02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 7

Journal of Health Are Techniques Used Copyright © 2005 SAGE Publications London, Thousand Oaks and New Delhi, in Cognitive Behaviour www.sagepublications.com Vol 10(1) 7–18 Therapy Applicable to DOI: 10.1177/1359105305048549 Behaviour Change Interventions Based on the Theory of Planned Behaviour? Abstract The Theory of Planned Behaviour (TPB) is increasingly being used to inform the development of interventions to promote health behaviour change. However, although the IMOGEN C. A. HOBBIS & STEPHEN theory can be used to identify SUTTON the determinants of particular University of Cambridge, UK health-related behaviours, it offers little guidance on how to change these determinants and hence how to promote IMOGEN C. A . HOBBIS is a Clinical Psychologist and behaviour change. There is Post-doctoral Research Associate at the Institute of evidence that Cognitive Public Health, University of Cambridge. She is (CBT) can interested in the application of clinical psychology be used to support health within health behaviour change settings and behaviour change. This article interventions. discusses the similarities and differences between the two STEPHEN SUTTON is Professor of Behavioural approaches, and considers Science at the Institute of Public Health, University of whether techniques used in Cambridge. His research interests include the CBT are applicable to application of theories of behaviour change in the interventions based on the TPB. development and evaluation of interventions to change behaviours such as smoking and physical activity.

ACKNOWLEDGEMENTS. The authors are grateful to Dr Susan Michie, Professor Ann Louise Kinmonth and Wendy Hardeman for their helpful comments on an earlier draft of this manuscript.

COMPETING INTERESTS: None declared. Keywords ADDRESS. Correspondence should be directed to: behaviour change interventions, I . C. A . HOBBIS, DClinPsych Programme, MED-HPP, Elizabeth Fry Building, UEA, Norwich, NR7 4TJ, UK. Cognitive Behaviour Therapy, [email: [email protected]] Theory of Planned Behaviour

7 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 8

JOURNAL OF HEALTH PSYCHOLOGY 10(1) Introduction health behaviours (Ogden, 2003). Second, the theory is clearly specified and there are clear THE THEORY of Planned Behaviour (TPB; published recommendations on how to measure Ajzen, 1991, 2002a) is widely used to inform the components of the theory (Ajzen, 2002b). studies to identify the predictors of behaviour, And third, there is a substantial body of particularly health behaviour (Conner & evidence supporting the theory as providing Norman, 1996). More recently, it has also been consistent prediction of intentions and, to a used to inform health behaviour change inter- lesser extent, behaviour (Armitage & Conner, ventions (Hardeman et al., 2002). However, 2001). However, although we focus on the TPB, although the TPB is consistently found to much of what we say in this article is applicable provide useful predictions of intentions and, to to other social cognition models, including stage a lesser extent, behaviour (Armitage & Conner, theories such as the Transtheoretical Model 2001; Godin & Kok, 1996; Sutton, 1998), it does (Prochaska & Velicer, 1997). Compared with not provide guidance on how to promote theories like the TPB, stage theories have a behaviour change. Other social cognition more complex structure and different impli- models, for example Social Cognitive Theory cations for intervention (Sutton, 2005; Wein- (SCT), have been shown to be successful in stein, Rothman, & Sutton, 1998). In particular, promoting change, utilizing primarily behav- they imply that interventions should be stage ioural techniques (Bandura, 1997). Such tech- matched, that is, that different interventions niques have been used for many years in other should be used depending on which stage a contexts, predominantly mental health care person is in. Techniques from CBT could be settings with people who have psychological used as part of a stage-matched intervention to difficulties. One therapeutic approach that is promote movement to the next stage in the currently attracting interest due to its reported sequence. However, in this article we focus on empirical testability, efficacy and cost-effective- the TPB. ness is Cognitive Behaviour Therapy (CBT; In sum, this article will address whether Department of Health, 2001; Nathan & certain therapeutic techniques used in CBT can Gorman, 2002; Roth & Fonagy, 1996), which offer useful additions to behaviour change inter- incorporates both behavioural and cognitive ventions based on the TPB. Before addressing techniques to support change. Given the appar- this question directly, we give a brief outline of ent success of CBT in promoting changes in each approach, and then discuss the similarities psychopathological behaviour (e.g. Clark et al., and differences between them. We devote more 1994; Whittal, Agras, & Gould, 1999), and the space to describing CBT and its underlying value of behavioural techniques in health cognitive theory than to the TPB, because we promotion interventions using SCT, it is timely assume that readers of this journal are likely to to consider whether such techniques, together be less familiar with the details of this approach. with others utilized in CBT, may be used to The description of the TPB ignores some of the promote behaviour change in interventions complexities of the theory (see Sutton, 2002a, based on the TPB. Using a CBT framework in 2004). this context rather than other therapeutic models is supported as CBT can be used to The Theory of Planned inform brief interventions, applicable to either Behaviour individuals or larger groups. This mirrors the application of the TPB to short-term, group- The TPB is an extension of the Theory of based interventions. Thus, to use CBT in this Reasoned Action (TRA; Ajzen & Fishbein, way seems feasible. 1980). According to the theory, the strength of The TPB is just one of many theories that can a person’s intention to perform a given behav- be used to predict and explain health behaviours iour is the proximal determinant of that be- and to guide behaviour change interventions. haviour (for behaviours that are at least We focus on the TPB for three main reasons. partially under the person’s control). Intention, First, it is currently the most widely used social in turn, is influenced by attitude towards the cognition model for predicting and explaining behaviour (overall evaluation of the behaviour),

