Abstract
Weight illusions demonstrate that an object's physical properties influence our subjective experience of its heaviness. The most robust illusion is the size-weight illusion (SWI), in which the smaller of two objects of equal mass feels heavier. Although several theories exist to explain the illusion, none account for all relevant findings. One set of theories attributes the SWI to bottom-up processes, proposing that certain object-related variables are directly perceived by the body during lifting, variables that must be interpreted consciously as heaviness. These theories include a critical role of haptic/somatosensory feedback of size in the illusion; however, there is mixed evidence regarding the contribution of haptic compared to visual input. We conducted a meta-analytic review of the SWI literature to quantify the contribution of visual and haptic information to illusion strength. Following literature searches and email communication with authors, we obtained data comprising perceptual heaviness estimates for stimuli with the same mass but different volumes. Data were included from 27 studies, including 4 unpublished datasets. Mean perceptual estimates were compared for stimulus pairs (e.g., smaller vs larger) within each study/experiment, and used to calculate effect size observations (d), weighted by sample size. Results demonstrated the SWI was comparable across the following conditions: with visual feedback, without haptic feedback (n=21; mean weighted d = 2.09, 95% CI 1.90-2.28), with haptic feedback, without vision (n=30, mean weighted d = 2.19, 95% CI 2.07-2.32) with both visual and haptic feedback (n=106, mean weighted d = 1.95, 95% CI 1.89-2.01). Findings suggest that both visual and haptic information elicit a SWI of comparable strength, with no additive effect of the modalities. The results do not support theories proposing a critical role of haptic/somatosensory information in the SWI, thus pointing to an alternative mechanism. Publication bias remains an issue and will be investigated in future analyses.
Meeting abstract presented at VSS 2018