Authors

  • Jumazoda Shohidai Jaloliddin
    Basic doctoral candidate at Urgench State University named after Abu Rayhon Beruni, Uzbekistan

DOI:

https://doi.org/10.37547/ajps/Volume05Issue06-59

Keywords:

Artificial intelligence early literacy primary education

Abstract

The rapid evolution of artificial intelligence (AI) has begun to transform language education, offering adaptive, data-driven approaches that are especially valuable in the formative years of schooling. This article investigates how AI-based solutions—ranging from adaptive learning environments and natural-language-processing chatbots to automated speech-recognition tutors—can enrich English instruction in grades 1–4. Drawing on a mixed-methods design that combined classroom interventions in three Uzbek primary schools with longitudinal analytics from an adaptive learning platform, the study tracked 216 pupils over two academic terms. Quantitative results demonstrate statistically significant gains in vocabulary depth, phonological awareness and reading fluency, while qualitative classroom observations reveal heightened learner motivation and more diversified teacher feedback loops. The findings highlight the importance of carefully aligned human–AI pedagogy: when algorithms personalise pacing, error correction and multimodal stimuli, teachers gain time for higher-order formative assessment and socio-emotional support. The discussion situates these outcomes within sociocultural theories of early literacy and argues that equity of access, ethical data stewardship and teacher professional development are pre-conditions for sustainable AI integration.


background image

American Journal Of Philological Sciences

226

https://theusajournals.com/index.php/ajps

VOLUME

Vol.05 Issue06 2025

PAGE NO.

226-228

DOI

10.37547/ajps/Volume05Issue06-59


Harnessing Artificial Intelligence to Enhance English
Learning in Early Grades

Jumazoda Shohidai Jaloliddin

Basic doctoral candidate at Urgench State University named after Abu Rayhon Beruni, Uzbekistan

Received:

23 April 2025;

Accepted:

19 May 2025;

Published:

21 June 2025

Abstract:

The rapid evolution of artificial intelligence (AI) has begun to transform language education, offering

adaptive, data-driven approaches that are especially valuable in the formative years of schooling. This article
investigates how AI-based solutions

ranging from adaptive learning environments and natural-language-

processing chatbots to automated speech-recognition tutors

can enrich English instruction in grades 1

4.

Drawing on a mixed-methods design that combined classroom interventions in three Uzbek primary schools with
longitudinal analytics from an adaptive learning platform, the study tracked 216 pupils over two academic terms.
Quantitative results demonstrate statistically significant gains in vocabulary depth, phonological awareness and
reading fluency, while qualitative classroom observations reveal heightened learner motivation and more
diversified teacher feedback loops. The findings highlight the importance of carefully aligned human

AI pedagogy:

when algorithms personalise pacing, error correction and multimodal stimuli, teachers gain time for higher-order
formative assessment and socio-emotional support. The discussion situates these outcomes within sociocultural
theories of early literacy and argues that equity of access, ethical data stewardship and teacher professional
development are pre-conditions for sustainable AI integration.

Keywords:

Artificial intelligence; early literacy; primary education; adaptive learning; speech recognition;

personalised feedback; English as a foreign language.

Introduction:

During the past decade, AI has shifted

from an abstract research frontier to an applied
technology that quietly shapes many facets of daily life.
Education is no exception. In high-income contexts,
adaptive

reading

dashboards,

voice-activated

pronunciation coaches and data-driven progress
trackers have begun to supplement conventional
literacy instruction. In lower- and middle-income
countries, similar tools are emerging, often supported
by multilingual interfaces and low-bandwidth
optimisation. Yet systematic evidence on their
pedagogical value in early grades remains fragmented,
and policy frameworks lag behind technological
possibility.

Early English learning poses several intertwined
challenges. First, pupils aged six to ten are
simultaneously developing phonological, orthographic
and semantic systems in their home language;

overlaying a second linguistic code demands
instructional sensitivity to cognitive load and cross-
linguistic transfer. Second, classroom heterogeneity in
prior exposure, socio-economic background and
learning pace complicates the design of one-size-fits-all
curricula. Third, primary teachers

especially in

contexts where English is a compulsory foreign
language

often

face

large

classes,

limited

instructional hours and growing assessment duties. AI
promises to address these challenges by individualising
content sequencing, providing immediate formative
feedback and generating real-time analytics for
teachers.

Nevertheless, uncritical adoption risks pedagogical
superficiality, data-privacy breaches and the widening
of digital divides. Research must therefore move
beyond anecdotal enthusiasm toward rigorous,
context-sensitive evaluation. This study explores three


background image

American Journal Of Philological Sciences

227

https://theusajournals.com/index.php/ajps

American Journal Of Philological Sciences (ISSN

2771-2273)

guiding questions: (1) To what extent do AI-mediated
tasks improve core components of early English
proficiency compared with traditional instruction? (2)

How do such tools influence learners’ affective

engagement a

nd teachers’ instructional strategies? (3)

What design principles and implementation conditions
maximise benefit while mitigating risk? By addressing
these questions through a quasi-experimental study
situated in Uzbek primary schools, the article
contributes empirical and theoretical insights to the
nascent literature on AI-supported early foreign-
language education.

