Original Research

Development and validation of a sentence-in-noise test for signal-to-noise ratio loss in adult normal-hearing and hearing-impaired listeners in South Africa

Alida Naudé, René Hornby, Rozelle Roets
South African Journal of Communication Disorders | Vol 73, No 1 | a1193 | DOI: https://doi.org/10.4102/sajcd.v73i1.1193 | © 2026 Alida Naudé, René Hornby, Rozelle Roets | This work is licensed under CC Attribution 4.0
Submitted: 11 May 2026 | Published: 29 September 2026

About the author(s)

Alida Naudé, Division of Speech-Language and Hearing Therapy, Department of Health and Rehabilitation Sciences, Stellenbosch University, Tygerberg Campus, Cape Town, South Africa
René Hornby, Rene Hornby Oudioloë Inc., Pretoria, South Africa
Rozelle Roets, Rene Hornby Oudioloë Inc., Pretoria, South Africa

Abstract

Background: Difficulties understanding speech in noise are common in adults with hearing impairment. Yet, validated sentence-based speech-in-noise measures remain limited in South African clinical practice.
Objectives: To validate two language-specific South African sentence-in-noise tests, the South African Sentence-in-Noise (SASIN) test and Suid-Afrikaanse Sin-in-Geraas (SASIG) test, by examining inter-list comparability and determining the signal-to-noise ratio (SNR) required for 50% correct sentence recognition (SNR-50).
Method: A quantitative, retrospective validation study used routinely collected clinical data from adult first-language speakers of South African English (SASIN; n = 36) or Afrikaans (SASIG; n = 51) across multiple audiology sites. Participants completed five sentence lists in four-talker babble using fixed 5 dB SNR steps. Keyword scores were analysed for inter-list comparability and for deriving SNR-50 and SNR-loss values.
Results: Friedman analyses indicated significant overall list effects for both tests. Bonferroni-adjusted pairwise comparisons indicated that only selected comparisons, primarily involving List 1, reached statistical significance, whereas most pairwise comparisons were not statistically significant. Overall, the findings supported broad inter-list comparability.
Conclusion: The SASIN and SASIG tests demonstrated acceptable psychometric properties as language-specific sentence-in-noise measures for South African English- and Afrikaans-speaking adults, supporting their clinical use.
Contribution: This study provides validated first-language sentence-in-noise tools and test-specific SNR-50 and SNR-loss metrics for adult audiological assessment in South Africa.


Keywords

sentence-in-noise; speech-in-noise; audiology; Afrikaans; South African English; sentence recognition in babble noise; 4-talker babble noise; SNR loss.

Sustainable Development Goal

Goal 3: Good health and well-being

Metrics

Total abstract views: 245
Total article views: 199


Crossref Citations

No related citations found.