Foreword by Dragan Gašević

My introduction to learning analytics was unexpected. A brief conversation with George Siemens in early June 2010 sparked a transformative journey for me, with lasting impact on many. George proposed co-organizing a conference on “learning analytics.” It was the first time I’d heard of the phrase. The phrase resonated deeply, connecting with my previous collaborations with Jelena Jovanovic, Chris Brooks, Marek Hatala, Griff Richards, Gord McCalla, Colin Knight and others colleagues in Canada’s LORNET research network (mid-2000s).

Excited to collaborate, I was initially a bit skeptical about George’s ambition to organize the conference by September 2010. However, I learned to appreciate his eagerness to implement ideas swiftly. We eventually reconnected and agreed on a late February 2011 date, giving us extra six months. Despite remaining nervous, I underestimated the potential impact. The experience proved me wrong in the most positive way, and I’m forever grateful to George, a dear friend, for his unwavering vision and belief.

The conference we organized, the 1st International Conference on Learning Analytics and Knowledge (LAK), was held in Banff, Canada, in late February and early March 2011. The frigid temperatures (-35°C) were matched by the conference’s energy. It attracted a large audience and fostered an open, vibrant, productive, and transformative community of researchers and practitioners. This conference is often considered the birth of learning analytics, and the definition we used in the chairs’ message in the conference proceedings remains widely used in the field.

Learning analytics is now a well-established field. Since our early conversations, we’ve emphasized the importance of methods. My dear friend and esteemed colleague, Shane Dawson, a co-founder of the Society for Learning Analytics Research and field pioneer, first referred to learning analytics as a “bricolage field,” one that borrows methods and theories from data science, artificial intelligence, network science, psychology, psychometrics, sociology, and natural language processing.

As founding editors-in-chief of the Journal of Learning Analytics, Shane and I recognized the need to introduce and promote diverse learning analytics methods. This led us to collaborate with Mykola Pechenizkiy, then-president of the International Educational Data Mining Society, to edit a special section on learning analytics tutorials featuring five papers on different methods. These papers, based on tutorials and workshops from the first two Learning Analytics Summer Institutes, became some of the journal’s most impactful publications. It demonstrated the need for learning analytics methods tutorials. However, eight years have passed, a significant timeframe considering the rapid advancements in artificial intelligence and the methods they’ve driven.

Therefore, I’m delighted to see “Learning Analytics Methods and Tutorials” edited by Mohammed Saqr and Sonsoles López-Pernas. This timely book addresses a critical need in learning analytics.

The book offers a comprehensive guide to data analysis methods for researchers and practitioners of all experience levels. It starts with the basics, equipping beginners with R programming and data analysis skills through chapters on data cleaning and exploration. These skills are fundamental for understanding student data and preparing it for further analysis. Even for experts, the book offers advanced methods while emphasizing the broader applicability of these techniques beyond education.

The core of the book explores various analytical approaches. Machine learning methods receive well-deserved attention, including introductions to commonly used methods – specifically, predictive modeling and clustering. Predictive modeling is frequently used in learning analytics to identify at-risk students or classify online discussions by analyzing past data patterns. This allows for early intervention to support students at different progress levels. Cluster analysis , another machine learning technique, groups students based on similar characteristics, behaviors, or learning outcomes. It’s commonly used to analyze learning strategies in different learning environments and can provide educators with valuable insights to tailor teaching support to diverse student needs.

We’ve long recognized the dynamic nature of learning, with many learning processes unfolding over time. This highlights the need for temporal analytic methods in learning analytics. I’m pleased to see the book provides a rich guide to various temporal methods that analyze the order and timing of events in data about learning and learners. Techniques like sequence analysis and process mining leverage student activity traces to understand how learning unfolds over time. This is crucial for studying longitudinal processes like student engagement throughout a program or explaining connections between cognitive and metacognitive processes in solo and group learning activities.

Network analysis has been a prominent methodological approach since the early days of learning analytics. The book offers excellent coverage of network analytic approaches that explore the relational aspects of data about learners and learning environments. By examining interactions between learners, teachers, and topics, researchers and practitioners can understand collaboration patterns and identify student communities. These methods can be further enriched by combining them with temporal analysis to explore how these relationships evolve over time. The book also covers recent advancements in quantitative ethnography, introducing epistemic network and ordered network analysis, techniques I’ve extensively used with collaborators in recent years.

I have recently been advocating for the need to establish stronger collaborative ties between learning analytics and psychometrics. Psychometrics is a well-established discipline that can offer many relevant methods to learning analytics. Specifically, psychometrics can help us address issues that have received insufficient attention in learning analytics – reliability and validity. Without addressing these issues, learning analytics can be jeopardized. Therefore, I am delighted that the final section of the book delves into psychometrics, a field that investigates relationships between psychological constructs (like metacognition) and observable data (like test scores). The book explores techniques like factor analysis and structural equation modeling, which help researchers test hypotheses and theories about these relationships. By mastering these methods, learning analytics researchers and practitioners can gain valuable insights from student data to improve learning experiences and outcomes.

The book editors and chapter authors must be commended for their valuable contributions to learning analytics. What is outstanding about this book is that it provides source code illustrating all the methods introduced. Readers can directly use the code to build hands-on skills on the many high-importance methods for learning analytics that are so thoughtfully introduced in this book. The book can be directly used in any graduate or postgraduate course on learning analytics. I wish this book had been available three years ago when we developed the graduate certificate in learning analytics at Monash University. Many of its chapters would have been directly used as part of the graduate certificate program. And, I can easily see that many similar programs around the world will greatly benefit from this outstanding book. The book is equally suitable for practitioners and researchers in education institutions, schools, or industry who either want to develop basic data analytic skills or advance their skills with methods they haven’t used before.

The overall learning analytics community is significantly richer because of this book. For that, we all owe immense thanks to Mohammed Saqr and Sonsoles López-Pernas, the book editors, for their incredible leadership and vision!

Dragan Gašević
Distinguished Professor of Learning Analytics
Director of Research in the Department of Human Centred Computing
Director of the Centre for Learning Analytics
Monash University