Universal Journal of Educational Research

ABSTRACT

Vocabulary acquisition is one of the fundamental components of second language learning, yet many students encounter difficulties in effectively using newly learned words in meaningful contexts. With the rapid advancement of educational technologies, innovative tools have emerged to support language learners in more interactive and engaging ways. One such innovation is text-to-image artificial intelligence, which has the potential to enhance vocabulary learning by providing visual representations of words, thereby fostering deeper cognitive associations. This study aimed to investigate the effectiveness of text-to-image AI tools in supporting English vocabulary acquisition among high school students learning English as a foreign language. Employing a mixed-method research design, the study included both quantitative and qualitative components. An experimental design was implemented, where the experimental group was exposed to vocabulary instruction through a text-to-image AI tool, while the control group received traditional instruction without visual AI support. Data were collected using vocabulary recognition and production tests, along with semi-structured interviews to capture the implementing teacher’s perceptions of the process. The results indicated that students in the experimental group demonstrated significant improvement in vocabulary use in the production test, although no statistically significant difference was observed in the recognition test. Moreover, the teacher's feedback highlighted the motivational benefits of AI-enhanced instruction, while also noting practical challenges, such as the occasional generation of irrelevant or misleading images and time constraints during classroom use. These findings suggest that while text-to-image AI tools hold promise for enriching vocabulary instruction, careful implementation and tool refinement are necessary to maximize their educational value.