The Scottish Government

« Previous | Contents | Next »

Listen

Changing Speeding Behaviour in Scotland: An evaluation of the 'Foolsspeed' campaign

CHAPTER FOUR: THEORY OF PLANNED BEHAVIOUR COMPONENTS: RESULTS OVER 4 YEARS

As outlined in Section 1.3 above, each of the 40-second television advertisements was designed to influence one of the three main components within the Theory of Planned Behaviour, which predict behaviour via behavioural intentions (Figure 4.1):

Figure 4.1: Relationship of Foolsspeed advertising to TPB

In order to assess whether any desired changes occurred over the campaign period in drivers' attitudes, subjective norms and perceived behavioural control, as well as in other components of the Theory of Planned Behaviour, the survey took full measures of all of the components at the baseline survey, and repeated these measures at each follow-up survey (see Chapter 2 above). By comparing mean scores for all respondents who matched between the baseline survey and the follow-up surveys, it was possible to assess whether changes occurred in any of the TPB components over the three years. In order to assess whether any changes could be associated with the Foolsspeed campaign, results were analysed by awareness of the Foolsspeed advertising. They were also analysed by baseline reported speeding behaviour, to assess whether there were particular trends in response for frequent, occasional and infrequent speeders. Multiple regression analyses were conducted to assess the predictive strength of the TPB at each stage and over time, and to examine any possible contribution made by awareness of the Foolsspeed advertising.

The analysis focused on investigating potential changes in attitudes following the Mirror ad, at the 2 ^nd survey (1999); potential changes in subjective norms following the Friends & Family ad, at the 3 ^rd survey (2000); and potential changes in perceived behavioural control following the Simon Says ad, at the 4 ^th survey (2001). It also examined whether any changes in attitudes and subjective norms were sustained over subsequent survey(s).

Findings in this section compare the baseline and relevant follow-up survey scores for each of the TPB components: attitudes (4.1), subjective norms (4.2), perceived behavioural control (4.3), additional elements of the TPB (4.4), behavioural intentions (4.5), and behaviour (4.6). Section 4.7 reports results of the multiple regression analyses.

4.1 ATTITUDES

Within the Theory of Planned Behaviour, attitudes are made up of behavioural beliefs about the consequences of speeding (i.e. the likelihood of particular consequences occurring), and corresponding outcome evaluations of the desirability or otherwise of those consequences. These are summed together to produce a composite Attitude towards the Behaviour score. Table 4.1 shows mean scores for each of the behavioural beliefs, with the baseline results in italics and the 2 ^nd survey results - obtained after Mirror's first screening - in bold.

Table 4.1: Behavioural beliefs: comparison of baseline and 2 ^nd survey

The most salient beliefs (i.e. the consequences considered 'most likely' to occur as a result of speeding) at both survey stages were that speeding would make it 'difficult to stop quickly in an emergency' and would 'cause danger to pedestrians'. 'Causing an accident', 'getting caught for speeding' and 'being able to keep up with the flow of traffic' were also considered likely consequences, although to a lesser extent. Beneficial consequences such as 'getting to one's destination quicker', driving at what 'would feel like a comfortable speed' and 'saving time' were, on average, considered neither likely nor unlikely.

Table 4.1 indicates that the strength of two of the beliefs changed in an anti-speeding direction between the baseline and 2 ^nd survey. There was, on average, an increase in the perceived likelihood that it would be 'difficult to stop in an emergency' (baseline 1.22, 2 ^nd survey 1.50). There also seemed to be a decrease in the belief that driving at 40mph in the situation described 'would feel like a comfortable speed' (baseline -0.01, 2 ^nd survey -0.24).

Table 4.2 presents mean scores for the corresponding outcome evaluations, again comparing baseline results, given in italics, with the 2 ^nd survey results, given in bold.

