Objective: The aim of this study was to investigate the use of verbal response time (RT) as a measure of listening effort within clinical settings. This was carried out in the context of speech-in-noise evaluations using staircase adaptive procedures.

Methods: A total of 58 participants were involved in the study, who were divided into three groups based on age and hearing thresholds: young participants with normal hearing (YL_NH, n=36, mean pure tone average PTA=3.5 dB nHL), older listeners with normal hearing (OL_NH, n=14, PTA=14.1 dB), and listeners wearing hearing aids (HA, n = 8, PTA=51.3 dB). The speech-in-noise test was presented through two loudspeakers placed in front of the listener and consisted of sequences of four disyllabic words. It was carried out under four listening conditions, created by combining two reverberation conditions (anechoic, reverberant with Tmid = 0.56 s) and two background noise types (stationary, fluctuating). An adaptive staircase procedure was used to determine the speech reception threshold for 80% correct word identification. During the staircase procedure, the verbal RT for each sentence was recorded, defined as the time elapsing between the audio offset and the onset of the participant’s verbal answer. The RT in a specific listening condition was calculated as the median value of the last 10 trials. Following each listening condition, participants provided a rating of their perceived listening effort using a visual analog scale.

Results: Statistical analyses of RTs revealed interactions between population, reverberation condition and background noise. In conditions with stationary noise, RTs were longer in reverberant compared to anechoic conditions. In all listening conditions, YL_NH had shorter RTs compared to HA. Additionally, RTs were significantly longer for OL_NH compared to YL_NH, but only in anechoic conditions and in the presence of fluctuating noise. Self-ratings of effort disclosed a significant main effect of the reverberation condition (lower perceived effort in anechoic conditions) and population (higher effort for HA in comparison to NL_NH). A significant positive correlation between RTs and self-rated effort was only found within the HA group.

Conclusions: Our findings provide an indication of the potential utility of incorporating verbal RT as a measure for assessing listening effort in clinical settings, particularly for speech-in-noise tests administered using a staircase adaptive procedure. At a high-performance level, RT is sensitive to changes in the auditory environment (background noise, reverberation) and disclose effects beyond intelligibility for the HA group. The absence of a correlation between RTs and self-rated effort, along with their different sensitivity, reinforce the argument that the two measures tap into separate cognitive dimensions. These results have practical implications for the clinical practice, specifically in defining new protocols for HA best-fitting procedures.