Skip to Content

Most usability professionals know when something is usable, they just don't know how to systematically do it

The Performance Technologies Group

Category: ui-design | Author: Craig Errey | Date: 04/11/2004

Summary

The ISO 9241 (part 11) and ISO 13407:1999 standards give us a definition and measurement framework for determining whether something is usable, but does not tell us how to make something usable. The usability and user interface design profession suffers from a lack of a consistent operationalisation of usability. Guidelines are often ambiguous, being descriptive, not prescriptive. All of this means too much variability between designers, too much iteration to get it right the first time and no robust teaching framework.

 

So what really makes something usable?

Over the last 6 months we’ve been interviewing a couple of people per week, coming from diverse backgrounds such as psychology, graphic design, communications, IT, user centred design, organisational consulting, and ranging from no experience to 10 years. Sometimes they’ve brought along things to show us — generally what they expect to happen in an interview.

We’ve taken a different angle to our interviewing and evaluating whether a candidate is a good user interface designer, causing much surprise and, (un)fortunately, discomfort for candidates.

What are we doing in interviews now?

After they tell us their goal is to make user friendly interfaces, we ask: ‘So what really makes something usable?’

They all reply using words like ‘easy to use’, ‘simple’, ‘clear’, ‘logical’, ‘intuitive’, ‘understandable’. Unfortunately, these words are just self referential, circularly tautological, rhetoric.

We then say to them ‘you’ve just used a bunch of words that mean the same thing. What really, really makes something usable?’ After that, we get some more vagaries until they confess, feeling quite embarrassed, they can’t really describe what it is. Some think that it’s a creative process where we should increase the variability between designers to get more ideas and input to the process. But this is not tenable because we don’t have the luxury of deploying several designers on the job and then spending time figuring out which one is right, or trying to combine them in some arbitrary way.

Our experience is further cemented when these ‘creatives’, who have brought samples of their work, are asked: ‘Why did you design it that way?’

We then hear answers like ‘that’s how I answered this type of brief before’ or ‘that’s what I thought was best for the user’ or ‘I tested it on my a colleague (or friend) and that’s what worked for her’ or, sometimes, silence. Not being able to explain usability does not give our clients the confidence they need to know that our work will deliver for their business or customers.

All of this begs the question: ‘How can we carry on our work if we can’t actually explain what we do at an operational level?’

 

The problem of operationalisation

My own formal profession, psychology, has also experienced definitional problems. One of the most limiting was the narrow definition of intelligence as ‘the thing which is measured by intelligence tests’. This is of no real use in either designing better tests, or developing a better understanding of the nature of intelligence. It obviously didn't last.

Similarly, the usability profession has a problem of defining exactly what makes something usable, without resorting to ‘usability is what people find usable’.

 

The problem with standards and guidelines

The job interview experience is an indicator of a greater challenge for the usability profession.

Doing a Google search returns a page consolidating various definitions of usability, all saying essentially the same thing by defining usability in terms of synonyms.

The ISO 9241 (Part 11) standard defines usability as ‘the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use’, but it doesn’t really say how you actually do that. This guideline basically describes how to measure usability, but not how to design for it.

ISO 13407:1999 takes the same approach of measurement and reinforces the notion of prototyping and testing designs iteratively to see if they comply with the definition of usability. For a definitive standard, there is no systematic method for coming up with the prototypes to test!!!

This definition gives a notion of measuring something to determine how usable it is, much like intelligence. But it is still predicated on the designer having to come up with a worthwhile design to test, in order to see if it is usable. The language of ISO 13407 perpetuates the notion of trial and error in design, as exemplified by the text of Section 7.4.5 'Alter the design in response to the user feedback and iterate this process until design objectives are met'. Surely there must be a process to short-circuit this roundabout approach and get it right the first time.

The well researched site, Usability.gov, explains that usability engineering (in contrast to usability testing) is ‘…a practical and systematic way to deliver a product that works for users [involving] several methods, each applied at appropriate times, including gathering requirements, developing and testing prototypes, evaluating design alternatives, analyzing usability problems, proposing solutions, and testing a site (or other interface) with users.

