Sun protective behaviour, optimism bias, and the transtheoretical model of behaviour change

Abstract

This study investigated the relationship between sun protective behaviours and three psychological variables influencing health behaviour: decisional balance, optimism bias, and the transtheoretical model of behaviour change. Two hundred participants completed the ‘Readiness to Alter Sun Protective Behaviour’ questionnaire, and a short questionnaire investigating optimism bias, decisional balance, attitudes, and experiences of sun protection and skin cancer. Participants were evenly distributed between the precontemplation, contemplation, and action stages. Participants in the action stage were significantly more likely to endorse the perceived advantages associated with sun protective behaviour than participants in the precontemplation and contemplation stages. They also reported sunbathing significantly less—and being more concerned about contracting skin cancer—than participants in the earlier stages. Decisional balance and optimism bias scores varied systematically across the stages of change; however, decisional balance was the only significant psychological predictor of sun protective behaviours. Optimism bias was greatest in the precontemplation stage, whereas the action stage was characterised by more positive attitudes to sun protection. This suggests that knowledge of the real risks of skin cancer might be a precursor to behaviour change, but only a change in attitudes results in a move to the action stage and measurable behaviour change.

Key words:

• decisional balance
• optimism bias
• RASP‐B
• sun protection
• transtheoretical model

The dangers associated with regular or prolonged exposure to the sun are well documented, and include premature aging, sunburn, and most seriously, skin cancer (Arthey & Clarke, 1995; Kripke & Ananthaswamy, 2003; Rossi, Blais, Redding, & Weinstock, 1995). Australia has the highest incidence of skin cancer of any country in the world, and the state of Queensland has the highest incidence within Australia (Australian Institute of Health and Welfare and Australasian Association of Cancer Registries, 2004). With a predicted lifetime prevalence of 67% (Cancer Council Australia, 2009), more than 2,000 new cases of skin cancer are diagnosed and over 200 deaths are attributed to the disease in Queensland annually (Buettner & Raasch, 1998; SunSmart, 2003). It is the behavioural link between sun exposure and skin cancer that provides the most promising avenue for decreasing the incidence of skin cancer. In recent years, several models attempting to explain the underlying process of health‐mitigating behaviour change have been proposed. Few studies, however, have examined the relationship between sun protective behaviours and specific psychological variables, such as optimism or the perceived advantages and disadvantages of sun exposure.

The transtheoretical model of behaviour change (hereafter TTM) has provided a useful explanation of the pattern of acquisition and cessation of a variety of potentially health‐compromising behaviours, including smoking (Prochaska & DiClemente, 1983), excessive alcohol consumption (Rollnick, Heather, Gold, & Hall, 1992), drug misuse (Sutton, 1996), and sun exposure (Kristjansson, Helgason, Rosdahl, Holm, & Ullen, 2001). Prochaska and DiClemente posited that individuals in the process of reducing or eliminating health‐compromising behaviours progress through six specific stages. Individuals in the first stage, precontemplation, are characterised as either having no intention to change the target behaviour or denying a need to change. Those in the second stage, contemplation, are considering change but have not yet made a commitment. In the third stage, preparation, a decision has been made to change the target behaviour within a defined period of time (usually the next 30 days). Individuals in the fourth stage, action, have changed their behaviour, but it remains unclear whether the changes will be long lasting. The fifth stage, maintenance, is characterised as relapse prevention. The final stage is termination, and individuals in this stage display total mastery over the target behaviour, experiencing no temptation to revert to old habits (Kristjansson, Branstrom, Ullen, & Helgason, 2003). While many individuals progress sequentially through the stages, relapse (particularly during the action stage) is a common hurdle during such progression.

In addition to the stages of change described earlier, the TTM also contains several processes of change and psychological constructs that help explain an individual's progression through the stages (Prochaska & Velicier, 1997). Decisional balance, one of the integral constructs of the TTM, has been consistently found to correlate with the proposed stages of change across a variety of health‐related behaviours, including sun protective behaviour (Herrick, Stone, & Mettler, 1997; Kristjansson et al., 2001; Littell & Girvin, 2002).

