Data is increasing with the use of learning technologies, and data is being produced at virtually every learning footprint. The next step in the process is to take the data and analyze the connections to improve the entire learning experience.
Learning analytics is the measurement, collection, analysis, and reporting of data about the learners and their contexts for the purpose of understanding and optimizing learning and the environment in which it occurs. 
Learning analytics has been around for some time. Its origin can be traced to business intelligence and to predicting consumer behavior. Learning analytics in education has emerged in the last few
decades, and it follows similar analysis and predictive relationships. Learning analytics is growing to keep pace with deciphering patterns from huge data sets to further support and personalize the learning experience.
My interest in learning analytics stems from my research on learning style preferences. The hypothesis was that, if you could determine a user’s learning style preference, then you could optimally display content in a form to best suit the way a learner could interpret it; you could support their success. At that time, most analysis had to be completed prior to the learning, and then you could track users accordingly. Real-time data analysis was in its infancy. The vision then was that, in the future, this could be done via machine learning, with data analysis and dynamically serving up content in a format that learners best understood. Today, those capabilities exist in some learning management systems in the form of learning analytics and adaptive learning.
Currently, most learning management systems are able to track a student’s footprint throughout a course. It can document when a user logs in and logs out, and they can determine the type of content they viewed and for how long. They can also alert students to assignments, assessments and most course requirements, including their status within each course. Some learning systems have dashboards that indicate the students’ progress compared to their expectations and compared to their cohorts’ performance.
In my opinion, most learning management systems are good at data reporting, but they fall short in data analysis and in relationships. The challenge is to harness the data and to make reasonable connections, so that meaningful, positive and proactive interventions can be made; ultimately, we hope to improve the instructional process and student success.
Why use learning analytics:
Learning analytics has relevance and usefulness across various groups, including instructors, students, instructional designers and institutions.
Instructors can use learner analytics to gain insight into student progress:
- Course navigation paths
- Most popular content
- Reflection time
- Measurement of student engagement and participation
- Assignment and assessment completion
Analytics can also be used as an early warning system for at-risk students; they can trigger appropriate messaging.
Students can use learner analytics to gain insight into their progress:
- Seeing their progress and grades
- Tracking their progress against course requirements
- Comparing their progress with their cohorts
- Tracking content and resources
As computer technologies develop and more learning components are online, it is essential for learning specialists to evaluate the impact of each emerging technology and to investigate the strengths, weaknesses and appropriate applications for the learners. Sometimes, this is in the form of a retrospective analysis, but increasingly this analysis can be done closer to the time of the event for more agile course adjustments.
Learning analytics can also be used for continuous improvement of the learning design, such as increasing learner engagement, expanding knowledge retention and improving course and program
Learning analytics can be applied at the institutional level for reporting usage trends. In the future, courses could have personality profiles based on course metadata. These items could include tags, such as “projects-based learning,” “discussions,” “hybrid” and “synchronous.” Each metadata tag could also have an associated strength. Each student would also have his or her own evolving learning personality profile.
This data matching would be similar to how Amazon recommends products based on a customer’s purchasing history and behavior. To optimize student success, the recommendation engine architecture could suggest courses that best match the profiles and that mesh with individual learning styles.
Learning analytics—one view but not the whole picture:
It would be short sighted to think that the landscape of learning analytics is only within the confines of an online learning management system. It is increasingly apparent that the majority of learning
occurs outside of the learning management system; it is only the tip of the iceberg. Learning also occurs informally, such as through social media, experiences and discussions. Learning analytics should be inclusive, capturing all learning opportunities. The Experience API (xAPI) has been developed as a mechanism to record and track all types of learning experiences. Ultimately, inclusion of this learning data will broaden analysis and connections. However, in my experience in piloting the xAPI, it is more elusive than reality. It will take time for the experiential footprints to be folded into the mix of the learning data.
Learning analytics is not a one-time, one-size-fits-all approach. It is dynamic, as the parts of the system change and grow. Learning analytics is an emerging field that can benefit many; it has the potential of being a significant factor in improving the overall learning experience in educational institutions or in corporate training.
 Society for Learning Analytics Research, 2011.
 Low, G. (1995). A study of the effects of learning style preference on achievement in a medical computer simulation (Doctoral dissertation). Retrieved from UMI Dissertation Database (Accession No. ALMA BOSU1 21625699380001161)
Not subscribed to our blog?