Children on ChatGPT Readability in an Educational Context: Myth or Opportunity?
Open accessAbstract
In this work, we present the results of a preliminary exploration aiming to understand whether the use of ChatGPT in an educational context can be an asset to meet the specific needs of the students. In particular, we focus on the possibility of adapting the responses to online inquiries related to the primary school curriculum to meet the expectations of readers with different literacy levels. The analysis of feedback elicited from children (9- to 10-year-olds) in three 4th grade classrooms indicates that ChatGPT can adapt its responses to the 4th grade level. However, it still needs improvement to reach the right level of readability. Outcomes from this work can inspire future research directions involving technologies like ChatGPT to adapt learning paths to suit a broad range of students with varied cognitive skills. The potential of such tools to support teachers in their effort to adapt to individual learning needs is still to be fully exploited.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
In a third configuration, AI functions as a materials planner, mass-producing concept maps, leveled texts, or gamified problem sets and predefining reading order and pacing, with classroom control largely ceded to AI after teacher screening.
“In a third configuration, AI functions as a materials planner. Systems mass-produce concept maps, leveled texts, or gamified problem sets and predefine reading order and pacing; after teachers screen and correct these resources, classroom control is largely ceded to AI.”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [76]
Split into 2 atoms — 0 confirmed, 1 partially supported, 1 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredGenAI was employed to deliver personalized and leveled reading materials, helping to reduce obstacles such as comprehension difficulties, anxiety, or lack of confidence resulting from cognitive differences among students.
“GenAI was employed to deliver personalized and leveled reading materials, helping to reduce obstacles such as comprehension difficulties, anxiety, or lack of confidence resulting from cognitive differences among students [75-80].”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [76]
The Introduction (page unavailable) describes using ChatGPT to adapt responses to students’ literacy levels, and the Conclusions say it supported teachers by adapting results to students’ language and literacy needs. But the study does not show that this reduced anxiety or lack of confidence, and its preliminary findings say children did not always fully comprehend the texts. Extending the observed results to obstacles arising from cognitive differences goes beyond what the study established. (LLM, conf=0.88)
Experimental — accuracy not yet measuredIn school settings, students repeatedly interact with AI assistants across English writing, reading comprehension, mathematics problem-solving, physics, quantum, and cognitive science, and computer programming.
“In school settings, students repeatedly interact with AI assistants (such as ChatGPT, custom bots, and social robots) across English writing [85], reading comprehension [76,79,80,107,109], mathematics problem-solving [99,105], physics, quantum, and cognitive science [78,91,104,127], and computer programming [97,98,140,141].”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [76]
The Data Collection Protocol describes a classroom exercise in which fourth-grade students read ChatGPT-generated answers and assessed whether they understood them, supporting the reading-comprehension part of the cited sentence. But the researchers generated the answers and handed out paper task sheets; the students did not repeatedly interact with an AI assistant. The study focused on history questions, not the other subjects listed in the sentence. (LLM, conf=0.86)
Experimental — accuracy not yet measuredReadability issues, insufficient explanatory depth, and cognitive overload can further impede knowledge construction and amplify over-reliance risks.
“Readability issues [76], insufficient explanatory depth [91,92], and cognitive overload [73] can further impede knowledge construction and amplify over-reliance risks [94,119].”
From A Systematic Review of Generative AI in K–12: Mapping Goals, Activities, Roles, and Outcomes via the 3P Model (2025) · cites this paper as [76]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326 · cited 1× · see citing works
“Instead, we take another viewpoint and focus on the opportunities that arise from the existence of a technology that, to some degree, could help personalize students’ learning paths at school [2, 10, 12].”
BackgroundWhat Is the Impact of ChatGPT on Education? A Rapid Review of the Literature
Lo, C. K. (2023) · Education Sciences · doi:10.3390/educsci13040410 · Cited by 2,024 · see citing works
ChatGPT User Experience: Implications for Education
Zhai, X. (2023) · doi:10.2139/ssrn.4312418 · Cited by 935 · cited 1× · see citing works
“Lastly, it emerges how teachers have to be properly trained to be proficient in the use of these types of tools and be aware of the opportunities these bring for enhancing teaching at large by accounting for children’s individual needs [31].”
