Using Artificial Intelligence to Choose a Research Topic: A Comparative Study of Tax Evasion in Turkey and Greece

Abstract

This paper examines the role of artificial intelligence (AI) in academic research by detailing how Grok, an AI assistant, was used to select a comparative study on attitudes toward tax evasion in Turkey and Greece. The authors queried Grok for Turkey's bordering countries and recommendations for comparison, leading to the selection of Greece due to shared geographic, cultural, and economic factors. However, Grok's cited data on tax morale from the World Values Survey (WVS) proved inaccurate—potentially an AI hallucination or outdated information—highlighting the need for verification. A literature review synthesizes studies on tax evasion drivers in both countries, including institutional, cultural, and demographic influences. Using WVS Wave 7 data (2017-2022) from 1,193 Greek and 2,395 Turkish respondents, the empirical analysis employs Welch's t-tests to compare views on whether tax evasion is justifiable (1=never, 10=always). Results show similar strong opposition overall (means: Greece 1.67, Turkey 1.69; p=0.7108). Demographic breakdowns reveal: no gender differences; age significant only for Greeks (older more opposed); education inversely related to opposition in Greece; income inversely related in Turkey. The study underscores AI's utility and limitations in research while contributing to comparative ethics of tax evasion.

Keywords

• Tax evasion
• artificial intelligence
• World Values Survey
• demographics
• tax compliance
• comparative study

Yapay Zeka Kullanarak Araştırma Konusu Seçimi: Türkiye ve Yunanistan'da Vergi Kaçakçılığına İlişkin Karşılaştırmalı Bir Çalışma

Abstract

Bu makale, yapay zeka (AI) asistanı Grok'un Türkiye ve Yunanistan'da vergi kaçakçılığına yönelik tutumları karşılaştırmalı bir çalışmada nasıl kullanıldığını detaylı olarak anlatarak, akademik araştırmalarda yapay zekanın rolünü incelemektedir. Grok'a Türkiye'nin komşu ülkeleri ve karşılaştırma için öneriler hakkında sorular sorulmuş, coğrafi, kültürel ve ekonomik faktörlerin ortak olması nedeniyle Yunanistan'ın seçilmesine karar verilmiştir. Ancak, Grok'un Dünya Değerler Araştırması'ndan (WVS) aldığı vergi ahlakı verilerinin yanlış olduğu tespit edilmiştir. Bu veriler, muhtemelen bir yapay zeka yanılsaması veya güncel olmayan bilgilerdi ve doğrulama ihtiyacını ortaya koymaktadır. Literatür taraması, her iki ülkedeki vergi kaçakçılığının nedenleri üzerine yapılan çalışmaları, kurumsal, kültürel ve demografik etkileri de dahil olmak üzere sentezlemektedir. 1.193 Yunan ve 2.395 Türk katılımcıdan elde edilen WVS Wave 7 verilerini (2017-2022) kullanan ampirik analiz, vergi kaçakçılığının haklı olup olmadığına dair görüşleri karşılaştırmak için Welch'in t-testlerini kullanmaktadır (1=asla, 10=her zaman). Sonuçlar, vergi kaçakçılığına karşı genel olarak benzer güçlü bir muhalefet olduğunu göstermektedir (ortalamalar: Yunanistan 1,67, Türkiye 1,69; p=0,7108). Demografik dağılımlar şunu ortaya koymaktadır: cinsiyet değişkeninde anlamlı bir farklılık bulunmamıştır. Yaş değişkeninin ise sadece Yunanlılar için önemli olduğu sonucuna varılmıştır (yaşlılar daha fazla muhalefet etmektedir); eğitim değişkeni Yunanistan'da vergi kaçakçılığına muhalefetle ters orantılıdır; gelir değişkeni ise vergi kaçakçılığına bakışta Türkiye'de ters orantılıdır. Çalışma, vergi kaçakçılığının karşılaştırmalı etiğine katkıda bulunurken, araştırmada yapay zekanın yararlarını ve sınırlarını vurgulamaktadır.

Keywords

• Vergi kaçakçılığı
• yapay zeka
• Dünya Değerler Araştırması
• demografi
• vergi uyumu
• karşılaştırmalı çalışma

References

1. Berk, H. (2024). Beware of Artificial Intelligence hallucinations or should we call confabulation? Acta Orthopaedica et Traumatologica Turrcica, 58(1), 1-3. DOI: 10.5152/j.aott.2024.130224
2. Beutel, G., Geerits, E. & Kielstein, J.T. (2023). Artificial hallucination: GPT on LSD? Critical Care, 27: 148. DOI: 10.1186/s13054-023-04425-6.
3. Colasacco, C.J. & Born, H.L. (2024). A Case of Artificial Intelligence Chatbot Hallucination. JAMA Otolaryngology-Head and Neck Surgery, 150(6), 457-459.
4. Kamel, H. (2024). Understanding the impact of AI Hallucinations on the university community. Special Issue for 3rd International Conference on Information Literacy, Issue 73, 111-135, December 2024. DOI: 10.70000/ci.2024.73.622
5. Park, Z, (2025). AI for KOS Discovery: Refining Search, Recommendation, and Hallucination Mitigation. Knowledge Organization, 52(4): 45889. DOI: 10.31083/KO45889
6. Wu, K., Wu, E., Ho, D.E. & Zou, J. (2024). Generating Medical Errors: GenAI and Erroneous Medical References. February 12. Stanford University, Human-Centered Artificial Intelligence. https://hai.stanford.edu/news/generating-medical-errors-genai-and-erroneous-medical-references
7. Xu, Z., Jain, S. & Kankanhalli, M. (2024). Hallucination is Inevitable: An Innate Limitation of Large Language Models. Preprint. arXiv:2401.11817v1 [cs.CL] 22 January. https://doi.org/10.48550/arXiv.2401.11817
8. Litina, A. & Palivos, T. (2015). Corruption and Tax Evasion: Reflections on Greek Tragedy. Working Paper 193. Bank of Greece. June. https://www.bankofgreece.gr/Publications/Paper2015193.pdf

