Different humanlike design dimensions cue distinct classes of anthropomorphism inferences¶

Claim · All claims
Evidence · 1 study · theoretical r3 · q2
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q2 i? People attribute affective, social, and moral minded capacities to robots based on facial features, while attributions about physical interaction are tied to bodily features and manipulators. → Phillips 2026

Evidence¶

Phillips 2026¶

Phillips, E. K., & de Visser, E. J. (2026). Humanlikeness as design, anthropomorphism as inference: a conceptual framework for human–robot interaction. Frontiers in Cognition. https://doi.org/10.3389/fcogn.2026.1786256

q2 · i? · theoretical · r3

The review cites prior empirical work by Zhao et al. (2019) showing that facial features alone cue affective, social, and moral attributions, whereas "bodily features and the presence of manipulators" cue attributions about physical interaction. The review reports this as cited prior work, not its own study.

"Prior work (Zhao et al., 2019) demonstrates that people attribute affective (e.g., positive and negative feelings), social (e.g., understanding others' minds), and moral (e.g., telling right from wrong) minded capacities to robots based solely on facial features, whereas attributions related to physical interaction with the world are more strongly tied to bodily features and the presence of manipulators such as arms, legs, and hands."

Discussion¶

Related Claims¶

• A conceptual framework paper reports, citing a meta-analysis (Roesler et al., 2021), that anthropomorphism benefits human-robot interaction with effects varying across task contexts and design implementations — related
• Manipulating anthropomorphism shows no clear causal effect on trust in controlled post-generative-AI studies, though long-term use and cue type condition its effects — related
• Trust evaluations were strongly and positively correlated across all robot–task combinations, suggesting trust in social robots may be a stable, generalized attitude — related
• Robot communication cues affect sense of agency: predicted-path display and averted gaze reduce it, direct gaze raises it, and facial expression shows no effect — related
• Participants generally felt humanlikeness should not be a design goal for a writing-support agent used as a tool — a narrower finding that bears on this claim
• RALL studies raise general issues regarding robots' social behavior in learning situations — related