Decisional balance

Decisional balance involves an individual ‘weighing up’ the perceived advantages and disadvantages associated with changing their current behaviour against the perceived advantages and disadvantages associated with choosing not to change (Herrick et al., 1997). Research investigating decisional balance and the stages of change has shown that individuals are typically able to identify a greater number of perceived advantages, and fewer perceived disadvantages, associated with changing their problematic behaviour as they progress through the stages, which may be indicative of the cognitive adjustments that occur during the process of behaviour change (DiClemente et al., 1991; Prochaska, 1994). Decisional balance has been found to vary with the stages of change in relation to several health‐related behaviours, including smoking (Herrick et al., 1997; Prochaska, 2000), physical exercise (Marshall & Biddle, 2001; Wakui et al., 2002), and oral health self‐care behaviour (Tiliss et al., 2003). At least one study has investigated the relationship between decisional balance and the stages of change in relation to sun protective behaviour. Kristjansson and colleagues (Kristjansson et al., 2001) interviewed a Swedish sample and reported that individuals in the precontemplation stage perceived more advantages to sun exposure than disadvantages, while the opposite was true for individuals in all other stages. These findings indicate that decisional balance may vary in systematic ways across the stages; however, the degree to which this finding is limited to the Swedish population is unknown. Individuals from Queensland may possess different attitudes and beliefs towards sun protection than members of the Swedish population, as the potential consequences of sun exposure in Queensland are considerably greater.

Numerous personal and situational factors can influence an individual's decision‐making process when engaging in health‐compromising behaviours, such as sun exposure. Accordingly, such factors also play a part in the reduction or elimination of such behaviours. For example, one's previous experience with skin cancer (including the removal of cancerous or non‐cancerous skin lesions) may alter one's decisional balance, and hence subsequent sun protective behaviours. Similarly, there is a strong association between age and sun protective behaviours, with younger people often engaging in more frequent intentional sun exposure than older people. In order to design effective programs to encourage individuals to implement preventive measures and reduce their levels of health‐compromising behaviours, a detailed understanding of their decision‐making process is required. The perceived cost‐benefit ratio of the behaviour, in addition to psychological constructs such as optimism, is likely to play a role in this process.

Optimism bias

Optimism is the tendency to expect the best possible outcome and focus on the most hopeful aspects of a given situation (Scheier & Carver, 1985). Optimism bias is the tendency for individuals to believe, often erroneously, that they are less likely than the average person to experience negative life events and more likely than the average person to experience positive life events (Weinstein, 1980). Thus, it refers to a favourable perception of risk relative to others' risk, rather than their actual objective risk (Williams & Clarke, 1997). This bias about one's perceived invulnerability could potentially hinder the adoption and maintenance of preventive behaviours because if one believes that certain health risks typically apply more to others than to oneself, there may be little motivation to take preventive action (van der Pligt, 1996). To date, no study has investigated the extent to which optimism bias correlates with the stages of change in the TTM. Such an investigation may contribute to a better understanding of the cognitive processes involved as people progress through the stages of the TTM.

Aims and hypotheses

The first aim of the present study was to investigate whether decisional balance and optimism bias scores varied across the stages of change in relation to sun protective behaviour. The second aim was to explore the relationship between these psychological constructs and sun protective behaviours. Based on previous decisional balance research (Kristjansson et al., 2001; Tiliss et al., 2003), it was hypothesised that individuals in the action stage would endorse more strongly the advantages associated with sun protection than participants in the earlier stages. It was also hypothesised that individuals in the precontemplation stage would report a stronger optimism bias than participants in the contemplation and action stages, as they would consider their chances of contracting skin cancer to be lower than participants in the later stages. Finally, it was hypothesised that participants with higher optimism bias would engage in significantly less sun protective behaviour than participants with lower optimism bias.

