Health Education Research, Vol. 18, No. 3, 267-277,
June 2003
© 2003 Oxford University Press
Predicting adolescent pedestrians’ road-crossing intentions: an application and extension of the Theory of Planned Behaviour
Department of Psychology, University of Wales, Swansea SA2 8PP and 1 Department of Psychology, University of Sheffield, Sheffield S10 2TP, UK. E-mail: p.norman{at}sheffield.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
The present paper reports an application of the Theory of Planned Behaviour (TPB) to the prediction of road-crossing intentions among adolescents. In addition, the sufficiency of the TPB was assessed by examining the additional predictive utility of moral norms, anticipated affect and self-identity. A sample of 1833 adolescents completed a questionnaire containing a scenario depicting a potentially hazardous road-crossing behaviour, followed by items measuring the TPB constructs, moral norms, anticipated affect and self-identity. Regression analyses revealed that the TPB was able to explain 25% of the variance in road-crossing intentions, over and above the influence of age and gender, with perceived behavioural control emerging as the strongest predictor. The additional variables were found to increase the predictive utility of the TPB. The results have a number of theoretical and practical implications. In particular, interventions should focus on perceptions of control in order to encourage safer road-crossing behaviour among adolescents.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
Pedestrian accidents represent the single, biggest accidental killer of children and adolescents in Britain. (Avery and Jackson, 1993)
In 1997 alone, over 18 000 children were involved in pedestrian accidents in the UK (Department of Transport, 1998). Moreover, 10–15 year olds have the highest pedestrian injury rates of any age group (Ward et al., 1994) and one child in 15 is injured in a road traffic accident before his or her 16th birthday (Jones, 1990). Against this backdrop, research has attempted to identify some of the factors that are associated with child pedestrian injuries and a review by Tight has highlighted a number of commonalities in child pedestrian injuries (Tight, 1996). First, child pedestrian injuries typically peak between 8.00 a.m. and 9.00 a.m. in the morning, and again between 3.00 p.m. and 6.00 p.m. in the afternoon. These times coincide with the journey to and from school. In fact, Grayson reports that during school term time 50% of child pedestrian injuries occur between the hours of 4.00 p.m. and 5.00 p.m. (Grayson, 1975). Second, the majority of child pedestrian injuries occur on urban roads (Southwell et al., 1990) and within a quarter of a mile of the child’s home (Grayson, 1975). Third, most pedestrian injuries occur when the child is alone (72%) or with other children (25%) (Tight, 1996). Fourth, boys are twice as likely than girls to be injured as a pedestrian (Pless et al., 1989), which may reflect the greater freedom afforded to boys and their greater tendency to play in the street. Fifth, child pedestrian injuries peak around the age of 12 years; a time that coincides with the transition into secondary school (Lynam and Harland, 1992). Sixth, the majority of pedestrian injuries occur when the child is intent on crossing the road (Southwell et al., 1990).
A number of researchers have reported that many children do not have the requisite perceptual or cognitive skills to cross the road safely (Avery and Jackson, 1993; Demetre and Gaffin, 1994). For example, in a study by Whitebread and Neilson, children were presented with a series of tasks to test their pedestrian skills (Whitebread and Neilson, 1996). Children under the age of 8 years were found to have little or no understanding of what constituted a safe place to cross the road or whether it was safe to cross the road in sight of oncoming traffic. However, Whitebread and Neilson noted that the majority of older children (i.e. 8–11 year olds) in their study performed well on the tasks and therefore had acquired the skills necessary to cross the road safely (Whitebread and Neilson, 1996). This implies that among children in the 10–15 year age range (i.e. adolescents), it is not the lack of perceptual or cognitive skills, but rather the failure to deploy these skills that contributes to the increased vulnerability of adolescent pedestrians. Therefore, in order to help reduce the level of pedestrian injuries among this age group a greater understanding of their road-crossing behaviour and decisions is required. As highlighted above, accident research has shown that adolescent pedestrian injuries occur, in the main, on journeys to and from school, when the adolescent is crossing the road away from a facility. Together these findings suggest that it would be useful to identify the motivational determinants which lead to the ‘risky’ decisions that appear to increase young people’s vulnerability to injury. Such information is likely to add to our understanding of the road traffic injuries among adolescent pedestrians.
One model of decision making that may be usefully employed in this context is the Theory of Planned Behaviour (TPB) (Ajzen, 1988, 1991). According to the TPB, the proximal determinant of behaviour is the individual’s intention to perform the behaviour. Intention is in turn determined by three constructs. (1) The individual’s overall positive or negative evaluation of the behaviour (i.e. attitude), which is determined by behavioural beliefs focusing on the perceived likelihood that the behaviour will lead to salient outcomes weighted by an evaluation of the outcomes. (2) The perceived social pressure from important others to perform or not perform the behaviour (i.e. subjective norm). Subjective norms are determined by the perceived social pressure from salient referents to perform the behaviour weighted by the individual’s motivation to comply with the referents. (3) The individual’s perception of the ease or difficulty of performing the behaviour (i.e. perceived behavioural control). Perceptions of control are based on a consideration of the perceived power of salient control factors to inhibit or facilitate performance of the behaviour and their perceived frequency of occurrence. According to the TPB, adolescents who decide to cross the road in a relatively risky manner are more likely to evaluate such behaviour in a positive light, believe that important others would approve of the behaviour and perceive the behaviour to be easy to perform.
