THE EFFECT OF AI PORTRAIT PROMPT WRITING ON VOCABULARY DEVELOPMENT OF SECONDARY ESL LEARNERS

Abstract

English vocabulary knowledge is considered as the prerequisite for second language competence and in Pakistan the instruction of English vocabulary in public secondary schools is mostly text book-based, formal and rote learning which results in superficial and superficial language knowledge in second language learners (ESL). Generative text-to-image AI creates a novel pedagogical task, AI portrait prompt writing, where learners create text-based prompts to be translated into visual portraits, potentially involving both comprehensible input, pushed output, and tool-mediated scaffolding, and dual verbal-visual coding. The aim of the present study was to examine the impact of using AI-generated writing prompts for portraits on the acquisition of the vocabulary of secondary English learners and its comparative effectiveness with traditional vocabulary teaching. A true experimental pre test – post test control group design was used in which the experimental group (n = 30) and the control group (n = 30) Grade 9 learners were randomly selected from a government secondary school in Tehsil Sadiqabad, District Rahim Yar Khan. The experimental group was taught for 8 weeks using AI portrait prompt writing whereas the same content was taught to the control group using traditional ways. The researcher derived 40-item Vocabulary Achievement Test (VAT) was used to assess vocabulary development with the test administered both before and after intervention. The two groups were statistically equivalent at pretest, t(58) = 0.162, p > .05, but significantly different by the posttest, t(58) = 7.3, p < .001, with a large effect size (Cohen's d ≈ 1.9). The results show that AI-generated portrait prompt writing results in a substantial increase in vocabulary compared to traditional methods, suggesting its viability for use in ESL vocabulary instruction in disadvantaged public schools.