International Journal of Pedagogics
18
https://theusajournals.com/index.php/ijp
VOLUME
Vol.05 Issue06 2025
PAGE NO.
18-20
10.37547/ijp/Volume05Issue06-05
Fostering Higher-Order Thinking Skills in ESL Classrooms
Through Ai-Supported Writing Tasks
Jurayeva Zuhra
Chirchik State Pedagogical University, Uzbekistan
Received:
10 April 2025;
Accepted:
06 May 2025;
Published:
08 June 2025
Abstract:
The rapid diffusion of generative artificial-intelligence (AI) tools such as ChatGPT is reshaping language-
learning ecologies. While early adopters have focused on gains in fluency and accuracy, a more compelling
pedagogical question is whether AI can cultivate higher-order thinking skills (HOTS)
—
analysis, evaluation and
creative synthesis
—
in English as a Second Language (ESL) learners. The present study examines an intervention
in which undergraduate ESL students completed a sequence of argumentative and reflective writing tasks
mediated by an AI co-writer that provided dynamic scaffolding, metacognitive prompts and automated discourse
analysis. Grounded in Bloom’s revised taxonomy and socio
-constructivist learning theory, the mixed-methods
design combined quasi-experimental pre- and post-testing with thematic analysis of learner journals. Quantitative
results showed statistically significant improvements (p < 0.05) in students’ HOTS rubric scores compared with a
control group engaged in traditional peer-review cycles. Qualitative data revealed heightened metalinguistic
awareness and strategic risk-taking in idea development. The findings suggest that, when carefully orchestrated,
AI-supported writing can transcend mere linguistic assistance and become a catalyst for deeper cognitive
processing. Pedagogical implications and design principles for AI-enhanced assessments that preserve academic
integrity are discussed.
Keywords:
Higher-order thinking; AI-
supported writing; ESL pedagogy; ChatGPT; Bloom’s taxonomy; academic
integrity; mixed-methods.
Introduction:
Educators have long aspired to move
second-language writing instruction beyond sentence-
level accuracy toward the cultivation of analytical and
creative reasoning. Yet empirical evidence indicates
that ESL classrooms often remain anchored in lower-
level cognitive activities, partly because teachers must
devote considerable time to error correction. The
emergence of large language models (LLMs) offers
unprecedented opportunities to redistribute this
cognitive load. Recent studies conducted in
undergraduate writing courses demonstrate that
students who receive AI-mediated formative feedback
outperform peers on measures of evaluative reasoning
and argumentation quality. However, scepticism
persists regarding over-reliance on algorithmic text
generation and the potential erosion of original
thought.
Higher-order thinking skills, situated at the apex of
Bloom’s taxonomy, encompass analyzing patterns,
evaluating evidence and producing novel syntheses.
Contemporary scholarship urges a re-examination of
these taxonomic categories in light of generative AI,
positing that strategic prompting can prompt learners
to operate at “create” and “evaluate” levels more
consistently. Nevertheless, there is limited classroom-
based research that operationalises HOTS explicitly
within AI-supported task design for ESL populations.
Addressing this gap, the present study investigates
whether structured human
–
AI collaboration during
writing tasks can measurably enhance HOTS while
maintaining language-development objectives.
This investigation adopted an explanatory-sequential
mixed-methods design that combined a quasi-
experimental component with qualitative process
tracing to elucidate how artificial-intelligence
mediation affects higher-order thinking during second-
language writing. The study unfolded over a fifteen-
week seme
ster in two intact sections of “English for
International Journal of Pedagogics
19
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
Academic Purposes II” at a public university in Central
Asia. Prior to group assignment, students (N = 54)
completed the TOEFL ITP to ensure baseline
comparability (M = 487, SD = 21); Levene’s test
confirmed homogeneity of variance (p = 0.64). The
experimental cohort (n = 27) engaged in AI-supported
writing cycles, whereas the control cohort (n = 27)
followed a traditional peer-review model.
