The morning session
Balder, Stefan, and Erin
After a description of the three different projects and the approach they’ve taken, then a discussion of how we tend to simply select the methods we’re familiar with that are part of our disciplinary training. The example of Erin’s work with how poets are creative and then comparing that to the design process. She shows some examples of a recorded session of a study and a transcript then we break into groups and talk about how we would analyze it.
The second morning session with Nigel Cross
The title is Understanding Design Cognition:
Case Study: Philippe Starke as written about by Lloyd and ? from Delft.
Alessi asks him to design a kitchen tool, then Phillipe goes to a small island off Italy and has dinner, the Primo Piatto has baby squid and he begins sketching the lemon squeezer progressively and over the course of the meal he goes through several stages.
We are able to get a more general idea of the process from this even if Starke does not describe it himself it is something like Analogy, Evaluation, Improvement, Precedant (bringing in things from one’s past) then final product sketch.
Goel & ? (Neuroscientists) have a patient who’s an architect and they ask him to redesign an office space. Essentially he’s unable to complete this fairly simple task, for control they have a similarly trained and experienced architect who is able to do it. 2 90 minute sessions.
Protocol studies:
Design a new litter colection system for the dutch railways system. Again done with colleagues at Delft. They were familiar with the current system. People are brought in for 90 minute design session. They had a large file of information from which they could draw if they asked for. There were lots of stakeholders, the railways, the company that produced it, the janitors, and others as well. 9 experienced industrial designers. Their designs were then normalized in terms of how they were presented and then these designs were given to others to assess. These people were from the University and used to grading design concepts. Creativity, materials were all factors. The problem that emerged early was that of newspapers. This came from the survey of cleaners they identified this as a problem. It’s the largest ingredient of the litter (bulkiest). Fill up the bines, they are left behind on the seats. The railway company wants to be environmentally friendly.
All of the designers thought that separating the newspapers would be a good idea, they all thought that this was a good idea and no one else would have this. This is a good example of how it may be new to you, but it’s certainly not new to the world.
Designers tend to force some kind of pattern in the information like the example of the pixelated dog. This shows the co-evolution of the problem and solution space. You get a formulation of the problem that is partial then a partial solution then back to a developed problem structure then back to the solution and back again. A scatter plot of creativity v overall and they don’t always correlate.
From an analysis of think aloud protocol one researcher found the following activities.
- Gather data
- Assess value and vialifity of data
- Identify constraints and reqs
- mofel behavbioud and environment
- Define problems and possibilities
- Generate partial solutions
- Evaluate solutions
- Assemble a coherent solution.
Then some scattergrams of some of the levels of some of these activities. The solution-driven ones have more generate and the problem driven ones talk more about identifying.
The Role of Sketching in Design
Norman Foster sketch of the Gherkin and the restaurant at the top.
Jack howe says, “If I’m stuck I draw something. Even it’s is silly I draw it. The act of drawing seems to clarify my thoughts.”
Sketches of Leonardo Davinci, the sketch is not of the thing being designed but many aspects of it, possibly in rapid succession or simultaneously. Today architects do similar things, James Sterling is an example used here with a page from his sketchbook. Alvar Aalto is the next example with lots of different parts of it all in a big sheet. Frank Gehry is the next example, and there are two different kinds of sketches one that seems to include details, but the other is more of the overall form. Richard MacCormac and then Gordon Murray. The drawing is a way of finding out what is good.
Sketches handle different levels of abstraction simultaneously.
Sketches enable identification and recall of relevant knowledge.
Sketches assist problem structuring through solution attempts.
Sketches promote the recognition of emergent features and properties.
What expert designers say about designing:
Santiago Calatrava
“To start with yhou see the thing in your mind and it doesn’t exist on paper, and then you start making simple sketches and organizing things, and then you start doing later after layer.
This is Design Thinking as Reflection.
Kenneth Grange
The designer’s job is to produce the unexpected. No brief of itselft…… you have to find the plums.
Design Thinking as opprtunitistic.
Richard MacCormac
I don’t think you can deisgn anything just by absorbving information and then hoping to synthesise it into a solution.
I or someone else will come up with an idea that seems powerful enough” then
Design thinking is an exploratory process, and as conjectural.
Geoffrey Harcourt
The solution I came up with wasn’t a solution to the problem at all. But when the chair was actually put togetehr ina way it solved the problem quite well. but from a completely different point of view.
Design Thinking as emergent.
Ted Happold
I have perhaps one real talent, which is that i don’t mind at all living in the area of total uncertainty.
