Tin Can API and Experience API (xAPI)

From E-Learning Faculty Modules


Module Summary

Advanced Distributed Learning, the makers of SCORM (sharable content object reference model), have been working hard on a Training and Learning Architecture (TLA) to enable the extension of human learning and capabilities. There are four main components to the TLA: (1) the tracking of learning experiences, (2) content brokering, (3) learner profiles, and (4) competency networks. The Experience API (xAPI, formerly known as Tin Can API) is a technology standard that enables the capturing of a learner’s learning experiences online or offline, from any number of experiences, formal and informal (ubiquitous learning, if you will). This module introduces the Experience API (xAPI) broadly.


Learners will...

  • explore some of the animating ideas behind the creation of the xAPI (Experience API, formerly known as the Tin Can API)
  • gain a sense of what the xAPI is and how it was designed to work
  • learn about how the Experience API compares with SCORM, the sharable content object reference model
  • probe how the xAPI may affect learning analytics, learning design, teaching and learning, the hiring of new employees, and job seeking, among other things
  • consider how the xAPI may affect the design of online instruction, online teaching, and data analytics based on learning
  • think about some extant questions about the Experience API

Module Pretest

1. What is the animating vision behind the Experience API (formerly the Tin Can API)? What does this idea suggest about learners today and where learning happens? What are the perceived workplace needs for understandings about people’s capabilities? What are the understood needs for employees today?

2. What is the Experience API (xAPI, formerly Tin Can API)? What are the component parts of this API? How does this interoperability standard “work”? (How does the Experience API support the evolution to the Semantic Web or Web 3.0)?

3. How does the Experience API (xAPI) compare with SCORM (sharable content object reference model)? What are the apparent benefits to going with the Experience API over SCORM?

4. What enablements may the Experience API (formerly Tin Can API) afford…in terms of learning analytics? Learning design (online and offline)? Teaching and learning? Hiring and human resources vetting? Job seeker empowerment?

5. What does the advent of the xAPI possibly mean for those who design online instruction? Those who teach online and offline? Those who conduct data analytics on learning data?

6. What are some extant questions about the Experience API?

Main Contents

This section contains some basic information about the Experience API (xAPI), which was formerly known as the Tin Can API. The information comes from a number of publications (cited below).


1. What is the animating vision behind the Experience API (formerly the Tin Can API)? What does this idea suggest about learners today and where learning happens? What are the perceived workplace needs for understandings about people’s capabilities? What are the understood needs for employees today?

Essentially, the makers of SCORM decided to move to a new e-learning specification because they found that SCORM was not sufficient to capture learning as it is happening in the current age, with learners experiencing a range of informal learning and changing up employment every few years (with requisite needs for a range of disparate skill sets).

Two of the individuals responsible for the creation of xAPI described the realization of how the learning environment had changed:

“Much has changed since SCORM was architected. The Internet and the World Wide Web exploded in growth and online computing became ubiquitous. As social networking services, mobile devices and gaming influence more and more how people work, play and learn, the role of “traditional content” in the entire learning experience is becoming less dominant. While still important, traditional content no longer tells the whole story” (Murray & Silvers, 2013, p. 8). Their ambition was to capture a more inclusive sense of how people learn in ways that are “organic, emergent…messy” (Murray & Silvers, 2013, p. 9). Their approach conceptualized people with “ownership over their learning experiences and the data that reflects what they’ve learned and what they’ve achieved” (Murray & Silvers, 2013, p. 9).

Wouldn’t it be great if there was a system that could record people’s learning, no matter whether that learning occurs from book reading, watching a video on YouTube, having a conversation on a social media platform, playing an online or offline game, experiencing a virtual lab, taking part in a community in a virtual world, or a for-credit course at a college or university? The data from that could be shared by the respective learner and verified by whichever entity offered the training. As Experience API (Tin Can API) is conceptualized, learner data would be private and belong to the individual learner. The learning data would be controlled by the Learning Records Store (LRS), which apparently serves as a kind of learning records warehouse and data brokerage. The records store has been explained this way: “The LRS is designed so you can host your Tin Can enabled content anywhere that content is accessed and whatever device is accessing the content, the LRS will record tracking, completion and other information” (Spindell, 2014, p. 23).