Instructional materials were fully parallel across
conditions: three genre modules
—
argumentative
essay, problem
–
solution report and reflective blog
—
each allotted four weeks. In the experimental group,
students interacted with ChatGPT-4 on a closed
intranet at three scaffolded checkpoints within every
module: (1) ideation, where an AI prompt encouraged
abductive reasoning and analogical transfer; (2) outline
refinement, during which students negotiated thesis
–
evidence alignments through Socratic questioning
supplied by the AI; and (3) metacognitive reflection,
where learners interrogated the cognitive strategies
they employed. All AI dialogues were logged and later
subjected to discourse-analytic coding. Teacher
facilitation was standardised through a protocol that
limited
direct
linguistic
correction,
thereby
foregrounding cognitive mediation.
Data collection occurred at three junctures. First, pre-
and post-intervention essays were evaluated with a
four-dimension HOTS rubric adapted from the Critical
Thinking Assessment Test; two doctoral-level raters,
blind to condition, achieve
d κ = 0.86. Second, weekly
learning journals elicited introspective accounts of
reasoning processes; 312 entries were thematically
coded via NVivo, following Braun and Clarke’s six
-phase
procedure. Third, screen-capture recordings of AI
interactions (≈ 27
h) enabled fine-grained analysis of
epistemic moves. Quantitative effects were tested with
ANCOVA, using initial HOTS scores as covariates; effect
sizes were interpreted according to Cohen’s
conventions. Credibility of qualitative findings was
bolstered through investigator triangulation and
member checking in post-semester focus groups.
HOTS were assessed through a validated rubric
adapted from the Critical Thinking Assessment Test,
encompassing four dimensions: (1) depth of analysis,
(2) evidence evaluation, (3) creative integration of
sources and (4) reflective self-regulation. Two blind
raters scored pre- and post-semester essays with inter-
rater reliability of κ = 0.86. Additionally, weekly learner
journals
captured
perceptions
of
cognitive
engagement. Thematic coding followed Braun and
Clarke’s six
-phase approach. Quantitative data were
analysed in SPSS using ANCOVA, controlling for
baseline proficiency.
The adjusted post-test means revealed that
experimental-group students achieved higher overall
HOTS scores (M = 22.3, SD = 2.1) than control peers (M
= 18.7, SD = 2.8), yielding a medium effect size (η² =
0.21). Dimension-level analysis indicated the greatest
gains in evidence evaluation and creative integration,
aligning
with
earlier
findings
on
human
–
AI
collaborative writing. Journal analysis uncovered three
dominant themes. First, students reported a “dialogic
push,” noting that AI prompts forced them to articulate
warrants for their claims rather than accept surface-
level paraphrases. Second, learners described
heightened confidence to experiment with unfamiliar
disciplinary vocabulary because immediate AI feedback
reduced fear of lexical errors. Finally, several
respondents articulated a nuanced awareness of
authorship, deliberately negotiating which AI
suggestions to accept, modify or reject.
The statistically significant improvement in HOTS
corroborates arguments that generative AI, when
embedded within scaffolding that foregrounds
metacognition, can elevate cognitive complexity in L2
writing tasks. This outcome resonates with global
research calling for AI-resistant yet AI-enhanced
assessments that privilege reasoning over rote output.
Beyond quantifiable gains, qualitative insights highlight
the role of reflective monitoring
—students’ capacity to
interrogate AI output emerges as a vital literacy in an
era where content generation is no longer the sole
domain of humans. Contrary to critiques that AI tools
homogenise discourse, the present study observed
increased
rhetorical
originality
as
learners
appropriated or contested AI-generated ideas.
Nevertheless, sustainability hinges on pedagogical
conditions. Unstructured exposure risks cognitive
offloading; thus, teachers must design purposeful
prompts aligned with curricular goals and transparently
address ethical dimensions of assisted writing.
CONCLUSION
The evidence indicates that strategically-orchestrated
collaboration with large language models can
recalibrate ESL writing from a predominately form-
focused exercise to a cognitively demanding enterprise
that nurtures analysis, evaluation and synthesis.
