104

							Overview of teen driver risk and intervention approaches: Past, present, and future directions
                                   Teresa M. Senserrick, Ph.D.
                           TraumaLink, Children’s Hospital of Philadelphia


Statement of problem
In the United States, young drivers have the highest per capita and per mile driven crash rates of any
age group. Nearly 4,000 die each year. The estimated annual economic cost of their crashes is $41
billion. They are also one of the fastest growing segments of the population, which urges
improvement and innovation in intervention developments.
What are the key skills and behaviors to address?
While driver age/maturity alone is related to crash risk, inexperience is the major contributor,
particularly regarding visual search and hazard recognition skills, choice of headways and speeds
relative to conditions, and errors in attention. Young novices also drive in risky conditions more
often than adults: on weekends, at night, with peers, and in recreational circumstances. While
intentional risk behaviors clearly increase crash risk, it is unclear what proportion of teens engages in
these behaviors and for what proportion of their total driving. Seat-belt use is particularly low
among US teens, especially at night and with alcohol use. Underage and unlicensed driving are not
an insignificant part of the problem.
The most common intentional risk behavior is speeding. While drunk-driving is less common, when
novices do drive under the influence, their risk increases with increasing BAC to a much greater
extent than adults, and is further inflated at night and with each additional passenger. Drunk-driving
is associated with drug-driving among young drivers, and both behaviors are associated with
increased reckless driving.
Driver education and training
Among the earliest teen driver interventions were school-based education and group training
programs. As was true for early attempts, recent approaches fail to be associated with crash
reductions. Moreover, programs focusing on vehicle-handling skills in emergency scenarios have
had counterproductive outcomes. In contrast, individual or small group-based programs focusing on
truly defensive driving, such as effective visual scanning, safer headways and stopping distances are
effective in graduated driver licensing (GDL) programs, but inconsistent effects are found in small
experimental studies.
A recent advance has been the development of CD-ROM training products to improve key cognitive-
perceptual skills, primarily hazard perception, which show effectiveness in simulated drives. It has
long been argued that such skills should be a key focus of training, although early attempts to
encourage development of these skills, such as GDL hazard perceptions tests, showed limited
success.
A potential new approach is environmentally-friendly driving courses, which have been found not
only to reduce fuel consumption and emissions, and therefore costs, but also to reduce crash risk and
increase positive driver motivation among fleet drivers. Some of these concepts are currently
included in compulsory training in Scandinavia
While some researchers argue that education and training do not work, these new developments are
offering support for how training can be improved and can work. The need to address “higher-order
skills”, cognitive-perceptual, attitudinal-motivational and social aspects of driving, and follow basic
learning principals, such as allowing training to be personalized, self-paced, over longer periods,
allowing practice and repetition of learned skills, and including tailored feedback, is increasingly
recognized. It is generally acknowledged, however, to be most effective, programs should be
incorporated into GDL.
Graduated Driver Licensing
GDL has undoubtedly been the most successful teen driving intervention of its day, with significant
reductions in crashes and fatalities of novices drivers reported after implementation. The most
effective components, restrictions on nighttime driving, passenger carriage and a zero BAC limit,
work primarily by reducing driving exposure, rather than changing behavior per se. There is,
however, some indication that good driving record and reduced demerit point systems provide
motivation to avoid violations. Initiatives to increase supervised practice driving are also showing
that those that use the opportunity to undertake more and varied practice are at reduced risk of a
crash in the months following licensure.
There is, nonetheless, much room for development. Not all states have implemented the most
effective components or implemented them to their full potential, and some allow earlier licensure
following driver education/training - a counterproductive component. A new promising restriction,
mobile phone use, has had limited application. Better guidelines are needed on how to best structure
the supervised practice period and many existing theory and practice driving tests inadequately
address safety. There is also a need to evaluate the long-term benefits of GDL and whether some
aspects of crash risk are shifted to later stages.
The role of parents
GDL systems that encourage a high level of parental involvement are associated with greater
involvement by teens, such as obtaining more driving practice than required. Moreover, general
parental monitoring and involvement in their teens’ activities (such as knowing where they are
going, with whom, and when they will be home), is associated with lower risk-taking among young
teens in terms of smoking, alcohol and other drugs use, as well as riding with an impaired driver.
A more specific initiative is the use of parent-teen contracts, requiring commitments from both
parties to help teens progress from less risky to more risky driving. These have shown success,
including over time. Dr. Bruce Simons-Morton, Chief, Prevention Research Branch, DESPR,
NICHD, will focus on this initiative in his presentation, the second in this Special Session.
