Memory-related perceptual illusions directly affect physical activity in humans

Abstract

Perceptual illusions help us understand deficits in human perception, but they also have the potential to serve as treatment methods; e.g., to alleviate phantom limb pain. Treatment effects are usually the direct result of a mismatch between false visual feedback and somatosensory/proprioceptive feedback. We aimed to influence physical activity (walking distance) using a memory-related perceptual illusion that relies on a mismatch between a spatially manipulated virtual reality environment and a weakness of memory for a similar, previously experienced environment. Participants' main task was to reproduce a baseline distance three times, by walking on a treadmill while moving through a virtual reality environment. Depending on condition, the environment was either stretched or compressed relative to the previous session, but participants were not informed about these manipulations. Because false, suggestive information can lead to alterations in memory, especially when conveyed through 'rich' forms of media such as virtual reality, we expected each manipulation to alter memory for the previous environment(s) and we hypothesized that this would influence walking distance. The results for the first time showed that memory-related perceptual illusions can directly affect physical activity in humans. The effects we found are substantial; stretching previously experienced virtual environments led participants to almost double their initial walking distance, whereas compressing the environments resulted in about half of the initial distance. Possible clinical applications arising from these findings are discussed.

Fig 1. Screenshot of the VR environment; an animal crosses the trail while holding a crystal.

Fig 2. Illustration of how the part of the virtual trail that participants walked in the first (baseline) session stretched or compressed over sessions (factor 1.2).

Fig 3. Mean walking distance (m) for each walking session per condition (cf. Fig 2).

The error bars represent standard errors.

Fig 4. Mean step length (distance divided by steps) for each walking session per condition.

The error bars represent standard errors.