Intuited interfaces: remembering that people don’t know

About why digital devices can never be inherently ‘intuitive’.

We were all beginners once. I was reminded of this when I was given a new cellphone (from my work, meaning I had no idea what it would do, and thus hadn’t done any research). As I struggled to figure out how to lock the keypad, and turn on predictive text, I found myself going through an extreme cycle of emotions which, as a technophile, I didn’t realise I did.

  1. Neutral (tad of anticipation): what’s it going to be like?
  2. Disruption/confusion. Try to lock the keypad like I did on my other phone, doesn’t work. Same with predictive text. What the …?!
  3. Impatience. Trying other options. Still doesn’t work. Getting annoyed.
  4. Very annoyed. I hate this thing.
  5. Resignation. I give up. Where’s the manual?
  6. Discovery. Find page. Ooooh. That’s how.
  7. Anticipation (again). OK, let’s try this again….
  8. Revelation. It works! Why didn’t I try that before?
  9. Complacence. Of course that’s how you’d do it. I can’t remember not knowing it.

These cycles of learning happen so fast, that once they’re over, it’s easy to forget they even happened. But they’re there. Turns out that going back on the same ‘intuitive’ functions for my other phones showed that they were equally arbitrary. And come to think of it, I’d had similar initial experiences with the iPod, Mac OSX, Windows Vista … and of course, Microsoft Office 2007.

What’s the point I’m trying to make?

Digital devices can never be inherently ‘intuitive’, as the fact that they deal in abstraction automatically means that actions must be arbitrary. (An aside: for those who argue that much of gestural and time based interactions are intuitive, remember that this assumes a Western way of looking at space and time. Anthropologists would tell you that there are other ways). In other word, interfaces aren’t ‘intuitive’, they’re ‘intuited’: before that, there’s nothing ‘intuitive’ about them at all. A famous example of this is when Star Trek’s Scotty lands on earth and tries to use a computer mouse as a microphone:

Usability guru Jared Spool calls this discrepancy ‘the knowledge gap‘, and provides the following advice:

The knowledge your users have when they arrive at the design (current knowledge), what knowledge they’ll need to complete their tasks (target knowledge), and what the design needs to do to help them complete the task (the gap) are the key ingredients for making an interface that seems ‘intuitive’ to your users.

He continues that the conditions of an interface being intuitive are either that the user already has the target knowledge, or that design helps bridge the gap so that they are trained “in a way that seems natural”. I believe that this can be expanded upon. One way would be through phenomenology, but another framework I’ve come across in the social sciences is known as the knowledge matrix.

Separating out what people think they know (perceived knowledge) and actually know (attained knowledge) creates four quadrants, all with potential applications in interfaces:

  1. They Know They Know: The general standards – for example with a mobile phone that is currently the power button, call and hangup systems, keypad method of dialing number etc. Mess with these at your peril – unless you’ve got a really good reason (see next header)
  2. They Know They Don’t Know: In other words, they’re ready to learn. Think iPhone touch screen or on a lesser scale, keyboard shortcuts. The good thing about these is that people expect some training, so if there is a great enough reason you can challenge some of the known-known attributes: e.g. iPhone doing away with the keypad. Manuals, help menus, and hints can all help.
  3. They Don’t Know They Know: These can be the little design features that make a project loved. As Spool describes it: “the user is completely unaware the design is helping them bridge the gap. The user is being trained, but in a way that seems natural”. However, for this to work, it requires dedicated user testing. Scott Jenson from Google Mobile has a great presentation about how Google Mobile Maps used the buttons in an unconventional way that users picked up after around 30 seconds. This can also be a matter of incorporating other standards in so that the user may inadvertently use them and get a pleasant surprise. I got a pleasant surprise using Twitteriffic when I absentmindedly hit Return-R as I would to refresh a page in Firefox, to find that it did exactly the same thing.
  4. They Don’t Know They Don’t Know: This can mean things completely out of the user’s frame of reference (“it does that?”), or more ominously, unpleasant surprises (like my phone example above). The frame of reference issue involves some form of enlightenment, be it a friend, tutorial, or just browsing through the options. By making options available for discovery, the user is likely to find it. The unpleasant surprises is a more difficult one, normally involving changing behaviour (as opposed to knowing you don’t know which is more about starting from scratch). The underlying is that users initally think “they know they know”, when in fact it’s the opposite. Hence the intial reaction will be pretty bad. The answer is, in the words of Steve Krug, to make them become “kayak shaped problems” that right themselves. Having a dramatically different look to a particular function/removing a button etc. can also get around this in a way by making it something that people know they don’t know. For more on this,  Jensen Harris’s comprehensive presentation about the redesign of Microsoft Office 2007 is well worth looking at.

What can also be interesting is imagining how people move through these phases of knowledge. Again, going back to my phone example, it would look like the model on the right. A bit like a game of hopscotch, the aim is to get to the top right, no diagonals allowed!

While this model has a few kinks in it (e.g it can’t quite deal with the slippage while someone has to relearn something that they didn’t know they did’t know, and doesn’t in this state separate out task from action), it’s still a useful way – along with user testing of course! – of remembering how users learn to ‘intuit’ a device.

Top image by futurestreet

Vicky Teinaki

An England-based Kiwi, Vicky is doing a PhD at Northumbria University into how designers can better talk about touch and products. When not researching or keeping Johnny Holland running, she does the odd bit of web development, pretends her TV licence money goes only to Steven Moffatt shows, and tweets prolifically about all of the above as @vickytnz.

5 comments on this article

  1. Pingback: Intuitive Interfaces? No Such Thing » Devan Being Manny

  2. Pingback: Interface != intuitiv — topotropic

  3. Pingback: Unintuitive Interfaces » Lone Gunman

  4. Steven on

    What an interesting find…it reminds me of a debate I prompted with an associate at a large design firm pitching to my employer. The premise of the debate: The firm, in all its buzz-word glory, touted intuitive and innovative in the same sentence. To this day, the thought makes me cringe. In my book, if it’s intuitive…it’s not new. ‘Nuff said.

  5. Zeb Reynolds on

    I thoroughly enjoyed reading this article Vicky. Thank you!

    P.S. (to the editor)
    The blockquotes in this article are crying out for more leading. Frankly, it’s painful.
    4-5 px would make a world of difference.

    Thanks again.