METHOD

Sample

The sample of 200 participants (65% female) was recruited from the Far North Queensland community. University undergraduates made up 55% of the sample (N = 110), and the remainder were members of the wider community. Participants ranged in age from 17 to 86 years (M = 42.0, standard deviation (SD) = 20.1). Student participants were recruited after viewing a brief presentation about the study in tutorial classes. Participants from the community were recruited from a variety of organisations and institutions, including social clubs, four independent retirement villages, and the University of the Third Age. Of the 275 questionnaires distributed, 200 were completed and returned, representing a response rate of 73%. All participants read an information sheet outlining the details of the study before signing an informed consent form and completing the questionnaires in private. Student participants received nominal course credit for participating, but no incentive was offered to community members.

Materials

The questionnaire administered in this study consisted of three parts. The first asked a series of simple demographic questions (age, gender, occupation, nationality, education level, how long they have lived in Far North Queensland), and if they had had any lesions removed. The second set of questions comprised the ‘Readiness to Alter Sun protective Behaviour’ (RASP‐B) questionnaire, a 12‐item instrument to allocate respondents to either the precontemplation, contemplation, or action stage of the TTM in relation to sun protective behaviour (Borschmann & Cottrell, 2009). Reported Cronbach's alpha for the RASP‐B ranged between 0.67 and 0.72 for allocation to stages of change. Finally, 19 items designed by the authors measuring attitudes and beliefs towards sun protection and skin cancer, current patterns of sun protective behaviour, and measures of decisional balance and optimism bias were also administered. Attitude questions included ‘How concerned are you about developing skin cancer?’ Belief questions included ‘How preventable do you think skin cancer is?’ Behaviour questions included ‘How often do you sunbathe?’

Decisional balance

Decisional balance scores were obtained on the basis of participants' responses to six items, each of which they were asked to endorse on a 5‐point scale ranging from ‘0 = Strongly Disagree’ to ‘4 = Strongly Agree’. Three of the items probed the advantages associated with sun protective behaviour and positive general health (e.g., ‘Minimising my level of sun exposure protects my overall health’), while the remaining three items probed the advantages associated with deliberate sun exposure (e.g., ‘It's worth getting sun‐burnt if it turns into a tan the following day’). The decisional balance score was the total of each participant's responses to the six decisional balance items, three of which were reverse‐scored. Decisional balance scores, therefore, ranged from 0 to 24, with higher scores reflecting a more positive attitude towards sun protective behaviours and a more negative attitude towards deliberate sun exposure. The obtained Cronbach's alpha for the decisional balance scale was a modest 0.6.

Optimism bias

Optimism bias scores were obtained on the basis of participant responses to three questions: (1) ‘Compared to other people of your age, what do you think your chances of contracting skin cancer in the future are?’; (2) ‘Compared to other people of your age and gender, what do you think your chances of contracting skin cancer in the future are?’; and (3) ‘Compared to other people in general, what do you think your chances of contracting skin cancer in the future are?’ Participants were required to endorse each item on a 5‐point scale ranging from ‘0 = Much less than average’ to ‘4 = Much greater than average’. The optimism bias score was the total of each participant's responses to these three items (all of which were reverse‐scored). Optimism bias scores, therefore, ranged from 0 to 12, with higher scores indicating an optimistic tendency to underestimate the chances of contracting skin cancer in the future. The obtained Cronbach's alpha for the optimism bias scale was a high 0.9.

Procedure

The study was approved by the James Cook University Human Ethics Committee (ref: H1590). All participants read an information sheet outlining the details of the study before signing an informed consent form and completing the questionnaires in private.

RESULTS

Frequency data were analysed using a series of chi‐square analyses, while analyses of variance and corresponding post hoc procedures were used to examine differences in sun protective behaviours and attitudes between groups based on stage of change. Logistic multiple regression analyses were used to examine the usefulness of psychological variables as predictors of sun protective behaviour. As there is a strong relationship among sun exposure behaviours, optimism bias, and age, separate regression analyses were conducted for participants younger than 50 years of age and participants older than 50 years of age. Previous research has shown that younger people are more likely to engage in voluntary sun exposure behaviours than older people, possibly due to perceived social and aesthetic benefits (e.g., Hillhouse & Turrisi, 2002). Thus, there is good reason to regard these groups as qualitatively different populations.

