A Short Guide to Behavioral Science (Part 1 )

A guide for practitioners by a practitioner

Behavioral science - the way I like to think about it - combines psychology, economics, anthropology, human centred design, systems thinking among other disciplines/methodologies to help us understand holistically how people think, and make choices.

Credit: Public Health England (which has now become a new agency)

This is not how many people see it (thought it’s changing) - I think this is because it got popular with the rise of Behavioral Economics (this is how I got first into it myself) which made more emphasis on cognitive biases, and talks about both System 1 & 2, biases, nudges and irrationality.

Even today, if you ask ChatGPT - the all knowing, but sometimes making it up, emoji writing bot to give you some core principles - you will keep getting the same social proof, loss aversion etc.. of the BE hay-day.

CHATGPT’s First default of whal Behavioral Science Principles are

My worry as someone who now has experience (but will never be an expert) is that people may think this is what Behavioral Science is all about and it may leave them a bit disappointed.

I am not saying that some social proof is not useful, but it is so limited, to think that this is all you have to help you design experiences that bring about better outcomes through change.

Who is using behavioral science?

Well, if you are building products and services, communications, policies or working on anything that has to do with people - you are already working on behavior change- however you are probably not doing so with purpose and in a systematic or scientific way.

There are of course teams in governments such as the UK, US, Australia, New Zealand and even organizations like the OECD and the World Bank who do it. In fact the OECD has a website where they have been mapping the different teams around the world.

Image credit: https://oecd-opsi.org/blog/mapping-behavioural-insights/

There are also many companies that have been applying it with or without explicitly having roles or teams like Facebook, Instagram, Google, Uber, Noom, Natwest, Disney and so many others.

My views on having separate roles or small teams in a company is that it could cause silos and lack the resources to show value.

A good way to get the most from a team or unit is to have them focused on up-skilling other teams in the org, building tools and mixing methods, while applying their skills on challenges with cross functional teams.

Why should it matter?

When looked from a point of view that goes beyond cognitive biases, behavioral science can gives us thinking tools to understand how people behave, think, see others and interact with the systems around them

These thinking tools can be made of models and frameworks -such as COM-B, ISM, PRIME, The Behavior Change Wheel, BASIC, Self Determination Theory etc… and they allow us to navigate complexity and analyze our insights more efficiently.

The truth is that I have worked with and without these thinking tools, and I can say that without explicit, well-chosen models I tended to rely only on experience, my gut and intuition which can accidentally lead us all to take on the wrong problems, to ask the wrong questions and to miss out on the core insights we need to develop good solutions.

It is important to say that these are - as I said thinking tools - and that you have to apply critical thinking (which is hard in the times of GPT’s) to all of the ones you use. They will not all be appropriate to ever single context as they are representations, so they are not 100% representative of everything all the time.

Remember - behind models or frameworks are things like theories - which means you may need to dig deeper to understand if they will be helpful for you and your projects in that context.

If you want to explore more of them, this free tool I made can be helpful

COM-B and TDF - Michie and colleagues UCL

So why should we care about behaviors anyways?

Behaviors are important, because they are the actions we can see and do - which means they are observable and for the most part measurable - and that also means you can design to enable them.

Behaviors can then be:

-Eating fruit instead of cake
-Taking the stairs instead of lift
-Going to the gym three times a week

Now, before we get excited about behaviors, let’s make sure they are not mistaken with outcomes.

If your outcome is to have a healthy lifestyle and lose weight, that will need to be achieved by specific actions such as eating fruit instead of cake, taking the stairs instead of lift or going to the gym three times a week which I mentioned earlier.

If we can start getting specific about the behaviors we want to enable, then we can think more tangibly about them, and be more precise about what is making them hard to do.

How do we use a model to understand what is making behavior hard?

I think this is a good place to look at an example from one of our models - The ISM Model is based on theory and evidence made by the Scottish Government and influenced by the work of Southerton et al (2011)

This model, like the COM-B model can help us start thinking about what is making a behavior hard to do. ( We have mapped both of these models, so we can use either or when needed and still be able to work systematically with say behavior change techniques)

ISM Model - Scottish Government

ISM breaks down drivers into 3 contexts:

1. The individual context: e.g. the beliefs, values, skills and emotions

2. The Social Context: e.g. the norms, roles, networks, institutions

3. The Material Context e.g. the rules, objects, time, technologies

With a model like this we can go deeper and get to the root causes of what is making it hard to do behavior.

You are probably already collecting insights with interviews, surveys etc… - so why not use a model to group them in an evidence based way.

How would this look like in an example? well, lets use one from the the ISM authors - (If you want to see the whole guide and example then so you can go here)

The example they used is related to buying an electric vehicle (it is a bit older, so some drivers may not be 100% valid anymore)

What they found that was relevant for the individual would be:

-Range anxiety - This is definitely something that influences my own decision making on this topic
- Lack of critical mass (chicken and egg between vehicles and charging network)
-Range anxiety plays into concerns about completing journey/not running out of charge
-Refuelling routines (new charging habits need forming)

What they found that was relevant in a social context that could drive decisions for buyers:

-Personal contact with someone who drives an EV (this would be an enabler if they were advocates)
-Car attachment very strong – sense of social conversation: ‘my car is me’

What they found that was relevant in a material context that could drive decisions for buyers:

-Financing Arrangements (e.g. government grants/discounts)
-Charging points (and parking spaces)
-Scheduling breaks in long journeys to recharge

What you design for will be driven by what level of importance and impact each one of these have - so it is important to prioritise.

It will also help inform what you make, by making the right thing - meaning you can’t educate me on how good an EV is if I have no charging stations around me to use it - so prioritising a nice video campaign to educate me will be a waste of your money.

Ok, that’s it for this first part :)

To set the stage for the next parts I want to be clear, that this is not an academic guide, I am just looking at the best ways to solve design, innovation and change challenges.

I always share and write as much as possible on what I have learnt, so that you can get inspired and learn with me when possible.

---

If you don’t know me, I’m Robert Meza a practitioner who is running a consultancy called Aim For Behavior - I teach courses and work with clients around the world on fun but challenging behavioral focused projects.

Have a great day, Robert