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

The DOI matches its Crossref record; 41 of the 45 references it cites were found in a registry or an index; 33 of the 72 sources its claims cite were checked for support.

From tool to scaffold: structured human–AI collaboration and its effects on academic writing and digital critical thinking among Saudi EFL learners

Open access

Abstract

This mixed-methods quasi-experimental study examined whether pedagogically structured, AI-supported writing instruction is associated with improvements in EFL academic writing performance and digital critical thinking/AI literacy among Saudi university students. Fifty-three undergraduates were assigned by intact classes to an experimental group ( n = 31) or a control group ( n = 22). Over 10–11 weeks, the experimental group practiced guided human–AI collaboration workflows—encompassing problem framing and prompt design, iterative drafting, revision cycles, verification of claims and citations, and responsible-use regulation—using a generative AI assistant and a language-feedback tool, while the control group completed the same syllabus without systematic AI integration. Both groups completed parallel pretest/posttest writing tasks (0–20 scale) and the Digital Critical Thinking Scale (DCTS; 20 items; 20–100; α = 0.97 pretest, α = 0.99 posttest). At posttest, the experimental group scored significantly higher on writing proficiency ( M = 15.23 vs. 11.91; large effect) and DCTS total ( M = 89.84 vs. 56.18; very large effect). Writing proficiency and DCTS were strongly associated at both posttest and in gain scores. Thematic analysis of experimental-group reflections revealed that students employed AI primarily as a planning and revision scaffold, while simultaneously enacting verification routines and articulating ethical self-regulation strategies. This study contributes context-specific empirical evidence demonstrating that agency-preserving, critically guided AI integration can simultaneously advance writing quality and digital critical thinking in EFL higher education contexts where such paired outcome evidence remains scarce. However, results should be interpreted as preliminary and institution-specific, given the modest sample, intact-class assignment, and single-institution context, which limit generalizability to other EFL populations and settings.

Overview

CoChat Summary

TL;DR

This mixed-methods quasi-experimental study found that structured, agency-preserving human–AI writing instruction was associated with large gains in EFL academic writing and digital critical thinking among Saudi university students compared with conventional instruction. The authors interpret the results as preliminary and institution-specific because of the modest sample, intact-class assignment, and single-institution context.

Key findings

  • Writing proficiency at posttest was higher in the experimental group (M = 15.23, SD = 1.65) than the control group (M = 11.91, SD = 1.44), with a mean difference of 3.32 points, 95% CI [2.44, 4.20], Hedges’ g = 2.09, p < 0.001; gains were 4.03 vs. 0.73 on the 0–20 scale.
  • DCTS total at posttest was higher in the experimental group (M = 89.84, SD = 10.25) than the control group (M = 56.18, SD = 8.94), with a mean difference of 33.66 points, 95% CI [28.21, 39.11], Hedges’ g = 3.41, p < 0.001; gains were 32.81 vs. 0.05 on the 20–100 scale.
  • Writing proficiency and DCTS were strongly correlated at pretest (r = 0.81, 95% CI [0.69, 0.89]), posttest (r = 0.94, 95% CI [0.90, 0.97]), and in gain scores (r = 0.93, 95% CI [0.88, 0.96]), all p < 0.001.
  • Thematic analysis of experimental-group reflections yielded four themes: AI as a flexible writing scaffold (100%), detecting and repairing AI errors (100%), verification as digital critical practice (81%), and metacognitive regulation/ethical positioning (100%).

Methods

  • Convergent mixed-methods quasi-experimental pretest–posttest control-group design; 53 Saudi undergraduate EFL students in a required academic writing course at a Saudi public university, assigned by intact classes to experimental (n = 31) or control (n = 22) groups; intervention implemented over 11 weeks.
  • Experimental group received guided human–AI collaboration workflows covering problem framing/prompt design, iterative drafting/revision, verification/triangulation, and responsible-use regulation; control group completed the same syllabus without systematic AI integration.
  • Measures included a timed 45-minute academic essay scored 0–20 with an analytic rubric and the 20-item Digital Critical Thinking Scale (DCTS; 20–100; α = 0.97 pretest, α = 0.99 posttest); analysis used independent- and paired-samples t-tests, Hedges’ g, Pearson correlations, and thematic analysis of experimental-group reflections.

Limitations

  • The paper states that the modest sample (N = 53), intact-class assignment, and single-institution context limit generalizability; observations may not be fully independent due to intraclass clustering, standard t-tests may underestimate standard errors and produce optimistic p-values, and with only two clusters multilevel modeling was not estimable.
  • The paper also states that posttest t-tests do not statistically control for residual pretest variability, ANCOVA would be more precise, writing was assessed via one task per time point without reported inter-rater reliability or blinded scoring, DCTS relied on self-report with high alpha possibly reflecting item redundancy, and qualitative reflections came only from the experimental group.
  • Evident from the text: the exceptionally large effect sizes may partly reflect novelty effects, greater instructional intensity in the experimental condition, DCTS responsiveness to practiced intervention content, and item redundancy; the authors therefore describe effect magnitudes as preliminary upper-bound estimates pending replication.

