Evaluating English Teachers’ Pronunciation Skills (E-TPS) Using Praat Technology and Human Raters

Abstract

When the goals in teaching pronunciation are that students attain “near-native” and “socially acceptable” pronunciation, and are assessed with a valid and reliable instrument using the Praat device and human interpretation raters, a new way to assess pronunciation skills is needed. This article proposes such a new way. Through suprasegmental identity (rhythm), English teachers’ pronunciation skill (E-TPS) will be correctly assessed. Using Praat as a device and human raters for interpretation to assess E-TPS stands on several premises. First, the measures of pronunciation constructs should be valid and reliable. Second, the role of raters in English pronunciation assessment presents unique challenges in drawing valid inferences from performance, in scoring assignments, in making decisions for which a pronunciation assessment was intended, to the pedagogical and social consequences beyond, and to connect concerns in giving judgments. Third, the raters should have familiarity with both accent and content. Fourth, a suprasegmental (stress, rhythm, and intonation) approach is an important complement to segmental teaching. Fifth, rhythm plays a big role and is defined as a “continuum” of functions and effects in pronunciation assessment. Sixth, the trend among the technology-minded for testing using automated assessment of pronunciation has as its goal accuracy in imitating native speakers, as is reflected in the use of Praat as a device. Therefore, the blending approach (Praat and raters) represents the future in assessing E-TPS.