Traditional behaviour-based safety (BBS) programmes are based on the model popularised by the American psychologist BF Skinner in the 1930s. Skinner said if you punish people when they do the wrong thing but reward them for doing the right thing, they will do more of the right one and less of the wrong. The rat will learn which way to turn in the maze to find the cheese, and the worker will learn to wear the right personal protective equipment (PPE).
There have been many criticisms of this approach to BBS: that it has the potential to lead to a blame culture, for example, and can focus too much on the behaviour of individuals at the sharp end rather than that of their managers and organisational decisions. BBS programmes will achieve little if the workplace, the tools and the processes are unsafe.
Programmes that measure the safety culture or climate in which the behaviour takes place have become popular, but there is little firm evidence showing how to use these measures to improve safety. However, there is an approach to behaviour change that has been tried and tested in another domain and can be applied to the workplace.
The behaviour change wheel
“It is difficult to determine how effective BBS programmes are,” says Maria Matthews, “because studies that have attempted to evaluate their effectiveness are usually flawed both in how they measure interventions and the outcomes. This makes it difficult to determine whether effects are significant and what interventions have actually had an impact.”
As a health psychology academic, Matthews is currently applying paradigms from that discipline to OSH behaviour.
“These models have been extensively applied to myriad health-related behaviour such as smoking, weight loss, dental hygiene and sexual health, but surprisingly they have rarely been used in predicting, explaining or modifying workplace safety-related behaviours.”
Matthews sees a direct parallel between health-related behaviour such as taking exercise or brushing teeth, and safety behaviour, such as using the correct equipment or wearing personal protection.
“In both cases,” she says, “the right behaviours reduce the risk of harm and offer protection from known hazards. It’s a logical step to use health psychology to reduce accident rates.”
One reason for the lack of cross-fertilisation may be that though BBS offers a fairly simple model of behaviour, there are multiple versions of behaviour change theory in health psychology. But Matthews points to the behaviour change wheel developed by Susan Michie, Lou Atkins and Robert West (www.behaviourchangewheel.com) as a tested model (see the “behaviour change wheel” figure above).
Organisations need to use a combination of interventions. Even then you are unlikely to move everyone to the new behaviour
“We looked at 19 behaviour-change frameworks,” says Professor Michie of University College London. “Each one had value, and there was much overlap between them, but none could be considered comprehensive. The behaviour change wheel was developed by synthesising the best aspects of these frameworks and establishing a coherent mode of behaviour change at the centre.”
The concern that both Michie and Matthews express about most behaviour change interventions is that they are prompted by inspiration rather than evidence.
“A group of people sit around a table discussing a problem,” says Matthews, “for example, there has been a spike in eye injuries reported. Someone has a lightbulb moment [and says], ‘We’ll design a poster telling people to wear eye protection’.”
Informal measures might also be applied locally – a supervisor watching an operation more closely, for example. “Some time later, the spike has gone, but whether the poster made a difference, or another intervention, no one knows.”
The health psychologist’s approach is to consider evidence, form a hypothesis about how an intervention might change an outcome, apply the intervention and then measure the result to see whether the hypothesis is supported. A tried and tested behaviour change framework provides a structured approach to collecting data, forming a working hypothesis, and choosing from a range of possible interventions.
At the centre of the wheel is the COM-B model of behaviour, developed by Michie and colleagues to overcome problems they identified with theories of behaviour change. In COM-B each behaviour (B) results from an interaction between physical and psychological capabilities (C), opportunities provided by the physical and social environment (O) and reflective and automatic motivation (M).
Motivation is often considered in a one-dimensional manner – if workers are motivated (by money, fear, social pressure) to do something (wear PPE, or not wear PPE) they will do that thing. However, some motivation can be regarded as automatic. We clean our teeth and wash our hands through habits formed long ago, but it feels odd if we skip those routines. Other habits require some thinking about – some reflection. I should go for a walk now instead of eating cake because I want to be healthy. Motivation is still involved, but one is automatic, and one needs consideration. Michie and colleagues give an example of hand washing by hospital staff. The automatic motivation comes when there are established routines and habits for hand cleaning; reflective motivation involves enhancing the belief that using alcohol gel more often will reduce infection transmission. The COM-B model of behaviour is represented by the green hub of the behaviour change wheel.
The red ring in the wheel relates to the choice of intervention functions. The nine intervention functions are listed below with sample interventions for changing manual handling behaviour.
- Education – provide information about correct handling techniques.
- Persuasion – induce negative feelings about poor handling by showing images of damaged spinal discs.
- Incentivisation – offer an extra day’s leave to the team with the fewest reported handling injuries.
- Coercion – threaten to discipline anyone found handling incorrectly.
- Training – show staff how to use handling aids, have them practise and give feedback on their techniques.
- Restriction – refuse to let people into a storeroom or warehouse without a trolley.
- Environmental restructuring – make trolleys more readily available in the right locations, or put a notice on a storeroom door reminding people to use a trolley.
- Modelling – use champions to set a good example of trolley use.
- Enablement – go beyond education, training and environmental restructuring, perhaps involving workers in the selection of new handling equipment that better meets their needs
The next step in using the wheel is to determine whether the behaviour change is linked to capability, opportunity or motivation (COM-B).
Michie and colleagues map sources of behaviour to intervention functions. For example, if the behaviour requires a physical capability, interventions that fall under the training or enablement function are more likely to be effective; if reflective motivation is needed, interventions using education, persuasion, incentives and coercion might be more successful. This mapping can avoid some of the mistakes that have been made in behavioural safety efforts – using rewards and punishments when the problem is not motivation, but lack of training or an ergonomic mismatch.