Table 4.2: Outcome evaluations (behavioural beliefs): comparison of baseline and 2 ^nd survey

The most undesirable consequences at both survey stages were judged to be causing an accident, getting caught for speeding, causing danger to pedestrians, and finding it difficult to stop quickly in an emergency, all of which had high negative scores. The potentially beneficial consequences of speeding, listed in the lower half of the table, had relatively weak positive scores, suggesting that although they were all perceived as desirable outcomes, they were potentially outweighed by the negative consequences.

Between the baseline and 2 ^nd survey, three of the consequences of speeding increased in undesirability ('finding it difficult to stop quickly in an emergency,' 'causing an accident' and 'being able to keep up with the flow of traffic').

Within the Theory of Planned Behaviour, a measure of overall Attitude towards the Behaviour is computed by calculating, for each item, the product of the behavioural belief and the outcome evaluation and then taking the sum of the products for each item, i.e.

Table 4.3 displays Attitude towards the Behaviour at an individual item level, and the composite Attitude to the Behaviour score, comparing the baseline and 2 ^nd survey scores.

At this combined level, two of the belief items displayed a significant change, in an anti-speeding direction, between the baseline and 2 ^nd survey ('finding it difficult to stop quickly in an emergency' and 'driving at what you feel is a comfortable speed').

The composite attitude score also displayed a significant change in an anti-speeding direction.

Table 4.3: Attitudes towards the Behaviour: comparison of baseline and 2 ^nd survey

Changes in attitudes over time

In order to examine whether any changes in beliefs and attitudes found at the 2 ^nd survey would be sustained for the next two years of the study, similar comparisons were therefore made for those respondents matching between the baseline and 3 ^rd survey, and those matching between the baseline and 4 ^th survey (full results are given in Appendix B). A summary of the main changes over all survey stages is given in Table 4.4 overleaf. A p-value given in bold indicates that the change was in the desired direction (i.e. became more anti-speeding or less pro-speeding), while a p-value given in italics indicates that the change was in the 'wrong' direction (i.e. became less anti-speeding or more pro-speeding). A dash indicates that there was no significant change.

Table 4.4: Summary of changes in attitudes: comparison of baseline with follow-up surveys

Turning first of all to the behavioural belief items, the same two beliefs which displayed a change between the baseline and 2 ^nd survey again displayed changes in an anti-speeding direction between the baseline and the 3 ^rd survey ('difficult to stop in an emergency', baseline 1.17, 3 ^rd survey 1.40, p<0.05; 'would feel like a comfortable speed', baseline 0.07, 3 ^rd survey -0.26, p<0.01). At the 4 ^th survey, three of the belief items displayed a change in an anti-speeding direction when scores were compared with the baseline: 'would feel like a comfortable speed', baseline -0.04, 4 ^th survey -0.34, p<0.01; 'would get to my destination quicker', baseline 0.19, 4 ^th survey -0.17, p<0.01; and 'would save time', baseline 0.04, 4 ^th survey -0.37, p<0.001). One item displayed a change in the 'wrong' direction between baseline and 4 ^th survey - i.e. beliefs became less anti-speeding ('would cause an accident', baseline 0.68, 4 ^th survey 0.36, p<0.01). Overall, however, the impression from the behavioural belief scores is that desired changes in a small number of beliefs occurred and were sustained over the four years of the study.

As reported above, outcome evaluation scores for three of the beliefs increased in undesirability between the baseline and 2 ^nd survey. By the 3 ^rd survey, ratings of two of these outcomes had reverted to their baseline levels, although 'being able to keep up with flow of traffic' was still rated as less desirable comparing baseline and 3 ^rd survey scores (baseline 0.78, 3 ^rd survey 0.56, p<0.01). This item was also rated as less desirable at the 4 ^th survey (baseline 0.78, 4 ^th survey 0.52, p<0.01), as was 'finding it difficult to stop quickly in an emergency' (baseline -2.27, 4 ^th survey -2.49, p<0.05). Overall, it appears from the comparisons that there were only minor changes over the survey stages in how respondents rated the desirability of particular consequences of urban speeding, but that these changes were in the hoped-for anti-speeding direction.

