车险理赔文献翻译.doc

2012届本科毕业设计（论文） 文 献 翻 译题目 某公司车险理赔数据分析与研究 学 院 商学院 专 业 信息系统与信息管理 班 级 信管081 学 号 084770538 姓 名 周小诗 指导教师 谢友才 开题日期 2011年11月6号 文献一： Effects of Pay-As-You-Drive vehicle insuranceon young drivers speed choice1. IntroductionWorldwide, across all age groups road traffic accidents currently account for 1.27 millions deaths a year (World Health Organization, 2004World Health Organization, 2004. Fact sheet: the top ten causes of death. Retrieved from . World Health Organization, 2004). Furthermore, traffic accidents constitute the second leading cause of death among people in the age of 20 to 24 years, right after HIV/AIDS (Toroyan and Peden, 2007). Young drivers are strongly overrepresented in road accident statistics. In the Netherlands, young adults (aged 1824) make up 8% of the driving population, but are involved in 22% of all severe traffic accidents. This means that per kilometer travelled, young adults in the Netherlands are 5.5 times as likely to be involved in a severe traffic accident as older adults (Schoon and Schreuders, 2003). The differences in accident involvement are often attributed to specific characteristics of young adults. For example young drivers, compared to older drivers, have been found to have a more positive attitude towards taking risks, display stronger motives for risky driving (Hatfield and Fernandes, 2009), and tend to drive at higher speeds (Boyce and Geller, 2002).Vehicle speed is commonly seen as the most important determinant of crash risk (Salusjrvi, 1981), and crash severity ( Joksch, 1993 and OECD/ECMT, 2006 ). Driving at higher speeds leaves less time to respond to unexpected events and increases stopping time, thus decreasing the possibility to avoid accidents (Aarts and van Schagen, 2006). So, reducing the driving speed of young adults, and in particular the amount of time spent above the speed limit, holds the potential of dramatically reducing accidents, and saving lives. The current study aimed to test whether a new type of car insurance (Pay-As-You-Drive vehicle insurance) is effective in reducing speed violations of young adults. Before explaining the study setup, we will discuss the (dis-)advantages of Pay-As-You-Drive vehicle insurance as compared to current road safety interventions.2. Method2.1. ParticipantsIn January 2007, we sent a letter with the request to fill out an online questionnaire on PAYD (duration 20 min) to approximately 6000 policyholders from five Dutch insurance companies (all policyholders were younger than 30 years). By filling out the questionnaire, participants were eligible to win a car navigation system. By July 2007, 706 policyholders had completed the online questionnaire. In the last part of the questionnaire, participants indicated whether they were interested in participating in a field experiment on PAYD. A substantial proportion of the people who filled out the questionnaire (n = 337) volunteered. However, because not all volunteers could be accommodated in the field experiment, a random selection of 228 people were eventually invited to participate.In the time between the participant selection and the installation and testing of the GPS device, some participants withdrew. Additionally, some participants were excluded from the analyses due to missing data (e.g. their GPS device did not work properly or was removed before the end of the experiment). By the end of the experiment (July 2008), we could establish the prevalence of speeding behavior throughout all phases of the experiment for 141 participants. The participants who dropped out (n = 87) did not differ in age, car size, income, gender, and intention to speed from the final set of participants (n = 141), but did report a somewhat lower yearly mileage. Self-reported mean yearly mileage of the final set of participants was 13,985 km, which is comparable to the 2007 Dutch average of 13,877 km (Centraal Bureau voor de Statistiek, 2009a). In July 2007, the mean age of participants was 24.4 years (SD = 2.2), and on average participants had possessed a drivers license for 4.5 years (SD = 2.4). Men (60%) were slightly overrepresented in the sample. The majority of participants worked fulltime (69.3%), and 30.7% were otherwise engaged (studying, part-time employment). Net monthly income levels were somewhat higher than the average income of Dutch young adults, with 25.7% of the sample earning less than 1000 Euros net per month, 46.2% earning between 1000 Euros to 1500 Euros net per month, 24.3% earning between 1500 and 2000 Euros net per month, and 3.7% earning more than 2000 Euros net per month (Centraal Bureau voor de Statistiek, 2009b).2.2. Design and procedureThe field experiment ran from November 2007 until June 2008. Prior to the start of the field experiment, participants cars were equipped with GPS devices. These GPS devices allowed for the monitoring of where participants were driving, what speed they were driving at and at what time, and how many kilometers they were travelling. Participants were randomly assigned to be either in incentive or control groups. The experiment comprised four phases, pre-measurement, intervention phase 1, intervention phase 2, and post-measurement, each phase lasting two months. During pre-measurement (NovemberDecember 2008) and post-measurement (MayJune 2008), participants driving behavior was monitored, but had no financial consequences. During the intervention phases, participants in the incentive group could earn a reward for adapting their driving behavior (a discount on their insurance premium of maximally 50 Euros per month), and received feedback on their driving behavior. Participants in the control group did not receive an incentive and feedback. The intervention period was divided into two phases, each lasting two months (intervention phase 1: JanuaryFebruary 2008, intervention phase 2: AprilMay 2008). This setup allowed us to compare driving behavior during four phases of equal length in the analyses (pre-measurement, intervention phase 1, intervention phase 2, and post-measurement).The monthly 50 Euros discount was divided into three components: 30 Euros was designated as a reward for keeping the speed limit,1 15 Euros for reduced mileage, and 5 Euros for avoiding driving during weekend nighttime hours.2 So, the maximum discount participants could earn during the two intervention phases was 200 Euros. Based on loss aversion (Tversky and Kahneman, 1981), we reasoned that financial losses should have more impact on behavior than equivalent financial gains. So communicating the reward drivers are foregoing (losing) by driving unsafely, may be a more effective frame than communicating the reward drivers are receiving (gaining) by driving safely. Accordingly, in the intervention phases, we used two ways to communicate how much of discount participants in the incentive group had earned. Participants in the gain framed incentive group (n = 50) were told that they started each month with 0 Euros, and could gain up to 50 Euros discount