In an earlier blog, we discussed how integration is essential for the next generation learning systems to break the silos and be truly distributed systems. Learning Management Systems (LMS) need to embrace multiple devices, offer offline support and work across multiple learning platforms. Tracking and measuring every learning experience to maintain a detailed log of learner’s activities is thus critical. Data capture and learning analytics, therefore, play a central role in the next generation learning systems.
Experience API (xAPI), along with the Learning Record Store (LRS), provides a mechanism to track all such activities. It employs a standard format and vocabulary to maintain learning records. One can also complement it with a custom vocabulary. It thus facilitates the measuring of a variety of data that business use cases demand. To know more about adding support for xAPI in your learning system, read this blog.
Analytics offers deeper insights than reports. For example, a report could tell you the grades a learner secured in a course. Analytics would help you understand the correlation of those grades with the performance of the learner, as well as the performance of your business. To generate detailed analytics, your application would need to capture a learner’s activity to a much finer granularity. For example, how much time the learner spent on a page, which type of content is the learner skipping, and how many attempts did the learner need to pass in a course. The steps involving in extracting analytics have been covered in the following section.
Four Steps for Rich Analytics
Here are the four steps that you would generally come across while extracting analytics:
Measurement: It’s the simple act of tracking things and recording values. It doesn’t demand any complex algorithm – just the discipline of capturing data that matters. There can’t be any analytics without data.
Evaluation: The next step is extracting meaning from the data. With small sets of data, simple statistical values like the most or least popular courses across the organization can be extracted on an ad-hoc basis.
Advanced evaluation: When data sets get large, you can use advanced techniques like data mining and machine learning for statistical insights on essentially any parameter(s). This can help you extract information like region wise course popularity, courses popularity by employee age group, or course performance based on content providers.
Predictions and recommendations: The next step is to discover patterns and make predictions and recommendations. Technologies like machine learning and artificial intelligence ease this process and help to automatically identify patterns in your data and make better predictions.
Varied Learning Analytics for Different Stakeholders
Learning analytics can serve varied purposes for different stakeholders, i.e. admins, content writers, learners, managers, and business leaders.
Every learner action is logged by xAPI. For instance, it tracks learner interaction with other people through social media, task status updates, stages and milestones in games, play and pause actions over video content, and switching between different sites during quizzes. Such detailed logs of the learner’s’ activity make it possible for L&D teams to design dashboards that reveal patterns of the following kind.
- Statistics of viewed v/s skipped a video
- Which sites or additional help sources do learners switch between during the course of learning?
- How much time is spent on each learning section?
- How many learners answered a specific question correctly/incorrectly?
- What was the average number of attempts for each question?
- How did the demonstration video perform (if it was used)?
For content writers
Analytics extracted by xAPI can be useful for content writers to improve the effectiveness of the courses. For example, they can get answers to the following questions.
- What kind of content is most effective – text, audio, or video?
- What kind of interactivities drives the most engagement?
- Are learners interested in the depth or breadth of the content?
For administrators, xAPI can provide standard analytics, just like traditional learning systems. These include insights on which learners have completed a given course, percentage of course completion for each learner, and assessment results.
Just knowing a learner’s progress is not enough for any organization. Businesses need to know how learning impacts employee performance. This performance is measured by a number of parameters, like industry knowledge, product knowledge, problem-solving skills, and technology skills. Weightage to each parameter varies by situation. These parameters are typically available in multiple external systems. SCORM doesn’t allow integration with multiple systems. With xAPI one can fetch information from multiple external systems and store in LRS, thus allowing learner’s performance measuring across systems.
Analytics can reveal business insights – e.g. which content provider is working and which one is not, should the system aim for wider breadth or have more specialized content, which is the trending topics, and so on. The next level of analytics is powered by machine learning, where, based on historic patterns, the system can predict which courses will be sought after. The system can also make recommendations for course selection based on the learner profile and objectives.
The global education and learning analytics market is expected to grow from USD 2.6 billion in 2018 to USD 7.1 billion by 2023, at a compounded annual growth rate (CAGR) of 22.6%. Is your learning system roadmap designed to keep pace with this explosive growth? Can it track every learner action, their progress and correlate it with business performance, and provide actionable insights? Do share how you plan to deal with this challenge!