Perceptual uncertainty and action consequences independently affect hand movements in a virtual environment

Area of Science:

• Cognitive Science
• Human-Computer Interaction
• Robotics

Background:

• Virtual and augmented reality (VR/AR) are increasingly used for behavioral research and motor skill training.
• Effective transfer of learned skills and insights from VR/AR to real-world settings is a key assumption.
• Motor behavior is significantly shaped by perceptual uncertainty and expected action outcomes.

Purpose of the Study:

• To investigate whether motor behavior in virtual reality (VR) environments transfers to real-world settings.
• To determine if differences in perceptual reliability or expected action consequences contribute to transfer failures.
• To examine hand movements in an obstacle avoidance task under varying environmental conditions.

Main Methods:

• Utilized mirror reflections to create a virtual environment with minimized perceptual artifacts.
• Compared hand movements in an obstacle avoidance task between real and virtual obstacle interactions.
• Manipulated the quality of perceptual information to isolate its effect on behavioral differences.

Main Results:

• Systematic differences in hand movements were observed between real and virtual obstacles.
• These behavioral discrepancies persisted regardless of the perceptual information quality in the virtual environment.
• The findings indicate that action correspondence does not automatically follow perceptual correspondence.

Conclusions:

• Disparities in expected action consequences between virtual and real environments are a primary driver of behavioral differences.
• Even when virtual environments achieve perceptual fidelity, motor skill transfer may be limited.
• Future VR/AR applications should consider the role of action consequence expectations in skill transfer.