on trustworthiness so far have used only static and often unnatural images, focusing only on faces. Here, we investigated whether observers can continuously and dynamically infer a person’s trustworthiness, based entirely on context, even when the target face is masked and invisible. Observers continuously rated the trustworthiness of target characters in natural movies using a Likert-scale. Participants were split into one of three conditions: the context-only condition, which masked-out the target character; the character-only condition, which masked-out the background contextual information; and, the baseline condition, which was the fully-informed movie with no mask. Dynamic trustworthiness ratings based entirely on context-only information (no visible target character) correlated significantly with trustworthiness ratings in the baseline (no-mask) condition. Ratings of the character-only condition (no visible context) also correlated with the baseline condition, but to a lesser degree. Examining individual observer performance revealed that observers in the context-only condition had significantly higher performance than individuals in the character-only condition. Finally, we also found higher between-subject agreement in the context-only condition compared to the character-only condition. Our results suggest that 1) observers can rely entirely on background context information to make accurate trustworthiness judgments; 2) observers more accurately infer the trustworthiness of another individual using contextual information than isolated face information; and 3) observers agree with each other more when using contextual information compared to face/body information when perceiving trustworthiness. The results demonstrate the importance of context in trustworthiness perception and raise the possibility that individual observer biases interfere with judgments of face trustworthiness more than context-based judgments of trustworthiness.


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.