Why Context Matters: The Influence of Application Domain on Preferred Degree of Anthropomorphism and Gender Attribution in Human–Robot Interaction

Abstract

The application of anthropomorphic design features is widely believed to facilitate human–robot interaction. However, the preference for robots’ anthropomorphism is highly context sensitive, as different application domains induce different expectations towards robots. In this study the influence of application domain on the preferred degree of anthropomorphism is examined. Moreover, as anthropomorphic design can reinforce existing gender stereotypes of different work domains, gender associations were investigated. Therefore, participants received different context descriptions and subsequently selected and named one robot out of differently anthropomorphic robots in an online survey. The results indicate that lower degrees of anthropomorphism are preferred in the industrial domain and higher degrees of anthropomorphism in the social domain, whereas no clear preference was found in the service domain. Unexpectedly, mainly functional names were ascribed to the robots and if human names were chosen, male names were given more frequently than female names even in the social domain. The results support the assumption that the preferred degree of anthropomorphism depends on the context. Hence, the sociability of a domain might determine to what extent anthropomorphic design features are suitable. Furthermore, the results indicate that robots are overall associated more functional, than gendered (and if gendered then masculine). Therefore, the design features of robots should enhance functionalities, rather than specific gendered anthropomorphic attributes to avoid stereotypes and not further reinforce the association of masculinity and technology.

Robot's Gendering Trouble: A Scoping Review of Gendering Humanoid Robots and its Effects on HRI

“…However, it might be interesting to explore more qualitative and indirect ways to measure the robot's genderedness so as to circumvent the described conundrum. For instance, Roesler et al [72] used naming frequency to understand how the robots in their study were attributed a gender, which gave participants the possibility not just to give robots traditional names, but also technical and more object-oriented ones. As such, we put forward the following guideline: Guideline 9: Explore more subtle and less biasing ways of checking whether gender is attributed to the robot and even relevant for participants, for instance, through qualitative and indirect measures.…”

Artificial intelligence in marketing: A meta‐analytic review

“…Anthropomorphism refers to imbuing human‐like traits, attributes, motives, intentions, or emotions into nonhuman agents, such as products, devices, robots, and service machines (Epley et al, 2007; Fan et al, 2020; Roesler et al, 2022). When people interact with such inanimate life‐like technologies, they engage in social cognitive processes, which further helps them understand the technology better, and therefore, enhances their adoption of these technologies (Yoganathan et al, 2021).…”

“…Anthropomorphism refers to imbuing human-like traits, attributes, motives, intentions, or emotions into nonhuman agents, such as products, devices, robots, and service machines (Epley et al, 2007;Fan et al, 2020;Roesler et al, 2022). When people interact with such inanimate life-like…”

Robot Voices in Daily Life: Vocal Human-Likeness and Application Context as Determinants of User Acceptance

“…A few empirical studies have so far addressed potential interaction effects of anthropomorphism and application context. In Roesler and colleagues' recent experiment (Roesler et al, 2022), participants had to choose one out of various robot pictures that differed in visual human-likeness based on different context descriptions. A lower degree of humanlikeness was found to be preferred for industrial application and a higher degree of human-likeness for social application, while there were no clear preferences in the service domain.…”