Hello Stuart and list!
A design space is the space of possible variation in any design. In
professional practice, the term "design space" is rather commonplace.
Designers often sketch out the space of possible design variability as a
way to generate ideas and avoid fixation on a particular point solution.
Design spaces are also used for classifying existing designs, in order to
suggest unexplored options (Shaw, 2012). However, because the potential
design space of any endeavor is infinite, significant creativity is
required to bound design spaces (with constraints) and to identify which
(of an infinite many) dimensions to vary, in order to produce meaningful
outcomes. Design dimensions don't come for free -- they must be discovered.
Design dimensions include both input parameters (e.g., font size) or its
output functions (e.g., the "readability" or "playfulness" of a font).
The first instance of the term “design space” that I could find was in
1965, when it was already commonplace in structural engineering.
“Many modern techniques for the optimum synthesis of structures are based
upon the concept of a design space. In an S-dimensional design space, each
Cartesian coordinate axis represents a design variable, and thus a point in
this space D = (d1,d2,…ds) represents a design. Furthermore, each design
has associated with it a value of the objective function (weight, cost,
etc.) and has a behavior as determined by the application of an analysis
method.” (Fox, 1965)
This work was preceded by the notion of a "Design Parameter Space" (Schmit,
1960). But even before that, statisticians had developed a generalized
statistical understanding of how to determine the effects of design factors
on outcomes, via experimentation. They didn't call it a design space, but
talked about manipulating the "factor space", the "design matrix" or even
the "independent variable matrix."
For instance, in 1951, Box and Wilson write:
“In the whole k dimensional factor space, there is a region R, bounded by
practical limitation to change in the factors, which we call the
experimental region. The problem is to find, in the smallest number of
experiments, the point ...[where the outcome is] a maximum or a minimum. In
our field, yield, purity or cost of product are the responses which have to
be maximized (or minimized in the case of cost). The factors affecting
these responses are variables such as temperature, pressure, time of
reaction, proportions of the reactants.”
Thus, the core idea of a design space is that a single design can be seen
as a point solution in a vast space of possible solutions that are
organized by dimensions (design factors) of design variability. This is a
very simple and powerful idea. In no small part, the power comes from the
fact that there are long established statistical mechanisms for
demonstrating the causal effects of various design factors on outcomes. As
design is often seen as a field bereft of empirical evidence, the notion of
a design space is provides a powerful framework for statistical theory
development.
In 1971, Bell and Allen Newell described the design space of computers,
which they referred to as a computer space:
“There are, then, three main ways to classify or describe a computer
system: according to its function, its performance, or its structure. Each
consists in turn of a number of dimensions. It is useful to think of all
these dimensions as making up a large space in which any computer system
can be located as a point.”
They asserted that these terms could be applied to any designed system: for
instance, an automobile might have a function like goods transport or
racing, performance measures like speed or cargo capacity and structure
like the number of wheels or color. They note that a subsystem’s
performance might become a system’s structure; for instance, horsepower can
be the performance of an engine but treated as a structural element for the
car. They say:
“Structure determines performance, although from the standpoint of design,
of course, causality runs the other way: from function to performance to
structure.”
Note that a vast number of structural design dimensions can be simplified
into a much smaller number of functional design factors. For instance, in
my own experimental design work, I've found that the effects of many game
design structural features can be modeled using the functional dimension of
"difficulty" and "novelty." The opportunity afforded by large online
experiments is really phenomenal for the field of interaction design, as it
allows for much easier quantitative measurement of design spaces and
empirical theory development. In my work, I use large scale online design
experiments to explore the design space of video games and to develop
generalizable theory, as in:
1.
https://www.researchgate.net/publication/262350490_Optimizing_challenge_in_an_educational_game_using_large-scale_design_experiments
2.
https://www.researchgate.net/publication/282251802_Interface_Design_Optimization_as_a_Multi-Armed_Bandit_Problem?ev=prf_pub
3. my thesis:
https://www.researchgate.net/publication/282251834_Optimizing_Motivation_and_Learning_with_Large-Scale_Game_Design_Experiments?ev=prf_pub
I hope these references are useful. I'm currently working on a paper about
"Design Spaces, Experimental Design and Design Theory", which obviously
isn't complete or I would have just sent that :)
Thanks!
Derek Lomas
Design Fellow
The Design Lab, UCSD
*References:*
Shaw, M. (2012). The Role of Design Spaces. IEEE Software, 29, 46–50.
Schmit, L. A. (1960). Structural Design by Systematic Synthesis. In *2nd
Conference on Electronic Computation, American Society of Civil Engineers* (pp.
105–132).
R.L.Fox. (1965). Constraint surface normals for structural synthesis
techniques. AIAA J., 3(8), 1517–1518.Box, G., & Wilson, K. (1951). On the
Experimental Attainment of Optimum Conditions. *Journal of the Royal
Statistical Society. Series B (Methodological)*, *13*(1), 1–45.
Bell, C. G., & Newell, A. (1971). Computer Structures: Readings and
Examples.
> > On Feb 17, 2016, at 6:27 PM, Stuart Reeves <[log in to unmask]> wrote:
> >
> > I wondered whether anyone here had any references to any discussions
> about the term "design space".
>
>
>
-----------------------------------------------------------------
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
-----------------------------------------------------------------
|
