It’s here: xAPI is the emerging learning data and analytics specification that will replace—and completely supersede—SCORM in coming years. Also called Experience API, xAPI is an open, industry-wide specification for sending, storing, and retrieving activity about learning and performance experiences. It offers instructional designers (IDs) the ability to capture and use far more data about the learning experience than they ever could before. Even better: xAPI captures rich data in any ID model.
Until recently, an xAPI implementation required custom programming and integration, putting even proof-of-concept projects out of reach for many IDs. With the advent of “geek-free” tools offering xAPI out of the box, xAPI is ready for mass adoption.
This offers IDs a huge leap from the limited data set that they typically work with in a SCORM-based eLearning environment. SCORM tracks the same set of somewhat useful data for every piece of eLearning in the LMS: status, score, time, a bookmark, questions, and dates. SCORM is predictable. IDs can count on rapid authoring tools to send this data, and LMSs can report on it. The interoperability that the SCORM specification enabled has led to the incredible growth—and fragmentation—of the eLearning industry.
It’s also incredibly bland. With SCORM, the L&D team can run the same report on eLearning offered for sales training, for compliance, for safety, for soft skills, and so on. If all that matters is completion of assigned courses, this makes sense. When the focus shifts to actual learning, though, IDs need more and better data. That’s where xAPI comes in. With xAPI, IDs can:
- Gather data from more than just eLearning
- Learn more about the learning experience
- Learn more about learners’ performance
- Correlate learning with performance
- Offer more targeted training
- Support performance in better ways
- Use data to learn with others
- Compare performance and learning across learners
- Deliver and track training outside of the LMS
Note: The actions in this list have been achievable for a decade using basic web technologies; xAPI is not revolutionary in its technology so much as in the interoperability that occurs when the same shared specification is used for reporting and tracking data.
If I can send any data, what data should I send?
Leaving data strategy and analytics aside, the answer to this question can be deceptively simple. IDs are familiar with instructional design models and approaches that provide insights into the data that’s tantalizingly out of reach with SCORM. Let’s look at three common ID models and what xAPI makes possible.
70-20-10 and xAPI
The 70-20-10 model suggests that 70 percent of our knowledge comes from experience; 20 percent comes from interactions with others, such as peers, mentors, and managers; and a paltry 10 percent comes from formal learning.
With traditional SCORM-based LMSs, IDs can track only the portion of that 10 percent that happens in eLearning—leaving a reported 95 percent of learning and performing experience un-measured.
With xAPI, though, the instructional designer can track:
Within the 10 percent from formal learning: xAPI makes available more detail about what goes on during the learning experience. It also can bring in data from non-eLearning tools, including performance support tools.
Within the 20 percent from interactions with others: xAPI can capture social tool interaction, reports of coaching conversations, data from observation checklists and rubrics—and much more.
Within the 70 percent from experience: xAPI collects data from systems of work where employee performance is recorded (manufacturing systems, sales systems, customer relationship management, enterprise resource planning, logistics systems, service desk software, etc.).
Kirkpatrick Levels of Evaluation and xAPI
The Kirkpatrick Levels of Evaluation offer another example of what kinds of data L&D teams could be tracking with xAPI:
Level 1—Satisfaction: Most LMSs offer a basic level 1 evaluation on formal learning. With xAPI, this type of data can be collected on a variety of learning experiences, including social, informal, and performance support systems.
Level 2—Knowledge: SCORM 2004 does a decent job of tracking test results. With xAPI, it’s possible to track other types of hands-on or practical assessments and scored observations.
Level 3—Application: LMSs that track Level 3 typically rely on surveys of individuals or their supervisors. With xAPI, it’s possible to pull data directly from systems of work where employee performance is recorded.
Level 4—Results: It is beyond the capability of nearly all LMSs to track actual organizational results, so reporting at this level—if done at all—is external to the learning environment. xAPI enables us to bring outside data into the learning ecosystem to analyze Level 4.
