Authors: Shivani Sharma¹; Kavassery Venkateswaran Nisha¹
¹AIISH
Background: Speech perception in noise requires significant cognitive load, yet listening effort remains underexplored in multilingual contexts where linguistic factors interact with acoustic challenges. This study examined listening effort in native and non-native Kannada speakers across varying signal-to-noise ratios (SNR) and reverberation conditions.
Method: A mixed factorial design was used with 90 adults divided into three groups: Hindi L1, Malayalam L1, and native Kannada speakers. Listening effort was assessed using a dual-task paradigm involving Kannada sentences mixed with multi-talker babble. Stimuli were presented at two SNRs (0 and +4 dB) across three reverberation times (0.74s, 1.66s, and 4.37s). Participants recalled the penultimate word of sentences within blocks, and data were analyzed using linear mixed-effects models (LMER).
Results: Significant main effects were found for reverberation time, SNR, and language group (p < .001). Increased reverberation reduced performance, while higher SNR improved scores. Native speakers demonstrated the lowest listening effort; among non-native groups, Hindi speakers exhibited higher effort than Malayalam speakers. A significant three-way interaction revealed that SNR benefits diminished under high reverberation, particularly for non-native listeners. Malayalam speakers’ superior performance relative to Hindi speakers likely stems from greater linguistic similarity to Kannada.
Conclusion: Listening effort is heavily influenced by the interplay between acoustic degradation and linguistic background. Native proficiency serves to reduce cognitive load, whereas non-native listeners face disproportionate difficulty in degraded environments. these findings have significant implications for audiological assessment and communication in multilingual societies.
