Our thanks to William for a stimulating talk that offered rare insights into the history and development of HCI Research, presented significant challenges, raised the title question and left us with some points to ponder.
I personally have worked in, or close to, the computer industry since 1975. With my inherent interest in the history of the industry, William’s CV is a matter for admiration and some envy. William worked in HCI research with Xerox for twenty years. His work at the renowned Palo Alto Research Center and then in their Cambridge Research Centre put him in the very heart of HCI research and development. Many of the fundamental mechanisms of interaction that we now use every day, were born in these centres. William was able to recount key developments in the history of HCI from the perspective of an insider.
William took us back to the first attempts by Xerox to bring together engineering science with the emerging interactive computing technology. The vision was to develop a formal predictive model of task-oriented human behaviour that could be used to guide the design of interactive processes. In response, Stuart Card and Tom Moran created the GOMS model. William gave the impression that this looked, at the time, to be a very useful contribution to the field until the anthropologist Lucy Suchman (now a professor and co-director of the Centre for Science Studies at Lancaster University) came along and exposed the complications that arise from the fact that human behaviour is contextual; that behaviour changes under the influence of factors external to the man-machine interface.
The next significant milestone William spoke about was the installation by Xerox of Alto computers at everyone’s workstation. These machines used a GUI with icons, menus, mouse and keyboard. They were connected by Ethernet and became the platform for application development by end-users for themselves and to share across the community. This first instance of ubiquitous computing offered prompt user feedback and seemed to signal the start of a shift away from the theoretical approach in the development of HCI to an empirical approach based on usability testing.
William went onto explore the forces that have shaped the subsequent development of HCI research. He proposed that HCI research is being driven, to differing degrees, by four key factors:
· Moore’s Law
· Canned interaction
· Software crafting tools, and
· The requirement to publish.
These forces have led HCI researchers to focus their efforts in
· Developing radical solutions;
· Deriving new heuristics from experimentation, and
· Developing better tools for designers.
All of this supports commercial activity and the requisite volumes of publishing, but as a result much less has been done to enhance existing solutions or develop new models that can be applied more broadly. This was contrasted with engineering research, where most of the work targets the development of new models and theories to support future design activity.
This dearth of pure research links directly to the need for reliable and appropriate metrics. One of the challenges of HCI Research in general and of Usability is the difficulty of providing objective measurements of problems and of improvements. This measurability is an essential component in mobilizing funding and motivating new research and development.
William indicated that these innovative developments and radical solutions tend to give people the ability to do things that we couldn’t do before. He suggested that greater long-term value might be found by working to enhance existing human capabilities, helping us to do what we already can do, but better, more effectively, faster or more cheaply. I found this distinction difficult to apply since most of the technology I work with has simply enhanced our capabilities, allowing us to do the same things faster, at greater distance and without the need to develop all of the specialised skills.
William then raised several good examples of real-world issues such as improving the effectiveness of rescue teams and providing support for carers. Whilst these are clearly worthwhile goals, they don’t necessarily attract the funding needed to support the required research. He also raised examples of more mundane issues that may warrant research. Two key examples were the tendency for writing tasks to overrun (a tendency I have experienced intensely while writing this) and the growing use of laptops in meetings. He suggested that work could usefully be done to address both of these issues. In the first case designing writing tools to help authors avoid overruns. In the second, to explore the causes of the behaviour and to find better ways of addressing these causes that don’t have socially undesirable consequences.
William left us with the thought that it may be our role as User Experience professionals that must expand to address some of these issues. The UX community is actively engaged with existing products and methods and seems more involved in achieving the sort of incremental improvement required. We can look to the HCI Research community for metrics, models and tools, but we need to apply these in our ongoing empirical research and continuous improvement of usability