What Can #EdTech Business Learn from Uber?
In the past several years, the proliferation of learning technologies is exciting, many EdTech products use “freemium” business model to enter classrooms. But this also results in severe data silos issue, not to say workflow integration is left to users to handle. When a new type of EdTech product gets popular, it’s not too hard to copy or mix with other ideas to come out another new product. New products are booming for years, the competition might make many tools more affordable, but this makes data silos even worse. According to an EdSurge report “Schools and Software, What’s Now and What’s Next” (2015):
Many K–12 software programs offer limited value on a stand-alone basis and must be integrated with other software, typically from different vendors, to realize their full potential. Manual integration across various software platforms and programs causes major headaches and inefficiencies for school systems. Automated provisioning of software licenses and single sign-on technologies are slowly addressing some of these pain points. Neither of these, however, helps to integrate data coming out of disparate software systems.
Data Silos Problems
For an instructor, she/he faces a lot of challenges in using new technologies and products.
- As she/he is using “Blended Learning”, “Flipped Classroom”, “Connected Learning”, “Self-Regulated Learning”, “Inquiry-Based Learning” and more pedagogies made possible by digital products, how can he/she monitor the learning process with minimized interruption?
- As she/he is using new technologies or products, how to know learners’ behaviors in using them? Or, does it work for everyone? Any improvement can be done in content design or learning design?
- As she/he use so many free/paid products, how can she/he see a whole picture of how students are doing with those applications? What works? What doesn’t? (it might be different for each learner) The biggest pain will be to put data together. Data are locked in different tools and platforms.
- Also new technologies (Augmented Reality, mobile Apps…) raise challenge of learning experience tracking.
Not only new EdTech products and mobile applications, but also social media platforms and informal learning happening everywhere are making the learning experiences diversified, distributed and, hopefully, personalized. But without hard evidences, hoping personalized learning will happen just because of new products is not responsible. When driving a car, everyone needs a dashboard and a map(or GPS). Why it’s not so in learning? What schools need is a data standard and integration of data flow and workflow across systems. Instructors and learners can have real-time data feedback loops, so that they can make adjustments as soon as possible. (not wasting time pulling data together)
What is xAPI?
XAPI is the latest learning standard from ADL (an R&D team under DoD, also the creator and steward of SCORM). xAPI was born to solve data silos problem. It provides the data structure interoperability of learning records and the freedom of data flow. XAPI can track granular learning process data (behaviors and contextual data) so it can empower learning analytics and enable creative applications developed based on the independent data layer. In the past, many assessment and attendance data are collected and analyzed, of course. But they are not revealing behaviors in the learning process, and the feedback loops are not in real time. And, in order to “probe” the learning, tests are used to evaluate the learners, it’s invasive and time-wasting. (it takes away learning time, and bad scores make learners frustrated)
In the “EdTech Developers’ Guide“(2015) from Office of Educational Technology, US Department of Education, it introduced xAPI as:
When students are learning across multiple technology applications, that information is scattered among many different systems. xAPI is an approach to describe learning data between systems in a uniform way.
xAPI helps collect experiences from online learning activities (such as watching a video or completing a quiz). These data can then be used to help other learning systems give students a more personalized experience, as well as help a teacher understand a student’s unique needs based on his or her activity across multiple learning tools. Because xAPI is simply a protocol to describe learning actions (John watched a video, Kelly completed a quiz, etc.), any type of learning experience can be placed into this format. This means that it could be used to account for more social and emotional skill sets (resilience, perseverance, etc.), which is a potentially big improvement over a reliance on hard outcome data in most digital systems.
In digital learning age, if we fully utilize digital records as data-enriched assessment for learning (for example, John watched video V on platform A from timeline t1 to t2 twice, and then he answered quiz items Q on platform B, taking 2 min. 15 sec. with a hint revealed, and then… all process records put together), the assessment isn’t separated from learning(means no time wasted) and it’s in stealth mode(non-invasive). It can give instructors and learners immediate, continuous and multifaceted feedback. The sample size is all, this means no speculation is involved.
Standardized Profiles, Recipes and Visualizations
XAPI is really about data integration and data analytics, but xAPI spec. only provides data structure interoperability. We’ve worked hard to build semantic interoperability based on the vocabulary best practices from ADL to build AcrossX vocabulary and recipes.
We makes it easy to implement xAPI for EdTech vendors so that they can focus on their own specialties. Just follow our recipes to send xAPI statements according to use cases (activity types), and then the real-time dataviz is available. It can be embedded into user’s current dashboard, or standalone. Only after we standardize basic elements and building blocks, we can develop hierarchical dataviz, higher level analysis and innovative applications.
In order to help users get actionable insights quickly, we also designed visualizations for users to catch patterns efficiently. The new service is called VisCa – means “Visual Catch“, visualization is crucial as a cognitive agent to process volume data. We assume teachers only have 5 minutes before and after classes. It must allow quick pattern recognition of both top view and drilling down, users can zoom in and out easily. Self-describing, structured visualizations make meanings immediately evident to learners, teachers and more. Actually recipe design and visual design are considered at the same time. (learn more)
Learning and performance standard alignment is one of our recipe design consideration, it’s coded in our dataviz too. XAPI is really about learning analytics, so a lot of relevant considerations have been put in our recipe design. We are also connecting with Learning Analytics community to strengthen the foundation.
For an example of integrated data flow and analytics leveraging VisCa, please read:
- A Teacher with AcrossX Solution Enabled by #xAPI
- How #xAPI Enables Analyzing Connected Learning Beyond the LMS
A WIN for Learning Technology Providers Too
VisCa is currently being implemented for about 350,000 users. Now we are looking to work with EdTech providers and developers for integration. 8 steps to grow your learning technology business with VisCa are explained here.
What Can EdTech Business Learn from Uber
In software world there is a saying “API is eating software” in recent years, which has replaced another saying “software is eating the world” in earlier years.(read this widely cited write-up by a Google employee) APIs make partnering up much more efficient, so that each party doesn’t re-invent the wheel some other party has already made. Same or similar products might create competitions without adding any value to the whole ecosystem. But right partnership will serve end users much better, also results in stronger business impact.
According to EdSurge’s report, teachers are willing to exchange some features with better integration! Here is another quote from EdSurge report:
The gaps in the current market represent significant pain points for teachers and administrators and divert attention away from teaching and learning. Until technology providers start cooperating by opening their APIs to integrate across disparate products, proprietary interfaces will force costly integrations on school systems. Until better integrations and coherent data standards facilitate reliable data transfer from academic software programs, the education technology market will face an uphill battle of gaining the trust of educators trying to implement these products to drive student learning.
With an independent xAPI data layer, and standard RESTful HTTP protocol based communication between systems, integrated workflow and data flow across systems are possible.(no need to move your contents into a new system, they can be anywhere) More creative applications out of xAPI data to empower learners and liberate learning designers are limited only by our imagination, we’ll talk about it in next article. If you think of Uber, which is a buzzword recently, it doesn’t create any asset, but only connect the right person with the right car at the right time. And it relies heavily on data and algorithms. The full picture of xAPI ecosystem is alike.
Connect, don’t re-invent the wheel and compete. Let’s talk!