BSCoP Networking Event

Yes, the presentations can be found here.

The guide can be found on our webpage as part of our online repository.

Our mission is ‘Providing specialist expertise on behavioural science, and developing the application of it, to improve health & wellbeing in Wales.’

We provide proactive and responsive support to stakeholders across the public health system, to optimise impact on health and well-being and reduce inequities.  We offer specialist expertise, and connection to a network of academic and wider partners.  Networking, collaborating, and understanding stakeholder needs is allowing us to underpin activity to build capability, opportunity, and motivation to routinely use behavioural science. 

The APEASE criteria prompts consideration of unintended positive and negative consequences beyond the desired impact on the target behaviour. APEASE can be used at any stage in the intervention development process where a decision needs to be made whether it is deciding whom to target, what behaviour(s) to target, whether it is better to focus on capability, opportunity or motivation, what intervention type(s) to adopt, what implementation option(s) to adopt or the details of the intervention

Further information on APEASE and a checklist to support consideration of all the APEASE criteria can be found in ‘Improving health and wellbeing: A guide to using behavioural science in policy and practice.’

We define behaviour as an observable and measurable action.

There are several methods for identifying behaviours relevant to the problem you are trying to address, the approach taken will be determined by the time and resources available to you as indicated below:

• Produce a list of relevant behaviours/populations on the causal pathway to your desired outcome, using existing knowledge of the problem. 
• Have a conversation with topic matter experts, key stakeholders, and/or the population whose behaviour you are trying to influence.
• Do a quick survey with topic matter experts, key stakeholders, and/or the population whose behaviour you are trying to influence.
• Conduct a review of the literature to identify the behaviours/populations associated with the problem you are seeking to address.
• Undertake more rigorous structured engagement with topic matter experts, key stakeholders, and/or the population whose behaviour you are trying to influence to generate a list of relevant behaviours.
• Undertake/commission behavioural systems mapping to identify the behaviours, the people enacting those behaviours, and the influences on the behaviours. This can help identify the most effective intervention point in the system and the best target behaviour.

Gaining insights from a specific population using the COM B model can be done in several ways including interviews, focus groups, surveys and questionnaires, and observations. The approach will very much depend on the behaviour and population of interest and the time and resources available. Undertaking an evidence review to explore drivers of the target behaviour can be beneficial and help to inform how you gain insights from your specific population.The guide provides examples of questions that can be adapted and applied to explore the behaviour of interest with your target population.

We would recommend the recently published guide developed by Professor Robert West and the Behavioural Science Unit.

The Behaviour Change Wheel book provides further guidance on using the COM B model to design interventions.

The COM B model can be applied to influence behaviours in several ways including starting a novel behaviour, stopping a behaviour, and increasing or decreasing a behaviour.

The COM B model can help in considering what staff might need to apply behavioural science in practice:

Systematic reviews of the literature can be completed at anytime throughout the process. For example, you may want to carry out a review to identify specific factors potentially influencing the behaviour(s) of interest when you’re completing your behavioural diagnosis. If you’re specifically looking at effective interventions during your review, it’s a good idea to have your behaviour defined, and your behavioural diagnosis complete first so that you’re able to search for the correct type of interventions. Your behavioural diagnosis may reveal that there are barriers relating to psychological capability, and that may guide you to review only interventions that are focused on education, training and enablement.

We encourage others to review the evidence when developing their behaviour change interventions, and so we try to keep evidence at the centre of all that we do too. Behavioural science is a quickly growing discipline and the scientific literature to support out work is growing too. As a team, we regularly review the evidence to help inform work. This could be when we’re completing a behavioural diagnosis, or when we’re searching for interventions that have targeted similar behaviour change techniques to the ones we’d like to us. We also draw on the literature when we’re looking for frameworks and models, such as the COM-B model and Behaviour Change Wheel.

Start small! Do a little bit of what you know first. That could be defining a behaviour or reviewing the literature to see if there is any insight into a behaviour you’re already working on. The more you practice, the more skilled you’ll become at regularly applying behavioural science… and don’t do it alone! Share your work with colleagues, or us if you’d like to. Sharing your thinking with others or discussing next steps as a team can help to clarify your thinking. Reach out to others who you know may be doing something similar, perhaps someone you connected with at the network – If you don’t know who that might be, we’re more than happy to connect the dots if we can!

Yes.  In a sentence – structures, strategy, systems, skills, styles, (motivated) staff and shared values (adapted from Mckinsey’s 7S) would need to be in place to underpin the approach. Ultimately, it would be great to have behaviourally enabled/informed organizations where knowledge and skills around behavioural science are diffuse, where processes reflect these knowledge/skills and support their deployment, and the workforce believes behavioural science adds to their efforts.

The routine application of behavioural science relies on sufficient levels of capability (people with knowledge and skills); opportunity (systems, processes and norms); and motivation (belief that this adds to our efforts) – these are extremely brief ‘foundations’ to build from.   

