Expecting Alternation With Gratings

Bayesian theories of perception propose that percepts are formed from a precision weighted combination of sensory input and expectations (Clark, 2013; de Lange, Heilbron, & Kok, 2018; Bar, 2004). However, violations of learnt statistical regularities need to be perceived well in order for our expectations to be updated accordingly (Behrens et al., 2007; Press et al., 2020). Our previous studies have found that the hitrate for events presented on stimuli at novel locations, or at novel orientations, is reduced compared to that of expected locations or orientations, respectively, in line with surprise-induced blindness (Asplund et al., 2010; Ward & Press 2024). In this study we ask whether the surprise-induced impairment in hitrate is maintained when change is expected. Specifically, on all expected trials, the location and orientation of the stimulus will alternate. Then, the novel trial is one where the stimulus is presented twice in the same location, or twice with the same orientation. This allows us to test whether surprise still leads to a decrease in hitrate when the set of possible stimuli does not increase.