Abstract
Many common tasks such as reading and visual search require the rapid updating of working memory with successive selections of visual inputs. This situation has been mimicked in the psychological laboratory using partial report procedures that require observers to identify two targets (T1 and T2) embedded in a sequence of distractors. The striking outcome from these studies is that memory for T2 is often poor when T2 is presented within less than 500 ms of T1, an effect known as the attentional blink (AB). Previous work shows that the failure to recall T2 can be prevented by precuing the target, suggesting that the allocation of attention to T2 is delayed during the AB (Nieuwenstein, Chun, Van der Lubbe & Hooge, 2005). Here, we examined whether the cause of this delay lies in selecting or encoding of T1. We presented sequences of six letters and asked observers to report either as many letters as they could (whole report), or only two letters presented in a particular color (partial report). The results showed a substantial AB for T2 report in partial report, while report of the corresponding letter was relatively accurate in whole report. A similar effect was observed in a second experiment in which selection for partial report was based on letter-identity. These findings show that the delay in selecting targets presented during the AB is caused by selection, not encoding, of T1.
NieuwensteinM.R.ChunM.M.Van der LubbeR.H.J.HoogeI.T.C. (2005). JEP:HPP.
Grant MH47432 from the National Institute of Mental Health