BackgroundA Prompt Pattern Catalog to Enhance Prompt Engineering with ChatGPT
White, J., Fu, Q., Hays, S., Sandborn, M., Olea, C., Gilbert, H., Elnashar, A., Spencer-Smith, J., & Schmidt, D. C. (2023) · arXiv · doi:10.48550/arxiv.2302.11382 · Cited by 838 · see citing works
A Conversation on Artificial Intelligence, Chatbots, and Plagiarism in Higher Education
King, M. R., & chatGPT (2023) · Cellular and Molecular Bioengineering · doi:10.1007/s12195-022-00754-8 · Cited by 602 · see citing works
“What Can ChatGPT Do?” Analyzing Early Reactions to the Innovative AI Chatbot on Twitter
Taecharungroj, V. (2023) · Big Data and Cognitive Computing · doi:10.3390/bdcc7010035 · Cited by 489 · see citing works
AlAfnan, M. A., Dishari, S., Jovic, M., & Lomidze, K. (2023) · Journal of Artificial Intelligence and Technology · doi:10.37965/jait.2023.0184 · Cited by 402 · see citing works
Hong, W. C. H. (2023) · Journal of Educational Technology and Innovation · doi:10.61414/jeti.v5i1.103 · Cited by 328 · see citing works
Iversen, O. S., Smith, R. C., & Dindler, C. (2017) · Proceedings of the 2017 Conference on Interaction Design and Children · doi:10.1145/3078072.3079725 · Cited by 306 · see citing works
Crossley, S. A., Skalicky, S., Dascălu, M., McNamara, D. S., & Kyle, K. D. (2017) · Discourse Processes · doi:10.1080/0163853x.2017.1296264 · Cited by 182 · see citing works
Online searching and learning: YUM and other search tools for children and teachers
Azpiazu, I. M., Dragovic, N., Pera, M. S., & Fails, J. A. (2017) · Information Retrieval · doi:10.1007/s10791-017-9310-1 · Cited by 50 · see citing works
Self-citationEffect of age and gender on children’s reading performance: The possible neural underpinnings
Vlachos, F. M., & Papadimitriou, A. (2015) · Cogent Psychology · doi:10.1080/23311908.2015.1045224 · Cited by 28 · see citing works
Query Formulation Assistance for Kids
Fails, J. A., Pera, M. S., Anuyah, O., et al. (2019) · doi:10.1145/3311927.3323131 · Cited by 24 · see citing works
Self-citationEnhancing classroom instruction with online news
Ekstrand, M. D., Wright, K. L., & Pera, M. S. (2020) · Aslib Journal of Information Management · doi:10.1108/ajim-11-2019-0309 · Cited by 16 · see citing works
Self-citationChildren's Perspective on How Emojis Help Them to Recognise Relevant Results
Aliannejadi, M., Landoni, M., Huibers, T. W. C., et al. (2021) · Conference on Human Information Interaction and Retrieval · doi:10.1145/3406522.3446058 · Cited by 13 · see citing works
Self-citation
Flagged first, then by citations
Research works printed but not linked: 9
- Geneva Gay. 2013. Teaching to and through cultural diversity. Curriculum inquiry 43, 1 (2013), 48-70.
- Ion Madrazo Azpiazu, Nevena Dragovic, and Maria Soledad Pera. 2016. Finding, understanding and learning: Making information discovery tasks useful for children and teachers. In CEUR Workshop Proceedings -Search as Learning (SAL) Workshop, co-located with ACM SIGIR 2016, Vol. 1647. CEUR-WS. https://ceur- ws.org/Vol-1647/SAL2016_paper_9.pdf
- Emiliana Murgia, Zahra Abbasiantaeb, Mohammad Aliannejadi, Theo Huibers, Monica Landoni, and Maria Soledad Pera. 2023. ChatGPT in the Classroom: A Preliminary Exploration on the Feasibility of Adapting ChatGPT to Support Chil- dren's Information Discovery. In Adjunct Proceedings of the 31st ACM Conference on User Modeling, Adaptation and Personalization (UMAP '23 Adjunct). 8 pages.
