The Ikea effect – why we should build a flat–pack V[R]LE
For some time now my colleagues in the Warwick E-learning Advisor Team have been arguing that we should create some kind of 'templating' system within the web publishing component of our V[R]LE (Sitebuilder). Chris Coe has a great name for it: Elaborate (quick, file for a trademark).
The idea is that we could create templates based on generic and discipline specific learning designs. A template would embody the structure, functionality, and flow of the learning design in a set of pages and interactions. The template could then be copied and, following a set of instructions, filled-out with learning content. For example, there could be a template for a peer-review process. This could even be combined with a 'wizard' approach that builds the detail of the learning activity based upon answers provided by the learning designer (tutor) to a series of questions.
We justify this argument with these claims:
- Speed and efficiency;
- Promoting consistency and good practice;
- Developing a shared language for describing different learning designs;
- Giving a focus for evaluating different learning designs;
- Allowing us to tailor designs for departments and courses.
I've just discovered a further significant reason for working in this way, given in an article in the February 2009 edition of the Harvard Business Review:
The Ikea Effect: When Labor Leads to Love by Michael I. Norton of Duke University
...labor enhances affection for its results. When people construct products themselves, from bookshelves to Build-a-Bears, they come to overvalue their (often poorly made) creations.
The success of Ikea, they claim, is to some extent based upon the flat-pack principle - do much of the work for the customer, but allow them to feel as if they are still investing their labor (and hence their love) in the product. Perfect flat-pack makes people love you! And Ikea certainly do have some ingenious ways with flat-pack construction.
A templated learning-design is flat-pack.
Norton gives a warning: if the customer puts too much work into the construction, they may end up loving the product too much, thus making future change impossible.
The second rule of Ikea is to make the construction easy enough for the product to feel potentially disposable.
The same must be true of learning activities built from a template.
But what of customisation? We know that people love to customise. Indeed, Ikea offer that to some extent, but the customisation is never fundamental - it's simply a matter of combining products and adding ornaments (that's why they sell cheap bits and pieces from which they can't possibly profit).
So then, how to build a V[R]LE that is used and cherished by the masses: use the Ikea effect.
If you're involved in building learning content, or are interested in the construction of V[R]LEs, you are welcome to comment on this idea.
Update: I've just thought of another reason why templating is good. Here's the use case:
- Tutor A creates a useful template.
- Other tutors in their department see the pages created from the template, and want to do similarly.
- Each page is marked with "Created using the template TEMPLATE NAME, designed by PERSONS NAME .
- The tutors can then create their own pages using the template, and get advice Tutor A.
24 comments by 7 or more people
[Skip to the latest comment]Chris May
I think another part of the Ikea analogy which is pertinent here is that Ikea work hard to make the process of construction not be so difficult that people just give up, or are unsatisfied with their creations. I once tried to make a bookshelf from ‘scratch’ (From a selection of bits of wood from the timber merchant) but my carpentry skills weren’t up to the job, and the result, even after a lot of effort on my part, was so wobbly that it went on the bonfire. Ikea stuff may require some effort, but it’s always do-able by the vast majority of their target market.
That said, I’m not sure that I now buy flat-pack furniture because I enjoy making it (or even, because I enjoy having made it); I buy it because it’s cheap, and it’s easy to get home. If I could get the same designs and quality pre-made, delivered, and fitted, for the same price, I would.
24 Feb 2009, 09:20
Robert O'Toole
Yes, and every teacher would like their own personal e-learning advisor to do their work for them.
We could have a compromise position, in which we create templates for small groups. I suppose that’s customisation. Still very satisfying, but achievable. To a great extent that’s what i’ve spent a lot of time doing already. We just don’t have a mechanism for making it simple and scaling it up.
24 Feb 2009, 09:38
Robert O'Toole
I recently went on a shelf building frenzy after (by luck) successfully completing a major set of home made book shelves. Now my attachment to my shelf building skills is so fundamental to my psyche that I’m putting up shelves in every conceivable location, despite the fact that they are totally un-unecessary. This is dangerous. I’m mentally locked-in to building my own shelves.
If i’d gone for flat-pack, i would have put the shelves up and quickly moved on to some other furniture construction problem, for example a better desk.
