Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 69 of the 69 references it cites were found in a registry or an index; none of the 47 sources its claims cite is checked for support yet.

‘AI-Learner Partnership’: Psychological mechanisms and developmental trajectories of EFL learners’ writing agency in GenAI-assisted courses

Open access
Zhang, Lei, Wei, Rining, Jiang, Chunli · 2026 · Acta Psychologica

Abstract

This study explores how EFL learners construct, negotiate, and develop writing agency through engagement with generative artificial intelligence (GenAI) in second language writing courses. GenAI is conceptualized as a cultural mediational means within an integrated sociocultural and positive psychological framework. Eleven Chinese EFL university students were followed over a 16-week semester as they completed multiple academic writing tasks with GenAI support. Drawing on a qualitative multi-method design, data were collected through semi-structured interviews, screen recordings of writing sessions, and reflective journals. The analysis identified three interrelated, non-linear phases in learners’ agency development: initial exploration (characterized by novelty-driven engagement, dependency on AI output, and emergent self-efficacy), strategic adaptation (marked by more selective, task-contingent tool use and the strengthening of writing-specific psychological capital), and deliberate appropriation (signaled by personalized integration of GenAI into writing workflows and the emergence of a more agentic techno-authorial identity). Across these phases, cognitive processes, affective experiences, positive psychological resources, and task design features jointly shaped how learners mobilized technological affordances to enact writing agency. The findings suggest that writing agency in human–AI collaborative writing is a dynamic, context-sensitive accomplishment rather than a fixed individual trait. By connecting sociocultural accounts of mediated action with positive psychological resources, the study offers a process-oriented account of how EFL learners negotiate assistance, authorship, and autonomy in GenAI-supported academic writing.

Overview

CoChat Summary

TL;DR

This qualitative case study of 11 Chinese EFL undergraduates over a 16-week GenAI-supported writing course traces how writing agency develops through initial exploration, strategic adaptation, and deliberate appropriation. It finds agency is mediated by cognitive, affective, and task-design factors and manifests in transitional, distributed, reflective, and transformative patterns.

Key findings

Learners’ agency trajectories evolved through three developmental phases: initial exploration (novelty, dependency, emergent self-efficacy), strategic adaptation (selective tool use, growing writing-specific psychological capital), and deliberate appropriation (personalized integration, flourishing techno-authorial identity).

  • Four patterns of agency manifestation emerged: transitional agency, distributed agency, reflective agency, and transformative agency.
  • Cognitive processes, affective dimensions, positive psychological resources, and task design characteristics significantly mediated how learners constructed agency through interactions with technological affordances.
  • The study involved 11 Chinese EFL university learners observed over a 16-week period; interviews lasted 30–45 minutes and yielded 59,613 words of transcribed data, and the transition to strategic adaptation typically emerged after 2–3 writing tasks.

Methods

Qualitative case study in an undergraduate English writing course at a university in China; participants were 11 EFL undergraduates (7 males, 4 females) with intermediate to advanced English proficiency, for whom this was the first experience using AI tools for academic writing; GenAI tools (primarily Doubao) were optional resources throughout planning, drafting, and revision.

  • Primary data were semi-structured interviews conducted at the end of the semester, each 30–45 minutes, in Mandarin Chinese, yielding 59,613 words of transcribed data; supplementary sources included self-reported summaries of emotional responses to AI-assisted writing and coded analyses of self-regulated writing strategies across cognitive, metacognitive, social, and motivational dimensions. The abstract also lists screen recordings, stimulated recall interviews, reflective journals, and sequential drafts with AI interaction logs.
  • Interview data were analyzed using Braun & Clarke’s six-phase thematic analysis informed by sociocultural theory; coding followed open, axial, and selective stages, and individual case narratives mapped agency trajectories over time.

Limitations

The paper does not state an explicit limitations section.

  • Evident from the text: the study is a small qualitative case study of 11 EFL undergraduates at one Chinese university in one semester-long course, so findings are not generalizable; primary data are end-of-semester semi-structured interviews plus supplementary self-reports and coded strategy analyses, making the study reliant on self-report.
  • Evident from the text: the data-collection description is inconsistent—the abstract mentions screen recordings, stimulated recall interviews, reflective journals, and sequential drafts with AI interaction logs, while the methods section says the primary data source was semi-structured interviews plus supplementary self-reported summaries and coded analyses of self-regulated writing strategies.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.actpsy.2026.107983

Cites

69 of 69 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 Valid · 2 Not valid · 33 Not checked · 91 Own results and interpretations

68 more claims

Create a free account to see all 126 claims and to verify the 11 references still open.

See all claims — free account
  • It may state two separate claims as one. 40
  • The quote only partly states this claim. 1
  • Probably not a substantial claim. 17

Create a free account to see these claims.

Citations

Cites

Show all 65 references

Research works printed but not linked: 2

  1. Dattathrani, 2023, Machine agency and human learning: Emerging AI mediations in education, Information and Organization, 33
  2. Teng, 2016, A questionnaire-based validation of the self-regulated learning strategies in writing scale, Language Testing, 33, 493

Metadata

Authors
Zhang, L., Wei, R., & Jiang, C.
Affiliations
Beijing University of Technology (CN)
Corresponding author
Jiang, Chunli
Venue
Acta Psychologica
Publisher
Elsevier BV
Published
2026-10
Volume
270
Pages
107983
ISSN
0001-6918 (ISSN-L)
OpenAlex
W7222418727
Cited by
2
References
65
Citations by year
2026: 2
FWCI
13.98
Type
journal article
Language
en
Open access
Licence
Locations
  • Acta Psychologica (publishedVersion, cc-by-nc-nd)
  • Research Square (submittedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
China Scholarship Council (202506540103)
Topics
  • Social Sciences › Arts and Humanities › Language and Linguistics › EFL/ESL Teaching and Learning
  • Social Sciences › Education › English Language Learning and Teaching
  • Social Sciences › Education › Artificial Intelligence in Education
Keywords
generative AI, EFL writing, AI dependency, self-efficacy, psychological capital, authorship, learner autonomy, mediation theory, positive psychology, academic writing, Chinese university students
MeSH terms
Generative Artificial Intelligence, Sense of Agency, Adult, Female, Humans, Learning, Male, Students, Writing, Multilingualism, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-18, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-18 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: ‘AI-Learner Partnership’: Psychological mechanisms and developmental trajectories of EFL learners’ writing agency in GenAI-assisted courses. https://app.cochat.ai/papers/10.1016/j.actpsy.2026.107983
The paper
Zhang, L., Wei, R., & Jiang, C. (2026). ‘AI-Learner Partnership’: Psychological mechanisms and developmental trajectories of EFL learners’ writing agency in GenAI-assisted courses. Acta Psychologica. https://doi.org/10.1016/j.actpsy.2026.107983

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.