Stage of change

Each participant was allocated to the precontemplation, contemplation, or action stage of the TTM on the basis of their responses to the RASP‐B. Participants were relatively evenly distributed between the three stages (see Table-1). Some, but not all, of the behavioural measures of sun protection reflected the stages of change. There were significant differences in reported frequency of sunbathing χ²_{(6,N = 200)} = 15.3, p = .018 across participants in the different stages. Participants in the action stage typically reported sunbathing less than participants in the earlier stages. However, the frequency of application of sunscreen did not differ significantly between participants in the different stages, χ²_{(8,N = 200)} = 13.5, p = .097, nor did the frequency of medical skin check‐ups, χ²_{(8,N = 200)} = 9.9, p = .27. The level of concern over contracting skin cancer also varied systematically across the stages of change, with participants in the action stage being more concerned about contracting skin cancer than those in the earlier stages, χ²_{(6,N = 200)} = 27.67, p < .001.

Table 1 Reported (and expected) frequency of sunbathing, sunscreen use, and concern over contracting skin cancer across the stages of change

	Precontemplation	Contemplation	Action	Total
	n (Exp)	n (Exp)	n (Exp)
How often do you sunbathe?
Never	39(42.8)	38(39.7)	49(43.5)	126
Yearly	13(9.2)	6(8.5)	8(9.3)	27
Monthly	12(10.5)	8(9.8)	11(10.7)	31
Weekly	4(4.5)	11(5)	1(5.5)	16
Total	68	63	69	200
How often do you apply sunscreen when exposed to the sun?
Never	11(6.8)	3(6.3)	6(6.9)	20
Rarely	13(10.9)	10(10.1)	9(11)	32
Sometimes	17(20.1)	26(18.6)	16(20.4)	59
Often	15(19.4)	16(18)	26(19.7)	57
Always	12(10.9)	8(10.1)	12(11)	32
Total	68	63	69	200
How concerned are you about contracting skin cancer?
Not at all concerned	21(10.9)	10(10.1)	1(11)	32
Slightly concerned	26(24.1)	21(22.4)	24(24.5)	71
Moderately concerned	19(27.5)	27(25.5)	35(27.9)	81
Very concerned	2(5.4)	5(5)	9(5.5)	16
Total	68	63	69	200

Decisional balance and optimism bias

It appears that the decisional balance and optimism bias scales are measuring different constructs, as the correlation between these scales was small and not significant (r = −0.13, p = .06, N = 200). Decisional balance scores differed significantly between the three stages, F_(2, 197) = 6.4, mean squared error (MSE) = 12.4, p = .002, η²_p. = 0.06. Post hoc comparisons using the Dunnett procedure to control the familywise error rate indicated that participants in the action stage had a significantly more positive attitude towards sun protective behaviours (M = 17.1, SD = 3.5) than those in either the precontemplation (p = .003, M = 15.2, SD = 3.9) or contemplation (p = .007, M = 15.3, SD = 3.1) stages.

Optimism bias scores also varied over the three stages (F_(2, 197) = 11.4, MSE = 6.7, p < .001, η²_p. = 0.10) (see Fig. 1). Post hoc comparisons indicated that the participants in the action stage (M = 5.7, SD = 2.4) were significantly less optimistic (p < .001) than those in the precontemplation stage (M = 7.5, SD = 2.6) but did not differ significantly (p = .98) from those in the contemplation stage (M = 5.6, SD = 2.8). Thus, as hypothesised, the earlier stages of change were associated with more positive attitudes towards sun exposure and more optimism about the chance of avoiding skin cancer. Fig. 2 shows that participants' perception of the benefits of sun avoidance increased from the precontemplation and contemplation stages to the action stage. The relationship between optimism bias and stages of change is slightly different; participants in the precontemplation stage were more optimistic about their chances of avoiding skin cancer than those in either the contemplation or action stage.

