Auditory streaming and short-term memory: Effects of talker variability on serial recall and auditory distraction
* Presenting author
Abstract:
Immediate serial recall of spoken items is better with single-talker sequences when compared to multiple-talker sequences. This “talker variability effect” (TVE) has been attributed to a reduction of the stream coherence in the multiple-talker sequence. Stream coherence also impacts the “irrelevant sound effect” (ISE), which is defined as the impairment of serial recall performance for visually presented verbal items by task-irrelevant background sounds such as speech or music. It has been argued that ISE is most pronounced when the irrelevant sound is composed of changing auditory tokens emitted by a single source, as pre-attentive, obligatory processing of such auditory streams evoke serial order representations which interfere with deliberate retention of the serial order of the list items.In this experiment, immediate serial recall of visually-presented digits is impaired by task-irrelevant background speech. We analysed whether the ISE is attenuated when the coherence of the irrelevant speech stream is reduced by multiple talkers. The task-irrelevant speech consisted of letter names, produced by single talker vs. multiple talkers. In the second part, we explored the TVE on serial recall of spoken letter names. Data, however, show no sign of a relationship between the TVE (auditory task) and the changing-state effect (visual task).