Communication and Behaviour Change

8/3/2019 Communication and Behaviour Change

1/70

Communicationsand behaviour

change

Co

mmonGood


2/70


3/70

Foreword ..................................................4

Introduction ..............................................6

What influences peoples behaviour?.......10

Embedding behavioural theory ................40

Conclusions and future implications .......58

Next steps ...............................................62

AppendicesAppendix 1: The five-step process .............................. 65

Appendix 2: References ............................................. 66

Appendix 3: Acknowledgements................................. 68

33


4/70

Communications and behaviour change

Foreword

4


5/70


6/70


IntroductionHuman behaviour is a very complex area. This document draws onkey sources from the disciplines of social psychology, economics and

behavioural economics (where the first two disciplines overlap).We have sought to distil this information into some key factorsthat are important to consider for anyone developingcommunications that seek to influence behaviour, andto develop a framework for applying these factors to thedevelopment of a communications strategy.

6


7/70

Introduction

About this documentThe document is designed primarilyfor those working in governmentcommunications and draws on theGovernment Social Research (GSR)

Behaviour Change KnowledgeReview.1

The GSR Review was developed primarilyfor those in analysis, research andpolicy roles. It is therefore importantthat this document covers the samekey theories and principles, so thatthose working across all types ofbehaviour change interventions candevelop a common understanding.

Broadly speaking, most governmentcommunications seek to encourageor enable people to act in one or

more of the following ways:to start or adopt a new behaviour;to stop doing something damaging;to prevent the adoption of a negative

or harmful behaviour; and/or

to change or modify an existingbehaviour.

In each case, the aim is to get peopleto behave in a certain way. Insightsfrom social psychological theory and

behavioural economics, both of whichprovide us with a deeper understandingof human behaviour, are therefore relevantto all government communications.

A brief introduction to

some key behavioural

theory disciplines

Many disciplines have somethingto say about human behaviour,including economics, psychology,sociology and anthropology. Withingovernment, behaviour change(which is often applied through socialmarketing campaigns) tends to bedominated by social psychological and(behavioural) economics thinking.

Behavioural models are designed tohelp us better understand behaviour.Those used within government tend to be

social psychological models that explainbehaviour by highlighting the underlyingfactors influencing the individual orgroup. Behavioural economics combinesinsights from economics and psychology togenerate principles that show how peoplesdecision-making can be less rational.

Over recent years, behaviouraleconomics has attracted much interest,with authors such as Richard Thalerand Cass Sunstein2 exploring someof the principles and looking at how

these can be used to nudge ustowards making better decisions.

1 Darnton A (2008), GSR Behaviour Change Knowledge Review. Reference Report: An overview

of behaviour change models and their uses, HMT Publishing Unit, London.

2 Thaler R and Sunstein C (2008), Nudge: Improving Decisions About Health, Wealth 7and Happiness, Yale University Press, New Haven, Connecticut.


8/70


While behavioural models and economicsboth seek to explain why people behavethe way they do, theories of changeseek to explain how behaviour changes.There are many theories of behaviourchange, drawn from a wide range of

disciplines. This document discusses afew of the most commonly cited theories.

We have provided an overview of some ofthe main factors from social psychologicalmodels, the key principles of behaviouraleconomics and the best-known theories ofchange. For a more in-depth discussion,see the GSR Review,3 which covers over60 social psychological behaviouralmodels and theories and includes anappendix that matches behaviour typesor domains (for example environment,

health or transport) to models. Youmay find this a helpful starting point inidentifying those models that are relevantto you and warrant further exploration.

Within some government departmentsand agencies, teams (usually eitherpolicy or analysis) have alreadyundertaken significant research intobehavioural models and theoriesrelevant to their policies. It is thereforewell worth investigating whether anywork has already been carried out inyour department or agency beforestarting your own exploration.

In this document

What influences peoplesbehaviour?This section outlines some of the key

factors that influence behaviour. Itdraws on a range of social psychologicaltheories and includes three examplesof behavioural models. The section alsogives an overview of the key principles ofbehavioural economics and of the best-known theories of change. Case studiesprovide a practical illustration of howmodels and theories have been used toinform government communications.

Embedding behavioural

theoryA five-step framework shows how, byincreasing our understanding of behaviour,behavioural theory can help to definethe role for communications and build acommunications model. The Departmentof Healths Tobacco Control campaignis used to show how each step of theprocess might work in practice. Thesection concludes with a summary of thesteps and a series of questions designedto stimulate thinking at each stage.


of behaviour change models and their uses, HMT Publishing Unit, London.8


9/70

Introduction

Conclusions and futureimplicationsThis section lists the main conclusionsemerging from the report, thengoes on to consider some of the key

implications for communicators.

Next stepsFinally, this section suggests someareas for future discussion aimed atembedding behaviour change theoryin communications development.

9


10/70


Whatinfluencespeoples

behaviour?Human behaviour is influenced by a huge range of factors. In thissection, we seek to distil the ever-increasing body of evidence aboutwhy we do what we do into some key factors and principles that areimportant to consider when designing communications aimed atinfluencing behaviour change.

1010


11/70

What influences peoples behaviour?

Individual behaviours are deeply embedded in social and institutional

contexts. We are guided as much by what others around us say and

do, and by the rules of the game as we are by personal choice.4

Social psychological

models of behaviourThere are many different socialpsychological models that seek toexplain human behaviours. Broadly

speaking, the factors in most of themcan be split into three levels:

1. Personal (micro) factors which areintrinsic to the individual, such astheir level of knowledge or their beliefin their ability to change their behaviourand their habits.

2. Social (meso) factors which areconcerned with how individuals relateto each other and the influence of otherpeople on their behaviour.

3. Environmental factors over whichindividuals have little control. Theseinclude both:

a. local (exo) environmentalfactors, for example the area inwhich an individual lives andlocal shops and facilities, and

b. wider (macro) environmentalfactors such as theeconomy or technology.

Most models tend to focus on factorsat the personal and social levels, withfew explicitly referencing those at theenvironmental level. It is, however,essential to consider factors from allthree levels. For this reason, we suggestthat social psychological models are

used primarily to identify personaland social factors and that additionalwork (see the discussion of systemsmapping, page 13) is undertaken toidentify the relevant external factors.

There are a number of reasons whyit is so important to identify factors atall three levels. Seeking to understandand influence behaviour by addressingpersonal factors alone, for example, isunlikely to work, because it fails to takeinto account the complex and interrelated

nature of the factors that influence whatwe do: we do not act in isolation, andmost people are influenced to a very greatextent by the people around them and theenvironment in which they live. Equally,it would be overly simplistic to focus onenvironmental factors, such as access toservices or levels of taxation, while ignoringthe social and personal factors at play.

In seeking to influence behaviour inthe context of health, for example,it is generally accepted that anecological approach that is, one thatidentifies and addresses the factorsinfluencing behaviour at all threelevels is likely to be most effective atbringing about behaviour change.

4 Jackson T (2005), Motivating Sustainable Consumption: A Review of Evidence on Consumer Behaviour

and Behavioural Change, report to the Sustainable Development Research Network, London. 11


12/70


Key learnings

It is essential to identify factorsinfluencing behaviour at the

personal, social and environmental

levels.

An ecological approach one thattakes account of and addresses

factors at all three levels is likely

to be most effective in bringing

about behaviour change.

A note aboutbehavioural modelsWhile models can help us understandmore about human behaviour, theydo have some limitations:

Models are deliberately simplified toaid understanding, and do not accountfor all the complexities of behaviour.5

Most models do not segment the targetpopulation. In reality, different factorsare likely to carry different weight fordifferent people.

Models are usually based on data from aspecific audience and/or designed witha specific behaviour in mind and maynot travel well. Any model used must be

explored for relevance to the behaviourthat is being addressed.

It is also important to note that behaviouralmodels and theoretical insights are nota substitute for primary research. Ouraim is to help you incorporate theoryinto your planning alongside primaryresearch and other evidence.



12


13/70


Behavioural models

and communicationsUnderstanding all the factors thatinfluence the behaviour you want to

change is an essential starting point.It will enable you to start identifyingthe most effective interventions and toestablish where communications (inthe broadest sense) sit within the mix.

