Health Psychology and applied psychology, Study notes of Psychology

Download Health Psychology and applied psychology and more Study notes Psychology in PDF only on Docsity! In this chapter we examine the way in which personality dimensions or traits can determine health outcomes. These effects of personality on health include direct effects through physiological mechanisms and indirect effects through health behaviours. Personality traits refer to stable individual differences in thinking, feeling and behaving across a range of different situations. Research in the health domain has found that particular dimensions of personality are associated with poor health and reduced longevity while others are linked to good health and increased length of life. The magnitude of these effects can be similar to those of known biological risk factors such as cholesterol (Caspi et al, 2005). The personality dimensions associated with poor health outcomes include neuroticism (or negative affect), type A personality and hostility. The dimensions associated with good health outcomes include optimism, extraversion and conscientiousness. In this chapter we consider the evidence linking these personality variables to health outcomes and some of the mechanisms by which personality affects health. For example, personality traits might lead to greater exposure to stressful events, to a reduction in the effectiveness of coping strategies, or a change in coping resources such as social support. These explanations of the personality– health link build on the discussion of the impact of personality on coping in Chapter 5. Other personality traits such as hostility may affect health through changing the intensity and duration of physiological reactions to stress, linking personality to the biopsychosocial pathways considered in Chapter 2. In this chapter we will consider: 1) optimism and health; 2) type A behaviour and coronary heart disease; 3) hostility and coronary heart disease; 4) neuroticism and health; 5) extraversion and health; 6) conscientiousness and health. Introduction This chapter reviews evidence suggesting that stable individual differences in the way people think, feel and behave (i.e. personality) are predictive of various health outcomes. We explore how these stable individual differences can predispose individuals to respond to life challenges in a manner which, over time, 6 Personality and health 110 | Personality and health damages or protects their health. Much research on personality in recent years has focused on five broad personality types: openness to experience (or intellect), conscientiousness, extraversion, agreeableness and neuroticism (or emotional stability) (McCrae and Costa, 1987; Digman, 1990). This is often referred to as the Big Five Taxonomy or the OCEAN model of personality. A growing body of research now relates traits from the Big Five Taxonomy to various health behaviours and health outcomes. For example, Booth-Kewley and Vickers (1994) suggest that the Big Five personality traits may determine the extent to which people engage in general clusters of health-related behaviours such as substance use risk behaviours (e.g. smoking). However, there has been less research on how openness and agreeableness link to health outcomes so we will focus on extraversion, neuroticism and conscientiousness. The Big Five model is based on the assumption that a range of more specific personality traits can be understood as blends of the different Big Five traits. Some of the best evidence for the impacts of personality on health outcomes arises from work looking at more specific personality traits. So, for example, work has examined the impact of optimism or positive affect on health. We will consider work on optimism, type A behaviour pattern and hostility as important areas of research relating personality traits to health outcomes. Some research has also suggested a cancer or type C personality type (see Focus box 6.1) and a distressed or type D personality type (see Focus box 6.3). Since the magnitude of personality effects on health outcomes can be comparable to known biological factors, these effects must be taken seriously by health psychologists (Hampson et al, 2006). In the western world the leading causes of morbidity and mortality in middle and later life are now various chronic diseases such as coronary heart disease, cancer and diabetes, while in children, adolescents and young adults unintentional injuries are the leading causes of death. Personality traits may have important roles to play in both these periods of life. In relation to the development of chronic diseases, personality traits may play an important role in the maintenance of behaviours that are health promoting or health damaging when engaged in over time (e.g. smoking and unhealthy eating). Learning outcomes When you have completed this chapter you should be able to: 1. Explain how optimism is related to positive health outcomes and the role of attributional styles in this relationship. 2. Discuss the effects of type A personality and hostility on coronary heart disease and potential mediation of these effects through physiological reactions. 3. Describe the role of neuroticism (or negative affect) on poor health and the explanations of this effect through perceptions of stress, ability to cope and social support. 4. Describe the impact of extraversion on positive and negative health outcomes through effects on mood and health risk behaviours. 5. Describe the relationship between conscientiousness and positive health outcomes and the mediating effects of health behaviours. 6. Evaluate the different mechanisms through which personality variables affect health outcomes. to external, unstable and specific causes while pessimists see the same events as resulting from internal, stable and global causes (Peterson et al, 1988). For example, an optimist might believe that they got a minor illness because they were ‘run down’ after an unusually busy time at work. In contrast, a pessimist might believe they contracted a minor illness because they are always susceptible to such things no matter what they do. The outcomes of optimism include increased psychological well-being, better physical health and even greater longevity. For example, in relation to psychological well-being, Litt et al (1992) found that optimistic individuals were less depressed after unsuccessful in vitro fertilization. Similarly, Carver et al (1993) reported that optimism in women with breast cancer was associated with less distress following surgery and that this effect persisted one year later. Alloy, Abrahamson and Francis (1999) found that students with a pessimistic explanatory style were more likely to subsequently experience depression. Research also demonstrates that high levels of optimism are associated with better physical health. Those with high levels of optimism have fewer infectious illnesses and report fewer physical symptoms even during periods of stress (Peterson and Seligman, 1987). They are also more likely to recover from surgery more quickly and less likely to be re-hospitalized (Scheier et al, 1999). Peterson and colleagues provide an impressive demonstration of the effects of optimism on physical health. In a sample of men, attributional style measured at age 25 was found to predict health status 35 years later as assessed by doctors; the optimists were more likely to be in better health even when initial physical and mental health were statistically controlled for (Peterson et al, 1988). Most impressively, those with high levels of optimism may even live longer. Danner, Snowdon and Friesen (2001) coded pieces of text that a sample of 180 Catholic nuns had written about themselves on entering the church as young women, for emotional content. The research then examined the survival rates of these same women when they were 75–95 years of age. Those who wrote sentences containing self-descriptions with the most positive emotions (e.g. happiness, pride, love) were more likely to live longer than those containing the fewest positive emotions. Comparison of the top and bottom 25 per cent (quartiles) indicated that 24 per cent of those in the top quartile had died compared to 54 per cent of those in the bottom quartile. Similar results have been reported for men. Everson et al (1996b) examined the relationship between hopelessness and health outcomes in a large sample of men. Comparing the top and bottom 33 per cent (tertiles) showed that those in the top tertile compared to the bottom tertile for hopelessness were 3.5 times more likely to die from all causes of death, 4 times more likely to die from cardiovascular disease, and 2.5 times more likely to die from cancer. Similarly, men with AIDS who are optimistic live twice as long as men who are pessimistic (Reed et al, 1994). More generally, among older individuals, those with positive attitudes towards ageing live an average of 7.5 years longer than those with more negative attitudes (Levy et al, 2002). These studies on health outcomes highlight an important but subtle emerging distinction between optimism and positive affect. Whereas optimism refers to Optimism | 113 114 | Personality