Low-Resource Strategies for IELTS Writing Preparation using Free Generative AI
DOI:
https://doi.org/10.3991/ijac.v19i1.58935Keywords:
IELTS Academic Writing, generative AI, scoring agreement, feedback quality, low resource settingsAbstract
The study assesses whether free, web accessible generative AI can produce IELTS Academic Writing band scores and feedback credible for formative use in low-resource contexts. A corpus of 110 official Task 1 and Task 2 responses was scored by four systems accessed at no cost: ChatGPT with GPT 4, Gemini 1.5 Flash, Claude 3.5 Sonnet, and DeepSeek. Agreement with official scores was quantified using Lin’s concordance correlation coefficient and the Bland-Altman method. Feedback was coded for structure, depth, specificity, and tone, with intercoder reliability estimated by Krippendorff’s alpha. Two systems showed minimal central bias with limits of agreement, while one system systematically underestimated bands. Agreement within a one-band tolerance was high. Feedback was criterion-linked and actionable but often lacked precise phrasing and had uneven clarity. Qualitative coding reliability was acceptable. Findings suggest that free AI can offer indicative banding and revision cues when combined with rubric checking and human oversight, especially in situations where bandwidth is limited. The study logs access conditions to support replication and sets priorities for self-directed learning.
References
[1] J. Hattie and H. Timperley, "The Power of Feedback," Review of Educational Research, vol. 77, no. 1, pp. 81-112, 2007, https://doi.org/10.3102/003465430298487
[2] V. J. Shute, "Focus on Formative Feedback," Review of Educational Research, vol. 78, no. 1, pp. 153-189, 2008, https://doi.org/10.3102/0034654307313795
[3] M. Younas, D. Abdel Salam El-Dakhs, and Y. Jiang, "A Comprehensive Systematic Review of AI-Driven Approaches to Self-Directed Learning," IEEE Access, vol. 13, pp. 38387-38403, 2025, https://dx.doi.org/10.1109/access.2025.3546319
[4] F. Jin, C.-H. Lin, and C. Lai, "Modeling AI-Assisted Writing: How Self-Regulated Learning Influences Writing Outcomes," Computers in Human Behavior, vol. 165, p. 108538, 2025/04/01/ 2025, https://doi.org/10.1016/j.chb.2024.108538
[5] J. Huang and A. Mizumoto, "The Effects of Generative AI Usage in EFL Classrooms on the L2 Motivational Self System," Education and Information Technologies, vol. 30, no. 5, pp. 6435-6454, 2025/04/01 2025, https://doi.org/10.1007/s10639-024-13071-6
[6] J. Kim, S. Yu, R. Detrick, and N. Li, "Exploring Students’ Perspectives on Generative AI-Assisted Academic Writing," Education and Information Technologies, vol. 30, no. 1, pp. 1265-1300, 2025/01/01 2025, https://doi.org/10.1007/s10639-024-12878-7
[7] C. Song and Y. Song, "Enhancing Academic Writing Skills and Motivation: Assessing the Efficacy of ChatGPT in AI-Assisted Language Learning for EFL Students," Frontiers in Psychology, vol. 14, 2023, https://dx.doi.org/10.3389/fpsyg.2023.1260843
[8] B. J. Zimmerman, "Becoming a Self-Regulated Learner: An Overview," Theory Into Practice, vol. 41, no. 2, pp. 64-70, 2002, https://dx.doi.org/10.1207/s15430421tip4102_2
[9] J. Read, "Test Review: The International English Language Testing System (IELTS)," Language Testing, vol. 39, no. 4, pp. 679-694, 2022, https://doi.org/10.1177/02655322221086211
[10] Z. Ahmed et al., "The Generative AI Landscape in Education: Mapping the Terrain of Opportunities, Challenges, and Student Perception," IEEE Access, vol. 12, pp. 147023-147050, 2024, https://dx.doi.org/10.1109/access.2024.3461874
[11] C. Rose et al., "A Conference (Missingness in Action) to Address Missingness in Data and AI in Health Care: Qualitative Thematic Analysis," Journal of Medical Internet Research, vol. 25, p. e49314, 2023, https://dx.doi.org/10.2196/49314
