Purchase this article with an account.
John W. Philbeck; Rapid recalibration of locomotion during non-visual walking. Journal of Vision 2005;5(8):308. doi: https://doi.org/10.1167/5.8.308.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Blindfolded walking has increasingly been used to measure perceived target location. In this technique, observers view targets and attempt to walk to them without vision. This response involves a mismatch between visual and proprioceptive self-motion signals. As such, walking could become recalibrated over the course of successive trials, resulting in increasingly biased responding. This study tested the idea that small amounts of non-visual walking might systematically bias indications of target location in blindfolded walking tasks.
Method: During a Pre-Test phase, subjects (N = 20) walked for 3 min at 5 km/hr, either with or without vision. Then, they binocularly viewed targets in a well-lit room and either verbally estimated the target distance or attempted to walk to it without vision. Target distances ranged from 1.0 to 6.5 m. In a second session occurring at least 5 days later, subjects completed a similar set of trials, with the exception that the availability of vision during the Pre-Test phase was switched.
Results: When participants walked WITH vision during the Pre-Test, the mean signed error in subsequent blindfolded walking trials was −2, −8, −23 and −25 cm for targets at 1.5, 3.0, 4.5 and 6.0 m, respectively. When the same participants walked WITHOUT vision during the Pre-Test, the analogous errors were 13, 13, 2, and 30 cm. On average, blindfolded walking responses preceded by non-visual walking during the Pre-Test phase were 29.2 cm greater than those preceded by walking with vision during the Pre-Test phase (p < .001). Verbal responses were not affected by vision during the Pre-Test (p = .83); the mean signed error was −10 cm.
Conclusion: Even 3 min of exposure to non-visual walking is sufficient to significantly recalibrate subsequent indications of target distance using blindfolded walking. This suggests that blindfolded walking responses are themselves likely to change the calibration of walking during lengthy experiments.
This PDF is available to Subscribers Only