and health positive beliefs and feelings about the future, positive affect reflects a level of pleasurable engagement with the environment such as happiness, joy, excitement, enthusiasm and contentment. The two tendencies may overlap substantially as is shown in the Danner et al (2001) study. However, recent research has begun to examine the effects of positive affect independent of optimism (see Pressman and Cohen, 2005, for a review of positive affect and health). One important issue here is the extent to which positive affect is just the opposite of, or alternatively truly distinct from, negative affectivity, a personality trait we consider below. Currently there is evidence supporting both views. The explanation for the relationship between optimism and various health outcomes is still unclear. One interesting suggestion is that those high in optimism may be more likely to avoid certain high risk situations. Some supporting evidence for this view comes from Peterson and colleagues (1988) who showed that those with an optimistic attributional style were less likely to die from accidental or violent causes than those with a pessimistic style, while the two groups did not differ in respect of mortality from cancer or cardiovascular disease. A further explanation for the relationship between optimism and health is through the effects of optimism on coping strategies. Those high in optimism are more likely to use adaptive and functional strategies for coping with problems such as acceptance, rational thinking, social support and positive reframing. For example, Scheier, Weintraub and Carver (1986) conducted a study in which students had to write about coping with stressful Figure 6.1 We can all show different ‘faces’ to the world. Personality traits tap consistencies in how we respond. Such consistencies have been found to play important roles in the development of illness. situations. Optimists were found to be more likely to use strategies such as making a plan and sticking to it, focusing intently on the problem, and seeking social support. Optimists were also less likely to distract themselves from thinking about the problem. Scheier et al (1989) reported similar differences between optimists and pessimists in the way they coped with recovery from surgery that resulted in faster recovery among the optimists. The use of more constructive coping strategies may lead to better health outcomes, partly by helping individuals to avoid negative life events and also by helping them to confront and deal with problems earlier and more effectively. A further explanation focuses on the effect of pessimism on physiological reactions to stress in terms of immune functioning and cardiovascular response (Scheier and Carver, 1987). Some support of this explanation can be found in studies that have shown immune responses to be lower in pessimists (Segerstrom et al, 1998). Type A behaviour Type A behaviour pattern is typified by a competitive drive, aggression, chronic impatience and a sense of time urgency (Rosenman et al, 1976). This type of behaviour is contrasted with the opposite cluster of characteristics, type B behaviour pattern, which leads to a more relaxed, laid-back approach to life. The concept of type A behaviour originated from the work of two cardiologists, Meyer Friedman and Ray Rosenman, who realized that their patients’ disease was not fully explained by conventional risk factors such as dietary cholesterol and smoking. For years they failed to look beyond the physical symptoms and to consider the signs of stress in their patients, even though patients tended to sit on the edge of waiting room chairs to the extent that an upholsterer commented that the front edges of the waiting room chairs were unusually worn (Friedman and Rosenman, 1974). Eventually, however, they sent out a questionnaire asking 150 businessmen what they believed had precipitated a heart attack in a friend. Few thought it was due to diet or smoking and most felt it was due to ‘excessive competitive drive and meeting deadlines’ (Rosenman et al, 1964: 73). A subsequent study suggested that physicians agreed even though this was not a recognized cause in the medical literature of the time. This and subsequent research ultimately led to the identification of the constellation of characteristics described above, and its long-term investigation in a large prospective study known as the Western Collaborative Group Study. This examined risk factors for coronary heart disease (CHD) in a sample of over 3000 healthy middle-aged men. The study started in 1960 and followed participants for more than 27 years. Rosenman and colleagues assessed the participants in the study using a structured interview. This interview involved the interviewer asking questions in a confrontational manner (including interrupting the participant) with the aim of provoking the participant in order to assess aggression and time urgency (Chesney et al, 1980). The men were then followed up at 8.5 and 22 years. The researchers found after 8.5 years that those men who were classified as type A had around twice the risk of developing CHD as those who were type B, even after controlling for other risk factors. At this stage it appeared that type A was Type A behaviour | 115 expressive component of hostility (i.e. verbally and physically aggressive behaviour). These studies suggested that the effects were at least similar in magnitude to those reported for traditional risk factors such as smoking, high blood pressure and cholesterol. Even among studies using self-report measures (the Cook–Medley scale: Cook and Medley, 1954), the review found small but consistent relationships with heart disease. It should be noted, however, that a more recent meta-analysis offers a less positive interpretation based on a smaller subset of papers (Myrtek, 2001), i.e. they suggest that while the effects are significant they are very small indeed. In the main, subsequent studies and reviews continue to suggest that hostility plays a role in causing CHD (e.g. Gallo and Matthews, 2003) and also hypertension (Rutledge and Hogan, 2002). Focus box 6.2 discusses how we can reduce hostility. Interventions to reduce hostility There is now evidence suggesting that a hostility-reduction intervention aimed at CHD patients with high levels of hostility may reduce risks for heart disease. Gidron, Davidson and Bata (1999) conducted a randomized controlled trial in which 22 hostile male patients were assigned to either a treatment or control group. Hostility was assessed by observation during a structured interview and by self-ratings. The hostility-reduction intervention involved eight 90-minute weekly group meetings using cognitive behaviour techniques. Participants were taught skills to reduce antagonism, cynicism and anger. They were also asked to rate their hostility in a daily log and to record their use of skills. The control group had a one-session group meeting giving information about the risks of hostility and about basic hostility-reduction skills. The participants were followed up immediately after the trial and again after two months. Those in the intervention group were observed to be, and rated themselves to be, less hostile at follow-up than the controls. They also had lower diastolic blood pressure. Furthermore, reductions in hostility were correlated with reductions in blood pressure. In a subsequent paper, Davidson et al (2007) conducted secondary analysis of the data from the above study. They found that patients who received the intervention tended to have fewer hospital admissions in the six months following the intervention, and, importantly, had significantly fewer days in hospital (a mean of 0.38 days compared with a mean of 2.15 days for the control group). Consequently, their hospitalization costs were less. While more studies are needed with larger and more diverse groups, these findings suggest there may be potential to design efficacious and cost-effective hostility-reduction treatments. 118 | Personality and health Focus 6.2 Hostility | 119 The possible mechanisms underlying the effects of hostility have also been discussed in some detail. Smith et al (2004) discuss five possible models: 1. Psychophysiological reactivity model. This suggests that hostile individuals show exaggerated cardiovascular and neuroendocrine responses to stressors. 2. Psychosocial vulnerability model. This model suggests that hostile individuals experience more interpersonal conflict. Hostility may lead to more stress and also be associated with less social support. 3. Transactional model. This combines the above two models and suggests that hostile individuals experience more interpersonal conflict and also have greater physiological reactivity – a ‘double whammy’ effect. 4. Health behaviour model. This suggests that hostility affects health via the mediating impact of poorer health behaviours. For example, hostile people may be cynical about health warnings or resistant to medical advice. 