The TPB has been successfully applied to a wide range of health-related behaviours among adults [see (Conner and Armitage, 1998; Sutton, 1998; Armitage and Conner, 2001)] and adolescents [e.g. (Rise and Wilhelmsen, 1998; O’Callaghan et al., 1999; Berg et al., 2000; Hagger et al., 2001)]. These applications have included a number of road safety-related behaviours among car drivers, cyclists and pedestrians. Considering car drivers, Parker et al. presented drivers with scenarios depicting four driving violations (i.e. speeding, drink-driving, close following, risky overtaking) followed by items measuring the main constructs of the TPB (Parker et al., 1992). The TPB was able to explain between 23 and 47% of the variance in intentions to commit the driving violations with the attitude, subjective norm and perceived behavioural control constructs all emerging as significant predictors. Turning to cyclists, Quine et al. found that the TPB was able to explain 34% of the variance in intentions to wear a safety helmet and 43% of the variance in reported safety helmet use at 1-month follow-up among a sample of 11- to 18-year-old schoolboys (Quine et al., 1998). Subjective norm and perceived behavioural control emerged as the most important predictors of intention, with intention and perceived behavioural control predicting safety helmet use at follow-up. Finally, Evans and Norman (Evans and Norman, 1998) used a similar methodology to that employed by Parker et al. (Parker et al., 1992) to examine the motivational determinants of adults’ road-crossing decisions. Respondents were presented with three potentially dangerous road-crossing scenarios (e.g. crossing a busy dual carriageway) followed by items measuring the TPB constructs. The TPB was able to explain between 37 and 49% of the variance in road-crossing intentions, with all three components of the TPB emerging as significant predictors in two of the three scenarios.
One of the attractions of the TPB is that it provides a relatively parsimonious model of the proximal determinants of individuals’ decisions (i.e. intentions). However, a number of researchers have proposed that further variables should be added to the model to increase its predictive utility (Conner and Armitage, 1998). In fact, Ajzen [(Ajzen, 1991), p. 199] has conceded that:
...the theory of planned behaviour is, in principle, open to the inclusion of additional predictors if it can be shown that they capture a significant proportion of the variance in intention or behaviour after the theory’s current variables have been taken into account.
In relation to adolescents’ road-crossing decisions, there are a number of variables that may be usefully added to the TPB. First, a distinction has been made between the subjective norm construct which refers to the perception of social pressure to perform or not perform a behaviour and personal or moral norms that tap the individual’s perception of the moral correctness or incorrectness of performing the behaviour (Ajzen, 1991; Manstead, 2000). Thus, moral norms focus on feelings of personal responsibility for performing or not performing a behaviour and, as such, may be an additional form of normative pressure. Conner and Armitage report that the addition of moral norms typically leads to a small, but significant, increase in the amount of variance explained in intention (Conner and Armitage, 1998). Second, a number of researchers have highlighted the importance of anticipated affective reactions (Parker et al., 1995; Van der Pligt and de Vries, 1998). For example, Richard et al. report that anticipated affect was predictive of intentions to engage in a range of behaviours (e.g. using soft drugs) over and above the influence of the TPB variables (Richard et al., 1996). Third, self-identity (i.e. the labels individuals use to describe themselves) may also predict intentions (Biddle et al., 1987; Charng et al., 1988). For example, Sparks and Shepherd found that individuals who held a strong self-identity as a ‘green consumer’ were more likely to intend to consume organic vegetables (Sparks and Shepherd, 1992). This effect was over and above the influence of the TPB variables, although Conner and Armitage report that on average self-identity only explains an additional 1% of the variance in intention (Conner and Armitage, 1998).
Few studies have examined the additional predictive utility of the above variables in relation to road safety-related behaviours. Parker et al. tested an extended version of the TPB that included moral norms and anticipated affect to examine intentions to commit three different motorway driving violations (Parker et al., 1995). Moral norms and anticipated affect explained an additional 11–15% of the variance in intentions, with both variables making significant contributions to the final regression equation. Evans and Norman’s study on adult pedestrians’ road-crossing decisions included a self-identity measure that focused on the individual’s perception of himself or herself as a ‘safe pedestrian’ (Evans and Norman, 1998). This measure made a significant additional contribution to the prediction of intention in two of the three scenarios. However, the amount of additional variance explained was modest, ranging between 1 and 3%.
