Assessing the Behavioural Intention of Individuals to Use an AI Doctor at the Primary, Secondary, and Tertiary Care Levels

The research has been designed to fit into the framework of AI doctors and used previously validated measurements. Social influence (SI), was derived from Teo (Teo, 2009) and Park et al. (Park, Baek, Ohm, & Chang, 2014). Perceived task technology ft (PTTF) was adapted from Lu and Yang (Lu & Yang, 2014). Performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), high-tech lifestyle (HTLS), behavioural intention to use AI doctors primary level (Primary BI), behavioural intention to use AI doctors secondary level (Secondary BI), behavioural intention to use AI doctors tertiary level (Tertiary BI) which were adapted from Venkatesh et.al (Venkatesh et al., 2012). The last two indicators of performance expectancy “AI will provide up-to-date health info that I need”, and “AI will provide new instructional content that I need” were adapted from Lee and Letho (D. Y. Lee & Lehto, 2013). Privacy concern (PC) was adapted from Martins et al. (Martins, Oliveira, & Popovič, 2014). A five-point Likert scale on a rating of 1 (strongly disagree) to 5 (strongly agree) was used to evaluate each indicator.
People who knew more about how AI technology was used in daily life, and believed AI technology to be a significant improvement in healthcare were more at ease with the medical use of AI (Armero et al., 2022). Participants with existing health data collecting experiences with devices, such as a smartwatch, band, bracelet, glasses, or fitness trackers have been chosen as the sample. These high-tech devices can instantly measure and report many things about individuals. These participants have been deliberately chosen because they can get a clear picture of their health, including their real-time heart rate, blood pressure, sleep time and quality, activity tracker, body mass index, and calories spent. Since many features of these people have become measurable, they have been asked how they would like this data to be used by AI technology.

The data has been collected from 432 participants.

本研究旨在贴合AI医生（AI doctors）的研究框架，并采用经先前验证的测量工具。社会影响（Social Influence, SI）的量表改编自Teo（Teo, 2009）与Park等人（Park, Baek, Ohm, & Chang, 2014）的研究。感知任务技术适配（Perceived Task Technology Fit, PTTF）改编自Lu与Yang（Lu & Yang, 2014）的研究。绩效期望（Performance Expectancy, PE）、努力期望（Effort Expectancy, EE）、社会影响（Social Influence, SI）、促成条件（Facilitating Conditions, FC）、高科技生活方式（High-Tech Lifestyle, HTLS）、使用AI医生的一级行为意向（Primary BI）、二级行为意向（Secondary BI）与三级行为意向（Tertiary BI）均改编自Venkatesh等人（Venkatesh et al., 2012）的研究。绩效期望的最后两个题项“AI将提供我所需的最新健康信息”与“AI将提供我所需的新型指导内容”改编自Lee与Letho（D. Y. Lee & Lehto, 2013）的研究。隐私顾虑（Privacy Concern, PC）改编自Martins等人（Martins, Oliveira, & Popovič, 2014）的研究。本研究采用1（完全不同意）至5（完全同意）的李克特五点量表对各题项进行评估。

过往对人工智能技术在日常生活中的应用具备了解，且认为人工智能技术是医疗保健领域的重大改进的人群，对人工智能在医疗场景中的应用更为安心（Armero et al., 2022）。本研究的样本选取为曾有使用智能手表、智能手环、智能手镯、智能眼镜或健身追踪器等设备收集健康数据经历的参与者。这类高科技设备可即时测量并报告个体的多项健康指标。选取该类参与者是因为他们能够清晰掌握自身健康状况，包括实时心率、血压、睡眠时长与质量、活动追踪数据、身体质量指数以及消耗热量。由于这类人群的多项健康指标均可被量化测量，本研究询问了他们对于人工智能技术如何使用这些数据的态度。

本研究共收集到432份参与者数据。