Facing Trust: How Anthropomorphism and Individual Differences Shape Perceptions of Robot Faces
Abstract
Trust is essential to the development and maintenance of human relationships, and perceptions of trust based on the face alone can occur in as little as 100ms. Humans also tend to make trust estimations of robots based on their physical design, and these assessments can lead to successful acceptance of the robot or result in rejection. As robots are being designed to integrate into our daily lives, examining the factors that contribute to robot trustworthiness is a crucial component to understanding Human-Robot Interactions (HRI). Anthropomorphism is the attribution of human-likeness to objects, and it is suggested that imbuing robots with human characteristics can facilitate perceptions of trustworthiness. This study investigated the association of anthropomorphism with perceptions of trustworthiness of robot faces and whether there is a potentially predictive role of the tendency to anthropomorphise that would mediate trustworthiness perceptions. 43 participants rated their impressions of 30 robot images on trustworthiness and human-likeness. Participants also completed the individual differences in anthropomorphism questionnaire (IDAQ; Waytz et al., 2010). We hypothesised that anthropomorphic ratings on the Godspeed Anthropomorphism questionnaire (Bartneck, Kulić, et al., 2009) would be associated with perceptions of trustworthiness on a single item measure of trustworthiness. We did not find support for this hypothesis (r(41) = .041, p = .795, N = 43). We hypothesised that individual differences in the tendency to anthropomorphise would be associated with human-likeness of robot faces. We found that IDAQ scores were not associated with anthropomorphism (Estimate = .219, S.E. = .286, C.R. = .766, p = .444). Finally, we hypothesised that individual differences would mediate the relationship between anthropomorphism and perceptions of trustworthiness of robot faces. We found that there was no significant relationship between IDAQ scores and trustworthiness perceptions (Estimate = -.005, S.E. = .135, C.R. = -.039, p = .969). These findings suggest that participants accessed prior negative knowledge about robots and applied them to the robot images in this study. Additionally, the results suggest that trust affordances cannot be detected simply in the physical features of a robot. Limitations include the experimental design, especially the one-item measure of trustworthiness and the lack of contextual cues.
Related
- Facing Empathy: Reducing Facial-Differences Bias Through a Story
- Respiration and individual differences shape behavioral and inter-brain synchronization
- Individual Differences in Dopaminergic Modulation of Exploration–Exploitation Behavior
- Why we hug: Individual differences predict emotional touch memories
- Subtle positive facial expression recognition is enhanced by comedy viewing: individual differences in empathy and depressive tendencies
BibTeX
@misc{fingal2026facing,
title = {{Facing Trust: How Anthropomorphism and Individual Differences Shape Perceptions of Robot Faces}},
author = {Lian Fingal and Samuel Bennett},
howpublished = {PsyArXiv},
year = {2026},
month = sep,
doi = {10.31234/osf.io/dzq28_v1},
url = {https://osf.io/preprints/psyarxiv/dzq28_v1/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Facing Trust: How Anthropomorphism and Individual Differences Shape Perceptions of Robot Faces
AU - Lian Fingal
AU - Samuel Bennett
PY - 2026
DA - 2026/09/20
PB - PsyArXiv
DO - 10.31234/osf.io/dzq28_v1
UR - https://osf.io/preprints/psyarxiv/dzq28_v1/
AB - Trust is essential to the development and maintenance of human relationships, and perceptions of trust based on the face alone can occur in as little as 100ms. Humans also tend to make trust estimations of robots based on their physical design, and these assessments can lead to successful acceptance of the robot or result in rejection. As robots are being designed to integrate into our daily lives, examining the factors that contribute to robot trustworthiness is a crucial component to understanding Human-Robot Interactions (HRI). Anthropomorphism is the attribution of human-likeness to objects, and it is suggested that imbuing robots with human characteristics can facilitate perceptions of trustworthiness. This study investigated the association of anthropomorphism with perceptions of trustworthiness of robot faces and whether there is a potentially predictive role of the tendency to anthropomorphise that would mediate trustworthiness perceptions. 43 participants rated their impressions of 30 robot images on trustworthiness and human-likeness. Participants also completed the individual differences in anthropomorphism questionnaire (IDAQ; Waytz et al., 2010). We hypothesised that anthropomorphic ratings on the Godspeed Anthropomorphism questionnaire (Bartneck, Kulić, et al., 2009) would be associated with perceptions of trustworthiness on a single item measure of trustworthiness. We did not find support for this hypothesis (r(41) = .041, p = .795, N = 43). We hypothesised that individual differences in the tendency to anthropomorphise would be associated with human-likeness of robot faces. We found that IDAQ scores were not associated with anthropomorphism (Estimate = .219, S.E. = .286, C.R. = .766, p = .444). Finally, we hypothesised that individual differences would mediate the relationship between anthropomorphism and perceptions of trustworthiness of robot faces. We found that there was no significant relationship between IDAQ scores and trustworthiness perceptions (Estimate = -.005, S.E. = .135, C.R. = -.039, p = .969). These findings suggest that participants accessed prior negative knowledge about robots and applied them to the robot images in this study. Additionally, the results suggest that trust affordances cannot be detected simply in the physical features of a robot. Limitations include the experimental design, especially the one-item measure of trustworthiness and the lack of contextual cues.
ER -