Assessing Interventions for Behavioral Change in Context

With each new year comes a wave of good intentions as people aim to be better. They want to lose weight, exercise more, be nicer, drink less and smoke not at all. They want to change behavior, and as Susan Michie knows well, “behavior is related to absolutely everything in life.”

Michie is a clinical and health psychologist who leads the Centre for Behaviour Change at University College London. She specializes in behavior related to health – for behavior or health practitioners, patients and population as a whole – and in looking at how behavior impacts the natural environment. And while you might think that the essentials of human behavior are pretty similar, one of the things Michie quickly tells interviewer Dave Edmonds in this Social Science Bites podcast is that it can be unwise to jump to conclusions when studying behavior (or trying to change it).

She notes, for example, that lots of behavioral research is done in North America, where there’s relatively abundant funding for studies, “but the biggest need [for research] is often where there’s the least investment. There’s no point in developing an intervention based on research evidence conducted in parts of the world that are very far away from the type of context we want to implement the findings in – only to find out it’s not going to work.”

So yes, she says, do look at both the rigour of the research, but also base any potential application of the findings on deep understanding of local conditions and using local knowledge.

Michie and her team describe this using a model, COM B, to account for the ‘capability, ‘opportunity’ and ‘motivation’ necessary to change behavior.

Changing behaviors is important – “In order to solve any of these big social challenges we need people at different positions in society to change their behavior” -- so these considerations matter. But that begs the questions of what behaviors need changing – and who decides what those selected behaviors are..

“There’s a big issue about who decides what the key issues are,” Michie says. “But I think there are certain problems which are very self-evident – there are people dying unnecessarily as a result of smoking, obesity but also environmental conditions – poor housing, etc. There are areas where the social consensus is that things needs to change, and I’d say those are the ones we start with.”

In the interview, Michie also addresses the ethics of behavior change and how algorithms and machine learning will be “absolutely vital” to parse through all the relevant data . Her own Human Behaviour Change Project is a collaboration between behavioral scientists and computer scientists combing the global literature to see what works, with an initial focus on smoking cessation. A comprehensive tobacco control strategy, she details, involves those infamous “nudges” beloved of policy makers, but also the legislation, services and taxation, that need to work synergistically to effect real change.

Michie had a long career as a research fellow and clinician before joining the Psychology Department of University College London in 2002. She’s a fellow of the Academy of Medical Sciences, the Academy of Social Sciences, the Academy of Behavioral Medicine Research, the Society of Behavioral Medicine, the European Health Psychology Society, the British Psychological Society and a Distinguished International Affiliate of the American Psychological Association.