5. Constitutional vulnerability model. This model raises the possibility that individual differences (which might be genetic) are associated with both the personality tendency and the disease risk, i.e. the association between hostility and CHD is due to a third variable. Overall, Smith et al conclude that there is considerable support for a number of these models. Hostile people do display heightened physiological responses; they also experience increased levels of conflict and less social support. However, research has not yet established whether these tendencies mediate the relationship between hostility and health. There is some evidence that hostile people do display poorer health behaviours but it is also clear that this does not wholly account for the relationship between hostility and health. Finally, the development of molecular genetics offers opportunities to explore the constitutional vulnerability model. Further research is awaited on these mechanisms. However, it is possible that several mechanisms play a part in explaining the association between hostility and health. An interesting possibility in relation to the development of hostility is suggested by the work of Matthews et al (1996). In this work, negative behaviours during parent–son discussions aimed at resolving disagreements were observed in 51 Caucasian adolescent (12–13 years of age) boys. Results showed that the frequency of negative behaviours in the family discussions predicted hostility and expressed anger assessed three years later even after controlling for baseline hostility. This would suggest that hostility may be nurtured within particular family backgrounds that are characterized by negative behaviours during interactions. In contrast, work by Caspi et al (1997) shows that measures of temperament taken at 3 years old predict later health-related risk behaviour in early adulthood and that this effect is mediated by personality measures taken in late adolescence. This would appear to be good evidence that personality traits are something we are born with or at least develop very early in life and remain stable throughout our lives (i.e. a nature explanation of personality). Together these studies suggest that, while certain aspects of personality may be stable from a very young age, other aspects change and develop over time as a result of our interaction with our environment. This would appear to suggest the relevance of both a nature and a nurture explanation of personality trait development. Type D personality Similar to type A, type D personality is a risk factor for coronary heart disease. The type D, or distressed, personality refers to individuals who experience high levels of negative emotions (negative affectivity) and inhibit the expression of these negative emotions in social interactions (social inhibition). The concept was introduced by Johan Denollet, of Tilburg University in the Netherlands. Type D personality can be assessed by a self-report questionnaire containing items that tap negative affectivity (e.g. ‘I often make a fuss about unimportant things’ or ‘I often feel unhappy’) and social inhibition (e.g. ‘I often feel inhibited in social interactions’ or ‘I find it hard to start a conversation’). A type D individual would be someone who scores highly on both of these dimensions. This is important because previous research has shown negative affectivity or neuroticism to be related to various negative health outcomes. Denollet and colleagues have shown the type D personality to be a risk factor for adverse health outcomes in cardiac patients. So, for example, Denollet et al (1996) assessed type D in a sample of 286 cardiac patients who were receiving treatment. Approximately one-third of the sample were classified as type D. Approximately eight years later, the patients were followed up. Among those classified as type D a total of 27 per cent had died compared with a total of 7 per cent of the rest of the sample. A majority of the deaths were due to heart disease or stroke. This translates into an odds ratio of almost four (i.e. being four times more likely to die if classified as type D compared to those not classified as type D). These effects have been replicated in several studies. Type D is also related to various forms of psychological distress in cardiac patients including depression and anxiety. The explanation for the relationship between type D and risk of death is not entirely clear. Those with type D personalities appear to have more highly activated immune systems and more inflammation (perhaps indicating more damage to blood vessels in the heart and throughout the body). They also show greater increases in blood pressure in reactions to stress. Recent research has suggested that type D individuals engage in fewer health behaviours and experience lower levels of social support and that these effects remain after controlling for neuroticism (Williams et al, 2008). Neuroticism Neuroticism is one of the Big Five personality traits. It refers to the tendency to commonly experience negative emotions such as distress, anxiety, fear, anger and guilt (Watson and Clark, 1984). Because of the focus on negative emotions 120 | Personality and health Focus 6.3 Conscientiousness | 123 behaviours, extraversion appears to be associated with both health-protective behaviours like exercise (Rhodes et al, 2002) and health-risking behaviours like smoking (Booth-Kewley and Vickers, 1994)! Figure 6.2. The Rorschach ink blot test for personality. Most modern personality tests employ questionnaires to assess different personality traits. Earlier tests employed more projective tests such as the Rorschach ink blot test shown here where respondents were required to interpret ink blots and their responses used to classify their personality. Such tests however have low reliability and scores tend to vary depending on the psychologist doing the interpretation. Source: © Science Museum. Activity 6.1 The accident-prone personality? Some research has addressed the idea that certain personality traits are precursors of accidents or unintentional injuries, i.e. the accident-prone personality. The best evidence supporting such a personality type comes from studies focusing on impulsivity. This research shows that childhood impulsivity predicts injuries both during childhood and later life (Caspi et al, 1997; Cooper et al, 2003). What mechanisms might explain how impulsivity is related to injuries? Try reading these two articles and coming up with a list of potential mechanisms. Conscientiousness Conscientiousness refers to the ability to control one’s behaviour and to complete tasks. Highly conscientious individuals are more organized, careful, dependable, self-disciplined and achievement-oriented than those low in conscientiousness (McCrae and Costa, 1987). High conscientiousness has also been associated with a greater use of problem-focused, positive reappraisal and support-seeking coping strategies (Watson and Hubbard, 1996), and a less 124 | Personality and health frequent use of escape-avoidance and self-blame coping strategies (O’Brien and Delongis, 1996). Conscientiousness is also associated with a propensity to follow socially prescribed norms for impulse control (John and Srivastava, 1999; Bogg and Roberts, 2004). In the last few years measures of conscientiousness with good levels of reliability and validity have become available. For example, the International Personality Item Pool contains statements such as ‘Am always prepared’, and ‘Am exacting in my work’ in order to tap conscientiousness. Those high in conscientiousness are more likely to consider these statements as accurate self-descriptions. A growing body of research shows conscientiousness to have impacts on health behaviours, health outcomes and even longevity. The key evidence for the impact of conscientiousness on longevity comes from the Terman Life-Cycle personality cohort study. In this highly regarded study, a sample of over 1000 children born around 1910 completed various measures every five to ten years from the age of 11. The original sample of children were selected to have above average IQ and were drawn from the area around the Californian cities of San Francisco and Los Angeles. The personality assessments included measures of conscientiousness, optimism, self-esteem, sociability, stability of mood and energy level. Friedman et al (1993) reported that of these variables only conscientiousness was significantly associated with lower mortality over time. The degree of association was such that those high in conscientiousness were more likely to live longer (by about two years) compared to those low in conscientiousness. Comparing the top and bottom 25 per cent (quartiles) on conscientiousness indicated that those in the bottom quartile were one and one half times more likely to die in any one year compared to those in the top quartile. Figure 6.3 shows the survival curves for participants in the Terman sample separately for males and females and for those with and without divorced parents among those with high and low levels of conscientiousness. 