Dirk Ifenthaler

The paper proposes a three‑level framework (micro/meso/macro) that synthesizes learning theories to explain how AI can promote learning processes. It articulates a causal learning mechanism and describes how information is aggregated and dissipated across levels. Fourteen roles for AI are outlined: four at the individual level, four for teams/knowledge communities, and six at the macro level of cultural‑historical activity, with implications for research, practice, and evaluation.

This systematic review synthesized empirical evidence on learning analytics interventions for supporting study success in higher education. From an initial pool of 6,220 articles, 46 key studies were analyzed. Evidence indicates numerous approaches that effectively support retention and completion, though rigorous, large‑scale trials remain scarce. The review catalogs variables, algorithms, and methods to guide future design and implementation of learning analytics systems aimed at improving student success.

Responding to limited models for technology adoption at the organizational level in education, this paper develops a theoretically grounded digital transformation maturity model and reports a case study in a European educational organization (N=222). Findings show generally positive IT attitudes and classify the organization as digitally pragmatist. The study recommends transparent model development, rigorous validation, and attention to mediating factors such as organizational culture and human resource development.

Combining qualitative interviews (n=20) and a survey (n=216), this study investigated students’ expectations for learning analytics features and their willingness to use them. Students expect features that support planning and organization, self‑assessment, adaptive recommendations, and personalized analyses of learning activity. The findings link desired features to phases of self‑regulated learning and inform design of learner‑facing analytics tools and dashboards.

Using a learning analytics benefits matrix, this study investigated institutional capabilities, perceived importance of data sources, and the value of insights from learning analytics in higher education. Results reveal shortages in staff capacity and technology to support projects, underscoring a need for empirical research on framework validity and on benefits for learning and instruction to meet high expectations for this emerging area.

Based on deliberations of EDUsummIT 2015 Working Group 5, this synthesis focuses on technology‑enhanced formative assessment to support inquiry‑ and problem‑based learning. It discusses feedback sources (teacher, self, peer), evidence‑centered design, and automation to strengthen assessment for and of learning, concluding with recommendations for future research and practice.

This exploratory study examined students’ perceptions of privacy principles in the context of learning analytics, including data collection, storage, analysis, and presentation to stakeholders. With 330 university students, findings show that learners expect adaptive, personalized dashboards but are generally conservative about sharing data. The results suggest that all stakeholders should be involved in the design and implementation of learning analytics to ensure alignment with privacy expectations and acceptance.

Synthesizing outcomes from EDUsummIT 2015, this paper discusses potentials, concerns, and issues regarding technology‑enhanced assessment as, for, and of learning, with an emphasis on formative assessment. It identifies design considerations for feedback and dashboard technologies that support inquiry‑ and problem‑based approaches and concludes with recommendations for advancing research and practice in technology‑enhanced formative assessment worldwide.

This empirical study examined how motivation and engagement relate to complex problem‑solving in a massively multiplayer online game‑based learning environment. Analyses revealed interrelationships among motivational constructs, engagement behaviors, and problem‑solving performance, offering evidence that well‑designed digital game‑based contexts can foster both motivation and complex problem‑solving competencies.

This article proposes and validates a learning analytics framework through two institutional case studies. Using support vector machines, the study develops student and learning profiles and tests their validity and accuracy. Findings indicate that educational data are highly context‑specific and that variables may carry different meanings across institutions and programs. The paper discusses benefits, concerns, and challenges of learning analytics and argues for frameworks that are sensitive to local idiosyncrasies and stakeholder needs.

Across two experiments, the study compared preflective and reflective prompts in online learning, while controlling for motivation and learning preferences. Results indicate that directed preflective prompts benefit novice learners and can activate positive motivation in online environments. The paper highlights the need for personalized and adaptive realization of preflective prompts and automated feedback to better support self‑regulated learning.

This empirical study examined relationships among learners’ motivation, engagement, and complex problem‑solving outcomes during gameplay in a massively multiplayer online game. A model is presented linking problem representation, motivational constructs (e.g., interest, competence, autonomy, relatedness, self‑efficacy), and engagement. Findings suggest that motivation influences engagement, which in turn shapes development of complex problem‑solving competencies, and that task design strongly affects these dynamics.

Limited research has examined teachers’ acceptance of Tablet‑PCs in classroom instruction. This qualitative study used semi‑structured interviews with 18 teachers during a pilot project introducing Tablet‑PCs at three schools. Findings reveal diverse attitudes toward the technology as well as differences in performance expectancy and perceived facilitating conditions. Results complement technology acceptance research by providing a fine‑grained qualitative perspective on conditions that shape adoption in K‑12 settings.

An experimental study (N=98) investigated instructional interventions for self‑regulated learning during problem solving and introduced an automated assessment method for evaluating learning outcomes. Results indicate that generic prompts are a valuable aid for developing cognitive structures while solving complex tasks, supporting learners’ planning, monitoring, and evaluation processes in computer‑based environments.

The paper introduces Highly Integrated Model Assessment Technology and Tools (HIMATT), a web‑based suite for assessing development of complex cognitive skills via model‑based representations. The authors describe functions of the system, illustrate applications in educational and workplace settings, and argue that HIMATT scales beyond laboratory use to support practical, automated assessment of mental models.