Üretken Yapay Zekâ Modellerinin Eğitimsel Sinirbilim Okuryazarlıklarının Belirlenmesi

Öz

Bu çalışmada üretken yapay zekâ (YZ) modellerinin beynin yapısı ve işlevleri hakkında doğru bilgi sağlama yetenekleri ile yanlış bilgi ve yanlış anlamalar (nöromitler) karşısındaki performansları incelenmiştir. Bu bağlamda, Eğitimsel Nörobilim anketindeki 41 soru, beş üretken YZ modeline (ChatGPT, Claude, Copilot, Deepseek ve Gemini) hem İngilizce hem de Türkçe olarak yöneltilmiştir. Modellerden her bir maddeye “Doğru”, “Yanlış” veya “Bilmiyorum” olmak üzere üç seçenekten birini kullanarak yanıt vermeleri istenmiştir. Anket, iki kez uygulanmıştır. Bulgular, üretken YZ modellerinin beynin yapısı ve işlevleri hakkında doğru bilgi sağlamada iyi performans sergilediklerini göstermiştir. Ancak, nöromitleri içeren sorular söz konusu olduğunda, performanslarının orta düzeye gerilediği bulunmuştur. Her iki alanda da anketin dili, yanıtların doğruluğunu etkileyen önemli bir faktör olarak belirlenmiştir. Modellerin yanıt örüntülerine daha yakından bakıldığında, yanlış bilgiye güvenme ve yaygın inançları doğru olarak kabul etme eğilimi ortaya çıkmıştır. Bu durum, YZ modellerinin nöromitlerle mücadeleye yardımcı olmakla birlikte, bunların yayılmasına da yol açma potansiyeli gösterebileceğini vurgulamaktadır. Sonuç olarak, çalışma, özellikle nöromit konusu başta olmak üzere, üretken YZ tarafından üretilen yanıtların eleştirel bir şekilde değerlendirilmesi gerektiğini vurgulamaktadır.

Anahtar Kelimeler

• Eğitimsel sinirbilim
• sinirbilim okuryazarlığı
• yapay zekâ asistanları
• sinirbilim mitleri
• eğitim.

Determining the Educational Neuroscience Literacy of Generative AI

Abstract

This study examines the ability of generative artificial intelligence (GenAI) models to provide accurate information about the structure and functions of the brain, as well as their performance in the face of misinformation and misconceptions (neuromyths). In this context, 41 items from an Educational Neuroscience questionnaire were presented to five GenAI models—ChatGPT, Claude, Copilot, Deepseek, and Gemini—in both English and Turkish. Each model was asked to respond to each item using one of three options: “True”, “False”, or “I don’t know”. The questionnaire was administered twice. The findings showed that the GenAI models performed well in providing correct information about the brain's structure and functions. However, when it came to questions involving neuromyths, their performance was found to be at a moderate level. In both areas, the language of the questionnaire was identified as a significant factor affecting the accuracy of responses. A closer look at the models' response patterns revealed a tendency to trust incorrect information and to accept popular beliefs as accurate. This suggests that GenAI models may help counter neuromyths, while also potentially contributing to their dissemination. Consequently, the study emphasizes the need to critically evaluate GenAI-generated responses, particularly those concerning neuromyths.

Keywords

• Educational neuroscience
• neuroscience literacy
• AI assistants
• neuromyths
• education.

Kaynakça

1. Adıgüzel, O. C., Kara, D. A., & Küçükkayhan, S. (2024). Doğru bildiğimiz yanlışlar. Eğitim sistemlerinde nöromit salgını. Enstitü Sosyal.
2. Amran, M. S., & Sommer, W. (2025). Seen through teachers’ eyes: Neuromyths and their application in Malaysian classrooms. Trends in Neuroscience and Education, 38, 1-7. https://doi.org/10.1016/j.tine.2025.100250
3. Avşar, B. (2025). Artificial intelligence and disinformation: A comprehensive analysis of global disinformation campaigns underpinned by the OpenAI report. TRT Academy, 10(23), 208-237. https://doi.org/10.37679/trta.1562358
4. Barman, D., Guo, Z., & Conlan, O. (2024). The dark side of language models: Exploring the potential of llms in multimedia disinformation generation and dissemination. Machine Learning with Applications, 16, 1-17. https://doi.org/10.1016/j.mlwa.2024.100545
5. Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021, March). On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 ACM conference on fairness, accountability, and transparency, New York, USA. https://doi.org/10.1145/3442188.3445922
6. Blanchette S. J., Riopel, M., & Masson, S. (2019). Neuromyths and their origin among teachers in Quebec. Mind, Brain, and Education, 13(2), 100–109. https://doi.org/10.1111/mbe.12193
7. Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P.,... & Amodei, D. (2020). Language models are few-shot learners. Advances in Neural Information Processing Systems, 33, 1877-1901. https://arxiv.org/abs/2005.14165v4
8. Can, A. (2019). SPSS ile bilimsel araştırma sürecinde nicel veri analizi. Pegem Publishing.

