Behavioural change approaches are often used to address societal challenges. However, the resulting behavioural interventions often evoke inconsistent and heterogeneous results. Ambiguity is a potentially key lens in helping to explain different interpretations of singular behavioural interventions, and hence unexpected outcomes, yet remains overlooked in current behavioural change research. In response to this, we take a case study approach to examine the role of ambiguity in designing behavioural interventions in a highly ambiguous context. We conduct a thematic analysis of the data collected, and identify three major types of ambiguity that each evoke heterogenous responses in specific ways. We synthesise the results in the form of the ‘Ambiguity in Behavioural Design framework’. Our framework highlights the need to acknowledge ambiguity when designing for behavioural change and is the first to identify this element within the field of behavioural design. It sets a foundation for understanding ambiguity in behavioural design and has implications for design theory and practice.

Behavioural design has been gaining momentum to address critical societal challenges such as elderly care. At the same time, it struggles to deal with complex challenges and integrate multiple contextual factors' influence in domains like healthcare. Behavioural design processes lack guidance on how to prevent this and instead define problems that are ecologically valid. Conventional design methods do guide designers to do this. Thus there is a need to investigate how design methods can support behavioural designers to explore the context and integrate multiple perspectives on the problem, resulting in context-sensitive problem definitions. To respond to this need, we present a case where designers used a combination of design methods to investigate the complex context of hygiene in nursing homes. We show how combining different design methods supported the designers in advancing their understanding of a complex context and the problematic behaviours that occurred in it. We conclude by discussing the importance of using design methods to reveal important insights at the early stages of the behavioural design process.