Learning-dependent evolution of spatial representations in large-scale virtual environments.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human Spatial Navigation

Background:

• Understanding how humans use environmental boundaries to anchor spatial representations during navigation is crucial.
• Existing behavioral and neurophysiological models offer conflicting predictions.
• Technical challenges in testing navigation in realistic, large-scale environments have hindered research.

Purpose of the Study:

• To investigate how humans utilize environmental boundaries for spatial anchoring during navigation.
• To determine the experience-dependent dynamics of employing boundaries in large-scale virtual environments.
• To bridge the gap between rodent neurophysiology and human spatial behavior models.

Main Methods:

• Participants freely ambulated in virtual reality town-sized environments using an omnidirectional treadmill.
• Judgments of Relative Direction (JRD) were used to assess spatial knowledge and boundary use.
• Structural equation modeling analyzed learning-dependent pointing accuracy aligned and misaligned with environmental axes.

Main Results:

• Initial pointing accuracy showed no directional bias towards environmental boundaries.
• Alignment effects, favoring boundaries, emerged after four learning blocks.
• Map pre-exposure without navigation also induced alignment effects, suggesting rapid schema formation.

Conclusions:

• The use of spatial boundaries as an organizing schema for large-scale navigation is experience-dependent.
• Spatial knowledge accumulation, rather than immediate environmental cues, drives boundary utilization.
• Findings reconcile rodent and human models by highlighting learned route knowledge in spatial anchoring.