When the combined items (behavioural belief x outcome evaluation scores) were compared, there appeared to be an improvement at the 3 ^rd survey, when combined scores for four of the items became significantly more anti-speeding: 'finding it difficult to stop quickly in an emergency' (baseline -2.69, 3 ^rd survey -3.41, p<0.05), 'being able to keep up with the flow of traffic' (baseline 1.30, 3 ^rd survey 0.89, p<0.05), 'driving at what you feel is a comfortable speed' (baseline 0.18, 3 ^rd survey -0.52, p<0.01) and 'saving time'(baseline 1.08, 3 ^rd survey 0.45, p<0.01). By the 4 ^th survey, however, only two of the items displayed a significant change, and one of these changes was in the 'wrong' direction (i.e. more pro-speeding): 'causing an accident': baseline -1.84, 4 ^th survey -1.11, p<0.05, change in pro-speeding direction; 'driving at what you feel is a comfortable speed': baseline 0.02, 4 ^th survey -0.65, p<0.05, change in an anti-speeding direction.

This suggests that the favourable changes in belief items observed between the baseline and 3 ^rd survey may have been wearing off by the 4 ^th survey. When the composite Attitude to the Behaviour scores over the four surveys are examined (Table 4.5), this is further confirmed.

Table 4.5: Composite Attitude to the Behaviour: comparison of baseline with each subsequent survey stage

The changes in individual behavioural beliefs and outcome evaluation scores reported in Tables 4.1-4.3 were strong enough to produce a significant change (in an anti-speeding direction) between the baseline and 2 ^nd survey in the overall Attitude to the Behaviour score, and this change was sustained at the 3 ^rd survey. This change in attitudes was no longer apparent by the 4 ^th survey, once again suggesting that the favourable changes previously observed may have been wearing off. However, further examination of the data revealed that this lack of observed change between the baseline and 4 ^th survey was possibly a symptom of sample attrition. Non-respondents at the 4 ^th survey had shown a shift in attitudes between baseline and 2 ^nd survey and baseline and 3 ^rd survey. Respondents at the 4 ^th survey had not shown a significant change in attitudes at those earlier surveys. It seems, therefore, that rather than changes wearing off, those who had earlier shown attitude change were less represented in the final survey.

Comparison by awareness of Mirror ad

In order to assess whether the observed changes in beliefs and attitudes could be associated with exposure to the Foolsspeed campaign, scores for those who had seen the Mirror ad were compared with scores for those who had not, at each survey stage.

Overall, the analysis suggests that there is some evidence for a modest effect by the Mirror ad on beliefs and attitudes. This effect appears to have been slightly stronger at the 3 ^rd survey than the 2 ^nd, but to be weaker by the 4 ^th survey.

The change in the composite Attitude to the Behaviour, observed at the 2 ^nd and 3 ^rd surveys, was not associated with awareness of the Mirror ad at the 2 ^nd survey, but was associated with having seen Mirror at the 3 ^rd survey (seen Mirror: baseline -5.85, 3 ^rd survey -7.96, p<0.05; Not seen Mirror: baseline -3.70, 3 ^rd survey -5.91, ns).

In addition, several of the changes in individual beliefs which make up attitudes were associated with awareness of the 'Mirror' ad. These were:

• the decrease at the 2 ^nd survey in the belief that 'driving down this road at 40 mph would feel like a comfortable speed'
• the increases at the 2 ^nd survey in the extent to which 'finding it difficult to stop quickly in an emergency' and 'causing an accident' were rated as undesirable outcomes of speeding. However, the decrease in the extent to which 'being able to keep up with the flow of traffic' was rated as desirable occurred only in the group who had not seen the Mirror ad.
• the increase at the 3 ^rd survey in the belief that speeding would make it difficult to 'stop quickly in an emergency', and the decrease in the belief that 'driving down this road at 40 mph would feel like a comfortable speed'.
• the decrease in the extent to which 'being able to keep up with the flow of traffic' was rated as desirable occurred only in the group who had seen the Mirror ad, at this survey stage. Overall, those who had not seen the Mirror ad at the 3 ^rd survey displayed no changes in behavioural beliefs or outcome evaluations.
• all the four observed changes in behavioural beliefs at the 4 ^th survey, including one change in a pro-speeding direction. However, none of the changes in outcome evaluation scores were associated with awareness of the Mirror ad at this stage.

Comparison by speeding behaviour

Attitude scores were also analysed to assess whether those who were categorised as frequent speeders at the baseline survey demonstrated similar or different patterns of response, in relation to attitudes, to those categorised as occasional and infrequent speeders. Table 4.6 presents the composite Attitude towards the Behaviour score for the three categories of speeder, comparing baseline and 2 ^nd survey.

Table 4.6: Attitude towards the Behaviour: comparison of baseline and 2 ^nd survey results, by baseline frequency of reported speeding

The results indicate that although the sample as a whole became more negative, between the baseline and 2 ^nd survey, in attitude towards speeding, the difference was only significant in those who were categorised as infrequent speeders at the baseline. Occasional and frequent speeders did not demonstrate a significant change in attitude, although scores did tend to become more negative. A similar pattern was observed at the 3 ^rd survey. However, at the 4 ^th survey, none of the three groups displayed a significant change in attitudes. Rather than changes in attitudes wearing off by the 4 ^th survey, it seems that those who had earlier reported attitude change were less likely to have participated in the 4 ^th survey.

4.2 SUBJECTIVE NORMS

Within the Theory of Planned Behaviour, subjective norms are made up of normative beliefs about how much significant others (or referents) would approve or disapprove of speeding and corresponding motivation to comply that is, to drive in a manner of which each of these referents might approve. These are summed together to produce a composite Subjective Norms score. Table 4.7 shows mean scores for each of the normative beliefs, with the baseline results given in italics and the 3 ^rd survey results - those obtained after the first screening of 'Friends and Family' given in bold.

Table 4.7: Normative beliefs: comparison of baseline and 3 ^rd survey

At both stages, the referents rated as most likely to disapprove of speeding was the police, followed by a competent driver, mother, partner/spouse and father. All the other referents were also rated as anti-speeding, with the exception of 'a typical young male driver', who was rated as slightly likely to approve of speeding.

Between the baseline and 3 ^rd survey, there were few changes in normative beliefs for most of the referents, apart from 'boss' and 'people you work with', both of whom were rated as significantly more likely to disapprove of speeding at the 3 ^rd survey compared to the baseline.

Motivation to comply scores are presented in Table 4.8, again comparing baseline and 3 ^rd survey scores.

Table 4.8: Motivation to comply: comparison of baseline and 3 ^rd survey

At both stages, respondents displayed the strongest motivation to comply with the police, a competent driver, passengers and closest friends, and lowest motivation to drive in a way which would be approved by a 'typical young male driver'.

Between the baseline and 3 ^rd survey, motivation to comply with four of the referents - partner/spouse, father, closest friends and boss - increased significantly.

Within the Theory of Planned Behaviour, a measure of overall Subjective Norms is computed by calculating, for each item, the product of the normative belief and motivation to comply and then taking the sum of the products for each item, i.e.

Table 4.9 displays Subjective Norms at an individual item level, plus the Composite Subjective Norms comparing the baseline and 3 ^rd survey scores.

Table 4.9: Subjective norms: comparison of baseline and 3 ^rd survey

The results suggest that the minor changes in individual elements shown in Tables 4.7 and 4.8 were not, unsurprisingly, substantial enough to produce a change between baseline and 3 ^rd survey, in Composite Subjective Norms or in any of the referent item scores other than 'people you work with'.

« Previous | Contents | Next »