Although artificial intelligence (AI) tools are now deeply embedded in Turkish university students’ academic lives, students’ cognitive representations of AI remain instrumentally oriented and structurally fragmented. This study presents a needs assessment for designing AI literacy micro-credential programmes grounded in learners’ actual knowledge structures. Using the Word Association Test (WAT) as a psycholinguistic cognitive diagnostic instrument, data were collected from N = 436 undergraduate students enrolled at four public universities in Turkey (Kırklareli, İnönü, Niğde Ömer Halisdemir, and Hatay Mustafa Kemal), representing social sciences and humanities (80.3
Online Education in Higher Education is rapidly evolving through the integration of Large Language Model (LLM)-powered intelligent systems, which enable personalized tutoring, dynamic content generation, and automated assessment. However, the widespread adoption of LLMs in education is hampered due to their inherent limitations, including susceptibility to hallucinations, insufficient domain-specific knowledge validation, and output inconsistency. These deficiencies can lead to misleading or erroneous content, potentially causing significant negative learning outcomes. A core challenge lies in ensuring that such errors are immutably logged and traceable, thereby establishing a mechanism for accountability among the entities deploying these LLM services. To address these challenges, this paper proposes a novel framework that Integrates LLM with consortium blockchain for personalized and verifiable online education. Our design features a synergistic architecture in which LLM based services provide the intelligent educational interface, while a permissioned consortium blockchain serves as a secure and tamper proof ledger. This blockchain infrastructure records critical educational transactions ranging from learning process data and academic credentials to the outputs generated by the LLMs. This integration not only secures academic credentials but also establishes a fully auditable trail, making it possible to trace responsibility for educational deficiencies caused by AI errors. Collectively, this work demonstrates a robust and accountable framework for leveraging LLMs in education, effectively mitigating the risks of AI inaccuracies through the verifiable and immutable nature of consortium blockchain.
Generative artificial intelligence (GenAI) is increasingly used for feedback in higher education, yet evidence remains limited on how alternative human–AI feedback designs shape learning processes and durable outcomes. This study addresses that gap through a multisite, cluster-randomized, longitudinal field experiment comparing four feedback designs in introductory university science courses: peer feedback only, direct GenAI-supported feedback, reflective GenAI-supported feedback, and a hybrid design combining self-evaluation, peer feedback, and GenAI critique. The analytic sample comprised 1,176 first-year undergraduate students from 48 course sections across four universities and three science domains. Primary and secondary outcomes were argument-quality gain on four shared rubric dimensions—claim quality, evidence relevance and sufficiency, coherence of reasoning, and treatment of limitations or alternative explanations—conceptual learning, and delayed AI-free transfer; feedback uptake and self-regulated learning during revision were modeled as process mediators. Direct GenAI-supported feedback improved immediate argument-quality gain relative to peer feedback, whereas reflective and hybrid designs produced stronger feedback uptake and self-regulated learning. The hybrid condition yielded the highest adjusted mean for immediate argument-quality gain and showed the clearest advantage on conceptual learning; the reflective condition showed a positive but non-significant adjusted contrast on conceptual learning relative to direct GenAI-supported feedback. Both reflective and hybrid conditions outperformed direct GenAI-supported feedback on delayed AI-free transfer. Multilevel mediation analyses indicated that feedback uptake and self-regulated learning partially explained these advantages. By comparing four feedback designs, modeling revision processes, and assessing delayed AI-free transfer in a multisite field experiment, the findings suggest that the educational value of GenAI in higher education may depend less on AI access per se than on whether feedback environments preserve student agency, evaluative judgment, and ownership during revision.
