Your Source for Learning
Technology, Strategy, and News
ARTICLES       RSS feed RSS feed

Share:

email print

eLearning Authoring: Taking the Next Step with xAPI

Feature

by Craig Wiggins, Peter Berking, Andy Johnson, Steve Foreman

August 17, 2015

“Enhanced xAPI support is likely to manifest itself in the true spirit of authoring tools, so that you, as author, will not have to know much about the technical details of how xAPI works. You will be able to create more complex xAPI- enabled instructional designs faster and easier—you may not even realize you are using xAPI.”

Authoring tools enable instructional designers (ISDs) without programming skills to create eLearning products. ISDs use these tools to create and sequence pages, enter content, add media and interactivity, and assemble courses.

The authoring tool marketplace comprises a wide variety of constantly evolving tools. Responsive design, gamification, avatars, animation, social networking, and many other capabilities are available. Yet many of these tools have not done much with the Experience API (xAPI). Some tools simply use built-in xAPI statements to replicate SCORM’s tracking mechanisms and give you the choice to export your course to xAPI instead of SCORM. This approach tracks page visits, module completions, and test scores, but it does not tap into the full power of xAPI.

In this article, we will discuss a new vision for how authoring tools can support xAPI, enabling ISDs to create more effective and engaging online courses.

Some typical scenarios

Let’s start with some scenarios.

Scenario one: A stockroom course

Stockroom employees must be able to retrieve items from the stock room quickly and accurately. A new-hire classroom course covers stockroom procedures, equipment, and safety. But after training, employees are on their own, spending the first two months on the job gradually discovering how to receive, organize, and store stock in ways that make it easier to retrieve.

Using xAPI and a handheld package-scanning device, it is possible to track package check-in times and entries in the stockroom computer system. Comprehensive collected data provides information about the accuracy and speed of each employee's performance. This data becomes the basis for just-in time performance support, personalized for each employee, including a competitive workplace game that motivates employees and rewards rapid and accurate job performance with a spot on the leaderboard.

Scenario two: A first-aid course

An organization determines that it has a need to teach non-medical emergency responders how to administer first aid to someone suffering from a life-threatening asthma attack. Identifying the problem accurately, performing the right steps in the right order, and working fast are all critical to saving a life. The ISD would like to deliver this training through a series of online simulations, but typical SCORM-based authoring tools do not provide adequate tracking capabilities.

The ISD creates a series of virtual-patient simulations and uses xAPI to measure speed and accuracy of aid administered. If the user hesitates, takes incorrect action, or does things in the wrong order, the simulated patient may not survive. After each simulation, the system can provide detailed feedback to help the user improve performance.

Scenario three: a hazmat course

Novice welders must learn how to secure and store gas tanks, mix gases, and light and extinguish the welding torch safely. In the past, welders have taken this training one-on-one through an apprenticeship with an expert. However, the organization wants to minimize the amount of time experts must spend training novices.

With the Experience API, the ISD can track the use of performance support that incorporates video demonstrations, procedural walk-throughs, and just-in-time coaching. This allows the novice to train on the safe handling of tanks and receive feedback without the need for an ever-present expert. Once the novice has mastered the process, an expert is called in to observe the end-to-end process and sign off on the novice’s competence.

The future of authoring tools and xAPI

In this section, we share our thoughts on how authoring tools can provide more robust support for xAPI in the future.

Viewpoint: Andy Johnson

Future xAPI tools will:

  • Ask me questions: Think TurboTax, asking basic questions about the last tax year. This would set the stage for automated rules throughout the creation process. Rather than writing statements for each “object,” we’d only have to create statements and code requests for exceptions.
  • Take advantage of online vocabularies: Future tools would allow users of that tool to self-identify with online domains such as medical, military, K-12, etc. The tool would then import verbs, activities, and results based on that domain. A tool’s ability to pull in data and dynamically change is going to be crucial.
  • Have an intuitive and dynamic UI: Building on the integration of vocabularies, the xAPI component of the authoring tool UI would be customized based on the subject matter and intended audience. While there would be some deviation from the “standard” way to use the tool, the iconography would be much more familiar to the user based on their profession or field or study, making the tool more intuitive.
  • Integrate self-reporting: A tool of the future wouldn’t only generate a course. It would produce a log of the author's experience while creating the course. This log would give the author ideas on how to improve in using the tool or, if the author is an expert, something to publish to novice authors.

