Edutech Research Latest Issue

The primary aim of this study is to examine the methodological characteristics of postgraduate theses on artificial intelligence in mathematics education in Türkiye. Within the scope of the study, theses published between 2013 and 2025 in the Council of Higher Education National Thesis Center database were systematically reviewed and analyzed in detail based on predetermined criteria. The theses were evaluated in terms of variables such as publication year, university, sample group, research method, data collection tools, and the types of artificial intelligence technologies used. The findings indicate that there has been a growing interest in artificial intelligence within mathematics education in recent years, with a particular increase in applied studies conducted at the high school level. This research aims to identify current trends, highlight gaps in the literature, and provide guidance for future studies in the field.

This study aims to investigate the impact of students' Information and Communication Technology (ICT) self-efficacy (ICTEFFIC) and subject-related ICT use during lessons (ICTSUBJ) on mathematics achievement. Framed within Self-Determination Theory and ICT Engagement Theory, this quantitative secondary data analysis is based on a sample of 6,599 15-year-old students (3,286 males, 3,313 females) from the Program for International Student Assessment (PISA) 2022 Turkey cohort. Data were analyzed using multiple regression analysis based on 10 pooled plausible values in accordance with Rubin's rules, while controlling for socioeconomic status and gender, and employing robust standard errors to account for heteroscedasticity. The findings revealed that even when ESCS and gender were held constant, both ICT self-efficacy and subject-related ICT use significantly and positively predicted mathematics achievement. The regression model explained 15.6% of the variance in mathematics performance. Comparative results demonstrated that pedagogical, subject-related ICT use (B = 12.109, β = 0.146) had approximately twice the predictive power on mathematics achievement compared to digital self-efficacy (B = 6.284, β = 0.080). Furthermore, socioeconomic status was confirmed as the strongest predictor in the model. Regarding gender, although male students scored statistically higher on ICT variables, the effect sizes (Cohen's d) indicating practical significance remained at a negligible level. In conclusion, this study demonstrates that students’ mere psychological competence (self-efficacy) is not solely sufficient for digital technologies to positively contribute to academic achievement. For academic benefit, it is imperative that this digital belief translates into action through curriculum-integrated, teacher-guided, and pedagogically purposeful use in the classroom environment.

This study comprehensively examines the issue of authenticity triggered by the integration of artificial intelligence systems into content production, addressing its philosophical, sociological, epistemological and cultural dimensions. The central argument is as follows: the concept of authenticity has historically been grounded in human experience, subjective intention and bodily existence; artificial intelligence outputs cannot be subsumed within this definition. However, this does not imply that such outputs are worthless; on the contrary, it necessitates a rethinking of the ‘authenticity’ category itself. The study discusses this issue through the lenses of education, digital identity, epistemic trust and digital content production, highlighting how technologies such as deepfakes challenge perceptions of reality and drawing attention to the inadequacy of existing conceptual frameworks. Two alternative frameworks are evaluated: process-oriented authenticity and a human-AI collaboration model. Ultimately, the issue is positioned not merely as a matter of technology policy, but as a profound cultural and ethical problem concerning how humans understand themselves and others.

In the digital transformation of the 21st century, the field of education is undergoing profound structural change through the integration of big data analytics and artificial intelligence (AI) technologies. Learning management systems, adaptive platforms and AI-supported tools offer considerable opportunities for monitoring student performance, personalising learning processes and developing early intervention mechanisms. At the same time, these technologies carry multilayered ethical risks, including privacy violations, algorithmic bias, lack of transparency and the deepening of digital inequalities. This study aims to analyse the ethical problems arising from the use of big data and AI in digitalising educational environments from a normative perspective, and to develop an applicable ethical framework. Conducted within a qualitative research paradigm using document analysis, the study examines academic literature from the 2016–2025 period alongside UNESCO, OECD and EU documents; following thematic coding, a normative evaluation was carried out through deontological and teleological/utilitarian ethical perspectives. The findings are concentrated around six themes: (i) data privacy, confidentiality and surveillance; (ii) algorithmic bias and discrimination; (iii) gaps in transparency, explainability and accountability; (iv) digital inequality and the erosion of equal opportunity; (v) automation bias and the decline of human oversight; (vi) the inadequacy of international frameworks in addressing education-specific dynamics. Whilst deontological ethics positions the student's privacy and human dignity as non-negotiable boundaries, utilitarian ethics requires that social benefit be demonstrated within those boundaries. The study recommends the development of data minimisation, algorithmic explainability, independent oversight and binding governance standards specific to the educational context.

The shift toward hybrid learning in university level EFL classrooms has introduced unique challenges, particularly regarding the engagement gap between physical and virtual environments. While face to face sessions tend to facilitate natural interaction, the online component often suffers from a participation challenge that limits students' opportunities for authentic language practice and critical thinking. This conceptual paper aims to identify factors contributing to this decline engagement, ranging from technical barriers and diminished motivation to the sense of psychological isolation students experience when separated from their peer. By synthesizing the TPACK framework and Social Presence Theory, this study proposes a more cohesive hybrid pedagogical approach. Th proposed model transcends basic digital literacy by advocating for the strategic use of interactive online methods, gamification elements to boost intrinsic drive, and Project Based Learning (PjBL) to foster collaborative inquiry. The discussions emphasize that bridging this engagement gap requires educators to be not only technically proficient but also pedagogically wise in creating inclusive digital spaces. Ultimately, this conceptual framework provides practical guidance for EFL practitioners to transform passive online presence into active and meaningful learning experiences.