8 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 9

HOBBIS & SUTTON: COMPARING CBT WITH THE TPB subjective norm (perceived social pressure to shown (Hardeman et al., 2002), most TPB- perform the behaviour) and perceived behav- based interventions attempt to change beliefs by ioural control (PBC, perceived control over providing information (e.g. leaflets). performing the behaviour). PBC can also The set of modal salient beliefs may not predict behaviour directly to the extent that adequately represent the salient beliefs held by perceived control accurately reflects actual individuals in the target population (Sutton, control. Underlying attitude, subjective norm 2002b; Sutton et al., 2003). Thus, in an inter- and PBC are considered to be specific beliefs, vention that targets modal salient beliefs, many referred to as behavioural beliefs, normative individuals in the target group will be presented beliefs and control beliefs respectively. Beliefs with information designed to change beliefs that salient to the individual are held to determine are not salient to them. This may limit the their attitude, subjective norm and PBC. efficacy of the intervention. According to the theory, changing behaviour An alternative approach is to elicit and requires changing these underlying beliefs. target individually salient beliefs. Each indi- (Another way of changing behaviour that is vidual would receive a different version of the suggested by the theory is to increase actual intervention, the exact content depending on control over the behaviour. However, we are his or her own idiosyncratic set of salient not aware of any TPB-based interventions that beliefs. Individually tailored interventions have used this approach.) based on the TPB are rare but they are entirely Developing an intervention based on the TPB consistent with the theory that explains behav- involves a number of steps (Fishbein & iour at an individual level. This approach is Middlestadt, 1989; Sutton, 2002b). The first step being used in the ProActive trial, an inter- is to define the target behaviour and population. vention designed to increase physical activity For example, in a physical activity intervention among people at high risk of diabetes, in which the target behaviour could be defined as walking trained facilitators are using the TPB to elicit for at least 30 minutes a day and the target and modify individuals’ beliefs about being population as sedentary adults aged 30–50. The more physically active (Hardeman et al., second step is to identify the salient behav- submitted). ioural, normative and control beliefs with respect to performing the target behaviour in a Cognitive Behaviour Therapy sample from the target population. The salient beliefs are those that first come to mind in Cognitive Therapy1 arose from Beck’s (1976) response to standard open-ended questions cognitive-behavioural hypothesis of emotion. such as ‘What would be the advantages for you This hypothesis states that emotions arise not of walking for at least 30 minutes every day?’ because of events but from how they are The most frequently reported beliefs are desig- appraised or interpreted, which is influenced nated the modal salient beliefs. The next step is by underlying cognitive structures that cause to conduct a quantitative study using a struc- faulty or biased interpretations of events. tured questionnaire incorporating measures of was first described in terms the modal salient beliefs to determine which of the cognitive theory of (Beck, components of the theory (attitude, subjective, Rush, Shaw, & Emery, 1979), which sees early PBC) are the most important in influencing life experiences as influencing the develop- intentions and behaviour in the target popu- ment of core beliefs (‘schemas’ or ‘schemata’). lation. The relevant set of beliefs is then exam- Core beliefs are held to be at a level of uncon- ined to see which individual beliefs best sciousness such that an individual is not fully discriminate between intenders and non- aware of their significance and influence on intenders. These are selected as the key beliefs current cognitions, emotions and behaviours to target in the intervention. until their attention is drawn to this through According to the TPB, anything that changes therapy. Considered to be stable personality these key beliefs in the appropriate direction traits, core beliefs are global, rigid, absolute will increase the likelihood of behaviour change. statements that organize information and In practice, as a recent systematic review has allow individuals to interpret experiences and

