|
I am a firm believer in evidence-based everything. Design is not an
exception. So let me try to answer Birger's questions and perhaps address
some of his concerns.
But there are many confusions about what is meant by that term, so let me
give you my views. Of course, I have no evidence to support these views, so
take them as opinions, and interpret them any way you wish.
I will borrow from an unpublished article I am writing for "Design X" (not
to be confused with DesignX):
The Future of Design: When You Come to a Fork in the Road, Take It." Chapter
prepared for the tenth anniversary of the University of the Republic of San
Marino’s offering the degree of Bachelor in Design. The book is tentatively
titled *Design X. 10 years of design at San Marino with a look at the next
100. *
About craft design:
Traditional craft-based design had no need for formal evidence: the proof
of their efforts was visible to all who viewed it. The designs were guided
by the finely-honed intuitions of the designer and could be appreciated by
any discerning viewer. This approach worked as long as the designs were of
relatively simple things such as watches, home appliances, and furniture.
The introduction of computers, communication networks, powerful sensors and
display, even our most common everyday devices became more complex. People
became confused and frustrated: a new form of design was needed to cope
with these issues.
...
Modern HCD applies the findings of many fields. It is the bridge between
technology and people, applying the findings of the cognitive, behavioral
and social sciences through a process of doing and making, testing and
probing, experimenting to make things better, working with the specialists
from the relevant disciplines as well as the people for whom the designs
are intended. HCD designers do their research by continual designs,
carefully analyzing the situation, using each design as a way to test their
ideas in small, controlled ways, using the resulting evidence to guide
further, continual refinement.
Human-centered design moves us away from designer as guru. It moves us into
an important profession where we have systematic methods for discovering
true needs of people and society, for developing proposed solutions,
testing and refining them. We used to be an opinion-based field: today we
are an evidence-based field. We have become human-centered.
The point is, when we deal with complex items, intuitions no longer
suffice, even the intuitions of the most brilliant among us. One of the
most important parts of the iterative cycle of Human-Centered Design is the
use of evidence that allows for continued testing and refinement based upon
evidence.
Styling, shape and form can probably be done without evidence - in fact, I
can even argue that evidence-based design for aspects such as styling form,
composition, and layout can be damaging.
But where the intention of the design is the development of a
useful, understandable result, evidence is required. That is why when I
write, I always test --asking discerning readers (editors) who do not know
the topic i am writing about, but who can correct areas where i am unclear,
confusions, redundant, and wrong. (Yes, wrong because people who know
nothing about the topic often ask wonderful questions about what I said --
and in the case of the chapter I am quoting from, in one case questioning
why I didn't talk about some issues. )
Now, the testing I am talking about is informal, but in the traditional HCD
process, it is formal and rigorous. There is an entire field, usability
testing, dealing with part of these issuers.
Interaction design, for example, requires strong evidence based assessment.
Apple was once the leader in this, but it seems to have disbanded that
component of their design process and the result shows; people are now
confused and frustrated by Apple's iOS product (yes, I have evidence for
this).
It is relatively easy to devise something that behaves properly when
everything goes correctly. But when some unexpected event occurs, or when
the user does some operation not expected by the device, this is where the
design becomes difficult. Here is where the careful interactive designer
must determine how to make clear what has happened, what actions
are possible, how the previous state can be recovered, etc. This requires
good evidence-based design and extensive testing.
Just this past week, the CEO of a startup for a cooking device was proudly
demonstrating the very simple application on his cellphone that made
control of the cooker extremely simple. I agreed that it looked effective.
But when we asked someone to use it to cook a piece of meat, we
immediately discovered a flaw in the design. Although the design
looked completely clear and understandable to both the CEO and me, it
confused the user. The fix was simple, but it was yet another illustration
of the importance of using evidence to assess a design.
----
I am now working in healthcare. They care about metrics. They will not
release a change in procedure or organizational structure without evidence.
But here is an interchange i just had with a senior
medical administrator in the UC San Diego hospital system (who is also a
physician). (ED = Emergency Department)
(this interchange has been truncated from a much larger emails, but no
editing aside from that
Admin: ... For the ED project, how will you measure impact before and after
a design change related to the experience? ...
Don: When we start any project, we are careful not to start with predefined
measures or goals. The first step is observational. ...
Once we decide upon we decide upon the focus of the work -- the root core
issues -- we then can start more focussed observations and rapid
prototyping to detrmine the direction of our proposed interventions and
changes. ...
This may take another few months -- maybe longer depending upon
the complexity of the site we are assessing.
After that we consider a variety of suggested interventions -- we prototype
and test these.
Metrics show up here. Note we select metrics only after we have decided
upon the actions we will be testing. We do not wish to start with metrics.
We need to determine what issues are of maximum concern to us in order
to determine the metrics.
As for what metrics we use -- that all depends.
... we are careful not to jump to conclusions, to make sure we are studying
the right issues, and not to rush to metrics -- metrics come after we know
what we will be doing.
