Forum Higher Ed & Training Forum: Re: SAKAI GUEST THREAD: When is workflow *not* appropriate?

By: James Dalziel

[ Reply | Forward ]

Another really important set of questions - let me take them from a few angles (using my words):


Learning Design/LAMS doesn't do everything

While I think it can make sense for every course to have a course website in a LMS (so you can post announcements, upload documents, provide students with an email system, etc) this is not true for using LAMS - not every course should necessarily have a LAMS sequence. LAMS represents a particular teaching method (or more accurately, a suite of teaching methods), and LAMS may not be appropriate to a given course. Faculty should only use LAMS when the things it achieves are *useful* to student learning in the relevant course.

So the real question is not "should I always use LAMS?", but "when does it make sense to use LAMS?" The answer to this will depend on your teaching context, teaching style, course goals, student expectations, technology facilities, etc.

A different question is "should faculty consider using a tool like LAMS more in their teaching?" - again, this depends entirely on context, but I would add this - many university courses are little more than stand and deliver lectures + essay/exams. If this is appropriate to the learning that faculty expect from their students, then fine.

But most employers, and many faculty would agree that generic skills such as problem solving, good communication skills, teamwork, etc, are important qualities that universities should teach. Given this, we can question the extent to which university courses provide activities that will develop these skills, and in my experience, there is much that the university sector could do to improve this situation - there is a great mismatch between the way our courses actually run, and the qualities we expect of our students at the end. You can't expect to graduate students who are good communicators, great at teamwork, etc, if all they did in their courses was listen to lectures and do essays/exams.

So my answer to the question "should faculty consider using a tool like LAMS more in their teaching?" is "yes", but not because of some fancy technology issue, but rather because Learning Design technology encourages the use of activities that foster teamwork, good communication skills, etc. If these activities are used in an appropriate way that matches course goals and the expected qualities of university graduates, then tools like LAMS can be beneficial.


Current LAMS limitations make it inappropriate

LAMS will continue to be a moving target in terms of functionality for quite some time to come - there is still so much more to do to fully explore the potential of Learning Design as a concept.

However, I've seen a number of LAMS users decide that LAMS/LD/workflow is not appropriate to their context as a result of specific issues with LAMS V1, for example:

- Faculty want students to be able to "see" all the activities that are ahead of them, regardless of whether they should be working on them now or later (within LAMS development, we call this the "look ahead" feature).
- Faculty want to be able to change their sequence while it is running - perhaps because they had a new idea, or the class has taken a different turn to what was expected, etc.
- Faculty want to be able to improvise a sequence as they go along - they don't want to have to plan it first, then run it - they want to decide on each step in real-time with their students.
- Faculty want to have some sequenced tasks, but also some "non-sequenced" tasks that act as supports to students working through the sequence
- A million small features that didn't work the way faculty wanted (eg, the option for students to go back and change their answer to a Q&A after they have submitted it).

LAMS V1.X (and V2.0) can't do the above, despite the fact that these are all perfectly reasonable pedagogical requirements, and so people assume that LAMS or Learning Design is not useful to them. The limitation here is not that we have a narrow view of pedagogy, but simply the practicalities of difficult software development.

However, all of the features above are things we plan to do with LAMS once the V2.0 release is out and stable. The reason we'll do them is because they represent important aspects of teaching, and I believe Learning Design/LAMS should at least aspire to describe any widespread aspect of the teaching process.

An important principle for me, though, is that many of these things should be settings that a teacher decides on based on their pedagogical purpose - if you don't want students looking ahead at future activities (because it might give away some important idea that you're building up to - like the results of Milgram's famous obedience experiments - something I used to teach in psychology), then don't click the "permit look ahead" option. If you don't want students to change their first answer to a Q&A (because you want to build on this and get them to reflect on how their view might change over the course of a sequence - eg, what percentage of average people would go to the end of the electric shock scale in Milgram's obedience experiment), then don't click the "allow students to edit answer once submitted" button. (NB: In both cases, these options don't yet exist, and may read a bit differently when we finally add them!).


LAMS/LD/workflow isn't relevant, because we don't use technology to teach X

One of the important new features of LAMS V2 is that every task can be defined as "online" or "offline". In the case of offline tasks, you still have a sequence that represents the tasks (a model if you like), but instead of running online activities such as Forums, Chats and Quizzes, LAMS will provide you with printed materials instead, such as teacher instructions, student worksheets, etc.

This is a major conceptual shift for LAMS - it is now a generalised lesson planning tool (with e-delivery as one option), rather than an e-learning system. As a result, it's potential reach is much broader than the traditional domain of e-learning.


When workflow/Learning Design really isn't appropriate

There will be plenty of teaching contexts where a structured flow of tasks (even a real-time teacher improvised sequence) will not be appropriate. As Karen as pointed out in another post, if there is no practical collaboration (either because collaboration is not part of the course, or because in reality no one bothers), then sequences of collaborative activities will not be appropriate.

