Dear Gunnar,

David Durling interpreted your question differently than I did. There was some ambiguity to it. David Durling wrote: “One American colleague insisted that ALL PhD students should undertake a class on statistics.” Gunnar Swanson asked, “Including English literature, art history, theology, geometry. . . ?”

As I read it, you were not asking whether doctoral students in design needed these other disciplines. Rather, I thought you were asking whether all students in all disciplines need statistics. 

While I’m not sure that all students in all fields need statistics, I think that there is a reasonable answer with respect to statistics in the PhD curriculum for design. 

For the PhD to serve as a research training degree, it should provide students with an overview of the methods they are likely to encounter in their field. Since people with a PhD are later expected to develop the skills that they will need to supervise a later generation of doctoral students, this includes a far greater range of research skills and methods than the student may ever need in her or his own research. 

In the contemporary university, the PhD degree is, in effect, a license to teach research methods and train research students. This establishes criteria for the PhD that may serve no purpose in the research of the student who earns the PhD. People who may not need to use statistics in their own work may need to know *about* statistical problems and statistical inference to serve as skilled doctoral advisors for others. Therefore, when we earn a PhD, we must often understand and be able to demonstrate skills that we may never use in our own research. We require these skills for our future students.

Let me offer an analogy. I once spent two weeks with a master chef who prepared magnificent meals for a conference while he ate only tuna sandwiches on toast. I asked him why. He told me that cooking was an art form for him, but he did not want to eat gourmet food after working with it all day. He told me that he tasted most recipes only twice. The first time was when another chef taught him to cook it so that he would know how it should taste. The second time was when he prepared it to make sure that it tasted the same way. He had the equivalent of a photographic memory for tastes, and tasting was a tool in his work. 

Every evening after he finished cooking, he ate a tuna sandwich. This was the case for the two weeks that I watched him cook. The “twice only, never again” principle may have been a slight exaggeration. Perhaps he ate gourmet food on vacation — or when visiting other chefs. Even so, the principle was clear.

Some of the skills we learn to earn a PhD are like the recipes in my friend’s repertoire. We master them so that we can cook with them for our students. We may never eat them again ourselves. And to be honest, we may never really master the methods we only study without using. I struggled with statistics. I rarely use statistical inference in my work, and I’m not very good at statistics. Nevertheless, I recognize problems where statistics may be useful to others — and I understand enough about some of those problems to turn to people with the appropriate expertise. 

If we do not understand these issues enough to work with the skills our doctoral students may need, we are not properly prepared to teach research methods or to train research students. As research teachers and supervisors, our own research needs come second to the needs of our students.

As long as the PhD is a license to teach and supervise, training for the PhD degree has specific criteria that may affect no other research we ever do. This is why PhD students in fields such as design may benefit from statistics. 

Because this is the PhD-Design list, I don’t think that we have many people getting a PhD in English literature, art history, theology, and geometry — or supervising such a degree. Even so, it may interest you to know that there are some forms of statistical research in each of these fields.

For design, however, many research students will need to draw responsible statistical inferences. This is also true in teaching and even in professional practice. 

Designers often need to know whether the statistical claims are responsible or irresponsible, correct or incorrect. In some cases, they’ll need to know whether statistical claims are truthful or dishonest. In the business world, statistical inference is often a key difference between the argumentation of people in finance or marketing as against the arguments of designers. Designers also face a problem when they address issues in statistical quality control for industrial processes.  

Relatively few PhD programs in design require statistics in the curriculum. For that matter, nearly none of the PhD programs in design have people on staff who can properly teach statistics or help students to master them. This leaves the vast majority of our PhD graduates incapable of using or understanding a valuable tool — and it is a handicap when they move into doctoral supervision and curriculum design.

Yours,

Ken

Ken Friedman | Editor-in-Chief | 设计 She Ji. The Journal of Design, Economics, and Innovation | Published by Tongji University in Cooperation with Elsevier | URL: http://www.journals.elsevier.com/she-ji-the-journal-of-design-economics-and-innovation/

Chair Professor of Design Innovation Studies | College of Design and Innovation | Tongji University | Shanghai, China ||| Email [log in to unmask] | Academia http://swinburne.academia.edu/KenFriedman | D&I http://tjdi.tongji.edu.cn 

—


-----------------------------------------------------------------
PhD-Design mailing list  <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------