The Antecedents of Consumer Acceptance Technology Agency–Model<\/strong> (Morosan & Dursun-Cengizci, 2024) is a multidimensional psychometric instrument formulated to capture consumer cognitive, affective, and ethical appraisals regarding the delegation of decision-making authority to artificial intelligence (AI<\/a>) systems. Grounded in the context of autonomous service systems within the hospitality and tourism sectors, this 22-item measurement battery evaluates the critical antecedents governing an individual\u2019s willingness to surrender consumer agency to automated non-human agents. The instrument systematically operationalizes six core latent dimensions: Perceived Ethics<\/em> (5 items), Perceived Benefits<\/em> (3 items), Perceived Risks<\/em> (3 items), Perceived Loss of Competence<\/em> (3 items), Perceived Unpredictability<\/em> (4 items), and Acceptance of Technology Agency<\/em> (4 items). Evaluated along an authentic 7-point Likert scale<\/a> ranging from 1 (“Strongly disagree”) to 7 (“Strongly agree”), the instrument demonstrates exceptional psychometric rigor. Confirmatory factor analysis (CFA<\/a>) supports the proposed structural specification, yielding robust goodness-of-fit indices (\u03c7\u00b2 = 370.279, df = 188, \u03c7\u00b2\/df = 1.97, CFI = 0.960, TLI = 0.951, RMSEA = 0.050). Internal consistency is verified via composite construct reliability (CCR) values exceeding the 0.80 benchmark across all latent constructs, and convergent validity is corroborated through average variance extracted (AVE) values surpassing 0.50, paired with standardized factor loadings predominantly above 0.70. By bridging ethical AI governance, cognitive appraisal theory, and agency theory, the scale serves as a standardized empirical apparatus for investigating autonomous human-AI collaboration, algorithmic aversion, and consumer technology adoption in high-contact service ecosystems<\/span>.<\/p>\n Technology Agency, Artificial Intelligence Acceptance, Perceived Ethics, Hospitality Technology, Algorithmic Decision-Making, Consumer Autonomy, Loss of Competence, Perceived Risk, Psychometric Validation, Measurement Model<\/p>\n The scale was developed, calibrated, and validated by:<\/p>\n The exponential advancement of artificial intelligence and machine learning technologies has precipitated a paradigm shift from passive assistive systems\u2014which merely present information for human consumption\u2014to autonomous agentic systems capable of independent decision-making and execution. In contemporary consumer environments, such as smart hotels, travel booking platforms, and automated personal finance, artificial intelligence is increasingly tasked with making binding decisions on behalf of users. While traditional technology acceptance frameworks (such as the Technology Acceptance Model<\/a> [TAM] or the Unified Theory of Acceptance and Use of Technology [UTAUT<\/a>]) examine an individual’s intent to interact directly with software tools, they remain fundamentally inadequate for capturing the psychological complexities of delegation<\/em>\u2014the intentional relinquishment of behavioral control and decision-making sovereignty to an artificial entity.<\/p>\n The primary purpose of the Antecedents of Consumer Acceptance Technology Agency–Model is to provide an empirically validated, psychometrically robust diagnostic instrument designed to isolate, measure, and explain the cognitive, psychological, and normative factors that drive or inhibit consumer acceptance of autonomous technology agency. Specifically, the scale was engineered to address the critical research gap concerning how ethical perceptions interface with traditional functional evaluations (benefits versus risks) and profound psychological frictions (such as perceived loss of competence and systemic unpredictability).<\/p>\n From an applied perspective, the instrument fulfills several critical research and managerial objectives:<\/p>\n The construct assessed by this instrument is Consumer Technology Acceptance of Technology Agency<\/strong>, conceptualized not merely as user adoption or behavioral intention to use an interface, but specifically as an individual’s cognitive willingness and behavioral intention to delegate discretionary choices, actions, and autonomous authority to an artificial intelligence entity. Within this framework, technology is no longer viewed as a static tool (an artifact) but as an independent proxy agent (an actor) operating on behalf of the principal (the consumer).