Influences on pedestrian risktaking in young adolescents

1
Influences on pedestrian risk-taking in young
adolescents

• Andy Tolmie
• Dept of Psychology and Human Development
• Institute of Education University of London
• Jimmie Thomson
• Dept of Psychology
• University of Strathclyde

Contact details 25 Woburn Square London WC1H
0AA Tel [phone] Fax [phone]
6304 Email a.tolmie_at_ioe.ac.uk
2
Other members of research team

• Hugh Foot, University of Strathclyde
• Rory OConnor, University of Stirling
• Eleni Karagiannidou, University of Strathclyde
• Maggie Banks, University of Strathclyde
• Pepi Sarvary, University of Strathclyde
• Chris ODonnell, University of Paisley
• Work funded by UK Department for Transport

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
3
Adolescents as pedestrians

• Evidence suggests children first approach adult
  levels of pedestrian skill at 10-12 years1
• However, peak pedestrian injury rates in UK occur
  later, between 12 and 15 years2
• But, evidence on skills misleading?
• Transition to secondary school in UK results in
  increased demands3
• Young adolescents regard road safety concerns as
  childish4
• 1 Thomson et al, 1996 Tolmie et al, 2003
• 2 TRL, 2004 DfT, 2005
• 3 Platt, 1998
• 4 Thomson et al, 2001 Lupton Bayley, 2001

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
4
Possible hypothesis

• Young adolescents overestimate their capabilities
  in more challenging traffic environments because
  they are less used to these
• ie their perception of their competence fails to
  match their actual competence
• They pay less attention to the effectiveness of
  their judgements, because they assume they are
  able to cope
• They make many poor or marginal decisions as a
  result

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
5
Tolmie et al. (2006)

• Assessed actual and perceived pedestrian skills
  of 11 to 15 year olds and adults using
  simulations in four skill areas1
• Validated simulations against performance at
  roadside
• Found adolescents do not perform at adult levels
  of skill under more challenging conditions
• But, much more likely than adults or younger
  children to over-estimate relative competence,
  especially post-performance
• Effect was restricted to 13 to 15 year olds, so
  not simple function of transition to secondary
  product of peer influenced disregard of risk?
• 1 safe route planning visual timing/gap
  selection perceiving cues to drivers
  intentions using designated crossings

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
6
Present study

• Does misperception of performance lead to
  hazardous behaviour?
• What attitudes and norms operate for adolescents
  in relation to pedestrian behaviour?
• Is risk-taking perceived as a positive
  characteristic?
• What is the relative impact of skill, attitude
  and identity on pedestrian decision-making?
• Present study addressed all these issues, by
• - collecting data from one sample of 12 to 15
  year olds on all elements
• - relating measures to self-reports of actual
  behaviour (cf. Theory of Planned Behaviour
  framework, Ajzen Madden, 1986)

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
7
Methodology

• Computer-based measures of actual skill and
  perceived difficulty as in Tolmie et al. (Session
  1)
• On-line questionnaire (Session 2), collecting
  data for eleven target behaviours, some cautious
  (eg waiting for green man) and some risky (eg
  running through tight gap)
• - attitude to, and intention to perform target
  behaviour
• - perceived approval of others, PBC, plus parent
  and peer norms
• - also measure of self-identity (Q-sort) and
  general risk-taking1
• Self-report measure of subsequent performance of
  target behaviours, plus accident/near-miss
  history (Session 3)
• 1 modified version of Attitudes Towards
  Risks Questionnaire (Franken, Gibson Rowland,
  1992)

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
8
Sample

• c.100 participants (n307) in each of three age
  groups
• - 12/13 year olds (S1)
• - 13/14 year olds (S2)
• - 14/15 year olds (S3)
• Participants drawn from four secondary schools in
  West Central Scotland
• Balanced for gender, full range of SES

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
9
Results attitude, approval and norms
Measure mean(cautious) mean(risky)
(standardised indices)
1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
10
Results risk-taking and norms
Measure Q-sort/general questionnaire
composite from factor analysis
1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
11
Results risk-taking, intentions, behaviour
1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
12
Summary of profile results

• Adolescents are in general positive about
  cautious pedestrian behaviours, and negative
  about risky
• Perceived approval and parental norms follow very
  similar profile
• Peers are seen as substantially less
  cautious/more risky
• Self-identity/risk-taking profile lies between
  parent and peer norms, and trend is towards
  greater risk with age
• Intentions and behaviour follow similar pattern
  to identity, but behaviour tends to be less
  cautious than intended, suggesting additional
  influences at work
• Averages mask considerable individual variability
  

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
13
Regression analyses

• Given variability, what predicts less
  cautious/more risky intentions and behaviour?
• Separate regression analyses carried out for
  intention and self-report of behaviour for each
  of the 11 target actions
• Variables entered in following order
  demographics, intention (for behaviour only),
  TPB, skill, perceived skill, norms,
  self-identity/risk-taking
• Results very consistent across target behaviours,
  models account for moderate to large proportion
  of variance (adj R2 .28 to .69 for intentions,
  .17 to .43 for behaviours)

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
14
Regression results - intentions

• Attitudes and approval are weak to modest
  predictors of intentions (final beta .07 to
  .30, .02 to .18 respectively)
• Approval is initially stronger predictor for
  cautious intentions (beta .09 to .35), but
  values drop when norms included
• Peer norms are initially modest predictors of
  risky intentions (beta .22 to .30), parent
  norms weakly predict cautious (beta .04 to .17)
• Self-identity/risk-taking is strongest overall
  predictor (final beta .20 to .65), and
  influence of peer norms drops when this is
  included
• Parent norms are associated with approval (beta
  .09 to .35), peer norms with self-identity/risk-ta
  king (beta .08 to .37) ie effects of norms on
  intention appear to be partially indirect

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
15
Regression results - behaviours

• Intention is modest predictor, not always that
  (final beta .05 to .20)
• Peer norms predict risky behaviours (final beta
  .03 to .23), parent norms cautious (final beta
  .07 to .26) ie effects on behaviour appear to be
  primarily direct (sheep-following vs established
  habit?)
• Self-identity/risk-taking is only modest
  predictor here, and not for all behaviours (final
  beta .01 to .28)
• Risky behaviours were associated with near misses
  (beta .11 to .34), and thence with accidents (r
  .24) ie impact is tangible
• Misperceptions of competence were more strongly
  associated with self-identity/risk-taking (r
  .14) than with behaviour (rav .03)
• Skills (especially safe route planning) were
  positively associated with more cautious
  behaviour (beta .09 to .15)

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007
16
Implications and conclusions

• Adolescents are pulled towards riskier intentions
  and behaviour by perceptions of peers, attempts
  to be like them
• Misperceptions of competence are symptomatic of
  pull towards risk
• Parental behaviour provides pull in opposite
  direction, via modelling of safe habits, approval
  of cautious behaviour
• Support for parental modelling of safe behaviour
  during primary school likely to be productive
  intervention
• Better skills provide a further measure of
  protective influence
• Training adolescents might be counterproductive
  due to perception of childishness, but if they
  contributed to training of younger children, this
  might improve awareness of skill gaps

1st Injury Prevention Conference for the UK and
Ireland, Bristol, 13th September 2007