Select Page

3 Things You Need To Know About xAPI Data (Before You Get Started)

by | Mar 15, 2017

In March, 2014, I was at a small un-conference with a group of learning professionals; it was a group of creative innovators, many of whom were early adopters of xAPI.  During one afternoon, I was lucky enough to stumble into a conversation that is still affecting my work today…  

We were talking about what was needed to support the adoption of xAPI; not what was needed technically but what was needed to get value from the data.   

Among others in that conversation were Ben Betts, who just a year later became my boss (when I joined HT2 Labs as Data Scientist), and Sean Putnam, with whom I’ve since had the opportunity to collaborate on a number of xAPI projects.

Collectively, we agreed that for xAPI to realise it’s potential, L&D professionals need to design for data, and design from data; we need to make sure we’re getting the data we need to do meaningful analysis, and we need to learn from our data and use it to inform future design improvements.

To do that, L&D professionals don’t need to be developers or statisticians, but there are a few things we need to know.

3 Thing You Need To Know About xAPI Data Before You Get Started

1. Know the Spec (at least a little)

To work with xAPI you do need to know a few things about the spec; things that go beyond the basic Actor → Verb → Object format.  

Knowing the basic vocabulary, and the format of xAPI statements allows you to design for data collection that will provide the maximum value when it comes time to do analysis.   

Understanding how data is collected in xAPI helps to select meaningful verbs, and include contextual data to your statements; it also helps you understand how to balance the potential value of data against the potential effort involved involved, particularly when one needs to work with both xAPI and non-xAPI enabled data sources (a common situation in L&D).

2. Know Your Customers

We collect better data when we know how it’s going to be used; and there can be a whole host of data customers to satisfy, from end users to stakeholders, to instructional designers.  Taking the time to find out what each group’s goals are  helps to set priorities for data collection.  

And if we’ve already got a decent handle on the xAPI spec, we can let our data customers know what is and is not possible, as well as helping them set priorities based on the level of effort involved in collecting various data.    

3. Know Your Data

If we have a foundational understanding of xAPI and we’ve built an understanding of our data goals, we’re on our way to some data  content strategy.  

Getting to really know our data, even before we’ve collected it can be time consuming, but it pays major dividends when it comes to having the data we need for analysis.  

Data content strategy for L&D exists at the intersection of the technical elements (“What data can we get from our systems?”),  instructional design (“What activities can provide measurable data?”), and analysis for end users (“What data can we, and should we, take action on?”).

After we’ve collected our data, we get to know it through analysis.  This doesn’t have to be a massive exercise in statistics; even a very basic exploration of data can yield useful insights or point to questions that merit pursuit.  

We may also be tasked with creating some dashboards. If we did our data content strategy well, that will play a key role in selecting dashboard elements that are useful and relevant for decision makers.  A careful exploration and evaluation of our data will tell us if the measurements on the dashboard are actually valid indicators of the performance we’re trying to represent.

None of this happens perfectly on a first pass; it’s an iterative process where each prior version can inform the improvements for the next.  But taking the time to do things well on the first iteration gives a strong foundation to build off of as you  begin to put your learning data to work.

Getting Value From Your xAPI Data

Things have come a long way with xAPI, since that un-conference in 2014.   In those years, I’ve seen amazingly innovative Learning Locker users demonstrate real business value from their learning initiatives.  

I’ve also had the chance to co-author a book (with Sean) about some of the things we’ve learned along the way.  Investigating Performance: Design & Outcomes with xAPI looks at how data can take a broad role in course design, along with the potential benefits of understanding and supporting user performance – and is out soon.

If you’re interested in finding out more about the the xAPI and Learning Analytics, download our free Learning Technology Manager’s Guide to xAPI.

About the Author

Janet Laane-Effron

Janet Laane-Effron

Data Scientist

Janet is our in-house data scientist. She has a penchant for applying diverse fields such as cognitive science and competitive intelligence to develop best practices in learning design & developing meaningful metrics for performance analysis. Basically, if you want some clever insight into your organisation’s learning, Janet’s your go-to person! Although not based in the UK, Janet makes up for it with some virtual cycling around rainy London each morning.

View all posts by Janet

Related Reading

shiftnav-red-panda

solutions-menu-mobile

about-menu-mobile

red-panda-menu

curatr-menu

learning-locker-menu

blog-menu

contact-us-menu

home-menu

Pin It on Pinterest

Thanks for reading!

Help us spread the love and share this post with your network