He was the chief engineer of Sydney Opera House and Pompideiu center.
Design Thinking as ambiguous.
Denys Lasdun
Design Thinking as appositional. It doesn’t follow logically from what was said before, but it fits the solution.
Mies van der Rohe
The cleint wasn’t very happy at first. But then we smoked some good cigars and we frank some glasses of good rhine wine and he liked it.
He convinced the client over cigars and wine.
Design Thinking is rhetorical.
Expertise in problem solving
- tackle the problem in the ‘easiest’ way
- Accept teh problem rules
- adopt standard problem representation
- re-use previous solutions
Expertise in Design Thinking
- Tackle the problem in a difficult way
- Challenge problem rules (Murray)
- Construct novel problem representation
- Create new solutions from first principles.
Jean Bernard is asking a question about how sketches may be part of working memory, helping them remember things, Nigel says that yes, there has been some work on that, and sometimes it is and sometimes not.
Lunch, in the hotel restaurant
First afternoon session
Real-world studies of design cognition: using cognitive ethnography to understand design behavior
by Linden J Ball & Bo Christensen
Ethnography can be a useful thing for design research and this is from the cognitive science so-called cognitive ethnography, a term Linden coined about 10 years ago. Reliability and validity is very different in this kind of ethnography.
Overview
1) Antropology and the origins of ethnography
2) Sociology and Ethnography
3) Ethnomethodology and ethnography
4) Ethnography as a ‘radial category’
5) Cognitive ethnography
Traditional Ethnography: what is it?
A form of investigative fieldwork and analysis: Malinowski was the prototypical example. He was forced to live there initially alone, but then he got lonely enough to start interacting with them. It involved immersive observation. It invovled an analytic mentality in something that was “in situ”.
Sociologists from the 1940s-1970s adopted this same method, often studied smaller, disposessed groups, giving them a voice. i.e. hell’s angels etc.
William Foote Whyte (1943) Street Corner Society is an example.
Garfinkle’s 1967 book critiqued ethnography, saying it misses the activities the “interactional what” of how/what groups do together. John Hughes’ book “making work visible” is a good example.
Ethnomethodology and systems design
HCI and CSCW has taken up this kind of thing in the 80s and 90s. Every grant that came through seemed to have a sociologist attached to the project doing ethnography, where this would produce requirements for the computational people could then turn into a system. Bently et al’s air traffic controller study would be a prototypical example.
Linden’s paper “Putting ethnogaphy to work….. in the IJHCI. The idea of a radial category from Lakoff is used. What if enthnography is a radial category? There are some prototypical characteristics of ethnography: situatedness, richness, participant autonomy, openness, personalization, reflexivity, self-reflection, intensity, independence, historicism.
Cognitive Ethnography identifies the following problems with traditional ethnography for design research: Intensity (You don’t have years, usually only weeks) independence (Whereas theories need to e tested and validated) non-verifiability.
This is importnat if you’re a congnitive psychologist, but I wonder how much this matters if you are not. This doesn’t lessen the interest however for those that are, as well as knowing more about the history of it.
A cognitive ethnography of design re-use by ball & Omerod (2000). Identifying the information unit of re-use. This is a prototypical example of cognitive ethnography.
Stefan asks what about Ed Hutchins Cognition in the wild and his notion of cognitive ethnography, Linden says that there are similarities but perhaps his notion is a bit more laid out but they seem to be compatible.
Studying deisgn coginition int he real-world- combingin ethnography with protocol analysis.
Overview: 1) Studying design cognition in the real world 2) th in vivo methodology 3)DTRS7 4)The use of analogy in design 5) the use of mental simulation in design.
DTRS7 design meetings, 4 camera angles recording in situ real design meetings lasting 1-3 hours and then all materials are collected. 7 hours of total video and then segmentation by turn taking for a total of 3886 data points.
So what is in-vivo methodology? Dunbar 2000-2001, involves ethnography & Protocol Analysis (PA). Finding and recording suitable real-world design situations, where natural design dialogue occurs., transscription, segmentation and coding along PA lines. No special instruction to ‘think aloud’ involved.
Quite often the amount of data is massive so you need to make selections of the data to code. Segmentation will depend on what is being studied.
So the advantages of in vivo are that it has ecological validity, the possibility of stdying cognition live in the real world etc. the disadvantages are that it may turn into a single case study, it takes a long time etc.
A coffee break then practical exercises with cognitive ethnography and in vivo methodologies.
