Relation of eye movements to errors on the Rod-and-Frame Test.

By a selective-attention model of rod-and-frame test performance frame-independent (small error) subjects are better able than frame-dependent (large error) subjects to select out relevant cues (the rod) while suppressing irrelevant ones (the frame). However, certain clinical studies have implied that smaller errors can be obtained by employing eye-movement strategies that ignore rather than actively suppress irrelevant cues. In this study the number, magnitude, and rate of eye movements were measured in 15 subjects actively engaged in the test. While no discernible relationship of these variables to test errors was determined for frame-dependent subjects, significant negative correlations of rate and magnitude of movement with errors were demonstrated for frame-independent subjects. Group differences were observed in number and magnitude of eye movements. These findings are discussed as support for the hypothesis of selective attention.