24 Feb 2009, 09:42
Interesting idea, though would we find that at the end of three years study students would find they had this odd module left over that seems important but doesn’t seem to fit anywhere?
24 Feb 2009, 13:21
BTW – the last comment may be a bit jokey, but I do feel that this is a valuable way of looking at it – just a bit frazzled to comment properly right now!
24 Feb 2009, 13:23
Robert O'Toole
Läs jävla manual!
24 Feb 2009, 13:24
Jenny Delasalle
I think you’ve had a great idea and I’d love to see the templates working.
My experiences with flat-pack furniture, rather like software, are that we really need good instructions to go with all the lovely pieces and the pretty picture of the end product.
There will always be a role for a proper carpenter (techie) who can rescue those flat pack pieces (electronic resources) that go wrong!
24 Feb 2009, 17:33
Robert O'Toole
Instructions, pictures and enabling constraints. Donald Norman talks about this in The Design of Everyday Things.
Enabling constraints direct and help without increasing the cognitive burden.
We can include constraints within an object that direct the object to be assembled or used in a certain way. Great designs do that. The constraints imply that the design can only be used in a certain way. The 3 pin plug is a classic. You can only plug it in one way (in some countries they have plugs that can be plugged in wrongly). In web design, there are design patterns that when implemented well include enabling or directing constraints – the process funnel is the classic. Similarly, in the emerging field of learning design patterns, the best patterns use constraints.
24 Feb 2009, 18:13
Catherine Fenn
Interesting use of flat pack analogy – final build quality/functionality determined by; quality of base components, quality of binding/fixing components, skills of ‘builder’, interpretation of instructions…
Have used ‘templating’ principle to structure SiteBuilder based resources for a while now – currently takes deep investment of time in manual interventions to keep content generation going which sadly restricts time available for more innovative stuff… would be really very interested in deployment of ‘templating’ system that could support a more automated approach?
Examples of our current stuff available for Warwick staff to look at:
Our VLEsCurrent approach is to “supply” templates as skeleton structures and then support/guide addition of ‘flesh to the bare bones’.
What do you think?
25 Feb 2009, 13:23
Robert O'Toole
Interesting.
I’ve worked in a similar way.
Do you use the personalisation tags to include instructions for staff only? That’s a good way of providing guidance.
25 Feb 2009, 15:40
jeremy hunsinger
But… there is nothing quite as nice having custom built-in bookshelves and really great antique furniture too. The ikea model is ‘ok’ but it is just an efficiency and standards argument. You have limited creativity, but the real problem with the idea model is detraditionalization form its generic style. That is, you lose certain aspects of identity construction. People make fun of other people’s homes when they see they are filled with ikea style furniture. It exemplifies a certain cultural lack, a lack of identity, a lack of valuing traditions. It is sort of like the Corbusier influence on public housing and urbanism, the generic, modular, sameness begets a certain violence toward its inhabitants embodied as a lack of care and personality. I suspect that if you had an ikea model, many professors would just not use it, or use their own tools, which we already see happening in blackboard and sakai.
25 Feb 2009, 16:18
Catherine Fenn
Yes, include [if-staff] [/if-staff] tags around staff guidance notes on various template pages. Tend to highlight these staff notes in some form of boxed format in hope that they will be noticed!
25 Feb 2009, 17:05
Robert O'Toole
Maybe a standard icon would be useful, so people know that there are ‘editor notes’ available.
25 Feb 2009, 17:27
Just to endorse Catherine’s comments (Catherine and I are colleagues, by the way)...
We started using a ‘template’ approach in 2005 and in various contexts have proposed the development of management tools to enable the construction of a template that, exactly as you say, ‘would embody the structure, functionality, and flow of the learning design in a set of pages and interactions’. We’d be keen to work with others on taking this idea forward.
Lack of such a facility is slowing down both development and dissemination.
25 Feb 2009, 18:32
Robert O'Toole
Hi Jeremy. Good point. We already support the custom built approach. Out of the +/- 3000 people who edit content every month, we have hundreds of people creating content in that way. Some of them are very advanced. But we also have thousands who aren’t quite able to do that (lack of skills or lack of time), and perhaps are limited to simple operations, such as editing text on a page that someone else has made.