Students exposed to AI scaffolding not only
outperformed peers on a validated HOTS rubric but
also demonstrated metacognitive vigilance in deciding
when to appropriate, modify or reject machine-
generated text. These outcomes challenge the binary
discourse that positions AI either as pedagogical
panacea or existential threat, suggesting instead that
learning gains hinge on the design of dialogic prompts
and the presence of an informed instructor who can re-
channel algorithmic affordances toward epistemic
International Journal of Pedagogics
20
https://theusajournals.com/index.php/ijp
International Journal of Pedagogics (ISSN: 2771-2281)
goals.
Several limitations temper the generalisability of the
findings: the sample was confined to business majors
at a single institution; writing genres were limited to
academic expository forms; and the study relied on
short-term measures of cognitive growth. Future
research should pursue longitudinal multicentre trials,
explore creative genres such as digital storytelling and
investigate automated analytics capable of issuing real-
time HOTS diagnostics without compromising data
privacy. Pedagogically, the study underscores the need
for explicit instruction in AI literacy, ethical citation of
machine assistance and assessment designs that
privilege reasoning over surface features. By weaving
reflective human judgment into every phase of AI-
mediated writing, educators can cultivate the higher-
order competencies that underpin lifelong learning in a
knowledge economy increasingly shaped by generative
technologies.
REFERENCES
Wu Y., Li S., Chen H. Effects of
ChatGPT on students’
learning performance // Humanities & Social Sciences
Communications.
–
2024.
–
Vol. 11, № 5. –
P. 1
–
14.
Lu K., Zhang J., Gao M. ChatGPT-assisted teacher
training and higher-order thinking outcomes //
International Journal of Educational Technology in
Higher Education.
–
2024.
–
Vol. 21, № 2. –
P. 45
–
63.
Khan R., Ahmad N. Impact of ChatGPT as a formative
feedback tool on ESL academic writing skills // Smart
Learning Environments.
–
2023.
–
Vol. 10, № 31. –
P. 1
–
18.
Oregon State University E-
Campus. Bloom’s taxonomy
revisited
–
artificial intelligence tools: faculty guide.
–
2024.
–
Mode
of
access:
https://ecampus.oregonstate.edu.
–
Accessed:
05.06.2025.
Jackson B. Higher-
order prompting: applying Bloom’s
revised taxonomy to the age of AI // Scholarship of
Teaching and Learning in the East.
–
2025.
–
Vol. 4, №
1.
–
P. 23
–
38.
Chen Q., Lee S. Human
–
AI collaborative writing
strategy effects on EFL argument essays // Computer
Assisted Language Learning.
–
2025.
–
Vol. 38, №
6.
–
P. 589
–
607.
Rahman J., Lim W. Meta-analysis of ChatGPT in L2
writing instruction // International Review of Research
in Open and Distributed Learning.
–
2024.
–
Vol. 25, №
3.
–
P. 120
–
139.
Smith T., O’Connor D. Triad of artificial intelligence
,
Bloom’s taxonomy and critical assessment literacy //
Education and Information Technologies.
–
2025.
–
Vol.
P. 4115
–
4134.
Pérez L., Santos V. AI-resistant assessments in higher
education: practical insights from the Global South //
Frontiers in Education.
–
2024.
–
Vol. 9.
–
Article
1499495.
–
P. 1
–
17.
Ali M., Zhang Y. Evaluating ChatGPT in EFL writing
assessment through G-theory // Humanities & Social
Sciences Communications.
–
2024.
–
Vol. 11.
–
P. 1
–
13.
Brown E. Exploring the impact of AI on critical thinking
development in ESL settings.
–
Rochester, NY: SSRN,
2025.
–
28 p.
Bloom B. S. Taxonomy of educational objectives.
Handbook I: Cognitive domain.
–
New York: David
McKay, 1956.
–
207 p.
