Shaping the Educational Technology Innovation Ecosystem

A steady stream of market developments — including the EDUCAUSE ELI research report, Next Generation Digital Learning Environments (NGDLE)¹ — are providing evidence that an ecosystem of innovative digital products shaped by educational leaders is becoming a "mainstream" idea and movement. Another significant report, from SURFnet [https://www.surf.nl/en/the-surf-cooperative/surf-operating-companies] in the Netherlands, has a title that says it all: "A Flexible and Personal Learning Environment: From Single Components to an Integrated Digital Learning Environment."² This report delves even deeper into IT architecture issues.

Here, I offer an overview of where we are in terms of developing an educational technology (edtech) ecosystem based on community-owned interoperability standards. Generally speaking, the news is good. Historic levels of collaboration among institutional leaders and suppliers are laying the foundation for something similar to the NGDLE and achieving some impressive results. However, we need additional leadership and ownership from higher education institutions to take this foundation to the level of sophistication and pervasiveness required for every institution to be able to leverage NGDLE concepts toward pragmatic improvements in teaching and learning.

NGDLE: A Call to Action

The NGDLE report calls out a relatively broad set of functional dimensions of IT systems and enterprise software that will define the future: personalization, analytics, advising, learning assessment, collaboration, accessibility, and universal design. That's a high bar.

Looking past all the details, the NGDLE concept clearly calls for an evolution from "one size fits all" education models to an education process focused on greater diversity and personalization. It also calls for a technology environment to enable the use of these new models.

Figure 1 shows a candidate NGDLE architecture in relation to product categories that already exist in many higher education institutions. Looking at figure 1, many (if not most) IT and academic leaders would say, "This is something we need sooner rather than later." Clearly, NGDLE is both a great concept and a call to action, but to realize it across all institutions, we must work together.

Figure 1. Candidate NGDLE architecture

Personalization at Scale Requires an EdTech Ecosystem

As the NGDLE report states, personalization is NGDLE's most "important user-facing functional domain." The report discusses two critically important aspects of personalization: the ability to configure the learning environment, and the ability to configure learning objectives and pathways throughout an academic program. The latter can influence student achievement and support new educational models.

The NGDLE report goes on to note the following:

Personalization is highly dependent on interoperability. Whereas the mechanisms of interoperability (such as data standards) are largely invisible to the user, personalization is highly tangible and is the most important factor shaping the user experience. A learning ecosystem that enables learners and instructors to act as the architects of their environments is a powerful tool.

Defining "learning ecosystem" is an endeavor worthy of its own article. But regardless of the ecosystem's exact form, it is clear from our current experiences with edtech that the technical ecosystem must consist of products (that is, learning resources, applications, tools, and platforms) that work well together to enable effective course experiences and pathways to degree completion.

As figure 2 shows, a mature ecosystem requires development and deployment of strategies for both interoperability (IT leadership) and instruction (academic leadership). The product ecosystem provides technology support for diverse educational needs, while interoperability enables those products to work well together. Both IT and academics are NGDLE stakeholders, and getting to the top right of the maturity model requires strong collaboration between and leadership from both parties on an institutional or departmental digital learning strategy.

Figure 2. The ecosystem maturity model

An Ecosystem of Our Own

At the forefront of the edtech ecosystem "movement" is the IMS Global Learning Consortium, through which more than 400 education organizations work together to enable an expansive set of interoperable products. IMS works with some 100 higher education institutions, including 13 of the top 100 ranked institutions worldwide and 25 of the largest 100 US school districts.

Prior to the existence and adoption of IMS standards as an integration strategy, even the best resourced institution or district that has developed and supports custom integrations have been unable to achieve a good, seamless user experience across even a handful of products. This is compelling evidence that anything short of one-click, zero-time, and zero-cost deep integration across the edtech spectrum won't meet the diverse, long-term needs of all types of institutions at all levels.

Both the NGDLE and SURFNet reports use a Legos metaphor to describe the construction of educational IT. How do we get the edtech Legos to fit together easily (one-click), but with the versatility that education requires (personalized)? Two words: interoperability standards. These terms are often misunderstood, as is the role of popular integration approaches.

Interfaces and Interoperability

The words, "interoperability" and "standards" are not synonymous; interoperability is the goal, while standards are the means. Two software applications interoperate by exchanging information that enables them to "work together" from the end-user perspective. Any two applications can be made to "interoperate" by creating custom data extraction and ingestion adaptors or by leveraging programmatic interfaces, also known as APIs, which are built into many of today's products.

To understand the API approach better, think of an API as a designated way that one application can "chat" with another. Each application defines its own approach to chat and "exposes" it to other applications that wish to chat. These "well-defined" interfaces ease the amount of custom programming required, but APIs still require custom programming — both in terms of establishing the API in the product and interfacing with the API to integrate with other applications. Hundreds of products, each with their own APIs, still require thousands of integrations (integration programming for every pair of products).

A scalable edtech innovation ecosystem requires a ubiquitous approach to interoperability that greatly reduces the need for any integration programming. What we need are common APIs implementing agreed upon exchange standards that every product implements, rather than different APIs for every product. The APIs also must be well engineered to enable efficient and productive communications between apps. A more crucial aspect of interoperability standards is information modeling and how it flows among various NGDLE products. Well-engineered standards seek to get the information model correct; the APIs tend to flow from there.

The IMS Experience

Is the hope for a set of interoperability standards that underlie the edtech ecosystem a fanciful dream? It seems so, until we look at the certified product directory on the IMS website. This product directory details the set of products that have gone through integration testing specified by the IMS community. As anyone who has followed the evolution of interoperability standards in the education sector can tell you, while not perfect the interoperability we are achieving today is years ahead of past attempts. The education sector organizations leading this work have proved it is possible to succeed with community developed and owned open standards.

Stakeholders in the product ecosystem (institutions and suppliers) have the incentive to encourage standard interoperability that is widely used and widely supported. The IMS experience shows that there is a will and a way forward to rally the K–20 educational community around pervasive standards, including APIs and information models. However, we cannot underestimate the magnitude of this endeavor and the evolution of our own mindset in terms of our ability to work together to grow and sustain this work. In particular, educational institutions must come to grips with a key new idea: we are the inventors of our own standards-based architecture and ecosystem facilitating educational innovation. At stake is an ecosystem of our own, working to enable the education of the future.

The LTI Standard: Can Lightning Strike Twice?

The IMS Learning Tools Interoperability (LTI) is the most successful interoperability standard in the edtech sector. LTI is pervasive in the US higher education sector, and it is moving rapidly into the US K–12 segment and to other parts of the world. LTI allows learning tools of all kinds to be "plugged in" to learning platforms of all kinds. Learning platform refers to almost any software that faculty and students use to facilitate access to educational experiences, including any digital content or software applications that teachers and students use. The most widely used platforms are LMSs, but IMS currently certifies approximately 70 products as "LTI consumers."

LTI v1

LTI version 1 has been a catalyst, enabling plug-and-play integration in the edtech space. Discussions about architectures such as the NGDLE would be extremely hypothetical if not for LTI's success. LTI v1 anticipated some of NGDLE's key requirements, including a trusted connection between learning platforms and learning applications, which is crucial to scaling an IT infrastructure that supports personalized approaches. From that core, many of the NGDLE's features can be supported (see table 1).

Table 1. Learning Tools Interoperability support for NGDLE

NGDLE Dimension	LTI Support
Personalization	1. LTI enables an ecosystem of products that can be easily integrated into learning platforms, thus enabling choice and flexibility. 2. LTI provides trusted and secure passing of context and user information that enables application to provide personalized user experiences.
Analytics	LTI establishes trusted connections, enables tracking of application launches and sessions, and collection of scores, grades, and artifacts.
Learning assessment and advising	LTI allows assessment tools, outcomes, and alert messages to be linked into an integrated user experience for teachers and students.
Collaboration	LTI enables an ecosystem of pluggable collaboration tools.
Accessibility and universal design	1. LTI provides a user-experience mash-up. 2. LTI lets applications exchange users' personal preferences and usage context, enabling the experience to be adapted to both.

 

LTI is a unique standard in that it offers three features required for educational institutional integration: an integrated user interface, a trusted connection, and basic messaging between educational apps. As such, it is a great example of an interoperability standard created — and now "owned" — by the education community. Despite its success, however, two LTI-related issues have yet to be addressed.

First, LTI is often used in a "sloppy," outdated way; this "sort of an LTI-like" usage "enables only mediocre integration. It is important to solve this issue as sloppy work leads to dissatisfaction, which in turn leads to less support for standards-based interoperability. IMS has gone to great lengths (and investment) to provide the tools and information developers need to "do LTI right." While IMS will continue to improve these resources, leading institutions must help the process along by collaborating on procurement, app vetting, and integration-acceptance testing. To achieve that, community involvement and resolve is needed.

Second, LTI v1 is too limited to meet the vision of NGDLE or similar personalization concepts. Despite all its glory, LTI v1 is a "dumbed down" version of the LTI that its developers envision — that is, LTI version 2.

LTI v2

LTI v2 is currently available, works well, and supports the scale NGDLE requires, offering adaptive integration "at the speed of now." That is, it gives tools and consumers the ability to declare their capabilities and auto-negotiate integration features.

Much of the LTI support for the NGDLE shown in table 1 actually requires LTI v2, which offers automated, scalable, and configurable tool and platform integration. Table 2 shows some key differences between LTI v1 and v2. Many suppliers are currently engaging in an incremental transition to v2 using the IMS resources for implementing services and features that can work with LTI v1 and LTI v2.

Table 2. Comparison of LTI versions

Function	LTI v1	LTI v2
Integration setup	Manual	Automated
UX mashup	Limited	Flexible
Messaging	Limited to basic launching and scores	Gradebook, adaptive, and extensible services
Technology	XML	JSON, REST level 3
Security	OAUTH message signing	Adds JSON web tokens
Privacy	Control of passing of user IDs	Control of all message types and content

To realize the NGDLE or similar personalization architectures, the education community must be motivated to transition en masse to the more sophisticated LTI v2. Advancing to LTI is a major topic of the IMS Board of Directors, as well as several leadership groups that IMS established for suppliers and institutions to encourage adoption. LTI v2 has the potential to be an even more ubiquitous and enabling standard than LTI v1 — if the education community can make lightning strike again.

The Next Vista: Educational Insights

As noted earlier, an ecosystem of pluggable tools and consuming platforms is a cool idea — one that has given rise to at least 70 LTI-enabled platforms. In the edtech ecosystem, the LTI consumer or learning platform is the launching point for users to find tools. This platform is similar to the Learning Hub's LMS component in figure 1. Thus, an LTI-enabled platform in the education ecosystem is analogous to a web browser in the worldwide web ecosystem. Because LTI has been successfully applied to a wide range of learning platforms at all levels of education, researchers are confident that it can handle any platform evolution that the NGDLE needs.

Integrated Gradebook Services

One potentially promising feature of digital learning environments is the ability to provide real-time information on student progress. Various tools, such as e-text readers, assessment applications, collaboration tools, or adaptive learning systems, might offer various views on what constitutes this "progress." Some record time-on-task, completion of a task, and scores, and many offer analytics dashboards to help students compare themselves against each other and let teachers see how all of their students are progressing.

Regardless of the features a tool provides, LTI makes that tool easily launchable and accessible. As some are beginning to discover, LTI now also provides a standard way to mash up gradebook-style data directly in a learning platform or portal, thus giving instructors a view of data across apps in a single user interface. Because many people have been using custom APIs to achieve this same functionality, tool integrations have become nonstandard across platforms. With LTI integrated gradebook services, we can now move away from these custom integrations in an elegant, extensible way.

Currently, several leading digital content providers and LMSs are pushing for integrated gradebook services adoption, which will be a big step forward in providing more insights for faculty, students, and even advisors. An LTI-enabled gradebook can show up almost anywhere it is needed in the educational enterprise.

Learning Analytics

Learning analytics is also a hot topic and a key feature of the NGDLE. IMS is bringing together a critical mass of leading suppliers and institutions to design and implement an analytics data standard called Caliper Analytics.

Caliper defines a rich information model and APIs to capture user actions across a range of common learning activities via real-time event streams. Although we are still in the early phases of unlocking the potential power of analyzing the real-time streams and big data collections that learning platforms and tools generate (both individually and in aggregate), both institutions and suppliers have validated the importance of standards-based data interoperability. The primary value is that the custom integrations required to get at data are simply too expensive. As a result, data analytics in education does not scale well. Caliper will solve that issue.

Caliper emphasizes four complementary areas:

• Achieving agreement on a core set of data that enables high-value use cases
• Broadening the Caliper ecosystem so that data is easier to obtain from a variety of products
• Enabling interoperability of data stream visualizations and alerts to make the data more actionable for students, teachers, administrators, and other users (the NGDLE data flow in figure 1 shows the importance of interventions)
• Enabling an ecosystem of data warehouse and analytics products to take advantage of what Caliper has to offer

All four of these areas will be shaped by the institutional and supplier organizations involved. As yet, Caliper has not addressed data reception or processing, which is key to completing the learning analytics picture.

The Third Rail: Microcredentials

The power of personalized education pathways lies in capturing student achievement. If any educational leader believes that transcripts are sufficient at any level of education — well, I haven't met that person. Meaningful personalization cannot happen without doing a better job of specifying and capturing achievement.

Figure 3 shows a high–level view of digital microcredentials, while figure 1 shows their importance in the NGDLE scheme.

Figure 3. Microcredentials interoperability structure

Microcredentials require interoperability standards, including a standard for defining and transmitting microcredential frameworks that can be shared across various digital applications to tag relevant activities and accomplishments. We also need a standard for compiling, curating, and publishing microcredentials via a transcript or other means, as well as one to enable searching and matching of accomplishments to educational or career opportunities.

IMS has a coherent set of interoperability standards in the works that cover all three types of microcredential interoperability. Further, IMS recently became the "home" for the open badges specification and movement worldwide, with full support of the movement's previous home, the Mozilla Foundation.

IMS has developed open-badge extensions for education that add features for accreditation, evidence, discovery, and analytics. It has also developed a standard for exchanging competencies and one for mapping one set of competencies to another (including derivations). These microcredentials can then be "rolled up" into an extensible digital transcript standard from IMS.

For the NGDLE or a similar concept to support personalization with a purpose — namely, student achievement — and support attainment in a way that institutions can track, requires digital microcredentials. Making the connection between academic progress with analytics and interventions driven by data is also needed. Understanding the effectiveness of learning tools in terms of progress reporting and the resulting interventions is the ultimate goal of learning analytics. In other words: How effective are our interventions under various circumstances, and how do we improve them?

Summary: A New Day Is Dawning

Educational IT is seeing the advent of a new day. This new IT generation has a crucial role to play in enabling academic transformation by providing an agile platform that users can shape and reshape. Clearly, IT concepts such as the NGDLE, based on interoperability standards and an ecosystem owned by the education community, will serve as the foundation for continuous improvement in innovation while reducing integration costs.

Notes

1. Malcolm Brown, Joanne Dehoney, and Nancy Millichap, The Next Generation Digital Learning Environment: A Report on Research, ELI Paper (Louisville, CO: EDUCAUSE, April 2015); for a summary, see Malcolm Brown, Joanne Dehoney, and Nancy Millichap, "What's Next for the LMS?" EDUCAUSE Review, Vol. 50, No. 4, 2015.
2. SURFNet, "A Flexible and Personal Learning Environment," November 7, 2016.

---

Rob Abel, EdD, is chief executive officer of the IMS Global Learning Consortium.