The approach outlined in this sectionis aligned with the early steps in the nineprinciples for developing interventionsusing behavioural models outlinedin the practical guide section of theGSR Review.6 This is to ensure thatpolicy and communications teams

share a consistent framework.

Identifying behavioursSome government communicationsfocus on influencing a single behaviour,such as getting people to drive moreslowly. For many issues, though forexample, climate change and obesity government will need to bringabout changes in multiple behavioursin order to meet policy goals.

Where the specific behaviour orbehaviours in question have already beenidentified and agreed, you can movestraight on to identifying the influencingfactors (see page 15). If, however, thisthinking has not yet been done, it is

important to identify all the relevantbehaviours and other factors contributingto the issue you are seeking to address.Otherwise, it will not be possible to selectthe appropriate behavioural model(s).

The most complex behaviours mayrequire a systems thinking approach.Systems thinking is a way of mappingout all the factors influencing a particularissue and the relationships betweenthem, so that issues can be seen aspart of an overall system rather than inisolation. Systems diagrams providea visual representation. At the coreof systems thinking is the concept offeedback, the idea that changing onefactor will often affect one or more ofthe other factors within the system. It is

therefore particularly useful in helpingto anticipate the possible consequencesof interventions aimed at a specificfactor or factors. While it may not becommunications teams who undertakesystems mapping, such exercises cannevertheless yield valuable insights intothe range of interventions needed toaddress the various factors and the partthat communications can play in the mix.

Where the issues involved are lesscomplex, a full systems mappingexercise may not be needed. However,an exercise to identify the factors andbehaviours contributing to the issueyou are seeking to address will providea useful starting point in developing acomprehensive picture of influences.

6 Darnton A (2008), GSR Behaviour Change Knowledge Review. Practical Guide: An overview


13


14/70


Mapping the causes of obesity

Obesity is a complex issue with extremely wide-ranging causes operating at the personal,

social and environmental levels.

In the absence of a model incorporating all the relevant factors, a team of experts from

a range of disciplines, including psychology, sociology, food sciences, genetics andepidemiology, produced an Obesity System Map highlighting the full range of contributory

causes.7

The Obesity System Map shows a vast number of factors operating at the personal, social

and environmental levels with multiple linkages and no one factor dominating. This clearly

points to the need for a sustained and wide-ranging programme of interventions in order to

facilitate change.

7 Vandenbroeck P, Goossens J and Clemens M (2007), Tackling Obesities: Future Choices

Building the Obesity System Map, Government Office for Science, London.14


15/70


Understanding

the factors that

influence behaviour

The factors that influence behaviourfall into the following broad levels:personal; social; local environment;and wider environment.

This section provides an overview offactors at all levels that can play a majorrole in influencing behaviour. It doesnot and cannot cover all the factorsthat influence behaviour. Instead, wehave chosen to focus on those thatappear across a number of differentbehavioural models and/or that we believe

are particularly important to considerwhen designing communications aimedat influencing behaviour change.

The section also includes three examplesof behavioural models to show howthe factors selected work in context.

Personal factors

Knowledge and awareness

When we ask people to change theirbehaviour, we need to clearly set out ourexpectations. This might be, for example,the speed limit we want them to observewhen driving in a built-up area. Standardeconomic theory assumes that if peopleare provided with information, they willact on it in such a way as to maximisepersonal benefit and minimise theircosts, a concept often referred to asrational choice theory.

Personal

Social

Local environment

Wider environment

The AIDA (Attention, Interest, Desire,Action) marketing model is an example ofan information deficit model. It is basedon the idea that providing informationwill spark interest, which in turn leadsto desire and subsequently to action.

Sometimes for example, when tellingpeople how to deal with swine flu orabout a new piece of legislation it isappropriate simply to give them therelevant information. But knowledgeand awareness are rarely enough by

themselves to bring about behaviourchange. Other factors can overrideour rational selves, and we maymake systematic errors in our rationalcalculations (see the section onbehavioural economics, page 20).

Providing information is therefore a firststep towards influencing behaviour changerather than an end point. For example, aswell as explaining how eating too muchand doing too little can lead to obesity, theChange4Life campaign (see case study

on pages 2829) also aims to increaseself-efficacy (see page 18) by showingpeople that they can incorporate moreactivity into their daily lives. Informationcan also be used to direct people toother communication channels orservices such as a website that aimmore directly at changing behaviour.

15


16/70


Attitudes

Attitudes are specific to particularbehaviours. Early psychological modelsshow attitudes leading to intention in apredominantly linear fashion. In latermodels, attitude still plays a role but

appears alongside a range of other factors.

While attitudes can influence behaviour,evidence now suggests that the link isnot as strong as we might previouslyhave thought. The so-called ValueAction Gapdescribes those situationswhere a person holds values that areinconsistent with their behaviour.

The Value Action Gap can be particularlyevident with regard to attitudes tothe environment. While people may

believe that it is important to protect theenvironment, other factors may takeprecedence when it comes to actuallychanging their behaviour. Darntoncites research into pro-environmentalbehaviours which found that at least80 per cent of the factors influencingbehaviour did not stem from knowledge orawareness.9 It is also important to bear inmind that although attitudes can precedebehaviour, the opposite can also be true.

According to Festingerstheory of cognitivedissonance,8 a personholding two inconsistentviews will feel a senseof internal conflict(cognitive dissonance),which will promptthem to change theirviews and so bring theirperceptions into line. Thishas also been found to

apply to inconsistenciesbetween perceptionsand behaviours.

A good example of cognitive dissonanceis evident among smokers. Mostsmokers know that smoking causeslung cancer and other health problems,but they also want to live a long andhealthy life. Smokers can seek to reducethis dissonance either by giving up

smoking or by finding ways to justifytheir habit, for example by claimingthat cigarettes keep them slim or thatthey know someone who smoked 30cigarettes a day and lived to be 100.

8 Festinger L (1957), A theory of cognitive dissonance, Stanford University Press, Palo Alto, California.




17/70


Communications can help to shiftattitudes. For example, governmentcampaigns have succeeded in changingattitudes to drink driving and this,combined with legislative change andother interventions, has succeeded in

bringing about behaviour change (seethe case study on page 19). However, aswith knowledge and awareness, shiftingattitudes alone may not be enough tobring about behaviour change. It istherefore rarely advisable to view attitudeshift as a precursor to behaviour change,or to use it as a proxy for behaviourchange for the purposes of evaluation.

Habit and routine

Habit or routine can be a key factor in

influencing frequent behaviours. In recentyears, it has become an emerging area ofstudy in social psychology and in fieldssuch as neuroscience. Here, experimentson animals have given insights into thedual process cognitive system10 wherebythe automatic mind handles most of ourday-to-day functioning while the executivemind both monitors the automatic mindand takes on focused mental tasks.

The more we repeat a particular behaviour,the more automatic it becomes. As time

passes and the behaviour is undertakenmore and more frequently, habit cantherefore become the key factor drivingbehaviour. As government communicators,many of the behaviours we try to influencewill be habitual (for example, leavingappliances on standby or having thetap running while we brush our teeth).Such behaviours are often unconscious,and difficult to explain or justify.

Kurt Lewins theoriesof change maintainedthat breaking habitsrequired an emotionalstir-up11 to raise the habit

to conscious scrutiny.Smoking is both anaddictive and a habitualbehaviour. The BritishHeart Foundations FattyCigarette campaign setout to tackle the habitualelement by associatingthe cigarette with thedamage it was doing. Theaim was to ensure thateach time the smoker

thought about having acigarette, they would alsothink about the damage itcaused. This constitutedthe emotional stir-upneeded to turn smokingfrom an unconscious habitinto a conscious action.

10 Graybiel (2008), Habits, rituals, and the evaluative brain in Annual Review of Neurosciencevol 31, 35987.

11 Lewin K (1951), Field Theory in Social Science, Social Science Paperbacks, London. 17


18/70


A number of behaviour change theoriesstress the importance of subjectinghabits to scrutiny as a first step towardschanging them, by raising them out ofthe unconscious mind. Communicationsdesigned to influence habitual behaviours

must therefore look at which strategies willbe most effective in turning a habit intoa conscious behaviour. Ambient mediaused at the point where the behaviouractually takes place is an example ofcommunications seeking to do this andcan be very effective for example,posters and stickers near washbasinsreminding people to wash their hands,as used in the swine flu campaign.

