The effect of masker type and masker timing on the serial recall of speech

Abstract

Understanding speech in noise involves a complex set of processes including segregation of sources, selection of the target talker, and recognition and comprehension of the ongoing message. Previous studies often have focused on the segregation, selection and recognition processes by measuring the intelligibility of short strings of words presented concurrently with different types of maskers. However, these studies may not fully capture the effects of masking on all aspects of ongoing speech processing. For instance, masking can cause poorer recall for sequences of words even if a noise/competitor leads to near-perfect recognition of the individual items [e.g., Cousins et al., 2014, Mem Cogn (42)]. Furthermore, in contrast to the typical design of speech recognition tasks, background noise usually does not cease during response to the message. In the present study, we examined listeners’ serial recall of target speech with different maskers presented at different points during the trial. Speech-shaped no...