Humanlikeness as a multidimensional design construct extending beyond physical appearance¶

Theory · All theories
Evidence · 2 claims (1 for, 1 mixed) · 1 study (1 theoretical), q1 · 0 of 1 report an effect size · 2 claims rest on one study

Description¶

The article argues humanlikeness should be defined broadly, extending beyond outward physical appearance: "Humanlikeness may vary along multiple dimensions, including the number, salience, and specificity of features; designed behaviors and interaction conventions; interface characteristics; and even the inspiration underlying processes of autonomy." For example, reinforcement learning represents a humanlike way for robots to learn. The construct is also inherently relational and perception-dependent, operationalized via consensus in human judgment on a spectrum from less to more humanlike.

Design Implications¶

Context¶

Requirements¶

• Operationalization relies on consensus among human judgments as a pragmatic solution to the lack of an objective standard human

Constraints¶

• Prior bottom-up feature frameworks were limited to outward static physical appearance; extension to movement, voice, and language needs additional research

Target Learners¶

• HRI researchers
• robot designers

Target Learning Objectives¶

• Specifying and measuring robot humanlikeness as a multidimensional design property

Claims¶

• Different humanlike design dimensions cue distinct classes of anthropomorphism inferences [+W]
• 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 [~W]

Related Theories¶

• Four-quadrant inferential framework for interpreting human attribution responses to robot design
• Integrated conceptual framework distinguishing humanlikeness as design property from anthropomorphism as attribution process
• DiSalvo et al. taxonomy of four anthropomorphic design forms
• Six theories of anthropomorphism explaining why learners attribute human characteristics to nonhuman entities

Examples¶

• Treat humanlikeness as a tunable design parameter and anthropomorphism as an evaluation metric

Key Sources¶

• 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