Skip to main content
. 2021 May 12;64(6 Suppl):2134–2153. doi: 10.1044/2021_JSLHR-20-00257

Figure 6.

Figure 6.

Mean phrasal stress classification accuracy across speakers for tokens produced by natural voices (NV), synthetic voices matched to speakers with typical voices (SVC), electrolaryngeal voices (EV), and synthetic voices matched to speakers with laryngectomy (SVL). Error bars are 95% confidence intervals. *p < .05.