Debunking misconceptions: a defense of the COM-B Model
A recent article argued that the COM-B model has become an overused and unhelpful default in behavioral diagnosis, overly focusing on individual-level factors and being more appropriate for simple health behaviors than complex societal issues. I've wanted to write a response to it for a while, so in this I'll take a look at some of the arguments by providing a more comprehensive overview of COM-B, its relationship to the Behavior Change Wheel, and its applicability to a wide range of behavioral challenges.
I'm going to focus on three main claims from the original article:
Is the COM-B model being overused and applied in contexts where it might not be suitable?
Does the COM-B model overly focus on individual-level factors, neglecting the broader systemic factors that shape behavior?
Beyond simple or health behaviors, is the Behavior Change Wheel (BCW) applicable and useful in understanding and addressing complex behaviors?
By the end of this article, you'll have a better grasp of why COM-B is such a powerful tool and how it can be applied to a wide range of behavioral challenges, from the simple to the complex.
Is the COM-B model being overused and applied in contexts where it might not be suitable?
The COM-B model has been around for over a decade, but it has only gained significant attention in the behavioural science community relatively recently. The Behaviour Change Wheel method was originally introduced in 2011 and the book was published in 2014, so the framework has existed for over a decade. Although some of the more academically-minded practitioners were familiar with it, it remained fairly unknown 2020 when Amy Bucher's Engaged featured it and since then it has rapidly become more "mainstream".
Personally, I'm very happy to see more people learning about COM-B and the Behaviour Change Wheel. While I can see why some might argue that it has become overused or an unhelpful default, there are significant benefits to having a common framework in applied behavioural science. I've worked as a BeSci practitioner since 2012, much of it as a co-founder of an agency specialising in applied behavioural science. I have seen dozens of agencies develop their own proprietary models of behaviour or decision making because it's what we are expected to do to differentiate ourselves and demonstrate our intellectual prowess.
Unfortunately, this is a "tragedy of the commons" situation, because when stakeholders speak to different organisations and see different models enthusiastially claimed to be equally true, it portrays our discipline as less scientific and credible. In short, the proprietary models that are meant to give agencies a competitive advantage over each other eventually create a less favourable impression of our profession.
Here, COM-B and the BCW offer a practical solution: using one framework as our lingua franca reduces the risk of confusion or misinterpretation between different models and theories.
COM-B can provide a solid foundation for everyone to work from which gives consistency and clarity across projects and contexts. Those who worry it might erase their competitive advantage should ask themselves how is it possible that every software development agency in the world works with the same programming languages and code repositories, yet some are more successful than others? You could also say the same about any professions which operate within regulatory constraints - it's good to ask why ours would be different.
Now, going back to the original claim: while some may argue that excessive reliance on COM-B can lead to missing important nuances or overlooking key factors, it is crucial to understand that COM-B is not intended to be a one-size-fits-all solution. Instead, it can serve as a starting point to help us organize our thinking and ensure comprehensive coverage of relevant factors as the core component of the larger Behavior Change Wheel ecosystem, which guides the entire process of behavioral diagnosis and intervention design.
Sure, you'll probably get lost if you start from the long list of constructs in the Theoretical Domains Framework but that would be just as inefficient and foolish as using Wikipedia's list of 100+ cognitive biases as your guide to the human mind. The BCW is a logic model, so there is a logical process to follow - doing so will allow you to focus on the relevant domains of the TDF without feeling overwhelmed or missing important nuance.
So, to suggest that COM-B is overused or limiting is to misunderstand its nature, purpose and its role within the larger framework as a tool that supports a systematic, comprehensive approach to behavior change. More fundamentally, the BCW is a part of a broader programme of work that aims to gather and use evidence about behaviour change more effectively than it is currently (see e.g. here and here).
Does the COM-B model overly focus on individual-level factors, neglecting the broader systemic factors that shape behaviour?
Another critique of the article was that the COM-B model places too much emphasis on individual-level factors while neglecting broader systemic influences on behaviour. This criticism , too, is misguided because overlooks the fact that COM-B explicitly incorporates environmental and contextual factors through the 'Opportunity' component.
Within the COM-B framework, the 'Opportunity' component is divided into two distinct categories: Physical Opportunity and Social Ppportunity. Physical Opportunity encompasses the systemic factors present in an individual's environment that can shape their behavior, such as infrastructure, resources, and technology. Social Opportunity, on the other hand, refers to the social and cultural factors that influence behavior, including social norms, peer influence, and interpersonal relationships (see table below).