This study, conducted using an explanatory sequential mixed-methods design, aims to examine the views of faculty members and students on ethical issues related to the use of artificial intelligence (AI) in higher education. In the quantitative phase, data were collected via a survey from 971 students and 135 faculty members, followed by semi-structured interviews with 23 students and 14 faculty members to obtain qualitative data. The findings show that both groups generally expressed supportive views regarding the ethical use of AI, but these views took different forms at the individual, technological, institutional, and societal levels. Faculty members emphasized the importance of ethical principles but pointed out the lack of institutional guidelines and support. Students, on the other hand, stated that AI tools were beneficial in their learning processes, but expressed uncertainty about the sharing of ethical responsibilities and the assessment of the ethical appropriateness of the processes. Qualitative findings showed that perspectives were shaped in six themes, revealing a multidimensional ethical structure in the use of AI in education. Participants emphasized that while AI facilitates learning, its excessive use can have negative effects on cognitive skills. The results indicate that, in order for AI to be used ethically in higher education, professional development support should be provided to faculty members, courses covering ethical dimensions should be added to curricula, clear ethical usage guidelines should be established at universities, and these guidelines should be updated through interdisciplinary collaboration.
Understanding how students cognitively engage with generative artificial intelligence (GenAI) has become a pressing concern in educational technology research. This study investigated how motivational orientations and GenAI role perceptions are associated with the quality and structure of metacognitive engagement in authentic student-GenAI dialogues. Using purposive sampling, 24 Chinese undergraduates drawn from 20 academic disciplines participated in individual retrospective interviews. They narrated the context and reasoning behind 120 authentic GenAI interaction logs comprising 688 messages in total, of which 344 were student-initiated. Our analyses reveal that intrinsically motivated students predominantly positioned GenAI as an instructor or collaborator, whereas extrinsically motivated students disproportionately treated it as a replacement tool. Intrinsic motivation was associated with significantly higher engagement in higher-order metacognitive dimensions (evaluation and elaboration), forming reflective-iterative co-occurrence networks, while extrinsic motivation produced task-completion-oriented patterns concentrated in lower-order processing. Our analyses further suggest that students who saw GenAI as a collaborator had a complete metacognitive chain (i.e., high engagement) and those who saw GenAI as a replacement tool exhibited no higher-order metacognitive connections. Overall, these findings suggest that the metacognitive consequences of GenAI use are neither uniform nor inherent to the technology. They are systematically conditioned by motivational orientations and perceived GenAI roles during interactions.
The widespread adoption of generative Artificial Intelligence (GenAI) among researchers has intensified the need for guidelines that safeguard academic integrity and ethical scholarly practice. However, existing GenAI policies tend to focus on teaching, learning and assessment rather than on research, where ethical concerns are considerably more complex. Given this complexity, such guidelines should move beyond rule enforcement to scaffold the AI literacy researchers need when deciding how, when, and whether to use GenAI across the full research process. To address this gap, this study conducted seven focus groups with 28 postgraduate research (PGR) students, who are among the most active adopters of GenAI tools in academic work. It examined how they used GenAI in research, how they interpreted ethically acceptable practice, what concerns they hold, and what institutional support they expected. Drawing on a four-dimensional AI literacy framework, the study analysed PGR students’ practices and perceptions through thematic analysis to examine how each dimension of AI literacy was reflected within the research setting. The findings showed that students’ awareness of GenAI’s capabilities, limitations, and risks corresponded to the “Know Understand AI” dimension, while their diverse, discipline-specific applications of GenAI across the research workflow reflected the “Use Apply AI” dimension. Their nuanced boundaries between ethical and unethical use demonstrated the “Evaluate Create AI” dimension, and concerns about output accuracy, originality, data privacy, and skill degradation highlighted the “AI Ethics” dimension. Building on PGR students’ research practices and support expectations, the study proposed researcher-oriented GenAI guidelines that foreground each AI literacy dimension across diverse research tasks, with the aim of informing more applicable policy and sustaining researchers’ AI literacy development.