Self-efficacy

Agency, self-efficacy and perceivedbehavioural control (different terms areused in different models) all describean individuals sense that they can carryout a particular action successfully andthat action will bring about the expectedoutcome.12 What is important is thebelief, not whether or not the individual isactually capable of achieving a particulargoal. This will determine the effort aperson is prepared to put into changingtheir behaviour and even whether theywill attempt it at all. Peoples sense ofagency can be driven by many things,

including past experiences and personalbeliefs (for example, some people arenaturally more pessimistic than others).

This factor appears in many socialpsychological models. Lack of agencycan be a strong barrier to behavioural

change. Again, environmentalissues are a useful example. Publicresponses to climate change arecommonly characterised by a lackof agency, for instance, the sensethat the problem is too large forindividuals to make a difference.13

Communications can help to increaseindividuals sense of agency, for instanceby providing clear instructions thatmake a particular behaviour seem moreachievable, by using testimonials to show

how other people have made the changeor by helping to teach relevant skills. Forexample, one element of a recent sexualhealth campaign set out to teach youngpeople negotiation skills which they couldthen use to initiate discussions about,for example, contraception. A furtherexmple of a campaign seeking to increaseagency is the Home Office vehicle crimecampaign (see page 20). However, it isessential that any such communicationsare seen as trusted and credible and that

the behaviour is depicted as achievable.



13 Ibid.1818


19/70


Moment of Doubt campaign

The Department for Transports Moment of Doubtcampaign aims to convince young men aged between17 and 29 that the consequences of drink driving arerelevant to them.

In the past, drink-drivingcampaign messages havebeen based on a risk andreward model, contrastingthe pleasure of drinkingwith the risk of causing

injury or death by drivingunder the influence. Whilethe number of peoplekilled has fallen from 1,600in 1979, it has stayedrelatively stable at above500 a year since 2000.

The advertising agency,Leo Burnett, set out to re-evaluate the assumptionsbehind the campaign.Attitudinal research

found that a small butgrowing number of people,particularly men aged 1729, refused to acknowledgethe risk of having acrash when driving afterdrinking, while qualitative

research suggested thattrying to shock viewerswith the most extremeconsequences wasbecoming less effective forthis group, who did not see

drink drive-related crashesas relevant to them.

From general

attitude to

personal response

These findings, along withinput from behaviouralpsychologists, pointedto a move away fromcampaigns that build

public outrage towardspersuading the targetaudience that drink drivingcould have immediate,negative consequencesfor them personally.

Consumer research, asemiotics analysis andbehaviour theory identifiedthe key intervention pointas occurring when thedrinker was deciding

whether to have a secondpint, ie when they were stillin control. The campaigntherefore focused oncreating cognitivedissonance (see page16) between the desirefor another drink and aset of credible, relevantconsequences such asgetting a criminal record,being banned fromdriving and damaging

relationships with a partneror family members.

Impact

Six months after the launchof the campaign, youngmens perception that theywould be caught by thepolice had risen from 58 to75 per cent.14 The numberof people breathalysedduring December 2007rose by 6.4 per cent, whilethe number testing positivefell by 19.5 per cent.15

The number of deaths andserious injuries caused bydrink driving fell for the firsttime in six years, from 560in 2006 to 410 in 2007.

14 BMRB Tracking, January 2008.

15 Accociation of Chief Police Officers. 1919


20/70


Home Office vehicle

crime campaign

The Home Office vehicle crime campaign

sought to increase peoples sense of

agency by persuading them that theycould outsmart the criminals.16

Where previous campaigns had used

images of circling hyenas to play on

peoples fear of crime, the strapline

Dont give them an easy ride aimed

to empower people and convince

them that through their own actions

they could lessen their chances of

becoming a victim. The campaign also

sought to change habits by placing

ads reminding people to lock their carsand not leave valuables on show in

strategic places such as parking meters

and petrol pumps where opportunistic

theft was most likely to happen.

Emotion

Emotions may be triggered without ourknowing and can have a strong influenceon our behaviour, both consciousand unconscious. While most models

see emotion as influencing behaviourchange indirectly (by altering attitude,habit or agency), there are occasionallysituations where behaviour is driven byemotion alone. For example, a phobiaor fear can determine how an individualbehaves in certain situations.17

Where communications do aim tostimulate emotion, it is important tounderstand which factors are likely to beaffected by the emotional response, giventhat often, emotion will not have a direct

influence on behaviour. Emotion can bea useful hook but it should not be theend point. Otherwise, there is a risk thatthe communication will not lead to anychange in behaviour, as in the HomeOffices original vehicle crime campaign(see box), where images of hyenas circlinga car inspired fear in some sectors of theaudience and left them feeling powerless.

Behavioural economicsIn the previous section, we looked at some

of the most common factors that drivebehaviour at the personal level. These allappear in social psychological models.Behavioural economics can also be usedto understand individual behaviour.

16 COI/IPA (2008), How Public Service Advertising Works, World Advertising Research Centre, London.




21/70

What influences peoples behaviour

Behavioural economics involves applyingpsychological insights to economic modelsto account for the systematic errors inour decision-making. It suggests thatthe decisions we make are dependenton context and that natural biases and

mental shortcuts (heuristics) can lead toimperfect decision-making. Behaviouraleconomics does not generate drivers,but it does offer principles that canhelp us understand why people deviatefrom the assumed rational standard.Communicators seeking to influencebehaviour in a way that involves making adecision or choice should therefore havean understanding of those principles.

Mental shortcuts (heuristics)

We take mental shortcuts or use rulesof thumb hundreds of times each day,and the more pressure we are under,the more shortcuts we take. Where, forinstance, we dont have time to make adecision by calculating the pros and consof various options, we often make aneducated guess based on, for example:

How easily we can recall (availability)or imagine (simulation) somethinghappening: We tend to believe thatevents we can easily think of orimagine happen more often and areconsequently more likely to happento us.... people tend to be more nervousabout flying than driving becauseairplane crashes are easy to recall.Similarly, it is found that the larger thejackpot in a lottery, the more tickets thatare bought, because the consequencesof a large prize attract more attentionand are easy to imagine.18

Communications can harness thisbias by making it easier for peopleto imagine the consequences of aparticular behaviour or reminding themof the (negative) consequences of apast action.

What has happened before(representativeness): We generallymake decisions based on how similaran outcome is to something that hashappened before, not by weighing upall the possibilities.

The heart of the gamblers fallacy isthe misconception of the fairness ofthe laws of chance. The gambler feelsthat the fairness of the coin entitleshim to expect that any deviation in one

direction will be cancelled out by acorresponding deviation in the other This fallacy is not unique to gamblers.19

BiasesInternal biases mean that our naturalresponses are not always fully rational.Recognising this can help us understandwhy people make the choices they doin the real world and identify whetherthere is a role for communicationsin overcoming these biases, someof which are discussed here.

We tend to prioritise short-termreward over long-term gain (hyperbolicdiscounting). For example, some peopleprefer to have more money now thanto pay into a pension plan. This is animportant principle for governmentcommunicators, as government is oftenseeking to persuade people to makechoices that involve a long-term pay-offbut little immediate gain.

18 Halpern D, Bates C, Geales B and Heathfield A (2004), Personal Responsibility and Changing Behaviour: the

state of knowledge and its implications for public policy, Prime Ministers Strategy Unit, Cabinet Office, London.

19 Kahneman D, Slovic P and Tvesrsky A (1974), Judgement Under Uncertainty: Heuristics and Biases in Science

vol 185 No. 4157, 112431. 21


22/70

?Communications and behaviour change?

Cass Sunstein andRichard Thaler in Nudge:

Improving DecisionsAbout Health, Wealth andHappiness20 use the termchoice architecture todescribe how decisionscan be influenced by theway in which they areframed. They argue thatthe choice architectcan nudge peopletowards better choiceswithout compromising

individual freedom.This concept isparticularly relevant tocommunications andcentral to the aims ofmany communicationsstrategies, suggestingas it does that the waychoices are framed canmake them more effective.

