9 Comments

  1. Adaptive learning is a popular topic at the moment. As learners have more access to a broader range of resources and learning experiences this is something that will become increasingly important to help them find the most effective ways to learn.

    It’s odd to set a learning methodology/strategy like adaptive learning in contrast to a data interoperability specification like xAPI though. It’s a bit like saying that we need to move on from speaking English and use mobile phones instead. They aren’t mutually exclusive, and really it’s a nonsense comparison. You can speak English on your mobile phone, and an adaptive learning ecosystem can (and should) leverage xAPI to enable the different components to communicate.

    In fact, in order to support useful adaptive learning, we’re going to need a whole load of data about the learner and their peers from a wide range of sources. We need to know about all the learning that people are doing, and how that learning affects their job performance so that we can make recommendations to other learners that want to improve in the same way. A standard way of communicating between learning products (like xAPI) is a vital piece of technical plumbing in order to make adaptive learning a reality.

    For example. the article says that “An adaptive approach will even allow an organization to take information from other pieces of technology they may be using such as HRIS or CRM systems and incorporate it into the learning architecture.” Certainly an adaptive approach should seek to pull data from whichever systems hold the data needed, including HRIS and CRM systems. But what technologies will be used to get that data?

    Let’s not confuse our overall pedagogical approach with the underlying technologies required to implement the solution our approach has led us to design. Both are vitally important to carefully consider.

    Andrew Downes
    Learning and Interoperability Consultant
    Watershed

    • Andrew: Well said!. You captured my reaction to this article.

    • You make some great points Andrew! I think the author may have been referring to adaptive software architecture not solely an instructional design methodology (although I don’t want to put any words in anyone’s mouth).

      When I attended one of their presentations (at either DevLearn or ATD) I was intrigued by the focus on learning analytics so I jumped on a public webinar. It was really good, quite innovative and definitely not like the more traditional digital learning practices. A lot of food for thought as they say!

      I think xAPI is great too (certainly better than SCORM) but it has its limitations. At the end of the day, right tool for the job based on clients/organizational needs. Data certainly seems to be the direction L&D needs to go though.

    • I didn’t read into this in the same way you did, but perhaps that was due to our terrible experiences with xAPI so far. I think what the author was alluding to was that xAPI has a number of downfalls if you’re trying to use it for complex analytics – which adaptive learning would certainly be if done properly. I would whole-heartedly agree with this. xAPI seems like a reasonable solution for simplistic tracking, however it’s far from useful as a backbone for anything sophisticated.

      I work as a data analyst in a big data team for a Fortune 100 company and first came across xAPI when working with one of our training teams in 2014. They were trying to analyze data based on xAPI statements from more than 20,000 learners. The statements had been implemented by several of the team’s instructional designers. When we looked at the type of analytics they were hoping to accomplish, it was clear they would not achieve it through xAPI. What they were tracking was on the right path, but instead of capturing around 40 different learner events, they really needed to be tracking hundreds. The thing is, given the skill set of the average instructional designer, there was little chance they would have been able to recognize the type of data that needed to be captured. Even if that was overcome, the incredible inefficiencies of xAPI make it a terrible choice for anything other than simplistic tracking. To achieve what the training team really desired with xAPI would have been like trying to hammer a nail into a piece of wood with a teddy bear.

      So while I agree that xAPI is reasonable for basic tracking and as a new evolution of SCORM, it’s far from useful as a tool for those trying to achieve serious and credible business insights or learning measurement. For that, you need to implement solutions designed around big data technologies and methodologies. I find it really concerning that there is such hype around this standard so it’s nice to see this author attempting to point some of these issues out.

      • Thank you for sharing your experience on xAPI. I am an instructional designer and would love to learn more about big data technologies and methodologies that are relevant for L&D. I was wondering whether you could recommend some basic reading materials to start with.

  2. Adaptive learning is a popular topic at the moment. As learners have more access to a broader range of resources and learning experiences this is something that will become increasingly important to help them find the most effective ways to learn.

    It’s odd to set a learning methodology/strategy like adaptive learning in contrast to a data interoperability specification like xAPI though. It’s a bit like saying that we need to move on from speaking English and use mobile phones instead. They aren’t mutually exclusive, and really it’s a nonsense comparison. You can speak English on your mobile phone, and an adaptive learning ecosystem can (and should) leverage xAPI to enable the different components to communicate.

    In fact, in order to support useful adaptive learning, we’re going to need a whole load of data about the learner and their peers from a wide range of sources. We need to know about all the learning that people are doing, and how that learning affects their job performance so that we can make recommendations to other learners that want to improve in the same way. A standard way of communicating between learning products (like xAPI) is a vital piece of technical plumbing in order to make adaptive learning a reality.

    For example. the article says that “An adaptive approach will even allow an organization to take information from other pieces of technology they may be using such as HRIS or CRM systems and incorporate it into the learning architecture.” Certainly an adaptive approach should seek to pull data from whichever systems hold the data needed, including HRIS and CRM systems. But what technologies will be used to get that data?

    Let’s not confuse our overall pedagogical approach with the underlying technologies required to implement the solution our approach has led us to design. Both are vitally important to carefully consider.

    Andrew Downes
    Learning and Interoperability Consultant
    Watershed

    • I didn’t read into this in the same way you did, but perhaps that was due to our terrible experiences with xAPI so far. I think what the author was alluding to was that xAPI has a number of downfalls if you’re trying to use it for complex analytics – which adaptive learning would certainly be if done properly. I would whole-heartedly agree with this. It seems like a reasonable solution for simplistic tracking, however it’s far from useful as a backbone for anything sophisticated.

      I work as a data analyst in a big data team for a Fortune 100 company and first came across xAPI when working with one of our training teams in 2014. They were trying to analyze data based on xAPI statements from more than 20,000 learners. The statements had been implemented by several of the team’s instructional designers. When we looked at the type of analytics they were hoping to accomplish, it was clear they would not achieve through xAPI. What they were tracking was on the right path, but instead of capturing around 40 different learner events, they really needed to be tracking hundreds. The thing is, given the skill set of the average instructional designer, there was little chance they would have been able to recognize the type of data that needed to be captured. Even if that was overcome, the incredible inefficiencies of xAPI make it a terrible choice for anything other than simplistic tracking. To achieve what the training team really desired with xAPI would have been like trying to hammer a nail into a piece of wood with a teddy bear.

      So while I agree that xAPI is reasonable for basic tracking and as a new evolution of SCORM, it’s far from useful as a tool for those trying to achieve serious and credible business insights or learning measurement. For that, you need to implement solutions designed around big data technologies and methodologies. I find it really concerning that there is such hype around this standard so it’s nice to see this author attempting to point some of these issues out.

      • Thank you for sharing your experience on xAPI. I am an instructional designer and would love to learn more about big data technologies and methodologies that are relevant for L&D. I was wondering whether you could recommend some basic reading materials to start with.


Add a Comment

Your email address will not be published. Required fields are marked *

Comment *
Name *
Email *
Website