Abstract Despite the rapid proliferation of Artificial Intelligence (AI)certificate programs in higher education, systematic frameworks for evaluating their pedagogical transformation potential remain absent—in contrast to the rich literature on AI tool adoption, which seldom addresses how credentials reshape teaching at the program level—and existing assessments rely on unidimensional methods that overlook both expert judgment and cultural variation. Grounded in the Technological Pedagogical Content Knowledge (TPACK) framework, we employ a hybrid Analytic Hierarchy Process (AHP) and Fuzzy AHP (FAHP) methodology, supplemented by Monte Carlo (MC) simulation (N = 10,000) for robustness verification, drawing on structured pairwise ratings and fuzzy assessments from 18 domain experts across Chinese universities. We identify and prioritize five dimensions—curriculum design, instructional implementation, faculty expertise, technological support, and cross-cultural adaptability—and quantify their relative weights and perceived performance. We find a clear hierarchy in expert priorities: student AI competency achievement (weight = 0.1438, rank stability = 99.8%) and curriculum alignment with AI frontiers (weight = 0.1056, rank stability = 98.5%) emerge as the paramount strategic levers, whereas cross-cultural adaptability receives the lowest weight (0.0735), signaling a technology-first bias in early-stage credential development. Strikingly, faculty professional competence exhibits the largest gap between its perceived importance (weight = 0.1976) and current satisfaction (score = 3.2171)—a finding that challenges the widespread assumption that technological infrastructure is the primary barrier to AI education. By revealing a developmental asymmetry in which content knowledge (CK) outweighs integrated pedagogical capacity, these findings extend TPACK theory into the credential evaluation domain and offer a robust, expert-informed strategic roadmap with explicit guidance for international adaptation.
Abstract With the rapid advancement of large language models, the demand for intelligent and fine-grained automated essay scoring in educational assessment has increased significantly. However, existing methods still face challenges in maintaining scoring alignment and output consistency, making it difficult to consistently approximate human scoring standards. To address these issues, this paper proposes a unified framework named RACES (Reward-Aligned Consistent Essay Scoring), which integrates LoRA-based parameter-efficient fine-tuning, reward modeling, and proximal policy optimization reinforcement learning. The framework establishes an offline inference–feedback–optimization pipeline, enabling optimization toward proxy preference signals simulated via LLM-generated feedback while constraining policy drift through KL regularization. Experimental results on the ASAP 2.0 dataset show that RACES improves QWK and auxiliary SimCSE metrics compared with the evaluated pretrained and fine-tuned model configurations, achieving rapid convergence with limited training iterations. The framework improves scoring accuracy under the evaluated settings, while consistency is examined through KL-regularized optimization behavior and auxiliary proxy-feedback analysis rather than direct deployment-level robustness tests. These findings suggest the practical potential of RACES for supporting more controlled preliminary essay scoring in educational assessment, particularly as an auxiliary tool for reducing grading workload and improving the reliability of large-scale writing evaluation.
Abstract Despite widespread adoption of large language models (LLMs), most students cannot effectively prompt them. The core challenge is teaching students how to ask: transforming prompting from trial-and-error guessing into a systematic, transferable skill. Existing solutions, such as static templates, rule-based hints, and automated rewriting, either ignore individual learning needs or optimize outputs without building competence, leaving students dependent and unable to generalize. ARPG+ is a real-time coaching system grounded in cognitive load theory and zone of proximal development that senses when learners struggle, delivers calibrated just-in-time interventions, and fades support as skills develop. The system tracks learner capability with uncertainty quantification, estimates cognitive overload from behavioral signals, diagnoses prompt quality across six dimensions, and adapts scaffolding intensity through a dynamic schedule with periodic skill probes. A lightweight-deep dual architecture ensures fast responsiveness for routine interactions while reserving richer analysis for critical moments. Evaluation with simulated learners shows ARPG+ produces improvements: prompt quality increases 143% beyond unguided practice, learners achieve independence in 91% of final interactions versus 59% under fixed support, and the approach generalizes to other domains without retraining. Our work establishes that principled real-time coaching can improve prompt quality, accelerate learning, prevent cognitive overload, and foster durable autonomy. All reported evaluations are conducted on LLM-based simulated learners; the present work does not yet constitute empirical validation with real students, and classroom validation in authentic educational settings is identified as a necessary subsequent step.
