As Generative Artificial Intelligence (GAI) applications become more widespread, those working in education are increasingly examining how to employ GAI in educational contexts. The current focus is on preparing teachers for rapid AI developments by identifying models that maximize learning benefits while maintaining learner agency. This study examined how two AI integration models—substitution and complementarity—were associated with preservice teachers’ active participation and cognitive extension. A mixed-methods approach was employed, combining quantitative comparisons of post-intervention scores collected using the researchers' developed Active Participation Scale (APS) and Cognitive Extension Scale (CES) with qualitative analysis of post-training reflections. Independent- samples t- tests showed that the complementarity group scored significantly higher than the substitution group across all dimensions (p <. .001), with very large effect sizes indicating substantial practical significance. Qualitative findings revealed three overarching themes distinguishing substitution and complementarity environments: from cognitive replacement to cognitive partnership, highlighting differences in learner agency; from cognitive offloading to cognitive extension, reflecting differences in reflection, evaluation, and metacognitive engagement; and from efficiency- seeking to responsible integration, illustrating contrasting future orientations toward GAI use. Complementarity-oriented participants perceived GAI as a cognitive partner that supported ownership, knowledge refinement, and professional development, whereas substitution-oriented participants demonstrated greater reliance on AI-generated outputs, reduced personal involvement, and efficiency-driven use of GAI. Guided by constructivism, the SAMR model, and extended cognition, these findings suggest that GAI functions as a scaffold for higher- order thinking when used as a complementary cognitive partner rather than a replacement for human thinking.
Storyboarding is widely used in instructional design, eLearning development, curriculum planning, and training; however, it is typically described as a tool, strategy, or framework rather than a formalized methodology. This conceptual article addresses that gap, establishing storyboard methodology as a disciplined, theory-informed, repeatable process for externalizing and testing instructional design coherence. Drawing on research in narrative cognition, cognitive load theory, multimedia learning, and systematic design, this paper articulates theoretical foundations of storyboard methodology, distinguishing it from flexible visual frameworks. A dual-component model, with story as narrative engine and board as visual architecture, is proposed, with four defined phases, fidelity indicators, and a practitioner-facing comparison framework. Illustrative applications in eLearning development, curriculum design, and doctoral research are described. The article identifies limitations and directions for empirical validation and offers implications for instructional designers and educational technologists.
Generative AI can give learners polished artifacts before they understand the underlying ideas, creating a risk of apparent competence without defensible knowledge. This conceptual article introduces the Reverse Learning Framework, an artifact-first model in which learners move from an AI-generated output toward understanding through skepticism, verification, iterative prompting, contextual integration, human reconstruction, and explainable ownership. Its central mechanism is the fluency–validity gap: a potential discrepancy between an artifact's apparent adequacy and its warranted validity, contextual fit, and the learner's capacity to justify it. Five propositions specify how motivation activates inquiry, how verification and prompting interact, why reconstruction can trigger regression, how ownership functions as an exit criterion, and how expertise and task stakes moderate how explicitly the cycle is externalized. Explainable ownership comprises accountable authorship, epistemic ownership, and identity-level ownership, with epistemic ownership as the required educational core. The framework translates these relationships into scaffolds and assessment evidence while identifying disciplinary, equity, and implementation boundaries.
Artificial intelligence (AI) tools are rapidly expanding in K–12 education, yet systematic analyses linking their structural design features to pedagogical frameworks remain limited. This study examines the pedagogical potential of 26 widely used AI educational tools by integrating affordance theory with the Substitution, Augmentation, Modification, Redefinition (SAMR) model. Rather than assessing enacted classroom implementation, the analysis focuses on structural affordances and their alignment with degrees of task change. Using directed content analysis and crisp-set Qualitative Comparative Analysis (csQCA), core affordance indicators were coded and examined configurationally. Results indicate that 85
Online education in Japanese universities expanded rapidly during the COVID-19 pandemic, exposing both the genuine value of flexible learning formats and the structural conditions required to make them effective. While students consistently appreciated the temporal and spatial autonomy of online study, research has identified persistent challenges including isolation, physical and psychological fatigue, assignment overload, declining motivation, and inequities in digital access and ICT literacy. This paper presents a narrative synthesis of recent empirical studies on online learning in Japanese higher education, arguing that student success is associated with the coordinated integration of three interdependent dimensions. Drawing on national surveys, institutional reports, and case-based research, the paper proposes the Flexibility–Structure–Support (FSS) Model, in which flexibility without adequate pedagogical structure may increase the risk of disengagement, while structure without sufficient institutional support may contribute to uneven and inequitable learning experiences. Effective instructional design—integrating active learning, motivating assignments, meaningful feedback, and blended formats—must be accompanied by investment in infrastructure, systematic faculty development, and explicit support for student self-regulated learning. Implications for course design, program coordination, and institutional policy are discussed, along with limitations and directions for future research.