The extent to which we disregard futuregain (the discount rate) will increase themore remote the issue appears to be.So, generally speaking, the younger theperson, the less likely they are to prioritiseinvesting in a pension plan. Disadvantaged

groups also tend to be more likely todiscount future gain, focusing insteadon getting by in the short term.

... policies that are based on individualsinvestment in their future (eg personalpensions, adult education) have atendency to widen inequalities as thosewith high discount rates will be lesslikely to take on these opportunities.21

Were loss-averse. We tend to put moreeffort into avoiding loss than ensuring gain.Evidence suggests that communications

focusing on potential losses rather thangains are more motivating: in other words,disincentives are more effective thanincentives. According to this principle, amessage that states You will lose X eachyear if you dont insulate your loft will havemore impact than one that states Youwill save X each year if you do insulateyour loft.

We have a natural preference for thestatus quo (inertia). When faced with a

difficult or complex choice, our tendency isto carry on doing what weve always doneand avoid making a decision.

In recent years, there has been muchdiscussion about how governmentcan harness the power of inertia byadjusting societys default settings, forexample by requiring people to opt out

20 Thaler R and Sunstein C (2008), Nudge: Improving Decisions About Health, Wealth and Happiness, Yale University

Press, New Haven, Connecticut.

21 Halpern D, Bates C, Geales B and Heathfield A (2004), Personal Responsibility and Changing Behaviour: the

state of knowledge and its implications for public policy, Prime Ministers Strategy Unit, Cabinet Office, London.22


23/70


of becoming organ donors or makingpension contributions something to optout of rather than opt in to. Indeed,in the case of pensions this is alreadyhappening: from 2012, all employerswill have to automatically enrol their

eligible employees into a good qualityworkplace pension scheme and providea minimum contribution unless theemployee chooses to opt out.

Such decisions tend to be taken bypolicy teams (although communicationsmay have a part to play both inhelping to develop the policy and ingetting messages out to the relevantaudiences). But where this approachis not possible, communicators shouldgive some thought to strategies for

overcoming inertia, for example bymaking the behaviour seem easier toundertake than people perceive it to be.

Choices are influenced by the way theyare presented (framing). As the discussionof biases demonstrates, many of thechoices we make are hugely influenced byhow they are presented to us or how theyare framed.

Key learnings

Understanding the naturalbiases and mental shortcuts that

shape peoples thinking should

inform the nature and content

of communications.

Social factors an extremely important task during theformative stages of the strategic planningprocess is to gain an understanding of theextent to which interpersonal influencesare likely to be important for one or more

target groups.22

Other peoples values, attitudes,beliefs and behaviour can have astrong social influence on our ownbehaviour, a phenomenon that has beenwidely discussed in recent years.

Social norms are the group rules thatdetermine what is deemed acceptablebehaviour. Social norms can have ahuge influence on our thoughts andbehaviours and therefore appear in

many different social psychologicalmodels. Social norms vary by group,so what the norm is for one group ofyoung people may well be different fromthat adopted by another group living indifferent circumstances. Failure to actin accordance with these rules canlead to exclusion from the group.

When we are unsure of how to act in socialsituations we often assume that othersaround us know more and look to them forpointers on how to behave, a phenomenonknown as social proof. An example ofthis would be a party where food is laidout but guests are unsure whether itsacceptable to help themselves. As soon asthe first person helps themselves, otherswill quickly follow: the behaviour has beenshown to be acceptable to the group.

22 Andreasen A (1995), Marketing social change: Changing behaviour to promote health, social development and the

environment, Jossey-Bass, San Francisco.

23


24/70


We tend to underestimate the extentto which we are influenced by others.Most people will deny any influenceat all, an important factor to consider,particularly for researchers. Directquestioning alone is unlikely to be the

best way of establishing the influenceof social norms on an individual.

Communications and

social norms

Communications can be effective inhighlighting social norms and promptingpeople to act in accordance with them.An experiment involving hotel guestsprovides an illustration.23 Half the guestshad signs in their bathrooms with amessage about how reusing towels

could benefit the environment. Theother half had the same sign but withan extra message stating that most hotelguests reused their towels at least onceduring their stay. Guests exposed to theadditional message were 26 per centmore likely to reuse their towels: a cleardemonstration of the power of social proof.

Further examples of howcommunications can utilise thepower of social norms include:

peer-to-peer approaches such asonline forums or communities wherepeople can connect to others in similarcircumstances. This can be particularlyhelpful with regard to less common ormore sensitive issues, as social proofand reassurance can be provided in asafe and anonymous way;

positive testimonials from others whohave used a service or adopted aparticular behaviour (thus showing thatsuch behaviours are acceptable). Notethat testimonials can also help to boostself-efficacy (see page 18);

targeting campaigns at respected,authoritative opinion leaders orrecruiting them as ambassadors for abrand or behaviour. These people canhelp both to spread the message and toprovide social proof of the acceptabilityof a particular behaviour; and

driving word of mouth, for exampleby using PR techniques to generatenews stories that describe other peoplebehaving in a particular way.

Social norms can be particularlystrong in relation to some of the mostintractable behaviours that governmentis seeking to change. For example, ifthe norm within a peer group is to goout and drink to excess, it is unlikelythat lasting behaviour change will beachieved unless that norm is addressed.For the young people in that group,the views and behaviour of their peerswill be a more powerful influence thaninformation provided by other sources.

Descriptive and injunctive normsThe social pyschologist Robert Cialdinidistinguishes between two types ofsocial norm: descriptive norms andinjunctive norms.24 Descriptive normsfit the description of social norms set outabove: that is, we base our own behaviouron how other people act. Injunctive

23 Goldstein N, Martin S and Cialdini R (2007), Yes! 50 Secrets from the Science of Persuasion, Profile, London.

24 Cialdini R, Kallgren C and Reno R (1991), A Focus Theory of Normative Conduct: a theoretical refinement and re-

evaluation of the role of norms in human behaviour in Advances in Experimental Social Psychologyvol 24, 20134.24


25/70

70


norms are the rules and regulations thattell us what we should do. Often, thetwo types of norm align but sometimes,as in Tim Jacksons motorway drivingexample (see right), this is not the case.

The risk of unintendedconsequences

Communicators must be aware of the riskof making the behaviour they are seekingto change seem as if it is widespreadand therefore acceptable. For example,in seeking to highlight the problem ofmissed appointments, hospitals and GPsurgeries will often cite the number ofpeople that fail to turn up. It has beenargued that this approach runs the risk ofinadvertently legitimising the behaviour

other people are doing it, so why shouldntI? An alternative would therefore be tofocus instead on the fact that the vastmajority of people do turn up on time.

Similarly, it has been argued that raisingawareness, particularly of risky healthbehaviours such as heavy drinking ordrug taking, may serve to exacerbate theproblem. Research in the US into thedifferent approaches taken by collegesto try to reduce alcohol consumptionfound that most campaigns highlightedthe descriptive norm, that is the highconsumption of alcohol. Throughthe use of trial communications, theteam found that such messages couldactually encourage those students whodrank less than the amount describedto increase their consumption to thelevel presented as the norm.25

70Tim Jackson gives theexample of motorwaydriving (Jackson 2005):if other drivers aroundhim are driving over thespeed limit, he may be

likely to do the same(following the descriptivenot the injunctive norm).If he sees a police carup ahead, he is likelyto reduce his speed;the police car performsa focusing function,making the injunctivenorm salient. The policecar also provides anelement of surveillance,

the sense of forebodingthat sanctions may beimposed which is arequirement for adherenceto injunctive norms. Thefocusing function is alsorequired in activatingdescriptive norms.26

25 Schultz PW, Nolan JM, Cialdini RB, Goldstein NJ and Griskevicius V (2007) The Constructive, Destructive,

and Reconstructive Power of Social Norms, in Psychological Science, vol 18 No. 5, 42934.

26 Darnton A (2008), GSR Behaviour Change Knowledge Review. Reference Report: An overview of behaviour change

models and their uses, HMT Publishing Unit, London. 25


26/70


While there are instances of socialnorms changing rapidly, this has oftenbeen as a result of legislative change,for example the ban on smoking inpublic (smoking had already becomeless acceptable, but the ban made it

even more so). More commonly, socialnorms change slowly, and therefore,in the absence of legislative change, along-term approach will be required asit will take time for any changes to filterthrough and become commonplace insociety. In other words, a short burst ofcommunications activity alone is unlikelyto be effective in shifting a social norm.