The main aim of the present study was to apply the TPB to the prediction of adolescents’ road-crossing decisions. In line with previous research on road safety-related behaviour (Parker et al., 1992, 1995; Evans and Norman, 1998), respondents were presented with a scenario depicting a potentially hazardous road-crossing behaviour, followed by items measuring the TPB variables. The scenario described the target person engaging in a road-crossing behaviour corresponding to the conditions under which the majority of adolescent pedestrian injuries occur, i.e. on a journey home from school (Tight, 1996), intentionally crossing the road (Southwell et al., 1990) away from a road-crossing facility (Avery and Jackson, 1983). An additional aim was to test the sufficiency of the TPB by examining the additional predictive utility of moral norms, anticipated affect and self-identity. Finally, given that both age and gender have been related to pedestrian injuries (Pless et al., 1989; Lynam and Harland, 1992), the influence of these variables was also considered.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
Respondents and procedure
All 23 local authority comprehensive (i.e. secondary) schools in the County of West Glamorgan, Wales were invited to participate in the study, of which 16 agreed to take part in the research. Pupils from Years 7, 8 and 9 (i.e. 11–14 year olds) in these schools were recruited into the study, and a sample of 1833 schoolchildren (962 males, 871 females) (mean age = 12.09, SD = 0.86) was obtained. In each school, the first author introduced the broad aims of the research and explained the format of the questionnaire prior to the pupils completing the questionnaire in their classroom under the supervision of their class teacher. Pupils were instructed that their participation was voluntary and were reassured that their questionnaire responses would be anonymous. No pupils declined to complete the questionnaire. The first author remained in the classroom while the questionnaires were completed in order to answer any queries and to assist pupils as required. The questionnaire took approximately 20 min to complete.
Scenario
The questionnaire outlined a potentially hazardous road-crossing behaviour. The scenario was written in the second person singular to encourage respondents to imagine themselves in the scenario described. The scenario was as follows:1
You are on your way home from school. It is cold and getting dark very quickly, so you are hurrying home. About halfway from home you have to cross a busy road. There is a crossing further down the road, but that will take you a lot longer to get home because you will then have to walk back up on the other side. You step out onto the road. You cannot see any vehicles coming. Taking a chance, you run across the road.
Measures
After reading the scenario, respondents were asked to complete a series of items based on the TPB and the additional variables following a similar format to that employed in previous studies on road safety-related decisions (Parker et al., 1992, 1995; Evans and Norman, 1998). In line with Ajzen and Fishbein’s recommendations, a pilot study was conducted with 61 schoolchildren in order to generate the modal salient behavioural, normative and control beliefs for the questionnaire (Ajzen and Fishbein, 1980). All items were followed by seven-point response scales, with descriptive labels to aid comprehension.
Behavioural intention was measured using two items. Respondents were asked to indicate how likely it was that they would cross the road (1) as depicted in the scenario and (2) if they came across a similar situation in the next few weeks (e.g. ‘If you really were the person in the description, how likely is it that you would take a chance and run across the road?’) (scored -3 ‘very unlikely’ to +3 ‘very likely’). The mean of the two items was used as a measure of behavioural intention (
= 0.85). Attitude was measured using a belief-based measure. Respondents were presented with four behavioural beliefs, i.e. ‘My taking a chance and running across the road would...get me home faster, be easier, get me run over, get me killed or injured’. The strength of each belief was assessed using response scales ranging from ‘very unlikely’ (-3) to ‘very likely’ (+3). For each belief there was a corresponding outcome evaluation item (e.g. ‘Getting home faster would be...bad/good’) (scored -3 to +3). The products of these ratings were averaged to produce a belief-based measure of attitude. Subjective norm was also measured using a belief-based measure. Four referents were used: i.e. my friends, my parents, the police and the drivers/motorists. Respondents were asked to indicate the likelihood that each referent would want them to cross the road as depicted (e.g. ‘My friends would think I should take a chance and run across the road’) (scored -3 ‘very unlikely’ to +3 ‘very likely’), and whether they were motivated to comply with their wishes (e.g. ‘I generally like to cross the road in a way that my friends think I should’) (scored 1 ‘not at all’ to 7 ‘a great deal’). The products of these ratings were averaged to produce a belief-based measure of subjective norm. The pilot interviews indicated that there was a considerable overlap between the content of the control beliefs and behavioural beliefs. As a result, three items were used to provide a direct measure of perceived behavioural control: i.e. ‘If I were the person in this description, I think taking a chance and running across the road would be...difficult/easy’, ‘If I were the person in this description, I think not taking a chance and running across the road would be...difficult/easy’ and ‘It would be mainly up to me whether or not I took a chance and ran across the road’ (‘strongly agree’–‘strongly disagree’). However, subsequent analyses revealed that it was not possible to create a reliable scale with these items. As a result, only one item, asking how difficult or easy it would be to cross the road, scored -3 to +3, was used a measure of perceived behavioural control.