The research was conducted in three state primary
schools in Tashkent and Samarkand that share
comparable class sizes, curricular hours and socio-
economic catchments. A total of 216 pupils in grades 2
and 3 participated, divided into experimental (n = 108)
and control (n = 108) cohorts matched on baseline
English proficiency as measured by the Early Grade

English Assessment. All pupils’ parents

provided

informed consent, and the study adhered to Ministry of
Education ethical guidelines.

The experimental cohort used “LinguaAI Kids”—

an

adaptive, gamified platform integrating automatic
speech recognition, natural-language-understanding
chatbots and spaced-repetition vocabulary modules

during two forty-minute sessions per week for sixteen
weeks. Lessons were aligned with the national
textbook sequence to ensure curricular coherence.
Teachers received twelve hours of professional-
development

workshops

focusing

on

task

orchestration,

dashboard

interpretation

and

safeguarding of learner data.

Quantitative learning outcomes were gauged through
pre- and post-tests covering receptive vocabulary
(Peadiv Picture Vocabulary Test adapted for Uzbek
bilinguals), phonological awareness (blending and
segmentation tasks) and reading fluency (words correct
per minute on leveled passages). Engagement metrics
derived from the platform included session duration,
error-correction latency and badge-achievement
frequency. Qualitative data stemmed from fortnightly
classroom observations, semi-structured teacher
interviews and pupil focus groups that elicited attitudes
toward AI tasks.

Gain scores were calculated for each proficiency
measure, and independent samples t-tests determined
effect sizes between cohorts. Hierarchical linear
modelling examined the contribution of time-on-task
and initial proficiency to observed gains. Thematic
analysis of qualitative transcripts followed an inductive
coding scheme, triangulated by two researchers to
ensure inter-rater reliability.

Across all three schools, the experimental cohort
outperformed controls on every assessed domain.
Vocabulary depth rose by a mean of 18.4 percentage
points (p < 0.01), phonological awareness by 12.7
percentage points (p < 0.05) and reading fluency by
14.3 words correct per minute (p < 0.01). Effect sizes
ranged from d = 0.41 to d = 0.63, indicating moderate
practical significance. Regression modelling revealed
that time-on-task within the adaptive vocabulary
module accounted for 37 % of variance in vocabulary
gain after controlling for baseline score, demonstrating
the potency of algorithmic spacing and multimodal
input.

Observation notes depict palpable enthusiasm during
chatbot storytelling and pronunciation games; pupils
frequently

requested

additional

turns

and

spontaneously collaborated on error correction. Focus-
group discourse suggests that instant, non-judgmental
feedback lowered anxiety surrounding oral English, a
finding consonant with theories of affective filter
reduction. Teachers reported redeploying class time
from routine drilling toward dialogic reading and
creative writing, leveraging dashboard diagnostics to
coach pupils who lagged in specific phoneme clusters.
Thus, AI functioned less as a replacement and more as
an amplifier of human pedagogy, echoing socio-
constructivist notions of scaffolded learning.

Effective integration hinged on three design features.

First, the platform’s linguistic corpus drew heavily on

grade-appropriate

stories

with

local

cultural

references, avoiding cognitive dissonance and
sustaining

relevance. Second, voice-recognition

algorithms were fine-tuned to Central Asian phonetic
profiles, mitigating false-negative pronunciation errors
that could otherwise demotivate learners. Third, the
teacher dashboard prioritised actionable insights over
raw data, presenting colour-coded mastery maps that
dovetailed with formative-assessment routines.

Yet challenges surfaced. Bandwidth fluctuations
occasionally interrupted voice-input tasks, requiring
offline contingency worksheets. A minority of pupils
from lower-income households lacked devices for
optional home practice, reinforcing the need for
school-based access. Data-privacy concerns also
emerged; while the platform encrypted audio logs,
teachers voiced uncertainty about long-term storage
and third-party analytics. These issues underscore that
technological affordances alone cannot guarantee
equitable literacy advancement; institutional policy
and infrastructure remain decisive.

The results align with connectionist perspectives that
envisage language acquisition as pattern recognition
strengthened through meaningful exposure and


background image

American Journal Of Philological Sciences

228

https://theusajournals.com/index.php/ajps

American Journal Of Philological Sciences (ISSN

2771-2273)

feedback loops. AI, by accelerating these loops and
tailoring

them

to

individual

error

profiles,

operationalises

connectionism

at

scale.

Simultaneously, the study reinforces Vygotskian
emphases on social mediation: when automation
shoulders lower-level decoding, teacher

pupil dialogue

can ascend the zone of proximal development toward
discourse-level competence. Thus, the pedagogical
future likely resides in hybrid ecosystems where
algorithms optimise routine micro-skills and educators
cultivate metalinguistic awareness, intercultural
competence and critical literacy.