Proportionality (doablity) is important – the guide describes (on page 11) a sliding scale of use – from using a personal understanding of behavioural science – which could happen (and probably does already in some places) on a daily basis; through to commissioning stand-alone behavioural science advice, reports or research – which might need a budget (and a good specification!).  Developing capability, opportunity and motivation to use behavioural science to maximum effect along that sliding scale is key.

Thanks for reinforcing the links, and yes, there should be a range of ways of measuring impact of behaviourally-informed interventions – the guide advocates for “using the best methods that are practicable and affordable, and with a clear focus on measuring the target behaviour in the target population” – that’s not always an RCT, page 35 has a starter list of options.  

In collaboration, with similar units, academics, policy makers and practitioners; with a clear recognition of ethics; and in an open and transparent way.

The BCW shows that ‘interventions’ include policies that can be implemented by national, regional or local governments, as well as by organisations such as the NHS. Some BCW policy options naturally operate at a system-wide level, such as fiscal policies, comms and marketing campaigns, regulation, etc. When it comes to others such as provision of services or products, it is a matter of ensuring that the resources and infrastructure are present or can be developed. A very successful example of this is the stop-smoking services, but there are plenty of other examples and it’s important to remember that educational provision and the criminal justice system are essentially behaviour change interventions.

The tried and tested way to do this is by calculating the Return on Investment (ROI). This is a well-developed method in policymaking and healthcare. It can get quite technical when one is estimating benefits that accrue many years down the line and depends on what parameters one wants to use to discount benefits that do not occur immediately, but the principles are quite simple, and it is a good way of focusing attention on best value options.

One example was drugs education campaigns in schools which appeared to increase drug use. Human behaviour is more like the weather than an electrical circuit – not always predictable even with the best science because there are so many moving parts. That is why it is so important always to keep monitoring the impact of interventions and be ready to make changes if things do not seem to be going according to plan. That is one of the big lessons from the application of behavioural science: apply the best available methods to gather data and keep gathering it as required to shape interventions in the light of changing contexts.

COM-B appears to be the only one so far that covers literally everything that underpins behaviour. There are lots of more specific models that focus on different aspects of motivation or social influence or cognitive biases and so on. But COM-B is a very general model and will not be enough on its own – you will need to drill down into the specifics of the behaviour change scenario (using COM-B as a guide) to develop the intervention approach. For example, COM-B doesn’t tell you whether varenicline or NRT are better at aiding smoking cessation – there is a whole literature on that, but a COM-B analysis can make it clear that tackling craving is the key to successful quitting.

Number one: Treat the ‘science’ in ‘behavioural science’ seriously. That means always be thinking about how you use the best available models and theories to structure your data collection and use data to update or abandon your generalisations, models and theories. Number two: Do not allow yourself to get attached to generalisations or theories. Treat everything as a hypothesis about which you have a certain degree of confidence and be willing to revise your confidence and your hypotheses as new data emerge.

It’s best to think of behaviour change in terms of multiple levels of actors with individuals at the bottom, then social groups such as families and peer groups, the structural groups such as schools, organisations and so on all the way up to whole populations. These levels interact so that, for example, individual preferences for different types of food influence organisations to cater for those preferences and market their products in such a way as to increase demand among individuals. That is why the starting point for any behaviour change intervention should generally be a behavioural systems map so that you can see where and at what levels to intervene.

There is no substitute for evaluation. The question is only what is the most rigorous and useful evaluation that is practicable and affordable and to do that, and then to adjust one’s confidence in the results according to the rigour of the evaluation. For example, if the only thing you can do is a pre-post comparison among a sample of people whose behaviour you wish to change, try and make sure your sample is as representative as possible, keep an eye out for anything that might have happened at the same time as your intervention that might have had an influence, and be appropriately cautious in your conclusions about what kind of effect the intervention may have had. In my experience there is a tendency to overinterpret the findings of evaluations.

COM-B aims to provide a framework in which you can plug and compare other theories and see where there gaps are and how best to fill them.

As a researcher I develop and evaluate interventions and so the emphasis is on finding out whether or how well they work. One that failed to work was one in which midwives were trained to deliver brief smoking cessation advice to pregnant smokers. What I learned is that it is vital to involve those who would be delivering the intervention in the development process.

Treat it as a behaviour change problem. Think about what needs to change to get them on board in terms of, for example, their understanding of the costs and benefits (capability) and how it can be incorporated into the workflow in a way that would be easy (opportunity) and deliver quick wins for them in their role as managers (motivation). Don’t overpromise but also point out examples close to home where applying behavioural science has delivered improved interventions.

It’s always about getting the best data one can in the circumstances and then being appropriately cautious about the interpretation of those data. If at all possible, go for objective behavioural markers and include a comparator.

There are. I put your query into Chat.GPT as an experiment and this is what it came up with!

It is useful and relatively easy to incorporate more behavioural science thinking into developing behaviour change interventions for health and wellbeing. They key is: 1) always to respect the data and whatever research has been done to date that is relevant, and 2) to use a systematic approach to applying the science in a step-by-step manner going from problem specification to solution evaluation.

Systems thinking is essential, and the guide provides a clue as to how to apply it using a behavioural systems mapping approach.