Anticipated affect was measured using two items focusing on anticipated feelings after having crossed the road as depicted in the scenario, i.e. ‘My taking a chance and running across the road would...make me feel big, make me feel good’ (scored -3 ‘disagree strongly’ to +3 ‘agree strongly’). The mean of the two items was used as a measure of anticipated affect ( = 0.79). The mean of two items was used as a measure of moral norms: i.e. ‘It would be quite wrong for me to take a chance and run across the road’ and ‘I shouldn’t really take a chance and run across the road’ (scored -3 ‘disagree strongly’ to +3 ‘agree strongly’) ( = 0.63). Finally, the extent to which the respondents saw themselves as safe pedestrians was measured using two items: i.e. ‘I like to think of myself as someone who always thinks carefully about how to cross the road’ and ‘I like to think of myself as a careful pedestrian’ (scored -3 ‘disagree strongly’ to +3 ‘agree strongly’). The mean of the two items was used as a measure of self-identity ( = 0.85).
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
Descriptive findings
The means, SDs and correlations between the study variables are presented in Table I
. Inspection of the means indicates that respondents held a slightly negative intention to cross the road as depicted in the scenario, held a negative attitude towards the behaviour, felt that the behaviour would attract social disapproval, perceived it to be relatively difficult to perform, did not anticipate experiencing positive emotions following the behaviour and felt that it would be wrong to perform the behaviour.
|
All of the study variables were found to correlate with intention to cross the road as depicted in the scenario, with perceived behavioural control and self-identity emerging as the strongest correlates of intention. Thus, intention to cross the road as depicted in the scenario was positively associated with attitude, subjective norm, perceived behavioural control and anticipated affect, and negatively associated with moral norms and self-identity. In addition, older and male respondents were also more likely to have strong intentions.
|
Predicting road-crossing intentions |
|---|
|
TopAbstractIntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
A hierarchical linear regression analysis was used to predict intention. The independent variables were entered in three blocks: (1) age and gender, (2) attitude, subjective norm and perceived behavioural control, and (3) anticipated affect, moral norms and self-identity. In this way it was possible to assess the predictive utility of the TPB constructs and the additional predictive utility of the additional variables after controlling for the influence of age and gender (see Table II
).
|
Age and gender were able to explain 6% of the variance in intention (F = 57.74, d.f. = 2,1813, P < 0.001), with both variables emerging as significant independent predictors. The TPB was able to explain an additional 25% of the variance in intention to cross the road (
R2 = 0.25, F = 219.95, P < 0.001), with attitude, subjective norm and perceived behavioural control all emerging as significant independent predictors along with age and gender. Together these variables were able to explain 31% of the variance in intention (F = 163.45, d.f. = 5,1810, P < 0.001). Entering the additional variables produced a further significant increment in the amount of variance explained (R2 = 0.06, F = 61.28, P < 0.001), with anticipated affect and self-identity emerging as significant independent predictors along with the three TPB variables and age. Together the variables under consideration were able to explain 37% of the variance in intention (F = 135.32, d.f. = 8,1807, P < 0.001). |
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
The present study sought to apply the TPB to the prediction of adolescents’ road-crossing decisions in response to a potentially hazardous road-crossing scenario. The TPB was able to explain 25% of the variance in road-crossing intentions, after controlling for the influence of age and gender, with all three components of the TPB emerging as significant independent predictors. Thus, those adolescents who intended to cross the road in a potentially hazardous manner were more likely to evaluate such behaviour positively, to believe that important referents would approve of the behaviour and to perceive the behaviour as being easy to perform. The present findings therefore confirm the utility of the TPB in relation to road safety research (Parker et al., 1992
, 1995; Evans and Norman, 1998; Quine et al., 1998). It is noteworthy that the perceived behavioural control component emerged as the strongest predictor of adolescents’ road-crossing intentions. Similar findings have been reported by Evans and Norman in relation to the road-crossing decisions of adults (Evans and Norman, 1998).
A number of researchers have suggested that the predictive utility of the TPB may be augmented by the inclusion of a range of additional predictors. In the present study, three such variables were considered, i.e. moral norms, anticipated affect and self-identity. The addition of these variables to the TPB led to a significant increment in the amount of variance explained in intention. However, of the three variables, only anticipated affect and self-identity made significant contributions to the regression equation. Similar findings have been reported in previous studies on road safety-related behaviours. Evans and Norman found that a measure of self-identity produced significant increments in the amounts of variance explained in adults’ road-crossing intentions (Evans and Norman, 1998), and Parker et al. reported significant increments for moral norms and anticipated affect in relation to intentions to commit motorway driving violations (Parker et al., 1995). Together these findings question the sufficiency of the TPB to predict road safety-related decisions. However, the amount of additional variance explained by these variables is typically modest and, as a result, it is necessary to weigh the additional predictive utility afforded by these variables against the relative parsimony of the TPB. Moreover, there are a number of issues that warrant further attention for each of the additional variables considered here.
