Abstract
etection of a secondary target (T2) in a Rapid Serial Visual Presentation (RSVP) task is usually impaired if it appears 100 – 500 ms after a primary target (T1) has been detected (the “attentional blink”, AB). Attending to T1 prevents attentional processing of T2. Previous research has found that in order for an AB to occur, both T1 and T2 need to be masked. Various theories have been put forward to explain this dual task deficit and why the masks are needed to produce it, including an interference account (Shapiro, Raymond & Arnell, 1994) and a two-stage capacity limited processing model (Chun & Potter, 1995). Interestingly, a recent study showed that task switching could also induce an AB effect even if T2 is unmasked (Kawahara, Zuvic, Enns & Di Lollo, 2003). What about T1? We hypothesized that we could induce an AB deficit without masking by manipulating the attentional demand of the T1 task. In these experiments we used a skeletal RSVP stream, which contained only T1, T2 and their respective masks. Participants had to (a) engage in a calculation to answer T1 correctly and (b) respond to the identity of T2. The T1 calculation required addition (easy task, e.g. 6 + 7) or multiplication (hard task, e.g. 6 × 7). T1 could be either masked or unmasked. T2 was always masked. When T1 was unmasked, the results showed an AB effect in the hard, but not the easy condition. However, T1 processing difficulty did not affect the AB when it was masked. We propose that in the unmasked condition more attentional resources were needed to process the hard calculation, producing a deficit in T2 report, but, with the addition of a mask, T1 processing difficulty increased to give a similar AB, regardless of calculation type. The data are discussed in terms of the similarities and differences between the different AB and dual task paradigms in order to better understand the effect.
ESRC