9. Chambers, R., & Clark, R. (2012). An introduction to model-based survey sampling with applications. Oxford University Press.
10. Cho, Y. C., & Yeh, T. K. (2024). Dissecting neuromyths in education: an analysis of educators, students, and parents in Taiwan. International Journal of Science Education, 1-21. https://doi.org/10.1080/09500693.2024.2415094
11. Clinton-Lisell, V., & Litzinger, C. (2024). Is it really a neuromyth? A meta-analysis of the learning styles matching hypothesis. Frontiers in Psychology, 15, 1-19. https://doi.org/10.3389/fpsyg.2024.1428732
12. Danry, V., Pataranutaporn, P., Groh, M., Epstein, Z., & Maes, P. (2024). Deceptive AI systems that give explanations are more convincing than honest AI systems and can amplify belief in misinformation, 1-31. https://doi.org/10.48550/arXiv.2408.00024
13. Dekker, S., Lee, N. C., Howard-Jones, P., & Jolles, J. (2012). Neuromyths in education: Prevalence and predictors of misconceptions among teachers. Frontiers in Psychology, 3(429), 1-8. https://doi.org/10.3389/fpsyg.2012.00429
14. Feiler, J. B., & Stabio, M. E. (2018). Three pillars of educational neuroscience from three decades of literature. Trends in Neuroscience and Education, 13, 17-25. https://doi.org/10.1016/j.tine.2018.11.001
15. Ferrero, M., Garaizar, P., & Vadillo, M. A. (2016). Neuromyths in education: Prevalence among Spanish teachers and an exploration of cross-cultural variation. Frontiers in Human Neuroscience, 10(496), 1-11. https://doi.org/10.3389/fnhum.2016.00496
16. Gillespie, N., Lockey, S., Curtis, C., Pool, J., & Akbari, A. (2023). Trust in artificial intelligence: A global study. https://doi.org/10.14264/00d3c94
17. Giuffre, M., Kresevic, S., You, K., Dupont, J., Huebner, J., Grimshaw, A. A., & Shung, D. L. (2024). Systematic review: The use of large language models as medical chatbots in digestive diseases. Alimentary Pharmacology & Therapeutics, 60(2), 144-166. https://doi.org/10.1111/apt.18058
18. Gülsün, Y., & Köseoğlu, P. (2020). Determining biology teachers' neuromyths and knowledge about brain functions. Education and Science, 45(204), 303-316. http://dx.doi.org/10.15390/EB.2020.8456
19. Hireche, A., & Belkacem, A. N. (2024, May). Integrating pepper robot and gpt for neuromyth educational conversation. In 2024 IEEE Global Engineering Education Conference, Kos Island, Greece. https://doi.org/10.1109/EDUCON60312.2024.10578778
20. Joshi, P., Santy, S., Budhiraja, A., Bali, K., & Choudhury, M. (2020, July). The state and fate of linguistic diversity and inclusion in the NLP world. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, Stroudsburg, USA. https://doi.org/10.48550/arXiv.2004.09095
21. Kaya, P. S., & Açar, Y. (2021). Evaluation of caffeine consumption in university students. Journal of Samsun Health Sciences, 6(3), 595-606. https://doi.org/10.47115/jshs.996974
22. Koçak, G. (2020). Reflections of brain researches to education: Toward education of future. Yeditepe University Faculty of Education Journal, 9(11), 1-16. https://dergipark.org.tr/tr/download/article-file/1086362
23. Lample, G., & Conneau, A. (2019). Cross-lingual language model pretraining. Computation and Language, 1, 1-10. https://doi.org/10.48550/arXiv.1901.07291
