Remote Matrix Sentence Test on Uncalibrated Consumer Devices: Test-Retest Reliability and Laboratory Agreement in Normal Hearing.
Thomas Schwarz, Lena Schell-Majoor, Samira Saak and 2 others
Remote German Matrix Sentence Test scores on uncalibrated consumer devices did not significantly differ from lab scores (mean gap 0.29 dB) and were stable across repeated sessions.
All 20 participants had normal hearing and were mostly in their 20s–30s, so this does not extend to hearing-impaired patients.
Key findings
1The remote GMST did not produce significantly different speech-in-noise scores from the laboratory version: the overall test effect was non-significant (p = 0.13) and the paired lab–remote difference was 0.29 dB (p = 0.08, r = 0.76).
2Repeated remote sessions were stable: the test–retest difference was 0.28 dB (p = 0.15), with a correlation of r = 0.82 and single-measure agreement ICC(2,1) of 0.80.
3Participants who had previously taken the GMST scored about 1.5 dB better than those who had not, a difference that was statistically significant (all p < 0.05).
Still to come
How it was doneWhat they found
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Only 20 participants, all with normal hearing and mostly in their 20s–30s (mean 33.5); the authors state generalisability to hearing-impaired listeners is limited. Eleven of the 20 had already taken the GMST before the study, and this familiarity produced a significant 1.5 dB advantage, inflating the between-subject spread. The remote platform was hosted on a university laptop and accessed only via the university network, not the public internet, so real-world connectivity variability was not tested. Participants used uncalibrated consumer devices and headphones; absolute presentation levels were not recorded and adherence to the 'quiet room' instruction could not be verified. The sample was predominantly male (16 of 20) and all native German speakers; findings may not transfer to other languages or demographics. The supplementary dichotic-vs-diotic comparison used 15 participants, 12 of whom overlapped with the main sample, limiting independence.
Declared interests
Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy (Project IDs 390895286 and 512607743). The authors declared no conflicts of interest.
The easy way to misread this
Do not read this as validation of remote speech-in-noise testing for hearing-impaired patients — all 20 participants had normal hearing, were mostly in their 20s–30s, and 11 of 20 had already taken the GMST before the study, so the agreement and stability you see here may not hold for the patients you actually see in clinic.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →