Don Norman: Designing For People

Nielsen Norman Group

Why Design Education Must Change

This is the first of several essays on design education:

  1. This was originally published on Core77.
  2. The second essay was also published on Core77: Design Education: Brilliance Without Substance.
  3. (Also available at this website, jnd.org)
  4. The third essay was written with Scott Klemmer and published on LinkedIn: State of Design: How Design Education Must Change . (Also available at this website, jnd.org)
Traditionally what designers lack in knowledge, they make up for in craft skills. Whether it be sketching, modeling, detailing or rendering, designers take an inordinate amount of pride in honing key techniques over many years. Unfortunately many of these very skills have limited use in the new design domains. (Core 77 columnist Kevin McCullagh.)

I am forced to read a lot of crap. As a reviewer of submissions to design journals and conferences, as a juror of design contests, and as a mentor and advisor to design students and faculty, I read outrageous claims made by designers who have little understanding of the complexity of the problems they are attempting to solve or of the standards of evidence required to make claims. Oftentimes the crap comes from brilliant and talented people, with good ideas and wonderful instantiations of physical products, concepts, or simulations. The crap is in the claims.

In the early days of industrial design, the work was primarily focused upon physical products. Today, however, designers work on organizational structure and social problems, on interaction, service, and experience design. Many problems involve complex social and political issues. As a result, designers have become applied behavioral scientists, but they are woefully undereducated for the task. Designers often fail to understand the complexity of the issues and the depth of knowledge already known. They claim that fresh eyes can produce novel solutions, but then they wonder why these solutions are seldom implemented, or if implemented, why they fail. Fresh eyes can indeed produce insightful results, but the eyes must also be educated and knowledgeable. Designers often lack the requisite understanding. Design schools do not train students about these complex issues, about the interlocking complexities of human and social behavior, about the behavioral sciences, technology, and business. There is little or no training in science, the scientific method, and experimental design.

Related problems occur with designers trained in engineering, for although they may understand hard-core science, they are often ignorant of the so-called soft areas of social and behavioral sciences. The do not understand human behavior, chiding people for not using technology properly, asking how they could be so illogical. (You may have all heard the refrain: "if only we didn't have people, our stuff would work just fine," forgetting that the point of the work was to help people.) Engineers are often ignorant of how people actually behave. And both engineers and designers are often ignorant of the biases that can be unwittingly introduced into experimental designs and the dangers of inappropriate generalization.

The social and behavioral sciences have their own problems, for they generally are disdainful of applied, practical work and their experimental methods are inappropriate: scientists seek "truth" whereas practitioners seek "good enough." Scientists look for small differences, whereas designers want large impact. People in human-computer interaction, cognitive engineering, and human factors or ergonomics are usually ignorant of design. All disciplines have their problems: everyone can share the blame.

Time to change design education
Where once industrial designers focused primarily upon form and function, materials and manufacturing, today's issues are far more complex and challenging. New skills are required, especially for such areas as interaction, experience, and service design. Classical industrial design is a form of applied art, requiring deep knowledge of forms and materials and skills in sketching, drawing, and rendering. The new areas are more like applied social and behavioral sciences and require understanding of human cognition and emotion, sensory and motor systems, and sufficient knowledge of the scientific method, statistics and experimental design so that designers can perform valid, legitimate tests of their ideas before deploying them.

Designers need to deploy microprocessors and displays, actuators and sensors. Communication modules are being added to more and more products, from the toaster to the wall switch, the toilet and books (now called e-books). Knowledge of security and privacy, social networks, and human interaction are critical. The old skills of drawing and sketching, forming and molding must be supplemented and in many cases, replaced, by skills in programming, interaction, and human cognition. Rapid prototyping and user testing are required, which also means some knowledge of the social and behavior sciences, of statistics, and of experimental design.

In educational institutions, industrial design is usually housed in schools of art or architecture, usually taught as a practice with the terminal degree being a BA, MA, or MFA. It is rare for in design education to have course requirements in science, mathematics, technology, or the social sciences. As a result the skills of the designer are not well suited for modern times.