Viewpoint: Craig Wiggins

The future of authoring tools in light of xAPI is one that:

  • Moves beyond the idea of a massive, all-encompassing authoring tool for learning experience realization. Modern authoring tools are more or less built with the same goal in mind: courses with quizzes, markedly limited interaction—the antithesis of multimodal. Authoring tools of the future will also focus on specific experiences, but those tools will be more like what we today think of as smartphone applications or plugins: contained and of limited means—but with specific effect.
  • Focuses on tools made for a specific context or use case. A focused tool is made for a specific use case or context—e.g., a tool to design realistic coaching or scenario dialogue (with realistic feedback and data inputs) or one that allows a designer to make video footage and create interactive video experiences that track all sorts of user activities. This doesn’t necessarily mean that bundles of context-bound tools couldn’t be made into a single offering (think video editing tools plus a myriad of plugins).
    • (Another possibility is the equivalent of industry or context-specific downloadable content, much like those for video games: function profiles that you download for a tool when you wish to create experiences for a specific context or activity. These function profiles would include things like xAPI domain profiles, including vocabulary sets, tracking presets, context-specific content, etc.)
    • Seeks to guide the design process. The ideal authoring tool might be a design guide as much as a tool equipped to execute design imperatives. This can range from checklists of things to consider when addressing “X” to full-on option-recommender engines.
    • Addresses new problems. Working with xAPI, your learning experience design can spread in any direction. These new options come with their own new issues. For example, figuring out where all of your learning experience elements are located (across servers and times) and connecting them can be very difficult. Abstracting this process and making it easier to manage would be valuable.
    • Sufficiently abstracts the user from having to know how the sausage is made. The success of the major authoring tool vendors comes not simply from their feature sets, but rather from creating a sufficient abstraction from having to know the underlying workings (SCORM, AICC, manifests, XML, etc.) and allowing the designer to focus on creating the experience that they think is necessary for performance improvement to occur. (Again, in most of these cases this is traditional eLearning.) While xAPI does require that learning designers know more than they historically have in order to reap the benefits of the specification (e.g., fundamentals of data science, familiarity with learning analytics, the anatomy of xAPI statements, the basic care and feeding of xAPI-enabled systems, etc.), there is no reason to believe that xAPI-enabled tools cannot one day soon be simple enough to use that one will not have to have knowledge of JSON, XML, linked data, etc.

Viewpoint: Peter Berking

Authoring tools five years from now will have the following xAPI-enabled characteristics:

  • They will focus on performance-based assessment rather than multiple-choice testing. This includes stealth assessment, where the learner is constantly being compared behind the scenes (possibly without the learner being aware of it) to a “gold standard” of interaction with the content, based on the path and responses an expert takes through the content. Rather than a simple test score result, a much more nuanced profile of the learner’s ability will be available through use of xAPI statements and data analysis engines. Test scores will be replaced by probabilistic statements such as “Mary performed at 83 percent of the level of an expert.”
  • Adaptive, personalized content will be standard. xAPI will be the key to create a feedback loop that captures and analyzes the user’s performance. This allows the content to determine optimal topics, media objects, learning paths, etc. for the learner based on their past performance.
  • Authoring tools will achieve bi-directional communication between the content and the LRS. This means that a designer can insert a dashboard into the authored content that pulls data from the LRS about who is doing what with the content and how well, so that the learner’s interactions are “socialized” with other learners. Instructors or moderators, etc. can then monitor learner progress and intervene appropriately.
  • Media objects (graphics, video, text layouts, etc.) and interactive functionality objects (widgets, scripts) will provide standard xAPI data communications capabilities within the scope of their functions. For instance, videos will contain the code to be individually configurable to report durations and locations of pauses via xAPI. The authoring tool (providing the overarching “container” for these objects) will then overlay the xAPI-enabled sequence, tracking, and communicative aspects to interconnect and manage them as a single learning experience and allow it to communicate with external systems
  • Authoring tools will focus not only on learner performance tracking, but iterative, continuous content improvement through xAPI-enabled capture and aggregation of usage patterns. In fact, one of the key purposes of the authoring tool will be to author, integrate, and automate formative and summative evaluation of the learning experience.

Viewpoint: Steve Foreman

In the near future (I hope):

  • Authoring tools will provide more robust support for xAPI. One way this could work is by enabling authors to define any interactive object as an xAPI trigger. Once an object is defined as a trigger, the author would use a natural language xAPI editor to select one or more predefined statements to associate with the trigger. The author would use the selected statement or customize it. Advanced authors would be able to create xAPI statements from scratch.

