Emotional dimensions of feedback: How AI and human responses shape ESL learning outcomes
The provision of feedback remains one of the most potent instructional interventions within second language acquisition, yet the affective mechanisms underlying its efficacy are still poorly understood. This study investigates how feedback type, specifically AI-generated versus teacher-provided feed...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-12-01
|
| Series: | Ampersand |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2215039025000190 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The provision of feedback remains one of the most potent instructional interventions within second language acquisition, yet the affective mechanisms underlying its efficacy are still poorly understood. This study investigates how feedback type, specifically AI-generated versus teacher-provided feedback, interacts with learners' Trait Emotional Intelligence (TEI) and Foreign Language Enjoyment (FLE) to influence language proficiency development. Adopting a quasi-experimental design with a purely quantitative methodological orientation, the research recruited 63 intermediate-level English as a Second Language (ESL) learners and assigned them randomly to either an AI feedback group or a teacher feedback group. Participants completed five academic writing and speaking tasks over six weeks, each followed by an immediate feedback and revision cycle. Measurements included pre- and post-intervention language proficiency assessments, alongside the administration of validated scales for TEI and FLE. Structural Equation Modeling (SEM) was employed to examine both direct and mediated pathways between variables. Results revealed that TEI significantly predicted learners' levels of FLE, which, in turn, significantly mediated the relationship between feedback type and language proficiency improvement. Teacher feedback demonstrated a stronger positive effect on FLE compared to AI feedback. The SEM model exhibited excellent fit indices, confirming the robustness of the hypothesized structure. These findings underscore the importance of addressing emotional dimensions in feedback practices, suggesting that optimal language learning outcomes arise not merely from the cognitive correction of errors but also from the emotional resonance that feedback generates. Implications are discussed for pedagogical practices, AI design in language education, and the broader field of affective second language acquisition research. |
|---|---|
| ISSN: | 2215-0390 |