If you’re walking through the hallowed halls of power and have your ears low to the ground, you might just hear people whispering the mysterious phrase ‘Tin Can API.’ Alright, maybe not in Australian Parliament or the White House, but perhaps in San Francisco’s Silicon Valley, where similar mysterious expressions — think ‘bitcoin’ or ‘virtual reality’ or even ‘Facebook’ to name but a few — turned relatively unknown tech-heads into multimillionaires overnight.
But let’s put the profit margins to one side and lead with an educational positive. And here it is: Tin Can API, or xAPI as it is more formally known, promises to capture big data about human performance and will fundamentally overhaul how education is delivered in the 21st century.
And here’s why.
I. What is Tin Can API — A Layman’s Perspective
For those of you technologically inclined, Tin Can API is a new specification for learning management systems designed to capture data about all of a student’s interactions — both online and offline — within the learning course. This includes interactions on a student’s desktop, tablet or smart phone. Hypothetically, rather than a student’s grade being calculated in the conventional manner that most of us are used to — for example, formal and summary tests and assignments — Tin Can API could record every single interaction the student has with the learning management system in order to calculate a student’s knowledge. Not just tests and assignments! Rather: simulations, virtual worlds, mobile learning, experiential learning, even games.
Those interactions will produce some serious data, and in the education sector, data is King. Specifically, so called “big data” — those extremely large data sets that are capable of revealing patterns trends, associations and relationships relating to human beings and how they interact with the world around them. Essentially, how human beings think. Pioneering marketing companies are already analysing ‘big data’ to produce targeted campaigns that can zero-in and engage with specific demographics with advertising messages.
Think: How long did your student spend reading your course material? Did your student encounter difficulty with the administrative structure of the course? What was the exact order of different parts of the course information opened by your student? Literally every click the student makes will be measurable, as will the amount of time they spend reading a page, watching a video or completing a quiz. The data obtained is going to be big.
If your brain is hurting, perhaps an analogy can help. Imagine a world where human evolution occurred in a deliberate way, rather than the hit-and-miss nature of random mutation as postulated by Charles Darwin. If the human race could evolve based on identifying every recorded flaw in their genetic structure and subsequently eliminate that flaw in each successive generation, then it would attain genetic and physiological results that current medicine can only dream of.
Education, for all of its innovations, has thus far progressed along the same haphazard route as our own evolution. Even with modern computer and online learning modules, we cannot pinpoint exactly which strategies work, and which don’t, in training top-class students. It’s a problem compounded by classroom sizes, by economics and geography, by language barriers, and by differently-skilled trainers operating in fluid environments.
Luckily, these arehurdles that educators and computer scientists have been taking seriously for quite some time.
II. History Check
In the late 1990s, the United States Department of Defense established the Advanced Distributed Learning initiative (or ADL, for short). One of the body’s chief aims was to engender cultural change to move from ‘classroom-centric’ models to ‘learner-centric’ models, with an emphasis on developing e-Learning systems that used traditional web-based courseware.
But ADL faced a big problem. The USA’s military — the world’s biggest and most advanced — was still essentially silo’d into its differing factions. The navy, air force and army all delivered educational programmes that contained similar course content, but were incapable of sharing this information between each other. The result was much of the training and knowledge imparted in one silo tended to dissipate into the ether when different parts of the military communicated.
In response, ADL developed SCORM (the Sharable Content Object Reference Model), a program that integrated a set of related technical standards, specifications and guidelines into a standard operating system. When we look at learning management systems today, much of the interoperability is a direct result of the research and innovation achieved by ADL in 2000 when SCORM was first released. But SCORM had real limitations too. It could only recognise traditional learning management systems.
III. How It Works: The Engine of Tin Can API
The ‘tin can’ reference in the title refers to the first advanced communication device many of us ever owned — the so-called ‘tin can telephone’. It was essentially two tin cans (or similarly shaped items) attached together at their base by a thin, taut string or wire. As the speaker speaks into the base of their tin can, the air vibrates and creates soundwaves that manipulate the tension of the string and thereby cause the bottom of the listener’s tin to vibrate in a similar manner, recreating the sound of the original speaker. Children would often set one up in their cubby house, as the device would allow them to communicate those most important secrets with each other over small distances with the smallest of whispers.
Similarly, Tin Can API operates to capture truly secretive, subtle data that previously faded into the educative ether, and redirects those learning activities back to the ‘learning management system’ (LMS). It does so in the form of short declarative statements that follow an ‘Actor-Verb-Object’ formula. All statements are captured in the ‘Learning Record Store’. For example, the API might capture “Fred (Actor) Analyses (Verb) Venn Diagram (Object)”, which in turn can be sent to the LMS where it becomes a primary source of data for other systems within the LMS to display, analyse and use.
What kind of actions can Tin Can API monitor? You name it. Reading a book, watching a tutorial, chatting with a trainer — they’re all measurable. Not impressed? OK, what about a learner’s real world performance in an actual operational context? Trainers might want to monitor a student’s heart rate during a physical exam, or assess their quickness of decision-making between high-intensity challenges. If it can be measured, Tin Can API can measure it. The limitations are only subject to the imagination of educators.
So we can measure that Sarah analysed that Venn diagram, but we might also wish to know when she analysed it, for how long, what actions she undertook immediately before and immediately afterwards, and whether her prior study will correlate with her future success. In sum, we can begin to illuminate Sarah’s engagement with the subject material, and her inner motivations for studying in the first place.
Motivation and engagement are incontrovertibly linked within an education context. By honing in “statements” via an ‘Actor, verb, object’ formula, Tin Can allows the end-user, i.e. the RTO itself, to choose the level of complexity that it wishes to record. If an RTO wants to know whether students are accessing supplementary educational materials, they can. But they can also scrutinise their own trainers too. Want to know whether your trainers are accessing a marking guide before they assess the submitted tests of students? Tin Can will record ‘Sally Reads Marking Guide’ in the same way. Literally, every component of your student management system will be integrated using Tin Can API.
The benefits should be immediately obvious. Every element of a provider’s course delivery will be transparent, accountable, and improvable. Educators will no longer be stabbing in the dark. They can plot the evolution of their students’ knowledge, bit by bit, subject by subject, in a manner that has never been possible before.
IV. The Future of Learning
By creating a direct information exchange between the student and the student management system, education providers will soon direct the evolution of their education strategy. Each student will input powerful data into Tin Can API receptors, thereby enabling course coordinators to modify their course delivery accordingly, in order to produce better education outcomes for the next iteration of students undertaking the course. The best news? Education isn’t going to be simply tests and assignments. Tin Can API promises to deliver mobile learning, collaborative learning, blended and social learning, all into a cohesive and engaging learning experience for the student. Finally, even games will be educational.
And that’s something to get excited about.