if they drove safely (which entailed keeping the speed limit, avoiding driving during on weekend nighttime hours, and reducing mileage). Conversely, participants in the loss framed incentive group (n = 50) were told that they started each month with 50 Euros discount, and could lose up to 50 Euros a month if they failed to drive safely.Participants in the incentive groups could track their performance during the intervention phases (from January to April 2008) by logging in to a personalized website (www.savedriver.nl). The website provided detailed feedback on speed violations, mileage, and nighttime driving, and showed, by default, how much discount the particular participant would receive at the end of the month if they continued the driving behavior of the previous week. Specifically, participants in the gain framed incentive group saw a graph that denoted how much discount they would gain, and participants in the loss framed incentive group saw a graph that denoted how much discount they would lose.Payments were made via bank transfer. The first of four monthly payments were made in March of 2008. Subsequent monthly payments were made in March, April and May 2008. A control group was used to assess the impact of extraneous variables on driving behavior. Right before the start of the first intervention period, participants in the control (n = 41) group were told that they would always receive the maximum discount of 200 Euros at the end of the experiment, irrespective of their driving behavior. They did not receive feedback on their driving behavior, and were paid 200 Euros during post-measurement.2.3. Data analysisDuring the field experiment, in-vehicle GPS devices logged X and Y coordinates every 100 m. On some occasions (e.g. in case of a loss of GPS signal, for instance due to driving through a tunnel), the GPS device logged on longer intervals (that is, less often than every 100 m). These observations (which amount to 23.4% of the total distance travelled) were removed from the analyses, as vehicle speed for these longer intervals could not be reliably determined. The occurrence of speed violations was established by matching vehicle location and vehicle speed to a map with local speed limits. Data was recorded on all types of roads, but analyses were limited to the most common road types in the Netherlands, with speed limits of 30, 50, 80, 100 and 120 km/h, respectively, and to observations for which local speed limit information was available (98.9%). Many Dutch roads tend to be congested around peak hours on weekday. During congestion, participants would have to deal with other vehicles ahead of them, which would prevent them from freely choosing their speed. This means that speed choice is more likely to be under volitional control during off-peak hours than during peak hours.3 Driving behavior recorded on peak hours (from 7 to 9 am and from 4 to 7 pm on weekdays) was therefore excluded from our analyses.The percentage of total distance travelled at 6% or more above the local speed limit (percentage of speeding) was used as the main dependent variable in our analyses. Violations of less than 6% were excluded from this definition, based on the assumption that such violations are often non-volitional and therefore most likely not influenced by a financial incentive and feedback. Furthermore, small thresholds are quite common in speed enforcement. Dutch police, for instance, only issue speeding tickets at 80 km/h roads for violations of 7 km/h