(The “tin can” refers playfully to two cans connected by a cord through which children may share messages. The API enables data exchanges between various entities—with much or most of the data exchanged via computers without the need for human intervention, once this is fully set up with the proper permissions.)

People may offer electronic portfolios of their work histories to potential employers; they could document informal learning and attain credit; they may engage in lifelong learning (cradle to grave) and have documentation of their learning experiences in real space as well as online. Potential employers may check out the training and skill sets of potential new hires, and current employers may keep track of the various trainings and skill sets of those in their respective organizations.

One of the major features of xAPI involves standards for data transfer: “The Experience API also includes a set of REST services for data transfer (including POST, PUT, GET and DELETE). The services do not only allow sending statements to the LRS, but also information about activities and actors. The Experience API uses either OAuth or HTTP Basic Authentication to authenticate access to LRS services. Therefore, the LRS API can be accessed by any system or digital content with the necessary credentials” (del Blanco, Serrano, Freire, Martínez-Ortiz, Fernández-Manjón, 2013, p. 1259).

Big data interoperability. From a “big data” angle, the xAPI may be a game changer. IEEE defines interoperability as the following: the “ability of a system or a product to work with other systems or products without special effort on the part of the customer. Interoperability is made possible by the implementation of standards” (“interoperability,” 2015). Data (or even informational) interoperability refers to the ability of two or more systems to exchange usable information. Information that may be collected from a variety of sites in a live or near-live way enables the awarenesses of a “sensor network”. Learning activities may be accessed in process (Schmidt, et al., 2013, p. 66).

xAPI enables capturing learning data from various learning architectures at various organizations (Sclater, Berg, & Webb, 2015) instead of leaving such data siloed. This approach enables the capturing of learning from non-formalized and informalized contexts. This technological specification enables the broad collecting of learning data that may be used beyond individual use cases.

Such capabilities may support the broadening of learning analytics beyond the boundaries of institutions of higher learning:

“The interest in deploying learning analytics services at the campus level is increasing but there are many barriers to deployment such as: breaking down data silos, understanding the predictive models and interventions, data governance, and the lack of competences within an organisation needed to manage interventions. There is therefore a growing need for guidance to institutions which wish to develop their learning analytics capabilities as to how to integrate multiple data sources and the new systems they need to build, procure or co-develop as part of a wider community” (Sclater, Berg, & Webb, 2015, p. 1).

Currently, most analytical tools that capture learner data are internal to particular learning management systems. This means that there are siloed clusters of student data that may not generally be compared to datasets from other institutions of higher education (even if the political will to share were there). Learning analytics may be more effective if they are multi-sourced and possibly networked.

Learning data (likely anonymized) may be used to understand how particular desirable skill sets are attained, which learning objects are most efficient for acquiring particular skills (as well as those which may have unintended consequences), the optimal learning paths of learning peers, and other insights. Such collections of data in the Learning Records Store (LRS) may be used to extract learning models using machine learning advances of late (with much recent learning from massive open online courses or “MOOCs”).

In terms of adaptive learning, learners will be able to benefit from others’ learning experiences and follow the learning paths of others like themselves (based on profiles or other data) and attain their highest and best learning performances. Learning objects may be “generative” and dynamically generated on-the-fly based on the particular requirements of the individual learner (based on computerized observations).


2. What is the Experience API (xAPI, formerly Tin Can API)? What are the component parts of this API? How does this interoperability standard “work”? (How does the Experience API support the evolution to the Semantic Web or Web 3.0)?

The Experience API (xAPI) is comprised of technology specifications and data interoperability standards that may be used to enable learning data to be captured from a range of sources—offline learning, social media content-sharing sites, mobile learning platforms, social networking sites, virtual worlds, serious games, and others—while protecting privacy, the integrity of the data, and so on. Rustici Software describes the xAPI or Tin Can API specification:

“The Tin Can API is a brand new learning technology specification that opens up an entire world of experiences (online and offline). This API captures the activities that happen as part of learning experiences. A wide range of systems can now securely communicate with a simple vocabulary that captures this stream of activities enables software components to share information.” (Kelly & Thorn, March 2013)

