Why are VLE Interfaces so poor? What can be done about it?

Part of my old job was helping people to use a virtual learning environment (VLE). I'd spend much of my time warning about the gotchas. Except, of course, that most of these were entirely avoidable user interface issues. Here's an example:

Auto-focusing on the first form element on a page. If it's a select menu, when people mouse-scroll to move down the page, they alter that option instead; fairly important if that option is also the course category... Basic usability testing would have picked this up, but it remains unfixed to my knowledge. Think of the number of hours wasted with staff "losing" courses and admins having to reinstate them.

That's just the easiest to describe in terms of consequence, but it's far from the only one. If you're used to using anything by say, 37Signals, then VLEs look positively antiquated.

For a while, I thought it was just that particular eLearning environment. But talking to colleagues from other universities, and watching them use their systems, showed basic usability and interaction design problems are endemic. Not minor problems either, "Oh, I can't change that here, I have to go out, back, find it again in admin view and then click edit". If this was a shopping basket on an ecommerce site, you'd be losing sales by now.

And this isn't even getting into the quality of visual design. If you want to see atrocious typography, lack of visual hierarchy, mixed visual cues, poor affordances and confusing wording, then check out an eLearning platform. The markup isn't much better. Bad cases of divitis, tables for form layout, mid-90s style inline Javascript... It's like Designing with Web Standards never happened. But I digress...

How did this situation come about?

Part of the problem is that those making procurement decisions often don't have a critical framework for assessing the quality of web systems. No code auditing, no interface guidelines, no formal usability testing methodology pre adoption and no meaningful success metrics post adoption. Occasional surveys asking if people "like" a system and measuring "hits" just don't cut it when five minutes doing a cognitive walkthrough would reveal a raft of serious issues.

It's particularly odd when you consider the amount of peer review and critique an academic paper has to receive: why isn't a similar level of rigour applied to web learning systems? It's not as if the tools haven't been out there for at least 10 years (longer if you look at generic HCI, rather than web specific). But a few key buzzwords about facilitating pedagogy and people seem to ignore the fact that these systems don't facilitate use.

Universities have a very limited time slot to solve these issues. The current crop of students are used to the user experience design provided by modern commercial web applications. They're far less patient with inadequate systems. Some staff are already moving their teaching to external tools because they provide better experiences. With these newer staff building awareness of what's out there, existing staff are realising they didn't have problems with technology, just certain pieces of it.

So what can be done?

Universities need to start taking usability and user experience more seriously, and quantify more clearly how much bad design is costing them and their students. In a way we need an HE equivalent of the $300 Million Button to get people's attention. There's a great opportunity for a piece of research there, if funding is available. Investing in this area will have long term benefits for the institutions willing to do so.

Ultimately, more Usability, Interaction Design and HCI professionals are needed in HE, and with the ability to influence those providing and choosing the systems. That's no small thing when such issues can lack advocates in management. Getting action on improvement is easier for the open source solutions like Moodle, than closed systems like Blackboard/WebCT, but relying on the producers of systems to fix things hasn't worked so far.

Existing user-centric designers also need to be more vocal about these systems' shortcomings, or everybody will continue ignoring the elephant in the room. Once people are willing to raise these issues then it's easier to fix them. It doesn't take much effort; as Jakob Nielsen says, most usability issues can be found with five study participants. But you need to admit there's a problem in the first place.

There're also some great opportunities here. With a recession on the way, a savvy web agency could earn a tidy sum by building a more user-centric VLE. You could build an eLearning system from scratch in Django in less time than it would take to fix most of the problems with the commercial systems. Build in import from a common VLE or SCORM format and you could laugh all the way to the bank.