9 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 10

JOURNAL OF HEALTH PSYCHOLOGY 10(1) information in personally meaningful ways. events are encouraged, and alternative They are seen to relate to oneself (‘I am outcomes are experienced. Thus, working at worthless’), the world (‘The world is a the level of conscious mediating cognitions competitive place’) and the future (‘Things (automatic thoughts) is the first line of will never get better’) (Beck, 1983). Core approach in CBT. For cases of more long-term beliefs lead to the development of dysfunc- and enduring difficulties, a greater emphasis is tional assumptions. These are conditional placed on the role of core beliefs. These are statements in the form ‘If . . . then . . .’, for challenged and restructured using the same example, ‘If I do X, then Y will occur’ (Beck, techniques as are applied to automatic 1987). Dysfunctional assumptions can be thoughts, although it is considered that working conceptualized as ‘rules for living’ in that they at this level of cognitive structure takes much guide how experiences are interpreted and longer given their perceived rigid and inflexible acted upon. They are considered to be nature. dysfunctional because they affect the interpre- tation of situations in a biased or exaggerated way. They, in turn, influence the content of the General applicability of the most conscious representation of these under- cognitive framework lying cognitive structures, automatic thoughts. This framework of cognitive structures deter- These thoughts are described as automatic as mining how incoming information is processed, they appear to come ‘out of the blue’ and to consisting of underlying core beliefs and be uncontrollable, characteristics that are assumptions and more conscious automatic particularly important in the treatment of thoughts, is deemed to apply to everyone, not mental health difficulties as they give the just those with psychological difficulties (Clark impression that the thoughts are facts and thus & Beck, 1999), though clearly in most cases, resistant to change. They are usually negative cognitions do not cause distress. Indeed, it is in content and are considered to play a role in considered that the negative cognitions and the development and maintenance of mental biased forms of cognitive processing character- health problems. Thus, cognitive theory is istic of psychological difficulties reflect an exag- formulated in terms of cognitive structures at gerated and persistent form of those seen in different levels of conscious awareness influ- normal emotional functioning (Beck, 1991). encing observable behaviour. For example, core beliefs are seen as having Cognitive Behaviour Therapy was developed positive/negative polarity so that those without from cognitive theory. It works to modify psychological difficulties will possess positive biased and dysfunctional cognitive processing. core beliefs (Clark & Beck, 1999), for instance Initially, CBT aims to educate patients about ‘I am a worthwhile person’. So, in reaction to the reciprocal relationship between thoughts, stimuli, appropriate functional and adaptive feelings and behaviours, and to increase aware- beliefs are applied to incoming data, which ness of the automatic thoughts that occur in elicit an appropriate response in terms of response to situations, events and interactions. behaviour, emotion or motivation (Clark & The accuracy of these thoughts is then evalu- Beck, 1999). Thus, underlying beliefs about the ated by assessing the available evidence outcomes of behaviours will be reflected in supporting or refuting them, and considering people’s actions, including health-related ones. their utility in allowing someone to function For example, someone may hold the core belief adaptively in everyday life. They are then modi- ‘I am a health-conscious person’ and the associ- fied accordingly. Patients are encouraged to test ated rule for living ‘If I take care of my health out and experience new ways of thinking and now, then this will benefit me in the future’. It behaving through the application of out-of- follows that their other thoughts and actions session homework ‘experiments’ to see if their will then be in accordance with this belief. This existing thoughts and beliefs are manifest in suggests that techniques used in CBT to reality and whether the feared outcomes do identify thoughts and beliefs are as applicable occur. Changes in behaviour are promoted as to those without mental health concerns as different ways of interpreting situations and those with.

10 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 11

HOBBIS & SUTTON: COMPARING CBT WITH THE TPB Cognitive Behaviour Therapy for or ingrained, as these cognitions are more easily health-related behaviours tested and thus more open to change than core CBT has, in addition, been demonstrated to be beliefs (Mooney & Padesky, 2000). Core belief applicable to health and health-related behav- work is usually considered appropriate for iours, in those with chronic illness and physical working with complex and enduring mental health problems, and in broader-based health health problems. Thus, working at the level of promotion initiatives. In the former inter- core beliefs may not be necessary to promote ventions, CBT works with illness-specific change in health behaviour interventions. In beliefs and cognitions that may be distorted or addition, the appropriateness of working at the unrealistic and aims to help the patient recon- level of core beliefs with people who do not ceptualize their beliefs in a more functional, suffer from psychological difficulties outside of adaptive or coping-orientated fashion. From specialized mental health care settings may be this, it is assumed that more adaptive behav- questioned. Recently, concerns have been iours in relation to their health status will be raised about the potentially damaging conse- adopted. Examples of the application of CBT in quences of working at the level of core beliefs chronic illness include diabetes (Henry, Wilson, with individuals who have no prior history of Bruce, Chisholm, & Rawling, 1997), obesity psychiatric problems, particularly by those who (Braet, Van-Winckel, & Van-Leeuwen, 1997; are unskilled in the approach (James, 2001). Liao, 2000) and myocardial infarction (Cowan, While eliciting core beliefs, including those not Pike, & Budzynski, 2001), all of which require related to psychopathological beliefs, may alteration to current lifestyle to improve health theoretically inform interventions, the exact outcomes. For example, cognitive behavioural nature of how this would be done, the particu- strategies have been shown to be helpful in lar beliefs considered important to the inter- supporting increases in physical activity in vention and the potential consequences of angina patients (Lewin et al., 2002). While a eliciting certain forms of beliefs must be seri- number of patients with chronic health prob- ously considered, particularly as it is not clear lems receiving CBT may have concurrent how such information would or could be acted psychological difficulties as well, this may not on within the constraints of many health behav- always be so and does not preclude the appli- iour interventions. These cautions may be even cation of CBT techniques to those without. The more significant for interventions conducted in focus of a CBT approach on the development ways other than face-to-face, and, indeed, group of a repertoire of self-management skills and interventions based on CBT tend to work the patient’s active participation and involve- primarily at the level of automatic thoughts and ment in change seems ideally suited to a rules for living (Morrison, 2001). broader health behaviour change context. That This brief description of the two models many health promotion approaches to behav- shows that similarities and differences in the iour change mirror a CBT approach has been approaches are apparent. These will now be previously described (Graham, 1985), and explored further. examples of the use of CBT in this context exist. For example, CBT has been applied in a mental Comparison of the two health promotion context to support stress approaches management (Brown, Cochrane, & Hancox, 2000; Kaluza, 2000), and ‘cognitive-behaviour Underlying theoretical modification’ has been used in interventions framework promoting physical activity (Marcus et al., A significant similarity between CBT and the 2000). TPB is that both approaches focus on beliefs In applying CBT to health-related behav- and see belief change as necessary for behaviour iours, it may not be necessary or desirable to change. In the TPB, the beliefs of interest are a elicit and modify core beliefs. Working at the limited subset of a person’s beliefs that concern level of automatic thoughts and underlying the target behaviour and are of three kinds: assumptions is considered more appropriate for behavioural, normative and control beliefs. psychological problems that are not long-term Although a person may hold many beliefs about