The Uninformed Are Training the Uninformed
My experience with some of the world's best design schools in Europe, the United States, and Asia indicate that the students are not well prepared in the behavioral sciences that are so essential for fields such as interaction and experience design. They do not understand experimental rigor or the potential biases that show up when the designer evaluates their own products or even their own experimental results. Their professors also lack this understanding.

Designers often test their own designs, but with little understanding of statistics and behavioral variability. They do not know about unconscious biases that can cause them to see what they wish to see rather than what actually has occurred. Many are completely unaware of the necessity of control groups. The social and behavioral sciences (and medicine) long ago learned the importance of blind scoring where the person scoring the results does not know what condition is being observed, nor what is being tested.

The problem is compounded by a new insistence by top research universities that all design faculty have a PhD degree. But given the limited training of most design faculty, there is very little understanding of the kind of knowledge that constitutes a PhD. The uninformed are training the uninformed.

There are many reasons for these difficulties. I've already discussed the fact that most design is taught in schools of art or architecture. Many students take design because they dislike science, engineering, and mathematics. Unfortunately, the new demands upon designers do not allow us the luxury of such non-technical, non science-oriented training.

A different problem is that even were a design school to decide to teach more formal methods, we don't really have a curriculum that is appropriate for designers. Take my concern about the lack of experimental rigor. Suppose you were to agree with me - what courses would we teach? We don't really know. The experimental methods of the social and behavioral sciences are not well suited for the issues faced by designers.

Designers are practitioners, which means they are not trying to extend the knowledge base of science but instead, to apply the knowledge. The designer's goal is to have large, important impact. Scientists are interested in truth, often in the distinction between the predictions of two differing theories. The differences they look for are quite small: often statistically significant but in terms of applied impact, quite unimportant. Experiments that carefully control for numerous possible biases and that use large numbers of experimental observers are inappropriate for designers.

The designer needs results immediately, in hours or at possibly a few days. Quite often tests of 5 to 10 people are quite sufficient. Yes, attention must be paid to the possible biases (such as experimenter biases and the impact of order of presentation of tests), but if one is looking for large effect, it should be possible to do tests that are simpler and faster than are used by the scientific community will suffice. Designs don't have to be optimal or perfect: results that are not quite optimum or les than perfect are often completely satisfactory for everyday usage. No everyday product is perfect, nor need they be. We need experimental techniques that recognize these pragmatic, applied goals.

Design needs to develop its own experimental methods. They should be simple and quick, looking for large phenomena and conditions that are "good enough." But they must still be sensitive to statistical variability and experimental biases. These methods do not exist: we need some sympathetic statisticians to work with designers to develop these new, appropriate methods.

When Designers Think They Know, But Don't
Designers fall prey to the two ailments of not knowing what they don't know and, worse, thinking they know things they don't. This last condition is especially true when it comes to human behavior: the cognitive sciences. Designers (and engineers) think that they understand human behavior: after all, they are human and they have observed people all their lives. Alas, they believe a "naive psychology": plausible explanations of behavior that have little or no basis in fact. They confuse the way they would prefer people to behave with how people actually behave. They are unaware of the large experimental and theoretical literature, and they are not well versed in statistical variability.

Real human behavior is very contextual. It is readily biased by multiple factors. Human behavior is driven by both emotional and cognitive processes, much of which is subconscious and not accessible to human conscious knowledge. Gaps and lapses in attention are to be expected. Human memory is subject to numerous biases and errors. Different memory systems have different characteristics. Most importantly, human memory is not a calling up of images of the past but rather a reconstruction of the remembered event. As a result, it often fits expectations more closely than it fits reality and it is easily modified by extraneous information.

Many designers are woefully ignorant of the deep complexity of social and organizational problems. I have seen designers propose simple solutions to complex problems in education, poverty, crime, and the environment. Sometimes these suggestions win design prizes (the uninformed judge the uninformed). Complex problems are complex systems: there is no simple solution. It is not enough to mean well: one must also have knowledge.

The same problems arise in doing experimental studies of new methods of interaction, new designs, or new experiences and services. When scientists (and designers) study people, they too are subject to these same human biases, and so cognitive scientists carefully design experiments so that the biases of the experimenter can have no impact on the results or their interpretation. All these factors are well understood by cognitive scientists, but seldom known or understood by designers and engineers. Here is a case of not knowing what is not known.

Why Designers Must Know Some Science
Over the years, the scientific method evolved to create order and evaluation to otherwise exaggerated claims. Science is not a body of facts, not the use of mathematics. Rather, the key to science is its procedures, or what is called the scientific method. The method does not involve white robes and complex mathematics. The scientific method requires public disclosure of the problem, the method of approach, the findings, and then the interpretation. This allows others to repeat the finding: replication is essential. Nothing is accepted in science until others have been able to repeat the work and come to the same conclusion. Moreover, scientists have learned to their dismay that conclusions are readily biased by prior belief, so experimental methods have been devised to minimize these unintentional biases.

Science is difficult when applied to the physical and biological world. But when applied to people, the domain of the social sciences, it is especially difficult. Now subtle biases abound, so careful statistical procedures have been devised to minimize them. Moreover, scientists have learned not to trust themselves, so in the social sciences it is sometimes critical to design tests so that neither the person being studied nor the person doing the study know what condition is involved - this is called "double blind."

Designers, on the whole, are quite ignorant of all this science stuff. They like to examine a problem, devise what seems to be a solution, and then announce the result for all to acclaim. Contests are held. Prizes are awarded. But wait-- has anyone examined the claims? Tested them to see if they perform as claimed? Tested them against alternatives (what science calls control groups), tested them often enough to minimize the impact of statistical variability? Huh? say the designers: Why, it is obvious - just look - What is all this statistical crap?

Journals do not help, for most designers are practitioners and seldom publish. And when they do, I find that the reviewers in many of our design journals and conferences are themselves ignorant of appropriate experimental procedures and controls, so even the published work is often of low quality. Design conferences are particularly bad: I have yet to find a design conference where the rigor of the peer review process is satisfactory. The only exceptions are those run by societies from the engineering and sciences, such as the Computer-Human Interaction and graphics conferences run by the Institute of Electronic and Electrical Engineers or the Computer Science society (IEEE, ACM and the CHI and SIGGRAPH conferences). These conferences, however, favor the researcher, so although they are favorite publication vehicles for design researchers and workers in interaction design, practitioners often find their papers rejected. The practice of design lacks a high quality venue for its efforts.

Design Education Must Change
Service design, interaction design, and experience design are not about the design of physical objects: they require minimal skills in drawing, knowledge of materials, or manufacturing. In their place, they require knowledge of the social sciences, of story construction, of back-stage operations, and of interaction. We still need classically trained industrial designers: the need for styling, for forms, for the intelligent use of materials will never go away.

In today's world of ubiquitous sensors, controllers, motors, and displays, where the emphasis is on interaction, experience, and service, where designers work on organizational structure and services as much as on physical products, we need a new breed of designers. This new breed must know about science and technology, about people and society, about appropriate methods of validation of concepts and proposals. They must incorporate knowledge of political issues and business methods, operations, and marketing. Design education has to move away from schools of art and architecture and move into the schools of science and engineering. We need new kinds of designers, people who can work across disciplines, who understand human beings, business, and technology and the appropriate means of validating claims.

Today's designers are poorly trained to meet the today's demands: We need a new form of design education, one with more rigor, more science, and more attention to the social and behavioral sciences, to modern technology, and to business. But we cannot copy the existing courses from those disciplines: we need to establish new ones that are appropriate to the unique requirements of the applied requirements of design. 

But beware: We must not lose the wonderful, delightful components of design. The artistic side of design is critical: to provide objects, interactions and services that delight as well as inform, that are joyful. Designers do need to know more about science and engineering, but without becoming scientists or engineers. We must not lose the special talents of designers to make our lives more pleasurable.

It is time for a change. We, the design community, must lead this change.

Now read my second and third essays:
  • The second essay was also published on Core77: Design Education: Brilliance Without Substance.
  • (Also available at this website, jnd.org)
  • The third essay was written with Scott Klemmer and published on LinkedIn: State of Design: How Design Education Must Change . (Also available at this website, jnd.org)
  • See all my Linked-In essays.

    Essays

    Books