Growing xAPI to Scale

by Aaron E. Silvers (President & Executive Director, The Data Interoperability Standards Consortium)

img_5104A week ago, through our Connections Forum, Megan and I saw our 9th xAPI Camp come to a close. This one was the first time we assembled presenters with case studies focused on a particular domain: Healthcare. Of the five presentations, there were three case studies of xAPI’s use related to resuscitation, preventative care or recovery.

Resuscitation. xAPI is being employed in a hospital to improve Code Blue responses. Consider that xAPI is literally being used to save lives. This is the original goal for xAPI and all of the Total Learning Architecture[1]. Watching Dave Bauer of SiTEL/MedStar present his case study, it struck me how remarkable an organization the Advanced Distributed Learning (ADL) Initiative is. In 2002, ADL released Version 1.2 of the Sharable Content Object Reference Model (SCORM) and it unified a hundred different ways people developed and distributed eLearning. In 2013, ADL released Version 1.0 of the Experience API.

In 2002 the feedback loops to improve eLearning such that it might influence performance outcomes were impossibly difficult to achieve. In 2016, xAPI is actually making that happen in visible ways. It bears repeating: we’re seeing xAPI employed to save lives. Only ADL could maintain the vision that brought people together to support this open source project this far. ADL is readying for other things in the Total Learning Architecture that need to make use of an industrial-strength xAPI. For that to happen, given their capacity, the xAPI community must now do its part to shoulder the responsibility of xAPI’s growth beyond ADL, and with thanks to ADL, it is doing so.

In this article, I’m going to talk to how xAPI is going to grow at scale: the market incentives that lay in wait to bring vendors and stakeholders along, the supporting efforts being undertaken by ADL and other organizations to make its implementation consistent, reliable, secure and easier, and the advancements to come to support professionals be consistent, reliable, ethical and secure in their design, development and analysis with xAPI.

Realizing xAPI’s Potential

In 2015 I was excited to see case studies of at least a dozen organizations leading their own investigations within, using xAPI. Since the beginning of 2016, there are easily 3x that many easily findable online. In Asia, Australia and Europe, ministries of education in multiple nations are developing common data infrastructure projects based on xAPI.

We are all now of a time when xAPI is no longer an applied research and development project. ADL has fostered an incredible open source success, such that xAPI’s adoption is now sufficient enough to maintain a small, but very sustainable growth market. With a few key structures in place to make it easier and practical to work with xAPI, adoption will grow exponentially.

Thus, we can now imagine the industry to emerge around xAPI. Much of what’s to come around xAPI intersects with a greater industry built around data standards, though xAPI won’t be completely enveloped by it. There are a few things that set xAPI apart as a de jure standard.

  • xAPI has a strong and active open source community.
  • xAPI is licensed as Apache 2.0.
  • xAPI offers both human- and machine-readable data,
  • xAPI is rooted in an established, well-researched pedagogy.
  • xAPI is intended, from its origins, to support data sovereignty.
  • xAPI is sponsored by the US Department of Defense.

The combination of these things makes xAPI very compelling to a number of organizations, in their own right, because those qualities translate to adopters to mean:

  • xAPI is supported.
  • xAPI is licensed favorably and flexibly.
  • xAPI is fairly easy to use.
  • xAPI is transparent in its biases.
  • xAPI is respectful.
  • xAPI is dependable.

It’s for these reasons that in the coming months, a US Department of Defense Instruction (DoDI) that mandated the use of SCORM[2] is likely to be revised to include xAPI[3]. When this Instruction was originally issued in 2006, it began a domino-effect for SCORM’s adoption. With the purchasing power of the US Government, every eLearning tool on the market, be it an LMS or an authoring tool, fully implemented SCORM within a year. With SCORM’s sudden ubiquity, every organization interested in purchasing an LMS or an authoring tool was going to be adopting SCORM. The same, I expect, will hold true for xAPI.

For xAPI to scale in its consistency, applicability and ubiquity, supporting structures are needed. I’m very happy to report that such supports are already underway.