Again, like the ISO standard, this describes what it is, not exactly how to do it.

 

Trial and error in user centred design

The sequence described above is to gather requirements and then develop and test prototypes. They go on to say ‘…the prototype can then be used to elicit user comments and observe the prototype's ability to lead the users through the tasks they need to perform ’.

This basically says, ‘We put something together and go round and round to see if it works. Sure, we can draw on our experience to get there more quickly, but what if we’re starting our career? What if we haven’t done one of those types of applications before? What if my experience is different to yours? What will we come up with? As a client or manager, how would I pick the right designer to work with?

Approaches like user centred design take the same approach of capturing requirements, doing a prototype and testing it out over and over again. But the big question is: ‘What happens in the middle?

 

Usability / UI design is not taught the right way

The few definitions that I’ve examined above suggest that as a profession, we are taught to say what usability is, but not definitively how do you do it. It’s a bit like the layperson art critic who says ‘I don’t know much about art, but I know what I like’. There are many usability hymn books and guidelines for all the major operating systems (Mac OS, Mac OS X, Windows, Linux [Gnome, KDE and combined Gnome and KDE], etc.), as well as general usablity (usability.gov, Yale Styleguide)and accessibility guidelines that we recite, but often these are about pixel placement of widgets on the UI and how to provide feedback in the navigation. Don't get me wrong - these are all very helpful, but none of them actually tell us exactly what makes something usable.

 

We must operationalise what we do

I feel very uncomfortable that there is no clear operational definition. How can we expect to be taken seriously as a profession if it looks a lot like guess work and trial and error until we get it right?

Imagine if you told the building developer what your requirements were for your new home, and s/he said ‘OK, that’s great, I’ll start building tomorrow. ’

Or as a property developer, your builder said ‘‘I don’t really know what makes it stand up, but I know that what I’m doing works’ would you trust them to do your next 200 story skyscraper?

Or if a change management consultant said, ‘I’ve got a few ideas on how we can re-engineer your business. Can I try a few things out at your different offices to see which one works the best?’

Or if the car engineer said ‘I want to put the brake and accelerator in different locations because I think it’s going to work better’ would you buy their car?

Or if the programmer said, ‘OK, I know what you want, I’ll start cutting code right now’.

It’s fair to say that in these circumstances, the first thing you’d say is ‘No Way!’ and promptly show them the door. You expect them to be systematic about their business. These professions, and most others, have various techniques that guide them through the process of understanding the problem space and arriving at a reasoned solution. That’s not to say that everyone who uses these techniques would come up with exactly the same solution, but there is generally significantly less variability between people, compared to user interface designers. Let’s face it, if building architects had no systematic process to take requirements to construction, then most buildings would fall down - probably with disastrous consequences.

 

What must happen to usability and user interface design?

At the base educational level the profession of usability and user interface design needs to develop in the following areas:

  • A clear operational definition of what usability really means. This is more than just the definition as defined by the ISO standards,
  • The development of a new standard to produce user interface designs,
  • An unambiguous set of usability measurement criteria,
  • A self-realisation that user interface design is not ‘creative’ exercise where you can do whatever you like and then test it out,
  • The development of a systematic process to move from requirements to design, in a way that reduces variance between people and increased confidence in their output,
  • To reject the popular belief that you need many iterations to get the design right.

These are not easy tasks, but represent some of the challenges for the industry. We have to justify what we do if we want our clients to take us seriously. This is the only way we can take pride of place along side change management and other large scale intervention activities aimed at business improvement through the successful deployment of technology.

You might think I'm being very unfair to the profession, but take the time to ask your team (developers and UI designers, alike) what really, really, makes something usable, and I guarantee you'll get very different answers from each of them. Then ask yourself 'if they all think it's something different, or can't define it precisely, how can I trust in the true quality and usability of what they're delivering?'.

 

About the Author

Craig Errey is the managing director of The Performance Technologies Group (PTG), an Australian consultancy specialising in high performance user interface deisgn. He is a registered Psychologist in NSW and has over 10 years of experience in usability and user interface design.

Top of page

 
dms stats recording image