Abstract
Stimuli presented in rapid serial visual presentation (RSVP) at lags shorter then half a second produce a phenomenon known as the Attentional Blink (AB) (i.e., a second target, T2, is poorly reported when close in time to a first target, T1). The AB occurs with almost every type of stimulus category. One exception is faces (Awh et al., 2004, Landau and Bentin, submitted). In the present study we investigated basic visual information that may lead to the face-AB-resistance phenomenon (fABr). We devised a mixed model design, manipulating spatial frequency of the stimuli in the RSVP as a between-participant factor with three levels: Low-pass, High-pass and Broadband. Each group participated in the same design; only the stimulus frequencies were varied. T2-type (face/watch) and Lag (1, 3, 7) were manipulated as within-participant factors. Participants performed a size discrimination task on T1 and a detection task on T2. There was a significant second-order interaction between spatial frequency range, T2-type and Lag. Specifically, while removing high spatial frequencies did not change the fABr effect, removing low spatial frequencies produced the AB phenomenon for faces. These findings suggest that the differential AB effects found for objects and faces may be attributed at least in part to differences in the spectral analysis by the visual system. The results will be discussed in the context of face specificity at both perceptual and attentional levels (Landau and Bentin, submitted).