Behavioural Prioritisation: what it is and how to do it

Authored by Peter Slattery and Sarah Kneebone (our education lead)

In Chapter 5 (published 1 May 2021), we discussed how to define, identify and prioritise behaviour. In this post we are going to revisit some of that content. We will also dive deeper into the process of prioritising behaviours and explore some other useful approaches.

What is a ‘behaviour’?

Before we discuss behavioural prioritisation – let’s quickly explain what a behaviour is and how we capture it precisely.

From our perspective a behaviour is an “observable action” performed in a particular time and place; what someone is doing, rather than thinking, believing or feeling (i.e., the internal state of mind).

To define behaviours we use a modified version of the framework provided by Fishbein and Ajzen (2010), and Presseau et al., (2019). Based on this we focus on:

WHO DOES WHAT with WHAT, WHERE and WHEN?

Table 1 (reproduced from our chapter) outlines and defines these components.

What is behavioural prioritisation?

Behavioural prioritisation is the process of taking a list of potential behaviours, such as those mentioned above, and picking the best ones to focus on.

What might this look like in practice? For an example see BWA’s work as part of the Cooperative Research Centre (CRC) for Water Sensitive Cities, led by Sarah and Liam Smith (see below for example of the output and Chapter 5 for the full discussion).

Why is behavioural prioritisation useful?

Behavioural scientists and practitioners often deal with problems where the options for behaviours to target seem likely to differ greatly in their impact on the problem.

Prioritisation is useful if we expect a high variance in the probability and size of potential impacts from picking different behaviours. For instance, cases where choosing a better behaviour to target (e.g., promoting eating fewer takeaway meals rather than more walking) might mean we have 2x or more positive impact on the problem (e.g., patient’s weight loss).

We might also prioritise between behaviours for feasibility reasons. Often there are some behaviours which you are not suited to address even if they are relevant. Similarly, there may also be various political or legal barriers to targeting certain types of behaviour which might rule them out even if they would be effective.

Prioritisation can also be useful for efficiency because it can reduce the behaviours targeted. Projects with a single behavioural focus arguably have a greater chance of success than those which target many behaviours. One reason for this is that, in general, a project’s risks and costs will scale up rapidly as the complexity increases. Each new person, object or process involved will add additional complexity and effort, and lower the chance of success.

For more on this idea, see why megaprojects fail and similar work.

Tools for prioritisation

In Chapter Five we discuss how to use Sarah’s Impact Likelihood Matrix (ILM) to prioritise between behaviours.

While the ILM is an effective and well tested approach for prioritisation, there are many other viable approaches, some which we quickly mention below.

Some of these approaches, mirror or expand on the ILM, as they primarily focus on factors such as impact and tractability. For instance, our Scale up Toolkit, suggests a more time intensive behavioural prioritisation process (SINTS) which involves rating potential target behaviours based on five criteria (see below).

Similarly, Selinske et al., 2020 built on Sarah’s ILM work and the work of Douglas Mackenzie-Mohr to suggest a four step process (see left side of below)

There are also approaches which focus on additional criteria which are instrumental towards creating impact. For instance, OECD’s BASIC approach uses with 10 criteria which factor in issues such as organisational accessibility and priority (see below).

Of course you don’t need to only stick to predetermined criteria. As we mention in our Scale up Toolkit, you will often need to generating your own relevant criteria. Similarly, in his book Designing for Behaviour Change, Stephen Wendel argues that you should pick and weight relevant criteria to prioritize behaviours based on organisational considerations.

Which prioritisation approach is best?

There is no universally best approach for behavioral prioritisation. The best choice depends, for instance, on the context, relevant factors at play (e.g., cost, time, optics), time available, and, of course, personal preference and familiarity.

Indeed, answering this question of ‘how best to prioritise’ is also a part of the prioritisation process. Peter Bragge, who leads our evidence review service, has worked extensively on prioritisation, and used this to develop a toolkit with 12 steps for organisations to follow.

His steps really get into many fundamental considerations, such as setting the scope of the prioritisation process and the criteria that you want to prioritise on. We won’t reveal them all now, because they are being prepared for a paper, but reach out if you are an organisation needing help with a big prioritisation project planned.

Conclusion

In this post we have explained what behavioural prioritisation is, why it is so important and discussed some ways to do it.

As discussed, it’s usually a good idea to prioritise between behaviours and particularly important to do so when there are a number of options available and reasons to believe that these differ significantly in expected impact.

This may all seem obvious but unfortunately, prioritisation isn’t our default as humans. We want to minimise mental effort so we generally go with familiar choices and rules of thumb. We may also tend to systematically underestimate the difference in our options.

The result is that a lack of prioritising has big consequences in every relevant domain: In health, charity, and business, billions of hours and dollars are spent tackling issues that are overshadowed by much more serious and easier to address problems.

There is a famous anecdote about the mathematician Richard Hamming, who, at great cost to his popularity, would ask his fellow researchers: “what are the important problems in your field, and why aren’t you working on them?”.

We too should ask ourselves a similar question whenever we are seeking to resolve a problem through behaviour change. If not we run the risk of tackling the wrong issues, like checking the coriander when we should be putting out a fire.

In Chapter 6 of our book we discuss how different factors influence behaviour. You can read about that here.