Four experiments tested conflicting predictions about which components of the serial-recall task are most sensitive to auditory distraction. Changing-state (Experiments 1a and 1b) and deviant distractor sounds (Experiments 2a and 2b) were presented in one of four different intervals of the serial-recall task: (1) during the first half of encoding, (2) during the second half of encoding, (3) during the first half of retention, or (4) during the second half of retention. According to the embedded-processes model, both types of distractors should interfere with the encoding and rehearsal of targets in the focus of attention. According to the duplex-mechanism account, changing-state distractors should interfere only with rehearsal, whereas deviant distractors should interfere only with encoding. Inconsistent with the latter view, changing-state and deviant distractor sounds interfered with both the encoding and the retention of the targets. Both types of auditory distraction were most pronounced during the second half of encoding when the increasing rehearsal demands had to be coordinated with the continuous updating of the rehearsal set. These findings suggest that the two types of distraction disrupt similar working memory mechanisms.
Keywords: Auditory distraction; attentional capture; irrelevant sound effect; selective attention; serial recall.