Literature Review
2.1 Behaviour Change
2.1.1 Behaviour and Behaviour Change
The field of behaviour change is concerned with describing and understanding the determinants of human behaviour, in order to comprehend what causes, influences, facilitates or hinders behavioural patterns. Ultimately, the aim is to enable actions or interventions to help change people’s behaviours for the better. Human behaviour can be broadly defined as ‘anything a person does in response to internal or external events’ (Michie et al. 2014). The central idea behind behaviour change is one that
6 A scoping approach to the literature review is largely the norm within the Participatory Design research tradition. Of all PD publications consulted throughout this thesis, only one reports having applied a systematic review of previous literature. That exception is Kujala, 2003. For a complete picture of the PD literature review, see appendices. In healthcare, Munn et al. (2018) support the adoption of a scoping strategy when the intentions include the clarification of concepts or definitions, the identification of factors and characteristics related to key concepts, the identification and clarification of knowledge gaps, as well as to examine methodological aspects of how research is done within a given field or area.
17 challenges the perception that people usually make rational, optimal decisions based upon their best judgement of the options available (The Behavioural Design Lab, 2013).
In order to better approach the complex nature of behaviour change, it is of use to differentiate between models of behaviour and theories of change. As distinguished by Darnton (2008), models of behaviour are descriptions of specific behaviours with respect to the underlying factors that help determine and influence such behaviours, while theories of change identify the mechanisms through which behaviours can be changed over time. Niedderer et al. (2014) have fostered a three-fold divide according to which models of behaviour are grouped into individualistic models, context driven models, and models that tread the middle ground. This approach complements two historically contrasting views on health-related behaviour, as explained by Guttmacher et al. (2010): the Medical Model, according to which individual choices have the greater impact over people’s health; and the Ecological Theory Model that defends behaviour as determined mostly by a ‘dynamic interplay between demographic variables and the physical and social environment’.
A recent publication by a group of influential scholars in the field of behaviour change has identified, and established relationships between, eighty-three different theories of change (Michie et al., 2014b). Underlying all theories of change and models of
behaviour is the assumption that the determinants of human behaviour can be singled-out, studied and understood, and these can, in turn, help to outline a description of their causes, manifestations and the factors that contribute to them occurring and the probability of certain patterns to occur more regularly than others.
Their study also indicated that the most frequently used theories are the
Transtheoretical Model of Behaviour Change, also known as Stages of Change, first developed by Prochaska & DiClemente in 1982; accompanied by the Theory of Planned Behaviour, outlined by Icek Ajzen in 1985. This conclusion is echoed by another
account that places these models in a privileged position in both the literature and the practice of behaviour change (Davis et al. 2015). Both models have also influenced design frameworks targeted at changing behaviours, such as the work of Coskun &
Erbug (2014), who incorporate Ajzen’s approach; and Ludden & Hekkert’s ‘Design for Healthy Behaviour’ (2014) – one of the few approaches of design for behaviour change developed specifically to tackle health issues – which is based on the Stages of Change theory of Prochaska & DiClemente.
18 The use of theory to describe and inform behaviour change research and practice has been problematized by different authors in recent years, with a special emphasis on the low incidence and inappropriate use of theoretical grounding in practical applications (Davies et al., 2010; Michie et al., 2015), the insufficient expertise regarding behaviour change (NHS, 2014), and the need for a more multidisciplinary, collaborative approach (Solomon, 2005).
As a consequence, there has been a growing movement to try and ensure that
behaviour change interventions are ‘evidence-based’ and ‘theory-based’ (Davies et al., 2010; Bartholomew & Mullen, 2011). Recent efforts have increasingly been pushing that agenda – such as the development of taxonomies (proposed by Abraham & Michie, 2008; Michie et al., 2011; Michie et al. 2015; Kok et al., 2015), as well as the
strengthening of a ubiquitous discourse clamouring for greater rigour and consensus about terminology, methodology, theory and practice (Orleans, 2005; Francis et al., 2012; Davis et al., 2015; Glasgow et al., 2016). Such research approaches emphasise a
‘hard science’, top-down (expert-focused), utilitarian view, dominated by strict rules ascertained, for example, by the rising adoption of templates for reporting
interventions. These templates determine guidelines for what should be reported and how, concerning the planning, design, implementation and evaluation of behaviour change interventions (as outlined by Hoffman et al., 2014; Börek et al., 2015; and Tate et al., 2016).