[12] Z. Lin, "Techniques for Supercharging Academic Writing with Generative AI," Nature Biomedical Engineering, vol. 9, no. 4, pp. 426-431, 2025/04/01 2025, https://doi.org/10.1038/s41551-024-01185-8
[13] T. K. Koo and M. Y. Li, "A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research," Journal of Chiropractic Medicine, vol. 15, no. 2, pp. 155-163, 2016, https://dx.doi.org/10.1016/j.jcm.2016.02.012
[14] I. K. L. Lawrence, "A Concordance Correlation Coefficient to Evaluate Reproducibility," Biometrics, vol. 45, no. 1, pp. 255-268, 1989, https://doi.org/10.2307/2532051
[15] L. Yan, S. Greiff, Z. Teuber, and D. Gašević, "Promises and Challenges of Generative Artificial Intelligence for Human Learning," Nature Human Behaviour, vol. 8, no. 10, pp. 1839-1850, 2024/10/01 2024, https://doi.org/10.1038/s41562-024-02004-5
[16] M. Al-Kfairy, D. Mustafa, N. Kshetri, M. Insiew, and O. Alfandi, "Ethical Challenges and Solutions of Generative AI: An Interdisciplinary Perspective," Informatics, vol. 11, no. 3, p. 58, 2024, https://dx.doi.org/10.3390/informatics11030058
[17] S. Herbold, A. Hautli-Janisz, U. Heuer, Z. Kikteva, and A. Trautsch, "A Large-Scale Comparison of Human-Written versus ChatGPT-Generated essays," Scientific Reports, vol. 13, no. 1, p. 18617, 2023/10/30 2023, https://doi.org/10.1038/s41598-023-45644-9
[18] I. K. L. Lawrence, "Corrections (A Concordance Correlation Coefficient to Evaluate Reproducibility)," Biometrics, vol. 56, no. 1, pp. 324-325, 2000, https://dx.doi.org/10.1111/j.0006-341x.2000.00324.x
[19] B. Efron and R. J. Tibshirani, An Introduction to the Bootstrap. New York: Chapman and Hall/CRC, 1994. https://dx.doi.org/10.1201/9780429246593
[20] J. M. Bland and D. G. Altman, "Agreement Between Methods of Measurement with Multiple Observations Per Individual," Journal of Biopharmaceutical Statistics, vol. 17, no. 4, pp. 571-582, 2007, https://dx.doi.org/10.1080/10543400701329422
[21] K. Krippendorff, Content Analysis: An Introduction to Its Methodology, 4th ed. Thousand Oaks, California: SAGE Publications, Inc., 2019. [Online]. Available: https://doi.org/10.4135/9781071878781. Accessed on: 2025/09/29.
[22] L. Li, Z. Ma, L. Fan, S. Lee, H. Yu, and L. Hemphill, "ChatGPT in Education: A Discourse Analysis of Worries and Concerns on Social Media," Education and Information Technologies, vol. 29, no. 9, pp. 10729-10762, 2024/06/01 2024, https://doi.org/10.1007/s10639-023-12256-9
[23] A. D. Samala et al., "Unveiling the Landscape of Generative Artificial Intelligence in Education: A Comprehensive Taxonomy of Applications, Challenges, and Future Prospects," Education and Information Technologies, vol. 30, no. 3, pp. 3239-3278, 2025/02/01 2025, https://doi.org/10.1007/s10639-024-12936-0
[24] H. Abu-Rayyash, "Revolutionizing Translator Training Through Human-AI Collaboration: Insights and Implications from Integrating GPT-4," Current Trends in Translation Teaching & Learning E, vol. 10, 2023, https://doi.org/10.51287/cttl20239
[25] M. S. P. Mohamed, "Exploring Ethical Dimensions of AI-Enhanced Language Education: A Literature Perspective," Technology in Language Teaching & Learning, vol. 6, no. 3, p. 1813, 2024, https://dx.doi.org/10.29140/tltl.v6n3.1813
[26] N. Zainuddin et al., "Responsible and Ethical Use of Artificial Intelligence in Language Education: A Systematic Review," Forum for Linguistic Studies, vol. 6, no. 5, pp. 316-325, 2024, https://dx.doi.org/10.30564/fls.v6i5.7092
[27] Z. B. Akhtar, "Generative artificial intelligence (GAI): From large language models (LLMs) to multimodal applications towards fine tuning of models, implications, investigations," Computing and Artificial Intelligence, p. 1498, 2024, https://dx.doi.org/10.59400/cai.v3i1.1498
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Rohib Adrianto Sangia
This work is licensed under a Creative Commons Attribution 4.0 International License.
