(PP1325) The Relationship Between Speech Recognition in Noise Ability and Listening Effort

Objectives: The goal of this study was to determine the relationship between speech recognition in noise (SRN) performances vs. self-reported listening effort. SRN ability was evaluated using the AzBio test with four listening conditions. Listening effort was based on self-report using a seven-point Likert scale. It was hypothesized that a statistically significant relationship would be found between speech recognition in noise abilities vs. self-reported listening effort and between listening effort ratings across the noise conditions. 

Rationale: Johnson et al. (2015) used a subjective listening effort scale to evaluate participants’ perceived listening effort in varying speech-in-noise conditions. SRN ability was evaluated using the R-Spin test. For each sentence list, the participants were asked to rate their listening effort on a subjective scale. According to Johnson et al. (2015) SRN scores presented at poorer SNRs would be more difficult to understand, which would coincide with greater listening effort. In the present study, the relationships between listening effort (using the scale from Johnson et al., 2015) and SRN ability (using the AzBio test) were evaluated across four masker conditions. It was expected that poorer SRN scores would reflect greater listening effort.

Design: Thirty-one native English speakers with normal pure-tone thresholds (≤ 25 dB HL, 0.25 – 4.0 kHz) with a mean age 20.9 participated in this study. The AzBio test was used to evaluate binaural speech perception in noise. The AzBio test was administered binaurally under supra-aural headphones in a sound-treated booth. The sentences and the maskers were presented at 65 dBA (0 dB SNR). The four masker conditions included four-talker babble forward, four-talker babble backwards (where the babble was presented backwards), conversational masker forward, and conversational masker backwards. All test conditions were randomized. After each noise condition, the participants were asked to rate their listening effort using the scale from Johnson et al. (2015), which ranged from 1-7(1=“No Effort” and 7=“Extreme Effort).” The Spearman rho correlation was used to evaluate the relationship between listening effort and AzBio scores and the relationship between listening effort ratings across masker conditions. 

Results: No statistically significant relationships were found between AzBio scores and ratings of listening effort. Out of the six comparisons, statistically significant relationships were found for listening effort between masker conditions: four-talker babble backwards vs. four-talker babble forward (r =.45, p = 0.012); conversational masker forward vs. four-talker babble forward (r = 0.44, p = 0.016); conversational masker forward vs. four-talker babble backwards (r=0.64, p=< 0.001); conversational masker backwards vs. four-talker babble backwards (r=0.43, p=0.016); and conversational masker backwards vs. conversational masker forward (r = 0.68 p=< 0.001).

Conclusions: Contrary to our hypothesis, no statistically significant relationships were found between the AzBio scores and effort ratings for each masker condition. Statistically significant relationships were found between listening effort ratings for five of the listening condition comparisons. This indicates that while listening effort was not related to SRN performance, listening effort was relatively consistent across masker conditions.