    • Behind the scenes, the tool would automatically convert the natural language statement into JSON format as required by xAPI. Authors would be able to toggle to a JSON view of the statement just like current tools allow authors to toggle between text and HTML in a WYSIWYG text editor.

    • The predefined statements would be intrinsic to the object defined as the trigger. For example, answer options for a multiple choice question could be defined as triggers. Predefined statements might include “<user> selected <answer text> to <question text>” and “<user> answered <question text> in <x> seconds.” The values in a dropdown list might have similar predefined statements. A drag-and-drop object might have predefined statements such as “<user> dragged <object label> to <placement label>.”

    • You will be able to track any user action in a lesson, assessment, game, simulation, eBook, or app. How might this new, fully configurable tracking capability affect your instructional designs?

Conclusion

In a few short years, we expect that authoring-tool xAPI functionality will grow significantly. In this article we have described a few ways this might happen. Enhanced xAPI support is likely to manifest itself in the true spirit of authoring tools, so that you, as author, will not have to know much about the technical details of how xAPI works. You will be able to create more complex xAPI-enabled instructional designs faster and easier. You may not even realize you are using xAPI. You will simply use xAPI-enabled authoring tools to design and build more effective and engaging learning experiences, free of technical distractions and complexities.

Want more?

In addition to the xAPI Camp pre-conference certificate workshop on Tuesday, September 29, concurrent sessions addressing the uses of xAPI (including ones by the authors of this article) are featured on the program at The eLearning Guild’s DevLearn 2015 Conference & Expo, September 30 – October 2 in Las Vegas. Register today!

Topics Covered

(23)
Appreciate this!
Comments

Login or subscribe to comment

This is heady stuff. I'm very glad that technically competent developers are thinking ahead. Couple of things: Good writers always spell out the term the first time they use acronyms. I really do hope developers work with instructional designers when they build these programs - users can always tell when things are built by techies alone.
Performance based assessment is KEY! I so agree with Peter Berking on his point. Some are there. Storyline 2 includes this but the LMS system we use does not allow xAPI as of yet. We have found a company that does an assessment within an application but are looking for others who can do the same function for comparison.
The article starts with "Authoring tools enable instructional designers (ISDs) without programming skills to create eLearning products." As someone who essentially knows nothing about programming, I'm curious to know how eLearning is built using programming? Up until recently, I assumed that all eLearning was built using authoring tools. I imagine that, by using programming, you can create much better eLearning since you're not limited to what the authoring tool can do. Can anyone explain this to me?

Thanks.
klwilcoxon: Kevin, what acronym did I miss when I edited this? If you mean SCORM, you can mouse over it or tap it (on a mobile device) to get it spelled out, since it's in our glossary. If you mean xAPI in the title, I seldom spell out acronyms in the titles (creates more problems than it solves). If it was something else, let me know and I will fix it. Not looking to confuse or bewilder anybody. -- Bill the (apparently) Inconsistent Editor.
Navedahm20: In response to your question, instructional games and simulations are sometimes developed using Java, C++, Visual Basic or other programming tools. Frequently, eLearning programs are created using web development tools such as HTML, HTML5 or Flash, XML, CSS and JavaScript. This is especially true in some of the more advanced vendor shops. Higher level authoring tools provide a friendly developer environment, hiding the technical complexity while often rendering HTML, HTML5 or Flash, XML, CSS and JavaScript behind the scenes as final output.
I like these ideas, in fact, in our Taipei city project, we are now developing an authoring tool for learning planning with hierarchy. And how to design xAPI statements when authoring is actually deciding what/how to report/analyze later. This is discussed in this post: http://classroom-aid.com/2015/01/20/xapi-profile-implementation-and-reporting-design/

While all these are possible, there shall be a crucial factor coming before they are possible to happen -- the community effort to build/agree with a consistent recipe/profile to describe some specific activity. -- Jessie
I'm very excited to read this because I can reveal that the future you envisage is actually here today :)
Our authoring technology called Conducttr was designed from the outset to create multiplatform experiences, track learner behaviour and personalise experiences. Without any programming knowledge anyone can create interactive experiences for teams or single learners and in real time read individual activity feeds such as "Robert's score increased from 5 to 10" or "Robert subscribed to character Janice using email".
All the activity data can be securely exported and mined; learner data can be exported with personally identifiable information removed for additional privacy.
Anyone with some programming skills can integrate Conducttr with existing systems or platforms.
We'll be at the DevLearn conference booth 146 or drop me a line now
Nice info!
Related Articles