In the contemporary wave of digitalization sweeping across higher education, this research article explores the critical role of academic leadership in steering and navigating this transformative tide to form a digital university. These institutions are identified by their comprehensive integration of digital technologies in educational and administrative frameworks. Moving beyond a traditional literature review, the study is grounded in qualitative findings from semi-structured interviews with middle-level leaders in universities. Nineteen interviews were conducted with professors specializing in the fields of Education and Management sciences, chosen from state universities in Iran. Applying grounded theory, as outlined by Strauss and Corbin (1998), the research constructs an order that captures the causal, contextual, and intervening conditions, alongside the strategies and consequences critical for fostering leadership in digital universities. These qualitative insights help in understanding how academic leaders can effectively direct the wave of digital transformation within a higher education landscape. The study provides in-depth insights into the essential attributes and competencies that academic leaders need to develop to guide their institutions effectively during the surging wave of digitalization. It offers a theoretical framework that serves as a guide for educational strategists to develop effective leadership suited for the digital era. Additionally, the framework contributes to the broader discourse on achieving integration, synergy, and efficiency in the digital transformation of higher education. The methodological credibility of this research is reinforced by validity and reliability measures, including the Content Validity Ratio, Content Validity Index, and Cohen's Kappa index. While the conclusions are contextually grounded and informed by expert opinions, they lay out a foundational framework for future research into the dynamic and complicated role of leadership in the middle of the ongoing wave of digitalization in higher education.
Abstract This study addresses the educational challenges of understanding how generative AI chatbots can equitably and effectively enhance student learning in higher education. Chatbot system analytics data and a mixed methods survey (n = 121) was used to examine four research questions related to the use of chatbots (academic value, frequency of use, and attitudes), technology experience and AI literacy (self-reported digital proficiency, device usage and AI familiarity), barriers to use of chatbots (ethical concerns, UX issues and policy ambiguity), and general user experience (clarity, relevance, accuracy, satisfaction and likelihood of recommendation). Descriptive statistics, one-sample t-tests, and ANOVAs revealed generally positive attitudes toward chatbots and perceived gains in knowledge and understanding, together with robust support for academic-integrity requirements. Usage analytics confirmed 24/7 usage requirements, with 36.8% of interactions occurring after hours. Users reported clear, relevant, largely accurate responses and an easy interface but overall satisfaction was mixed. The study recommends human-centred design (e.g., opt-in launch, clearer links to official resources), explicit AI policies and labels for assessment, staff and student training (including prompt literacy, and potential use cases), subject-specific knowledge bases, and continuous monitoring for errors and edge cases, and subsequent update of system instructions and knowledge bases. Study findings highlight conditions under which chatbots can equitably complement human instruction and methods for chatbot engagement across different student cohorts.
