Design Fixation

What I read

In this paper, the authors conducted a set of experiments to see if, when given an example, engineers focus on the example, which then limits their exploration. They concluded that this phenomenon does occur, and call the phenomenon “design fixation.”

First, the authors introduce a model for how engineering works during early stages of problem solving (conceptual design). In this model, an engineer moves between two “spaces” of thinking. The first is a configuration space, which focuses on exploration at a practical level: things that can actually be made. The second is the concept space, which is focused on ideas and relationships, abstractions, and new ways of thinking. In the model, engineers move back and forth between configuration and concept spaces and within spaces as they work through a problem concept definition. There may be barriers that slow down or stop movement between spaces, and this would be problematic in practice. This is the focus of the remainder of the paper.

Next, the authors describe a historic basis for the idea of “fixation,” citing Maier’s work in the 1930s and Luchins’ work. This previous work identified two forms of fixation: “functional fixedness,” where previous use of an object limits the ability to see new uses, and “mental set” fixedness, where blocks due to fixation occur in the context of a situation. The authors describe a third type of fixation, “design fixation,” which occurs in a specific situation related to solving an engineering problem.

The authors describe the experimental method used during this research. In four related experiments, participants were provided with a design problem to solve. In each experiment, a control group worked through the problem, while a “fixation” group was provided with an example, and then worked through the problem.

In the first experiment, participants designed a bicycle rack for a car. The fixation group mimicked the form and approach of the example. In the next experiment, participants designed a measuring cup for blind people. The example given had clear problems. The results showed that the fixation group mimicked many of the problems presented in the example. In the third example, the participants designed a spillproof coffee cup. The example provided again had clear problems. In this experiment, the authors also measured the flexibility of new ideas, and the originality of the new ideas. Again, the fixation group included the same problems present in the example, and their solutions were more inflexible and less original.

The first three experiments were conducted with student participants. In the fourth experiment, the participants were professional engineers. The study was "sent to the head of company's design department. The exercises and questionnaires were distributed to the engineers in the department... the engineers were asked to administer the exercises to themselves." The results mirrored the results with students: design fixation occurred in this group as well.

Next, the authors discuss the findings, and implications of the findings.

The primary finding is that engineers fixate on examples provided to them, and that fixation limits their creative exploration.

One question that emerges is related to the duration of fixation—if it is a short term effect or a long term effect. Another question is if fixation is only detrimental, or if there are positive benefits to it as well. There may be distinction between fixation and lack of knowledge: the participants may have focused on the example because they lacked real understanding related to the problem space. And, the experiment focused on small problems, but an engineering problem is of “high-order complexity” and it’s unclear how the phenomenon impacts problems in a more realistic setting.

The work also raises the question as to if fixation occurs more or only at the start of solving a problem, or if and how it continues throughout the rest of the problem-solving process.

Finally, the authors indicate that, more broadly, they “believe that engineering design researchers should be more willing to recognize the broad multi-disciplinary character of the subject matter and to meet this difficult head-on”—that researchers should embrace cognitive psychology research in addition to engineer research.

What I learned and what I think

I see design fixation show up in a number of ways in my work.

First, I see it in the use of competitive analysis or competitive research or inspiration research, which purposefully introduces existing design solutions as a starting point for exploration. I’m not sure if the examples “block” exploration or “guide it” or both, but viewing what others have done absolutely constrains a solution-set.

I also see it in the way compressed timeframes in consulting lead to a use of a “playbook” approach. This isn’t fixation as a small part of the process; it pushes an outright copy/paste of a solution from another client in another discipline. This is similar to the use of pattern-language in strategy work, where we apply a solution from one domain we’ve worked on in another and the outcomes are very much related to one-another.

It’s absolutely present in the use of design systems in software development; it’s almost mandated. Fixate on the solutions we’ve provided, and use them with only subtle changes, so we can save development time and reinforce consistency.

I think it’s related to tool use, too; designers fixate on the things a tool can do, instead of a solution to a problem.

And it’s reallllly obvious in students. They want an example, because otherwise, they flounder around and don’t do anything. And then they copy the example. Which is a really good idea, because it’s a first try, and it’s about gaining a mental space for what the process looks like and the output feels like. The problem, though, is if that’s the start and end of it. There has to be a second try, where they don’t use an example. There’s something in here about bootcamp trash education. And there’s definitely something in here about portfolio development.

The word seems broad, though, and maybe something else is going on entirely. Use of components is mandated; it’s not a problem solving phenomenon as much as an organizational culture phenomenon. The paper seems to imply that fixation is happening unconsciously, but the use of components is purposeful. Knowing a tool doesn’t necessarily mean “fixating” on the tool as a form of exploratory block; it should be an enabler, not a disabler. I’m not sure about that one.

There’s something in here about constraints. Fixation as a limiter is positive if it makes a design problem tractable, and the blocking may actually be beneficial in that it minimizes the white canvas effect.

I’m running into the same issue around innovation and uniqueness, too. There’s no reason innovation should be a goal for most design and engineering problems. In fact, I’m pretty sure engineers cling to “if it ain’t broke, don’t fix it” as a philosophy, and that really means just using a solution that already exists. If exploring existing solutions makes me fixate on those, and those are good solutions—that should be a huge benefit, as long as I understand it should be a stopping condition: fixate, find, move on to other things.

A few things on process. I love the idea of giving the experimental exercise to a manger at a company and letting them administer it. It solves so many logistical problems, privacy concerns, and so-on. I will use this.

I also really appreciate the lack of citations in this work. Maybe it’s because there are none to cite (this was from 1991), but maybe it’s because they used restraint in the paper and included only the things most relevant. It seems that their goal was not to anticipate and head off citation criticism in order to publish, but instead, to actually document their results—publishing doesn’t seem like the goal here, at least not publishing for publishing sake. I don’t know if one could get away with such sparse and thoughtful citation selection any more, at least not if a goal is actually to publish in a meaningful journal.

Additionally, the paper made a case for why a quantitative assessment was important (counting fixation across groups and comparing them), and then did that simply. There is no over-the-top statistical analysis of the data; it’s just there.

And maybe most importantly on the paper itself: it’s in plain language. It’s easy to read, to understand, and to think about.