Abstract
Most of studies of object-based attention use the classic two-rectangle paradigm (Egly et al., 1994) featuring a hybrid cue in which both exogenous and endogenous components draw attention to a location within an object. Response times in a simple detection task indicate the spread of covert attention along this object. Here, we tested whether (1) purely exogenous cueing evokes this object-based advantage; (2) this effect is also reflected in visual sensitivity. We measured sensitivity and response time in a visual discrimination task: Observers reported the tilt of an oriented noise patch (probe) briefly presented at the end of one of two rectangles, each filled with dynamic noise. At one of three time points before the probe, we flashed a cue at the end of one rectangle. This exogenous cue was 100% task-irrelevant, as the probe appeared either within the cued or the uncued object, but never at the cued location. By varying the contrast of the dynamic noise, we measured performance in five different signal-to-noise ratio (SNR) conditions. We expected a within-object benefit in response times for high SNRs, and in visual sensitivity for lower SNRs. Each participant completed four sessions of the experiment with response time prioritized over accuracy, followed by four sessions with accuracy prioritized over response time. Increasing SNRs substantially reduced response time and increased visual sensitivity. Moreover, for lower SNRs, responses were much slower for shorter cue-probe delays while sensitivity was unaffected. This effect suggests that our manipulation of SNR indeed affected the time required to extract visual information. Importantly, we did not find an object-based benefit in response time or sensitivity: Performance for probes presented in the exogenously cued vs. uncued objects was strikingly similar. Given the high power of our study, this finding suggests that object-based selection requires that cueing has an endogenous component.