In a presentation titled "Rethinking Learning Systems with the Tin Can API" at Google Corp., Mike Rustici (of Rustici Software) explained the vision of a learner-centered approach, with learning data captured in a way similar to the “quantified self” movement but also considering the importance of privacy protections in the management of people’s “personal data lockers”. [In one sense, these may be parts of people’s e-portfolios (Brouns, Vogten, Janssen, & Finders, 2013).] In Rustici's conceptualization, the lockers would enable sharing of select data with select groups—such as institutions of higher education and potential employers and employers. Learning data is part of people’s identities. [At one point, he asked rhetorically if people might want it known that they engaged in some politically sensitive learning. If all the trainings are recorded, would people have to then disavow that they took part in a training? Would they have to not only disavow but lie in order to disavow? And not only that, but would they have to lie in a recorded way to disavow?] A learning records store could inform learning systems and instructors of learners’ past learning experiences, which may inform their current learning needs. Rustici described a system which is “loosely coupled” and “tightly integrated”—with various dispersed sources of learning (loose coupling) but tightly integrated in terms of data sharing. He explained the importance of “strong assertion” with a “signing statement” asserting that the learning occurred (and not just an assertion by a learner). He suggests that this approach may provide business intelligence. For example, companies may identify high performing employees and work backwards to see what training they went through and identify critical learning paths…and use that information to head-hunt future employees.

What xAPI is to technologists will require someone with that expertise to fully address. Some general features have been described in the research literature though. The xAPI is REST (REpresentational State Transfer)-based, which uses the architecture of the Web (based on http and URLs). It also is built partially using natural language (to help humans read the descriptions of the documented learning.

Educational Theory Underpinnings : The Experience API is apparently informed by empirical research and various learning theories, particularly several that use a constructivist learning approach.

“The xAPI specification is highly influenced by the socio-cultural framework Activity Theory (Silvers 2014) as initially developed by Lev Vygotsky (1978) and re-envisioned by Engestro¨m (2001). Indeed, even the unit of analysis, the activity, is the same for both Activity Theory (Kuutti 1996) and the xAPI (Silvers 2014). Activity Theory’s close alignment with constructivist learning theory (Jonassen and Land 1999), which frames learning as individualized meaning-making, and social constructivist theory, which emphasizes the value of social interaction in that knowledge construction (Richey et al. 2011), has potential implications for future work. If the majority of instructional designers self-identify as constructivists, as suggested by Richey et al. (2011), then the xAPI provides a new approach for designers to implement constructivist-aligned strategies from design through evaluation” (Kevan & Ryan, 2015, p. 2).

The Actor-Verb-Object Triple to Describe Learning: Another feature of this tech specification is “RDF store” (triplestore) data. An RDF store database is one for storing and retrieving triples using semantic queries. A “triple” is defined as data which is comprised of a subject-predicate-object. The xAPI data are captured as “activity statements” composed of three parts: actor-verb-object. One author explained it as the following: {I} {did} {this}. The “actor” is the individual (or group of learners or learning teams). The “verb” is an action verb that describes the activity (“watching” a video, “viewing” a slideshow, “listening” to an audio podcast, “reading” a book, and so on). The “object” in the actor-verb-object triple is the form of the learning that was consumed (namely, the video, slideshow, podcast, book, serious game, course, presentation, or something else). The learning itself would be verified through a trusted content provider.

Another author team describes the data, which is both machine-readable and human-readable: “Actor data is unique information that describes a specific subject, such as a student or group of students. Verb data classifies the type of activity the actor participated in and often links to a human readable description of the event. Object data will link to an artifact that is typically a byproduct of or related to the activity…Since xAPI activity statements closely follow the syntax of English, the majority of xAPI data is expressed in human-readable format” (Kevan & Ryan, 2015, p. 2).

Intriguingly, as often happens with cutting-edge tools, the innovations of the time have cross-pollinated in order to enable new capabilities. One research team describes the xAPI as

“…an extension of the Activity Streams2 specification, a format for capturing activity on social networks, created by companies like Google, Facebook, Microsoft, IBM etc., that allows for statements of experience to be delivered to and stored securely in a Learning Record Store (LRS). These statements of experience are typically learning experiences, but the API can address statements of any kind of experiences a person immerses in, both on- and offline” (Megliola, De Vito, Sanguini, Wild, & Lefrere, 2014, p. 13).