Making an Industrial Strength xAPI

In November 2015, Megan and I established the Data Interoperability Standards Consortium (DISC) with friends, board members, and collaborators Kirsty Kitto, Eric Nehrlich, Brenda Sanderson and Robert Todd. The communication and evangelism we’ve been doing through Connections Forum has been wonderful for growing the energy around the spec and to celebrate works with xAPI that are done well. Like the xAPI Quarterly, having credible information about what’s happening with xAPI is important stuff

For xAPI to really scale in its adoption, more is required than just publishing and events.

There are some pretty basic things that are needed for xAPI to be easier for organizations to adopt. Right now, anyone can just say they use xAPI — there’s no easy way to demonstrate or prove it otherwise. Software that tests for conformance to the xAPI specification is a start, and in July 2016, ADL contracted with DISC to vet conformance requirements with the xAPI community as part of a mandate to ready xAPI’s adoption at this magnitude. The US Government and military, much like every corporation, buys learning technology and they want the cost of commercially available, off-the-shelf software to be reduced as well as the costs of software integrations, maintenance. Meanwhile, they want the quality of commercial, off-the-shelf software to continuously improve. To this end, DISC delivered to ADL xAPI’s Conformance Requirements Document for Learning Record Stores. From these requirements, vetted by the xAPI community over the past two months, ADL will develop a software conformance test. With every LRS conforming to the xAPI specification, and evaluating their conformance against the same test suite, vendors will be able to reliably approach data interoperability in explicit, transparent and measurable ways.

Software conformance testing is a start; having an organization that can reliably certify the interoperability of their software using xAPI is better. To that end, DISC is establishing software certification programs in 2017.

Software certification takes the guesswork for professionals involved in IT as well as acquisitions. Having a reliable third-party rigorously test and, when performance merits it, certifies xAPI software products — that certification makes it easy to identify software that works as it’s supposed to. Organizations can and should test software themselves, but the certification makes it easy to sift through all the noise vendors can make about xAPI and look closely at the products that have demonstrated their reliable, interoperable performance.

Software certification makes it possible for xAPI to scale and accelerate its adoption by (and while) ensuring the quality of data interoperability improves in actual and transparent ways. If I’m going to trust my data, I need to trust the systems that make it and manage it. That’s why, just like xAPI’s specification conformance requirements are going to be open source — DISC’s xAPI certification requirements are going to be open, too — likely, licensed Apache 2.0. The end result will be completely wide open for all to see, cite and use.

The process by which DISC’s certification requirements are created will be a different process than how xAPI’s specification development, and its conformance requirements, have been defined. Because software certification affects an industry, and only those who really are invested in such industry need to be involved — the stakeholders who are adopting or intend to adopt xAPI tools and the vendors who make them.

The Right Stuff

Who among us in adult learning, education and/or training have really held ourselves accountable for continued improvement of our craft? By what common measures have we done that? How often have we demanded such accountability for ourselves. I argue that as professionals, we haven’t demanded such accountability because we couldn’t imagine how it’d be possible. We live in an age where credible data to use in our own development is possible.

People who work with xAPI should have professional certification.

Credentialing is needed to elevate the work to be done with data to a professional class. In 2015, the Association for Talent Development (ATD) released the analysis of research conducted with 1,100 Instructional Designers. 40% of the Instructional Designers in the report expressed concern about the lack of skilled staff in their organizations; 29% cited “difficulty keeping pace with new developments in learning technologies/media.”[4] ATD is recognized by the US Department of Labor as a trade organization, and thus can award someone a certification designating her/him a Certified Professional in Learning and Performance (CPLP). The program, however, covers “the entire talent development profession” — meaning a range of professionals. There are many different roles within talent development, and the areas addressed by xAPI, while inclusive of many roles in talent development, span beyond it.

The Project Management Institute’s Project Management Professional (PMP) certification is a more apt model for what xAPI needs. While a Talent Development professional may include project managers, classroom trainers or instructional designers, it is quite specific to identify a project manager role, and the expectations shared across organizations and contexts of professional in such a role. Merideth Levinson cites a 2006 survey by PwC[5]

higher-performing projects are significantly more likely to be staffed with certified project managers and 80 percent of projects classified as high-performing use a certified project manager.[6]

The stakes for being able to provide and make appropriate use of credible data could not be higher for our personal health decisions, our management of risk within our organizations, and our decision-making as communities and, dare I say, nations.

