When Bad User Experience Design Costs Lives

Phil Jenkins

Designers constantly look for examples of good and interesting design for inspiration, but sometimes it’s useful to examine examples of bad design to reinforce the desire to make things better.  This was brought home to me recently at the “Designing Digital Futures” seminar delivered by Business Innovation - Design.

During a highly engaging presentation, Professor Thimbleby, Human Computer Interaction expert from the Department of Computer Science at Swansea University, managed to illustrate the importance of User Centred Design and Usability Testing through a series of compelling examples which, as a product designer, left me slightly disheartened that so many businesses continue to dismiss the value of improving the usability of their products.

The first example was an emergency coastal telephone – those located near beaches and along coastal paths to contact emergency services.  Having since see this first hand, this scarcely believable example of bad communication design would be laughable were it not so potentially serious.  The signage around the phone invites the user to dial 999 and ask for the coastguard. Yet the telephone keypad includes only 3 buttons and none of them is a 9! They are in fact labelled 1, 2 and 3 with no other obvious description. To work out how to use the phone, the user must refer to a small handwritten note at the base of the phone instructing the user to lift the handset and press any key. So, having deliberately confirmed the ‘999 convention’ and primed the user on approach, the telephone then expects the user to work out and engage in an entirely different and poorly communicated process. All this is happening presumably while someone else is in mortal danger (perhaps drowning in the sea). This ill thought out product is costing vital seconds, even minutes, as the user, in a highly stressed state, adjusts themselves to a new unfamiliar process. The ideas of priming, affordance, expectation and anticipation are all key to great user interaction and this product fails miserably by totally misdirecting the user.

Another, more subtle example was a study on automatic syringes – the type of device which delivers a set dose of drug (e.g. morphine) periodically to a patient. The focus of the study was to analyse the number of uncorrected errors in dosage setting. Two types of device were compared – one which used a typical 0-9 keypad for entering a dose and another which used up and down arrowed buttons to set a figure. Conventional thinking might suggest that the 0-9 keypad is easier to use – the nurse simply punches in the exact dose required. It’s quick, direct, precise and simple. However, the results of the study showed that the 0-9 keypad version showed twice the amount of uncorrected errors compared to the up and down arrowed version. Using eye tracking technology it became clear that with the 0-9 keypad product, the user’s focus was on the keypad, whereas with the up and down arrows the user’s focus is on the display, meaning that the user was more likely to pick up an incorrect dose value.

Another factor which possibly came into play with the 0-9 keypad product is that of transfer errors. These are mistakes made by a user due to familiarity with a similar but slightly different system. For example, although both do the same thing, a telephone keypad starts at the top whereas a calculator keypad starts at the bottom. A user very familiar with one can often make many mistakes when switching to the other. Muscle memory (physical rhythms learned through repetition) can easily cause mis-typing on this type of device and unless the user specifically checks the value on the display, an error will go unnoticed. This example highlights the fallacy that 'usability is easy'. It clearly is not. It is easy to disregard usability testing under the belief that it is just ‘common sense’. But as this example shows, usability has subsequent effects that may not be immediately obvious – in this case not just for the primary user (the nurse) but also for the patient. The right solution in the real world might actually be rather counter-intuitive to our perception of what the right solution should be. The only way to ever really know is to run usability research and test prototypes during the development process.

While these two examples illustrate the importance of good user interaction design in emergency and critical healthcare environments, product usability is a day-to-day issue affecting many businesses. How many businesses lose customers or repeat business because their product’s users are exposed to a complex, frustrating or downright confusing user experience? This is happening continually and many businesses may not be aware that poor user experience is costing them revenue. The irony is that bad user experience is often so subtle that customers do not complain or feedback about it. They tend to blame themselves for mistakes in the belief that they did something wrong.

This all serves to demonstrate that user experience and importantly usability testing are absolutely crucial, not just in maximising business opportunities but sometimes in saving lives.

Phl Jenkins is Senior Design Consultant at Kinneir Dufort