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

Integrity signals

Registry

Crossref · 10.3389/fpsyg.2026.1830103

Cites

41 of 45 references verified2 citation errors · 2 not checkable

Statistics

Figures

Figure 1 Posttest writing proficiency versus DCTS scores.
Figure 1 Posttest writing proficiency versus DCTS scores.CC BY · PubMed Central

Verification

Claims the paper makes

22 Valid · 3 Not valid · 7 Not checked · 67 Own results and interpretations

39 more claims · 1 with their cited sources checked

Create a free account to see all 99 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 46
  • Probably not a substantial claim. 16
  • The quote only partly states this claim. 4

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Consistentt(51) = 0.03reported p = .977 · recomputed p = .976

    “t(51) = 0.03, p = .977”

  • Consistentt(51) = 0.33reported p = .741 · recomputed p = .743

    “t(51) = 0.33, p = .741”

  • Consistentt(30) = 46.58reported p < .001 · recomputed p < .001

    “t(30) = 46.58, p < .001”

  • Consistentt(21) = 7.48reported p < .001 · recomputed p < .001

    “t(21) = 7.48, p < .001”

  • Consistentt(51) = 7.6reported p < .001 · recomputed p < .001

    “t(51) = 7.60, p < .001”

5 of 8 · the rest aren’t loaded

3 more recomputed tests

Create a free account to see every test the Stats check recomputed.

See all tests — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 41 references

Research works printed but not linked: 7

  1. Ahmed, 2024, The impact of ChatGPT on the writing skills of Saudi EFL university students: a mixed-methods study, Int. J. Interact. Mob. Technol., 18, 4, 10.3991/ijim.v18i06.48122 · doi:10.3991/ijim.v18i06.48122
  2. Al-Dossary, 2024, Attitudes and intentions toward using artificial intelligence for writing among Saudi university students, Int. J. Engl. Linguist., 14, 1
  3. Alghamdi, 2024, The effect of AI-based applications on enhancing vocabulary and reading skills among Saudi EFL learners, Engl. Lang. Teach., 17, 23
  4. Al-Harbi, 2025, ChatGPT in the Saudi EFL context: digital critical thinking and academic performance, Front. Psychol., 15, 1334512, 10.3389/fpsyg.2024.1334512 · doi:10.3389/fpsyg.2024.1334512
  5. Alhathli, 2024, AI-enhanced courses and the development of analytical reasoning and data literacy in Saudi universities, J. King Saud Univ. Educ. Sci., 36, 101
  6. Al-Housni, 2024, The impact of artificial intelligence tasks on developing critical thinking skills in gulf higher education, Int. J. Instr., 17, 405
  7. Aluthman, 2024, An investigation of artificial intelligence tools in editorial tasks among Arab researchers publishing in English, Eurasian J. Appl. Linguist., 10, 174, 10.32601/ejal.10116 · doi:10.32601/ejal.10116

Metadata

Authors
Alshehri, O. A. O., Lamouchi, A., Abdellatif, M. S., Al-Dosari, M. N. A., Mekheimer, M., & Nemt-allah, M. A.
Affiliations
  • University of Bisha (SA)
  • Prince Sattam Bin Abdulaziz University (SA)
  • Beni-Suef University (EG)
  • Al-Azhar University (EG)
Corresponding author
Lamouchi, Ali
Venue
Frontiers in Psychology
Published
2026-05-26
Volume
17
Pages
1830103-1830103
ISSN
1664-1078 (ISSN-L)
OpenAlex
W7162408299
Cited by
2
References
41
Citations by year
2026: 2
FWCI
40.21
Type
journal article
Language
en
Open access
Locations
  • Frontiers in Psychology (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
Prince Sattam bin Abdulaziz University
Topics
  • Social Sciences › Arts and Humanities › Language and Linguistics › EFL/ESL Teaching and Learning
  • Social Sciences › Education › Artificial Intelligence in Education
  • Social Sciences › Education › English Language Learning and Teaching
Keywords
human-AI collaboration, generative AI in education, EFL academic writing, AI literacy, writing scaffolding, AI-assisted writing instruction, Saudi university students, responsible AI use, ethical self-regulation

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.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

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: From tool to scaffold: structured human–AI collaboration and its effects on academic writing and digital critical thinking among Saudi EFL learners. https://app.cochat.ai/papers/10.3389/fpsyg.2026.1830103
The paper
Alshehri, O. A. O., Lamouchi, A., Abdellatif, M. S., Al-Dosari, M. N. A., Mekheimer, M., & Nemt-allah, M. A. (2026). From tool to scaffold: structured human–AI collaboration and its effects on academic writing and digital critical thinking among Saudi EFL learners. Frontiers in Psychology. https://doi.org/10.3389/fpsyg.2026.1830103

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.