Identifications of speaker-ethnicity: attribution accuracy in changeable settings

Abstract

Several studies have considered the auditory identification of foreign-accented/nonnative speech. Here, a finer-grained alternative to the traditional definition of speech based on the usual criterion of birthplace alone, yet accommodates complexities which arise when dealing with speakers from countries which are historically diverse, ethnically. Using the foregoing construct, this study examines the potential for listeners to accurately group different speaker-types, with respect their ethnicity, in low/high quality transmission conditions. The findings of the Ethnic Group Attribution (EGA) task confirms overall human competence. The n = 120) brings added generalisability to smaller-scale studies. It furthermore, allows a better understanding of contextualised performance, gender-wise.