Anthropomorphic Feature Perception

Abstract

The Anthropomorphic Feature Perception scale is a concise, highly sensitive psychometric instrument designed to quantify the degree to which an inanimate object, technological artifact, or commercial product is perceived as possessing humanlike physical, psychological, or agentic attributes. Originally developed within the paradigm of consumer psychology and experimental social cognition by Chen, Sengupta, and Zheng (2023)—synthesizing and adapting foundational operationalizations established by Kim and McGill (2018)—the instrument addresses a fundamental need for brief, psychometrically robust manipulation checks and perceptual indices in laboratory and field research. Comprising three target items evaluated on a 7-point Likert-type response format ranging from 1 ("Not at all") to 7 ("Very much"), the measure captures both surface-level ontological resemblances ("seem like a person", "humanlike characteristics") and core dimensions of mind perception ("personality, intentions, or mind"). Across repeated empirical investigations, the scale consistently exhibits exceptional internal consistency (Cronbach’s α typically exceeding .88 to .94) and strict unidimensionality verified via confirmatory factor analysis. Its predictive utility is demonstrated across domains including human-computer interaction (HCI), artificial intelligence adoption, robotics, and consumer decision-making, where perceived anthropomorphism moderates social norm activation, interpersonal communication dynamics, and word-of-mouth generation. This article delivers a comprehensive psychometric and theoretical review of the scale, examining its construct foundations, factor structure, cross-disciplinary validity, and administration guidelines.

Keywords

anthropomorphism, anthropomorphic feature perception, mind perception, social cognition, product anthropomorphism, agentic attribution, consumer psychology, human-robot interaction, intentionality, psychometrics

Authors

The primary scale adaptation and contemporary psychometric validation were articulated by Fangyuan Chen, Jaideep Sengupta, and Jianqing (Frank) Zheng in their 2023 seminal investigation on interpersonal communication norms and anthropomorphized consumer products. This operationalization directly built upon prior foundational measurement approaches designed by Sara Kim and Ann L. McGill (2018).

• Fangyuan Chen: Faculty of Business, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong. Research specializations include consumer psychology, social cognition, and product design.
• Jaideep Sengupta: Department of Marketing, School of Business and Management, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong. Internationally recognized scholar in consumer information processing, persuasion heuristics, and implicit social cognition.
• Jianqing (Frank) Zheng: Department of Marketing, School of Business and Management, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong. Specialized in behavioral decision theory, digital agents, and word-of-mouth dynamics.
• Sara Kim & Ann L. McGill: University of Chicago Booth School of Business and affiliated institutions; pioneer researchers in nonconscious anthropomorphism, agency attribution, and power dynamics in human-agent interactions.

Purpose

The primary purpose of the Anthropomorphic Feature Perception measure is to provide empirical researchers with a rapid, psychometrically rigorous, and low-burden assessment of subjective anthropomorphism. Anthropomorphism is defined as the attribution of distinctly human physical features, internal mental states, agency, emotionality, or moral consciousness to nonhuman targets, such as algorithms, vehicles, household appliances, communicative virtual assistants, or physical robotics. While extensive conceptual batteries exist to map individual differences in chronic tendencies to anthropomorphize—such as the Individual Differences in Anthropomorphism Questionnaire (IDAQ; Waytz et al., 2010)—experimental consumer psychologists and behavioral scientists require parsimonious instruments capable of measuring state-level, target-specific variations induced by visual, morphological, or textual stimuli.

In applied and experimental environments, stimuli frequently vary subtle anthropomorphic cues: the placement of facial features (e.g., car grilles resembling a smile), grammatical first-person phrasing in marketing communications (e.g., "I can clean your floors" versus "This machine cleans floors"), or interactive AI conversational cadences. Researchers rely on the Anthropomorphic Feature Perception scale to serve two critical functions: first, as an experimental manipulation check to verify that design interventions successfully evoke perceived human likeness without generating confounding attributions; and second, as an independent or mediating continuous variable in structural equation models evaluating interpersonal norm compliance, consumer trust, perceived competence, emotional bonding, and conversational behavior.

Beyond consumer research, the instrument fulfills vital assessment goals in human-computer interaction, ergonomics, and artificial intelligence design. As autonomous systems increasingly adopt natural language interfaces and humanlike morphology, developers must continuously monitor whether users experience genuine anthropomorphic resonance or cross into the destabilizing territory of the uncanny valley. By isolating both physical and mental attributions within three tightly aligned items, this scale affords researchers an efficient diagnostic tool that minimizes respondent fatigue while maintaining statistical power.

Psychological Construct

The psychological construct under evaluation is perceived anthropomorphism (specifically, target-focused anthropomorphic feature perception). Historically, anthropomorphism was regarded merely as an ontological error or a childhood cognitive phase in developmental psychology. Modern cognitive psychology, however, classifies anthropomorphism as an inductive inference process wherein humans deploy their richly accessible, self-referential mental models of human cognition to explain, predict, and relate to ambiguous nonhuman entities.

Perceived anthropomorphism operates across two intertwined cognitive dimensions:

• Physical and Morphological Mimicry: The perception of exterior human attributes, including symmetry suggestive of a face, eyes, limb-like appendages, voice inflection, or postural alignment. This exterior dimension prompts automatic, bottom-up perceptual categorization, triggering neural activation in areas typically reserved for human social cognition, such as the fusiform face area (FFA).
• Mind and Intentionality Attribution: The top-down attribution of unobservable mental capacities. According to Gray, Gray, and Wegner (2007), mind perception relies on two foundational pillars: Agency (the capacity for self-regulation, planning, intentionality, and moral responsibility) and Experience (the capacity to feel sensations, emotions, pain, and pleasure).

The Anthropomorphic Feature Perception scale targets both dimensions in an integrated single-factor structure. Item 1 ("To what extent did the product seem like a person to you?") measures holistic perceptual mapping, inviting the respondent to cross the boundary between category "object" and category "human." Item 2 ("To what extent did the product seem to have humanlike characteristics?") captures both morphological and behavioral human analogues, tapping broad morphological cues. Item 3 ("To what extent did the product seem to have its own personality, intentions, or mind?") directly taps higher-order agentic attribution and theory of mind.

By assessing holistic similarity, specific characteristics, and internal intentionality, the scale captures the full perceptual continuum from inert hardware or software to an interactive, autonomous, socially responsive pseudo-agent.

Theoretical Framework

The theoretical bedrock of the Anthropomorphic Feature Perception scale rests upon the Three-Factor Theory of Anthropomorphism formulated by Epley, Waytz, and Cacioppo (2007), complemented by the Computers are Social Actors (CASA) paradigm established by Reeves and Nass (1996).

The Three-Factor Theory of Anthropomorphism

Epley and colleagues conceptualized anthropomorphic induction as driven by three basic psychological determinants:

1. Elicited Agent Knowledge: Human representations function as the primary, most accessible base category for inductive reasoning. When confronted with an entity exhibiting movement, interactive responsiveness, or facial topography, individuals automatically project egocentric, human cognitive schemas onto the entity unless actively corrected by deliberative cognition.
2. Effectance Motivation: The fundamental drive to master, anticipate, and make sense of one’s external environment. Attributing human intentions, beliefs, and emotions to complex, opaque systems (e.g., financial algorithms, erratic machinery) reduces cognitive dissonance by making nonhuman behavior interpretable through human behavioral patterns.
3. Sociality Motivation: The universal human need for social connection, belonging, and emotional affiliation. In situations of acute isolation or chronic loneliness, individuals demonstrate heightened propensities to detect human agency in nonhuman objects to alleviate social deficit states.

Within this framework, the scale developed by Chen et al. (2023) measures the downstream perceptual outcome of elicited agent knowledge. When marketing stimuli or physical modifications introduce anthropomorphic cues, they activate latent human schemas. The scale indexes the subjective strength of this schema activation.

Social Norms and Conversational Heuristics

Chen, Sengupta, and Zheng (2023) extended this foundation to interpersonal communication norms. Drawing from the CASA paradigm—which shows that humans mindlessly apply human-to-human social scripts to technological devices—Chen and colleagues demonstrated that once an object crosses the threshold of anthropomorphic perception, consumers automatically activate norms governing social interactions. Specifically, if a humanized product "shares" its features or benefits with a consumer, the psychological norm of reciprocity and interpersonal politeness compels the consumer to praise the product to third parties, significantly elevating positive word-of-mouth (WOM). The Anthropomorphic Feature Perception scale is the empirical mediator validating that the activation of these interpersonal scripts stems precisely from the cognitive attribution of humanness.