Figure 1 Mean Optimism Bias (SEM) across the three stages of change.

Figure 2 Mean Decisional Balance (SEM) across the three stages of change.

Predictors of sun protective behaviours

The previous analyses indicated that stages of change are associated with sun‐related behaviours, such as sunbathing and applying sunscreen. The following two‐step logistic regression analysis explored the utility of the three psychological constructs (decisional balance, optimism bias, and stage of change) in predicting sunbathing and sunscreen use. In the first step, the three control variables of age, gender, and experience with skin lesions were entered; at the second step, the three psychological variables were entered into the equation. Age and gender were chosen as control variables because it has previously been demonstrated (Montague, Borland, & Sinclair, 2001; Woolley, Buetnner, & Lowe, 2004) that females engage in more sunbathing behaviour than males and that younger people engage in more risky sun‐related behaviours than older people. Preliminary analysis of the potential relationship between personal experience with skin lesions and optimism bias indicated that ‘personal experience’ with skin cancer was associated with less optimism bias, while ‘no experience’ was associated with a more optimistic view. As experience with skin lesions is related to optimism bias, it is reasonable to assume that it might also be associated with subsequent sun protective behaviours. Thus, experience with lesions was also used as a control variable in the first step of the sequential models. The psychological variables were entered on the second step of the models to ascertain their predictive ability after gender and experience with skin lesions were controlled for. Table-2 summarises the logistic regression models predicting sunbathing behaviour and sunscreen use for under 50‐year‐olds only, as no variable was a significant predictor of sunbathing or sunscreen use for participants over 50 years of age.

Table 2 Summary of sequential logistic regression analyses for variables predicting sun protective behaviours for participants under 50 years of age (N = 132)

Variable	B	SE B	p	R^2	Modelχ^2
Sunbathing step 1				0.020	1.971
Gender	−0.395	0.450	0.380
Experience of Lesions	−0.287	0.330	0.384
Sunbathing step 2				0.278	8.360
Gender	−0.933	0.526	0.076
Experience of lesions	−0.316	0.391	0.419
Decisional balance	−0.290	0.064	0.000
Stage of change	0.097	0.279	0.728
Optimism bias	0.137	0.089	0.122
Sunscreen application step 1				0.157	3.957
Gender	1.820	0.492	0.000
Experience of lesions	−0.365	0.328	0.266
Sunscreen application step 2				0.299	12.412
Gender	1.576	0.525	0.003
Experience of lesions	−0.210	0.364	0.565
Decisional balance	−0.185	0.057	0.001
Stage of change	−0.495	0.267	0.064
Optimism bias	0.045	0.084	0.589

Note. All χ² statistics are non‐significant, indicating a reasonable model fit.

Sunbathing behaviour was not predicted by the control variables of gender and experience with skin lesions, with the first step model only explaining 2% of the variance. In the second step, the psychological variables significantly improved the final model, explaining 27.8% of the variance in predicting sunbathing behaviour. Decisional balance was the only significant psychological predictor. Participants with more positive attitudes to sun protective behaviours and negative attitudes to deliberate sun exposure were more likely to report rare or no sunbathing. The model was 86% correct in predicting rare or no sunbathing, and 48% correct in predicting regular sunbathing.

The analysis of predictors of sunscreen use (see Table-2) indicated that gender, a control variable, was a significant predictor in both steps. The control variables contributed 15.7% of the variance in sunscreen use, with gender as the only significant predictor. Females were more likely to use sunscreen regularly than males were. The first step of the model was 89.4% correct in predicting regular application of sunscreen and 39.4% correct in predicting rare or sometimes applying sunscreen. In the full model, both gender and decisional balance were significant predictors of sunscreen use. Similar to the results for sunbathing, people with more positive attitudes to sun protective behaviours were more likely to regularly apply sunscreen. The full model accounts for 29.9% of the variance in predicting sunscreen use. The model was less useful than in step one for correctly predicting regular sunscreen use (75.8% correct), but better predicted those who rarely or sometimes applied sunscreen (71.2% correct).

