SCORM 2.0 Data Model and Structure
I just finished reviewing one of my new “Bird Dog” assignments for the SCORM 2.0 workshop. The white paper was by Chris Raasch, the Technical Lead for the ADL Academic Co-Lab.
Chris frames the need for an evolution in SCORM Run-Time Data Modeling, SCORM 1.x allows for data storage within an LMS, but this ability is limited in what it can store and how data can be used for groups of learners (synchronously or asynchronously). SCORM 2.0 must address an expansion of the uses of the Run-Time Data Model for synchronous and asynchronous collaborative learning experiences, facilitated and stand-alone.
Chris advocates for
- Revision of the Run-Time Data Model
- Maintenance of a “Simple” data model
- Strong content typing for easier content development
- Increased data storage bandwidth
- Rules for scoping learner data to content.
Chris proposes that this can be accomplished rather quickly by vetting out the extension of the Run-Time Data Model through the LETSI community (a practice I’d certainly endorse) and modeling the SCORM 2.0 Data Model after ECMAScript.
I generally agree with the points Chris has brought to bear in the paper. That said, I have some questions about the idea of how to make multi-user learning experiences available in SCORM-based systems, even going forward with everything up for grabs.
I’m a content developer who’s pretty familiar with the existing 1.x data model, I’ve watched the evolution of the authoring tools that help non-technical people build content that runs in Learning Management Systems –they may be the lowest common denominator, but I’d like to know who among the masses of content developers will be able to develop content to such a data model that supports interactions (small “i”) with the LMS as a persistent storage unit that will be used by other learners. One example comes to mind immediately: training sales teams on how to use a customer relationship management system (CRM). You could conceivably simulate a CRM with the LMS by having an interface that allowed sales learners to enter their contact notes as an instructor facilitates with the phone calls, and the sales team in training could then build off each other’s notes.
I get the instructional use of the model. Again… who can build to it?
It’s a question I would pose to Gafford/Blackmon, Ullery/Grata and several others talking to a collaborative data model. When the complexity of doing collaborative tasks is put on the server side, it improves the chances of implementation by content developers. Like Sequencing, however, if the complexity is placed on the shoulders of content developers –you’ll have a lot of templated systems writing to a select few data model elements.
Related posts:
Aaron, yeah I can see the value of this but it seems like the course developer would need to flag the data (under the new rules the lms follows in 2.0) so the LMS could parse what can be pulled for other users. It’s kinda like data mining via reports in the LMS’s now. The issue is that as you said-that gets complex fast. Maybe he’s thinking more along the idea of a curriculum in an lms so the sco can say this data should be bubbled up and in a standard format-then in the curriculum from the lms the data can be used in various group efforts-forums, group chats, etc?
My concern with the paper as described is that it’s messy. If one of the bigger themes in SCORM 2.0 is going to be interoperability, it’s got to be cleaner. There should be a space for ad hoc data storage at the content level — the problem is that what if the same developer is building a course for the same system and wants to use that same data space. What are going to be the security rules around that, which is different than the sandbox we’ve all come to know and/or accept? If you take content that is looking for this data and it moves to a new system — what happens when the content can’t find that specific data?
The paper has a lot of merit, but there’s work to be done.