Abstract

This study compares the effectiveness of Generative AI-supported instruction and strategic dictionary use on upper elementary students’ receptive and expressive vocabulary development. A quasi-experimental pretest–posttest design was employed with 136 students aged 10–12, including monolingual and bilingual learners, who were assigned to three conditions: Generative AI-supported instruction, strategic dictionary use, and a control group. Receptive vocabulary was assessed with the LEXiS standardized test, while expressive vocabulary was measured using the Greek adaptation of the Renfrew Word Finding Vocabulary Test. Both interventions focused on climate change vocabulary and were implemented over four weeks, incorporating parallel receptive and productive vocabulary-learning activities. General Linear Models revealed significant effects of intervention type on receptive vocabulary development, with both the AI-supported and dictionary-based groups outperforming the control group. No significant differences emerged between the two experimental conditions. For expressive vocabulary, no overall intervention effect was observed; however, a significant interaction between intervention type and bilingualism indicated that bilingual students in the dictionary-based condition achieved significantly greater productive vocabulary gains than their peers in the AI and control groups. The findings suggest that Generative AI and strategic dictionary instruction support different dimensions of lexical development and should be viewed as complementary rather than competing pedagogical approaches.

Keywords