Editorial: Actionable Learning Analytics in Education: An Opportunity to Close the Learning Loop

JohnPaul Kennedy ^1* Florence Claude Gabriel ¹ Malgorzata Korolkiewicz ¹ Irina Rets ² Bart Rienties ²

• ¹ University of South Australia, Adelaide, Australia
• ² The Open University (United Kingdom), Milton Keynes, England, United Kingdom

The need for actionable learning analytics LA has traditionally focused on descriptive, predictive or prescriptive approaches to student performance (e.g., Brass et al., 2023;Chiera et al., 2021;Gabriel et al., 2022;Herodotou et al., 2020). Data is collected, analysed, and presented to educators and policymakers to evaluate student outcomes and sometimes inform future educational delivery for other students. However, rarely does this feedback loop offer actionable insights that can be immediately used by current students to enhance their learning. For LA to truly impact learning, it must go beyond performance monitoring to facilitate self-reflection, goal setting, and adaptive learning strategies; that is, LA must be actionable.To achieve these outcomes, this research topic brings together articles that address three key themes: theoretical frameworks for integrating learning analytics into learning processes, empirical studies on the impact of learning analytics on student outcomes, and case studies demonstrating practical applications of learning analytics in educational settings. Together, these contributions offer a comprehensive view of how learning analytics can be transformed into an active agent of change in educational practice.Several articles in this special issue highlight methodological innovations that advance the field of LA by developing new tools and approaches for analysing educational data. For example, van Berk et al. build on the Train Track Task (Bryce & Whitebread, 2012) by digitising it to capture log data and analysing these to track students' monitoring and control behaviours during problem-solving tasks. This innovative approach employs finite state machines triangulated with think-aloud data to interpret complex behavioural patterns, offering educators actionable insights to support learning strategy development in real-time.Another methodological innovation comes from Vasco et al., who compare two analytical methods-Discrete Time Markov Chains and Epistemic Network Analysis-for evaluating group-level discourse in computer-supported collaborative learning. By analysing the affordances of each method, the authors demonstrate how these approaches can reveal patterns in collaborative interactions that might otherwise go unnoticed. These methods offer new ways for educators to interpret discourse data, improving their ability to foster metacognitive awareness and enhance collaborative learning outcomes.Designing effective LA systems requires careful consideration of both technological and pedagogical elements. Ouhaichi et al. explore using multimodal learning analytics to identify key design considerations that enhance the practical impact of these tools in educational settings. By stressing the human factor in LA, their Multimodal Design Framework provides practical guidelines for integrating data sources to support students' real-time learning, social engagement, and adaptive strategies.Tempelaar et al., explore how LA systems can be tailored to align with students' unique learning dispositions. By examining patterns in students' engagement with learning tasks and tools, their work highlights the importance of personalising LA to meet diverse learner needs. These insights underscore the potential of LA to intertwine traditional learning metrics with affective measures to identify and address deeper understandings of learning behaviours.The longitudinal case studies in this special issue provide valuable insights into the practical applications of LA in diverse educational settings. Johnson et al. explore how social network analysis can be used to monitor decision-making patterns in a large Australian school. The study demonstrates how these metrics can be translated into actionable insights to support an expansive perspective on education. This pragmatic approach to learning analytics emphasises the importance of contextualising data within human activity systems recognising the uncertain and ongoing nature of learning.Barragán and Marcelo examine the use of standardised government tests as an educational quality indicator in Colombia. The study highlights the challenges of relying solely on standardised assessments to evaluate educational outcomes and calls for a more comprehensive approach that incorporates LA to address discrepancies between policy goals and student performance. By expanding the scope of LA, this research demonstrates the potential for more context-sensitive approaches to educational evaluation.The contributions to this research topic collectively emphasise the need for a more holistic approach to LA. As widely noted (e.g., Dawson et al., 2019;Rets et al., 2023;Viberg et al., 2018), LA often tends to focus on outcomes, such as academic performance and knowledge acquisition. However, as several articles in this collection demonstrate, it is equally important to consider non-cognitive outcomes, such as motivation, self-efficacy, and metacognitive skills. These factors all play a critical role in students' ability to regulate their own learning and achieve long-term academic success. Furthermore, as highlighted by these articles, learning cannot be separated from its cultural and historical context; individuals and communities shape and are shaped by their environments. Therefore, LA needs to shift from tracking isolated learning transactions to understanding and adapting complex learning systems. These nuanced perspectives challenge the one-size-fits-all approach to learning strategies and underscore the need for personalised LA interventions.Despite the promising findings presented in this research topic, several challenges remain in the field of actionable LA. Of these, the key challenge is the need for greater collaboration between researchers, educators, and policymakers to ensure that LA systems are designed and implemented in ways that meet the needs of all stakeholders. The case studies in this special issue provide valuable examples of how such collaborations can be achieved, but more work is needed to scale these efforts across diverse educational contexts.This research topic has demonstrated that actionable LA has the potential to close the educational loop. That is, actionable LA has the power to bridge the gap between data collection and practical application. As the field of LA continues to evolve, we need to prioritise actionable insights that empower learners and other stakeholders to ensure that the innovation emerging from the field of LA addresses the real challenges of learning and teaching and not only the challenges that are (relatively) easy to understand and measure.