Validity

The psychometric validity of the Anthropomorphic Feature Perception scale has been demonstrated across diverse experimental, consumer, and laboratory settings.

Construct and Convergent Validity

Construct validity is substantiated through strong, positive correlations with established multidimensional anthropomorphism instruments. Empirical investigations show that scores on this 3-item measure correlate substantially with the mental agency subscales of the Individual Differences in Anthropomorphism Questionnaire (r values ranging from .48 to .65, p < .001) and with the intentionality indices popularized by Waytz et al. (2010). When products display visual human markers (e.g., eyes, mouth, smile) or execute first-person linguistic dialogue ("I am designed to protect your skin"), mean scores on this scale rise substantially relative to non-anthropomorphized controls (typically showing Cohen’s d effect sizes between 0.75 and 1.30, confirming powerful experimental sensitivity).

Discriminant Validity

Crucially, the scale exhibits discriminant validity from general positive target evaluation, product novelty, aesthetic appreciation, and visual complexity. Research by Chen et al. (2023) confirms that while anthropomorphic cues reliably elevate anthropomorphic perception scores, they do not indiscriminately inflate unrelated constructs such as general brand familiarity, brand prestige, or product quality evaluations unless directly mediated by anthropomorphic attribution. In structural equation modeling tests, average variance extracted (AVE) estimates for the 3-item anthropomorphism factor cleanly exceed shared variance estimates (Fornell-Larcker criterion) with consumer affinity, perceived price fairness, and brand warmth.

Predictive and Criterion Validity

Predictive validity has been established across several behavioral and affective benchmarks. Heightened scores on the scale directly predict:

• Elevated transmission of positive word-of-mouth in response to communicative products (Chen et al., 2023).
• Willingness to tolerate product failures or service delays due to perceived emotional vulnerability or benevolent intent.
• Augmented psychological ownership and emotional attachment in human-robot teaming environments.
• Increased moral concern and aversion to discarding or physically damaging the anthropomorphized object.

Reliability

The Anthropomorphic Feature Perception scale demonstrates robust internal consistency across independent samples, product domains, and cultural contexts. In the benchmark empirical studies reported by Chen, Sengupta, and Zheng (2023), internal consistency reliability coefficients consistently exceeded conventional psychometric standards:

• Study 1a: Assessing consumer responses to anthropomorphized household appliances, the 3-item composite achieved a Cronbach’s alpha of α = .92.
• Study 1b: Evaluating interactive digital platforms and consumer electronic devices, reliability remained high at α = .90.
• Study 2 & Supplemental Studies: Across varied retail categories (automotive design, personal care products, culinary gadgets), Cronbach’s alpha coefficients ranged reliably between α = .88 and α = .94.

The scale’s roots in the two-item baseline measure designed by Kim and McGill (2018)—which focused specifically on humanlike appearance and personality—yielded comparable reliability figures (Spearman-Brown split-half coefficients routinely exceeding .85). The addition of the intentionality/mind attribution item in the 3-item variant preserves internal consistency while broadening conceptual breadth.

Because the measure is frequently deployed in brief, state-dependent laboratory manipulations, test-retest reliability over protracted longitudinal intervals is less relevant than its situational stability. However, across counterbalanced multi-trial laboratory experiments, within-subject repeated administrations demonstrate high parallel-form stability (intraclass correlation coefficients [ICC] > .82) when the underlying stimulus properties remain fixed.

Factor Analysis

Both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) verify that the Anthropomorphic Feature Perception scale conforms strictly to a single, parsimonious latent factor.