Such learning data is verified for the Learning Records Stores, which are “data-endpoints that can safely collect and exchange learning activity traces” (Glahn, 2013, p. 268). In this sense, the roles of verified “activity providers” (Aps) is critical in the provision of activity statements. Aps are conceptualized as sensor networks which may identify individual learning actors and verify the (in)validity of the actor-verb-object learning statements, which depict “activity streams” that may indicate learning. One author explains:

“The xAPI data format describes an experience statement with the following 11 attributes.
  • Unique Identifier
  • Actor
  • Verb
  • Object
  • Result
  • Context
  • Timestamp
  • Stored (internal recording timestamp)
  • Authority
  • (Protocol) Version
  • Attachments
All information in xAPI statements can be separated into meta-data, descriptive information, and complementary data. The “unique identifier”, the “internal recording timestamp”, and the “protocol version” are such meta-data at the activity level.” (Glahn, 2013, pp. 268 - 269)

In the research literature, various diagrams have been drawn illustrating the xAPI. One research team shared a fairly effective figure titled “Semantic network of the statement model” which illustrates well the human-readable parts of the activity statement (Vidal, Rabelo, & Lama, 2015, p. 268). But for copyright considerations, the visual would be shown here.

As a standard, this has to be adopted by various entities. Those who run learning management systems (LMSes), massive open online course (MOOC) learning platforms, social networking sites, social media platforms, devices, and others, have to make this standard an integrated part of their technologies. The vendors who create authoring tools have to enable the addition of the xAPI standard, and then those who build learning objects (with those software tools) have to build in the xAPI features into their learning objects. [For example, Adobe Captivate, iSpring Suite, and Articulate Storyline 2 have already adopted the xAPI (Kevan & Ryan, 2015, p. 4). Various mobile-based social networks, e-book publishing tools, and learning record stores have been built using the integrated Tin Can architecture (Raghuveer & Tripathy, 2015). Rustici Software and Advanced Distributed Learning (ADL) maintain lists of early adopters on their respective sites: Tin Can API Adopters and ADL xAPI Adopters.]

Learners themselves have to choose to share their information with Learning Records Store (LRS). They may choose to retract their information as well.

The Semantic Web: In some ways, the Tin Can API / Experience API (xAPI) further advances the Semantic Web or Web 3.0 (a machine-readable Web), which was first conceptualized by Tim Berners-Lee in 2001. This xAPI enables machines to capture information automatically from the Web and Internet (once all the proper technologies and permissions are emplaced).


3. How does the Experience API (xAPI) compare with SCORM (sharable content object reference model)? What are the apparent benefits to going with the Experience API over SCORM?

SCORM (or the “sharable content object reference model”) is considered one of the most widely used e-learning standards. In January 1999, under Executive Order 13111, the DoD’s Advanced Distributed Learning initiative (of the Office of the United States Secretary of Defense) was tasked with developing common specifications for e-learning across federal and private sectors. SCORM 1.0 rolled out in January 2000, 1.1 in January 2001, 1.2 in October 2001, and SCORM 2004 (1st ed.) in January 2004, SCORM 2004 (2nd ed.) in July 2004, SCORM 2004 (3rd ed.) in October 2006, and SCORM 2004 (4th ed.) in March 2009. In September 2011, the Tin Can API was released as a next-generation SCORM. In April 2013, Tin Can API was re-named to “Experience API” (xAPI) as version 1.0.0. (“Sharable Content Object Reference Model,” Dec. 2, 2015)

Since its creation, SCORM has been built into authoring tools used to create digital learning objects. SCORM players are built into learning management systems (LMSes). They are parts of cloud-deployments for digital learning objects (and enable the capturing of pass-through grades which are then stored on learning management systems).

Comparing SCORM and xAPI: Between “SCORM” and the “xAPI,” SCORM itself provides very coarse data capture. In terms of greater granularity of data capture, the xAPI captures information that was not necessarily seen as directly critical when SCORM was created in 2000.

