Reinterpretation of the rod-and-frame illusion: a virtual reality study

In the Rod-and-Frame Test (RFT), participants align a pivoted rod with the vertical while viewing a tilted, coaxially mounted frame. In doing so, they can use the edge of the frame and its imaginary diagonal as visual cues. Relying on one of these cues leads to the RFT illusion-an error in determining the vertical. We investigated whether individuals who can use both cues perform more accurately at tilt angles, where errors typically peak.

Twenty-one young adults completed a Virtual Reality RFT. A bias function was defined to range from 1 (rod rotated consistently toward the edge cue) to -1 (toward the diagonal cue). We calculated the bias for the tilt angles ±35° (where the diagonal cue is visually salient) and alignment errors at ±15° (where errors are high).

The bias and error were strongly correlated (r = 0.75). Participants with bias values below -0.5 (indicating reliance on the diagonal cue) at ±15° exhibited errors nearly four times smaller than those with bias values above 0.5 (indicating reliance on the edge cue). For ±35°, the error for such groups was not statistically different.

Reliance on the diagonal cue at large tilt angles (e.g., ±35°) is associated with improved performance at smaller tilt angles (e.g., ±15°). These findings suggest that RFT errors-arising from multisensory integration of visual, vestibular, and proprioceptive inputs-also reflect individual differences in the processing of visual context.