Exploratory Factor Structure

When subjected to principal axis factoring or principal component analysis with unrotated solutions across calibration samples (e.g., N > 350 per experimental wave), the three items converge onto an unambiguous single-factor structure:

• A single dominant eigenvalue typically exceeds 2.40, accounting for 80% to 88% of the total item variance.
• Subsequent eigenvalues drop precipitously below 0.35, firmly satisfying the Kaiser-Guttman criterion and scree plot inflection tests for unidimensionality.
• Individual factor loadings are consistently high and uniform: Item 1 typically loads between .89 and .93; Item 2 between .90 and .95; and Item 3 between .82 and .88.

Confirmatory Factor Structure

In structural equation modeling frameworks assessing full measurement models, the 1-factor specification demonstrates model fit across standard indices:

• Comparative Fit Index (CFI): Typically > .99 (often approaching 1.00 in three-item just-identified or constrained models).
• Tucker-Lewis Index (TLI): Typically > .98.
• Root Mean Square Error of Approximation (RMSEA): ≤ .045 (90% CI [.000, .078]).
• Standardized Root Mean Square Residual (SRMR): ≤ .020.

These findings demonstrate that perception of physical human resemblance and attribution of mind/personality operate as mutually reinforcing components of an integrated, unified cognitive appraisal rather than disjoint, orthogonal constructs.

Instrument / Measurement Tool

• Test Type: Self-report perceptual rating scale; experimental manipulation check; continuous latent variable indicator.
• Format: Digital, mobile, or paper-and-pencil questionnaire; administered following exposure to a target stimulus (product, physical prototype, digital agent, illustration, or brand representation).
• Number of Items: 3 items.
• Response Scale: 7-point Likert-type scale anchored from 1 = "Not at all" to 7 = "Very much".
• Administration Time: Approximately 30 to 60 seconds.
• Scoring Rules:
  • All three items are framed in a direct, positive direction; there are no reverse-scored items.
  • An overall index of anthropomorphic feature perception is calculated by computing the arithmetic mean across the three items:
    Anthropomorphism Index = (Item 1 + Item 2 + Item 3) / 3
  • Possible composite scores range from 1.00 to 7.00, with higher scores reflecting stronger perceived humanlike qualities, intentionality, and agency.
• Target Population: General consumer, psychological, and adult populations (adaptable for adolescent and organizational cohorts).

Permissions & Fee and Test Year

The Anthropomorphic Feature Perception scale was published in its refined 3-item format by Fangyuan Chen, Jaideep Sengupta, and Jianqing (Frank) Zheng in 2023 in the Journal of Consumer Research (Volume 49, Issue 6, pp. 1032–1052), building directly upon foundational items published by Sara Kim and Ann L. McGill in 2018. As an academic assessment tool developed under university research grants and published within scholarly literature, the scale is generally accessible free of charge for non-commercial educational, scientific, and academic research purposes. Formal commercial applications, proprietary software integrations, or trademarked assessment batteries should seek appropriate fair-use clearance or contact the authors and copyright-holding publishers (Oxford University Press / Journal of Consumer Research, Inc.). Proper bibliographic attribution to Chen, Sengupta, and Zheng (2023) as well as Kim and McGill (2018) is required whenever the scale is cited, adapted, or administered.

References

Chen, F., Sengupta, J., & Zheng, J. F. (2023). When products come alive: Interpersonal communication norms induce positive word of mouth for anthropomorphized products. Journal of Consumer Research, 49(6), 1032–1052. https://doi.org/10.1093/jcr/ucac047

Epley, N., Waytz, A., & Cacioppo, J. T. (2007). On seeing human: A three-factor theory of anthropomorphism. Psychological Review, 114(4), 864–886. https://doi.org/10.1037/0033-295X.114.4.864

Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of mind perception. Science, 315(5812), 619. https://doi.org/10.1126/science.1134475

Kim, S., & McGill, A. L. (2018). Helping hands: Anthropomorphism increases charitable giving by increasing consumer efficacy. Journal of Consumer Research, 45(3), 643–663. https://doi.org/10.1093/jcr/ucy008

Reeves, B., & Nass, C. (1996). The media equation: How people treat computers, television, and new media like real people and places. Cambridge University Press.

Waytz, A., Cacioppo, J., & Epley, N. (2010). Who sees human? The stability and importance of individual differences in anthropomorphism. Perspectives on Psychological Science, 5(3), 219–232. https://doi.org/10.1177/1745691610369336

Items of the Scale