Gallup cites three factors in effective data-driven risk management, including

  • Real-time results
  • Local data
  • Predictive analytics[7]

xAPI can easily provide a common means to aggregate local data into real-time results that informs predictive analytics. Technology alone cannot mitigate risk, however. Without credible professionals entrusted to do the work, to ask the questions that need answers and to shape data into actionable insights, whether as individuals or collectives or organizations, we give up our agency and our accountability.

This is why DISC will establish one or more professional programs related to working with data, focusing at first on xAPI, starting in 2017.

Hard to scale? Yes. Impossible? No.

This seems too hard? Right? It sounds damn near impossible. Maybe for a lot of people it is, but the stakes could not be higher where data and trust are concerned — especially when it comes to people around the world whose work and lives are going to be more and more impacted by human and machine decisions, made with data.

The people supporting DISC are extraordinary people — professionals who over decades have dedicated their work and lives to the betterment of others. The xAPI Community, its friends and advocates, are extraordinary people living in extraordinary times. Truth and fact are under attack. We need systems for data analysis that we can trust. We need people working with data we can trust to not only know, technically, how to use the tools before them but also have the ethics to make good decisions about and with it. We need shared tools and services that make it easier to make those good decisions and carry them out.

On October 17 at 1:30pm Eastern, an informational webinar will kickoff a research effort that will inform DISC’s software certification program(s). Throughout October, November and December of this year, there will be a few meetings online and dozens of interviews conducted, ultimately informing the design of an xAPI software certification program.

With help of stakeholders already adopting xAPI and those who intend to do so, we can do something special with software certification and professional certification that the learning industry has never seen – quality, consistency, reliability and trust defined by both engineers and business stakeholders. There is only one organization that is stepping up to do that in a way that is both usable and useful, and that is DISC. That’s why it exists. DISC is a vehicle to take us to a better informed and educated tomorrow for everyone.


[1] Advanced Distributed Learning. “Total Learning Architecture.” Advanced Distributed Learning. Accessed September 28, 2016. https://www.adlnet.gov/tla/.

[2] US Department of Defense. “Instruction 1322.26: Development, Management, and Delivery of Distributed Learning.” Defense Technical Information Center. June 16, 2006. Accessed September 28, 2016. http://www.dtic.mil/whs/directives/corres/pdf/132226p.pdf.

[3] xAPI Conformance Requirements and Testing Strategy Kickoff Meeting. Jason Haag and Aaron Silvers. YouTube. August 12, 2016. Accessed September 28, 2016. https://www.youtube.com/watch?v=ATKBrgFZEac. (02:45/52:05)

[4] ATD Research. “ATD Research Presents — Instructional Design Now: A New Age of Learning and Beyond.” Association for Talent Development. March 2015. Accessed September 28, 2016. https://www.td.org/Publications/Research-Reports/2015/Instructional-Design-Now.

[5] Levinson, Meridith. “Why Project Management Certifications Matter.” CIO. January 20, 2010. Accessed September 28, 2016. http://www.cio.com/article/2421276/project-management/why-project-management-certifications-matter.html.

[6] PwC. “Insights and Trends: Current Programme and Project Management Practices.” 2014. Accessed September 28, 2016. doi:10.1002/9781118835531.

[7] Gallup, Inc. “Managing Employee Risk Demands Data, Not Guesswork.” Gallup.com. March 16, 2016. Accessed September 28, 2016. http://www.gallup.com/businessjournal/189878/managing-employee-risk-demands-data-not-guesswork.aspx.


Aaron E. Silvers (President & Executive Director, The Data Interoperability Standards Consortium)

Aaron E. Silvers copyAaron E. Silvers (@aaronesilvers) helped learning technologies like SCORM and the Experience API (xAPI) find massive adoption. Through MakingBetter, Aaron helps companies implement content, data and reporting strategies that yield powerful insights to many levels of stakeholders. He is the co-founder and co-organizer of the seminal community, “Up to All of Us.” Aaron co-founded the Data Interoperability Standards Consortium (DISC) with Megan Bowe – a consortium to steer xAPI’s growth and evolution through establishing the requirements for certifying software and professionals working with xAPI.

 

Speak Your Mind

*