Impact of ELSA Speak Application on English Consonant Pronunciation Accuracy among Grade 10 ESL Learners in the Educators School, Nowshera

Abstract

English pronunciation is a key skill for communicating with others, but it is among the least developed in a typical Pakistani schools ESL (English as a second language) class. One of the common challenges Pakistani learners face with English is its phonological system, which lacks consonant sounds from their L1. This study investigates English consonant accuracy gain between the use of ELSA Speak, an AI-based pronunciation application, and Grade 10 English as an additional language (ESL) students in a Private school located in Nowshera, Pakistan. Thirty-five Grade 10 students were involved during a 5-week period. The instruments used were a consonant word list, a pronunciation test and a perception questionnaire. Means, percentages, and a paired-samples t-test were used to analyze the data. Results have shown that consonant accuracy significantly improved after the intervention. There were some positive changes for the /θ/, /v/, /z/ and /ð/ sounds. Students' views about ELSA Speak were very positive. The study validates that ELSA Speak is an effective, easily accessible tool for pronouncing consonants at the word level in under-resourced Pakistani ESL contexts.