Moral norms were found to be unrelated to road-crossing intentions in the present study, thus failing to support the claim that they may represent an independent form of normative pressure (Ajzen, 1991; Manstead, 2000). It may be the case that the independent influence of moral norms is reserved for those behaviours that have a clear moral or ethical dimension (Beck and Ajzen, 1991). Manstead has argued that it is theoretically possible to differentiate between moral norms and the TPB constructs (Manstead, 2000), although Conner and Armitage have reported moderate average correlations between moral norms and both the attitude (r+ = 0.49) and subjective norm (r+ = 0.34) constructs of the TPB (Conner and Armitage, 1998). In the present study, moral norms were found to correlate significantly with attitude (r = -0.20, P < 0.001) and subjective norm (r = -0.33, P < 0.001), although in a subsequent factor analysis the behavioural belief, normative belief and moral norm items were found to load onto separate factors.
Anticipated affect emerged as a significant independent predictor of road-crossing intentions in the present study, thus confirming the importance of affective reactions in the prediction of health and safety-related behaviours. A number of researchers (Manstead and Parker, 1995; Van der Pligt and de Vries, 1998) have argued that the TPB fails to tap affective influences and that measures of anticipated affect should be added to the TPB, especially when the consequences of performing or not performing a behaviour are affect laden. However, Conner and Armitage suggest that anticipated affective reactions may be covered by the behavioural beliefs component of the TPB (Conner and Armitage, 1998). In other words, anticipated affective reactions can be conceptualized as potential consequences of performing a behaviour. Nevertheless, traditional methods for eliciting behavioural beliefs typically fail to tap such affective consequences (Manstead and Parker, 1995; Van der Pligt and de Vries, 1998). In the present study, the elicitation procedure used in the pilot study to identify modal salient beliefs failed to elicit any affective consequences of performing the road-crossing behaviour depicted in the scenario, as has also been reported by Parker et al. in their study of motorway driving violations (Parker et al., 1995). Moreover, a subsequent factor analysis revealed that the behavioural belief and anticipated affect items loaded onto separate factors. It remains for future theoretical and empirical work to determine whether measures of anticipated affect should be added to the TPB, as in the present study, or whether current elicitation procedures need to be modified to encourage the generation of affective outcomes.
Self-identity also emerged as a significant independent predictor of adolescents’ road-crossing intentions in the present study, in line with Evans and Norman’s study with adult pedestrians (Evans and Norman, 1998). Thus, those pedestrians who described themselves as ‘safe pedestrians’ were less likely to intend to cross the road as depicted in the scenario. Sparks has questioned the extent to which self-identity may be simply operating as a proxy measure for past behaviour (Sparks, 1994). For example, in relation to exercise, engaging in frequent exercise may lead an individual to form a self-identity in which exercising is a salient component. However, a number of studies have found that the relationship between self-identity and intention remains when past behaviour is controlled for (Sparks and Shepherd, 1992; Theodorakis, 1994; Terry et al., 1999). It is also possible that the relationship between self-identity and intention may be moderated by past behaviour. Given that self-identity is, in part, based on past behaviour, it follows that one’s self-identity in a particular domain will become stronger, and more predictive of intentions, as experience with the behaviour increases (Charng et al., 1988). In the present context, given that road-crossing is a frequently performed behaviour, one would expect self-identity to be predictive of road-crossing intentions. However, formal tests of this moderation hypothesis have produced mixed results (Charng et al., 1988; Conner and McMillan, 1999; Terry et al., 1999). More work is therefore required to establish the nature of the relationship between self-identity and past behaviour.
The present study has a number of limitations that should be noted. First, a number of the extended TPB variables were measured with two-item scales. While the internal reliabilities of these scales were satisfactory, the use of multi-item scales would be preferable, although this has to be balanced against increasing the length of the questionnaire to be completed by adolescent respondents. Moreover, it is worth noting that the anticipated affect measure used in the present study would benefit from the inclusion of further items as it only focused on two, positive, anticipated emotions that attempted to tap into the ‘thrill’ of crossing the road in a potentially dangerous manner. It is likely that other, negative, emotions are likely to be experienced following risky road-crossing behaviour such as fear and worry.
Second, it was not possible to construct a reliable scale for the perceived behavioural control construct. Other researchers have noted similar difficulties [e.g. (Parker et al., 1995; Sutton et al., 1999)]. Nevertheless, the single item used to measure perceived behavioural control (i.e. easy–difficult) has been used frequently in previous studies and was found to be the strongest predictor of road-crossing intentions in the present study. Moreover, Parker et al. reported that such an easy–difficult item correlated significantly with a belief-based measure of perceived behavioural control in their study on motorway driving violations (Parker et al., 1995).
