Attentional settings based on previous experience affect bias in visual comparisons

Previous experiments in visual comparison have shown that the spatial congruency bias (SCB), a bias to categorise two targets as "same" if they occupy the same location in successive displays, and the overall bias (OB), the average bias across all trials, vary with visual similarity; and that the OB also varies with the type of task being performed. In four experiments, we explore whether these results are best explained by a visual similarity account, an attentional zoom account, or a combination of the two. Using a shape comparison task, we manipulated the visual similarity and predictability of the target displays by varying the local position of the target letter, either between different blocks (i.e., predictable; Experiments 1a and 2a) or within a block (i.e., unpredictable; Experiments 1b and 2b); additionally, we varied the distractor letters such that they were the same between the to-be-compared displays in most of the trials (Experiments 1a and 1b) or they were different in every trial (Experiments 2a and 2b). Under conditions of low interference, the predictability of visual information has no effect on the OB, or on the SCB, but this may be because attentional demands are low in these conditions. As predicted by the attentional zoom account, the SCB is influenced by predictability when peripheral interference is high. These results suggest that both similarity and attention play a role in visual comparisons.