11 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 12

JOURNAL OF HEALTH PSYCHOLOGY 10(1) performing the target behaviour, the theory The process of questioning is continued until assumes that their intentions are determined the beliefs underlying the automatic thoughts only by those that are salient or accessible. By are elicited. From the viewpoint of CBT, the contrast, the cognitive theory on which CBT is elicitation techniques used in applications of based postulates different kinds of beliefs at the TPB yield very limited information. different levels of consciousness, from core Socratic questioning, by contrast, has the beliefs at the deepest level to automatic potential to yield much more detailed infor- thoughts at the most conscious level. mation with a broader content. There is no equivalent in the TPB to core In typical applications of the TPB, elicitation beliefs in cognitive theory, which are assumed to of beliefs is a separate stage that precedes the be unconscious and global in nature. However, attempt to modify beliefs. In CBT as it is usually there appears to be some similarity between conducted, on the other hand, the processes of salient beliefs in the TPB and automatic belief elicitation and belief modification thoughts in cognitive theory, where these are continue throughout the course of therapy and behaviour-specific, insofar as both are easily there is no clear separation between them. accessible and may be automatically elicited by relevant cues. Although dysfunctional assump- Setting and target group for tions or ‘rules for living’ in cognitive theory have intervention a conditional (‘If–then’) form like beliefs in the Most interventions based on the TPB tend to TPB, they may not be specific to the target focus on large groups within the general popu- behaviour. For example, someone who is think- lation and are targeted at everyone within this ing about stopping smoking may hold the salient group. CBT, conversely, has primarily been normative belief ‘If I stop smoking, my wife used with individuals or small groups. As such, would approve’. From the standpoint of cogni- the idiosyncratic, individualized focus of the tive theory, this belief may reflect a more global method is lessened in larger group settings. rule for living, for example ‘If I do things for However, the theory underlying the cognitive other people then they will like me’. approach and techniques for evaluating and modifying thoughts can be taught in group Method of eliciting beliefs settings, thus retaining certain essential aspects In the TPB, salient beliefs are elicited by a small of CBT (Morrison, 2001). For example, the number of standard open-ended questions. utility of a CBT approach in larger groups has CBT, on the other hand, elicits relevant beliefs been demonstrated by the successful use of in an unstructured manner. A style of ques- cognitive behavioural stress management tech- tioning that is often used in CBT is that of niques in a recent population-based large group ‘Socratic questioning’ or guided discovery study (N = 40) (Brown et al., 2000). While such (Padesky, 1993) whereby carefully worded interventions might not be considered as strict exploratory questions are used to help the CBT, as they do not incorporate the idiosyn- patient determine the personal and idiosyn- cratic nature of the approach, these results cratic meaning of beliefs that drive their behav- suggest that individual settings, implementing iours. Examples of Socratic questions include: the standard delivery of CBT are not needed for ‘What difference would making changes to X change to occur. Furthermore, limited contact make to you?’, ‘What makes you think that X TPB-informed interventions for use in one- would be difficult to do?’, ‘How might X change to-one settings are currently being evaluated, your day-to-day life?’ Similarities can be for example in the ProActive trial (Hardeman observed here between the content of these et al., submitted); CBT techniques may be questions and those used to elicit salient beliefs particularly applicable to interventions of this in TPB-based studies. Where differences do form. Thus, although in practice the two exist they are in the style and process of ques- approaches are used in different settings, this is tioning, rather than the content. In CBT, ques- not an inherent difference between them. tions are not pre-determined by a script, so the In TPB-based interventions, decisions about individual is free to bring to sessions the issues the target behaviour and target population are that are most pressing for them at that time. likely to be influenced by the interests and