As mentioned previously, the same analyses were conducted for the 68 participants who were over 50 years of age. The sunbathing analysis was not viable as all but one participant reported rarely or never sunbathing. Sunscreen use among participants in this age group was more varied with 23 participants regularly applying it and 45 participants rarely or sometimes applying it. The control step of the model (gender and experience of lesions) only explained 1.2% of the variance in sunscreen use, and the full model including the psychological variables only increased the variance explained to 7.5%. No psychological variable was a significant predictor of sunscreen use.

DISCUSSION

The first aim of the present study was to investigate whether decisional balance and optimism bias varied across the stages of change in relation to sun protective behaviour. The second aim was to explore the relationship between these psychological constructs and sun protective behaviours.

Decisional balance

Participants in the action stage were significantly more likely to perceive the advantages of sun protective behaviour than participants in both the precontemplation and contemplation stages, as indicated by their decisional balance scores. This finding supports the findings of previous studies that have shown that, for individuals in the action stage, the perceived advantages of change typically outweigh the perceived disadvantages of change (DiClemente et al., 1991; Herrick et al., 1997). It was also noted that decisional balance was the only significant psychological predictor of sun‐related behaviours. The logistic regression models indicated high levels of correctly identifying participants who regularly applied sunscreen and rarely or never sunbathed. This supports previous research that has shown that individuals with higher decisional balance scores typically engage in more health‐protective behaviour than those with lower decisional balance scores (DiClemente et al., 1991; Prochaska, 1994).

Optimism bias

Participants in the precontemplation stage were significantly more optimistic about their likelihood of not contracting skin cancer in the future than those in the action stage. There was no significant difference in optimism bias scores between participants in the action and contemplation stages. An optimism bias score of 6 would indicate perceived chances exactly equal to those of the average person of contracting skin cancer, while scores greater than 6 indicate the presence of an optimism bias. Therefore, participants in the precontemplation stage demonstrated a significant optimism bias: 95% confidence interval (6.9, 8.1), while those in the contemplation (4.9, 6.6) and action (5.1, 6.3) stages did not. This finding may be associated with an increased risk of contracting skin cancer, as the presence of such an optimism bias may hinder the adoption of appropriate preventive behaviours in future.

This represents the first time that optimism bias has been shown to vary systematically across the stages of change in relation to sun protective behaviour. After considering the decisional balance and optimism bias scores together, it becomes evident that the transition from the precontemplation stage to the contemplation stage is marked by a decrease in optimism bias to more realistic levels of beliefs about the likelihood of contracting skin cancer, while the transition from the contemplation stage to the action stage is marked by an increase in decisional balance. It is possible that an individual's belief that he or she is unlikely to contract skin cancer leads to a range of behaviours and attitudes incompatible with sun protective behaviours. Once the possibility of a susceptibility to the disease is accepted, a cost‐benefit analysis determines whether preventive action is taken.

Such an interpretation presupposes that the likelihood of individuals in the precontemplation stage engaging in sun protective behaviour is low, perhaps due in part to their misguided belief that they are unlikely to contract skin cancer in the future. Participants in the action stage reported sunbathing significantly less frequently than participants in each of the two earlier stages, adding weight to previous research indicating that individuals in the action stage have already implemented health‐protective lifestyle changes (Prochaska, 2000; Prochaska & DiClemente, 1983).

Despite the clear differences in optimism bias across the stages of change, it was surprising that this psychological variable was not a significant predictor of sun protective behaviours. Of the psychological variables, decisional balance was the only significant predictor of both sunscreen use and sunbathing frequency in under 50‐year‐olds. In this group, people with more positive attitudes to sun protective behaviours were more likely to engage in these behaviours. Of the non‐psychological variables, females were more likely to use sunscreen than males. For over 50‐year‐olds, sun protective behaviour does not seem to be predicted by psychological variables or experience with lesions. However, it is worth noting that while sunscreen use was somewhat variable, the majority of over 50‐year‐olds reported never sunbathing, suggesting that sun exposure was incidental rather than intentional.

