An analysis of the readability and main idea of texts generated by artificial intelligence applications

In this study, the readability levels and main idea characteristics of texts produced by artificial intelligence applications were examined. In the increasingly digitalized educational landscape, evaluating the educational value of AI-generated texts has become critical. The study aims to determine the extent to which such texts can be understood by fourth-grade primary school students and to assess their pedagogical appropriateness. The research was conducted within a mixed-methods design, employing both quantitative and qualitative data collection techniques. Quantitative data were analyzed using the Ateşman Readability Formula to determine the readability levels of AI-generated texts. Qualitative data were obtained through content analysis of the views collected from students and teachers. The data sources consisted of original stories generated by artificial intelligence and semi-structured interview forms. The findings indicated that AI-generated texts were generally at an appropriate level in terms of readability. However, in some texts, factors such as abstraction, lack of contextual clarity, and cultural inconsistency were observed to negatively affect the clarity of the main idea. Teacher opinions highlighted that these texts have potential for classroom use, yet emphasized the necessity of teacher-guided content adaptation for effective implementation. Data obtained from students showed that the main idea was more easily understood in texts where the problem situation and solution elements were presented clearly. The study concludes that AI-supported texts may serve as complementary tools in education; however, adapting the content to the cultural context and target grade level is essential. Future research is recommended to examine the effectiveness of AI-generated materials more comprehensively across different age groups, text types, and educational settings.