The basic side-by-side comparison is as follows:

+A basic comparison between SCORM and xAPI
... ...SCORM ...Tin Can API (Experience API / xAPI)
Track completion * *
Track time * *
Track pass / fail * *
Report a single score * *
Report multiple scores *
Detailed test results *
Solid security *
No LMS required *
No internet browser required *
Keep complete control over your content *
No cross-dimain limitation *
Use mobile apps for learning *
Platform transition (i.e. computer to mobile) *
Track serious games *
Track simulations *
Track informal learning *
Track real-world performance *
Track offline learning *
Track interactive learning *
Track adaptive learning *
Track blended learning *
Track long-term learning *
Track team-based learning *

(“SCORM vs The Tin Can API”)

Transitioning from SCORM to the Experience API (formerly Tin Can API):

According to Rustici Software, a company that “help(s) companies conform to e-learning standards like SCORM and the Tin Can API,” there are two main ways to use SCORM content in a Tin Can environment. One is through a SCORM Engine or a SCORM Cloud (https://tincanapi.com/scorm-to-tin-can-solution/). To use a SCORM Engine or SCORM Cloud, users may import SCORM packages, and those packages generate Tin Can statements to be stored in a learning records store (LRS). Developers may tap more complex tools to achieve the same aims (https://tincanapi.com/scorm-evolved/).


4. What enablements may the Experience API (formerly Tin Can API) afford…in terms of learning analytics? Learning design (online and offline)? Teaching and learning? Hiring and human resources vetting? Job seeker empowerment?

The Experience API (xAPI) may well be a game changer in a number of ways. One of the core ways involves the ability to capture data across a range of platforms, sites, and sources.

“The Experience API is a novel interface specification designed to link sensor networks together to enable the real-time collection and analysis of learning experiences conducted in different contexts, with the aim of providing better interoperability between the different types of participating educational systems and devices. Being precondition to mining and modelling, it forms a centrepiece in the canon of next generation techniques and technologies for capturing, codification, and sharing of hybrid learning experiences” (Megliola, De Vito, Sanguini, Wild, & Lefrere, 2014, p. 12).

Learning Analytics: “Learning analytics” refers to the analysis of learning data (such as log data for learner behaviors, such as performance or achievement data) in order to enable informed reasoning about teaching and learning and how to improve both. In terms of automated ways to apply learning analytics, this has been started already in the area of massive open online courses (MOOCs). One research team described the vision behind learning analytics linked to data from various types of learning:

“Objectives are the detection of interesting aspects and patterns in learner and learning data, building hypotheses based on these detected structures, confirming such hypotheses, drawing conclusions, and possibly communicating the results of this analytical process. In the context of this paper we will discuss how Learning Analytics relates to formative assessment, and what specific requirements can be stated on corresponding solutions from the perspective of a teacher-as-investigator” (Rebholz, Libbrecht, & Müller, 2013, p. 67).

In terms of learning analytics, if xAPI is widely adopted across a range of technology systems and it gains buy-in from the wider populations of learners, it is possible for researchers to enhance learning efficacies. They may use machine learning to identify patterns:

  • Based on the “big data,” what are variables that are correlated or associated with particular outcomes?
  • Are there demographic pattern differences of learning and learning preferences and learning performance outcomes between learners? Are there geographical patterns? Cultural patterns? Time patterns?
  • What are some types of effective messaging for instruction that empower learners? Which are types of ineffective messaging for instruction that should be avoided?

Various types of direct learning questions may be asked:

  • When is the best time to introduce particular learning in a particular sequence?
  • What sorts of informal types of learning activities maintain particular professional skillsets and awarenesses? Innovation? Resilience?
  • What sorts of learning objects, which have empirically observed “negative” effects, should be avoided?
  • How may particular skill sets be developed? What are various paths to acquiring these skills? Which paths suit which types of learners?
  • What are “optimal” learning paths for similar individuals (based on profiles, for example) with similar objectives?
  • What sorts of learning interventions should automated intelligent tutoring systems make at particular times based on the learners’ online behaviors (and their indicators of affect and cognition)?
  • What sorts of learning agents are the most effective for a particular type of learning? A particular learning domain?
  • Based on the training and learning data of various individuals and groups that are parts of organizations, such organizations and their high-level and sub-group objectives may be inferred?

Learning Design: Learning designs (online and offline) will have to take into consideration how the learning objects and learning sequences and courses will be documented in learning records stores. They will have to be “smart” in terms of being adaptable (on-the-fly) to learner needs based on their unique individual profiles and unique learning needs. The learning objects, sequences, and courses will have to be clearly defined, and the learning objectives and competencies (as statements, as rubrics) will have to be clearly defined as well. All digital learning elements will have to be designed in ways that align smoothly with not only the xAPI but other potential interfaces for information collection and design.