9. McGee, R.W. & Benk, S. (2016). Attitudes toward Tax Evasion in Turkey: An Empirical Study. In M. Mustafa Erdogdu & Bryan Christiansen (Eds.), Handbook of Research on Public Finance in Europe and the MENA Region (pp. 299-318). Hershey, PA: IGI Global. http://www.researchgate.net303737564.
10. Richter, H. (2015). Greece: Turkey as an Example? Comparing the political cultures of Turkey and Greece. The Globalist, June 28. https://www.theglobalist.com/greece-turkey-as-an-example/
11. Anastasiou, A., Kalamara, E., & Kalligosfyris, C. (2020). Estimation of the size of tax evasion in Greece. Bulletin of Applied Economics, 7(2), 97-107. https://doi.org/10.47260/bae/727
12. Diakomihalis, M. (2020). Factors of tax evasion in Greece: Taxpayers’ perspective. Journal of Tax Reform, 6(2), 180–195. https://doi.org/10.15826/jtr.2020.6.2.081
13. Kounadeas, T., Eriotis, N., Boufounou, P., & Sofia, D. (2022). Analysis of the factors affecting tax evasion in Greece. International Journal of Economics and Business Administration, 10(1), 140–158.
14. Marinakos, K., Pistikou, G., & Papaioanou, A. (2022). Tax evasion in hospitality industry: Institutional deficit, mentality or necessity? Bulletin of Applied Economics, 10(1), 69–79. https://doi.org/10.47260/bae/1013
15. Devos, K., & Argan, M. (2019). Factors that influence tax evasion perceptions in Australia and Turkey: A structural equation model study. Journal of the Australasian Tax Teachers Association, 14(1), 1-28.
16. Benk, S., Budak, T., Püren, S., & Erdem, M. (2015). Perception of tax evasion as a crime in Turkey. Journal of Money Laundering Control, 18(1), 99-111. https://doi.org/10.1108/JMLC-04-2014-0012
17. Karlinsky, S., Hughlene, B. and Blanthorne, C. (2004), “Perceptions of tax evasion as a crime”, E-Journal of Tax Research, 2(2), 226-236.
18. McGee, R.W., Benk, S., Yüzbaşı, B. & Budak, T. (2020). Does Religiosity Affect Attitudes toward the Ethics of Tax Evasion? The Case of Turkey. Religions 11(9): 476-488. Doi:10.3390/rel11090476
19. Geyik, O., Şeren, G. Y., & Yakşi, Y. (2023). The effect of tax education on tax awareness: An experimental research in Diyarbakır Province [Conference paper]. In Proceedings of the 37th International Public Finance Conference/TR (pp. 351-354). https://www.researchgate.net/publication/376990556
20. Geyik, O., Şeren, G. Y., & Yakşi, Y. (2024a). The effect of tax education on Generation Alpha’s tax awareness: An experimental study. International Journal of Public Finance, 9(2), 223-246.
21. Geyik, O., Şeren, G. Y., & McGee, R. W. (2024b). An Empirical Analysis on the Effect of Taxpayers’ Educational Level and Marital Status Factor on Their Attitudes and Behaviors Towards Taxes. Bulletin of Economic Theory and Analysis, 9(2), 427-449.
22. Haerpfer, C., Inglehart, R., Moreno, A., Welzel, C., Kizilova, K., Diez-Medrano J., M. Lagos, P. Norris, E. Ponarin & B. Puranen et al. (eds.). (2020). World Values Survey: Round Seven – Country-Pooled Datafile. Madrid, Spain & Vienna, Austria: JD Systems Institute & WVSA Secretariat [Version: http://www.worldvaluessurvey.org/WVSDocumentationWV7.jsp].
23. Delacre, M., Lakens, D. & Leys, C. (2017). Why Psychologists Should by Default Use Welch's t-test Instead of Student's t-test. International Review of Social Psychology, 30(1), 92–101. https://doi.org/10.5334/irsp.82
24. Derrick, B., Toher, D. & White, P. (2016). Why Welch’s test is Type I error robust. The Quantitative Methods for Psychology, 12(1), 30-38. https://doi.org/10.20982/tqmp.12.1.p030
25. Welch, B. L. (1938). The significance or the difference between two means when the population variances are unequal. Biometrika, 29(3/4), 350–362. https://doi.org/10.2307/2332010
26. Welch, B. L. (1947). The Generalization of `Student’s’ Problem when Several Different Population Variances are Involved. Biometrika, 34(1/2), 28–35. https://doi.org/10.2307/2332510
27. McGee, R.W. (Ed.). (2012). The ethics of tax evasion: Perspectives in theory and practice: New Yor; Springer.
28. McGee, R.W. (2025). Welch’s t-test: The Robust Default for Comparing Two Groups in Psychological Research. Sociology & Social Policy, 2(2), 14-20. DOI: 10.26855/ssp.2025.12.003