- Side effects/safety
Once the appropriate intervention functions have been identified, interventions falling within those categories can be considered. Matthews suggests using a range of techniques from the health psychology toolbox. “We look at existing research,” she explains, “use interviews, site visits, questionnaires and focus groups, and use these findings to design appropriately targeted interventions.”
Dr Jim Morgan of the psychology applied to safety and health research team at Leeds Beckett University points out that a sole intervention is seldom enough: “Safety behaviour is multifaceted, so even for well-defined interventions, the power of a single intervention is small. Organisations need to use a combination of interventions. Even then you are unlikely to move everyone to the new behaviour.”
Studies from health psychology illustrate another key point in determining whose behaviour to target. Dr Lou Atkins, who worked with Michie on developing the behaviour change wheel, says: “Getting obese children to eat less works better if parent behaviour is targeted – shopping for healthier food, not stocking unhealthy snacks and serving smaller portion sizes.” Similarly, trying to incentivise workers to take time to plan a job safely, carry out a pre-job check, or find and fit the right PPE will be less effective if supervisors reward those who finish first. The behaviour needed by all those who affect the outcome – from procurement and senior management, cleaning, maintenance and facilities, to customers and end users – should be considered.
If there is a long list of potential candidate interventions, Michie and colleagues suggest the APEASE criteria to select those to use (see the “APEASE criteria box). The first three: affordability; practicality; and effectiveness all relate to the reasonable practicability principle OSH professionals will already be familiar with (see “A is for ALARP”, bit.ly/2x2z5es). Reasonable practicability allows a balance of risk with the sacrifice involved in reducing the risk (in money, time or trouble). Applying the same measures to candidate interventions should come naturally in OSH.
The fourth criterion is acceptability, and perhaps comes less naturally in some businesses than reasonable practicability. Michie explains that acceptability “refers to the extent to which an intervention is judged to be appropriate by relevant stakeholders”. In the workplace, this includes the people whose behaviour you want to change, but may also need to take into account supervisors, managers, customers or end users. For example, an incentive to give auditors extra annual leave if they save time by carrying out audits directly on a tablet device, rather than on paper, which will entail typing up later, may be cost-effective in that the time saved in processing each audit was greater than the additional leave offered. But other staff who can’t be offered a similar chance to win extra time off may find this unacceptable.
The last two criteria: side-effects/safety and equity, are linked, because the side effects or unintended consequences of an intervention could result in increased risk for other people. If providing more accessible trolleys for handling staff creates extra trip hazards for other workers or blocks fire exits for customers, the intervention would be inequitable.
The behaviour change wheel’s outer rim is relevant mostly when governments are involved in public health campaigns. For OSH, legislation would include the government banning a particular material (such as asbestos), rather than trying to reduce its use through guidance. The Health and Safety Executive (HSE) has experimented with changing behaviour through service provision in the past, but the HSE information line and the worker safety advisers service were withdrawn.
The Fit for Work advice line is an example of current service provision, although the Fit for Work occupational health assessment service was discontinued earlier this year. There are different opinions on which behaviour is promoted by fiscal measures – such as fee for intervention charging. Higher fines in the 2016 sentencing guidelines were undoubtedly an intervention intended to convince organisations to control risk more vigorously. The decision to make all HSE guidance free online in the early 2000s is an example of another measure on the outer rim.
The guidelines element includes “documents that recommend or mandate practice”. Although regulations and legislation might be regarded as the same thing for OSH professionals, in the wheel regulations refers to more local rules that are not enforced by legislation, such as a site speed limit. At a local government level, environmental/social planning on the outer rim might include changes to road layout to slow drivers down or to separate lorries from cyclists. Communication/marketing takes us back to our original lightbulb idea – sometimes a poster, or an equivalent message via social media or company intranet is what is needed to help promote behaviour change.
Matthews has had the opportunity to test her approach in two organisations. Until May this year, she worked with Stuart Webster-Spriggs, health, safety quality and environment director at VolkerRail UK, to apply this approach to the safety behaviour of rail workers in safety critical roles. Matthews explains that, for the first part of the project, “we took a broad-brush approach, exploring potential predictors of safety attitudes and safety behaviour such as production pressure and employee wellbeing. Later we focused on specific behaviours that we identified would benefit most from designing targeted interventions.”
Webster-Spriggs had recognised the value of a psychological research base to his work and approached Morgan to provide a psychologist through the Knowledge Transfer Partnership programme.
“We are very interested in the application of psychological research methods,” Webster-Spriggs explains. “Deeper knowledge of behavioural factors influencing safety for our workers has allowed us to investigate system and process changes that, in combination with other safety measures, result in a steady decrease in adverse safety events and an increase in positive safety events therefore breaking the ‘plateau’ of stagnant performance statistics.”
Matthews is now applying her experience of health psychology at specialist engineering company, AMCO-Giffen. SHEQ director Lawrence Ling says “the company has recently experienced a period of organisational change and growth. The use of behaviour change theory, wider safety culture interventions and learnings from human factors will help us in influencing AMCO’s safety culture."
Applying the behaviour change wheel and other lessons from health psychology might seem like a longer process than just trying something out. But if you just try something out because it’s the first good idea that occurs, and you don’t measure the impact, in the long run you are likely to waste more time.
There is lot of information online on applying the behavioural change wheel, and though this is mostly targeted at public behaviour, for example in health and environmental management, with a little imagination it might inspire some new solutions to old workplace problems. Michie agrees readers should try applying it.
“Absolutely,” she says. “I think the framework would work well for occupational safety and health.”