The xAPI Barcamp at the end of the first day of the Learning Technologies conference attracted around fifty people, eager to talk xAPI over a few free drinks at the local pub! I was one of five invited experts alongside Andrew Downes from Rustici (@mrdownes), Mark Berthelemy from Wyver Solutions(@berthelemy), Ben Betts from Learning Locker (@bbetts) and Jonathan Archibald from Tesello (@jonarchibald). Moving around five tables in turn, each expert began by talking for a few minutes about what they were doing with xAPI, then the table held an open discussion.
— Aaron E. Silvers (@aaronesilvers) January 30, 2015
//platform.twitter.com/widgets.jsI found the event fascinating. Having worked on a few xAPI projects for clients I had some solid work to discuss, however I personally still have more questions than answers about xAPI so this event was the perfect forum to pose some of those questions and find out what other practitioners were doing and thinking.
Xapi and Moodle
This was my area of expertise and why i was invited, so no surprises here! I started by taking about what Leo have done with Moodle and xapi. We took an existing xAPI content launch plugin originally published by Andrew Downes as a prototype, and made it production-ready. We also built a new plugin that converted Moodle quiz results to xAPI statements. We pushed these statements out from the customer’s Moodle LMS to an external Learning Record Store (LRS) that the customer had built themselves. The hotel group we built this for also had their CRM vendor do likewise, so were able to compare learning data and sales data in the LRS reporting interface. Our work is available as open source for others to extend, of you do so please contribute back for everyone’s benefit.
Breaking out of L&D
The xAPI vision of tracking learning undertaken anywhere requires tools outside of the traditional L&D sector to start pushing out xAPI statements about user experiences. Khan Academy, YouTube, library management systems, classroom event systems, conference booking systems and more need to get on board if the vision is to be realised without being over-reliant on self reporting which only really appeals to the most self-motivated learners. All these vendors would need to adopt the standard and work out how to do authentication with the LRS against a unique personal identifier. No one on any table had seen evidence of this yet, but agreed that to achieve its vision, xAPI must break out of the L&D sector.
Ethics and privacy
There was a good deal of discussion over what data to track and the implications of doing so. We live in a world where it is becoming the norm for software service providers to track the living daylights out of everything, and there is a danger of xAPI being caught up in that. However, this has already started a backlash from users, with the increasing popularity of ephemeral web applications. There was a concern about the ethics of data collection and how user privacy can be supported by xAPI, but also concern about exporting organisational intellectual property via xapi interactions with external LRS tools. If you are tracking workplace learning, do you own the tracking data or does your employer? Are there circumstances where employment related data should remain in the workplace LRS and not be exportable to an external, personal LRS? And in the global workplace, how do we handle national laws like in Germany, which has strict personal privacy laws. For example, on a recent project at LEO for a German client, our technical solution had to be approved by workers unions due to personal data collection needs for a career planning tool. You can’t just track everything and expect no repercussions, and as early adopters these are important issues we need to grapple with to ensure we are working in an ethical and legal manner.
Starting the day with talk of takeaways from yesterday – for me – the xAPI fringe event was a definite highlight #LT15uk
— Nick Lee (@N1ckL33) January 29, 2015
This was a fascinating event and I hope the attendees learnt as much from it as I did, from the followup tweets it sounds like they did. Many thanks to Ben Betts and Martin Couzins for organising. Am really looking forward to the next one!