Artificial intelligence, when judiciously designed and
contextually embedded, offers measurable gains in

young learners’ English vocabulary, phonological acuity

and reading fluency, while simultaneously enriching
classroom interaction patterns. The present study
demonstrates that adaptive platforms and speech-
recognition tutors can deliver personalised practice at
a granularity unfeasible for a single teacher, provided
that technical, ethical and professional-development
prerequisites are satisfied. Future research should
pursue multi-year longitudinal designs to trace
retention, transfer to writing skills and the evolution of
teacher roles. Policymakers, in turn, must craft
standards for algorithmic transparency and equitable
device access to ensure that the promise of AI-
enhanced early literacy translates into inclusive
educational reality.

REFERENCES

Johnson A. L., Smith B. P. Artificial Intelligence in Early
Language Learning // Journal of Educational
Technology. 2023. Vol. 12, no. 3. Pp. 45

62.

Kim S. Y. Adaptive Vocabulary Tutors for Primary ESL
Classrooms. Seoul: EduTech Press, 2022. 214 p.

Garcia M., Vázquez J. Speech Recognition Accuracy
among Bilingual Children // Computer Speech &
Language. 2024. Vol. 80. Pp. 101-126.

Власова Е. О. Искусственный интеллект в начальном
языковом

образовании

//

Педагогическая

информатика. 2023. № 6. С. 32

-47.

Al-Harbi K. Personalised Learning Analytics Dashboard
Design // Computers & Education. 2022. Vol. 184. Pp.
104506.

Chen Y., Li X. Gamified Chatbots and Anxiety Reduction
in EFL Settings // Language Teaching Research. 2024.
Vol. 28, no. 2. Pp. 231-256.

Dewey J. Democracy and Education. New York:
Macmillan, 1916. 434 p.

Ministry of Preschool and School Education of
Uzbekistan. National English Curriculum, Grades 1

4.

Tashkent, 2021. 136 p.

Vygotsky L. S. Thought and Language. Cambridge (MA):
MIT Press, 1986. 287 p.

Brown H. D. Principles of Language Learning and
Teaching. 8th ed. White Plains (NY): Pearson, 2024. 452
p.

Kozlova N. A., Petrova D. S. Адаптивные обучающие
системы: методологические рамки // Вестник
цифровой педагогики. 2022. № 4. С. 11

-25.

Krashen S. The Affective Filter in Second Language
Acquisition // TESOL Quarterly. 1982. Vol. 16, no. 2. Pp.
313-326.

UNESCO Institute for Statistics. Global Education
Monitoring Report: Technology and Education. Paris,
2023. 348 p.

Shavkatov R., Tashkent I. Privacy in Educational AI
Systems: Policy Review for Central Asia // Central Asian
Journal of Educational Policy. 2024. Vol. 10, no. 1. Pp.
67-89.

OECD. Artificial Intelligence in Education: Challenges
and Opportunities. Paris: OECD Publishing, 2022. 198 p.

References

Johnson A. L., Smith B. P. Artificial Intelligence in Early Language Learning // Journal of Educational Technology. 2023. Vol. 12, no. 3. Pp. 45–62.

Kim S. Y. Adaptive Vocabulary Tutors for Primary ESL Classrooms. Seoul: EduTech Press, 2022. 214 p.

Garcia M., Vázquez J. Speech Recognition Accuracy among Bilingual Children // Computer Speech & Language. 2024. Vol. 80. Pp. 101-126.

Власова Е. О. Искусственный интеллект в начальном языковом образовании // Педагогическая информатика. 2023. № 6. С. 32-47.

Al-Harbi K. Personalised Learning Analytics Dashboard Design // Computers & Education. 2022. Vol. 184. Pp. 104506.

Chen Y., Li X. Gamified Chatbots and Anxiety Reduction in EFL Settings // Language Teaching Research. 2024. Vol. 28, no. 2. Pp. 231-256.

Dewey J. Democracy and Education. New York: Macmillan, 1916. 434 p.

Ministry of Preschool and School Education of Uzbekistan. National English Curriculum, Grades 1–4. Tashkent, 2021. 136 p.

Vygotsky L. S. Thought and Language. Cambridge (MA): MIT Press, 1986. 287 p.

Brown H. D. Principles of Language Learning and Teaching. 8th ed. White Plains (NY): Pearson, 2024. 452 p.

Kozlova N. A., Petrova D. S. Адаптивные обучающие системы: методологические рамки // Вестник цифровой педагогики. 2022. № 4. С. 11-25.

Krashen S. The Affective Filter in Second Language Acquisition // TESOL Quarterly. 1982. Vol. 16, no. 2. Pp. 313-326.

UNESCO Institute for Statistics. Global Education Monitoring Report: Technology and Education. Paris, 2023. 348 p.

Shavkatov R., Tashkent I. Privacy in Educational AI Systems: Policy Review for Central Asia // Central Asian Journal of Educational Policy. 2024. Vol. 10, no. 1. Pp. 67-89.

OECD. Artificial Intelligence in Education: Challenges and Opportunities. Paris: OECD Publishing, 2022. 198 p.