Understanding Trust in Automation: A Consideration of Human Factors Context
Authors: Hannah Biermann, Sophia Otten, Martina Ziefle
Abstract: People interact with modern technology in various areas of life, both private and professional. One major advantage is the relief provided to users by technical devices that progressively take over tasks and perform them autonomously. Increasing system automation is often accompanied by uncertainties among users, expressed in concerns about paternalism through technology and loss of control. Trust can help to overcome perceived uncertainties and is thus a key driver for the acceptance and successful implementation of innovative technology. The study aim is to understand how trust formation “works”, what technology trust means in different contexts, which perceived uncertainties exist, and how individual usage decisions are made.We designed a two-step empirical research approach. Our aim was to explore what trust and distrust mean to different individuals and groups – in general (e.g., interpersonal trust) and in the context of technology (e.g., trust in automation) – and what influence different dimensions of trust and distrust may have on the acceptance of technology in different application fields. In a preliminary qualitative study, guided interviews (language: German) were conducted to identify conceptual facets of trust and distrust. In sum, 21 adults of different age groups (21 to 86 years) participated. The interviews were transcribed verbatim and analyzed using content analysis. Results were operationalized and measured in an online questionnaire survey (data collection is ongoing). To investigate influences of human factors on perceptions of trust and distrust, demographic data, personality traits, and attitudes were measured (e.g., risk perception, trust disposition, openness to innovation). To evaluate how differences in context might influence trust perceptions, the fields of autonomous driving and ambient assisted living were considered.Interviews: Trust was described as a positive gut feeling, difficult to put into words, and associated with affection, love, and friendship. Distrust was described as a negative, acrid feeling and associated with doubt, uncertainty, and deception. Experience (good/bad) and knowledge were identified as key factors that influence (dis)trust in general. Regarding technology, technical reliability and functionality as well as the perception of service operators influence the perceived trustworthiness. Trust and distrust seem to influence behavior in terms of reliability and control and depend on general risk perception and attitude (open/skeptical) toward others and things, respectively.Questionnaire: Factor and regression analyses will be conducted to form distinct dimensions for trust and distrust and to model their influence on technology acceptance in different contexts (mobility vs. health care). Cluster analysis will be conducted to examine the extent to which trust and distrust are perceived the same or differently among diverse study participants. Our results contribute to a deeper understanding of trust in automation across contexts and help scientific research measure (dis)trust constructs more precisely, taking into account individual associations and context-specific predictors. This also strengthens the knowledge of trust as a prerequisite for technology acceptance. The practical relevance of our study lies in the possibility to derive design and action recommendations for the user-centred development of technology innovations considering different target groups and application fields.
Keywords: trust in automation, user diversity, human, automation interaction
Cite this paper: