A short history of SCORM

The basic problem

For many years, one of the major problems with e-learning has been the creation and deployment of quality e-learning content. E-learning content is actually software. Its development has typically been subject to the same issues as other software projects. In other words, e-learning content has been, like most software, expensive and time-consuming to develop.

To deliver and especially track results from the use of the content, it typically had to be custom programmed to work in a specific delivery environment. Different learning management systems had very different delivery environments. If an enterprise wanted to upgrade a learning management system or change vendors, often that meant abandoning very expensive content and starting over. If a content vendor wanted to distribute content widely, it was very expensive. A completely different version was often required to accommodate each different learning management system. Big content vendors, on the other hand, specified their own delivery environment and forced each learning management system to implement different delivery modules for each large content vendor.

The business case for learning objects

As learning requirements changed and overlapped within enterprises and government agencies, the importance of reusable content modules became apparent. In education as well, the usefulness of content objects or "content nuggets" to support all kinds of learning activities was demonstrated by various projects. Those are often called "learning objects."

There has thus been a strong market incentive for content that is durable, portable between systems and reusable in a modular fashion. In other words, content that is interoperable.

In the 21st century, new E-learning paradigms have evolved that use content that is developed quickly and that must be integrated in seconds, rather than in days or months like in the past. A standardized way to package and deliver such content to provide interoperability is highly desirable.

A win-win proposition

With interoperable content, content developers win because the same content can work in more different systems without modification. LMS vendors win because they can focus on the management aspects of learning, without having to constantly adapt the delivery environment to various content libraries. Enterprises and agencies using the learning management systems and the content win because instead of wasting money and time on integration of different libraries of content, they can easily mix off the shelf content with their own custom content using the same delivery environment.

Precursors of SCORM

AICC

A long time ago, when the delivery environments were still mostly DOS machines and Windows was still brittle, the Aviation Industry Computer-Based Training Committee (AICC) published guidelines and recommendations for interoperable content. The AICC has since been updating these specifications to work in a web based environment, sometimes with mixed results because of an ongoing concern for legacy content and systems. The AICC specifications did not provide a way to guarantee cataloguing metadata, or a robust way to package the content to make it portable. However, good lessons were learned from those early efforts and contributions.

IMS Global Learning Consortium

New useful specifications have emerged from other communities of practice, such as the Content Packaging specification and Content object Metadata profile from the IMS Global Learning Consortium. SCORM various IMS specification.

IEEE

An accredited international standards organization, the IEEE has been busy turning specifications into high quality standards. The standards initiatives often cooperate. For example, the IMS metadata specification was based on an early draft of the IEEE metadata standard, and other IEEE standards incorporate some of the functionality first contributed by the AICC. The current version of SCORM incorporates several of the IEEE standards.

Enter the ADL

Several years ago, the White House Office of Technology, the Department of Defense and the Department of Labor launched the Advanced Distributed Learning Initiative (ADL) in the United States. Industry and education partners, as well as various entities in other countries also joined the effort. One of the first ADL projects was for a practical profile of existing specifications and standards for content. A profile is a document that specifies a particular interpretation of a standard or specification. In the process, some gaps had to be filled in, and the end result was a set of "books", each describing a different aspect of the solution. The result was called the Shareable Content Object Reference Model, or SCORM.

The SCORM was born to take the best from the early efforts, specifications and standards, and achieve the goals of durability, portability, reusability, interoperability and accessibility for content.
Many people from the e-learning industry were involved in the genesis of the SCORM, along with the technical team funded by ADL. Each release of the SCORM has been tested in "Plugfest" events. A Plugfest is a meeting where competing vendors and developers get together to verify that their content and implementations interoperate as expected, and to iron out problems in a cooperative atmosphere. Plugfest events have attracted a very wide international participation, which is a clear indication of the impact and adoption of the SCORM well beyond its North American origins.

SCORM 1.1

The first release of the SCORM was a trial balloon, intended to discover unresolved issues. Test bed implementations revealed that SCORM 1.1 was less than fully functional, and interoperability was still mostly hit and miss. The lessons from SCORM 1.1 were put to good use in the subsequent releases.

SCORM 1.2

The first "real" release of the SCORM was SCORM 1.2. This was the first version for which a test suite was available, and thus the first version for which conformance could be verified.

SCORM 1.2 proved that content can be made portable and interoperable. For example, one leading LMS vendor has seen the cost and time of new content integration with their SCORM conformant drop to almost nothing when the content was verifiably SCORM conformant. Any remaining issue was a result of not conforming to SCORM, or the result of a misinterpretation of a SCORM feature that was still not quite pinned down completely.

The current version: SCORM 2004

SCORM 2004 improves significantly on SCORM 1.2, by eliminating even more ambiguities in the specification, and by making SCORM conformant with robust IEEE standards. The API now supports the wide range of human languages supported by ECMAScript. Besides improving on SCORM 1.2, SCORM 2004 also adds optional features for sequencing and navigation.

The addition of sequencing is a major functional milestone. SCORM 1.2 was all about making content portable, but left it to the learner to choose which part of the content to run. SCORM 2004 adds the ability to deliver activity-centered content packages that support guided or adaptive sequencing behavior.

The SCORM 2004 has been updated since its original release. You should always consult the ADL web site to verify that you have the most current release before engaging into any implementation or validation project. The most current Conformance Requirements document should be used for test plans and to verify conformance.

Compatibility between versions

Many content Learning Management System vendors will probably continue to support SCORM 1.2 content for a long time, along with SCORM 2004.

Tools are available or can be built relatively easily to convert SCORM 1.2 content packages to SCORM 2004.

It is possible to launch unmodified SCORM 1.2 content objects in a SCORM 2004 environment by using a "wrapper". Such a wrapper is a generic SCORM 2004 content object that is used to the SCORM 1.2 content object and that looks like a SCORM 1.2 environment to that object.

It is also possible to launch SCORM 2004 content objects in a SCORM 1.2 environment through such a "wrapper". However, in that case, and depending on the content, some tracking or session data may be lost because SCORM 1.2 does not support the full IEEE data model used by SCORM 2004. Obviously, SCORM 1.2 environments do not support SCORM 2004 sequencing.

In practice, the wrapper approach does not work with some SCORM 1.2 content that has been implemented using an incorrect algorithm to detect the SCORM interface. For that content, the only solution is to replace the defective algorithm in each content object, or to force the LMS implementation to provide a customized, nonstandard compatibility mode.

Another approach is to create content objects that can work in either SCORM 1.2 or SCORM 2004, with graceful degradation if the environment is SCORM 1.2. This approach is of course more expensive.

New implementations should focus on SCORM 2004 because this is the current version. There is little point in fighting the old battles of SCORM 1.2 in new implementations.

Compatibility with other specifications

Since SCORM is defines profile of IEEE standards and IMS specifications, SCORM content packages conform to those standards or specifications. However, packages that comply with those standards or specifications are not necessarily SCORM compliant. For example, an IMS compliant package that does not include the SCORM extension elements in its manifest is not SCORM conformant.

Although AICC has a long term plan to conform to the same IEEE standards as SCORM, basically AICC content is not SCORM conformant. For one thing, the AICC Guidelines and Recommendations do not include a packaging specification that supports the inventory of all components of a package, as provided by the IMS content packaging manifest. Also, most AICC compliant content uses an older, non-standard AICC defined communication protocol (called HACP by AICC), which is not compatible with the IEEE standard used by SCORM. Unlike SCORM, the AICC's HACP protocol is not compatible with offline delivery of content, because it requires an active web server component. However, some newer AICC content that uses the so-called "JavaScript" protocol can often be adapted and repackaged to work in a SCORM delivery environment.

SCORM beyond traditional CBT

SCORM is still being used mostly for fairly traditional CBT content, in the form of durable packages that get published and installed in more or less persistent ways in learning management systems. However, by design SCORM supports other kinds of learning content, including just in time content that might be aggregated on the fly to meet the requirements of a particular situation or individual. The content itself might combine high production value durable assets and simple assets created quickly to address a pressing need. The SCORM content might be used in a blended learning context or for just in time performance support rather than in a traditional LMS course assignment context. Some implementations and research projects aim at generating the SCORM content dynamically, using back ends that range from the very simple to the very complex and that may include databases, ontologies, simulation engines, assessment engines. The back ends for such systems may be using Web Services or some other service oriented architecture. While technically such SCORM content is not portable, because it is server-dependent, it can still take advantage of the interoperable features of SCORM packaging, delivery and tracking.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 2.5 License.