Optimism bias appears to be a salient predictor of stage of change. What leads to optimism bias in the face of so much health advertising with regard to the risks of skin cancer was not examined in this study. It does appear that personal experience of the most direct kind has a significant influence on optimism bias; however, this is something that is not amenable to modification. The salience of previous experience with skin cancer as a moderator of behaviour can be seen in the finding that those participants who reported having previously had a skin lesion removed (either cancerous or non‐cancerous) reported sunbathing significantly less frequently than those who had never had a skin lesion removed. They also reported having medical skin examinations significantly more frequently.

Although the failure for optimism bias to predict sun protection behaviours is unexpected, the relationship between optimism bias and stage of change suggests a possible explanation. The significant difference in optimism bias between the precontemplation and contemplation stage suggests that optimism bias plays a role in maintaining risky sun exposure behaviours. A move to the later contemplation stage is associated with a decrease to optimism bias, but there is no associated change in behaviour in the move from precontemplation to contemplation. Behaviour change occurs as people progress into the action stage, which is not associated with any change in optimism bias (see Fig. 1); however, decisional balance does differ between the contemplation and action stages (see Fig. 2). Thus, decisional balance is associated with behavioural change and optimism bias is not if one assumes that changes in health‐related behaviour occur as predicted by the TTM.

Concern of contracting skin cancer

Participants in the action stage were significantly more concerned about contracting skin cancer than participants in the contemplation stage, who, in turn, were significantly more concerned than participants in the precontemplation stage. These results suggest one of two things: (1) as individuals progress through the stages of change, their level of concern in relation to contracting skin cancer increases; or (2) individuals in the latter stages are more aware of the probability of contracting skin cancer (which in Queensland corresponds to 67% of the population at some time in their lives). It would seem that the latter is more likely, and thus that greater awareness is, in part, driving progression through the stages of change. If this is the case, then it suggests that public educational campaigns are of some value in changing sun protective behaviours

Our findings need to be considered in view of some limitations. First, a majority of the participants (65%) were females. The differences in sun protective behaviours and attitudes of males and females are well documented (Montague et al., 2001; Woolley et al., 2004) and, as males and females differed significantly in their scores on key constructs such as sunscreen application frequency, it is possible that the gender imbalance may have influenced the results obtained. It may be beneficial to duplicate the study using a sample with equal numbers of males and females. Second, data were obtained through the use of self‐report instruments; an inherent disadvantage associated with all self‐report instruments is that they are dependant entirely upon respondents' candour, awareness, and comprehension of items, and may be susceptible to (intentional or otherwise) reporting bias. Finally, some of the individual items used in the present study had not been validated prior to use, and as such, their reliability and validity remain unclear.

CONCLUSIONS

Results indicated that the TTM could be applied successfully to describe and predict sun protective behaviours. Participants in the precontemplation stage reported sunbathing significantly more frequently and being significantly less concerned about contracting skin cancer in the future than participants in both the contemplation and action stages. In addition to this, they inaccurately believed their chances of contracting skin cancer in the future were lower than average, unlike participants in the latter stages. The current findings also indicate that the sample from Far North Queensland were, at the time of assessment, typically not engaging in adequate levels of sun protective behaviours, despite living in the region with the single highest incidence of skin cancer anywhere on the planet. Additionally, many participants underestimated their own statistical likelihood of contracting skin cancer, indicative of a generally optimistic view of their chances of escaping the disease. Overall, the best predictor of sun protective behaviours was decisional balance. Positive views of tanning were associated with more dangerous sun exposure behaviours and attitudes. Interestingly, one of the differences between the contemplation and action stages was decisional balance. It is possible that it is a change in these attitudes that leads to moving from thinking about changing one's behaviour to actually putting change into action. In contrast, optimism bias differed between the precontemplation and contemplation stages. Thus, we suggest that optimism bias is a characteristic of the early stages of change, and a move to a less optimistic view of one's health is associated with a change in stage but no observable change in behaviour.