The effect of artificial intelligence-supported augmented reality application developed for the human and environment unit on the achievement and attitude of secondary school students

The use of Augmented Reality (AR) applications in education is steadily increasing, supporting the presentation of science content particularly topics that are impossible to observe directly or difficult to learn in an interactive and realistic classroom environment. Another technology utilized in learning environments is artificial intelligence (AI). Integrating the textbased interactive structure of AI, powered by large language models, into AR applications can elevate their level of interactivity to an innovative stage. The aim of this study was to develop an AI-supported AR instructional material (BioscienceAR) for the topics of Biodiversity, Human–Environment Interactions, and Destructive Natural Events in the Human and Environment unit of science education, and to investigate its effects on students' academic achievement, attitudes toward science, and attitudes toward AR applications. A mixed-methods approach was employed, with a dominant quantitative component supported by qualitative data. A sequential explanatory design was adopted, in which qualitative data were collected to explain the quantitative findings. The research process was conducted using the PIE model, which includes planning, implementation, and evaluation. In the planning and development of the instructional material, a five-phase dynamic design model commonly used by AR developers was used as the foundation. In the quantitative phase, data were collected through the Human and Environment Achievement Test, the Science Attitude Scale, and the AR Applications Attitude Scale. The experimental process was carried out over eight weeks with 56 fifth-grade students from a public school located in the Central Anatolia region of Türkiye. For the analysis of the quantitative data, it was first determined that the data met the assumption of normality, and parametric tests appropriate for normal distribution were applied. In the qualitative phase, semi-structured interview forms were administered to nine students and the course teacher. Researcher notes were vii also considered as part of the qualitative data. The data were analyzed using the triangulation method and were transformed into themes, categories, and codes. According to the quantitative findings of the study, there was no significant difference between the pre-test achievement mean scores of the experimental and control groups. However, when the post-test achievement scores were examined, it was found that the mean score of the experimental group was higher than that of the control group, although this difference was not statistically significant. Another finding revealed that, prior to the intervention, the experimental and control group students had equivalent scores on the science attitude scale; however, in the post-test, the experimental group's mean score was statistically significantly higher than that of the control group. Similarly, in terms of attitudes toward AR applications, the two groups had equivalent scores before the intervention, whereas post-test results indicated a statistically significant difference in favor of the experimental group. Based on these results, it can be said that the BioscienceAR material developed for the Human and Environment unit within the scope of the study can be used as a support for developing positive attitudes towards Science course and AR applications. Furthermore, it can be said that this research can serve as an exemplary study for integrating AR content into textbooks suitable for the purpose specified in the Turkey Century Education Model, which was implemented by the Ministry of National Education (MEB) in the 2024-2025 academic year, and that it can make significant contributions to the literature.