Abstract The integration of learning analytics with artificial intelligence represents a paradigm shift in educational decision-making, yet systematic frameworks for AI-enhanced learning design remain critically underexplored in creative higher education contexts where ethical considerations are paramount. Despite growing interest in AI-enhanced education, existing approaches lack systematic integration of learning analytics with ethical frameworks for evidence-based educational interventions in arts-based disciplines. This study develops and validates the Learning Analytics-driven Educational Decision-Making (LA-EDM) Framework, a comprehensive approach for AI-enhanced learning design through data-informed educational decision-making in creative education. A sequential mixed-methods design incorporated quantitative analysis of learning analytics data from 508 Chinese film students, qualitative interviews with 10 film educators, and systematic assessment of 10 student films. Structural equation modeling demonstrated strong model fit ( $$\chi ^{2}$$ /df=2.677, CFI=0.949), with mediation analysis revealing significant pathway relationships. The LA-EDM Framework demonstrates robust predictive validity, explaining substantial outcome variance (R $$^{2}$$ =30.6%−35.7%) in learning design effectiveness. Key findings reveal that Ethical Fitness significantly predicts successful AI integration ( $$\beta $$ =0.262 for technical-artistic balance) and indirectly influences Educational Effectiveness through Technical-Artistic Balance, with this pathway accounting for 19.834% of the total effect. Qualitative analysis identifies critical dialectical tensions including empowerment versus deskilling dynamics and efficiency versus creative depth considerations. This research extends learning analytics theory by providing the first empirically validated framework integrating ethical considerations with data-driven educational decision-making in creative disciplines. The findings offer evidence-based guidance for educators implementing AI-enhanced learning design in arts education, demonstrating how learning analytics can inform personalized and ethically-grounded pedagogical interventions.
Dialogic peer feedback in synchronous online discussions provides multiple educational benefits; however, its implementation is often constrained by students’ cognitive and interactional limitations. To address this issue, this study investigated the effects of argumentation scaffold and role assignment on students’ emotion, social metacognition, and discourse patterns in dialogic peer feedback. A quasi-experimental design was employed, involving 98 undergraduates assigned to four conditions: (a) dialogic peer feedback only (DF; control group; n = 24); (b) scaffolded dialogic peer feedback (SDF; argumentation scaffold only; n = 24); (c) role-assigned dialogic peer feedback (RDF; role assignment only; n = 25); and (d) scaffolded, role-assigned dialogic peer feedback (SRDF; combination of argumentation scaffold and role assignment; n = 25). Results showed that argumentation scaffold enhanced students’ positive emotion, social metacognition knowledge, and social metacognition judgment, whereas adding role assignment reduced these benefits. Discourse patterns further indicated that the SDF promoted deeper argumentation, the RDF increased participation, and the SRDF supported higher-level cognitive engagement. These findings suggest that argumentation scaffold and role assignment shape dialogic peer feedback in distinct ways, with scaffold strengthening the quality of reasoning and role assignment redistributing interactional participation.
Despite the pedagogical potential of Extended Reality (XR) to transform higher education, institutional adoption continues to be fragmented, marked by abandoned pilots and high failure rates. Existing maturity models, such as the Technology Readiness Level (TRL), prioritize hardware functionality but fail to account for the complexity of university environments. This study addresses this gap by developing a systemic framework for university XR readiness. Employing a Socio-Technical Grounded Theory (STGT) approach, semi-structured interviews are conducted with twenty diverse stakeholders - ranging from lab directors to technical specialists across Australia, North America, Europe, and Southeast Asia. The analysis reveals that readiness is not a binary condition defined solely by technology acquisitions, but rather an emergent property from a multitude of factors. The factors are organized into three stages: Foundational (Infrastructure/Logistics), Operational (Content/Governance), and Strategic (Equity/Convergence). XR adoption is propelled by visionary leadership and accelerated by the convergence of Artificial Intelligence. Our findings uncover systemic inhibitors and illustrate that sustainable integration requires alignment among people, pedagogy, policy, and technical infrastructure. The study highlights overarching principles to assist university administrators in implementing effective XR design and integration.
