Effect of geometrical structure on perception of the vertical in the Rod-and-Frame Test.

Male and female subjects were selected on the basis of their unsigned deviation scores (USD) in the rod-and-frame test to be retested and scored by a new method. The new method analyzes performance on this task in terms of the effect of the tilt of the frame (phi), the constant deviation (micron), the starting position of the rod (rho), and the response consistency (sigma) of each subject. The geometrical structure of the frame was varied by changing the number of identical stimulus elements in it. Variations in the geometrical structure of the frame affected phi-scores primarily in subjects with significantly high phi-scores (frame-dependent subjects). Variations in the frame's geometrical structure did not affect micron or rho. The relationship between USD-scores and the values for phi, micron, and rho for each subject is discussed.