Learning analytics is a far reaching broad encompassing idea that often sees varied definitions and equally varied applications. Learning data may be descriptive, diagnostic, predictive or prescriptive in nature as data is tied to specific contexts and questions. For the purposes of our discussions, learning data encompasses use of data to both answer questions about, and to improve, teaching and learning.

What are Learning Analytics?

Educators have used information to make innovative and iterative changes to the world of education for years. Grade Point Averages, transcripts, even grades themselves are all measures of learning, all forms of analytics that inform teaching practices, all ideas that came about likely through questions instructors had related to teaching and learning.

New in the last decades has been the specificity and accessibility of this data. As education has become increasingly more digitally mediated, as the use of learning management systems have increased, so too has the ability to capture and visualize aspects of student interactions and learning data and to do so in nearly real time. Instructors often have questions related to their course or practice. And now more than ever before, learning data may inform the answers to those questions.

Types of Learning Analytics

Using learning analytics, or learning data, allows instructors a tool to reflect upon their activities, their teaching, their course or program to identify areas that may be altered, improved or better understood. The use of learning data may yield immediate, distinct knowledge that is actionable, some change that could be made in a semester to an upcoming activity or order of activities. Or, data may yield information that affects future course design, programming decisions or a combination of all of these aspects. Generally, questions may be grouped into larger categories associated with different types of data.

What occurred?	Descriptive Learning Data
Why did it occur?	Diagnostic Learning Data
What might occur?	Predictive Learning Data
What should occur?	Prescriptive Learning data

Regardless of the question or goal, using learning analytics requires an understanding that the uniqueness of courses will lead to unique sets of data and that this data requires interpretation to be used successfully.