Some cautionary research has also emerged in relation to parent roles. French research suggests
some parents continue to share the driving taskload with the learning driver throughout extended
practice periods, rather than transferring these skills to achieve independent driving. Research has
also demonstrated child modeling of poor driving behaviors of their parents in relation to drunk-
driving and dangerous driving behaviors.
Overall, research highlights the key roles of parents, suggesting this as an important area to explore
further. Future research should consider issues such as: how successful parent-related interventions
could be effectively applied more widely; whether the success of parent-teen contracts during the
early license period can be adapted to the practice driving period; and how risky learning transfer
from parents to their children can be avoided.
Technological advances
The rapid development of technology has seen the emergence of innovative in-vehicle Intelligent
Transport Systems (ITS), which aim to assist with the driving task and reduce risk, as well as a range
of devices to provide feedback to parents on their children’s driving behavior and performance.
Professor Max Donath, Director, Intelligent Transportation Systems Institute, University of
Minnesota, will review these systems in his presentation, the third in this Special Session, with a
particular focus on monitoring devices.
Researchers have cautioned that research is needed to determine whether ITS developments will in
fact improve or worsen young driver crash involvement given their potential to add to the cognitive
workload and to provide additional distractions, especially for novices. Novices are found to be less
efficient than experienced drivers in cognitive processing of in-vehicle distractions and upkeep of
safe driving practices. This is of less concern for unobtrusive, automatic monitoring devices.
Other issues such as acceptability are also important. Early research showed young drivers were
generally in favor of ITS technologies and believed they could be particularly useful for young
drivers, but were not in favor of systems that could be circumvented, that compromised privacy or
were not compulsory for all drivers. The use of monitoring devices could also be viewed as
compromising trust relations between parents and their teens, or ignored or deactivated either by
teens or parents who are not committed their use.
While several available technologies could therefore be applied to the teen driver problem, it is likely
they would need to operate as part of an overall strategy, rather as an end in themselves to benefit the
majority. Further research is needed on a framework for how behavior can be changed effectively
and if and how these technologies fit within or support this framework.
Police enforcement, road engineering and other community initiatives
Young drivers, especially young males, are known to respond to actual detection, ticketing, warnings
and license revocations by Police, yet with limited resources, programs rarely target teens.
Compliance with differential regulations through GDL stages is likely to increase with increased
enforcement, as well as measures to inflate the perceived risk of detection. Improved awareness of
restrictions and penalties for non-compliance is also needed. General enforcement initiatives, such
innovative speed camera and red-light running programs, random breath testing for alcohol/other
drugs, and primary seatbelt laws also benefit teens.
Designated driver campaigns are problematic. Not only is the carriage of teen passengers a particular
high-risk combination, drivers are often selected during or after drinking (“the least drunk” driver),
with many above legal BAC limits, especially males. Compliance with impaired driving laws is
likely to increase when alternative transport options are provided in the community, such as “disco
buses” and other shuttle services, especially in rural and remote areas where options are limited.
Recidivist measures can also be improved for young drivers. Some current measures, such as
alcohol interlocks and vehicle immobilization/impoundment, are showing effectiveness in reducing
recidivism. In contrast, other measures, such as driver improvement programs rarely address issues
specific to novices and are less effective.
General engineering initiatives to improve the road environment for all drivers, will naturally help
reduce teen crashes, but some could be more effective given the most common teen crash types: e.g.,
improved sight-distances and blackspot treatments at intersections to reduce the common multiple-
vehicle crashes; and roadside and center barriers on long, straight, high-speed rural roads to reduce
the common loss-of-control crashes, including single-vehicle, run-off road crashes and head-on
crashes due to over-correction after lane departure.
Peer-focused programs that encourage teens to follow safe driving practices both as drivers and
passengers, have also shown success at the local (school or community) level, but lack a framework
for implementation at the wider level.
Lessons to learn from behavioral interventions in other domains?
Risky driving is not an activity that occurs in isolation, but in the context of a range of risky activities
considered normative during this developmental stage, such as smoking, alcohol, other drugs, unsafe
sexual practices, and gambling. Teen risk research demonstrates that many of these behaviors
covary, yet achievements gained within these domains are rarely applied to driving. Likewise, risky
driving is rarely considered in these other public health domains.
Intervention programs targeting other teen risk behaviors have shown recent successes in reducing or
preventing uptake, particularly in relation to smoking and unsafe sexual practices leading to teenage
pregnancies. This leads us to ponder whether these examples from the public health arena might be
applied to the driving domain. Peter Mitchell, President of Marketing for Change, will address this
issue in his presentation, the fourth and final presentation in this Special Session, with a focus on
successes of the anti-smoking Truth Campaign.