Third, the dependent variable in the present study was intention in response to a road-crossing scenario. Scenarios in which the respondent is depicted committing the risky behaviour are typically used in applications of the TPB to road safety behaviour [e.g. (Parker et al., 1992, 1995; Evans and Norman, 1998)], as directly asking respondents to report their intention to perform such risky behaviours may lead to under-reporting due to social desirability effects. However, it is possible that changes in the wording of the scenario may produce different findings, although in the present study a manipulation in which the respondent was depicted as being alone versus with friends had little effect on strength of the extended TPB variables or on their relationships with intention. It is also possible to question the extent to which responses to such scenarios may reflect actual road-crossing behaviour. Evans and Norman have argued that individuals may have tendencies to make typically safe or unsafe road-crossing decisions that are supported and reinforced by their beliefs (Evans and Norman, 1998). As a result, responses to scenarios may reflect typical road-crossing behaviour, and a number of studies have provided support for the link between such intentions and typical road safety-related behaviour. For example, Parker found significant correlations between car drivers’ intentions to speed when presented with a speeding scenario and observations of their speed along three sections of a road (Parker, 1997). Evans has also reported a significant correlation between road-crossing intentions and performance on a simulated road-crossing task among 12 year olds (Evans, 1999). Furthermore, Norman and Evans reported that schoolchildren who had previously been involved in a road traffic accident as a pedestrian were more likely to intend to cross the road in the manner depicted in potentially hazardous scenarios (Norman and Evans, 1996). These findings, together with the results of meta-analyses that suggest strong intention–behaviour relations across a range of behaviours (Sutton, 1998), provide support for a link between individuals’ responses to scenarios and their typical road safety behaviour.
The present study has a number of practical implications. The extended TPB was found to be predictive of adolescents’ road-crossing intentions and, as such, may provide an appropriate framework for the development of interventions to encourage safer road-crossing behaviour among adolescents (Fishbein, 1993; Hardeman et al., 2002). For example, the TPB has been used to develop a theory-based intervention that has been found to change attitudes and intentions towards drink-driving, and towards being a passenger of a drink-driver, among 14- to 15-year-old students (Queensland Drink Driving Project, 1990) and to also reduce the incidence of drink-driving passenger behaviour at 3-year follow-up (Sheehan et al., 1996). Similarly, Quine et al. (Quine et al., 2001) used the results of an earlier study (Quine et al., 1998) to identify the key beliefs that were predictive of cycle helmet use intentions and behaviour among school-age cyclists. These beliefs were then incorporated into a booklet, containing a series of persuasive messages, which was found to lead to increased cycle helmet use at 5-month follow-up.
In the present study, perceived behavioural control was found to be the strongest predictor of intention, suggesting that interventions should target perceptions of the ease or difficulty of crossing the road in potentially hazardous situations vis-à-vis using road-crossing facilities. Perceptions of control are seen to be based on both internal (e.g. skills, self-efficacy) and external (e.g. opportunities, constraints) factors (Ajzen, 1991). Thus, engineering interventions may focus on external factors such as the road layout in order to make it more difficult to cross the road in potentially dangerous situations. However, research has shown that such an approach can be counterproductive. For example, while remedial engineering work at accident blackspots may reduce the number of accidents at the original site it may also lead to an increase in the number of accidents in the surrounding area (Ebbecke and Shuster, 1977; Boyle and Wright, 1984). As a result, it may be more profitable to invest in educational interventions that make adolescents more aware of the difficulty of crossing the road in potentially dangerous situations (i.e. reduce perceptions of control). Such an approach has been followed by Evans and Norman (Evans and Norman, 2002) in a pilot intervention in which members of a school drama class were invited to develop a short drama piece on road safety awareness. The class was presented with the main findings of the present study and instructed to incorporate the information into the drama piece, thereby encouraging active processing of the material (Petty and Cacioppo, 1986). Encouragingly, at the end of the project the class members were found to be less likely to intend to cross the road in a potentially dangerous manner (as depicted in a scenario) and more likely to believe that doing so would be difficult. Further work is required to develop and evaluate theory-based interventions to encourage safer road-crossing behaviour given the continuing vulnerability of adolescent pedestrians (Department of Transport, 1998) and the lack of road safety education in secondary schools in the UK (Singh and Spear, 1989).
|
Notes |
|---|
1 A further manipulation was included in the study. The scenario was also written depicting the target person crossing the road with his/her friends. Half the respondents read the scenario with the target person depicted as being alone, whereas the other half read the scenario with the target person depicted as being with friends. Responses to these two scenario descriptions were first compared using a series of t-tests. Only one significant difference was found with the ‘alone’ description attracting more social disapproval (t = 2.83, d.f. = 1825, P < 0.01). A moderated regression analysis was conducted in order to examine whether the type of description moderated the relationships between the extended TPB and road-crossing intentions (Baron and Kenny, 1981). Separate regression analyses were conducted for each type of description and the unstandardized regression coefficients were compared (Cohen and Cohen, 1983
). No reliable differences were found between the unstandardized regression coefficients obtained in each regression analysis, thereby indicating that the type of description (i.e. alone or with friends) failed to have a moderating effect on the relationships between the TPB and additional variables and intentions to cross the road. As a result, subsequent analyses are reported for combined datasets. 
|
References |
|---|
|
TopAbstractIntroductionMethodResultsPredicting road-crossing...DiscussionReferences |
|---|
Ajzen, I. (1988) Attitudes, Personality and Behavior. Open University Press, Milton Keynes.