1.00 0.80 0.00 50 60 70 Age 80 Females whose parents did not divorce Males whose parents did not divorce Males whose parents divorced 90 0.20 0.40 0.60 P ro ba bi lit y of d ea th Low conscientiousness High conscientiousness Figure 6.3 Probability of dying (survival curves) at different ages for the Terman sample for different groups of males in comparison to females. Source: © 1993 Howard S. Friedman and Joseph E. Schwartz. Reprinted by permission. For further information about this project, see www.faculty.urc.edu/~friedman/. An important mechanism by which conscientiousness may influence health is through health behaviours. Friedman H.S. et al (1995) showed that the impact of conscientiousness on longevity in the Terman sample was partly accounted for by its effect on smoking and alcohol use, that is, conscientious children were less likely to become heavy smokers and drinkers. Consistent with these findings, Booth-Kewley and Vickers (1994) found that conscientiousness was more strongly correlated with clusters of health-related behaviours than the other Big Five traits, and was particularly strongly associated with health protection and accident control behaviours. Similarly, Courneya and Hellsten (1998) reported that, of the Big Five traits, conscientiousness was most strongly related to engaging in exercise behaviours, while Siegler, Feaganes and Rimer (1995) showed that regular mammography attendance was predicted by conscientiousness and extraversion. A comprehensive meta-analysis of work on the relationship between conscientiousness and behaviours (Bogg and Roberts, 2004) showed conscientiousness to be positively related to a range of protective health behaviours (e.g. exercise) and negatively related to a range of risky health behaviours (e.g. smoking). Table 6.2 shows the size of these effects for a range of different health behaviours (the impact of health behaviours on health is further considered in Chapter 7). A further way in which conscientiousness may impact on health outcomes is through modifying behaviour following illness. So, for example, some studies have demonstrated that individuals high in conscientiousness are more likely to follow health care advice and that this difference is particularly apparent when the advice is difficult or time consuming to follow (Christiansen and Smith, 1995; Schwartz J.E. et al, 1999). Recent research has begun to examine how personality traits may produce changes in health behaviours through shaping the way in which individuals think about these behaviours. This work suggests that our thoughts and feelings about performing a particular health behaviour (e.g. exercising) are a primary determinant of whether we perform that behaviour. That is, we tend to engage in behaviours that we have positive thoughts and feelings about (see Chapter 7). In this view, conscientiousness might, for example, influence the amount of exercise we do as a result of shaping our thoughts and feelings about exercising (i.e. thoughts and feelings mediate the impact of conscientiousness on exercise). Such mediation effects have been demonstrated by Siegler et al (1995) who found that the effect of conscientiousness on mammography attendance was mediated by knowledge of breast cancer and the perceived costs of seeking mammography. Similarly, the impact of conscientiousness on the self-care activities of patients with type 1 diabetes has been found to be mediated by treatment beliefs (e.g. Christensen et al, 1999). However, other research has found both mediated and direct effects for conscientiousness when predicting health behaviour (Vollrath et al, 1999; Conner and Abraham, 2001). In addition to mediation effects, conscientiousness might also operate as a moderator changing the relationship between health beliefs and health behaviour. A few studies have examined the moderating role of conscientiousness. For example, in a retrospective study, Schwartz M.D. et al (1999) found that conscientiousness moderated the relationship between breast- cancer-related distress and mammography uptake such that, among those with Conscientiousness | 125 128 | Personality and health A final set of explanations for the relationship between personality traits and health outcomes relates to stress and the variables that protect against the effects of stress. So, for example, individuals high in neuroticism may perceive themselves as experiencing more stress. Such individuals may also be less likely to employ appropriate coping mechanisms or have access to coping resources such as social support to deal with this stress. In this case it may be the stress that causes the negative health outcomes, but it is high levels of neuroticism that cause the stress and the inability to cope appropriately with the stress. Penley and Tomaka (2002) provide an interesting discussion of the relationship between all of the Big Five personality traits and both stress and coping. Correlation and inferences of causation When an independent (or predictor) variable (e.g. social support or attitude) is measured at the same time as a dependent (or outcome) variable (e.g. immune functioning or condom use) this is known as a cross-sectional study. When the dependent variable is measured at a later time then this is known as a longitudinal or prospective study. For example, if we measure job stress and then follow up our participants a year later this is a prospective study. Prospective studies offer more reassurance regarding the direction of causation because we know that the independent variable measure preceded the dependent variable measure in time. Prospective studies also allow us to control for levels of a dependent variable at time 1 so that we can predict change in the dependent variable from an independent variable. For example, we might find that lower reported social support (at time 1) predicts increases in stress over the following year. Thus while we might use analysis of variance (ANOVA) to test whether an association between an independent and dependent variable is likely to be replicable, we can use analysis of covariance (ANCOVA) to assess the degree to which an independent variable can predict change in a dependent variable over time by including a baseline measure of the dependent variable as a covariate. The direction of causation is ideally assessed in an experimental study in which we manipulate (rather than measure) the independent variable. In health psychology, interventions such as behaviour change interventions provide good examples of experimental methodology. For example, one group may receive an intervention to change attitudes or reduce work stress while another (control) group receives no intervention. If participants are randomly allocated to these two groups (to try to distribute confounding factors across groups) or matched (to balance confounding factors) then any difference in the dependent variable following the manipulation (that is, the intervention) can be reasonably attributed to that manipulation. The classic use of experimental methodology in health psychology is the randomized controlled trial (RCT). It is of course worth noting that such experimental methods merely establish one causal determinant of the dependent variable, they do not necessarily demonstrate that this is the one and only causal determinant. Research methods 6.1 Unfortunately, we often cannot manipulate independent variables in health psychology and so must infer underlying causal mechanisms from correlational data. For example, in relation to smoking, the majority of the evidence supporting an impact of smoking on cancer and cardiovascular disease outcomes is correlational, at least for studies in humans. Similarly, the relationship between personality and health is based on correlational data. There are well-known dangers in drawing causal inferences from correlational data. Two key issues are causal direction and the third variable problem. Causal direction refers to the issue of the direction of effect being unknown when two variables are correlated: did A cause B or B cause A? For example, in relation to personality and health this issue becomes one of whether a personality trait resulted in a health outcome or the health outcome produced the personality trait. So, for example, some patients with serious illnesses such as cancer may become anxious and neurotic. This might lead us to the erroneous conclusion that neuroticism played a role in causing the cancer when in fact the cancer had produced increased neuroticism. Third variable problems refer to the possibility that a correlation between two variables might be due to both variables being caused by a third variable. So, for example, there is some evidence that a hyper-responsive nervous system is an underlying factor in both the development of an anxious personality (high neuroticism) and the development of heart disease. Here an anxious, reactive personality would be related to (that is, correlated with) heart disease without being a causal determinant of heart disease (McCabe et al, 2000). Similarly, Eysenck (1967) argued that extraversion relates to differences in the sensitivity of the nervous system that influences emotional reactions and reactions to socialization. Extraverts may also be more likely to seek stimulation through behaviours such as smoking. In both these cases it is not the personality trait itself which causes the health outcomes but an underlying biological mechanism that causes both the personality