Like Behaviour & Information Technology, System Concepts has been in existence for 30 years. This is a long time and technology has changed dramatically although many of the issues which we addressed then are still current today. How to design interfaces which users can use effectively, efficiently and with some measure of satisfaction is still a challenge.
Another big change, which has had a dramatic impact on System Concepts' business as a human factors consultancy is that the world wide web has demonstrated beyond doubt that usability is critical to business success. No-one launches a website these days without at least performing some rudimentary user testing. For System Concepts, this means that we no longer have to spend time convincing our customers that usability is important. Indeed, we are increasingly selling our services to internal usability specialists and our task is now to persuade them that our services should be preferred over our competitors.
Fortunately we are quite successful at this and, as I write this editorial in April 2011, we are recruiting both to replace and add to our staff. One of our first questions is always why do you want to join System Concepts. I have been heartened to hear many candidates explain that they find our origins in ergonomics attractive both for providing a methodological basis for analysis, design and testing and also because we deal with so much more than websites.
When System Concepts was 25, we gave away a usb key marking our anniversary (and containing an interesting array of case studies). I was amused when a young ‘webstart’ told me that we could not have been involved in usability for that long because the web had only been around since the 1990s. I explained to him that one of our earliest projects involved helping design an interface for what were effectively call centre operators (they were not called call centres then). We were involved in such issues as use of colour (colour displays were large and new at that time), navigation and help messages. We even videoed users working through simulated tasks. Although our video camera and recorder filled two suitcases, the results were not unlike today's usability testing results.
Of course, although user testing is now well established, clients are increasingly getting us involved much earlier in the design process to identify potential products and research user needs and wants (which are not always the same!). We are also helping more clients with their overall customer experience strategy – digital products are no longer an aid to business for many organisations, they are the business.
The other link to our ergonomics heritage is that we believe the user experience is total and includes the hardware, software and service. Currently, we are working on mobile phones and tablets for several clients and although there are relatively straightforward usability issues to be overcome, the bigger challenge is to understand how such devices fundamentally change our relationship with technology.
But we do not just work in so called ‘new technology’, we also help clients with desktop software and non-web based applications. Many of the same issues apply but the biggest difference is usually the tasks of the users. Unlike web browsing with a bit of transaction when they want to purchase, other software users are often performing very complex tasks – at the moment I am working with an oil company on reservoir modelling and well management software, which is nothing at all like browsing the web. This is another area where our ergonomics heritage helps us analyse and understand complex tasks.
So it is a welcome relief to introduce an issue of Behaviour & Information Technology which primarily deals with software and system development. The papers in the second section do apply to web-based interfaces but they are also relevant to any graphical user interface.
Software and system development
Unified modelling language (UML) 2.0, introduced in 2002, is widely used in information systems analysis and design modelling. However, some users of the language find it complex and cumbersome when following an agile design approach. In the first paper of this issue of Behaviour & Information Technology, Dr Warnars from the Department of Computing and Mathematics at the Manchester Metropolitan University in the United Kingdom explains how such problems can be avoided using a simplified version of the UML diagrams for agile developments, even with novice UML users.
Netta Iivari from the Department of Information Processing Science, University of Oulu in Finland, reports two case studies on participatory design of open source software (OSS). The first case study examines a traditional community OSS development project; the second concentrates on the software development unit of a global corporation. The author characterises distributed participatory design which includes gaining an understanding of users' current practices, redesigning them together with users and gathering feedback from users related to the solutions. Online forum-based and intermediary-driven participatory design are identified as particularly important in this context.
OSS development is also the focus of the next paper by Shih-Wei Chou from the National Kaohsiung First University of Science and Technology and Mong-Young He from the College of Management, Kaohsiung, both in Taiwan. The authors investigated why OSS developers choose to contribute and how their dispersed efforts are controlled to lead to viable outputs. They found that OSS values are positively associated with collaborative elaboration and communication competence, which in turn influence task completion. They also provide OSS practitioners with guidelines on how OSS communities use OSS ideology to achieve better performance.
