Developing an online educational curriculum to enhance parent-supervised driving

描述(申请人提供)：车祸是青少年死亡和受伤的主要原因，每年造成3000多人死亡，100倍的受伤，以及超过140亿美元的相关成本。疾控中心已将交通事故和相关伤害列为公共卫生的首要任务。经验不足是青少年司机新手撞车的主要原因，但积累必要的经验以成为一名安全的司机可能需要多年时间。幸运的是，来自驾驶和其他领域的证据表明，通过基于情景的培训来增加经验性知识是可能的。目前为青少年提供练习的方法主要是父母监督驾驶。然而，家长们还没有做好应对这一角色的准备，许多家长只专注于驾驶技巧、法律和一般安全建议。很少有家长讨论驾驶决策方面的问题，甚至可能会传递不准确的信息。许多家长限制了青少年面临的驾驶环境的范围，错误地认为这会增加安全性，而不是危险地限制青少年积累重要驾驶经验的能力。我们的长期目标是设计和验证一种创新的工具，旨在加快青少年司机关键安全知识的获得。我们建议通过开发适合父母用来指导他们经验不足的青少年司机的场景来扩展我们团队过去的工作。这些场景将是在线教育课程的一部分，旨在帮助父母在青少年学习驾驶的过程中为他们提供指导练习。该在线工具将把基于情景的培训与机器学习技术的创新应用相结合，以评估情景反应，并提供量身定制的反馈，其中包含实用的、适合发展的提高安全性的战略，以及对父母监督的道路实践的建议。第一阶段将解决这些具体目标：1)进行基础性研究，通过对15-18岁青少年司机的半结构化访谈，确定新手青少年司机经常遇到的现实情景。2)收集具有不同驾驶经验的青少年和成年人对情景的代表性反应，以确定知识的进展情况，并确定适合发展的策略作为反馈。3)创建机器学习算法，以提供量身定制的反馈，并根据对场景的响应推荐道路驾驶体验。4)开发原型在线系统。第二阶段将侧重于额外的具体目标：5)将第一阶段的概念证明发展成具有改进算法的完整在线课程。6)通过驾驶模拟器实验和对青少年新手司机的有限现场试验，评估在线课程的可用性、通过焦点小组获得的接受度和有效性。拟议的产品代表着青少年司机培训方法的重大转变，通过第三阶段的传播，它将满足以证据为基础的家长教育司机的迫切需求。 与公共健康相关：车祸是青少年伤亡的主要原因，每年造成3000多人死亡，100倍的伤害，以及超过140亿美元的相关成本；尽管如此，许多青少年接受驾驶教育的主要来源是他们的父母。在第一阶段，我们建议扩展我们过去教授体验式知识的工作，并开发一个原型在线教育课程，允许家长利用创新的机器学习技术为青少年提供指导练习，而第二阶段将允许我们完成原型并评估其有效性。拟议的应用程序代表着培训方法的重大转变，通过第三阶段的传播，有可能改善青少年驾驶安全。

DESCRIPTION (provided by applicant): Motor vehicle crashes are the leading cause of death and injury for teens, accounting annually for over 3000 deaths, 100 times as many injuries, and over 14 billion dollars in associated costs. The CDC has identified traffic crashes and associated injuries as a top public health priority. Inexperience is the leading cause of crashes among novice teen drivers, but accumulating the necessary experiences to become a safe driver can take many years. Fortunately, evidence from driving and other domains suggests that it is possible to increase experiential knowledge through scenario-based training. Current methods for providing teens with practice focus on parent-supervised driving. However, parents are ill-prepared to handle this role, with many focusing only on the mechanics of driving, laws and general safety advice. Few parents discuss decision-making aspects of driving and may even pass on inaccurate information. Many parents limit the range of driving situations teens are exposed to mistakenly believing that this increases safety rather than dangerously limiting teens' ability to accumulate important driving experience. Our long-term objective is to design and validate an innovative tool that aims to accelerate the acquisition of critical safety knowledge for teen drivers. We propose to extend our team's past work by developing scenarios appropriate for parents to use to mentor their inexperienced teen drivers. The scenarios will be part of an online educational curriculum designed to help parents provide teens with guided practice as they learn how to drive. The online tool will combine scenario- based training with innovative applications of machine learning technology to evaluate scenario responses and provide tailored feedback containing practical, developmentally appropriate strategies for improving safety as well as recommendations for parent-supervised on-the-road practice. Phase I will address these specific aims: 1) Conduct foundational research to identify realistic scenarios commonly encountered by novice teen drivers through semi-structured interviews with teen drivers aged 15-18. 2) Collect representative responses to scenarios from teens with different levels of driving experience and adults to identify the progression of knowledge and identify developmentally appropriate strategies to use as feedback. 3) Create machine learning algorithms to provide tailored feedback and recommend on-the-road driving experiences based on responses to the scenarios. 4) Develop prototype online system. Phase II will focus on additional specific aims: 5) Develop the Phase I proof of concept into a complete online curriculum with refined algorithms. 6) Evaluate the online curriculum for usability, acceptance via focus groups, and effectiveness via a driving simulator experiment and limited field trial with novice teen drivers. The proposed product represents a significant shift in trainig approaches for teen drivers and through Phase III dissemination it will fill a critical need for evidence- based parent-taught driver's education. PUBLIC HEALTH RELEVANCE: Motor vehicle crashes are the leading cause of death and injury for teens, accounting annually for over 3000 deaths, 100 times as many injuries, and over 14 billion dollars in associated costs; despite this, many teens' primary source of driver's education is their parents. In Phase I we propose to extend our past work for teaching experiential knowledge and develop a prototype online educational curriculum that allows parents to provide teens with guided practice by utilizing innovative machine learning technologies, while Phase II will allow us to complete the prototype and evaluate its effectiveness. The proposed application represents a significant shift in training approaches that has the potential, through Phase III dissemination, to improve teen driving safety.