trait and the health outcome. Summary A number of personality traits show significant relationships with various health outcomes such as morbidity and mortality. Indeed some of these relationships are of a similar size to those reported for more well-known risk factors like blood cholesterol levels. Of the Big Five personality traits that form much of the focus in modern-day personality research there is good evidence relating low levels of neuroticism and high levels of extraversion and conscientiousness to health outcomes (e.g. lower levels of illness and greater longevity). Evidence also suggests that optimism is positively related to health outcomes, while type A behaviour and hostility (low agreeableness) are negatively related to health outcomes. Summary | 129 In Chapter 6 we examined the part that personality plays in determining health outcomes. In this chapter we examine health cognitions, foundational to motivation, which have been found to differentiate between people who do and do not perform health behaviours. This is followed by an examination of how health-related motivation can be changed in Chapter 8, and how health-related behaviours can be changed in Chapter 9. The prevalence of health-related behaviours varies across social groups. For example, smoking is more prevalent among those from more economically deprived backgrounds. This would suggest that these factors might be the focus of interventions to change health-related behaviours. However, socio- demographic factors may be impossible to change or may require political intervention at national or international levels (such as changes in income distribution or taxation). For that reason a considerable body of research has examined more modifiable factors that may mediate (and explain) the relationship between socio-demographic factors and health-related behaviours. A particularly promising set of factors are the thoughts and feelings the individual associates with the particular health-related behaviour. These are known as health cognitions and are the focus of this chapter. We will consider: 1) predicting health behaviours; 2) social cognition models; 3) a critical appraisal of social cognition models; 4) the intention–behaviour gap. 7 Learning outcomes When you have completed this chapter you should be able to: 1. Describe the key health cognitions associated with performing health behaviours. 2. Explain what the cognitive determinants of health behaviours are according to: a) the health belief model; b) protection motivation theory; c) theory of planned behaviour; and d) social cognitive theory. 3. Evaluate the contribution of stage models to the understanding of change in health behaviours. 4. Critically evaluate the contribution of social cognition models to understanding the determinants of health behaviours. 5. Describe the intention–behaviour gap in relation to health behaviours. Health cognitions and health behaviours models’ because of their use of a number of cognitive variables to predict and understand individual behaviours, including health behaviours. It is important to note at the outset that these models focus on behaviour-specific cognitions as determinants of the relevant behaviour. For example, on this view healthy eating is best understood in terms of cognitions about healthy eating rather than more general thoughts and feelings about health. How do health behaviours impact on health outcomes? A great many studies have now looked at the relationship between the performance of health behaviours and a variety of health outcomes (e.g. Doll et al, 1994). Large scale epidemiological studies have demonstrated the importance of a variety of health behaviours for both morbidity and mortality. For example, the Alameda County study, which followed nearly 7000 people over 10 years, found that seven key behaviours were associated with lower morbidity and longer life: not smoking, moderate alcohol intake, sleeping seven to eight hours per night, exercising regularly, maintaining a desirable body weight, avoiding snacks and eating breakfast regularly (Belloc and Breslow, 1972; Breslow and Enstrom, 1980). Health behaviours are assumed to influence health through three major pathways (Baum and Posluszny, 1999): first, by generating direct biological changes such as when excessive alcohol consumption damages the liver; second, by changing exposure to health risks, as when the use of a condom protects against the spread of HIV; and third, by ensuring early detection and treatment of disease, as when testicular or breast self-examination leads to early detection of a cancer that can more easily be treated. Can you think of further examples of the pathways through which health behaviours might exert their effects on health? Social cognition models Social cognition models describe the important cognitions that distinguish between those who do and do not perform health behaviours. This approach focuses on the cognitions or thought processes that intervene between observable stimuli and behaviour in real world situations (Fiske and Taylor, 1991). This ‘social cognition’ approach has been central to social psychology over the past quarter of a century. Unlike behaviourism, it is founded on the assumption that behaviour is best understood as a function of people’s perceptions of reality, rather than objective characterizations of environmental stimuli. 138 | Health cognitions and health behaviours Focus 7.2 Research into social cognition models can be seen as one part of what has been called ‘self-regulation’ research. Self-regulation processes are defined as those ‘. . . mental and behavioral processes by which people enact their self- conceptions, revise their behavior, or alter the environment so as to bring about outcomes in it in line with their self-perceptions and personal goals’ (Fiske and Taylor, 1991: 181). Self-regulation research has emerged from a clinical tradition in psychology which views the individual as striving to eliminate dysfunctional patterns of thinking or behaviour and engage in adaptive patterns of thinking or behaviour (Bandura, 1982; Turk and Salovey, 1986). Self-regulation involves cognitive re-evaluation of beliefs, goal-setting and ongoing monitoring and evaluating of goal-directed behaviour. Two phases of self-regulation activities have been defined: motivational and volitional (Gollwitzer, 1990). In the motivational phase costs and benefits are considered in order to choose between goals and behaviours. This phase is assumed to conclude with a decision concerning which goals and actions to pursue at a particular time. In the subsequent volitional phase, planning and action directed towards achieving the set goal predominate. Much of the research with health behaviours has focused on the important cognitions in the motivational phase, although recent research has begun to focus on the volitional phase. The key social cognition models in this area are: 1. The health belief model (HBM; e.g. Janz and Becker, 1984; Abraham and Sheeran, 2005). 2. Protection motivation theory (PMT; e.g. Maddux and Rogers, 1983; Norman et al, 2005). 3. The theory of reasoned action/theory of planned behaviour (TRA/TPB; e.g. Ajzen, 1991; Conner and Sparks, 2005). 4. Social cognitive theory (SCT; e.g. Bandura, 2000; Luszczynska and Schwarzer, 2005). A distinct set of models focus on the idea that behaviour change occurs through a series of qualitatively different stages. These so-called ‘stage’ models (Sutton, 2005) importantly include the transtheoretical model of change (Prochaska and DiClemente, 1984; Prochaska et al, 1992). In the following sections we consider these different models and what they say about how cognitions help direct health behaviours. These social cognition models (SCMs) provide a basis for understanding the determinants of behaviour and also provide important targets which interventions designed to change behaviour should focus on if they are to change motivation (see Chapter 8) and, thereby, behaviour (see Chapter 9). The health belief model The health belief model (HBM) is the earliest and most widely used SCM in health psychology (see Abraham and Sheeran, 2005, for a review). For example, Hochbaum (1958) found that perceived susceptibility to tuberculosis and the belief that people with the disease could be asymptomatic (so that screening would be beneficial) distinguished between those who had and had not attended for chest X-rays. Haefner and Kirscht (1970) took this research further by Social cognition models | 139 140 | Health cognitions and health behaviours demonstrating that health education interventions designed to increase participants’ perceived susceptibility, perceived severity and anticipated benefits resulted in a greater number of check-up visits to the doctor compared to controls over the following eight months. The HBM suggests that health behaviours are determined mainly by two aspects of individuals’ representations of health behaviour: perceptions of illness threat and evaluation of behaviours to counteract this threat (see Figure 7.1). Threat perceptions are based on two beliefs: the perceived susceptibility of the individual to the illness (‘Am I likely to get it?’); and