Moments of Learning Need and xAPI
Expanding from the formal training world to informal learning and performance support, xAPI can provide meaningful data from all five of Conrad Gottfredson and Bob Mosher’s Five Moments of Learning Need (as well as additions, bringing the total to Nine Moments of Learning Need):
Before learners know they need to know something, IDs can use xAPI to gather data about work performance and use that information to pinpoint learning needs.
When learners are preparing to enter the training environment, measuring the experience of a variety of preparatory exercises—such as scavenger hunts, pre-reading, discussion groups, and self-diagnostic tools—allows for a more personalized and targeted learning experience once formal learning begins.
When learners are learning something new or learning more, expanding the breadth and depth of what they have already learned, xAPI offers far more insight into what goes on in the formal eLearning experience than SCORM ever could.
1. Before learners know they need to know something, IDs can use xAPI to gather data about work performance and use that information to pinpoint learning needs.
2. When learners are preparing to enter the training environment, measuring the experience of a variety of preparatory exercises—such as scavenger hunts, pre-reading, discussion groups, and self-diagnostic tools—allows for a more personalized and targeted learning experience once formal learning begins.
3 and 4. When learners are learning something new or learning more, expanding the breadth and depth of what they have already learned, xAPI offers far more insight into what goes on in the formal eLearning experience than SCORM ever could.
5. When learners remember and review what they’ve learned, boosters, post-course drills, and other practice exercises can be tracked using xAPI; the data can be used to predict who might need additional support on the job.
6. When learners apply their new skills and knowledge on the job, they might need support in the moment. With xAPI, IDs can connect experience and usage statistics from performance support tools to the formal learning tools, drawing a more complete picture of who is learning and using what, and to what effect, on the job.
7. When problems arise or things break and the learner must solve a problem, xAPI can offer insights into who is using troubleshooting tools, what search terms they are using, what resources learners are sharing in discussion groups, and how the peer community is accepting these resources.
8. When something changes and an employee needs to learn a new way of doing something—which often takes place in a more formal learning environment—xAPI makes it possible to bring in data from the individual’s work experience. This results in more carefully targeted training.
9. Learners may need to teach what they know and help others apply it. Expert communities and mentoring supports can leverage xAPI data to better understand the learning experience among these experienced learners.
This is only a beginning; xAPI can apply to—and enhance—just about any instructional design model or approach. All instructional design models offer insight into what data about the learning and performance experience we can and should track.
What to do with all this data
Once the L&D team has figured out what data to record to inform the learning experience, the options for actually capturing and sending that data fall into four basic categories:
- Send data from another tool. A small—but growing—number of software tools offer xAPI integrations that do the heavy lifting of coding. These tools offer survey authoring, observation checklists, flash cards, social learning, learning pathway creation, and more, all capturing detailed data on the learners’ activities. When investigating new learning technology, ask whether there is an existing or in-development xAPI integration.
- Write custom code. If this is an opportunity to learn to code, take it! Otherwise, work with a software developer to write some custom code to do what’s needed. Using xAPI is not at all difficult for experienced developers.
- Export and convert data. Particularly when working with an organization’s work systems, IDs may find that the data they need is not set up for xAPI. It might be feasible to extract data in spreadsheet or .csv format and convert it to xAPI statements, which would allow for reporting on learning and performance data in one place.
Of course, sending the data is only half the story. Organizations will need to make decisions about the learning ecosystem in order to receive, store and analyze all this data—and in many cases, this is a much larger conversation, as it involves multiple departments and considerations, and generally more investment, to accomplish. An upcoming article on xAPI ecosystems will include several options.In the meantime, The eLearning Guild offers a number of opportunities to learn more about how xAPI captures rich data, meet xAPI experts, and talk with vendors about their xAPI-ready platforms and tools. DevLearn 2018 Conference & Expo will kick off with the xAPI Camp pre-conference workshop; it will also include an xAPI Central Showcase in the expo with stage programming that will feature vendor demos and real-life case studies.