or more over the speed limit (Goldenbeld and van Schagen, 2005).For each of the four phases of the field experiment, the percentage of speeding was calculated based on the weekly distance travelled at a speed 6% over the speed limit across all five road types (30, 50, 80, 100 and 120 km/h). Percentage speeding of the five road types were weighted by distance, and averaged across road types. The resulting variable represents an indicator of volitional speeding across all road types, and was used as a dependent variable in our analysis of the effect of PAYD on overall speeding. During off-peak periods, local speed limits were violated by 6% or more during 19.1% of the distance travelled. This figure is in line with previous literature that used GPS devices to measure vehicle speeds (Hultkrantz and Lindberg, 2003). Because the variable percentage of speeding was not normally distributed the natural logarithm of the scores was used in the analyses.3.ConclusionsAs noted before, the awareness of being monitored may initially have reduced speeding, but, as time passed, this awareness faded, and speeding returned to its natural level. As such, we reason that vehicle speeds during post-measurement are a better reflection of the speed at which participants would normally drive than speed levels during pre-measurement, where the GPS device had been recently installed and might have deterred participants from driving at their regular speedsFollowing this logic, the best possible estimation of the effect of PAYD on speeding might be reflected in the difference in speeding of the incentive group during the second intervention phase (March and April 2008) and after the incentive was taken away (post-measurement, May and June 2008). In other words, the financial incentive appears to have decreased the percentage of distance travelled during which drivers violated the speed limits by 6% or more from 20.5% to 17.6%. To put it differently, PAYD is estimated to have reduced volitional speeding by 14%, which roughly amounts to a reduction of 39.2 h per year for the average young driver.4 Although these numbers might not appear impressive, one should note that these numbers reflect reductions in actual speeding across all roads participants drove at. As such, the safety gains from PAYD could add significantly to those already made through the use of more conventional speed enforcement tools (e.g. speed cameras and mobile radars), especially as the effects of such tools on driving speed are typically limited to areas where enforcement is active. Also, small differences in driving speed are associated with large differences in crash risk (Elvik, 2006). A reduction of speeding by 5% may lead to as much as a 20% decrease of fatalities in road accidents (OECD/ECMT, 2006 OECD/ECMT, Speed Management, OECD Publishing, Paris (2006). OECD/ECMT, 2006). Thus, we conclude that PAYD leads to modest, but relevant reductions in the driving speed of young adults, and thus may prove a valuable tool in reducing the crash risk of young drivers.翻译一：付费驱动车险理赔对年轻驾驶员车速选择的影响1. 引言在世界范围内跨越所有年龄段的道路交通事故中，每年约有1.27百万人的死亡记录（世界卫生组织，2004年）。紧跟其后的是由艾滋病导致死亡的，年龄段大多在20至24岁之间，而道路交通事故统计数据中显示年轻司机所占比率最高。在荷兰，青壮年（18-24岁）的驾驶人口为8，但所占所有严重交通事故比率为22，这意味着每公里的行驶，在荷兰的年轻成年人身上发生交通事故的概率是在其他年龄段的5.5倍。涉及意外的差异往往是由于青壮年的鲜明的个性特点，例如，年轻司机相对与老司机来说，具有更积极冒险的态度，更愿意承担风险和在危险驾驶中表现出更强的动机，并且喜欢高速驾驶。高速驾驶速通常被视为导致交通事故的最重要因素，且和车辆碰撞的损坏严重程度有关，因为在高速驾驶中驾驶员有很短的时间来应对处理突发事件和减少伤害。所以使年轻驾驶员降低行驶速度和规定车速上限将有利于保障人民生命安全和减少交通事故的发生。目前这项研究的目的是测试的新型汽车保险（付费驱动的车辆保险）是否是有效地减少青壮年驾驶速度，从而减少交通事故发生。