The multilingual digits-in-noise (DIN) test: development and evaluation.

Abstract

Objective: To develop a methodologically uniform digits-in-noise (DIN) test in 17 different languages.

Design: The DIN test was developed for Android devices as an extension to the open-access Hearing Test^™ app, available on the Google Play store. It utilised professionally recorded female speech, speech-shaped noise, a digit scoring method and a variable step size. The test was adaptively optimised and evaluated as the results of tests taken online by users of the app became available.

Study sample: Optimisation using 35,534 ears, evaluation using 6012 ears.

Results: Optimisation improved the slopes of the psychometric functions for all languages by an average of 6.8%/dB. Evaluation included calculation of normative speech reception thresholds (SRTs) and estimation of test-retest standard deviations. Normative values for SRTs ranged from -14.2 dB SNR (95% CI -14.3 to -14.0) for Chinese to -11.2 dB SNR (95% CI -11.3 to -11.1) for Japanese, with reliability estimates ranging from 0.48 dB (95% CI 0.36-0.64) for Portuguese to 0.91 dB (95% CI 0.73-1.21) for Romanian.

Conclusions: The optimisation of each language version was confirmed by the improvement in the slopes of the psychometric functions. The normative values obtained from the test evaluation were in agreement with literature data.

Trial registration: Science Support Centre of Wroclaw Medical University BW-59/2020.