Scaffolded English-dominant interaction framework with young A1 EFL learners: an embedded mixed-methods classroom pilot in Ecuador

Abstract

Maximizing target-language exposure may increase opportunities for oral practice in English as a Foreign Language (EFL) classroom; however, rigid English-only practices may restrict beginning learners’ comprehension and participation. This classroom pilot study introduced and examined the classroom implementation of the Scaffolded English-Dominant Interaction Framework, which combines comprehensible input, teacher scaffolding, communicative interaction, and progressively independent oral production. A quantitative-dominant embedded mixed-methods design was employed with 30 fourth-grade A1 learners aged 8-9 years at a private school in Ecuador. Participants completed parallel speaking tasks before and after a 12-session intervention. Performance was rated using a three-point analytic rubric addressing reliance on the first language (L1), speech pauses, and hesitations. The qualitative process was obtained from 12 structured teacher reflection journals and analyzed thematically. The median composite score increased from 5 to 8. The primary two-sided Wilcoxon signed-rank test indicated a statistically significant within-group change, W   =   435.00, p   <   .001, r rb   =   1.00, with 29 positive differences and one tie. A paired sample t-test conducted as a sensitivity analysis produced the same conclusion, t (29) = 9.95, p  < .001, d z  = 1.82. Teacher journals described an overall but task-dependent pattern of greater English-mediated participation, less reliance on Spanish, more complete responses, and fewer extended pauses and hesitations during later structured tasks. The journals also showed that instructional scaffolding remained important throughout the intervention. The journal evidence complemented the quantitative findings by contextualizing the observed changes in classroom-rated oral performance across intervention. However, the one-group design, ordinal classroom ratings, and reliance on teacher-generated quantitative and qualitative evidence preclude causal conclusions and limit the independence of the two data strands. The study documents the classroom implementation of the framework and provides preliminary evidence to guide its evaluation through controlled designs and direct temporal-fluency measures.