A Scaffolded Data-Driven Learning approach to EFL synonym differentiation: an extended unit of meaning perspective

Abstract

This study investigates the effectiveness of an Extended Unit of Meaning (EUM)-informed scaffolded Data-Driven Learning (DDL) approach for helping EFL learners differentiate near-synonymous verbs. To support learners’ phraseological induction during concordance observation, the study proposes the Observe–Mark–Write–Induce (OMWI) framework as a scaffolded instructional model. A total of 60 first-year English-major students at a Chinese university participated in the study, with one group receiving EUM-informed DDL-based instruction and the other receiving traditional vocabulary instruction. Pre- and post-tests provided the primary evidence of learning gains, while a post-treatment questionnaire supplied complementary evidence concerning learners’ perceived awareness of collocation, colligation, and semantic preference. The results showed that learners receiving EUM-informed DDL-based instruction achieved significantly greater gains in synonym differentiation than those receiving traditional instruction. The questionnaire responses further suggest that learners perceived the OMWI-supported concordancing activities as helpful for noticing recurrent usage patterns and understanding contextual differences between near-synonymous verbs. Although learners reported certain challenges in interpreting concordance evidence, they generally expressed positive attitudes toward DDL-based synonym learning. The study contributes to corpus-informed vocabulary instruction by showing how scaffolded concordance observation can direct learners’ attention to phraseological patterns and support synonym differentiation in EFL classrooms.