Objectives: Task-irrelevant sounds can be distracting, impairing speech recognition and increasing listening effort. One way to measure distraction is to evaluate performance differences between trials that do and do not contain task-irrelevant sounds. Poorer performance following irrelevant sounds is interpreted as evidence of distraction. Using this general approach, previous work has demonstrated that the degree to which irrelevant non-speech sounds are distracting is related to stimulus valence. That is, unpleasant and pleasant sounds are more distracting than neutral ones. The purpose of the present study was to evaluate the effect of the presence of task-irrelevant, human vocalizations on word repetition performance and behavioral listening effort. Design: Eighteen adults with normal hearing participated in a dual-task paradigm where the primary task was word repetition, and the secondary task was word categorization speed, with speed being interpreted as a measure of behavioral listening effort. Immediately before 45% of the target words, a task-irrelevant vocalization was presented. Each list contained trials with and without irrelevant vocalizations; the difference in performance between trial types was interpreted as distraction. Conditions varied by background noise (quiet, noise) and by valence of the vocalization (pleasant, unpleasant, neutral). Conditions were blocked such that testing with each list of words was completed with a single combination of noise and valence vocalization. Data were analyzed using linear mixed-effects modeling to evaluate the effect of task-irrelevant vocalization presence, vocalization valence, and background noise on word repetition performance and behavioral listening effort. Results: As expected, background noise negatively affected word repetition performance (~6 percentage points) and increased behavioral listening effort (119 msec slowing of secondary task response times). The valence of the task-irrelevant sound within a condition did not affect word repetition performance, but did affect behavioral listening effort. Response times were longer (~40 msec longer) during the conditions with pleasant or unpleasant vocalizations than during conditions when neutral vocalizations were interleaved with target words. No distraction was noted in word recognition performance; performance was similar for trials with and without task-irrelevant vocalizations. Distraction effects were evident in behavioral listening effort; response times were ~30 msec longer in trials preceded by task-irrelevant vocalizations than in trials without task-irrelevant vocalizations. The calculated distraction effects were similar across conditions. Conclusions: Although the valence of the vocalizations generally increased listening effort, slowing secondary task times in conditions with valenced task-irrelevant vocalizations, the calculated distraction effects within a condition did not depend on valence or the presence of background noise. Moreover, the task-irrelevant vocalizations increased behavioral listening effort in quiet and in noise, despite minimal changes in word repetition performance. Future work is warranted to disentangle the methodological differences and the differences in effect sizes between the present study and previous work with non-speech sounds. The findings have implications for listening in complex, real-world environments, where distraction and valenced sounds are likely to co-occur.
Picou et al. (Fri,) studied this question.