Ajzen, I. (1991) The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179–211.[CrossRef][ISI]
Ajzen, I. and Fishbein, M. (1980) Understanding Attitudes and Predicting Social Behavior. Prentice-Hall, Englewood Cliff, NJ.
Armitage, C. J. and Conner, M. (2001) Efficacy of the theory of planned behaviour: a meta-analytic review. British Journal of Social Psychology, 40, 471–499.[CrossRef][ISI][Medline]
Avery, J. G. and Jackson, R. (1993) Children and their Accidents. Arnold, London.
Baron, R. M. and Kenny, D. A. (1986) The moderator–mediator distinction in social psychological research: conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.[CrossRef][ISI][Medline]
Beck, L. and Ajzen, I. (1991) Predicting dishonest actions using the theory of planned behavior. Journal of Research in Personality, 25, 285–301.[CrossRef]
Berg, C., Jonsson, I. and Conner, M. (2000) Understanding choice of milk and bread for breakfast among Swedish children aged 11–15 years: an application of the theory of planned behaviour. Appetite, 34, 5–19.[CrossRef][ISI][Medline]
Biddle, B., Bank, B. and Slavings, R. (1987) Norms, preferences, identities and retention decisions. Social Psychology Quarterly, 50, 322–337.[CrossRef]
Boyle, A. J. and Wright, C. C. (1984) Accident ‘migration’ after remedial treatment at accident blackspots. Traffic Engineering and Control, 25, 260–267.
Charng, H. W., Piliavin, J. A. and Callero, P. L. (1988) Role identity and reasoned action in the prediction of repeated behavior. Social Psychology Quarterly, 51, 303–317.
Cohen, J. and Cohen, P. (1983) Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Erlbaum, Hillsdale, NJ.
Conner, M. and Armitage, C. (1998) Extending the theory of planned behavior: a review and avenues for further research. Journal of Applied Social Psychology, 28, 1429–1464.[CrossRef]
Conner, M. and McMillan, B. (1999) Interaction effects in the theory of planned behaviour: studying cannabis use. British Journal of Social Psychology, 38, 195–222.
Conner, M. and Sparks, P. (1996) The theory of planned behaviour and health behaviours. In Conner, M. and Norman, P. (eds), Predicting Health Behaviour. Open University Press, Buckingham, pp. 121–162.
Demetre, J. D. and Gaffin, S. (1994) The salience of occluding vehicles to child pedestrians. British Journal of Educational Psychology, 64, 243–251.
Department of Transport (1998) Road Accidents in Great Britain 1997. HMSO, London.
Ebbecke, G. M. and Shuster, J. J. (1977) Areawide Impact of Traffic Control Devices. Transport Research Board Record 644, Washington, DC.
Evans, D. (1999) Understanding and changing road safety awareness in adolescents. Doctoral thesis. University of Wales, Swansea.
Evans, D. and Norman, P. (1998) Understanding pedestrians’ road crossing decisions: an application of the theory of planned behaviour. Health Education Research, 13, 481–489.
Evans, D. and Norman, P. (2002) Improving pedestrian road safety among adolescents: an application of the theory of planned behaviour. In Rutter, D. and Quine, L. (eds), Changing Health Behaviour: Intervention and Research with Social Cognition Models. Open University Press, Buckingham, pp. 153–171.
Fishbein, M. (1993) Introduction. In Terry, D. J., Gallois, C. and McCamish, M. (eds), The Theory of Reasoned Action: Its Application to AIDS-Preventive Behaviour. Pergamon, Oxford, pp. xv–xxv.
Grayson, G. B. (1975) Observations of Pedestrians at Four Sites. Department of the Environment Report 670. TRRL, Crowthorne.
Hagger, M. S., Chatzisarantis, N., Biddle, S. J. H. and Orbell, S. (2001) Antecedents of children’s physical activity intentions and behaviour: predictive validity and longitudinal effects. Psychology and Health, 16, 391–407.
Hardeman, W., Johnston, M., Johnston, D. J., Bonetti, D., Wareham, N. J. and Kinmonth, A. L. (2002) Application of the theory of planned behaviour in behaviour change interventions: a systematic review. Psychology and Health, 17, 123–158.[CrossRef][ISI]
Jones, D. (1990) Child Casualties in Road Accidents. Department of Transport, London.
Lynam, D. and Harland, D. (1992) Child pedestrian safety in the UK. Presented at the VTI/FERSI Conference, Berlin.
Manstead, A. S. R. (2000) The role of moral norm in the attitude–behavior relation. In Terry, D. J. and Hogg, M. A. (eds), Attitudes, Behavior and Social Context. Lawrence Erlbaum, Mahwah, NJ, pp. 11–30.
Manstead, A. S. R. and Parker, D. (1995) Evaluating and extending the theory of planned behaviour. In Stroebe, W. and Hewstone, M. (eds), European Review of Social Psychology. Wiley, Chichester, vol. 6, pp. 69–95.