This strong push towards greater scientific rigour is built on the premise that too often interventions ‘fail or succeed’ without reliable explanations about why it happened either way7, and the recognition that there is a great deal of uncertainty about ‘which methods work and under which conditions’ (Peters et al., 2015). This difficulty in ascertaining causation has propelled demands to strengthen the theory and evidence regarding intervention design, implementation and evaluation (in order to, among other things, enable proper scientific replication). Simultaneously, growing concerns about the effectiveness and cost-effectiveness of interventions – stemming from the traditional bounds of the field with public policy – further contribute to a research environment that is discouraging one of its fundamental intentions: to connect
7 In September 2016, a major project was launched by a coalition of three institutions in the UK, involving a multidisciplinary team of behavioural, computer and informational scientists, in order to develop an artificial intelligence system to gather and synthesise evidence from the overwhelming amount of available literature, and generate new ideas about strategies for changing people’s behaviours. The project has as its main goal to answer the big question: ‘What behaviour change interventions work, how well, for whom, in what setting, for what behaviours and why?’. More information can be found at:
https://www.ucl.ac.uk/human-behaviour-change (last access in 19/09/2016).
19 specialists and lay people when tackling complex, behavioural issues. In short, there seems to be an exaggerated movement toward the ‘scientification’ of the field, narrowing the space of opportunity for research efforts that involve qualitative methods and other forms of investigation that, recognising the highly complex nature of most behaviour change interventions, do not intend to reach degrees of universality through statistical inference.
This type of research ethos makes it particularly difficult for process-driven, qualitative efforts to arise, flourish and offer valued contributions to the ongoing discussion around behaviour change. The apparent disregard for this dimension of behaviour change interventions counters the propositions of many authors and organisations that have increasingly acknowledged the fundamental, moral and practical importance of improving the participatory aspects of health care in general (Cahill, 1998; NHS, 1998;
Crawford et al., 2002; NHS, 2013; Richards et al., 2013; Ridley & Jones, 2002), and of behaviour change interventions in particular (Darnton, 2008; Craig et al., 2008). This gap between knowledge and practice becomes evident with respect to the lack of explicit reference to participation within most frameworks and methods for designing and implementing interventions. This issue is aggravated by an apparent lack of awareness of many authors regarding the potential contributions that (Participatory) Design could offer, regarding the planning and implementing of interventions with the active participation of representatives of all relevant stakeholder groups.
David R. Buchanan (2006) argues that ‘research validity’ is by no means a topic beyond suspicion as some that stand for traditional ‘hierarchies of evidence’ may claim. He states that ‘values’, tend to be validated through dialogue – i.e. qualitative reflection – in the sense that some things hold more value than others to the extent that people agree upon comparative assertions. Comparisons also depend on contextual factors, specific to the time, place and social environment in which the observed phenomena happen. In his words (Buchanan, 2006):
[…] the criteria for establishing the validity of value claims are based on reaching reasoned agreement. We gain confidence in the validity of normative propositions to the extent that others exercise their considered judgment and find good reasons to concur with the analysis.
In order to be put into practice, the theoretical apparatus of behaviour change – the models of behaviour, theories of change, and change techniques – need to be brought together in actions that, respecting the specificities of the context and the target population, can ethically assist in altering ‘bad behaviour’ into ‘good behaviour’,
20 enhancing the probability of engagement with better behaviours, or restricting the likelihood of adherence to behaviours that are potentially harmful, wrong,
inappropriate, undesirable or counterproductive. The practical processes of
implementing change usually involve some type of planned intervention that, in turn, frequently relies on the use of Behaviour Change Frameworks – both will be further examined in the next section.