Recent societal issues require higher education institutions to engage students in interdisciplinary knowledge creation beyond knowledge acquisition. While the jigsaw method is expected to facilitate novice learners’ engagement in interdisciplinary collaborative knowledge creation, authoritative expert materials limit learners’ agency beyond the given knowledge. To overcome this constraint, this study proposes the Knowledge Expansive Jigsaw Method, which integrates Generative AI (GenAI) as a resource for expanding ideas beyond expert materials. Using a two-cycle design-based research approach with 70 freshmen using a conjecture map as our framework, we investigated how the implemented GenAI-supported jigsaw method promotes students’ ideation processes. Specifically, students designed illustrations of the “Future classroom of our university in 2050.” The instructional design introduced an expansion phase after a cognitive trigger and positioned GenAI as both a possibility engine and a co-designer. Students’ discourse, GenAI usage records, and final presentations were analyzed using the Knowledge-Building Discourse Explorer (KBDeX) and rubric-based evaluation. KBDeX analysis confirmed the co-occurrence of AI-inspired and human-generated idea expansion across groups. Comparison of the outcomes did not demonstrate improvements after design refinement in Cycle 2; whereas outcomes in Cycle 1 were diverse, those in Cycle 2 converged at an emerging level. The results suggested both the potential—the proposed design consistently promoted idea expansion beyond expert knowledge—and the challenge—high-level idea innovation were rare under the GenAI-supported jigsaw method. Future studies should focus on refining AI systems, instructional designs toward idea innovation and analysis to identify and validate other promising processes critical for idea innovation.
Abstract This systematic literature review analyses AI literacy, focusing on the required competencies for, the obstacles arising from, and the benefits of, Generative AI (GenAI) in the fields of education and business. The analysis uses the PRISMA 2020 methodology with data from the SCOPUS and ERIC databases. A total of 538 articles were identified; of these, 206 were included after the full-text screening phase. Of those 206, only 33% (education) and 29% (business) were based on empirical research, highlighting the predominantly conceptual state of research. Using a combination of inductive coding and GenAI (ChatGPT-4o) validation, we identified AI literacy as a multidimensional concept comprising technical competencies (e.g. algorithmic literacy and prompt engineering), personal and interpersonal competencies (e.g. adaptability and collaboration), and ethical and critical thinking competencies (e.g. awareness of bias and ethical reflection). While educational literature emphasised pedagogical applications such as adaptive feedback and inclusive curriculum design, business research focused on process automation and data-driven decision-making. Top three identified obstacles included hallucinations, ethics and plagiarism, which manifested differently in contexts such as student assessment and personnel selection. Addressing these challenges will require targeted training modules, ethical governance structures, and institutional support in the form of faculty development programmes or workplace reskilling initiatives. Top three identified benefits of GenAI literacy training are described as critical thinking, personalized teaching and learning and personalized feedback across sectors.
Abstract Purpose This study examines how different emotion-aware chatbot interaction modalities, particularly voice-based support augmented with an animated avatar, affect student engagement and code outcomes in undergraduate programming courses, compared with text-only and real-teacher support. Methods We conducted a controlled classroom experiment with Arabic-speaking computer science students at Birzeit University during a 60-minute Java programming task. Participants were randomly assigned to one of four instructional support conditions: text-based chatbot interaction, voice-based chatbot interaction, voice-based interaction with an animated avatar, or voice-based interaction with a real teacher. For all chatbot-based conditions, a Wizard-of-Oz methodology was employed in which a trained human operator simulated an emotion-aware chatbot to ensure consistent and realistic responses across modalities. Data collection included subjective measures (satisfaction, usability, and perceived engagement) and objective indicators (task completion, code readability, maintainability, and accuracy), along with affective signals captured throughout the session. Results The avatar-supported voice condition was associated with more stable positive affect, higher sustained engagement, and better-structured, more readable, and more maintainable code compared with text-based interaction. Across multiple objective outcomes, voice-based support showed advantages over text-only interaction, while the real-teacher voice condition demonstrated strong task completion performance. Differences in subjective usability and satisfaction followed similar trends but did not reach statistical significance. Conclusion The findings suggest that voice-based, emotionally expressive chatbot interfaces, particularly when combined with an animated avatar, can approximate aspects of supportive teacher presence while remaining scalable for large programming classes. Even when implemented using a Wizard-of-Oz approach, emotion-aware voice interactions show promise for improving engagement and selected code quality dimensions. These results offer design guidance for developing classroom-ready AI teaching assistants for programming education, with particular relevance for Arabic-speaking learners.