High risk systems are a special case where usability and user experience may need to be sacrificed to manage and reduce risk for the wider population. Earl H. McKinney Jr from the Department of Accounting and MIS, College of Business, at Bowling Green State University in the USA, describes how high risk systems use information to maintain control, and how IT systems should be designed to support this activity. Two variations in the systems view – the objective and constructive – are distilled and compared, and for each, the implications for crisis IT systems design are discussed. The limitations of the two variations of the systems view are presented, as is a brief annotated bibliography for further reading about the systems view.
Yuk Kuen Wong from Macquarie University in Australia investigated the impact of software developers' performance in the software review process. Two hundred and five companies took part (from 1380 who were invited to participate) and their responses suggest that software developers' experiences are the key driver for the software review performance. The author argues that the most critical factor in the software review process is ‘teamwork’, which can be affected by the characteristics of the task, the use of software documents and the software developers' motivation.
How to ensure password security is a problem common to desktop systems and many web applications. John Campbell, Wanli Ma and Dale Kleeman from the Faculty of Information Sciences and Engineering at the University of Canberra in Australia report a study on the efficacy of using a restrictive password composition policy. The results of their study show that although a restrictive password composition policy reduces the similarity of passwords to dictionary words, it does not reduce the use of meaningful information in passwords such as names and birth dates (a popular source for password breakers), nor did it reduce password recycling. As someone who hates having to find new passwords which I can remember but hackers cannot guess easily, I must confess I am a fan of biometric access controls. Using my laptop in public is no longer a security risk – unless someone decides to remove my finger to access my emails!
The last paper in this section deals with the management of system development knowledge (SDK). Shihchieh Chou from the Department of Information Management, National Central University in Taiwan has developed an approach to the management of SDK which takes into account the system developer's cognitive processes. The author reports a study of a real working situation where the implementation of an artificial knowledge structure was found to be useful as a means of decreasing the system developer's cognitive processing load.
Usability of input devices
In-vehicle systems are challenging the design of existing input devices. Studies already show that driver distraction can be significant and car manufacturers are looking at better ways of interaction with vehicle systems. Gary Burnett, Glyn Lawson and Laura Millen from the Human Factors Research Group, Faculty of Engineering, University of Nottingham, and Carl Pickering from Jaguar and Land Rover Technical Research, Coventry, all in the UK, describe a study to establish which tasks are best supported by an in-vehicle touchpad system. Eighteen participants (50:50 right/left handed) drove a right-hand drive simulator following a lead vehicle at a perceived safe distance. At specific points, participants were asked to carry out seven tasks of varying qualities using a prototype touchpad system, a touchscreen or a rotary controller interface. Results indicated which tasks were best suited to alternative input devices within vehicles and, in particular, the potential for a touchpad/touchscreen solution.
C. Sutter, J. Müsseler and L. Bardos from the Department of Work and Cognitive Psychology, RWTH Aachen University in Germany surveyed the use of sensorimotor transformations to control actions using graphical input. They argue that it is the distal action effect (the moving cursor) rather than the proximal action effect (the moving hand) which determines the efficiency of tool use. In their experiments, they found that comparing a touchpad and a mini-joystick, Fitt's law holds for distal action–effect movements, but less for proximal action–effect movements. In a second experiment, the dominance of the action effect on motor control was confirmed in an experiment with a digitiser tablet. The tablet amplitude was held constant, and movement times followed the perceived index of difficulty on the display. They concluded that Fitt's law did not rely on the movements of the motor system, but on the distal action effects on the display (changes in visual space).
The last paper in this issue of BIT deals with how usability flaws in graphical user interface (GUI) design increase mouse movements and consequently affect users' health. Nektarios Kostaras and Michalis Xenos from the Hellenic Open University, School of Sciences and Technology in Greece, compared three software products which had at least one usability flaw related to extra mouse movement (the selection of the software was made out of over 20 software programs that had been evaluated in our Software Quality Assessment Laboratory). For these products, all additional mouse movements were measured, involving actual users in various settings and computer configurations. The findings showed that even a single usability flaw can increase mouse movement dramatically. The authors conclude that better usability can help eliminate unnecessary mouse movement and reduce the fatigue and discomfort caused by musculoskeletal disorders.