Validation of the Korean Sentence-Word Identification and Recall Test: Assessing Cognitive Spare Capacity in Noise

The Sentence Final Word Identification and Recall (SWIR) test is a well-established tool for estimating listening effort under ecologically valid conditions. However, the original task targeting sentence-final words may be less suitable for topic-prominent languages like Korean. We present the Korean version (K-SWIR), which was adapted to assess sentence-first word recall to reflect the linguistic structure of Korean. This study aimed to validate the K-SWIR in normal-hearing adults by examining correlations with working memory capacity and analyzing pupillary responses.

Sixty-five normal-hearing adults participated in two experiments using 14 seven-sentence lists derived from the Korean Hearing in Noise Test (K-HINT). Participants performed the K-SWIR under three listening conditions: quiet, continuous noise, and four-talker babble. Cognitive function was assessed using the reading span (RS) test and the Korean Montreal Cognitive Assessment. Pupillometry was recorded during stimulus encoding to measure physiological listening effort.

Recall performance was significantly poorer in four-talker babble compared to quiet conditions (P = 0.031). The high-RS group recalled significantly more words than the low-RS group (P < 0.001). Notably, recall performance in four-talker babble showed a strong positive correlation with RS scores (P = 0.010), indicating high sensitivity to working memory capacity. Contrary to expectations, peak pupil dilation was significantly reduced in noise compared to quiet (P = 0.007).

The K-SWIR, adapted for sentence-first word recall, is a valid tool for measuring cognitive spare capacity in Korean speakers. The strong correlation with working memory in competing speech supports its utility for assessing auditory cognitive processing.