Exploring the Methodological Contexts and Constraints of Research in Artificial Intelligence in Education

In this paper, we present a Systematic Literature Review (SLR) on the state-of-the-art in Artificial Intelligence in Education (AIED) focusing on methodological contexts and constraints of the research landscape. To do so, we built on existing works and extended them to cover the latest research advancements in the field over the past five years. We aimed at covering all educational levels and retrieving important data regarding the planning and execution of research studies and the robustness of results. In total, we reviewed 181 papers and answered three research questions, relating to the educational context of AI use, the methodology and study design utilized in AIED research, and the type of AI algorithms and technologies used in education. Our findings suggest that research in AIED primarily focuses on formal, higher education and that there is a demand for robust and rigorous scientific evidence of the effectiveness and impact of AIED. Furthermore, the findings indicate that the most popular AI technologies currently studied are traditional AI algorithms, usually used for prediction, classification, or clustering. Based on our analysis, we discuss practical implications that can serve as inspiration and guidance for future research initiatives.