背景:COVID-19大流行引起了全球技术使用的变化,社会和经济。这场大流行危机带来了其他措施,如接触者追踪应用程序(CTA),以帮助对抗病毒传播。不幸的是,这些应用程序的低采用率影响了它们的成功。采用率低的原因可能有很多,包括对安全和隐私的关注,以及报告的CTA信任问题。一些担忧与CTA如何用作监视工具或涉及健康数据时对隐私的潜在威胁有关。例如,在爱沙尼亚,名为HOIA的CTA在2021年1月中旬有大约25万次下载。然而,2021年,只有4.7%的人口使用HOIA作为COVID-19CTA。采用率低的原因包括缺乏能力,隐私和安全问题。尽管欧盟在建立道德和可信赖的基于人工智能(AI)的应用程序方面做出了努力,但这种较低的采用率和缺乏可信赖性仍然存在。
目的:本研究的目的是了解如何衡量对健康技术的信任。具体来说,我们评估了人类-计算机信任量表(HCTS)用于衡量爱沙尼亚人对HOIA应用程序的信任以及缺乏信任的原因。
方法:主要研究问题是:HCTS可以用来评估公民对健康技术的信任感吗?我们建立了四个假设,并通过调查进行了测试。我们使用了一个方便的样本来收集数据,包括在社交网站上分享问卷,并使用滚雪球方法接触爱沙尼亚人口中所有潜在的HOIA用户。
结果:在78名受访者中,61已经下载了HOIA应用程序,其中包含有关使用模式的数据。然而,下载该应用程序的人中有20人承认,尽管大多数人声称经常使用移动应用程序,但它从未打开过。主要原因包括不了解它是如何工作的,隐私和安全问题。在以下三个属性方面,参与者对CTA的总体信任与他们对HOIA应用程序的感知信任之间存在显着相关性:能力(P<.001),风险感知(P<.001),和互惠(P=0.01)。
结论:这项研究表明,爱沙尼亚居民对HOIA应用程序的信任确实会影响他们使用该应用程序的倾向。参与者通常不认为HOIA可以帮助控制病毒的传播。这项工作的结果仅限于使用类似联系人追踪方法的HOIA和健康应用程序。然而,这些发现有助于获得更广泛的理解和意识到设计值得信赖的技术的必要性。此外,这项工作可以帮助提供设计建议,确保CTA的可信度,以及人工智能使用高度敏感数据和服务社会的能力。
BACKGROUND: The COVID-19 pandemic has caused changes in technology use worldwide, both socially and economically. This pandemic crisis has brought additional measures such as contact-tracing apps (CTAs) to help fight against spread of the virus. Unfortunately, the low adoption rate of these apps affected their success. There could be many reasons for the low adoption, including concerns of security and privacy, along with reported issues of trust in CTAs. Some concerns are related with how CTAs could be used as surveillance tools or their potential threats to privacy as they involve health data. For example, in Estonia, the CTA named HOIA had approximately 250,000 downloads in the middle of January 2021. However, in 2021, only 4.7% of the population used HOIA as a COVID-19 CTA. The reasons for the low adoption include lack of competency, and privacy and security concerns. This lower adoption and the lack of trustworthiness persist despite efforts of the European Union in building ethics and trustworthy artificial intelligence (AI)-based apps.
OBJECTIVE: The aim of this study was to understand how to measure trust in health technologies. Specifically, we assessed the usefulness of the Human-Computer Trust Scale (HCTS) to measure Estonians\' trust in the HOIA app and the causes for this lack of trust.
METHODS: The main research question was: Can the HCTS be used to assess citizens\' perception of trust in health technologies? We established four hypotheses that were tested with a survey. We used a convenience sample for data collection, including sharing the questionnaire on social network sites and using the snowball method to reach all potential HOIA users in the Estonian population.
RESULTS: Among the 78 respondents, 61 had downloaded the HOIA app with data on usage patterns. However, 20 of those who downloaded the app admitted that it was never opened despite most claiming to regularly use mobile apps. The main reasons included not understanding how it works, and privacy and security concerns. Significant correlations were found between participants\' trust in CTAs in general and their perceived trust in the HOIA app regarding three attributes: competency (P<.001), risk perception (P<.001), and reciprocity (P=.01).
CONCLUSIONS: This study shows that trust in the HOIA app among Estonian residents did affect their predisposition to use the app. Participants did not generally believe that HOIA could help to control the spread of the virus. The result of this work is limited to HOIA and health apps that use similar contact-tracing methods. However, the findings can contribute to gaining a broader understanding and awareness of the need for designing trustworthy technologies. Moreover, this work can help to provide design recommendations that ensure trustworthiness in CTAs, and the ability of AI to use highly sensitive data and serve society.