Digits-in-Noise test implementations for onsite and online testing with normal-hearing adults

Introduction: To investigate the effects of frequently used variations of the Digits-in-Noise (DIN) test on speech reception threshold (SRT) for onsite and online implementations. Methods: We tested various DIN test implementations in a within-subject design on 30 normal-hearing adults onsite, with 22 of them also online. We varied three parameters: (1) sound presentation, diotic or antiphasic; (2) starting signal-to-noise ratio (SNR), 0 or −16 dB; and (3) mixing method, with fixed presentation level of the speech, the noise, or the speech and noise mix (tested onsite only). Results: Antiphasic presentation yielded significantly lower DIN SRTs than diotic by around 6 dB. The effects of starting SNR and mixing method were significant but small, around 1−2 dB. These effects seemed more pronounced with antiphasic presentation. Overall, onsite and online DIN test results were comparable, with the largest observed difference being 1.30 dB. Conclusion: The selection of diotic or antiphasic sound presentation seems important, and antiphasic presentation may result in a more sensitive test. In contrast, the effects of the other parameters were small. The comparable onsite and online outcomes indicate that online testing via the internet could be a viable option for making the DIN test available to large populations.