12 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 13

HOBBIS & SUTTON: COMPARING CBT WITH THE TPB concerns of, for instance, researchers or policy increasing physical activity (and the beliefs makers, and may include behaviours such as about their ability to do so) in an achievable increasing physical activity or altering diet. way. This strategy of generating situations CBT, however, is primarily used with people through which an individual can gain experience who are actively seeking help to support behav- of making successful changes is akin to the iour change (or who have been referred for help guided mastery experiences of interventions through the instigation of another, for example, based on SCT (Bandura, 1997), and suggests parent or partner). Thus, health behaviour that such techniques can be used successfully change programmes based on the TPB are and effectively within health promotion and generally instigated by others who consider such lifestyle change interventions. programmes to be of benefit to populations they The similarities and differences between the deem to be at risk, rather than by those seeking two approaches are summarized in Table 1. help. Having outlined the two approaches and Population-based interventions such as these highlighted the similarities and differences raise concerns regarding the ethical implications between them, we now turn to the main ques- of such initiatives. Health promotion pro- tion posed by this article. grammes may ignore issues of personal choice and autonomy in behaviour change, for Are techniques used in CBT example, when others make decisions regarding applicable to TPB-based health status on behalf of a population (Webb, interventions? 1997). These issues would need to be considered within any large-scale intervention regardless of The conceptual frameworks underlying CBT the theoretical framework used. Approaches and the TPB are fundamentally different. that acknowledge the importance of individual- According to the TPB, intention to perform the ized care need to ‘accept the possibilities of indi- target behaviour (and hence the likelihood of vidual beliefs and attitudes to health’ (Thomas performing the behaviour) is influenced by & Wainwright, 1996). A CBT approach, with its salient beliefs about the behaviour. By contrast, focus on individual and idiosyncratic beliefs according to the cognitive theory that underlies may, in part, be able to achieve these aims. CBT, these behaviour-specific beliefs may not be the most important influences on intentions Behaviour change techniques and behaviour. Instead, behaviour may be influ- For both the TPB and CBT, the intervention enced by more global beliefs at a deeper level of aims to modify existing unhelpful beliefs, consciousness. For example, according to the strengthen pre-existing adaptive beliefs or TPB, the intention to walk for at least 30 create new ones. However, the TPB does not minutes a day will be determined by salient specify how beliefs are changed. In practice, behavioural, normative and control beliefs with most TPB-based interventions attempt to respect to this specific target behaviour change beliefs by presenting information (although these three sets of beliefs may not be (Hardeman et al., 2002). In contrast, CBT equally important). It is these beliefs that should targets behaviour change through a combi- be targeted in an intervention. Cognitive theory, nation of cognitive and behavioural techniques, on the other hand, holds that this behaviour may for example thought challenging and behav- be influenced by global beliefs such as ‘I am a ioural experiments in which patients try out lazy person’. Such beliefs may influence not only alternative ways of behaving based on new, the target behaviour but also other behaviours more adaptive beliefs. The presentation of related to physical activity. They may also have persuasive information alone is not considered wider implications, for instance contributing to sufficient to produce change within this para- a person’s sense of self-worth. CBT would aim digm; experience of both cognitive and behav- to elicit and modify these more global beliefs. ioural change is required (Persons, 1989). For The two frameworks can be brought closer example, a CBT intervention aimed at increas- together by arguing that the effect of global ing physical activity may encourage participants beliefs held at a deeper level of consciousness to conduct a behavioural experiment to test out on intentions to perform a given target

13 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 14

JOURNAL OF HEALTH PSYCHOLOGY 10(1) Table 1. Similarities and differences between the models and approaches of CBT and TPB CBT TPB Similarities: Primary focus on underlying beliefs Sees belief change as necessary for behaviour change Differences: Sees behaviour as influenced by different kinds of Sees behaviour as influenced by salient (accessible) beliefs at different levels of consciousness behaviour-specific beliefs Beliefs elicited using unstructured methods, e.g. Beliefs elicited using a small number of standard ‘Socratic questioning’ open-ended questions Traditionally used with individuals or small groups Most interventions have been at a population level (but can also be used with large groups) or with large groups (but individually-tailored interventions also possible) Interventions usually instigated by intervention Interventions usually instigated by others e.g., recipients researchers, policy makers Sees beliefs at different levels of conscious Sees salient beliefs as sole focus of an intervention awareness as potential targets of an intervention Provides guidance and strategies on how to change Focus is on what to change, no guidance included on behaviours how to instigate change

behaviour is mediated by salient behavioural, action planning, monitoring progress through normative and control beliefs. This would be diaries, self-reward and consistent with the TPB assumption that the strategies, including identification of high-risk effects of all variables that influence intentions situations and rehearsal of management are mediated by salient behaviour-specific strategies. While these techniques are integral beliefs. However, it would imply that TPB- to the application of CBT, their use is not based interventions should elicit and target such dependent on, or limited to, the use of CBT, and global beliefs rather than directly targeting indeed, interventions based on social cognition salient behaviour-specific beliefs, which would models have utilized behavioural techniques. represent a major shift of focus and would For example, studies based on SCT have proved require evidence supporting the proposed efficacious in promoting dietary and physical causal links. Furthermore, proponents of cogni- activity change (Anderson, Winett, Wojcik, tive theory would see this proposed constraint Winettt, & Bowden, 2001; Marcus, Owen, on possible causal pathways as unrealistically Forsyth, Cavill, & Fridinger, 1998). Studies restrictive. purporting to be TPB-based have also incor- It follows from this analysis that it would not porated behavioural techniques. For example, be appropriate to use CBT belief elicitation Hardeman and colleagues’ (2002) recent techniques such as Socratic questioning in inter- systematic review notes that after information ventions based on the TPB. From the perspec- giving and persuasion, skills learning, goal tive of the TPB, these techniques would yield setting and action planning were the most superfluous and irrelevant information. How- commonly used intervention techniques. ever, the use of other CBT techniques designed The use of such behavioural techniques is to change beliefs and behaviours is not broadly consistent with the TPB. For example, precluded. These fall broadly into cognitive and goal setting and action planning involve helping behavioural categories and will be discussed participants to form intentions, and self-reward separately, though the division between purely in effect creates a new behavioural intention (‘I ‘behavioural’ and ‘cognitive’ techniques is arti- will reward myself by going to the cinema if I ficial as they clearly overlap (Bandura, 1997). achieve my goal’) and a new behavioural belief (‘If I achieve my goal, I will receive a reward’). Behavioural strategies Another useful strategy commonly used Behavioural techniques to support behaviour within CBT is the ‘behavioural experiment’ in change used in CBT include goal setting and which participants are encouraged to test

