Shahaf Rocker Yoel, Ayelet Becher E39 Introduction and Literature Review Over the past decade, education systems have been shaped by two interrelated developments: the growing reliance on data-driven decision-making (DDDM) and the rapid advancement of artificial intelligence (AI) technologies. Both movements share a vision of improving instructional quality and organizational learning, yet they differ in how data are produced, interpreted, and acted upon. DDDM has long been promoted as a cornerstone of evidence-based reform, emphasizing teachers' capacity to collect, analyze, and apply data to improve teaching and learning (Mandinach & Schildkamp, 2021). AI, by contrast, introduces new modes of data generation and interpretation through algorithms capable of predicting, modeling, and even generating content (Holmes et al., 2022; Luckin, 2018). Together, these trends are reshaping the ways in which educators understand and exercise professional judgment (Sajja et al., 2025). Data-Driven Decision-Making (DDDM) Research on DDDM in education highlights both promise and persistent challenges. When teachers engage meaningfully with student data, they can identify learning needs, tailor interventions, and monitor progress (Mandinach & Gummer, 2016). However, studies consistently show that data use is constrained by limited time, data literacy, and systemic pressures emphasizing accountability over pedagogical improvement (Mandinach & Schildkamp, 2021). Teachers often find data practices burdensome or disconnected from classroom realities. The effectiveness of DDDM, therefore, depends not only on access to information but also on professional cultures that value inquiry, trust, and collaboration (Marsh & Farrell, 2014). School leaders play a crucial role in creating such conditions by providing structures for shared data interpretation and professional learning. Artificial Intelligence (AI) in Education Parallel to this evolution, AI has entered educational settings in diverse forms, from intelligent tutoring systems and learning analytics to generative AI tools such as ChatGPT. These technologies promise to reduce workload, reveal data patterns, and support personalization (Zawacki-Richter et al., 2019). Yet teachers' responses remain ambivalent. While many appreciate AI's potential to enhance efficiency and student engagement, they also express concerns about transparency, fairness, and professional autonomy (Williamson & Eynon, 2020). Because AI systems often operate as "black boxes," offering recommendations without clear explanations, teachers may struggle to trust them or to integrate algorithmic outputs into professional reasoning. Ethical issues such as algorithmic bias, privacy, and the risk of deskilling further complicate perceptions (Arantes, 2023; Lee & See, 2004). Ultimately, the educational value of AI hinges on teachers' trust that the algorithmic insights enhance, rather than override, their professional judgment and capacity for evidence-based decision-making (Holmes et al., 2022). The Intersection of AI and DDDM The intersection of AI and DDDM represents an emerging yet underexplored area of inquiry. Traditional DDDM required teachers to manually interpret datasets, whereas AI now automates aspects of analysis and prediction, shifting the balance between human and machine judgment. This convergence raises new questions about how teachers interpret and evaluate AI-generated educational data used for decision-making. Do teachers view AI as a partner that supports data interpretation and evidence-based instructional decisions, or as a system that constrains their professional autonomy? How do emotional and ethical factors affect teachers' trust in and use of AIgenerated data for decision-making? Despite the growing presence of AI in educational data systems, systematic synthesis of research on teachers' perceptions at this intersection remains limited (Romero & Ventura, 2020; Williamson & Eynon, 2020).
RkJQdWJsaXNoZXIy Mjk0MjAwOQ==