Norman, P. and Evans, D. (1996) Predicting school children’s road crossing intentions: it’s easy to feel good. Paper presented at the 11th General Meeting of the European Association of Experimental Social Psychology, Gmunden, Austria.
O’Callaghan, F. V., Callan, V. J. and Baglioni, A. (1999) Cigarette use by adolescents: attitude–behavior relationships. Substance Use and Misuse, 34, 455–468.
Parker, D. (1997) The relationship between speeding attitudes and speeding behaviour. In Grayson, G. (ed.), Behavioural Research in Road Safety VII. Transport Research Laboratory, Crowthorne, pp. 176–184.
Parker, D., Manstead, A. S. R., Stradling, S. G. and Reason, J. T. (1992) Intentions to commit driving violations: an application of the theory of planned behavior. Journal of Applied Psychology, 77, 94–101.[CrossRef]
Parker, D., Manstead, A. S. R. and Stradling, S. G. (1995) Extending the theory of planned behaviour: the role of personal norm. British Journal of Social Psychology, 34, 127–137.[ISI]
Petty, R. E. and Cacioppo, J. T. (1986) Communication and Persuasion: Central and Peripheral Routes to Attitude Change. Springer, New York.
Pless, I. B., Peckham, C. S. and Power, C. (1989). Predicting traffic injuries in childhood: a cohort analysis. Journal of Pediatrics, 115, 932–938.[CrossRef][ISI][Medline]
Queensland Drink Driving Project (1990) Development and Implementation of the Plan a Safe Strategy Drink Driving Prevention Program. NCADA Monograph Series 13. AGPS, Canberra.
Quine, L., Rutter, D. and Arnold, L. (1998) Predicting and understanding safety helmet use among schoolboy cyclists: a comparison of the theory of planned behaviour and the health belief model. Psychology and Health, 13, 251–269.
Quine, L., Rutter, D. and Arnold, L. (2001) Persuading school-age cyclists to use safety helmets: effectiveness of an intervention based on the theory of planned behaviour. British Journal of Health Psychology, 6, 327–345.[CrossRef][ISI][Medline]
Richard, R., Van der Pligt, J. and de Vries, N. (1996) Anticipated affect and behavioral choice. Basic and Applied Social Psychology, 18, 111–129.[CrossRef]
Rise, J. and Wilhelmsen, B. (1998) Prediction of adolescents’ intention not to drink alcohol: theory of planned behavior. American Journal of Health Behavior, 22, 206–217.
Sheehan, M., Schonfeld, C., Ballard, R., Schofield, F., Najman, J. and Siskind, V. (1996) A three year outcome evaluation of a theory based drink driving education program. Journal of Drug Education, 26, 295–312.[ISI][Medline]
Singh, A. and Spear, M. (1989) Traffic Education: A Survey of Current Provision and Practice in Secondary Schools. TRRL Report CR115, DTRRL, Crowthorne.
Southwell, M. T., Carsten, O. M. J. and Tight, M. R. (1990) Contributory Factors in Urban Road Accidents. Institute for Transport Studies, University of Leeds.
Sparks, P. (1994) Attitudes towards food: applying, assessing and extending the theory of planned behaviour. In Rutter, D. R. and Quine, L. (eds), Social Psychology and Health: European Perspectives. Avebury, Aldershot, pp. 25–46.
Sparks, P. and Shepherd, R. (1992) Self-identity and the theory of planned behavior: assessing the role of identification with green consumerism. Social Psychology Quarterly, 55, 388–399.[CrossRef]
Sutton, S. (1998) Predicting and explaining intentions and behavior: how well are we doing? Journal of Applied Social Psychology, 28, 1317–1338.[CrossRef][ISI]
Sutton, S., McVey, D. and Glanz, A. (1999) A comparative test of the theory of reasoned action and the theory of planned behavior in the prediction of condom use intentions in a national sample of English young people. Health Psychology, 18, 72–81.[CrossRef][ISI][Medline]
Terry, D., Hogg, M. and White, K. (1999) The theory of planned behaviour: self-identity, social identity and group norms. British Journal of Social Psychology, 38, 225–244.
Theodorakis, Y. (1994) Planned behavior, attitude strength, role identity, and the prediction of exercise behavior. The Sport Psychologist, 8, 149–165.
Tight, M. R. (1996) A review of road safety research on children as pedestrians: how far can we go towards improving their safety? ATSS Research, 20, 69–74.
Van der Pligt, J. and de Vries, N. (1998) Expectancy-value models of health behaviours: the role of salience and anticipate affect. Psychology and Health, 13, 289–305.
Ward, H., Cave, B. L., Morrison, A., Allsop, R. and Evans, A. (1994) Pedestrian Activity and Accident Risk. AA Foundation for Road Safety Research, Basingstoke.
Whitebread, D. and Neilson, K. (1996) Cognitive and Metacognitive Processes Underlying the Development of Children’s Pedestrian Skills. Department of Transport Report S2/141 Child Development. HMSO, London.
Received on September 3, 2001; accepted on April 20, 2002
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||