14 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 15

HOBBIS & SUTTON: COMPARING CBT WITH THE TPB alternative beliefs by trying things out them- That CBT techniques can successfully be taught selves. This provides personal experience of the in group settings to some degree has already outcomes of adopting an alternative behaviour, been discussed, and the cost-effectiveness of which is considered important in encouraging group CBT treatment has some support (Gould, the adoption of new behaviours. As noted previ- Buckminster, Pollack, Otto, & Yap, 1997). ously, this is akin to the use of guided mastery Although the impact of the strategies may be experiences in SCT. However, large-scale SCT diluted at a group level, this may be balanced by interventions do not seem to have used this the potential to treat more participants more approach (e.g. Marcus et al., 1998) and, in quickly and at a lower cost. In addition, the general, TPB-based interventions do not work applicability of CBT delivered through distance to increase personal experience of alternative media, for example via email or by telephone behaviours in the same way, though this CBT- has been supported (Burgess & Chalder, 2001; informed technique is currently being used in Robinson & Serfaty, 2001) potentially eliminat- the ProActive trial. Again, this technique can be ing the need for face-to-face meetings. Second, regarded as compatible with the TPB. Trying there are arguments for conducting health- out a new behaviour may lead to the develop- behaviour change interventions on a one-to-one ment of new salient beliefs and increase PBC. level, for example that large-scale interventions do not address individually salient beliefs. Using Cognitive strategies additional techniques such as those used in CBT The application of cognitive techniques in may alleviate these concerns and help to therapeutic settings arose partially from obser- improve outcomes. vations that behaviour therapy was not always successful, and that cognitive factors were inher- Implications for future ently involved in psychological difficulties. research Thus, an approach incorporating aspects of both seemed more likely to engender successful and Theory-driven research is needed to investigate lasting change. Consideration of behaviour the links between behaviour-specific beliefs in change interventions based on social cognition the TPB and the more global beliefs postulated models that have not produced positive results by the cognitive theory that underlie CBT. supports this. For example, it has been proposed Intervention studies should compare the that the negative outcomes observed in some efficacy of a ‘pure’ TPB-based intervention, in SCT-based studies may be the result of failure which beliefs are targeted by providing infor- to address underlying beliefs regarding, for mation, with an enhanced intervention incorpo- example, barriers to behaviour change (Hallam rating techniques from CBT. These could & Petosa, 1998). Therefore, if this were the case, include both intensive, one-to-one interventions using cognitive strategies to tackle these under- and distance approaches in which CBT is deliv- lying beliefs may be helpful in challenging ered remotely. Mediation analysis should be current ways of thinking. For example, CBT used to examine whether CBT techniques influ- focuses on evaluating how currently held beliefs ence intentions and behaviour by changing the influence behaviour and responses to situations; components of the TPB as the theory would whether these are helpful and constructive or predict. destructive belief systems; and on the develop- ment of alternative, more adaptive and func- Conclusion tional ways of understanding and interpreting events. This article has examined whether health To suggest that techniques from CBT could behaviour change programmes informed by the be used in TPB-based interventions in a one- TPB may be improved in their scope and effec- to-one, in-depth way may provoke questions tiveness by inclusion of techniques used within about the cost-effectiveness and feasibility of CBT. It is acknowledged that there are concep- doing so at the level of large-scale, population- tual differences between the TPB and the based interventions. However, first, CBT may underlying theory informing CBT, which may not need to be conducted on a one-to-one basis. preclude use of some techniques. However, it