Capability | Opportunity | Motivation |
Ability to engage in behaviour | Environments that enable the behaviour | Brain processes that energize and direct behaviour |
Physical skill, strength, or stamina Psychological skills, knowledge, or strength/stamina to engage in the necessary mental processes | Social opportunity: interpersonal influences, social cues and cultural norms that influence the way we think about things Physical opportunity: Opportunity afforded by the environment involving time, resources, locations, cues, physical ‘affordance’ | Automatic processes involving emotional reactions, habits, wants and needs, impulses, and reflex responses Reflective processes involving goals, plans (self-conscious intentions) and evaluations |
In recent years, there have also been efforts to further enhance the COM-B model's ability to capture systemic factors. One notable example is the development of the SeCOM-B model, which explicitly incorporates social and environmental factors into the framework. This extension of COM-B can help with incorporating the broader, systemic context for complex challenges. Alternatively, you could use the Implementation in Context (ICON) Framework - originally conceptualised as a meta-framework of the context domains, attributes and features that can facilitate or hinder healthcare professionals’ use of research evidence and the effectiveness of implementation interventions, but with some adaptations it can provide a useful lens in many organisational settings.
Even within the Behavior Change Wheel framework itself there are additional tools to expand the analysis of systemic factors. For example, behavioural systems mapping can be used to identify and visualize complex interactions between individual, social, and environmental influences on behavior. So, in short, it seems somewhat misleading to critique the COM-B model for only being suitable for individual-level factors.
Beyond simple or health behaviours, is the Behavior Change Wheel (BCW) applicable and useful in understanding and addressing complex behaviours?
The third critique suggested that while COM-B may be appropriate for simple or health-related behaviors in controlled contexts, the ISM model is better suited for understanding complex societal behaviours. Unfortunately, this comparison fails to recognize the comprehensive nature of COM-B and its ability to incorporate the factors considered by the ISM model.
Overall, comparing BCW/COM-B to ISM is like comparing apples to oranges - or more like comparing an entire orchard to a single orange.
Based on an analysis of 30 behaviour change initiatives focusing on reducing the carbon intensity of consumption practices - specifically, informational campaigns that sought to change consumer attitudes | Based on a systematic literature review of existing behaviour change frameworks, of which 19 were combined to create the BCW |
Not tested for validity or peer reviewed (commissioned by Scottish Government) | Extensively tested for validity before being published |
Google Scholar suggests 97 citations | Original paper cited 5770 times (and >10k citations according to Google Scholar) |
Most crucially, ISM is focused on characterising behaviour change interventions - it is not a tool for behavioural diagnosis like COM-B, but more like a list of ideas for interventions in the field of sustainability which may or may not be successful, as the report itself says.
Just for my intellectual curiosity, I thought I'd compare them anyway, and quickly discovered that the factors outlined in the ISM model can easily be mapped onto the BCW framework.
The individual, social, and material factors emphasized by ISM can also be translated into COM-B's Capability, Opportunity, and Motivation components, and classified with the intervention functions, policy categories, Behaviour Change Techniques and even hypothetical Mechanisms of Action, although the report did not have a huge amount of detail at this level.
It should be clear that the ISM model lacks the structured approach and granularity offered by COM-B and the broader Behaviour Change Wheel. So, while ISM may provide a simple, high-level categorization of behavioural influences, it does not offer the same level of guidance for systematic behavioural diagnosis and intervention design. This limitation can lead to inconsistencies and reduced effectiveness when applying the ISM model across different contexts and populations.
It is also worth noting that COM-B and the Behaviour Change Wheel have a strong foundation in implementation science, focusing on translating research findings into practical, real-world applications. In other words, despite its academic credentials and origins, the purpose of BCW is to be practical and help with making behaviour change interventions as effective as possible.
Conclusion
As we have addressed each of the criticisms, it should have become apparent that BCW and COM-B offer a flexible starting point for understanding behaviour instead of being a rigid or limiting model. They give practitioners and researchers a common language and framework for applied behavioural science and tackling various behavioural challenges.
With the risk of sounding like I'm in a cult, there is no real equivalent in the entire behavioural science/behaviour change literature that is as comprehensive, robust and extensively validated as the Behaviour Change Wheel. It isn't just a model or framework among others - it's an ecosystem of tools and evidence. It can't be just cast aside because we are bored of hearing about the same framework and want something new and exciting - just like we don't ask an architect what new and exciting theories of gravity they could work with for the house we want to build.
One of the reasons I wanted to write this detailed response is that I realised many people seem to have a shallow understanding of the BCW, so I will be following up this article with focused articles to shed light on the details. Stay tuned!