Environmental factorsIn order to understand why people behave

in the way they do, it is essential to takeinto account the physical conditionsand environment in which they live.

Environmental factors can be hugelysignificant in determining how anindividual will behave. Before behaviourchange can occur, the right facilitatingconditions27 must be in place in boththe individuals local (exo) environmentand the wider (macro) environment.

The local (exo) environment

However motivated we are to behave ina certain way, if our local environmentputs barriers in the way of that behaviourit is unlikely that we will succeed. Forexample, an individual may be stronglymotivated to use public transportinstead of a car, but if there is littleaccess to public transport in their area,it will be difficult if not impossible forthem to change their behaviour.

Another example is recycling.Before kerbside collections becamecommonplace, anyone wanting torecycle needed not only the motivationto recycle but also the commitment tofind a recycling facility and access to

transport. To be effective, communicationsaimed at persuading people to recycleneeded to go hand in hand withimproving access to recycling facilities.

The wider (macro) environment

Factors that operate at a nationalor even international level can alsohave a huge influence on individualbehaviour. Wider or macro environmentalfactors can include technology, theeconomy, taxation and legislation.

Most social psychological behaviouralmodels do not make explicit reference toexternal factors. The Theory of PlannedBehaviour (see page 31) looks atenvironmental factors within the contextof perceived behavioural control, whilethe Theory of Interpersonal Behaviour(see page 32) considers external factorsas part of facilitating conditions.

Although few models reference specificenvironmental factors, it is essentialthat communicators take them fully intoaccount. Systems mapping or similarexercises at the start of any interventionsplanning should identify the factorsinfluencing behaviour at this level. Socialpsychological models can then be usedto identify factors and plan interventionsat the personal and social levels.

27 Triandis H (1977), Interpersonal Behavior, Brooks/Cole, Monterey, California.26


27/70


The implications for

communicators

Broadly speaking, factors influencingbehaviour at the environmental level willoften be tackled through policy or delivery

interventions, for example by makingchanges to regulations or the provision ofservices. Communications neverthelesshave an important role to play.

Communications can provide vital supportfor policy interventions, for exampleby providing information about newlegislation (such as the reclassificationof cannabis) or promoting local services.In such cases, more traditional formsof advertising may be appropriate.

Communications can also influence

environmental factors by helping toframe the discourse on particular issues.While this may challenge traditionaldefinitions of the role of communications,it is an important point to consider,particularly where government is seekingto influence behaviours that are deep-seated and difficult to change.

Participative approaches such as citizenssummits, workshops or online dialoguedraw on theories of learning and changewhich argue that sustainable changecannot be imposed. Rather, peopleneed to play an active role in the changeprocess. Communications could beused, for example, to recruit peopleinto programmes that empower themto change the environment themselvesor to highlight local problems, thuscreating the momentum for change.

Even where hard policy levers areapplied, individuals may fail to respondin a rational way (see page 15). Litterpenalties are one example of this. Since2006, councils have had the power toissue fixed penalty notices to anyone

dropping litter in public. However, litteringremains a major problem, even in areaswhere those penalties are enforced. Insuch cases, communications can play acrucial role alongside policy interventionsby seeking to persuade people to adoptor advocate a particular behaviour.

Key learnings

We cannot rely on people to makerational decisions based on the

information provided, and we cannotassume that changing attitudes will

lead to a change in behaviour.

Social factors are powerful, andsocial norms are deeply entrenched.

Change in social norms is often slow

to happen, and communications

are likely to be most effective when

working with other interventions in

shifting social norms.

Communications will oftenprovide valuable support forpolicy interventions in influencing

environmental factors.

27


28/70


Change4Life

The cross-government Obesity Teams Change4Life

campaign aims to help families improve their long-termhealth by making positive changes to their lifestyle.

The Foresight reportTackling Obesities: FutureChoices28 concluded thatobesity is caused by a widerange of factors, includingenvironmental, economic,

media, educational andtechnological factors,and that tackling it callsfor a multi-faceted,cross-societal approach.Change4Life, whichwas launched in 2009,therefore forms part of awider cross-governmentstrategy, Healthy Weight,Healthy Lives, whichincludes initiatives suchas building safe places to

play, promoting healthyfood in schools andthe development of anactive transport policy.

There is no behaviourchange model for obesityprevention, so theChange4Life team basedits marketing campaign ona hypothetical behaviourchange journey derivedfrom a review of behaviourchange theories andresearch with the targetmarket. The plan is toadapt the hypothesis as thecampaign progresses.29

The right conditions

for change

The most relevant modelsincluded Robert WestsPRIME Theory, the Theory

of Planned Behaviour,Social Cognitive Theoryand Social Capital Theory.The team agreed that theprogramme should besplit into two steps: first,creating the right conditionsfor behaviour change;and second, supportingpeople on their behaviourchange journey. Findingsfrom the Tobacco Controlcampaign (see case study

on page 43) suggestedthat the conditions forbehaviour change includedbeing dissatisfied withthe present, having apositive image of thefuture and having specifictriggers for action.

The marketing plan wastherefore split into sixphases, which map onto the behaviour changejourney shown opposite:

Pre-stage: engagingwith workforces andpartners, includinglocal service providersand NGOs, both faceto face and through

direct marketing.Phase 1: reframing

obesity as relevant tothe target audiencegroups with the aim ofencouraging a socialmovement, using TV,print and outdooradvertising, PR, ahelpline, a campaignwebsite and fulfilmentmaterials, and by

building partnerships.Phase 2: personalising

the issue by makingpeople realise that theirbehaviours could beputting themselves andtheir families at risk. Themain mechanism forthis was a questionnaire,asking families abouta typical day in theirlives. Responding to this

questionnaire triggeredindividualised advice andguidance. Activity wastargeted at postcodeswith the highest riskof obesity.

28 Vandenbroeck P, Goossens J and Clemens M (2007), Tackling Obesities: Future Choices

Building the Obesity System Map, Government Office for Science, London..

29 Department of Health, Change4Life marketing strategy, 2009.2828


29/70


Phase 3: defining theeight behaviours familiesshould adopt andpromoting them throughadvertising and a rangeof partners, using

direct and relationshipmarketing and focusingactivity on areas wheretarget audience groupswere most likely to live.

Phase 4: inspiringpeople to changethrough real-life storiesin the local and nationalpress, and locallytargeted activities.

Phase 5: ongoingsupport andencouragementfor at-risk familiesgoing through thechange process,delivered by postand online.

Impact

Although it is too early inthe campaign to validateall the assumptionsmade about behaviourchange, early results

suggest that behaviouris being influencedby the campaign.The ultimate aim is toestablish a correlationbetween campaignresponse, behaviourchange, altered weightstatus and improvedhealth outcomes.

Measuring Change4Lifes impact through the behaviour change journey

Reachingat-risk families

Helping familiesunderstandhealthconsequences

Convincingparents thattheir childrenare at risk

Teachingbehaviours toreduce risk

Inspiring people

to believe theycan do thebehaviours

2929

Creating desireto change

Triggeringaction

Supportingsustainedchange


30/70


Behaviour models:

some examplesModels help us to understand theunderlying factors that influence

behaviour. There are many different socialpsychological models, some specific toparticular behaviours and others moregeneral, and it is worth spending timeidentifying which are most appropriate tothe behaviour you are seeking to influence.

It is likely that you will identify a numberof models that are relevant to thebehaviour you are seeking to influence.From these, you will be able to identifythose factors where communications arelikely to have an impact. As discussed

previously, models should be usedin conjunction with other evidence tobuild up as comprehensive a picture aspossible of the behaviour in question.

The three examples of models includedhere have been selected because theyrepresent the differing levels of complexityand scale that can be found in models:

The Theory of Planned Behaviour:provides an example of an intention-based model. It is also one ofthe best-known and widely usedbehavioural models.

The Theory of Interpersonal Behaviour:builds on intention-based models byincluding other factors, notably habit.As discussed above (see page 17) thisfactor is a major influence on manyof the behaviours that government isseeking to change.

The Needs, Opportunity, Ability modelrefers explicitly to environmental-levelfactors, something that is relatively rarein social psychological models.