2. 研究方法2.1参与者2007年1月，我们发出一封信给大约6000名来自五个荷兰保险公司（所有投保人的年龄不超过30岁）的投保人，要求他们填写一份关于“付费驱动车辆保险”的网上问卷（持续时间为20分钟）。通过填写问卷参加者将有机会赢得汽车导航系统。 2007年7月，已有706投保人完成了这份网上问卷。在问卷的最后一部分的调查项目是投保人是否有兴趣参与对付费驱动的车辆保险的田间试验。填写问卷的投保人大部分都表示愿意（N= 337），然而不是所有的志愿者都可以参加这次付费驱动的车辆保险田间试验，随机选择了228人最终被邀请参加在参与者的选择和GPS设备的安装和测试之间的时间内，有些参与者会退出试验，此外，一些参与者因弄丢失数据（如自己的GPS设备没有正常工作或实验结束前被删除）也会被排除出分析。到2008年7月实验结束时，我们可以建立整个实验阶段的超速行为的统计分析（141人参加，87人退出），分析因数有车大小（无差异），年龄，收入，性别，并将记录从参与者的最后一组速度，但不报告每年有所降低的行驶里程。每年自我报告的参与者最后平均里程为13,985公里。在2007年7月，荷兰参与者的平均年龄为24.4岁（SD= 2.2），平均参与者拥有驾驶执照为4.5年（SD= 2.4），男性（60），轻微的样本中所占比例过高。参与者多数是全职工作（占69.3），其余30.7有其他要事（学习，兼职）。净月收入水平较之荷兰年轻的成年人的平均收入有所增加，有25.7的样本收入低于1000欧元，月收入在1000欧元至1500欧元占46.2，24.3的样本收入在1500欧元至2000欧元，3.7的收入超过2000欧元。2.2程序与设计2007年11月至2008年6月的举行现场试跑实验，田间试验开始之前车辆都安装“汽车配备GPS设备，这些GPS设备对参加者驾驶进行监测，监测项目速度驾驶速度和当时时间以及驾驶路程。受试者是被随机分配成激励组或对照组，实验包括三个阶段，预测，干预分为干预第1阶段和第2阶段，每个阶段为期两个月。测量前（2007年11月-12月）和测量后（2008年5月）期间，举办方对参加者的驾驶行为进行了监测，但不涉及有关参赛者的财务状况。在介入阶段，激励组的参与者可以通过调整他们的驾驶行为来赚取有关奖励（奖励是对他们每月最大限度为50欧元的保险费折扣），并收到反馈的驾驶行为。而在在对照组的参加者没有得到激励和反馈。干预期间分为两个阶段，每个阶段为期两个月（干预1阶段：2008年1-2月，干预第2阶段：2008年4月至五月），这样的设置能使我们能够在长度相等的四个阶段分析（预测，干预，干预第2阶段第1阶段，并测量后）参与者的驾驶行为。每月50欧元的保险率折扣的奖励可通过三个部分获得：参与者保持驾驶速度在指定限速内可获得30欧元的奖励；减少驾驶里程可获得15欧元的奖励；周末夜间少于2小时的驾驶里程则可获得5欧元奖励，参与者可在两个干预阶段当中获得的最高优惠为200欧元。根据厌恶损失的理论基础，我们的理由是金融的损失应该会对驾驶行为产生更多的影响。所以对上述情况中没有获得奖励的对照组驾驶时可能不会按要求来理性驾驶，相比之下能获得奖励的激励组则按要求来安全驾驶。因此在干预阶段，我们使用两种奖励方式来进行对比观察，在激励组总共有多少人赢得奖励，在增益框激励组的参与者（N =50）告诉记者，从开始每月0欧元的奖励到如果他们安全驾驶（其中包括保持车速限制，限制上周末夜间驾驶时间，减少驾驶里程）可以获得高达50欧元的折扣，相反造成损失陷害激励组的参与者（N =50）告诉记者如果他们驾驶不按照规定安全驾驶则将造成每月50欧元的损失，即从每个月50欧元的奖励降为0欧元。在干预阶段（1月至2008年4月），可以登录到个性化的网站（www.savedriver.nl）对激励群体的参与者的具体表现进行踪登录到，该网站提供了显示详细的反馈速度侵犯，里程，夜间驾驶，默认情况下如果他们继续前一周的驾驶行为的参与者将在本月底他们赢得的奖励。具体来说，在增益框激励组的参与者看到的图形，是表示他们将获得多少奖励，而损失陷害激励组的参与者看到了一个图表示的是他们将失去多少奖励。所有奖励都是通过银行转账付款的，在2008年3月一次完成了四个月的按月支付，其后的将按每月付款。对照组是用来评估无关变量对驾驶行为的影响，第一干预期间开始前，在对照组中的41名参与者被告知，不论他们的驾驶行为如何他们将在实验结束时收到总额为200欧元的最高回报，他们没有得到他们的驾驶行为的反馈，并在测量后支付200欧元。2.3 数据分析在田间试验中，车载GPS设备记录的有X和Y坐标每单位为100米。在某些特殊情况下（例如，在GPS信号的损失情况下或则当驾车通过隧道），GPS设备记录较长的时间间隔（往往长于100米）。这些意见（数额为23.4的总距离前往）被拆除的分析，这些较长的时间间隔的车辆速度不能可靠地确定。建立与当地的车速限制的地图匹配的车辆的位置和车辆速度速度侵权行为的发生。所有类型的道路上，但数据记录分析在荷兰的最常见的道路类型是有限的，30，50，80，100和120公里/小时的车速限制，分别和意见，为当地的车速限制信息（98.9）。许多荷兰的道路往往是平日围绕高峰时段拥挤的。在拥塞期间，与会者将不得不处理与其他车辆在他们之前，这将阻止他们自由选择自己的速度。这意味着，速度的选择是更有可能比在繁忙hours.3意志控制下，在非高峰时段的驾驶行为记录高峰时段（7日上午9时和下午4至7平日），因此我们排除分析总距离的百分比前往当地的车速限制（“超速百分比”）的6以上或更多，在我们的分析为主要因变量。违反低于6被排除在这个定义基础上的假设，这种违法行为往往非意志，因此最有可能不的金融激励和反馈的影响。此外，小阈值速度执法的现象相当普遍。例如，荷兰警方，唯一的问题侵犯7公里/小时或以上的车速限制“（Goldenbeld和van Schagen，2005年）在80公里/小时的道路，超速行驶的罚单。对于每一个四个阶段的田间试验，超速百分比计算的基础上每周一次的距离在所有5个道路类型（30，50，80，100和120公里/小时的车速限制超过6的速度游）。的五个类型的道路超速百分比加权的距离，跨越道路类型平均。由此产生的变量代表的所有道路类型的意志超速的指标，并作为因变量在我们的整体超速驾驶，PAYD效果分析。在非高峰期，当地的车速限制在19.1的距离违反，乘坐6以上。这个数字是与以前的文献，使用GPS设备来测量车辆速度（Hultkrantz和林德伯格，2003年）。因为超速驾驶的变量的百分比是不是正态分布的自然对数的分数是在分析中使用3结论如前所述被监视的认识最初目的是建立减少超速的意识，但是随着时间的推移，这种意识的逐渐褪色，并加快恢复其原本“自然”的交通驾驶意识水平，由于这些原因在测量后参与者的驾驶速度通常会比测量前速度有所降低，而最近已安装GPS设备则并有可能阻止他们驾驶的参与者更好地反映常规的速度。按照这一逻辑，超速驾驶付费驱动的车辆保险效果最好的估计，在加快在第二次干预阶段（3月和2008年4月）和激励被带走（后测量后的激励组的差异，可能反映5月和2008年6月）。换言之，财政激励似乎有所减少距离的百分比前往在此期间，司机违反了6或以上的时速限制，从20.5到17.6。要换种方式，付费驱动的车辆保险估计减少了14，大致相当于减少了39.2的H平均年轻驾驶员.虽然这些数字可能不会出现令人印象深刻的意志超速，要注意的是，这些数字反映减少所有道路参与者在实际超速驾驶。因此，从付费驱动的车辆保险安全收益显着增加，尤其是行驶速度等工具的影响通过使用更多的常规速度的执法手段（如高速照相机和移动雷达）已经取得的，通常是有限的地方执法处于活动状态。此外，在行驶速度小的差异与崩溃的风险较大的差异（Elvik2006年）。减少5的超速驾驶可能导致的死亡道路交通事故减少20（经合组织/欧洲运输部长会议，2006年）。因此，我们认为付费驱动的车辆保险是一项能有效减少年轻成年人的驾驶速度从而减少交通事故发生有价值的方法。文献二： A study of safety in road transport in Hong Kong1.IntroductionRoad safety has become a major problem in modern and industr