Human-Centered Design
Many decisions require deciding whose preferences should influence these decisions. In some situations, there may be one ultimate decision maker, although this is very rare in public-private ecosystems. Success usually depends an understanding all stakeholders.
Human-centered design addresses the values, concerns, and perceptions of all stakeholders in designing, developing, deploying, and operating products, services, and policies. The basic idea is to delight primary stakeholders and gain the support of the secondary stakeholders.
This notion first occurred to me at a workshop in the late 1980s at the NASA Langley Research Center near Hampton, Virginia. Many participants were discussing pilot-centered design that focused on enhancing aircraft pilots’ abilities, overcoming pilots’ limitations, and fostering pilots’ acceptance. I suggested that we should do this for all the human stakeholders involved in the success of an aircraft program. People asked what I specifically meant.
I responded, “Pilots may fly them, but they don’t build them or buy them!”
In other words, pilots being supportive of design choices may be necessary for success, but it is not sufficient. The airlines have to want to buy the airplanes, the aerospace companies have to be willing to produce them, and regulatory bodies have to certify the use of the planes. The buyers, builders, and regulators have criteria beyond those important to pilots.
I elaborated the human-centered design construct and an associated methodology in a book, Design for Success (Wiley, 1991). Two other books soon followed, Strategies for Innovation (1992) and Catalysts for Change (1993), addressing innovation and organizational change. The human-centered design methodology has been applied many times and continually refined in subsequent books in 2007, 2015, and 2019.
The premise of human-centered design is that the major stakeholders need to perceive products, services, and policies to be valid, acceptable, and viable. Valid products, services, and policies demonstrably help solve the problems for which they are intended. Acceptable products, services, and policies solve problems in ways that stakeholders prefer. Viable products, services, and policies provide benefits that are worth the costs of use. Costs here include the efforts needed to learn and use products and services, and policies, not just the purchase price.
The overall human-centered design approach is intended to increase validity, acceptability, and viability beyond that usually experienced with the ways in which problems of the significant scope are usually pursued. This begs the question of what shortcomings plague existing approaches.
First and foremost are viability issues. Sponsors of change initiatives complain that they take too long and are too expensive. This is due in part to the business processes of sponsors. However, more fundamentally, much time and money goes into developing aspects of solutions that, at least in retrospect, were not needed to address the questions of primary interest.
Second are acceptability issues. Many key stakeholders in the types of challenges I have addressed are not educated in analytic methods and tools. Nevertheless, they are often highly talented, have considerable influence, and will not accept that the optimal policy, somehow magically produced, is X equals 12. We need methods and tools that are more engaging for these types of stakeholders.
Third are validity issues. There is often concern that overall analyses are of questionable validity. This concern is due in part to the possibility that assumptions are inconsistent across component analyses. There is also the issue of incompatible definitions of organizational states across component analyses, which can lead to misleading or incorrect results. This is particularly plaguing when one is unaware of these incompatibilities.
The overall human-centered design approach overcomes these issues in several ways. The early steps of the methodology focus on problem formulation, with particular emphasis on interactive pruning of the problem space prior to any in-depth explorations. In-depth analyses tend to be expensive, so it is important to be sure they are warranted.
Second, we have found that key stakeholders value being immersed in interactive visualizations of the phenomena, and relationships among phenomena associated with their domain and the questions of interest. This enables them to manipulate controls and explore responses. This is typically done in a group setting with much discussion and debate.
Third, the overall approach explicitly addresses agreeing on a consistent set of assumptions across analyses. This prompts delving into the underpinnings of each type of analysis. The overarching question is whether connecting multiple types of analysis will yield results that are valid in the context of the questions at hand.
This post has set the stage for next week’s post on Human-Centered Systems.