the perceived severity of the consequences of the illness for the individual (‘How bad would it be?’). Similarly, evaluation of possible responses involves consideration of both the potential benefits of, and barriers to action. Together these four beliefs are believed to determine the likelihood of the individual undertaking to perform a health behaviour. The particular action taken is determined by the evaluation of the available alternatives, focusing on the benefits or efficacy of the health behaviour and the perceived costs or barriers of performing the behaviour. Hence individuals are more likely to follow a particular health action if they believe themselves to be susceptible to a particular condition which they also consider to be serious, and believe that the benefits of the action taken to counteract the health threat outweigh the costs. For example, an individual is likely to quit smoking if he or she: believes him or herself to be susceptible to smoking-related illnesses; considers the illnesses to be serious; and that, of the alternative behaviours open to him/her, considers quitting smoking to be the most effective way to tackle his/her susceptibility to smoking-related illnesses. Two other variables often included in the model are cues to action and health motivation. Cues to action are assumed to include a diverse range of triggers to Health motivation Behaviour Cues to Action EXTERNAL VARIABLES Demographic Variables Age, sex, occupation, socioeconomic status, religion, education Personality Traits Extraversion Agreeableness Conscientiousness Neuroticism Openness Other Psychological Factors Peer pressure Self-efficiency Threat (motivation) Response Effectiveness Costs Benefits Perceived severity Perceived susceptibity Figure 7.1 The health belief model. Social cognition models | 143 Theory of planned behaviour The theory of planned behaviour (TPB) was developed by social psychologists and has been widely applied to understanding health behaviours (Conner and Sparks, 2005). It specifies the factors that determine an individual’s decision to perform a particular behaviour (see Figure 7.3). Importantly this theory added ‘perceived behavioural control’ to the earlier theory of reasoned action (Ajzen and Fishbein, 1980) which continues to be applied (Ajzen, 2001). The TPB proposes that the key determinants of behaviour are intention to engage in that behaviour and perceived behavioural control over that behaviour. As in the PMT, intentions in the TPB represent a person’s motivation or conscious plan or decision to exert effort to perform the behaviour. Perceived behavioural control (PBC) is a person’s expectancy that performance of the behaviour is within their control and confidence that they can perform the behaviour. PBC is similar to Bandura’s (1982) concept of self-efficacy used in the PMT. In the TPB, intention is itself assumed to be determined by three factors: attitudes, subjective norms and PBC. Attitudes are the overall evaluations of the behaviour by the individual as positive or negative (and so include beliefs about benefits and barriers included in the HBM). Subjective norms are a person’s beliefs about whether significant others think they should engage in the behaviour. PBC is assumed to influence both intentions and behaviour because we rarely intend to do things we know we cannot and because believing that we can succeed enhances effort and persistence and so makes successful performance more likely (see Chapter 8). Thus, according to the TPB, smokers are likely to quit smoking if they form an intention to do so. Such an intention to quit is likely to be formed if smokers have a positive attitude towards quitting, if they believe that people whose views they value think they should quit smoking, and if they feel that they have control over quitting smoking. Attitudes are based on behavioural beliefs, that is, beliefs about the perceived consequences of behaviours. In particular, they are a function of the likelihood of Behavioural intention Belief about outcomes × Evaluation of outcomes Normative beliefs × Motivation to comply Perceived likehood of occurrence × Perceived facilitating/ inhibiting power Subjective norm Behaviour EXTERNAL VARIABLES Demographic Variables Age, sex, occupation, socioeconomic status, religion, education Personality Traits Extraversion Agreeableness Conscientiousness Neutroticism Openness Perceived behavioural control Attitude towards behaviour Figure 7.3 Theory of planned behaviour. a consequence occurring as a result of performing the behaviour and the evaluation of that outcome (i.e. ‘Will it happen?’ and ‘How good or bad will it be?’). It is assumed that an individual will have a limited number of consequences in mind when considering a behaviour. Thus a positive attitude towards quitting smoking will result when more positive than negative consequences are thought to follow quitting. Subjective norm is based on beliefs about salient others’ approval or disapproval of whether one should engage in a behaviour (e.g. ‘Would my sexual partner approve?’, ‘Would my best friend approve?’). These beliefs are weighted by the ‘motivation to comply’ with each salient other on this issue (e.g. ‘Do I care what my sexual partner/best friend thinks about this?’). Again it is assumed that an individual will only have a limited number of referents in mind when considering a behaviour. Thus the more people (whose approval is seen to be important) who are thought to approve of the action, the more positive the subjective norm. Judgements of PBC are influenced by control beliefs concerning whether one has access to the necessary resources and opportunities to perform the behaviour successfully, weighted by the perceived power, or importance, of each factor to facilitate or inhibit the action. These factors include both internal control factors (information, personal deficiencies, skills, abilities, emotions) and external control factors (opportunities, dependence on others, barriers). As for the other types of beliefs it is assumed that an individual will only consider a limited number of control factors when considering a behaviour. So, for example, in relation to quitting smoking, a strong PBC to quit smoking would be expected when a smoker believes there are more factors that facilitate than that inhibit quitting smoking, especially if the inhibiting factors do not have strong effects on the feasibility of quitting. The TPB has at least two advantages over the extended HBM. First (as in PMT), health beliefs are seen to affect behaviour indirectly, in this case through attitude and intention. Thus the model outlines a mechanism by which particular beliefs combine to influence motivation and action. Second, the model takes account of social influence on action. The TPB has been widely tested and successfully applied to the understanding of a variety of behaviours (for reviews see Ajzen, 1991; Conner and Sparks, 2005). For example, in a meta-analysis of the TPB Armitage and Conner (2001) reported that across 154 applications, attitude, subjective norms and PBC accounted for 39 per cent of the variance in intention, while intentions and PBC accounted for 27 per cent of the variance in behaviour across 63 applications. Intentions were the strongest predictors of behaviour, while attitudes were the strongest predictors of intentions. The TPB has also informed a number of interventions designed to change behaviour. For example, Hill, Abraham and Wright (2007) employed a randomized controlled trial to test the effectiveness of a TPB-based leaflet compared to a control in promoting physical exercise in a sample of school children. The leaflet condition compared to the control condition significantly increased not only reported exercise but also intentions, attitudes, subjective norms and PBC. Additional analyses indicated that the impact on exercise was mediated (i.e. partly explained) by the increases the leaflet had produced (compared to the control group) in intentions and PBC. 144 | Health cognitions and health behaviours Social cognition models | 145 Recent work with the TPB (see Conner and Sparks, 2005) has suggested the value of dividing attitude, subjective norm and PBC each into two components to form the ‘two-factor TPB’ (Figure 7.4). Attitude is divided into an affective or feeling component and a cognitive or instrumental component. The first concerns beliefs and evaluations about how it will feel to perform the behaviour while the second includes beliefs and evaluation about other consequences. So, for example, quitting smoking might be perceived as both unenjoyable (affective evaluation) but beneficial (cognitive evaluation). As well as subjective norms (defined above), the two-factor model includes descriptive norms. Descriptive norms refer to perceptions of what others are doing (‘e.g. all my friends are doing it’) rather than beliefs about others’ approval of the target individual performing the behaviour. For example, a smoker might believe that important others approved of him or her quitting but those other individuals to have not quit smoking themselves. PBC is divided into perceived control and perceived confidence. So, for example, one