Abstract
Previously we showed that a subject's implicit knowledge of the direction of the symmetry axis biased the positions in which symmetry was detected. That is, when stimuli were presented in single-symmetry blocks, vertical symmetry was best detected on the vertical axis of the display and horizontal symmetry was best detected along the horizontal axis of the display. However, these biases were reduced when symmetry types were mixed within a block, thus implying a role of top-down influences. Here we investigate how explicit knowledge (cueing the direction of symmetry axis) influences these positional biases. Stimuli consisted of four 8×8 arrays of black and white checks positioned 4 deg from fixation along the cardinal axes, and presented for durations of either 100 or 400ms. One array, the target, was bilaterally symmetric (horizontal (H) or vertical (V)). Distractor arrays were colored at random. Subjects (N = 4) viewed these stimuli in single-symmetry blocks (H or V), and in three randomly mixed (H & V) block conditions (no-cue, pre-cue (200ms pre-stimulus onset), post-cue (200ms post-stimulus offset)). In cued trials, a vertical or horizontal grating was presented for 50ms at fixation to indicate target symmetry, but not target location. All trials were validly cued. We calculated a Symmetry Bias Index to determine the difference between the fraction correct when the target's symmetry axis matched the display axis, and the fraction correct in the off-axis positions. When symmetry types were mixed, the bias index decreased markedly compared to the single-symmetry indices (four-fold at 100ms; two-fold at 400ms). Pre-cueing restored the indices to the single-symmetry levels. Post-cueing indices were intermediate. No implicit priming effect was observed when the target appeared in the same location over subsequent trials. We conclude that prior stimulus information, whether implicit or explicit, rapidly deploys top-down influences when processing bilateral symmetry.
NIH EY7977