7th April 2018

Stealth Health: Behaviour Change By Design

Salad or Snickers? Brisk run or box set binge? Sometimes despite our best intentions, we don’t pick the healthiest options. And often it feels that you’re fighting not just your own learned behaviours but the might of a zillion companies trying to sell you their naughty but oh so tasty snacks. If only there was something that could be done…

Enter Professor Marcus Munafò, a biological psychologist based at the University of Bristol. Alongside a team at University of Cambridge, Marcus has been investigating the factors at play when we’re making decisions and whether design can help us to live healthier. He’s found that telling us fickle humans that things are really bad for us isn’t hugely effective, but that subtle changes (like changing the shape of pint glasses for example) might encourage us to make healthier decisions by stealth.

The team has just won four years of funding from the Wellcome to pursue the research in a project called Behaviour Change By Design. Below Marcus fills us in on why it’s so tricky to reprogram the brain and how the four-year project will unfold.

Can you start by telling us about your professional background and field of expertise…

I’m a psychologist, but probably not the kind of psychologist most people would immediately think of. My specific area is biological psychology, which focuses on how biological factors (e.g., our genetic makeup) contribute to our behaviour. I do a lot of work on health behaviours – things like smoking and drinking alcohol.

For example, smoking is an example of a complex behaviour under a very close degree of biological control. There is very good evidence that smokers smoke for the nicotine (even though it’s the other stuff in cigarettes that does them harm) – when smokers switch from high tar to low tar cigarettes they simply change their behaviour (for example by inhaling more deeply) to get the level of nicotine they need. Often this happens without the smoker even being aware of the change. And it means that low tar cigarettes are just as harmful as high tar cigarettes. For me, that’s a nice example of how biological factors can influence behaviour.

How did you come to be involved in the Behaviour Change by Design research programme?

I’ve worked with Theresa Marteau and other colleagues in Cambridge for a while. We’re interested in how our micro-environments influence our health behaviours – things like the size and shape of the plates we eat from. However, research needs funding, so we applied to the Wellcome for support for a programme of work to investigate these questions in more detail. Fortunately, we were successful, and were awarded funding for four years to explore these questions in laboratory studies and field trials.

Why do we need a collective change in behaviour?

Even small reductions in how much we eat or how much alcohol we consume would have a substantial positive impact on health across the population as a whole. Unfortunately, behaviour change is difficult, even though many people would like to be a bit more healthy. If we can find evidence that making small changes to our micro-environments (known as choice architecture, or “nudge”, interventions) could encourage healthier behaviour (consciously or unconsciously) then this could be a very useful additional strategy.

Our actions are a result of conscious and non-conscious processes. Tell me a little more about this?

One popular theory of how our behaviour arises is dual process theory – this proposes that there are two sets of processes that interact to influence how we act in different situations. The conscious processes are slow, deliberate and more directly under our control, whereas the non-conscious processes are faster, automatic and not directly under our control. An example of the first is the active desire to stop smoking, while an example of the second is the same person who finds themselves having lit a cigarette without even realising it…

Evidence shows that conscious processes (eg telling someone about the health implications of smoking) are not particularly effective. Why?

I don’t think it’s entirely ineffective – health education campaigns focusing on, for example, the harms of smoking have contributed to the decline in rates of smoking in many countries. But they’re not VERY effective. More importantly, if we accept dual process theory, these interventions only target one set of processes, so that if we can separately target the other (non-conscious) set of processes we should be able to do even better. Both approaches can be considered complementary.

What are some of your most promising interventions and how do they work?

One area we won’t be working on in our Behaviour Change by Design grant is tobacco labelling, because we’ve already made excellent progress in this area. Cigarette packs are now standardised, removing the imagery that can become associated with the effects of the drug and become a cue to smoke. Our earlier laboratory and field trials suggested that standardised tobacco packs would make health warnings more salient and change attitudes to smoking. We’re now hoping to apply a similar approach – combining laboratory and field studies to inform policy decisions – to other behaviours (diet and alcohol use, as well as smoking) and other aspects of design (size and shape, as well as labelling).

How will you test the effectiveness of these designs?

A key feature of our project is that we will use laboratory studies, which provide tightly controlled conditions, to initially test choice architecture interventions and understand how they might work, together with field trials in real-world settings. The latter are critical – interventions that look promising in the controlled conditions of a laboratory might not work in the noisier, messier environment of a bar or supermarket. We’ve been fortunate enough to be able to set up collaborations with partners such as the NUS, who will allow us to use their bars to test some of our interventions. One of the first interventions we’ll test, based on promising laboratory results, will be whether serving beer in straight glasses rather than curved glasses reduces consumption.

You will also be involved in the implementation of the evidence found during your research. Why is this important?

We have an implementation advisory panel as part of the project, which includes journalists, members of NGOs with an interest in public health, and policymakers. They will provide an important perspective on the feasibility and acceptability of the interventions we’re proposing. Even if an intervention is effective in principle, it still needs to be implemented to make a difference in practice. We’re planning a series of workshops and public engagement activities to promote choice architecture interventions as an important way of improving health.

The research project is four years long. How will the project evolve over that time?

The project is intended to be flexible and responsive to emerging evidence and research questions. For example, over the last five years electronic cigarettes have emerged as an important public health topic where there is currently very little evidence (something we’re hoping to address in this project). So things can change a lot in four years, and the flexibility of this grant will allow us to respond to new challenges and questions as they emerge.

What is your ultimate goal in terms of this project?

I think there are two main goals – one scientific and one practical. From a scientific point of view we want to understand how choice architecture interventions influence behaviour, from the perspective of dual process theory. From a practical perspective, we want to estimate what impact these interventions might have on behaviour in the real world, and use this information to help policymakers and others introduce these interventions for the benefit of public health.

How can the design community help or get involved?

I think this is something that could be really exciting. Many of the interventions we’re considering focus on everyday objects that, in principle, could be designed in such a way as to unconsciously promote healthier behaviour – plates and glasses are the obvious example. At the moment we’re assuming that changes may need to be enforced, for example through licensing regulations that encourage bars to serve drinks from certain kinds of glasses. However, the design community could have a really important role here, by helping us to design (let’s say) glasses that are attractive and appealing – that people want to use – but that also help to nudge us towards healthier behaviours. Given that most people would like to eat and drink more healthily, but find it difficult because of the unconscious processes we’re trying to target, this could be a really attractive option for many people.

Triangulation of evidence

Follow Marcus’ progress on Twitter @marcusmunafo.