might perceive that quitting smoking is within one’s control but not feel confident that one can easily quit smoking. The latter factor is most like self-efficacy and has been found to be the stronger predictor of intentions and behaviour (Rodgers et al, 2008). Social cognitive theory In social cognitive theory (SCT; Bandura, 1982) behaviour is held to be determined by three factors: goals, outcome expectancies and self-efficacy (see Figure 7.5). Goals are plans to act and can be conceived of as intentions to perform the behaviour (see Austin and Vancouver, 1996; Luszczynska and Schwarzer, 2005). Outcome expectancies are similar to behavioural beliefs in the TPB but here are split into physical, social or self-evaluative depending on the nature of the consequences considered. Thus, in this model, beliefs about others’ approval (subjective norms in the TPB) are grouped with beliefs about other consequences. Self-efficacy is the belief that a behaviour is or is not within an individual’s control and is usually assessed as the degree of confidence the individual has that they could still perform the behaviour in the face of various Affective attitudes Cognitive attitudes Injunctive norms Attitude Subjective norm Intention Actual behavioural control Behaviour PBC Descriptive norms Perceived confidence Perceived control Figure 7.4 The two-factor theory of planned behaviour. 148 | Health cognitions and health behaviours evidence for stage models would be where we showed that interventions matched to individuals’ stage of change were more effective in producing behaviour change than interventions mismatched to an individual’s stage (although see also Abraham, 2008). So, for example, in a matched intervention outcome expectancies might be targeted in individuals in the contemplation stage, while self-efficacy was targeted in individuals in the action stage, and this would be reversed in a mismatched intervention. Unfortunately, few such matched–mismatched studies have produced evidence supportive of stage models (see Littell and Girvin, 2002 for a systematic review of the effectiveness of interventions applying the TTM to health-related behaviours). Thus, at present, research findings do not support the added complexity and increased cost of stage-tailored interventions. West (2005) in reviewing stage models has recently suggested that work on the TTM should be abandoned. It is difficult to usefully categorize people as ‘pre-contemplators’ or those ‘in preparation’ because people frequently cycle between such states as their motivation to change shifts. Nonetheless, an individual at a particular time may be more focused on deciding whether or not to act or on ensuring that they act on a prior decision to act (i.e. an intention). This is captured by the terms ‘motivational phase’ and ‘volitional phase’, respectively. This two-phase conception of action readiness suggests that health promoters need to think about how they can consolidate people’s motivation to act and how they can help people to enact their intentions (see Chapters 8 and 9). In general, the social cognition models considered in this chapter have focused on the former. For example, the TPB does not help us distinguish between intenders who do and do not take action. Thus there is a need to better theorize the processes which determine which intentions are translated into action. As Bagozzi (1993) argues, the variables outlined in the main social cognition models are necessary, but not sufficient, determinants of behaviour. In other words, they can provide good predictions of people’s intentions (or motivation) to perform a health behaviour, but not always their actual behaviour. This area of research has been referred to as the ‘intention–behaviour gap’. Deciding between social cognition models Although a great deal of research has been devoted to testing individual social cognition models, little research has compared the relative predictive power of different SCMs. For example, Reid and Christensen (1988) found that while the HBM explained 10 per cent of the variance in adherence among women taking tablets for urinary tract infections to a tablet regimen, the variance explained increased to 29 per cent when cognitions specified by the theory of reasoned action were added. Another approach to the variety of SCMs is to integrate them. This may be valuable, especially since many include similar cognitions. For example, commentators agree that the key cognitions prominently include intention, self-efficacy and outcome expectancies (or attitudes). An important attempt to integrate these models was made by Bandura (SCT), Becker (HBM), Fishbein (TRA), Kaufen (self-regulation) and Focus 7.3 Social cognition models | 149 Triandis (theory of interpersonal behaviour) as part of a workshop organized by the US National Institute of Mental Health in response to the need to promote HIV-preventive behaviours. The workshop sought to ‘identify a finite set of variables to be considered in any behavioral analysis’ (Fishbein et al, 2001: 3). They identified eight variables which, they argued, should account for most of the variance in any (deliberative) behaviour. These were organized into two groups. First were those variables which were viewed as necessary and sufficient determinants of behaviour. Thus, for behaviour to occur an individual must: 1) have a strong intention; 2) have the necessary skills to perform the behaviour; and 3) experience an absence of environmental constraints that could prevent behaviour. The second group of variables were seen to primarily influence intention, although it was noted that some of the variables may also have a direct effect on behaviour. Thus, a strong intention is likely to occur when an individual: 4) perceives the advantages (or benefits) of performing the behaviour to outweigh the perceived disadvantages (or costs); 5) perceives the social (normative) pressure to perform the behaviour to be greater than that not to perform the behaviour; 6) believes that the behaviour is consistent with his or her self-image; 7) anticipates the emotional reaction to performing the behaviour to be more positive than negative; and 8) has high levels of self-efficacy. Figure 7.7 illustrates this integrated model. If you were trying to identify the determinants of condom use, which cognitions would you focus on? Would this be any different if you were trying to predict smoking cessation? Self-discrepancy Intention Behaviour Environmental constraints Self-efficacy Skills Social pressure Advantages/ disadvantages Emotional reaction Figure 7.7 The ‘major theorists’ integrated social cognition model. A critical appraisal of SCMs The use of SCMs to predict health behaviour has a number of advantages and disadvantages. Below we outline the main advantages of a social cognition approach before considering a range of specific and more general criticisms that have been made of this approach. There are four clear advantages of using SCMs to predict and understand health behaviours. First, they provide a clear theoretical background to any research, guiding the selection of cognitions and providing a description of the ways in which these constructs combine in order to determine health behaviours. Second, because the models have been repeatedly tested they provide reliable and valid measures of selected cognitions (for example, see Ajzen’s website for guidance on developing TPB measures at www.people.umass.edu/aizen/ tpb.html). Third, SCMs provide us with a description of the motivational and volitional processes underlying health behaviours. As a result, they add to our understanding of the proximal determinants of health behaviour and, because of this, they, fourth, identify key targets for interventions designed to change motivation (see Chapter 8). The use of SCMs could also limit our understanding of health behaviour. For example, because SCMs provide clearly defined theoretical frameworks, their use may lead to the neglect of other cognitions. For example, moral norms are not included in the main SCMs but have been shown to be important in behaviours such as blood donation (Godin et al, 2005). Another limitation of SCMs is that while they usefully identify cognition change targets, they commonly do not specify the best means to change such cognitions. Moreover, an over-exclusive focus on SCMs may lead to the neglect of other potentially effective behaviour change interventions, such as increased taxation or legislation, which may not or may not have their effects through the cognitions specified by the SCMs (see Chapter 9). There has been one widely cited critique of SCMs written by Ogden (2003) with a response by Ajzen and Fishbein (2004) (see also Greve, 2001; Norman and Conner, 2005). Ogden’s (2003) critique is based on a review of 47 empirical studies published in four main health psychology journals over a four-year period and focuses on the HBM, PMT and TRA/TPB. Ogden raised four issues: use in developing interventions, interpretation of empirical testing, analytical versus synthetic truths and mere measurement. Ogden first concluded that SCMs were useful to researchers and ‘. . . to inform service development and the