Purpose: This study investigated the relative outcomes of segmental and prosodic training of nonnative speakers of American English.
Method: The study used a single-subject, alternating treatments, multiple baseline design with replication across participants and counterbalanced for order effect. Participants were 4 adult male native Hindi speakers proficient in English. Two participants received ABABCACA (A = baseline/withdrawal, B = segmental training, C = prosody training), and 2 participants received ACACBABA, with a minimum of 5 sessions per phase. Segmental accuracy and prosodic accuracy were probed at each session, as were perception of accentedness and ease of understanding.
Results: Visual assessment of data and effect size calculation demonstrated that segmental and prosody training resulted in increased accuracy of pronunciation and prosody patterns, respectively, and those improvements appeared to be maintained over the short term. Listeners perceived lesser accent and easier understanding as a result of the combination of segmental and prosody training. The findings are uncertain with respect to the relative contribution of segmental and prosody training, and they may be speaker dependent, but the data do suggest that both components are important.
Conclusion: Accent management, consisting of both segmental and prosody training, yielded positive outcomes. Further research with native language speakers of other languages is important to verify and expand on these findings.