Applying behavioural science to a Design Challenge

How can we use methods from experimental psychology to design better services? 

I have to confess when I joined the Design Council to work on Knee High, a Design Challenge aimed at improving the lives of children under five, I wasn’t quite sure what my role would involve. Coming from an academic research background in experimental psychology, it wasn’t immediately obvious to me how I could help entrepreneurs test and prototype their products and services.  

I had taught behavioural science to students before but it seemed like a big jump to suddenly be thrown into the real world. I imagine that the innovators and entrepreneurs I met through my work felt similar when I approached them with the news that I would help them design and conduct experiments to evaluate the impact of their services and give them feedback on their prototyping. Most of them did not have an academic background, and the prospect of dealing with data collection, sample sizes and analysis seemed new, daunting and very time-consuming for most of them. 

Using insights from psychology in the development of ideas, as well as an experimental approach to prototyping and evaluation are extremely valuable tools in designing the best possible services.

As our work progressed, it became obvious that we were all wrong. So much of what we design is made with people in mind, yet the practice of combining current knowledge of human behaviour and experimentation with design is still uncommon. However, it is clear that using insights from psychology in the development of ideas, as well as an experimental approach to prototyping and evaluation are extremely valuable tools in designing the best possible services. And it’s not even as scary or difficult as it might seem!

Here are five lessons in applying behavioural science to a design process: 

1. Use existing research for what it’s worth

By taking existing research findings about human behaviour into account when you start designing your product or service, you can save a lot of time and money.

In the end, it’s all about good design. We want to create something that works well. Luckily, there are a lot of researchers that have spent a lot of time studying how people behave and why. Taking their findings into account at an early stage might save you several trips back to the drawing board.

For example, it was not necessary for the Good Enough Mums Club, a musical group aimed at helping mothers with post-natal depression, to spend time establishing the link between positive outcomes for parents and positive outcomes for their children – they can simply draw on existing research that demonstrates the importance of a mother’s mental health on her children. 

2. Be clear about what you want to change

In order to design an experiment to test your prototype, you need to carefully decide what your variables are.

This involves deciding what specific behaviour your service is trying to influence, and what your desired outcomes are. Although this can be surprisingly complicated, even the process of describing your theory of change allows you to notice gaps in your idea or other things that might need to be improved or clarified.

For example, the aim of the Creative Homes project is to reduce stress in the home through artistic interventions. Their research identified areas where a lack of routines led parents to experience unnecessary stress. Through delivering their artist-led home interventions they hope to reduce day-to-day stress by establishing fun routines. 

3. Plan ahead, but be flexible

When we test if something is working as well as we hope it will, we often use experiments where we compare our intervention to a control group.

Remember that even when you have to make changes to your experiment and start again, you have learned something new by trying. The learning you gain from it is invaluable in the development of your work.

These experiments need to be carefully planned ahead of time since it’s not a good idea to change them after you have started collecting data. However, when you run experiments outside of a controlled lab environment, a lot of things can – and usually will – deviate from your plan. One of the most common surprises is how difficult it can be to engage people to participate in your study.

This was something that all the Knee High teams experienced difficulty with. Several of them used financial incentives to get families to take part in their experiments. However, money is not the most appropriate incentive in every situation. Be creative and flexible in your approach to conducting your research.

Just keep in mind that if you need to make changes to your study it sometimes means that you have to do things again to avoid invalidating your results. When you report your findings, be honest about what went well and what was difficult. You are usually not the first one to come across a specific problem! 

4. Remember the bigger picture

You can often get stuck focusing on all the little details when you run an experiment, especially when your idea is constantly evolving and you are adapting to changes and the demands of different stakeholders.

Even the process of describing your theory of change allows you to notice gaps in your idea.

For example, a potential funder might ask you to demonstrate the impact of your service on something you hadn’t planned on measuring. Don’t forget to step back and remember what you started out wanting to achieve. What assumption are you trying to test, and what do you actually need to learn more about? It’s easy to go off in the wrong direction if you lose track of your key objectives and try to do too many things all at once. 

5. Don’t be afraid of getting it wrong

If something doesn’t go as well as you planned, don’t get discouraged!

Remember that even when you have to make changes to your experiment and start again, you have learned something new by trying. The learning you gain from it is invaluable in the development of your work.

EasyPeasy, an app aimed at improving child behaviour through fun games, and Crafty Explorers, who encourage parents to play outside with their children by giving them missions to complete in nearby parks, both tried to run their studies at the very end of term time and found it difficult to get parents to participate. In the end their samples were smaller than they had hoped. However, just from trying to reach parents EasyPeasy learned that they are easier to engage through brief text messages rather than emails. 

I hope you find these useful. If you have combined behavioural science with design and come away with other key points, do let me know in the comments below.

Find out more about the six Knee High Design Challenge teams.