Abstract
According to two-stage models of visual search, search slopes during conjunction search reflect increased distractor numbers – and that conjunction search involves serial, self-terminating search where distractor items are checked off one by one. There is however an inherent problem with measuring set-size effects with the single target search. When distractors are added, both absolute set-size (the number of stimuli on the screen) and the relative set-size (the target vs distractor ratio) increase. Do search slopes reflect increasing absolute set-set size or decreasing relative set-size? We addressed this question using a visual foraging task, where observers tap multiple target items among distractors. But this paradigm also entails a challenge, since as targets are selected, they disappear, changing both absolute and relative set-size. We therefore tested three foraging conditions, a 'classic' condition, where targets disappeared when tapped, a 'replace' condition where targets turned into distractors when tapped, keeping absolute set-size constant throughout the trial and finally a 'disappear' condition where a random distractor disappeared along with the target tapped, keeping relative set-size constant. We tested these three conditions during both feature and conjunction foraging. When absolute set-size was kept constant, but relative set-size changed throughout the trial (replace condition), switch-costs were higher, the RTs for the final foraging target slower and generally all indicators of task difficulty showed that manipulating relative set-size had a larger effect on foraging performance than manipulating absolute set-size. Furthermore, our previous results have shown that the final-target RT's during foraging tasks replicate the search slopes seen in single target search tasks. Here, we see this same traditional pattern in the replace condition, but not in the disappear condition, suggesting that absolute set-size is indeed not responsible for the traditional set-size effects seen in single target conjunction search.
Meeting abstract presented at VSS 2018