Skip to main content
. Author manuscript; available in PMC: 2024 Apr 4.
Published in final edited form as: Behav Res Methods. 2023 Jun 8;56(3):1433–1448. doi: 10.3758/s13428-023-02101-9

Figure 6. Validation of web-based testing using word-recognition scores and consonant confusion patterns.

Figure 6.

(A) Recognition scores for monosyllabic words embedded in spectrally matched (in the long-term average sense) 4-talker babble showed overall performance and SNR dependence consistent with literature (Miner and Danhauer, 1976). (B) Consonant confusion matrix for consonants presented in C/a/ context in speech-shaped stationary noise at −8 dB SNR from N = 295 samples across four talker voices. Confusion clusters extracted from this matrix match lab-based results (Phatak and Allen, 2007). (C) A summarization of diagonal entries from panel (B) revealed that recognition scores were not uniform across consonants. Differences across sets of consonants with low, intermediate and high scores match lab-based data (Phatak and Allen, 2007).