For a more comprehensive list ofmodels, see the appendices to theGSR Review30 which cover over 60social psychological and behaviouralmodels and theories and includes anappendix matching behaviour typesor domains (for example environment,health or transport) to models.

30 Darnton A (2008), GSR Behaviour Change Knowledge Review. Reference Report: An

overview of behaviour change models and their uses, HMT Publishing Unit, London.30


31/70


Ajzens Theory of Planned Behaviour (1986)31

Evaluation ofoutcomes

Beliefs aboutwhat others think

Beliefs aboutoutcomes

Attitudes towardsthe behaviour

Intention BehaviourRelative importanceof attitude and norm

Subjective norm

Perceivedbehavioural control

The Theory ofPlanned BehaviourThe Theory of Planned Behaviour (TPB)32

is a well-known model that has beenfairly widely used in the past, in part dueto its relative simplicity and ease of use.According to the model, the key factorsinfluencing behavioural intention are:

attitudes towards the behaviour;subjective norms and;perceived behavioural control or agency.Intention is seen as leading directlyto behaviour.

The TPB is an adjusted expectancyvalue model. Whereas an expectancyvalue model is based solely on attitude,

the TPB also recognises the influenceof the subjective norm (that is, howsocially acceptable an individual believestheir behaviour to be). It also includesperceived behavioural control, definedin this case as the ease (or otherwise)of performing the behaviour in question(see page 18, on self-efficacy).

For these reasons, the model is seen asproviding a more accurate predictionof behaviours than models based solelyon attitude. Nevertheless, the TPBremains an intention-based model.Given that behaviour is driven by manyfactors other than intention, the modelmay be more effective at predictingintention than actual behaviour.


overview of behaviour change models and their uses, HMT Publishing Unit, London.

32 Ibid. 31


32/70


Triandis Theory of Interpersonal Behaviour (1977)33

Evaluation ofoutcomes

Beliefs aboutoutcomes

Attitude

Intention

Habits

Behaviour

Facilitatingconditions

Roles

Self-concept

Norms

Social factors

Emotions Affect

Frequency ofpast behaviour

The Theory ofInterpersonal BehaviourHarry Triandis Theory of InterpersonalBehaviour (TIB) also defines intentionas one of the key factors influencingbehaviour. However, unlike the TPB,it also recognises the role of habitand acknowledges the importance ofenvironmental factors (called facilitatingconditions in the model). Triandis alsorecognised that behavioural intention isinfluenced by emotions and social factors,elements not included in the TPB.

Of the three directly influencingfactors (habit, intention and facilitatingconditions), Triandis saw habit as themost influential. He also believed thatits influence increased over time and,therefore, that the influence of intentiondecreased. The more times a behaviouris repeated, the more automaticand less deliberative it becomes.

While the TIB is less widely usedthan the TPB, it can be particularlyhelpful in relation to regular/habitualbehaviours such as car use.




33/70


Velk et als Needs Opportunities, Abilities model (1997)34

Technology Economy Demography Institutions Culture

Motivation Behavioural control

NeedsRelations, development, comfort, pleasure, work,

health, privacy, money, status, safety, nature,freedom, leisure time, justice

Intention

Consumer behaviour

Subjective well-being, environmental quality

OpportunitiesAvailability,

advertisement,prices, shops

AbilitiesFinancial, temporal,spatial, cognitive,

physical

Needs, Opportunity, AbilityThe Needs, Opportunity, Ability(NOA) model of consumer behaviouris a good example of a model thatexplicitly incorporates factors atthe environmental level.35

NOA consists of an intention-basedmodel of individual behaviour nestedwithin a model that shows the influenceof macro-level environmental factors.At the individual level, intentions areformed through both motivation (whichis driven by needs and opportunities) andbehavioural control (agency) (which isdriven by opportunities and abilities).

At the macro level, needs, opportunitiesand abilities are influenced by the fiveenvironmental factors at the top of themodel: technology, economy, demography,institutions and culture. The modelshows a two-way relationship between

environmental factors and consumerbehaviour, with a large feedback looplinking the top and bottom levels.

Perhaps because of its greater complexity,the NOA is used less frequently thaneither the TPB or the TIB. However, itprovides a valuable demonstration ofhow environmental factors can influencebehaviour and shows clearly thatfocusing only on personal factors willnot be enough to bring about change.

The NOA emphatically shows theinteraction between individualand society, and demonstratesthe need for interventions to workon multiple levels of scale.36

The FRANK case study (overleaf)provides a practical example of how abehavioural model has been used toinform a communications strategy.

34, 36 Darnton A (2008), GSR Behaviour Change Knowledge Review. Reference Report: An overview of behaviour change models

and their uses, HMT Publishing Unit, London.

35 Gatersleben B and Vlek C (1998), Household Consumption, Quality of Life and Envrironmental Impacts: A Psychological

Perspective and Empirical Study, in Noorman KJ and Schoot-Viterkamp, AJM (eds), Green Households? Domestic Consumers,

Enviroment and Sustainability, Earthscan Publications, London. 33


34/70


FRANKIn 2005, the FRANK campaign team set out to adopt amore targeted approach to communicating with youngpeople about drugs

Desk research into the best ways ofcommunicating with young peopleabout drugs37 identified Gibbons andGerrards Dual Path Theory38 as aneffective way of addressing risk issues

where there was no direct correlationbetween attitude and behaviour.

Dual Path Theorybehavioural modelGibbons and Gerrards model reflectsthe fact that drug taking involvesrational, irrational and social factorsand informs four potential roles forintervention communications:

boosting young peoples resistance(behavioural willingness);

encouraging young people to see druguse as marginal rather than mainstream(subjective norms peers behaviour);

emphasising the risks involved in takingdrugs (personal vulnerability); and

undermining the image of drug users(risk images).

Insights and implications

The model research yielded a numberof valuable insights. The first was that ifa young person has previously used adrug, this will inform their future drug-taking behaviour. Communicationstherefore work best when they aretargeted at young people before theybecome regular drug users. TheFRANK campaign is not interested inoutright rejecters, so resources havebeen focused on young people whoare thinking about taking a drug forthe first time and occasional users.

Secondly, younger adolescents (1114years old) are heavily influenced bytheir peers. Even where there is astrong intention not to use drugs, theirhigh level of behavioural willingnessmeans that they will look to their peer/

Previous behaviour

Attitudes(personal vulnerability)

Behaviouralintention

Risk behaviour

Behaviouralwillingness

Subjective norms(peers behaviour)

Risk images

Socialcomparison

37 Darnton A (2005), Understanding Young Peoples Drug Use, Desk Research Report One: Commentary, for FRANK

and COI, London.

38 Gibbons F, Gerrard M and Lane D (2003), A Social Reaction Model of Adolescent Health Risk in Suls J and

Wallstone K (eds), Social Psychological Foundations of Health and Illness, Wiley-Blackwell, Malden, Massachusetts.

34


35/70


social group when making a decision.Messages therefore encouraged thisgroup to contact FRANK for expert advicerather than listening to their friends.

They also indirectly promoted negativeperceptions of drug users andemphasised the risks of drug taking.The campaign focused on cannabis, themost prevalent drug for this age group.

Thirdly, young people aged 15+ tend to

think more rationally about the risks ofdrugs and be more drug experienced.Their decisions are more likely to beinformed by their own sense of personalvulnerability as well as by their peer/socialgroup. Messaging for this group thereforefocused on the risks involved in usingdrugs (mainly cocaine) and on buildingup trust and confidence in FRANK.

ImpactThis model-driven approach to the

campaign is reflected in its evaluation.Key performance indicators (KPIs)include the number of young peoplecontacting FRANK, increasing awarenessof the risks, strengthening resistanceand promoting negative perceptions ofdrug users. The campaign has achievedconsiderable, measurable success in theseareas since the model was adopted.

Knowing and trusting FRANKas the expert on drugs:39

Total awareness of the FRANKcampaign among the target groupis 90 per cent.

FRANK is the most trusted sourceof drug information and advice foryoung people

81 per cent trust FRANK to providethem with reliable information;

59 per cent would turn to FRANK forinformation about drugs, compared

with 44 per cent who would turn totheir mother, 22 per cent who wouldturn to their friends and 20 per centwho would turn to their doctor.