24. Lim, Z. W., Pushpanathan, K., Yew, S. M. E., Lai, Y., Sun, C. H., Lam, J. S. H.,... & Tham, Y. C. (2023). Benchmarking large language models’ performances for myopia care: A comparative analysis of ChatGPT-3.5, ChatGPT-4.0, and Google Bard. EBioMedicine, 95. https://doi.org/10.1016/j.ebiom.2023.104770
25. Mercan, G., Altun, A., & Köseoğlu, P. (2023). Determination of neuromyths regarding brain functions among prospective teachers: Hacettepe University sample. Ihlara Journal of Educational Research, 8(1), 1-16. http://ihead.aksaray.edu.tr/tr/pub/ihead/article/1128433
26. Mohammed, S., Budach, L., Feuerpfeil, M., Ihde, N., Nathansen, A., Noack, N.,... & Harmouch, H. (2025). The effects of data quality on machine learning performance on tabular data. Information Systems, 132, 1-18. https://doi.org/10.1016/j.is.2025.102549
27. Monteith, S., Glenn, T., Geddes, J. R., Whybrow, P. C., Achtyes, E., & Bauer, M. (2024). Artificial intelligence and increasing misinformation. The British Journal of Psychiatry, 224(2), 33-35. https://doi.org/10.1192/bjp.2023.136
28. OECD (2002). Understanding the brain: Towards a new learning science. https://www.oecd.org/en/publications/understanding-the-brain_9789264174986-en.html
29. Saygıner, Ş. (2024). Eğitimsel sinirbilim: NeuroSky MindWave ile beyin dalgalarının analizi. In F. Balaman & Ş. Saygıner (Eds.), Akademik araştırma süreçlerinde dijital teknolojilerin destekleyici rolü: Teoriden uygulamaya (Sect. 6, pp. 127–146). Nobel.
30. Schwartz, B. (2025). ChatGPT mostly source Wikipedia; Google AI overviews mostly source Reddit. seroundtable.com/chatgpt-google-aio-sources-39578.html?utm_source=chatgpt.com.
31. Sisman, G., Demirbulak, D., & Virlan, A. Y. (2025). The prevalence of neuromyths among K‐12 English Language Teachers. European Journal of Education, 60(2), 1-12. https://doi.org/10.1111/ejed.70070
32. Spennemann, D. H. (2023). What has ChatGPT read? The origins of archaeological citations used by a generative artificial intelligence application. arXiv, .https://doi.org/10.48550/arXiv.2308.03301
33. Stephens, L. D., Jacobs, J. W., Adkins, B. D., & Booth, G. S. (2023). Battle of the (Chat) Bots: comparing large language models to practice guidelines for transfusion-associated graft-versus-host disease prevention. Transfusion Medicine Reviews, 37(3), 1-5. https://doi.org/10.1016/j.tmrv.2023.150753.
34. Sullivan, K. A., Hughes, B., & Gilmore, L. (2021). Measuring educational neuromyths: Lessons for future research. Mind, Brain, and Education, 15(3), 232-238. https://doi.org/10.1111/mbe.12294
35. Torrijos-Muelas, M., González-Víllora, S., & Bodoque-Osma, A. R. (2021). The persistence of neuromyths in the educational settings: A systematic review. Frontiers in Psychology, 11. https://doi.org/10.3389/fpsyg.2020.591923
36. Tunga, Y., & Çağıltay, K. (2023). Myths or facts: Prevalence, and predictors of neuromyths among Turkish teachers. Education and Science, 48(216). http://dx.doi.org/10.15390/EB.2023.12089
37. Voulgari, A., Potvin, P., & Adıguzel, O. C. (2025). Neuromyths in Greek primary teachers and their sources. Review of Science, Mathematics and ICT Education. https://doi.org/10.26220/rev.5248
38. Yüzüak, B. N. & Yılmaz, F. N. (2025). Comparison of different artificial intelligence applications in question revision: An application on true-false items. The Journal of Buca Faculty of Education, 63, 500-520. https://doi.org/10.53444/deubefd.1532545