XIV.6 November + December 2007
Page: 64
Digital Citation

Observation and interaction design

Jonathan Arnowitz, Elizabeth Dykstra-Erickson

back to top 

Despite the ever-growing collection of books on interaction design, we have to note an element we want to see further emphasized: the role of observation in designing interactions.

Designing interactions, particularly between people and machines, has been around a long time. History reveals designers of interactions prior to the digital meaning of the word. We'll claim that to create a meaningful and useful design, one must acknowledge the context in which it will be used, including the people who will use it and their attitudes and beliefs. Therefore, we'd like to acknowledge three influences that serve as examples to us:

  • Frederick Winslow Taylor
  • The Hawthorne studies
  • Post-modern ethnography
  • May we create interactions with humility and benefiting from the lessons of our predecessors!

Frederick Winslow Taylor

Born in 1856, Taylor came of age in an era replete with innovations in materials science and technology. His contribution, later termed Scientific Management, held that enforced cooperation and enforced standardization of methods is imperative to determine and leverage "the one best way" to reach a given objective in a manufacturing environment. Some views on Scientific Management consider the worker to be a cog in the machine, but that was not Taylor's original intent, which is where Taylor and Taylorism part ways. Taylor's approach was often seen as mandating efficiency (not dissimilar to critiques of usability). In reality, Taylor was an idealist who strove to make life easier for the worker by introducing ergonomy to the factory floor: Make the factory floor better fit the working of the human body. Taylor designed the factory's physical layout, coupling the machine's interactions with humans into a workflow that ensured a worker never needed to unduly stretch or needlessly walk to and fro on the factory floor. He was proud of his work, claiming that designing industrial processes for efficiency could make men behave more like machines. This regrettable wording emphasized a strong distinction between workers, who don't generally know why they are doing activities, and management, upon whom it is incumbent to direct those activities. Similar efficiency procedures wend their way into software development processes today in texts such as: H. Kahler's "From Taylorism to Tailorability: Supporting Organizations with Tailorable Software and Object-orientation" in Symbiosis of Human and Artifact-Future Computing and Design for Human-Computer Interaction.

Taylor introduced time studies (later coupled with Frank Gilbreth's motion studies to produce time and motion studies) to closely observe how production is affected, in order to improve the efficiency of production. While some would view Taylor and his innovations as treating man as a machine, his focus on improvement required close observation of how the human body performs work. We can thank Taylor and these early studies for proving the value of observation.

The Hawthorne Studies

Between 1924 and 1932, a series of studies was conducted at the Hawthorne Works, a Western Electric Company factory in Illinois. The most well-known of the studies attempted to determine the optimum level of lighting for productivity in a factory setting—interesting, since Western Electric was at the time the largest manufacturer of lightbulbs in the U.S. The industrial engineers conducting the tests noted that no matter what variable was adjusted, productivity improved. They concluded that something besides lighting, noise, or other working conditions was producing this effect. The experiments began again under the direction of a Harvard professor, Elton Mayo, and continued until 1932, this time directly studying productivity and the variables that can affect it. Mayo and crew redesigned jobs, changed work hours, and observed the effect of wage changes through several experiments. They essentially started to tinker with the interaction design of the factory and see what interactions would improve productivity. Without going into the disputed details of experimental design and the accuracy of the reported results for these studies, today we refer to the Hawthorne Effect in a variety of ways:

  • Observer effect: awareness that one is being observed is sufficient (if unintended) incentive to produce change
  • Demand characteristics: a confounding variable in experimental design; participants in an experiment guess the experiment's purpose and subconsciously change their behavior in accordance with expectations
  • Implicit social cognition: actions speak louder than words; how the observed will react (sometimes as a group) and behave socially can be influenced in various ways
  • Informal organization: informal rules, rather than interaction design, affects productivity among workers

Far from discovering the keys to improved efficiency, the legacy of the Hawthorne studies is a great example of unintended and unanticipated outcomes influenced by social reactions on contextual factors that went far beyond interaction design. Discovering what influences and motivates behavior, particularly in a rich context, is at the core of the observations done then and still done today. The great value of the Hawthorne experience is in learning to observe and keep on observing, especially when an initial causal relationship doesn't quite account for the observed interaction.

Post-Modern Ethnography

Ethnography under that name has been around since at least 1922, if you take ethnography to be a development in cultural anthropology. If you look further back, Samuel Pepys's travel diary from the mid-1600s is a type of observational predecessor to the modern concept. Ethnography generally provides a picture of a culture, society, or community from a participant observer. The observer effect is addressed by long association with the observed in a participatory arrangement, where the ethnographer records attitudes, interactions, and the values that influence them.

Taylor's work on "modernizing" the factory floor is an excellent example of Modernism, a movement from the late 1800s suggesting reality is concrete, progress is good, and we should eliminate everything that holds it back. Then things got weird. From impressionism in art ("that's not real!") to Darwin's theory of evolution, disruptive thinking kept unseating the comfort and security of the modern thinker and, voilà, by the 1960s postmodernism as a movement and concept indicates a general revolt against previously held structuring beliefs.

Where the classic ethnographer with pith helmet and notebook makes observations, the postmodern ethnographer invents a reality in collusion with the observed: The world is a complex place, one can't always identify causal relationships, and there is much room in the world for contradictions that we apparently do not have a great need to resolve. Post-modern ethnography is the elaboration of observation without a central structuring set of rules.

Observation and Interaction Design

Regardless of how one chooses to conduct "design"-by exploration (we'll know it when we see it), by problem solving (disambiguating the problem space from the solution space), or by synthesizing problems and precedents—we have to rave about our historical predecessors who kept their eyes open and sought to learn from observing, and sometimes experiencing, the "situated user."

Observation is an integral part of the design process, and observation cannot be done in isolation. The observer needs to understand the confluence of social, political, economic, and environmental context; context of use; philosophy; belief system; and other factors. Taking a rich, multiple-perspective approach to observation provides a necessary complement to other design skills.

We hope to see the next books on interaction design treat observation with the great respect it deserves.

back to top 

©2007 ACM  1072-5220/07/1100  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2007 ACM, Inc.

Post Comment

No Comments Found