Another example I've seen is in the use of LAMS at Oxford University. The tutorial style of teaching there, where a tutor works with one or a few students, and provides them with great flexibility to explore different ideas, etc, may not be well suited to LD/LAMS. That's not to say you couldn't try to use it, just that the nature of the teaching experience is so fluid, individualised and contextual that any attempt at structuring it may be more trouble than it is worth. However, even in a context like Oxford, there are still other types of learning that may benefit from some structure (eg, learning statistics for use in other disciplines) - you'll see a large set of statistics sequences (from Oxford) for just this sort of approach in the LAMS Community sequence repository (NB: these were developed as examples - they're not taken from a current course).


Workflow/LD/LAMS and LMSs

Ever since we started LAMS, we've been asked whether LAMS is a LMS. The answer is no. LAMS is a Learning Design system. In software terms, the two systems are radically different - one is basically a content management system for protected course websites (with a few educational tools like quiz thrown in), whereas the other is a workflow engine for sequences of collaborative tasks.

At a practical level, we've worked hard to provide basic integrations between LAMS and LMSs like Blackboard, dotLRN, Moodle, Sakai and WebCT. In summary, these integration solve the single-sign-on problem, and allow a faculty member to jump directly into LAMS authoring from their LMS course page (without logging in a second time), author a sequence, and add it as a link to their course page; students then see the sequence under an appropriate title "eg, "Click here for activities for week 3", and then when they click on the link, they jump straight into LAMS (no second login), and when they are finished, return to their course page. For details on the various LAMS integration (including screenshots) see http://www.lamsfoundation.org/integration/

For those who've used these integration, there is one interesting issue - you now have two sets of tools (ie, you have a forum tool in your LMS, and a forum tool in your LAMS sequences). This is known as the "tools duplication" problem. Superficially these tools may look the same, but the key difference is that the LAMS tools are "workflow/LD aware", whereas the LMS tools are not. These differences are in the back-end of the tools, and the reason that LAMS can't currently drag and drop your Sakai (or any other LMS's) tools into a sequence is that the Sakai tools don't have all the necessary interfaces that LAMS needs to make them run in a sequence (and author and monitor them) like it does with its own tools.

This issue has a long history, and I won't try to recap it now, but when we started LAMS development almost five years ago, we had hoped to use existing tools to "plug in" to the LAMS workflow engine. In practice, we discovered that it would have been such a massive task to rebuild the existing tools to do what we wanted, it was simpler to build them from scratch. For some slides that try to illustrate this problem graphically (this is from a Cambridge MIT institute presentation in 2004), see http://www.lamsfoundation.org/CD/html/resources/presentations/LearningDesign.CambridgeExtra.ppt
If we'd been waiting for LMS tools to gain the required interfaces, we'd be still waiting today (as those who work on Coppercore - another LD engine - struggle with now according to Alex Little from the OU who has worked on Coppercore and his own run-time LD player "SLED").

The solution to this, in our view, was to build a generic "Learning Design aware" tools interface - this would allow you to then separate the activity tools from the workflow engine. About half of this interface is all the typical things you'd expect an LMS tool to do, but the other half is all the LAMS-related workflow elements. This is the heart of LAMS V2, and although it was very difficult and costly to do this (and there is still further work to do in the future), we believe this is an essential step for the future if the LD approach will be part of the wider requirements of "learning platforms"/LMSs. You can find out all about the LAMS "Tools Contract" at http://wiki.lamsfoundation.org/display/lams/Tool+Contract
While I can take no credit for the actual work to create this (I'm not a programmer, I'm a teacher), I think it may be one of our most important achievements.

So the LAMS Tools Contract provides half the solution to the "tools duplication" problem. The other half is that any 3rd party tools that want to "play" with the LAMS workflow engine need to implement at least a basic set of LAMS Tool Contract features (if not all of them). This is the next big step of LAMS and LMS integration - and everyone is interested in this, but no one (yet) has taken the plunge.

Once a LMS uses the tools contract to make its tools "LAMS/workflow enabled", then it will be possible to overcome the "tools duplication" problem by stripping out some or all of the LAMS activity tools, and replacing these with LMS tools. This way, you could still use the LAMS authoring environment to create sequences using drag and drop, but the tools would be LMS tools, not LAMS tools. When you run this with students, the LAMS workflow engine would co-ordinate the flow of activities for students, but the tools they would see would be the familiar LMS tools - just running within a sequence of tasks.

So given all this, what is the future for LAMS in relation to LMSs? If LMSs want LAMS-like functionality, and implement the LAMS tools contract, then I suspect you'll see LAMS merge into the "LMS of the future". It will still remain a distinct system in its own right for those who want to use it on its own - such as many of our K-12 users (and we'll keep building activity tools to allow this), but for those who want one big platform that does everything, the LAMS Tools Contract provides the technical roadmap for how to get there. I look forward to seeing who takes the plunge first!

Posted by James Dalziel