Real-time continuous perceived uncertainty annotation for spatial navigation studies in buildings

Abstract

In order to study the complex psychological processes involved in human wayfinding in increasingly complex built environments, it is crucial to have continuous self-report measures. In this study, we developed two different methods for continuously measuring perceived uncertainty during wayfinding, one is real-time continuous annotation (RCUA) using a joystick device and the other is post-hoc continuous annotation (CUA) involving annotations while watching video footage of recent wayfinding activity. To evaluate the usability, reliability, and validity of both approaches, we conducted a study with 54 participants. We assessed the measures' reactivity during different sign-seeing events. We also evaluated the convergent validity of both measures by comparing their outcomes with a self-report questionnaire, and assessed their discriminative and predictive validity by comparing uncertain values between known groups and correlating those values with wayfinding performance. Our findings suggest that both approaches were valid at the task level, but RCUA was better at capturing fine-grained dynamics of human experience. These continuous uncertainty measures can provide valuable insights into the fleeting nature of human experience and help identify "problem spots" for wayfinding in complex buildings.