15 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 16

JOURNAL OF HEALTH PSYCHOLOGY 10(1)

appears that there are enough similarities, Beck, A. T. (1983). Cognitive therapy of depression: particularly the shared focus on underlying New perspectives. In P. J. Clayton & J. E. Barrett beliefs, for some ideas to be of benefit. Appli- (Eds.), Treatment of depression: Old controversies cation of behavioural techniques and cognitive and new approaches, pp. 265–290. New York: Raven strategies, together with consideration of how Press. Beck, A. T. (1987). Cognitive models of depression. CBT works to influence the translation of inten- Journal of Cognitive Psychotherapy: An Inter- tions into actions could contribute to improved national Quarterly, 1, 5–37. outcomes in interventions informed by the TPB. Beck, A. T. (1991). Cognitive therapy: A 30-year Furthermore, TPB-based interventions to date retrospective. American Psychologist, 46, 368–375. have done little to encourage target populations Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. to act to increase their personal experiences of (1979). Cognitive therapy of depression. New York: altering their behaviour and, thus, experiencing Guilford Press. the consequences of this. The limits potentially Braet, C., Van-Winckel, M., & Van-Leeuwen, K. imposed by the setting and scope of future inter- (1997). Follow-up results of different treatment ventions should not outweigh the possible programs for obese children. Acta Paediatrica, 86, 397–402. benefits of drawing on CBT. Further research is Brown, J. S. L., Cochrane, R., & Hancox, T. (2000). needed to illustrate where these gains may Large-scale health promotion stress workshops for occur. the general public: A controlled evaluation. Behav- ioural and Cognitive Psychotherapy, 28, 139–151. Note Burgess, M., & Chalder, T. (2001). Telephone cogni- tive behaviour therapy for chronic fatigue syndrome 1. The terms Cognitive Therapy and Cognitive in secondary care: A case series. Behavioural and Behaviour Therapy are used interchangeably as it Cognitive Psychotherapy, 29, 447–455. is usual for ‘Cognitive Therapy’ to incorporate Clark, D. A., & Beck, A. T. (1999). Scientific foun- behavioural techniques as well. dations of cognitive theory and depression. New York: Wiley. References Clark, D., Salkovskis, P. M., Hackman, A., Middleton, H., Anastasiades, P., & Gelder, M. (1994). A Ajzen, I. (1991). The theory of planned behaviour. comparison of cognitive therapy, applied relaxation Organizational Behaviour and Human Decision and imipramine in the treatment of panic disorder. Processes, 50, 179–211. British Journal of Psychiatry, 164, 759–769. Ajzen, I. (2002a). The theory of planned behaviour. Conner, M., & Norman, P. (1996). Predicting health Available at http://www.people.umass.edu/aizen behaviour: Research and practice with social cogni- Ajzen, I. (2002b). Constructing a TPB questionnaire: tion models. Buckingham: Open University Press. Conceptual and methodological considerations. Cowan, M. J., Pike, K. C., & Budzynski, H. K. (2001). Available at http://www.people.umass.edu/aizen Psychosocial nursing therapy following sudden Ajzen, I., & Fishbein, M. (1980). Understanding atti- cardiac arrest: Impact on two-year survival. Nursing tudes and predicting social behaviour. Englewood Research, 50, 68–76. Cliffs, NJ: Prentice Hall. Department of Health. (2001). Treatment choice in Anderson, E. S., Winett, R. A., Wojcik, J. R., Winettt, psychological therapies and counselling. London: S. G., & Bowden, T. (2001). A computerized social HMSO. cognitive intervention for nutrition behaviour: Fishbein, M., & Middlestadt, S. E. (1989). Using the Direct and mediated effects in fat, fibre, fruits and theory of reasoned action as a framework for under- vegetables, self-efficacy, and outcome expectations standing and changing AIDS-related behaviors. In among food shoppers. Annals of Behavioural V. M. Mays, G. W. Albee, & S. F. Schneider (Eds.), Medicine, 23, 88–100. Primary prevention of AIDS: Psychological Armitage, C. J., & Conner, M. (2001). Efficacy of the approaches (pp. 93–110). Newbury Park, CA: Sage theory of planned behaviour: A meta-analytic Publications. review. British Journal of Social Psychology, 40, Godin, G., & Kok, G. (1996). The theory of planned 471–499. behaviour: A review of its applications to health- Bandura, A. (1997). Self-efficacy: The exercise of related behaviours. American Journal of Health control. New York: Freeman. Promotion, 11, 87–98. Beck, A. T. (1976). Cognitive therapy and the Gould, R. A., Buckminster, S., Pollack, M. W., Otto, emotional disorders. New York: International M. W., & Yap, L. (1997). Cognitive-behavioural and Universities Press. pharmacological treatment for social phobia: A

