Discourse-Based Grammar Instruction and Written Communicative Competence Among Intermediate EFL Learners: An Action Research Study in a Vietnamese Private Tutoring Context

Abstract

This qualitative action research study investigated how discourse-based grammar instruction (DBGI) influenced the written communicative competence of a small group of intermediate English-as-a-foreign-language (EFL) learners in Vietnam. The study was conducted over twelve weeks with five high-school students attending a private tutoring class in Dong Nai City. The instructional intervention consisted of ten DBGI sessions organized around five stages: Encounter, Noticing, Concept Mapping, Discrimination, and Production. Authentic or adapted texts, guided noticing, form–meaning mapping, collaborative analysis, contextualized grammar practice, and guided writing were used to help learners move beyond sentence-level rule memorization toward purposeful language use in discourse. Data were collected through pre- and post-intervention writing assessments, an open-ended learner questionnaire, and classroom observation. Learners’ writing was evaluated in relation to grammatical, sociolinguistic, discourse, and strategic dimensions of communicative competence. Descriptive results showed that all five learners improved after the intervention. The mean score increased from 5.35 on the pre-test to 7.90 on the post-test, representing a mean gain of 2.55 points. Individual gains ranged from 1.50 to 3.50 points. A descriptive paired-samples effect size was very large, , although this figure cannot be generalized because of the small sample. Qualitative findings indicated improved grammatical accuracy, more appropriate grammatical choices, clearer organization, stronger cohesion, reduced anxiety, increased confidence, and a shift from rule-based knowledge to contextualized understanding. The study suggests that DBGI can function as a practical bridge between systemic functional linguistics, communicative language teaching, and writing pedagogy. However, the findings should be interpreted as context-specific evidence rather than proof of causal effectiveness.