Image source: jnd.org
A couple of weeks ago, I finally finished, “The Design of Future Things,” by Donald Norman. I loved his popular book, “The Design of Everyday Things.” Norman is clearly an important thinker in the subject of design and usability. He tackles the intelligent systems that are being designed for our homes, offices, cars and personal devices. Instead of writing a full blown book review, I want to highlight on idea raises towards the end of his latest book, which is the entitled “The Science of Design.”
In this section, he cites that design is an interdisciplinary field, which often combines, art, social science, engineering, and business. Of the fields which comprise design, each field falls within a spectrum of formal methods of evaluating design. Engineering has quite formal approaches, and aesthetics tends to resisting them. Norman calls for a “science of design” because he feels that more rigor is needed in the intelligent systems he describes the book. The argument is easy to accept after reading about the many failures in the initial attempts at intelligent systems, which he documents in the book.
Norma does not offer the specifics of what this methodology would look like. While there are benefits to a formal approach to evaluating design, I would argue that we need to proceed with extreme care in creating an approach. My fear is that the easy route will be taken, which would blindly try to build evaluation tools based on medical experimental methods, which is where “quantifying research” usually ends up. This would be clearly wrong. Understanding if a cancer drug treatment worked is much more straight forward than if and more importantly why a design worked. (I won’t go into the problems of medical experimental methods, of which there are many.) Tom Reeves from the University of Georgia has some interesting thinking in this area when looking at methods of evaluating interactive educational tools.
Obviously, someone could create an experimental measure if a person used, learned, and understood the design properly. However, a simplistic efficacy rate (99% of testers used the design “properly”) may miss the big picture of, for example, a disenfranchised population who are not being designed for, which raises questions about the ethical and political responsibility of the designer. (Products such electronic voting booths, public transport, and educational tools are examples which readily come to mind.) Further, the leaps in innovation often require a lag time for people to understand and integrate the new design features into their lives. The temptation to overuse design evaluation tools will be great for companies who risk millions of dollars to roll out products. If the evaluation tools are poorly implemented, innovation may decrease as companies and innovators choose safer designers over ground breaking products such as GUI desktops, because they don’t initially “test” well. In the end, Norman’s call for a science of design is an important one, and it ties into the ethics of design that I’ve been thinking about lately. So, I suspect that there will be more posts in the future on the subject.
