How Schools Can Use Student Data to Improve Learning

Learning analytics, also known by its Spanish equivalent *analítica de aprendizaje*, represents a significant development in understanding and optimizing educational processes. At its core, it involves the systematic measurement, collection, analysis, and reporting of data specifically related to how people learn. This field aims to generate new insights from existing data, ultimately seeking to improve learning outcomes and the environments in which

learning occurs. Its global relevance is underscored by its use in numerous languages, including Arabic, Bulgarian, Catalan, Czech, Estonian, Persian, French, Hebrew, Korean, Swedish, and Xhosa, highlighting a worldwide interest in leveraging data for educational advancement.

Defining the Scope of Learning Analytics

One of the primary definitions, widely adopted since the 1st International Conference on Learning Analytics, describes it as the measurement, collection, analysis, and reporting of data about learners and their contexts. The ultimate goal is to comprehend and optimize both the learning process itself and the environments where it takes place. This perspective emphasizes a holistic view, aiming to understand the entire system rather than just isolated components. While this definition is prevalent, the precise scope and objectives of learning analytics remain subjects of ongoing discussion within the community.

Another way to conceptualize learning analytics is by viewing it as a specific type of "analytics." In this broader sense, analytics is the process of developing actionable insights through problem definition and the application of statistical models and analysis to existing or simulated future data. From this viewpoint, learning analytics emerges when the data, problem definition, and insights are all directly related to learning. This connection highlights its role in informing decisions that can enhance educational practices.

Distinguishing Learning Analytics from Educational Data Mining

The fields of learning analytics (LA) and educational data mining (EDM) are closely related, both dealing with the collection, analysis, and exploration of education-related data. However, they possess distinct focuses and methodologies. EDM often concentrates on developing automated methods to achieve the best possible prediction regarding the teaching-learning process. It tends to be reductionist, aiming to restrict systems to their components and investigate their relationships, often using techniques from statistics, machine learning, and data processing.

In contrast, learning analytics leans towards adopting techniques that incorporate human judgment to better understand social educational phenomena. It seeks a more holistic approach to grasp the entire system. While EDM might focus on automated adaptation through intelligent tutoring systems, LA emphasizes informing and empowering learning environments for instructors, teachers, and students, leading to personalized and adaptable settings. Despite these differences, both fields utilize similar analytical techniques, such as social network analysis and semantic discourse analysis, though their ultimate objectives may vary slightly.

The Role of Data and Prediction in Learning Analytics

Learning analytics heavily relies on various types of data and analytical techniques to achieve its goals. It utilizes descriptive analysis to represent data intelligibly through summarization, visualization, and interactive interfaces, allowing instructors to immediately see student activities. Diagnostic analysis is employed to identify relationships between student behavior and contextual variables, validating phenomena as components in further analysis. Predictive analysis is crucial for inferring future events, such as student behavior and performance, by establishing a foreseeable variable and applying it to a group.

These analytical methods enable learning analytics to determine several key aspects: the probability of a student successfully completing a course, the likelihood of a student dropping out, the probability of a student needing additional support, and the probability of a student achieving specific learning outcomes. By providing these insights, learning analytics offers a powerful tool for educators and institutions to proactively support students, personalize learning experiences, and continuously improve educational effectiveness. The integration of computational aspects with existing educational research is seen as vital for learning analytics to fully realize its potential in understanding and optimizing learning.