- Emiliana Murgia, Maria Soledad Pera, Monica Landoni, and Theo Huibers. 2023. Learning Technologies & AI: Who are we designing for. In CEUR Workshop Proceedings -Workshop 6: SOCIALIZE: Social and Cultural Integration with Person- alized Interfaces, co-located with ACM IUI 2023., Vol. 3359. CEUR-WS, 250-253.
- Jodi Pilgrim. 2019. Are we preparing students for the web in the wild? An analysis of features of websites for children. The Journal of Literacy and Technology 20, 2 (2019), 97-124.
- Cathy Ringstaff and Loretta Kelley. 2002. The learning return on our educational technology investment: A review of findings from research. (2002).
- Lauren M Singer and Patricia A Alexander. 2017. Reading across mediums: Effects of reading digital and print texts on comprehension and calibration. The journal of experimental education 85, 1 (2017), 155-172.
- Kari Stunell. 2021. Supporting student-teachers in the multicultural classroom. European journal of teacher education 44, 2 (2021), 217-233.
- Kevin R Sumayang, Kaycee Celendron, Neil P Declaro, and Deodato L Flandez Jr. 2022. MAINSTREAMING LEARNERS WITH SPECIAL NEEDS IN A REGULAR CLASSROOM: A SCOPING REVIEW. Universal Journal of Educational Research 1, 3 (2022), 1-1.
Cited by
Sorted by citations
Hadi Mogavi, R., Deng, C., Juho Kim, J., Zhou, P., D. Kwon, Y., Hosny Saleh Metwally, A., Tlili, A., Bassanelli, S., Bucchiarone, A., Gujar, S., et al. (2024) · Computers in Human Behavior: Artificial Humans · doi:10.1016/j.chbah.2023.100027 · Cited by 369 · see citing works
A systematic review of <scp>ChatGPT</scp> use in K‐12 education
Zhang, P., & Tur, G. (2024) · European Journal of Education · doi:10.1111/ejed.12599 · Cited by 280 · see citing works
Saihi, A., Ben-Daya, M., Hariga, M., & As'ad, R. (2024) · Computers and Education: Artificial Intelligence · doi:10.1016/j.caeai.2024.100274 · Cited by 75 · see citing works
Lin, X., & Tan, H. (2025) · Systems · doi:10.3390/systems13100840 · Cited by 22 · see citing works
Di Paolo, L. D., White, B., Guénin-Carlut, A., Constant, A., & Clark, A. (2024) · Philosophical Transactions of the Royal Society B: Biological Sciences · doi:10.1098/rstb.2023.0148 · Cited by 14 · see citing works
“With this in mind, LLMs could be powerful educative ‘tools’ ( sensu [187]) in the hands of teachers, augmenting the growing skilled capabilities of children (e.g. [232–234]).”
Background
- Authors
- Murgia, E., Pera, M. S., Landoni, M., & Huibers, T.
- Affiliations
- University of Genoa (IT)
- Delft University of Technology (NL)
- Università della Svizzera italiana (CH)
- University of Twente (NL)
- Venue
- Adjunct Proceedings of the 31st ACM Conference on User Modeling, Adaptation and Personalization
- Publisher
- ACM
- Published
- 2023-06-16
- Pages
- 311-316
- OpenAlex
- W4380881519
- Cited by
- 54
- References
- 15
- Citations by year
- 2026: 10 · 2025: 18 · 2024: 20 · 2023: 6
- FWCI
- 13.89
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- ACM Conference on User Modeling, Adaptation and Personalization (UMAP) (publishedVersion)
- University of Twente Research Information (submittedVersion, other-oa)
- Research Repository (Delft University of Technology) (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Information Systems › Artificial Intelligence in Education
- Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Text Readability and Simplification
- Keywords
- ChatGPT, readability, children, elementary school, literacy levels, response adaptation, online inquiry, personalized learning, individual learning needs, teacher support, generative AI in education
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.
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