In terms of serious games design, the xAPI stands to influence such endeavors:

“The potential of Experience Tracking for games resides in first place in the fact that it is being developed not only having in mind highly interactive learning activities but also taking into account the feedback from some SG (serious game) initiatives. The flexibility of the data model allows tracking complex situations as games and simulations can include and also defining specific verbs if needed for adapting the statements to the specific field of SG. The fact that tracking data is not centralized in LMS (learning management system) opens a wide range of possibilities. First, e-learning systems can take advantage of the rich assessment data to make detailed adjustments on the lesson's flow. Furthermore, the tracking data can also be used by games to perform adaptation in the game content (e.g. difficulty level, extra content, etc.). Finally, educators are not tied to the LMS limitations anymore, they can benefit from reporting tools external to the LMS which best fit their needs” (del Blanco, Serrano, Martinez, & Fernandez-Manjon, Apr. 25 – 26, 2013, p. 235).

Another vision is embedding e-learning applications in social networking sites to reach a broader audience and using the Tin Can API to track the learning and to report back to a learning record store (Fazamin, Hamzah, Ali, Saman, Yusoff, & Yacob, 2015).

Teaching and Learning: In terms of teaching and learning, instructors would do well to be aware of the new affordances and “machine awareness” of their instructional work and learning artifacts. To be professionally credible, they will likely have to demonstrate continuing learning and professional development—both formal and informal. xAPI may be used to integrated learning from remote and virtual labs (Wuttke, Hamann, & Henke, 2015). In one sense, ubiquitous learning will have truly arrived and become an integrated part of people’s learning and professional lives.

Learners would have to be savvy about their learning and possibly focus on lifelong learning in order to be professionally relevant. However, to these ends, they will have more information for learning “wayfinding”, and they may be able to lower time-to-competence. The xAPI may empower learners in terms of self-regulated learning (Vásquez, Rodríguez, & Nistal, 2015).

Employment and HR Vetting: For employers, they will have another stream to vet potential employees and current employees. They will have a greater sense of access to the various learning opportunities available online and offline, and they will have some metrics to use to assess the efficacy of various learning objects and sequences and courses.

Job Seeking: Those who are seeking jobs will likely be encouraged to continuously retool and train in new capabilities and skill sets. They will have an extra tool to validate their credentials and capabilities. (including something like credit for experiential learning).


5. What does the advent of the xAPI possibly mean for those who design online instruction? Those who teach online and offline? Those who conduct data analytics on learning data?

It may help to re-frame some of the expected changes to those who work in contemporary teaching and learning. The assumptions here are likely a little beyond the current state-of-the-art. The xAPI currently “lacks specific support for any student profile information” (del Blanco, Serrano, Freire, Martínez-Ortiz, Fernández-Manjón, 2013, p. 1260) although that is apparently in the plan.

The Training and Learning Architecture (TLA) of the ADL (Advanced Distributed Learning) has four key components:

1. Experience Tracking (also known as Experience API, xAPI, and Tin Can API)

2. Content Brokering

3. Learner Profiles

4. Competency Networks

Some Implications for the Design of Instruction: The release of the xAPI sparked questions of its effects on the design of instruction and the possible requirement that they apply coding skills, which many may not have and to consider a much wider range of data points available from learning that were previously not considered (Kelly & Thorn, 2013). Until various platforms and authoring tools enable scripting to enable the capture of various “states” of learning, instructional designers will be expected to do some basic development work and put in code to capture various learner interactions, the states of their learning, and the nuances of their behaviors and achievements. Researchers have also been looking at smart learning content, “smart” as in being aware of respective learners’ contextual information and being able to adapt pedagogical strategies and services according to learners’ contexts (Taamallah & Khemaja, 2014).

Some Implications for Online and Offline Instructors: For those who teach online (or even offline, for that matter), it will be critical to understand all the learning options in the environment and the new learning about learning (through the new data and data analytics methods available).

Some Implications for Data Analysts: Those trained in handling big data and with a background in education and learning (and other fields) stand to benefit from the large amounts of information that could be made available with the popularization of the xAPI (Experience API or “Tin Can API”). As more and more parts of the ADL’s Training and Learning Infrastructure (TLA) become available, there should be that many more affordances.

In terms of analytics, the inferring of (hidden) knowledge | (given) certain observations (such as based on Bayesian Knowledge Tracing) may be enhanced.