16 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 17

HOBBIS & SUTTON: COMPARING CBT WITH THE TPB

meta-analysis. Clinical Psychology: Science and Morrison, N. (2001). Group cognitive therapy: Practice, 4, 291–306. Treatment of choice of sub-optimal option? Behav- Graham, P. (1985). Psychology and the health of ioural and Cognitive Psychotherapy, 29, 311–332. children. Journal of Child Psychology and Nathan, P. E., & Gorman, J. M. (2002). A guide to Psychiatry, 26, 333–347. treatments that work, 2nd edn. Oxford: Oxford Hallam, J., & Petosa, R. (1998). A worksite inter- University Press. vention to enhance social cognitive theory Ogden, J. (2003). Some problems with social cognition constructs to promote exercise adherence. models: A pragmatic and conceptual analysis. American Journal of Health Promotion, 13, 4–7. Health Psychology, 22, 424–428. Hardeman, W., Johnston, M., Johnston, D. W., Padesky, C. A. (1993). Socratic questioning: Changing Bonetti, D., Wareham, N., & Kinmonth, A. L. minds or guiding discovery? Transcript of keynote (2002). Application of the theory of planned address delivered at the European Congress of behaviour in behaviour change interventions: A Behavioural and Cognitive Therapies, London, 24 systematic review. Psychology and Health, 17, September. 123–158. Persons, J. B. (1989). Cognitive therapy in practice: A Hardeman, W., Sutton, S., Griffin, S., Johnston, M., case formulation approach. New York: W. W. White, A., Wareham, N. J., & Kinmonth, A. L. Norton & Company. (submitted). Use of causal models in the develop- Prochaska, J. O., & Velicer, W. F. (1997). The trans- ment of theory-based behaviour change theoretical model of health behavior change. programmes for trial evaluation: Integrating American Journal of Health Promotion, 12, 38–48. psychology and epidemiology. Robinson, P. H., & Serfaty, M. A. (2001). The use of Henry, J. L., Wilson, P. H., Bruce, D. G., Chisholm, email in the identification of bulimia nervosa and its D. J., & Rawling, P. J. (1997). Cognitive-behavioural treatment. European Eating Disorders Review, 9, stress management for patients with non-insulin 182–193. dependent diabetes mellitus. Psychology, Health Roth, A., & Fonagy, P. (1996). What works for whom? and Medicine, 2, 109–118. A critical review of psychotherapy research. London: James, I. A. (2001). : The next gener- Guilford Press. ation, but should it carry a health warning? Behav- Sutton, S. (1998). Predicting and explaining intentions iour and Cognitive Psychotherapy, 29, 401–407. and behaviour: How well are we doing? Journal of Kaluza, G. (2000). Changing unbalanced coping Applied Social Psychology, 28, 1317–1338. profiles: A prospective controlled intervention trial Sutton, S. (2002a). Testing attitude–behaviour in worksite health promotion. Psychology and theories using non-experimental data: An examin- Health, 12, 423–433. ation of some hidden assumptions. European Lewin, R. J. P., Furze, G., Robinson, J., Griffith, K., Review of Social Psychology, 13, 293–323. Wiseman, S., Pye, M., & Boyle, R. (2002). A Sutton, S. (2002b). Using social cognition models to randomised controlled trial of a self-management develop health behaviour interventions: Problems plan for patients with newly diagnosed angina. and assumptions. In D. Rutter & L. Quine (Eds.), British Journal of General Practice, 52, 194–196, Changing health behaviour: Intervention and 199–201. research with social cognition models (pp. 193–208). Liao, K. L. (2000). Cognitive-behavioural approaches Buckingham: Open University Press. and weight management: An overview. Journal of Sutton, S. (2004). Determinants of health related the Royal Society of Health, 120, 27–30. behaviours: Theoretical and methodological issues. Marcus, B. H., Dubbert, P. M., Forsyth, L. H., In S. Sutton, A. Baum, & M. Johnston (Eds.), The McKenzie, T. L., Stone, E. J., Dunn, A. L., & Blair, Sage Handbook of Health Psychology. London: S. N. (2000). Physical activity behaviour change: Sage Publications. Issues in adoption and maintenance. Health Sutton, S. (2005). Stage theories of health behaviour. Psychology, 19(Supp. 1): 32–41. In M. Conner & P. Norman (Eds.), Predicting health Marcus, B. H., Owen, N., Forsyth, L. H., Cavill, N. A., behaviour: Research and practice with social cogni- & Fridinger, F. (1998). Physical activity inter- tion models (2nd edn). Buckingham: Open ventions using mass media, print media and infor- University Press. mation technology. American Journal of Sutton, S., French, D. P., Hennings, S. J., Mitchell, J., Preventative Medicine, 15, 362–378. Wareham, N. J., Griffin, S., Hardeman, W., & Mooney, K. A., & Padesky, C. A. (2000). Applying Kinmonth, A. L. (2003). Eliciting salient beliefs in client creativity to recurrent problems: Constructing research on the theory of planned behaviour: The possibilities and tolerating doubt. Journal of Cogni- effect of question wording. Current Psychology, 22, tive Psychotherapy: An International Quarterly, 14, 234–251. 149–161. Thomas, J., & Wainwright, P. (1996). Community

17 02 048549_hobbi_(jr/t) 11/11/04 2:09 pm Page 18

JOURNAL OF HEALTH PSYCHOLOGY 10(1)

nurses and health promotion: Ethical and political (1998). Stage theories of health behaviour: perspectives. Nursing Ethics, 3, 97–107. Conceptual and methodological issues. Health Webb, P. (1997). Some ethical issues in health and Psychology, 17, 290–299. patient education. In P. Webb (Ed.), Health Whittal, M. L., Agras, W. S., & Gould, R. A. (1999). promotion and education: A professional’s guide Bulimia nervosa: A meta-analysis of psychosocial (pp. 38–56). Cheltenham: Stanley Thornes (Pub- and pharmacological treatments. Behaviour lishers) Ltd. Therapy, 30, 117–135. Weinstein, N. D., Rothman, A. J., & Sutton, S. R.

18