Supplementary material for paper: "Passengers Perceive more Uncertainty in Fully Automated Cars than in Manually Driven cars : An Online Crowd-sourced Animation study under Mixed Traffic Conditions"

<strong>Abstract</strong>Automated cars have the potential to improve road safety and reduce traffic congestion. Not all objective benefits are perceived this way by society. Passengers may feel uncertain about the behavior, reliability, and intentions of fully automated cars, which can affect their overall experience. This study investigates the amount of psychological uncertainty that passengers perceive in fully automated and in manually driven cars. Using a crowd-sourced approach, participants reviewed scenarios in which the only difference was whether they were passenger in a fully automated car or in a manually driven car. Participants rated their perceived uncertainty after each scenario. The results show that passengers generally perceive higher psychological uncertainty in fully automated cars. The passenger's trust in the capabilities of a fully automated car seems to influence the psychological uncertainty that they perceive. Furthermore, the psychological uncertainty scores for different scenarios indicate a link between perceived risk and psychological uncertainty.<br><strong>Procedure &amp; Participants</strong>To study perceived psychological uncertainty, we chose to let the participants use their own understanding of what they perceive as the feeling of uncertainty, in order to measure it in its purest form as it would be experienced by passengers under those conditions. We clearly defined the action that the passenger could be uncertain about, by asking: "At the end of the scenario, the fully automated car (/the driver of your car) changes to the left lane. How uncertain would you feel?". Perceived psychological uncertainty was measured using a subjective rating scale. After viewing each scenario, participants were instructed to use a slider scale ranging from "Absolutely not uncertain" to "Absolutely uncertain", but were not shown numerical values to minimize anchoring bias. The slider ranged from 0 ("Absolutely not uncertain") to 100 ("Absolutely uncertain"), keeping track of a numerical value in the back end. To understand and control individual differences, participants were also asked to rate their attitudes toward fully automated cars at the beginning of the experiment and to give an indication of how much day-to-day uncertainty they think they experience. This additional data helped assess whether people with a prejudice towards fully automated cars were more likely to feel uncertain about being a passenger in a fully automated car.<br>A total of 1437 people (Valid N) (M = 705, F = 718, Other = 8, Prefer not to respond = 6) from 84 countries participated in the study. We allowed contributors from all countries to participate except Venezuela (after a ban), as there were serious cheating incidents as identified by the researchers. All participants were 18 years or older (M = 35.55, STD = 11.00) and digitally provided their informed consent prior to participation. Participants were recruited through the crowd-sourcing service Appen (https://appen.com) and were later redirected to Heroku (https://www.heroku.com) to review the stimuli.

# 摘要
自动驾驶汽车（Automated Cars）有望提升道路交通安全并缓解交通拥堵。但社会大众并未完全认可其客观优势。乘客可能会对完全自动驾驶汽车（Fully Automated Cars）的行驶行为、可靠性与意图产生不确定感，这会影响其整体乘车体验。本研究旨在探究乘客在完全自动驾驶汽车与人工驾驶汽车中感受到的心理不确定程度。研究采用众包方式，让参与者观看仅在乘车载体（完全自动驾驶汽车/人工驾驶汽车）上存在差异的场景，并在每个场景结束后对自身感知到的不确定程度进行评分。结果显示，乘客在完全自动驾驶汽车中普遍感受到更高的心理不确定感；乘客对完全自动驾驶汽车性能的信任程度，似乎会影响其感知到的心理不确定水平。此外，不同场景下的心理不确定评分表明，感知风险与心理不确定程度之间存在关联。

# 实验流程与参与者
为研究感知到的心理不确定感，我们选择让参与者基于自身对不确定感受的理解进行评分，以期以最贴近真实乘车场景的方式测量该感受。我们明确界定了乘客可能产生不确定感的具体行为："在场景结束时，完全自动驾驶汽车（/您所乘车辆的驾驶员）变更至左侧车道。您会感受到多大程度的不确定？"。感知心理不确定感采用主观评分量表进行测量。每个场景观看完毕后，要求参与者使用从"完全无不确定感"到"极度不确定"的滑块评分，但不显示具体数值以尽可能降低锚定偏差。该滑块的数值范围为0（"完全无不确定感"）至100（"极度不确定"），后台会记录对应的数值。为探究并控制个体差异，实验开始时还要求参与者对完全自动驾驶汽车的态度进行评分，并说明自身日常感受到的不确定程度。这些额外数据有助于评估对完全自动驾驶汽车持有偏见的人群，是否更易在乘车时产生不确定感。

本研究共招募到来自84个国家的1437名有效参与者（有效样本量N），其中男性705人、女性718人、其他性别8人，6人选择不披露性别。研究排除了委内瑞拉的参与者（因研究者发现存在严重的作弊行为），其余所有国家的贡献者均可参与。所有参与者均年满18周岁（平均年龄35.55岁，标准差11.00），并在参与前通过电子方式签署了知情同意书。参与者通过众包服务平台Appen（https://appen.com）招募，随后跳转至Heroku平台（https://www.heroku.com）观看实验刺激场景。