The Ed Techie

The dilemma at the heart of learning design

I've been up in Birmingham for a couple of days at a workshop for the JISC D4L projects (I'm the director on the D4LD project). The two pedagogic planner projects generated quite a bit of interest. I know about IOE's planner, so went to the session on Oxford's Phoebe. This is a wiki-based resource that helps users with two main pathways in - 'I want to do activity X' and 'I want to use technology Y'. This multiple perspective is essential I think in any design tool - I'd probably add in 'I need resources on Z'. It's a good resource, but as they recognise not something that is really stand alone, ie you wouldn't let any educator loose on it, since it doesn't have enough guidance. This is intentional, it is an open model, and this contrasts with the IOE planner which offers a level of guidance (based on Laurillard's Conversational model but this could be swapped out with any theory of your choice). This gets to one of the problems for me in developing a learning design aid (something we're trying to do for OU course teams) - people have many different ways of working, so you need to have an open, broad system that accommodates these, and yet if the system is too broad then it doesn't help the user enough. You need to add in a level of guidance and interpetation, but as soon as you do that you are imposing a method of working, which if it doesn't match the way you operate can become tedious.

I remember when we were developing a course for the ill-fated UKeU. Sun, who were developing their VLE, created a complicated workflow system that you had to go through. This didn't match the way any of us worked and the best feature they provided was a 'select all' button which would just tick all the workflow stages so you could bypass it. Obviously you don't want the same to happen to any design tool, so getting the balance right between flexibility and suitable guidance and affordance is the key issue I feel.

A dash of 2.0

I was up in London yesterday visiting Diana Laurillard to talk about their Pedagogic Planner project (which is part of the JISC D4L programme along with our own D4LD project). I was quite impressed with what they've done. They've taken a pragmatic approach which allows users to define a course using some of the standard data (e.g. learning outcomes, number of hours, etc), and then added a layer of pedagogic planning to this that builds on Diana's conversational framework (although it could be any approach and is likely to be extended). They have framed it around a number of questions learners want answers for and then matched these with exemplars. The next step is to represent these as LAMS sequences.

I could see how we in the OU might take such a tool and rebadge it, so the labels were much more specific (and thus meaningful) to our practice.

But all the way through I kept seeing opportunities for a bit more "2.0" in there. I think this will come, and they are seeding the database at the moment, but it struck me that I have become something of a monomaniac about this now. I remember someone saying they were a fan of Tabasco and they found it impossible to eat any meal without thinking what it would be like with a dash of the hot stuff. I find the same with 2.0 - and not just to do with technology. I have recently suggested an open, 2.0 solution to managerial styles, an evening out and children's entertainment. I now admit that web 2.0 doesn't have anything to do with these really. But even so, I can't stop looking at something and thinking, 'what would a dash of 2.0 do to it?'

D4LD progress

The children are back in school (in their expensive new uniforms in my case), the holidays are over, back to work...

First up was a conference call with the D4LD team at the OUNL and Liverpool Hope. Over the summer we have mainly been concentrating on looking at some of the performance issues of Coppercore and SLeD. It seems the iTunesesque (I know, I know, let it be Martin) response times we were experiencing have been replicated by the OUNL team when they have duplicated our set up. The likely suspect was optimisation of the HSQL database. By fixing this and some minor code tweaks and an improvement to the caching method response times now seem to be down to a couple of seconds. Not ideal but a world away from the minutes we were getting at one stage.

At the OU Juliette White has been concentrating on some interface changes and bug fixes.

What we will find out now is what happens with real users as Liverpool Hope will commence using the system next week.

Learning Design project

Had a meeting yesterday with the people from JISC, who are funding the D4LD project which I am project director for. Our colleagues from Liverpool Hope also came along and we had the OUNL on the telephone.

The main aim of the project is to improve our Learning Design player, SLED, and the underlying Coppercore Learning design engine from the OUNL. We are doing this in the light of feedback from Liverpool Hope who are using the system on real live students on four courses.

The improvements tend to fall in to three categories:

• Performance - this really degrades with a few users. We have found a few bottlenecks, probably in the database, but we are still unsure whether the performance issues come down to a fundamental architectural issue.
• Usability - there are a number of interface issues we need to address, but at the moment the performance one overshadows these.
• Bug fixes - the usual.

Apart from the LAMS system (which is not pure Learning Design, but is very usable) there aren't really any other Learning Design players around (the effort seems to have gone into authoring systems). In trying to promote the Learning Design approach the barrier one often faces is its relative immaturity, and thus lack of good examples. Having a usable system, with learning designs from an actual course will make this job easier.

This is our third iteration of the SLED project, and it is still not clear that Learning Design is the way to go. I think a second strand of evaluation in this project is that of the specification itself. While I remain convinced that tools and a methodology that is focused around pedagogy and allows the exchange of designs is necessary in e-learning I am less sure that the formal IMS Learning Design spec is the way forward. I admire the very pragmatic approach taken by the James Dalziel and the LAMS team. In the Learning Design community the debate is often characterised as that between Learning Design with a capital L and D (Ie the specification) and learning design with a small l and d (ie a learning design type approach).