• Developing Better Curb Management Strategies through Understanding Commercial Vehicle Driver Parking Behavior in a Simulated Environment

Three different data types were obtained from Oregon State Driving and Bicycling Simulator Laboratory for purpose of this report and they are as follow: 1) Speed data consists of subject number, average speed, minimum speed, and all the independent variables. Speed data were collected based on the truck’s speed while driving through a certain scenario (out of 24). For each scenario, the average and minimum speed (mph) of 12 drivers were recorded along each segment (scenario) from the start of the road to 150 feet before the intersection (traffic signals). 2) Eye tracking data consists of subject number, total fixation duration (TFD) in milliseconds, area of interest (AOI), and all the independent variables. TFD data were collected while the truck driver maneuvers through a certain scenario (out of 24). For each scenario, the TFD for each AOI was recorded for 11 subjects along each segment (scenario) from the start of the road to 150 feet before the intersection (traffic signals). AOI represent the area of interest that a driver fixates for a certain of time to generate the total fixation duration. 3) Eye tracking data consists of subject number, GSR in peaks per minute, and all the independent variables. GSR data were collected while the truck driver maneuvers through a certain scenario (1 out of 24). For each scenario, the peaks per minute data was recorded for 11 subjects along each segment (scenario) from the start of the road to 150 feet before the intersection (traffic signals). Peaks per minute represents the emotional arousal (i.e., something is scary, threating, joyful, etc.) that a driver generates when reacting to a particular event. Fourteen participants were recruited, two of them had a simulator sickness so they were excluded from the data and the analysis. While there are no quality or consistency issues with this data set, it should be noted that the sample is on the smaller side and that should be considered when interpreting derived results. The average values were calculated to apply robust statistical analysis for such data (speed and lateral position). As the experiment consists of 2x2x2x3 factorial design, each participant had to driver through 24 scenarios; therefore, 288 scenario observations were obtained and recorded in the excel file.

本报告所使用的全部数据均采集自俄勒冈州立大学驾驶与自行车模拟实验室（Oregon State Driving and Bicycling Simulator Laboratory），共涵盖三类不同数据类型，具体如下：
1) 速度数据：包含受试者编号、平均车速、最低车速及所有自变量。该数据采集自卡车通过24个预设场景之一时的行驶车速。针对每个场景，我们沿道路起点至交叉口（交通信号灯）前150英尺的路段，记录了12名驾驶员的平均车速与最低车速，单位为英里每小时（mph）。
2) 眼动追踪数据（Eye tracking data）：包含受试者编号、总注视时长（total fixation duration, TFD，单位：毫秒）、感兴趣区（area of interest, AOI）及所有自变量。该数据采集自卡车驾驶员通过24个预设场景之一时的眼动指标。针对每个场景，我们沿道路起点至交叉口（交通信号灯）前150英尺的路段，为11名驾驶员的每个感兴趣区记录了对应的总注视时长。其中，感兴趣区指驾驶员在特定时长内进行注视以生成总注视时长的区域。
3) 眼动追踪数据（Eye tracking data）：包含受试者编号、每分钟峰次的皮肤电反应（Galvanic Skin Response, GSR）及所有自变量。该数据采集自卡车驾驶员通过24个预设场景之一时的生理指标。针对每个场景，我们沿道路起点至交叉口（交通信号灯）前150英尺的路段，为11名驾驶员记录了每分钟峰次数据。每分钟峰次代表驾驶员针对特定事件做出反应时产生的情绪唤醒水平，例如恐惧、威胁、愉悦等。
本次研究共招募14名受试者，其中2人出现模拟器不适症状，故被排除在数据采集与分析流程之外。尽管该数据集不存在质量或一致性问题，但需注意其样本量偏小，在解读衍生结果时应充分考虑这一局限。针对速度与横向位置这类数据，我们通过计算平均值以开展稳健的统计分析。本实验采用2×2×2×3析因设计（factorial design），每名受试者需完成24个场景的驾驶任务，因此共计获取288条场景观测数据，并存储于Excel文件中。