A CASE STUDY ON THE IMPACT OF THE ELSA ENGLISH-SPEAKING LEARNING APP ON HOTEL STAFF’S COMMUNICATION PROFICIENCY AT PHAM GIA HOTEL

Tóm tắt

The hospitality industry demands strong English communicative proficiency from frontline staff, yet hotel workers frequently face compounding barriers including rotating shifts, language anxiety, and the absence of accessible training pathways. This study investigates the impact of the ELSA Speak application — an AI-powered Mobile- Assisted Language Learning (MALL) tool — on the English communicative proficiency of hotel staff at Pham Gia Hotel in Phan Thiet, Vietnam, and identifies which specific app activities most effectively support speaking skill development. The study is theoretically grounded in Self-Determination Theory (Ryan & Deci, 2000) and Canale and Swain's (1980) model of Communicative Competence. A mixed-methods case study design was employed, incorporating a quasi-experimental pre-test/post-test control group structure within the quantitative strand. Fifteen hotel staff members were purposively assigned to an Experimental Group (EG, n=8) and a Control Group (CG, n=7). The EG underwent a structured five-week ELSA Speak intervention of 15 minutes daily self-directed practice, while the CG continued normal work routines without supplementary training. Data were collected through three instruments: a hospitality-specific speaking test assessed against a four-component rubric; a structured Likert-scale questionnaire comprising two reliability-verified constructs; and semi- structured interviews with all EG participants. Quantitative data were analyzed through Pearson correlation analysis to establish inter- rater reliability, Normal Q-Q plots and Levene's Test to verify parametric assumptions, and Independent Samples T-tests to evaluate between-group differences at both pre-test and post-test stages. Internal consistency of questionnaire constructs was confirmed via Cronbach's Alpha. Qualitative interview data were examined through thematic analysis. The findings revealed statistically significant improvement in the EG's post-test speaking performance relative to the CG (p < 0.05), while pre-test equivalence confirmed a valid experimental baseline. Questionnaire results identified AI-powered pronunciation feedback, hospitality-specific interactive dialogues, and gamified self-assessment as the three most effective ELSA features. Thematic analysis produced five recurring themes, most notably reduced language anxiety, enhanced communicative confidence, and increased intrinsic motivation for self-directed study. These findings suggest that structured, short-duration MALL interventions represent a viable, psychologically supportive, and cost-effective supplementary training model for rotating-shift workers in Vietnam's hospitality sector.