6. What are some extant questions about the Experience API?

While the early rollout of xAPI has been met with some strong buy-in, it is not yet clear what the rate of overall adoption will be. SCORM, if it is comparable, required years before it was adopted (somewhat), and its popularity abroad (particularly in S. Korea) often overshadowed its popularity stateside. While the makers of the xAPI have pointed out some serious competitive advantages of their tool, there are some obvious headwinds:

  • Why would educational entities care to share their information with others (while opening themselves up to potential liabilities)? What would their incentives be?
  • What if such data (as many other types of data) may be prone to inference attacks? What if anonymized data may be re-identified? What happens if there are privacy challenges (such as from data leaks and hacks)?
  • Such sensor data is likely to reveal more information than those providing the data may want (which is a typical case in data)? Are there ways to head off potential secondary and tertiary uses of such data?
  • Would this API just be relegated to informal learning? Microlearning?
  • xAPI is one among several candidate standards related to e-learning (Cooper, 2013, p. 11). It is possible that something else may come to the fore.

As an open specification, it may be modified by others. That may be a channel for innovations.



How To

The adoption and integration of the Experience API (xAPI) has wide implications for many in the field of learning: coders/developers, system administrators, online instructors, instructional designers, content developers, and others. The “how to’s” are being developed and evolved at present, and there are no “must do’s” at present.

Possible Pitfalls

For the vision of the Experience API to work, there will be other pieces that need to be put into place. There is momentum in terms of the thinking work, the research work, technology integrations, and other endeavors. The work towards improving the documentation of learning will continue no matter which technology ultimately prevails through wide adoption.

Module Post-Test

1. What is the animating vision behind the Experience API (formerly the Tin Can API)? What does this idea suggest about learners today and where learning happens? What are the perceived workplace needs for understandings about people’s capabilities? What are the understood needs for employees today?

2. What is the Experience API (xAPI, formerly Tin Can API)? What are the component parts of this API? How does this interoperability standard “work”? (How does the Experience API support the evolution to the Semantic Web or Web 3.0)?

3. How does the Experience API (xAPI) compare with SCORM (sharable content object reference model)? What are the apparent benefits to going with the Experience API over SCORM?

4. What enablements may the Experience API (formerly Tin Can API) afford…in terms of learning analytics? Learning design (online and offline)? Teaching and learning? Hiring and human resources vetting? Job seeker empowerment?

5. What does the advent of the xAPI possibly mean for those who design online instruction? Those who teach online and offline? Those who conduct data analytics on learning data?

6. What are some extant questions about the Experience API?


Brouns, F., Vogten, H., Janssen, J. & Finders, A. (2013). E-portfolios in lifelong learning. In the proceedings of TEEM ’13, Nov. 14 – 15, 2013, Salamanca, Spain, 1 – 6.

Cooper, A.R. (2013). Learning analytics interoperability—A survey of current literature and candidate standards. Centre for Educational Technology, Interoperability and Standards. Blog. Retrieved December 16, 2015, from http://blogs.cetis.org.uk/adam/wp-content/uploads/sites/23/2013/05/learning-analytics-interoperability-v1p1.pdf.

del Blanco, A., Serrano, A., Freire, M., Martínez-Ortiz, I., & Fernández-Manjón, B. (2013). E-learning standards and learning analytics. (2013). Draft version. In the proceedings of the 2013 IEEE Global Engineering Education Conference (EDUCON). Technische Universitӓt Berlin. March 13 – 15, 2013. Berlin, Germany, 1255 – 1261.

del Blanco, A., Serrano, A., Martinez, I., & Fernandez-Manjon, B. (2013, Apr. 25 – 26). Integrating serious games into e-learning platforms: Present and future. The 9th International Scientific Conference eLearning and Software for Education. Bucharest, Romania.

Fazamin, W.M.A., Hamzah, W., Ali, N.H., Saman, M.Y.M., Yusoff, M.H., & Yacob, A. (2015). The use of Tin Can API for Web usage mining in e-learning applications on the social network. IEEE. 113 – 118.

Glahn, C. (2013). Using the ADL Experience API for mobile learning: Sensing, Informing, Encouraging, Orchestrating. In the proceedings of the 2013 Seventh International Conference on Next Generation Mobile Apps, Services and Technologies. 268 – 273.

