Abstract
As search difficulty increases, search performance decreases (e.g. Huang & Pashler, 2005), and the amount of target related detail held in visual working memory (VWM), as measured by contralateral delay activity (CDA), increases (Schmidt & Zelinsky, 2017). However, it is unclear how the expectation of search difficulty interacts with the actual level of search difficulty, and how the amount of target related visual detail maintained in VWM modulates the expectancy effects. This study aimed to isolate the effects of expected and actual search difficulty using eye movements and to measure the target representation using CDA. Participants pictorially previewed (200 ms) and then searched for a target Landolt-C which appeared among oriented distractor C’s. CDA was assessed during the delay following preview offset but before search onset (1000 ms). Target-distractor similarity was manipulated to generate easy, moderate, and difficult search blocks. Importantly, blocks contained 25% of trials from the unexpected difficulty level to assess how search performance changed when expected difficulty mismatched actual search difficulty. Consistent with previous findings, our results indicate decreased accuracy (p < .01) and increased reaction time (p < .05) as search difficulty increased. This difference was also modulated by expected search difficulty, such that if an easier search was expected, accuracy and RT decreased. However, if a more difficult search was expected, accuracy and RT increased. Significant CDA was also observed in all difficulty conditions (p < .01), however, differences between conditions have not yet risen to the level of significance. Our current eye movement data suggest that expected search difficulty modulates search performance, however, the relationship to the target template is currently unclear.