As GenAI technologies become more pervasive in higher education (HE), scholars call for guidance on AI governance. To meet this need, a Delphi technique and collective writing was used in gathering expert perspectives from across 22 countries/locations and six continents. This resulted in the development of a HE GenAI policy/guidelines framework with eight core areas: (1) academic integrity, (2) ethical use and responsible use, (3) privacy and protection, (4) equitable access, (5) GenAI literacy, (6) integration strategy, (7) human oversight and accountability, and (8) institutional support and infrastructure. In addition, a six-part framework was developed to ensure that policies remain current and relevant: (1) creating a dedicated GenAI Committee, (2) conducting regularly scheduled policy reviews, (3) providing ongoing professional development and support, (4) communicating with all stakeholders, (5) evaluating the effectiveness and impact of GenAI, and 6) monitoring external developments. By providing a robust, eight-part framework for policy and guidelines, alongside a six-part mechanism for continued review, this study offers faculty, students, administrators, educational leaders, policymakers, and funders a responsible, adaptable, and consensus-driven blueprint for navigating the integration of GenAI in HE, ensuring that technological innovation serves pedagogical excellence.
Abstract Recommender systems have become increasingly important in online learning platforms, where learners face a growing number of available courses and substantial variation in course quality and relevance. However, many existing course recommendation methods primarily rely on user–course interaction sequences and therefore often underutilize rich course semantic information, such as descriptions of learning objectives, covered concepts, and prerequisite-related content. This limitation is particularly problematic in educational settings, where learners’ next-course choices are influenced not only by historical behavior but also by semantic relationships among courses. To address this issue, we propose LLM-CR, a recommendation framework that leverages large language models (LLMs) to enrich course representations with structured semantic information. Specifically, course metadata and textual descriptions are processed offline by an LLM to produce semantically informative course summaries, which are then encoded into dense representations and incorporated into the recommendation pipeline through a lightweight fusion module. In this way, LLM-CR augments conventional behavioral representations with course semantics related to topics, knowledge progression, and prerequisite-relevant information. The resulting model is evaluated in the standard next-course recommendation setting using ranking-based metrics. Experiments on six subject-specific subsets constructed from the XuetangX platform show that LLM-CR consistently improves recommendation performance over strong baseline methods, with especially notable gains on relatively sparse datasets. Additional analyses indicate that the proposed semantic augmentation introduces only modest extra complexity because the most expensive LLM processing is performed offline and reused across training and inference. These results suggest that incorporating LLM-derived semantic features is an effective and practical way to improve course recommendation quality.
Abstract This research examines the impact of psychological and cultural factors on the academic use of digital technologies by Chilean pre-service teachers. Considering the global digital transformation and the influence of digital technologies, such as artificial intelligence (AI), in education, it is essential to prepare future educators to integrate technology pedagogically. Although the use of technology among pre-service teachers has been studied, the specific factors influencing their academic use require further investigation considering an integral approach. A study was conducted with 1191 pre-service teachers from various Chilean universities. Data were collected through questionnaires assessing academic use of digital technologies, motivation towards Information and Communication Technologies (ICT), technology attitudes, self-directed learning readiness, disposition towards learning and teaching, technological infrastructure availability, and portrait values. Structural Equation Modeling (SEM) was utilized to analyze the data. The results indicated that self-management, disposition towards pedagogical learning, and motivation were the most significant direct predictors of the academic use of digital technologies. Among the cultural factors, self-transcendence, openness to change, and the use of portable PCs had a significant but minor direct influence. The model explained 32.1% of the variance in academic technology use. These findings underscore the significance of both psychological characteristics and cultural values in influencing how future teachers utilize digital technologies for academic purposes. Understanding these influences is crucial for enhancing teacher education programs and equipping educators to integrate technology into teaching and learning effectively.