25 Feb 2009, 21:18
Robert O'Toole
Instructions, pictures, enabling constraints + good icons.
26 Feb 2009, 10:15
James Bateman
I think you’ve turned over a stone here Robert, and I am a worm crawling out from the mud…..
I think the idea is perfect. As a masters student, and NHS doc, I used Aidrian Stokes Templates in a Warwick module in Med Ed. Its one of the few modules that has really drawn us as students into the discussion, as he was able to update, analyse and maintain them carefully. I am guilty of not ‘buying into’ the ‘sign in and hello’ approach used in many of the Masters modules I’m sitting because I feel often that the resource isn’t really being utilised, and is more a tic box phenomenon.
I think one critical aspect of this is that, as a university with a migrating flux of stundents coming in and out of your services it poses a bit of a problem, in that if they want to get to grips with Sitebuilder, it can be a bit complex.
Contrast this with Moodle/ Blackboard (I personally run a Moodle based VLE with over 600 members, without an exactly massive IT support framework). I don’t think I could do the same with Sitebuilder in the same amount of time. I’m sure sitebuilder is more secure/ has numerous other benefits, once you get more used to it, but in a year I’ll be in Stoke/ Hereford/ somewhere else.
I draw the analagy (as a hospital doctor) of hospital IT systems. Whilst they may all work very well in their own way there is a need for standardisation. Its clear that this hasn’t worked in the NHS (and has been very costly) but the doctors that rotate round websites can’t be expected to master a new IT system once a year.
In summary I think sitebuilder is good, with templates it could be great, and why not have an option to embrace things like moodle inside the current structre to allow people like me to do research into behaviours of students without learning another system?
And are you sure the templates won’t just make it more complicated? :)
27 Feb 2009, 16:42
Robert O'Toole
Thanks James.
The comparison between Sitebuilder and ‘traditional’ VLEs is interesting, but potentially misleading. They really are two different kinds of system, with different (but overlapping) purposes and priorities. Before making any judgements about the value of different systems for your needs, it is important that one understands these differences.
Sitebuilder is a web content publishing and management system. It enables people to create and publish web pages with tools that go a long way towards guaranteeing the quality and consistency of those pages. In addition, we now have tools that make it easy for editors to create sophisticated dynamic, interactive, multimedia pages of an increasing range of types.
Pages are published within a ‘browser tree’ pattern. The position within the tree gives each page its context (relating it to other pages, to people and to organisations). This is very much an institution-centric design pattern, but with responsibility devolved widely. Importantly for our ‘research-based learning’ pedagogy, it includes an ability to devolve content publishing and editing to students, in a managed and structured way (this perhaps qualifies Sitebuilder to be a VRLE). The result has been:
1. A very high number of content items produced (I’ve forgotten how many, but it is a staggering number).
2. Around 3000 people creating and editing content each month.
3. Most content is well maintained over time, despite changes in staff (rather than posted and forgotten).
4. There’s a massive user community, with a tight feedback loop between users and developers.
5. A well managed diversity, with a wide number of approaches catered for in a system that supports very high activity levels (I think it’s over 200,000 page hits a day) with high reliablity (99.9%).
My thoughts on the Ikea effect are relevant to this content creation activity. They suggest that adding more of a ‘flat-pack’ approach might increase and improve content production even further.
Sitebuilder is not, however, a ‘content-push’ tool. It’s not about efficiently feeding each user with personalised information about resources, events, requirements and actions. It doesn’t directly support that design pattern (although a student could use it’s RSS mechanism to set up content-push). There’s a second significant question to be asked: should there be a content-push facility on top of Sitebuilder? Perhaps. Maybe it’s being worked upon right now. Maybe it will work if we can get good quality data on which to base the push.
Traditional VLEs start from a different perspective. They are principally content-push tools. The student signs in to a portal page, which pushes information to them about resources, events, requirements and actions. It’s all done quite efficiently (depending of course on the underlying student data). Over time, VLEs have grown to include tools by which tutors (and sometimes students) can add data about resources, events, requirements and actions. Perhaps because of the VLEs focus upon efficient and simple content push, the format and design of the this content tends to be either restricted. There are tools for creating more elaborate content, but these are neither standard nor core. Actual learning content tends to be “posted” to the VLE in the form of Word documents and Powerpoint presentations (with not guiding framework, pink comic sans text is not un-common).