development of health-related interventions to promote health behaviors’ (Ogden, 2003: 425). However, she also made three key criticisms. First she argued that SCMs cannot be empirically tested, that is, confirmed or disconfirmed. She supported this point by pointing out that researchers do not conclude that they have disconfirmed SCMs when they find that one or more of the theory’s constructs do not predict the outcome measure or that the findings do not explain all or most of the variance in intentions or behaviour. Ajzen and Fishbein (2004) highlight that the logic of this argument is unsound. For example in the case of the TPB, numerous descriptions of the theory make clear that the extent to 150 | Health cognitions and health behaviours including conditions without baseline (time 1) questionnaires for comparative purposes. Moreover, these mere measurement effects suggest that SCMs are indeed tapping psychological processes crucial to behaviour change. The intention–behaviour gap The intention-behaviour gap refers to the fact that intentions are far from perfect predictors of behaviour. In this section we review two areas of research exploring this gap. The first focuses on the stability of intentions across time while the second examines the volitional processes that might be important in determining whether intentions get translated into action. Intention stability In the vast majority of applications of SCMs the predictors of behaviour are measured by questionnaire (at time 1) and then behaviour is measured at a second time point, thereby employing a prospective survey method. One important implication of such a design is that the measured constructs (e.g. attitudes) will remain unchanged between the measurement and the opportunity to act. So, for example, in using the TPB the assumption is that intentions to exercise will remain the same from when the (time 1) questionnaire is completed to the time points at which the respondent has the opportunity to engage in exercise. This is one of the limiting conditions of the TPB. However, cognitions including intentions may indeed change in this time period and such change provides one important explanation of the intention– behaviour gap. Several studies have now demonstrated that the intention– behaviour gap is indeed reduced for individuals with intentions that are more stable over time. For example, Conner, Norman and Bell (2002) found that intentions were stronger predictors of healthy eating over a period of six years when these intentions were stable over a six-month time period. These findings show that intention stability moderates the relationship between intention and behaviour. A number of other factors have been found to influence the size of the intention–behaviour gap. For example, anticipating feeling regret if one does not perform a behaviour or perceiving a strong moral norm (that is, believing that one is morally obliged to act) have both been found to significantly reduce the intention–behaviour gap (see Cooke and Sheeran, 2004, for a review). Like Conner et al (2002), Sheeran and Abraham (2003) found that intention stability moderated the intention–behaviour relationship for exercising but, more importantly, found that intention stability mediated the effect of other moderators of the intention–behaviour relationship, including anticipated regret. This suggests that the mechanism by which a number of these other moderators may have their effect on intention–behaviour relationships is through changing the temporal stability of intentions. Hence, factors that might be expected to make individual intentions more stable over time would be expected to increase the impact that these intentions have on behaviour and so reduce the intention– behaviour gap. The intention–behaviour gap | 153 Implementation intention formation A variety of factors which affect the enactment of intentions have been investigated including personality traits, self-efficacy and planning. For example, we noted in Chapter 6 that conscientious individuals may possess skills that help them to enact their intentions (see Chapter 8 for more on self-efficacy and Chapter 9 for more on planning). However, another factor may relate to the nature of the intention formed. Gollwitzer (1993) makes the distinction between goal intentions and implementation intentions. While the former is concerned with intentions to perform a behaviour or achieve a goal (i.e. ‘I intend to do x’), the latter is concerned with if-then plans which specify an environmental prompt or context that will determine when the action should be taken (i.e. ‘I intend to initiate the goal-directed behaviour x when situation y is encountered’). The important point about implementation intentions is that they commit the individual to a specific course of action when certain environmental conditions are met. Sheeran et al (2005: 280) note that “to form an implementation intention, the person must first identify a response that will lead to goal attainment and, second, anticipate a suitable occasion to initiate that response. For example, the person might specify the behaviour ‘go jogging for 20 minutes’ and specify a suitable opportunity ‘tomorrow morning before work’”. Gollwitzer (1993) argues that, by making implementation intentions, individuals pass control of intention enactment to the environment. The specified environmental cue prompts the action so that the person does not have to remember or decide when to act. Sheeran et al (2005) provide an in-depth review of both basic and applied research with implementation intentions. For example, Milne, Orbell and Sheeran (2002) found that an intervention using persuasive text based on protection motivation theory prompted positive pro-exercise cognition change but did not increase exercise. However, when this intervention was combined with encouragement to form implementation intentions, behaviour change was observed (see Gollwitzer and Sheeran, 2006, for a meta-analysis of such studies). Thus implementation intention formation moderates the intention– behaviour relationship demonstrating that two people with equally strong goal intentions may differ in their volitional readiness depending on whether they have taken the additional step of forming an implementation intention. Implementation intention formation has been shown to increase the performance of a range of behaviours with, on average, a medium effect size. Implementation intentions appear to be particularly effective in overcoming a common problem in enacting intentions, that is, forgetting. Provided effective cues are identified in the implementation intention (i.e. ones that will be commonly encountered and are sufficiently distinctive), forgetting appears to be much less likely. 154 | Health cognitions and health behaviours Sample essay titles | 155 Summary There is considerable variation in who performs health behaviours. Demographic differences explain part of this variation, although such factors are not easily modifiable. Various modifiable cognitions have been identified which explain differences in who performs health behaviours. Key cognitions include intentions, self-efficacy and outcome expectancies (or attitudes). Cognitions have been incorporated in a number of social cognition models (SCMs) that describe the key cognitions and how they are interrelated in the determination of behaviour. The most important SCMs include the health belief model, protection motivation theory, theory of reasoned action/theory of planned behaviour and social cognitive theory. These models focus on the cognitive antecedents of motivation. Stage models attempt to describe the process of behaviour change from first consideration to maintenance of change but there is limited evidence to suggest that people remain in stable stages of action readiness over time. While SCMs have a number of advantages, criticisms of SCMs suggest the need for further sophistication in the testing of such models. SCMs are limited in their capacity to explain why some intentions are translated into behaviour while others are not. Various factors explaining this intention– behaviour gap have been explored and the important role of the temporal stability of intentions has been identified. Research has also investigated volitional processes which facilitate the enactment of intentions. Implementation intentions, that is, if-then plans situating an intended action in a specific context, have been shown to reduce the intention–behaviour gap. Critique of SCMs Health behaviours Health belief models Implementation intentions Intention–behaviour gap Mere measurement Protection motivation theory Self-regulation Social cognition models (SCMs) Social cognitive theory Stage models Theory of planned behaviour Transtheoretical model of change Key concepts and terms Sample essay titles ■ Critically evaluate the use of social cognition models in understanding health behaviours. ■ Compare and contrast the health belief model and the theory of planned behaviour as explanations of why people do and do not perform a range of health behaviours. ■ What do we know about the antecedents of intention? Discuss with reference to available empirical evidence.