Perceived risk of drugs (ofcannabis among the 1114s andcocaine in the 15+ group):

The percentage of 1114-year-oldsagreeing that cannabis is very likely todamage the mind rose from 45 per cent

in April 2006 to 63 per cent in April2009.

After the last burst of cannabisadvertising (February 2009), 74 percent said the advertising made themrealise that cannabis is more risky thanthey thought.

After the last burst of cocaineadvertising (January 2009), 67 per centsaid the ads made them realise thatcocaine is more risky than they thought.

Negative perceptions of drug users(negative average out of 10):

Drug user March April March 5 April2006 2007 2008 2009

Cannabis40 7.2 7.5 7.5 7.7

Cocaine41 7.4 7.6 7.8 8.0

Note: The higher the score the more negative

the perception is of users of that drug.

Building young peoples resistance skills

After the last burst of cannabisadvertising, 73 per cent said that theadvertising made them less likely to takecannabis in the future.

After the last burst of cocaineadvertising, 62 per cent said that thecampaign made them less likely to takecocaine in the future.

39 Tracking survey (Synovate).

40 Base: young people who have never used cannabis.

41 Base: young people who have never used cocaine. 35


36/70


Key learnings

Models should be used as tools toidentify factors that influence the

behaviour you are seeking

to change.

You may find that there area number of models that are

relevant to the behaviour.

Not all models include factors atthe environmental level, but it is

nevertheless critically important to

identify these and build them into

your planning.

Models should be used inconjunction with primary research

and other evidence to build upa full picture of the behaviour.

This can then be used to develop

an intervention/communications

strategy.

Understanding how

behaviour changesIn the section headed Understanding thefactors that influence behaviour (see page15) we looked at the various factors that

can influence behaviour. In this section,we will look at some theories on howbehaviour actually changes. Anyoneworking on communications aimed atinfluencing behaviour will need anunderstanding of the main theories ofbehaviour change and their implicationsfor communications.

The Stages of ChangeThe best-known theory of behaviourchange is the Stages of Change model,42

developed in the late 1970s and early1980s by James Prochaska and CarloDiClemente and based on smokersapproaches to giving up cigarettes.

Since then, this model has been used fora wide range of primarily health-relatedbehaviours, including weight loss andalcohol and drug problems. The centralidea is that behaviour change comprisessix stages, with people progressing fromone to the next at their own pace.

The six stages of change are:1. Pre-contemplation: The individual has

no intention of changing their behaviourin the foreseeable future (usuallydefined as the next six months),possibly because they have not yetacknowledged that there is a problembehaviour that needs to be changed.

2. Contemplation: The individualacknowledges that their behaviourneeds to change. While they are not yet

ready to change, they intend to do sowithin the next six months.

42 See Prochaska J and Velicer W (1997), The Transtheoretical Model of Health

Behaviour Change, American Journal of Health Promotion, 12(1), 3848.36


37/70


3. Preparation: The individual is preparingto change their behaviour within thenext month and may already havedevised a plan of action.

4. Action:The individual has changed

their behaviour within the past sixmonths.

5. Maintenance: The individual is activelymaintaining the changed behaviour(although they can still relapse at thisstage).

6. Termination: The new behaviouris embedded and the individual isconfident that they will not relapse.Their self-efficacy (see page 18) is high.

Individuals are segmented by stage.Staged-matched interventions have beendevised and provide suggestions as to howpeople at each stage can best be movedforward. Over recent years, however, themodel has been subject to criticism, withsome arguing that there is no evidencethat stage-matched interventions are anymore effective than other interventions,and that interventions can work atany stage of the change process.

Nevertheless, the model can providea useful way of identifying wherean individual is in their changejourney and understanding what thismeans for targeting and messaging.For instance, individuals who areinterested in changing their behaviour(ie who are in the contemplationstage) may be a useful primary targetfor any communications activity.

Rogers Diffusionof InnovationsThis well-known theory seeks to explainhow new ideas and technology spreadthrough society by diffusion, defined

as the process by which an innovationis communicated through certainchannels over time among the membersof a social system.43 Its basic premiseis that people fall into the following fivecategories, according to how likely theyare to adopt particular innovations:

1. Innovators: The first to adopt. As agroup they are more willing to takerisks. They also tend to be young andto belong to the higher social classes.

2. Early adopters: Tend to be seen asopinion leaders among the later adoptergroups and, like innovators, tend to beyounger and from higher social classes.

3. Early majority: Later to adopt thaninnovators and early adopters.Innovation spreads through contactwith early adopters. They are also seenas opinion leaders by some.

4. Late majority: Adopt innovations laterthan average and tend to be moresceptical than the above groups.They are less likely to show opinionleadership than groups 1, 2 and 3,and have less contact with earlieradopter groups.

5. Laggards: Adopt innovations later thanall the other groups. Laggards tend tobe older and are not usually opinionleaders. Their networks tend to belimited to family and close friends.

43 Rogers E (1995), Diffusion of Innovation, Free Press, New York.

37


38/70


It is notable that the groups most likelyto adopt an innovation early on tend todemonstrate higher levels of opinionleadership (that is, to be held in highregard and have their opinion respectedby others). Opinion leadership is often

subject-specific; it is possible to be anopinion leader in relation to one innovationbut a follower with regard to another.Rogers argues that adoption dependson access to information about theinnovation, and that this informationshould be tailored to the group.

According to the theory, innovationspreads gradually through the first groups.When the level of adoption reachesbetween 10 and 20 per cent (Gladwellstipping point44) the rate of adoption

suddenly increases steeply before levellingoff at around 80 to 90 per cent. Diffusionitself also follows a five-stage process:

1. Knowledge: The individual is aware ofthe innovation but lacks informationabout it.

2. Persuasion: The individual becomesinterested in the innovation and seeksinformation about it.

3. Decision: The individual weighs the

pros and cons and decides whether ornot to adopt the innovation.

4. Implementation: The individual adoptsthe innovation.

5. Confirmation: The individual decides tocontinue using the innovation.

The theory is based on a deliberativelinear process similar to that shown in theAIDA marketing model (see page 15). Theindividual decision as to whether or notto adopt the innovation is very rational.

This theory of change has been widely

used in commercial marketing, principallyin relation to the adoption of newtechnologies and products rather thanbehaviours. Nevertheless, the conceptof opinion leaders as important agentsof change is a useful one, as is anunderstanding of how change spreadsthrough society. Both these factors supportthe case for using communicationsto target opinion leaders (in the hopethat they will then spread the messageand encourage take-up among others)

and for peer-to-peer marketing thatexploits the power of social networks.

Key learnings

As well as understanding thefactors that influence behaviour,

it is important to understand how

behaviour changes.

44 Gladwell M (2000), The Tipping Point: How Little Things Can Make a Big Difference, Little, Brown, London.38


39/70


39


40/70


EMBEDDINGBEHAVIOURAL

THEORY

40

This section focuses on how behavioural theory can be used

to support the practical planning, delivery and evaluation ofcommunications. This will help to ensure that communicationsare designed with a realistic understanding of what they can beexpected to achieve, within what timescale and at what cost.


41/70

Embedding behavioural theory

What influences peoples behaviour?demonstrated some of the ways inwhich behavioural theory can helpus understand human behaviour.However, it also highlighted some of thecomplexities of the field, and showed

that there is no one simple way ofaddressing communications challenges.

The process set out here recognisesand reflects that complexity, but isnevertheless designed to be practical.It is not prescriptive; rather, its aim isto show how behavioural theory canbe used to support the development ofa robust communications model. Weappreciate that communications teamsare often working to tight time constraints.This process is designed to be usable

and achievable within tight deadlines,and can be adapted to work withinboth shorter and longer timescales.

For clarity and simplicity, we havesplit the process into five steps, asshown in the figure overleaf..

The five-step process

Many communications and policyformation projects are already broadlyfollowing all or some of these steps.However, what is often missing is anysystematic attempt to utilise the valueof behavioural theory throughout theprocess and to share knowledge betweenall of those involved in developinginterventions, including communicationsteams and their agencies.

At each stage of the process, anexample drawn from the Department

of Healths Tobacco Control projectbrings the theory to life by showing howthat step could operate in practice.