Perhaps the important questions to ask are:
1. What do you value most, efficient content push or rich content creation?
2. If you want both, is there a system that can give both? (I think not at the moment)
3. Which system is more likely to move to offering both? I think perhaps Sitebuilder.
03 Mar 2009, 10:26
Robert O'Toole
That’s in part why the university seems to favour what i call an ‘institution-centric’ approach. In a year you might be somewhere else, but the university wants teaching content developed and used by you for teaching at Warwick to stay at Warwick and to make sense to future teachers.
03 Mar 2009, 13:39
Steve Ranford
The thing I like about Ikea is the store! Many an hour can be spent walking through the upper floor trying out the furniture in those example rooms. I think a key to the Ikea experience, and their success lies in their ability to inspire people with what can be done with their furniture. You certainly don’t go to the warehouse to make your choices.
There are many great examples around sitebuilder such as the links Catherine posted above. These examples are going to have to be promoted to inspire and help people select the right tempates for the right job.
Not everyone likes the Ikea experience: Njarnia 6min clip from BBC3’s The Wrong Door.
03 Mar 2009, 17:25
James Bateman
Robert I think you make your point well re: the student creation, I hadn’t really thought of it specifically in those terms.
1.) The beauty of sitebuilder is that you allow students (like me) to publish on the site, without the capacity to “wreck” the site in terms of editing powers etc. I think you can do these sorts of things in other VLE’s using the standard heirachy approach, but its not as easy.
2.) I don’t know anything cheap that does both: although wordpress when managed well incorporated into VLE’s( I like what the NYtimes does with their blog) is coming close with the widgets etc
3.) Sitebuilder may well offer both, but to the floating graduate like myself, its the question is of usability for people on short term placements who will/ are likely to have been exposed to BB or Moodle.
As a final point (as a non IT expert)
4.) Also the commercial aspects of the Universities VLE’s (e.g. subscriptions to courses run externally through PayPal) is not as easy. I did actually meet a modern Languages lady to talk through our PayPal function on my moodle site, a simple plugin. Sitebuilder may have this, I’m not sure. Its not particularly romantic in terms of the student experience, but I guess its the reality of the current educational environment. This does open up another can of worms, but I think its still a (borderline) valid point!
03 Mar 2009, 23:16
Robert O'Toole
I work with Arts Faculty staff and students, many of whom are not exactly ‘digital natives’. And yet when we introduce them to Sitebuilder, the majority are able to create and edit content very quickly. There are things that could be done to make it easier, but it’s not the case that Sitebuilder is unusable to such people. It’s just a matter of making it more efficient.
However, I have found that some people do struggle with Sitebuilder. They are the people with more experience of VLEs and similar systems. Their struggle is with understanding Sitebuilder’s purpose and consequent design patterns.
Perhaps Sitebuilder might soon adapt to different purposes with different design patterns.
04 Mar 2009, 09:15
Robert O'Toole
Online payment?
In Sitebuilder you can create a Formsbuilder form with secure online payment (a proper card based system). You then have two options regarding subscription:
Option 1 is the correct and proper route. But of course it adds many complications. The student would have to be registered on an officially approved course. The Centre for Lifelong Learning run short courses, so there must be a way of doing it without going through endless committees. But there would probably be a time delay. This is the result of University policy, not Sitebuilder’s limitations.
04 Mar 2009, 09:23
Robert O'Toole
Update: I’ve just thought of another reason why templating is good. Here’s the use case:
1. Tutor A creates a useful template.
2. Other tutors in their department see the pages created from the template, and want to do similarly.
3. Each page is marked with “Created using the template TEMPLATE NAME, designed by PERSONS NAME .
4. The tutors can then create their own pages using the template, and get advice Tutor A.
04 Mar 2009, 09:30
Add a comment
You are not allowed to comment on this entry as it has restricted commenting permissions.