This interchange might also help clarify how I use evidence. I use all the
qualitative skills from my training and experience, coupled with
qualitative observations of current practices plus extensive
working sessions with all involved: physicians, nurses, staff, patients,
and families. We try to determine the root issues to
be addressed and then develop appropriate methods that might help solve
those issues. Only then do we develop the metrics.
=========
You might agree with me that metrics are needed for complex endeavors such
as restructuring activities for a Large Emergency department, but disagree
for other things. Yes, but I would take each case on its own merits.
The parts of design closest to art are least in need of evidence.
Radical new directions (radical innovation) can often not be
evaluated. Many radical innovations are poorly received at first and
thereby fail at most formal assessments even though they go on to change
the world. Here is where the intuitions of the designers and
business executives are critical. Note that going against data is
dangerous. Most radical ideas fail, and even those that succeed can take
decades to become successful, a long enough time that startups cannot wait
it out. The first hand-held scientific digital calculator (the HP 35) was
produced in 1972 despite all the evidence against it -- market studies and
other sources of evidence showed it would fail. But Hewlett and Packard
(who still ran the company) ignored the evidence and went forth with the
product: the rest is history. (HP never understood the consumer marketplace
and blew their lead in calculators. Among other things, they were
Engineers, so they insisted upon the use of reverse Polish Notation, which
severely limited sales to all except the most geekish among us (I plead
guilty to be one of those). So, HP was correct in its discounting the
initial evidence about the viability of the scientific calculator, but HP
was fundamentally wrong in ignoring marketing evidence about the future
direction of the calculator, especially when there was an established
market and competition. I can say more about their stupidity, but that
diverts me from the current discussion.)
If you feel an obligation to your client of company to help produce
successful products, you will use evidence. And if you don't feel that
obligation, your design firm will fail, or if you work for a company, you
will forever wonder why design is not properly respected.) If you teach in
a university, nobody will complain about your lack of
real-world understanding, but you will be doing a disservice to your
students.
In many fields, there is no choice. The design of a rocket engine? An
automobile engine? The dimensions of the inside of the vehicle?
The intelligibility of the navigation System? The addition to (or reduction
in) the danger of distraction for drivers caused by the design of the
interior discplays and controls of the automobile? The effectiveness of a
new meidal device?
I would not trust any design that did not have rigorous evidence behind it.
Do we require evidence for everything we do? No. But when designers make
claims, they need to back up those claims with evidence.
See my 2005 essays
Industrial Design: Claims Without Substance.
http://jnd.org/dn.mss/industrial_design_claims_without_substance.html
AND
"Industrial Design: Claims & Substance".
http://www.jnd.org/dn.mss/industrial_design_c.html
==========================
Aside: Some of you may wonder what is meant by the title of my essay "The
Future of Design: When You Come to a Fork in the Road, Take It." But that
remains the subject of a different discussion thread. And for that, I need
permission of the folks at San Marino to distribute copies or sections of
my essay. This is only appropriate courtesy.
On Sat, Jan 2, 2016 at 3:17 AM, Birger Sevaldson <[log in to unmask]
<[log in to unmask]" target="_blank">https://mail.google.com/mail/?view=cm&fs=1&tf=1&[log in to unmask]>
> wrote:
> Again I am getting worried when EBD is a theme here at this list. My
> worries are not caused by the use of evidence in design but by the
> spreading belief that the use of evidence in design will become main
> stream, replacing designing with problem solving based on evidence, and
> becoming the dominating way of designing.
>
> To me this is like doing a time travel back to the first generation of
> design methods. It is like nothing has happened in between. It’s like the
> design research community is in danger of drifting into an unjustified
> belief that we can solve complex socio-technical issues with hard methods.
>
> Or am I misinterpreting this?
>
My take:
I hope EBD does become dominant -- not for everything, but for interaction
and service design and for the design of complex systems.
And what is the matter of building upon the first generation of design
methods? They had good ideas. We have learned a lot since then, so we don't
act as if nothing has happened since then, but building upon one's history
is a time-honored methods of developing deep understanding and expertise. A
problem with the field of design is that history is often ignored. Science
builds upon previous work: testing, refining, keeping what works, modifying
what doesn't.
Of course I will now be accused of trying to make design into a science.
Well, yes: the use of evidence is the first step.
But remember, I did not say all of design, just those parts where evidence
and current scientific knowledge is relevant. Interactive design is one
such area. Experience design is only partially ready. Communication/graphic
design would really benefit from the use of the large base of scientific
knowledge of color, contrast, legibility, readability, comprehension,
learning sciences, and psychophysics of vision. (Apple's designs would
benefit by someone who knew that literature and who also tested their
designs.) But not all of communication/graphic design can be made into a
science. Creativity, aesthetics, etc, are not part of scisnce.
Don
Don Norman
Prof. and Director, DesignLab, UC San Diego
[log in to unmask]
<[log in to unmask]" target="_blank">https://mail.google.com/mail/?view=cm&fs=1&tf=1&[log in to unmask]>
designlab.ucsd.edu/ www.jnd.org <http://www.jnd.org/>
-----------------------------------------------------------------
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
-----------------------------------------------------------------
|
