When is behavioural science good design?

22 June 2018

By Ed Gardiner

I work between the worlds of behavioural science and design and often get asked the question 'so when is behavioural science good design?'

This article attempts to answer this question, reconcile some of the differences and discuss how and when these two fields can work together to be a powerful problem-solving force.

What do I mean by behavioural science and design?

Behavioural science is broadly the study of how and why people behave the way they do, drawing on insights and methods from psychology, economics and neuroscience.

It includes fields such as cognitive science, behavioural economics, and judgement and decision-making, providing the basic principles for what causes?- and how to cause - ?specific behaviours to occur.

Design has multiple definitions. At a functional level, design is a plan for arranging elements to accomplish a particular purpose. For example, while the purpose of a kettle is to boil water, one can better accomplish that purpose by arranging the elements differently, as a hot tap.

Design principles? - ?for example, Dieter Rams’ 10 principles for good design or the UK Government’s design principles - provide the foundations for what good looks like across different industries.

At a process level, design is a creative approach to problem-solving that brings together what is desirable from a human point of view with what is technologically feasible and economically viable (see Design Thinking). This approach - design thinking or human-centred design - provides a process and set of tools for people to follow in order to practice good design. The IDEO design process and the Design Council double diamond are among the best known of an ever expanding set.

Shining example: The Design Council's double diamond process

At a higher level, design is about determining people’s true, underlying needs, and creating products and services together that help them (see How Apple is Giving Design a Bad Name).

Certain abilities are associated with this design mind-set, for example navigating ambiguity, moving between concrete and abstract, and understanding collective needs (see Let’s stop talking about THE design process). Not everyone who is a designer has this mind-set, and not everyone who has this mind-set is a designer.

So, when is behavioural science good design?

From a functional perspective, one could argue that if the behavioural intervention - ?a product, service, communication or other interaction developed using behavioural science to cause a specific behaviour to occur? - performs the desired purpose by changing the desired outcome, then it is good design. ‘Behavioural’ or not, it’s astonishing how many products and services fail at this basic level.

However, judging an intervention purely based on outcomes may ignore the value of good design, both in terms of its contribution to improving outcomes, and broader enjoyment.

Is the Behavioural Insight Team’s seminal example of drawing attention to social norms in a tax letter to increase response rates good design (see Applying behavioural insights to reduce fraud, error and debt and Ivo Vlaev's research)? Did it change the desired outcome? Yes. Did it conform to all of Rams’ principles? No? - ?it’s not aesthetic (to be fair, most letters are ugly). Would it have been more effective - or enjoyable - if it had? Maybe.

It’s harder to make a judgement of good design based on process because no one process is suitable for all problems. Behavioural science is a top-down process: state the desired outcomes, research and list the behaviours needed to achieve the outcomes, match the behaviours to intervention types and empirically test the ideas.

Design is a more bottom-up process: explore the problem, system or needs to be addressed, define the problem and desired outcomes, create or support the development of ideas, prototype and test.

One key difference is when the outcomes are defined. This generally happens before the exploration stage in behavioural science, but after the exploration stage in design.

Designers might question how you can know you are solving the right problem without initial exploration. A good design process starts by understanding the system and empathising with the people you are trying to benefit.

Related course: Behavioural Science in Practice

Behavioural scientists might question how you can genuinely determine what people want and need. A good behavioural science process starts by defining a problem statement around a desired behaviour.

A good combined process might depend on how well defined the problem is.

Well defined problems have specific goals, clearly defined solution paths, and clear expected solutions. Ill-defined problems are the opposite, with many interdependent causes and multiple solutions.

Applying a top-down, behavioural science process to an ill-defined, complex - or ‘wicked' - ?problem, without understanding the broader system or prototyping potential solutions, begins to look like bad design.

On the flip side, applying a bottom-up design process to a well-defined problem, without any scientific evidence of what is influencing behaviour and without testing solutions using experimental methods, begins to look like bad behavioural science.

At a higher level, determining whether behavioural science is good design depends on the practice of the individuals involved. In a consulting setting, both behavioural scientists and designers serve the interests of themselves, the client or funder, the audience and society, depending on the commercial and social aims of the project.

Many designers champion the importance of design as an open, inclusive set of problem-solving skills. In 1964, Ken Garland along with 20 other authors, published the manifesto, First Things First, as a call for their skills to be put to more worthwhile use. It was republished in First Things First Manifesto 2000 by a new group, again calling for more useful, lasting and democratic forms of communication.

I’m unsure whether behavioural science ‘stands’ for anything beyond changing a specific behaviour to achieve a specific outcome. A common concern or criticism is that behavioural science can be used to for both good and bad. The same is true of design. The First Things First manifestos were a reaction to increasing consumerism and the use of design for trivial purposes that contribute little to national prosperity.

What is a good nudge?

Assuming that good design is associated with this positive mind-set, I think behavioural science can be judged against this standard of good design. To paraphrase Garland, should we be applauding the work of behavioural scientists who have flogged their skill and imagination to sell such things as detergent and deodorants, or should we be prioritising more useful and lasting interventions?

So what next? As behavioural scientists increasingly develop sophisticated products, services and communications, they are increasingly using design, whether implicitly or explicitly.

Rather than judging the quality of this design based on existing standards, maybe there is a need to update and combine existing principles, processes and manifestos in a more formal manner.

What is clear though, is there must be a stronger relationship between the two. Many years ago, the economist John Maurice Clarke warned that when economists attempt to ignore psychology, in doing so they will not avoid psychology but make their own, and it will be bad psychology.

The same could be said of behavioural scientists and design.

Ed Gardiner is Behavioural Design Lead and teaches Behavioural Design on the Doctorate in Business Administration, Innovation and Creativity in Organisations on the Executive MBA. He also lectures on Behavioural Science for the Manager on the Executive MBA (London) and Problem-Solving in Organisations on the Undergraduate programme.

Follow Ed Gardiner on Twitter @edwardgardiner

For more articles like this download Core magazine here.