“interoperability”. (2015). IEEE Standards Glossary. Retrieved Dec. 17, 2015, from https://www.ieee.org/education_careers/education/standards/standards_glossary.html.

Kelly, D. & Thorn, K. (2013, March). Should instructional designers care about the Tin Can API? eLearn Magazine.

Kevan, J.M. & Ryan, P.R. (2015). Experience API: Flexible, decentralized and activity-centric data collection. Technology, Knowledge and Learning. 1 – 7.

Megliola, M., De Vito, G.; Sanguini, R., Wild, F. & Lefrere, P. (2014). Creating awareness of kinaesthetic learning using the Experience API: current practices, emerging challenges, possible solutions. In the proceedings of the 4th Workshop on Awareness and Reflection in Technology-Enhanced Learning (ARTEL14 2014), 16 September 2014, Graz, Austria, pp. 11–22.

Murray, K. & Silvers, A.E. (2013). A learning experience. JADLET Journal of Advanced Distributed Learning Technology: 1(3-4), 7 – 13.

Raghuveer V.R. & Tripathy, B.K.. (2015). On demand analysis of learning experiences for adaptive content retrieval in an e-learning environment. Journal of e-Learning and Knowledge Society: 11(1), 139 – 156.

Rebholz, S., Libbrecht, P., & Müller, W. (2013). Interactive visual exploration of learning data: The role of teachers as learning analysts. X World Conference on Computers in Education. July 2 – 2013. Toruń, Poland, 65 – 73.

Rustici, M. (2015, Apr. 22, published). “Rethinking Learning Systems with the Tin Can API.” Google Tech Talks. Delivered Jan. 16, 2014. Retrieved Dec. 17, 2015, from https://www.youtube.com/watch?v=GRYGSI81NqI.

Schmidt, A.P., Kunzmann, C., Attwell, G., Bachl, M., Burchert, J., Deitmer, L., Gray, O., Kaschig, A., Kerr, M., Lübbing, M., Kӓmӓrӓinen, P., Müller, W., Sandars, J., Stieglitz, D., Treasure-Jones, T., & Zaki, D. (D3.1. Concept & prototype for creating and maturing instructional material. Learning Layers: Scaling up Technologies for Informal Learning in SME Clusters. European Commission Information Society and Media. 1 – 87.

Sclater, N., Berg, A., & Webb, M. (2015). Developing an open architecture for learning analytics. EUNIS-2015. 1 – 11.

SCORM vs. The Tin Can API. (2015). Retrieved Dec. 16, 2015, from https://tincanapi.com/scorm-vs-the-tin-can-api/.

“Sharable Content Object Reference Model.” (2015, Dec. 2). Wikipedia. Retrieved Dec. 16, 2015, from https://en.wikipedia.org/wiki/Sharable_Content_Object_Reference_Model.

Spindell, A. (2014, May). Mobile learning: Getting started. edChunk white paper. 1 – 26.

Taamallah, A. & Khemaja, M. (2014). Designing and eXperiencing smart objects based learning scenarios: An approach combining IMS LD, XAPI and IoT. In the proceedings of TEEM ’14, Oct. 1 – 3, 2014, Salamanca, Spain. 373 – 379.

Vásquez, M.M., Rodríguez, M.C., & Nistal, M.L. (2015). Development of a xAPI application profile for self-regulated learning. In the proceedings of the 2015 IEEE Global Engineering Education Conference (EDUCON). March 18 – 20, 2015. Tallinn, Estonia. 358 – 365.

Vidal, J.C., Rabelo, T., & Lama, M. (2015). Semantic description of the Experience API specification. In the proceedings of the 2015 IEEE 15th International Conference on Advanced Learning Technologies. 268 – 269.

Wuttke, H-D., Hamann, M., & Henke, K. (2015). Integration of remote and virtual laboratories I the educational process. In the proceedings of the 2015 12th International Conference on Remote Engineering and Virtual Instrumentation. Feb. 25 – 28, 2015. Bangkok, Thailand. 157 – 162.

Extra Resources

Open-Source Learning Records Store Options

Open-Source Learning Records Store from Advanced Distributed Learning: https://github.com/adlnet/ADL_LRS

Hosted LRS from ADL: https://lrs.adlnet.gov/xapi/

Open source LRS from learning locker: http://learninglocker.net/