The purpose of this study was to investigate the test-retest reliability of speech recognition thresholds (SRTs) measured by Korean standard bisyllabic word list for adults (KS-BWL-A). Thirteen normal-hearing adults (14 ears) and 67 hearing-impaired persons (93 ears) with sensorineural hearing loss (SNHL) participated in this study. Subjects with SNHL were categorized into four groups: 1) mild hearing loss, 2) moderate hearing loss, 3) moderately severe hearing loss, and 4) severe-to-profound hearing loss. Pure tone audiometry and the SRT test with KS-BWL-A were administered in all participants. The SRTs were mathematically calculated based on the national and international standards (ISO 8253-3, 2012; KS I ISO 8253-3, 2009). The retest of the SRTs was performed within two weeks with the same procedure as the initial test. Results showed that the first SRTs were highly correlated with the second SRTs for normal-hearing and hearing-impaired groups. There was no significant difference between the first and second SRTs for normal-hearing group. For hearing impaired groups, the second SRTs were slightly decreased than the initial SRTs in all four groups. The 95% confidence intervals for each group were between ± .82 ~ ±1.44 dB and the 95% prediction intervals of individual scores for each group were between ±3.10 ~ ±4.34 dB. Results recommend that SRT test using the KS-BWL-A be a reliable method in every category of SNHL as well as normal-hearing based on high correlation, narrow CI and PI. It is also suggested that a difference greater than 5 dB between the first and second SRTs be interpreted as significant in clinical situations.

The purpose of this study was to investigate the test-retest reliability of speech recognition thresholds (SRTs) measured by Korean standard bisyllabic word list for adults (KS-BWL-A). Thirteen normal-hearing adults (14 ears) and 67 hearing-impaired persons (93 ears) with sensorineural hearing loss (SNHL) participated in this study. Subjects with SNHL were categorized into four groups: 1) mild hearing loss, 2) moderate hearing loss, 3) moderately severe hearing loss, and 4) severe-to-profound hearing loss. Pure tone audiometry and the SRT test with KS-BWL-A were administered in all participants. The SRTs were mathematically calculated based on the national and international standards (ISO 8253-3, 2012, KS I ISO 8253-3, 2009). The retest of the SRTs was performed within two weeks with the same procedure as the initial test. Results showed that the first SRTs were highly correlated with the second SRTs for normal-hearing and hearing-impaired groups. There was no significant difference between the first and second SRTs for normal-hearing group. For hearing impaired groups, the second SRTs were slightly decreased than the initial SRTs in all four groups. The 95% confidence intervals for each group were between ± .82 ~ ±1.44 dB and the 95% prediction intervals of individual scores for each group were between ±3.10 ~ ±4.34 dB. Results recommend that SRT test using the KS-BWL-A be a reliable method in every category of SNHL as well as normal-hearing based on high correlation, narrow CI and PI. It is also suggested that a difference greater than 5 dB between the first and second SRTs be interpreted as significant in clinical situations.

Speech recognition threshold (SRT), Korean standard bisyllabic word list for adults (KS-BWL-A), Test-retest reliability, Confidence interval (CI), Prediction interval (PI)