We need an even better understanding of humanbehaviour, not just to arrive at the optimum solution

in terms of motivation but also to understand how

to get as close as possible to the behaviour under

investigation to stand the greatest chance

of influencing it. Judie Lannon45

45 COI/IPA (2008), How Public Service Advertising Works, World Advertising Research Centre, London. 41


42/70

Exploration wide range of thinking

Reducing focus down to solution

The five-step process


42

1 Identifying behaviours

2 Understanding the influences by audience

3 Developing a practical model of influences on behaviour

4 Building a marketing framework

5 Developing a communications model

Why use the five-stepprocess?Identifying, translating and applyingrelevant behavioural theory fromthe outset will help you to:

make sure that the right questions areasked and the right people are involved;

build close working relationshipsbetween communications and policyteams and external agencies;

set clear expectations as to whatcommunications can achieve;

identify new roles for communications;identify the types of agencies that

should be involved in the process andensure that their energies are focusedon addressing the challenge, rather thandefining it; and

support robust evaluation basedon a clear, shared understandingof objectives and the role ofcommunications in achieving them.


43/70


In practice: background to the strategic review of the

Tobacco Control campaign

The behavioural goal of the Department of Healths Tobacco Control

marketing and communications strategy 200810 is to help reduce

smoking prevalence to 21 per cent among the general adult populationand to reduce the prevalence among routine and manual workers to

26 per cent.

Before a recent strategy review, Prochaska and DiClementes Stages

of Change model (see page 36) played a major role in shaping

communications activity. The model was used both to focus the creative

targeting of campaigns (for example, campaigns about NHS Stop Smoking

services targeted those in the preparation phase) and to evaluate them

(for example, analysis of campaign tracking data was based on the stages

within the model).

For one part of the campaign, the Together programme, the Prochaska

model has formed the basis of the entire contact strategy. Together,which is still in use, offers remote support to move potential quitters from

preparation to maintenance (that is, from stage 3 to stage 5 of the model).

The blueprint for content creation and the timing of messages are entirely

dictated by the model.

However, while the Stages of Change model has been invaluable in

structuring thinking, it is a model about how behaviour changes, not what

influences behaviour. Many important influences in smoking cessation

such as the impact of legislation and pricing are off the model and not

explicitly described.

Moreover, there was growing clinical evidence that interventions tailoredaround the model were no more effective than untailored approaches.

The strategic review of Tobacco Control communications presented an

ideal opportunity to survey the behaviour change landscape and identify

a more appropriate model.

43


44/70


Step Key questions Outcome

Step 1 ExAgreed target behavioursIdentifying plobehaviours ratio

n

or preventing?

1 Identifying behavioursStep 1 involves identifying thebehaviour or behaviours that yourpolicy and/or communicationsactivities are seeking to influence.

As previously mentioned, sometimesgovernment communications will seekto influence a single behaviour forexample, getting drivers to reduce theirspeed. But many of the more complexissues government is seeking to address,such as climate change and obesity, willrequire changes in multiple behaviours.

If you are aiming to influence a singlebehaviour, or the component behaviourshave already been identified, go straightto step 2. If not, your starting point should

be to identify all the relevant behavioursrelating to your policy and communicationgoals (which may not necessarily bethe same) (see pages 1315 for moredetail). Otherwise, it will be impossibleto develop an understanding of thefactors that influence those behaviours.

Step 1 in practice: the

Tobacco Control campaign

The Department of Health identifiedtwo key behaviours that it would need

to influence in order to reduce smoking

prevalence:

whether or not an individual choosesto make a quit attempt; and

the method by which they chooseto quit.

By getting more people to make quit

attempts the aim is to increase the

market for quitting. In influencing howpeople quit, the aim is to increase the

overall success of quit attempts. Evidence

shows that some methods of quitting are

more successful than others: for example,

smokers who quit with NHS Stop Smoking

services are four times more likely to

succeed than those who go cold turkey.

44


45/70


Step Key questions Outcome

Step 2 ExDetailed understanding of allUnderstanding personal (for example habit, emotions, confidence, knowledge,

ploinfluencing factors

the influences biases or mental shortcuts)?rationby audience social (for example is the behaviour with or against social norms,

peer pressure, who will influence and how strongly)?

local and wider environmental (for example access, price,opportunity, services, proximity)?

vary across different audience groups?

2 Understanding theinfluences by audienceThe next step is to identify all the factorsinfluencing the behaviour(s) that relateto your policy and communication goals.

These factors will vary among differentaudience groups. Behavioural theoryshould play an important role in this task.

The factors influencing behaviour can bedivided into broad levels: personal, socialand environmental (both local and wider).(See pages 1527for more detail.) You willneed to consider factors at all these levels.Note that this may mean consideringmore than one behavioural model andlooking at other relevant theories.

The GSR Review46 describes over 60

models and theories and includesan appendix that matches behaviourtypes or domains (eg environment,health, transport) to models. This canbe a useful starting point for identifyingrelevant models and factors thatinfluence the behaviour(s) in question.

Generating behavioural insight

As noted above, behavioural models andtheories are often based on a theoreticaleveryman. Models do not always indicatethe relative importance of factors, andsome are developed with a specificbehaviour in mind. To develop meaningfulinsights into the reasons why peoplebehave as they do, you will need to drawon other information (including primaryresearch) as well as on behavioural theory.

In practice, steps 1 and 2are likely to be iterative.Exploring the factors that

influence one behaviourmight highlight anotherbehaviour that could beeasier to influence, orwhere communicationshave a more obviousrole to play. In this case,you will need to go backto step 1 and add thenew behaviour to yourlist before proceedingagain to step 2.

Personal

Social

Local environment

Wider environment


overview of behaviour change models and their uses, HMT Publishing Unit, London. 45


46/70
http://www.cabinetoffice.gov.uk/public_service_reform/contact_council/resources/segmentation.aspx


47/70


Segmentation in action: HMRC

When HM Revenue & Customs set out to segment its target audience by looking at the

factors that influence their behaviour in relation to tax, the first step was to scope the issue

through a large-scale data review and over 100 interviews. These were used to develop

a framework that identified the factors driving the way that people behave. Qualitative

research followed, and was used to design a major piece of quantitative research. The

findings pointed to five distinct segments, based on four dimensions: awareness (of

obligations); motivation (to comply); ability (to comply); and opportunity (to comply).

Step 2 in practice: the Tobacco Control campaign

The Department of Health used a number of different models and theories to gain furtherinsight into the factors that influence smoking behaviour, as well as primary research with

the target audience and interviews with academics and practitioners. A number of new

insights emerged from these sources.

Many theories identify the role of attitudes and beliefs in driving motivation. For smoking

cessation these operate in two ways: by shaping the desire to stop smoking (through beliefs

such as smoking harms me and my family); and by shaping positive images or a vision

of the future. This latter form of positive motivation was particularly important, as primary

research among the target audiences identified that while many smokers had a strong

desire to stop smoking, there were far fewer positive associations with the concept of being

a non-smoker.

Robert Wests PRIME Theory of motivation identified the need for a trigger as well as

motivation to drive a quit attempt. Previous marketing had sought only to drive motivation.

The concept of self-efficacy or agency (the extent to which an individual believes that

they can carry out a particular action successfully and that that action will bring about the

expected outcome) has been identified in many theories, including the Theory of Planned

Behaviour and the Theory of Reasoned Action. The role of self-efficacy was also highlighted

by NHS Stop Smoking service practitioners and is summarised in the model subsequently

developed (see below) as confidence in ability to quit.

The impact of social norms on behaviour is also explicit in many models, and research

among the target audience highlighted the role of peer pressure in driving both quit

attempts and relapse. This was captured in the model as an influence that drove positive

environmental pressure to quit.

Many other influences that operate at a cultural or environmental level such as legislation

and price play a very important role in driving cessation and influence the environment

for quitting.

Finally, in order to successfully change their behaviour smokers have to know how to

change successfully. In culture change models this is sometimes referred to as practical

steps for action.

47


48/70

Factors and behaviour


Step Key questions Outcomes

Step 3 To Developing a wafactorspractical model rds of influences on ameasurement of how thebehaviour factors work together and

solu

tiontheir impor

Documents

Communication and Behaviour Change