<\/p>\n The instrument deconstructs this overarching construct into six distinct, interrelated latent subscales:<\/p>\n This dimension evaluates the consumer\u2019s normative and moral appraisal of the AI system’s underlying operational logic. Drawing upon foundational literature in computer ethics and consumer trust, it gauges beliefs regarding whether the autonomous system adheres to basic ethical tenets. Specific sub-facets include:<\/p>\n Reflecting classical cognitive appraisal theory and utilitarian valence, this subscale captures the anticipated instrumental utility derived from allowing an AI agent to assume executive control. Rather than assessing simple ease of use, it measures advanced functional gains: qualitative enhancement of the service encounter, dramatic reductions in cognitive friction and decision-making time, and elevated operational convenience achieved through autonomous service orchestration.<\/p>\n This construct measures the consumer\u2019s subjective expectation of negative outcomes, systemic liabilities, and unintended adverse consequences resulting from delegated authority. Rooted in perceived risk theory, this subscale captures both generalized situational apprehension (“risky”) and specific anticipatory anxieties concerning algorithmic errors, cascading operational failures, or irreversible suboptimal selections executed by the machine without human intervention.<\/p>\n Anchored in psychological theories of personal efficacy and self-determination, this dimension assesses the psychological threat posed to an individual\u2019s identity as an active, capable consumer. When an algorithm preempts consumer deliberation, individuals may feel that their agency is degraded, their personal decision-making skills are undermined, and their subjective sense of self-efficacy is neutralized. This construct specifically quantifies feelings of perceived disempowerment, diminished personal control, and systemic consumer deskilling.<\/p>\n This subscale captures the epistemic uncertainty and perceived opacity surrounding the machine’s internal decision logic (frequently referred to as the “black-box” phenomenon). It measures the extent to which the consumer perceives the AI’s future actions to be volatile, erratic, impossible to forecast, or fundamentally outside cognitive comprehension. High perceived unpredictability induces psychological tension because the principal cannot reliably anticipate how their proxy will represent them.<\/p>\n The focal criterion construct of the model represents the consumer’s consolidated behavioral intention, readiness, and explicit willingness to authorize an artificial intelligence platform to make binding decisions and enact autonomous transactions on their behalf. It embodies the active surrender of operational jurisdiction to the digital proxy within a designated service ecosystem.<\/p>\n The Antecedents of Consumer Acceptance Technology Agency–Model is constructed upon an integrative theoretical architecture that synthesizes principles from Principal-Agent Theory<\/strong>, the Cognitive Appraisal Theory of Emotion and Coping<\/strong>, Self-Determination Theory<\/strong>, and Value-Belief-Norm (VBN) \/ Ethical Trust Paradigms<\/strong>.<\/p>\n Originally formulated within economics and organizational sociology by theorists such as Jensen and Meckling (1976)<\/a>, Principal-Agent Theory examines relationships where one party (the principal) designates another party (the agent) to perform service tasks that involve delegating decision-making authority. When translated to autonomous technological contexts, the human consumer acts as the principal, while the artificial intelligence architecture assumes the role of the operational agent.<\/p>\n Classical agency dilemmas stem from information asymmetry and conflicting self-interests: the principal is frequently unable to verify whether the agent is acting faithfully in the principal\u2019s best interests or pursuing opportunistic behaviors. Morosan and Dursun-Cengizci (2024) recognized that consumer acceptance of AI agency hinges directly on resolving this fundamental agency problem. Because algorithms are non-sentient, their perceived agency is inextricably linked to consumer trust in the institutional, structural, and moral boundaries coded into the system\u2014represented within the scale by Perceived Ethics<\/em>, Perceived Risks<\/em>, and Perceived Unpredictability<\/em>.<\/p>\n According to Lazarus and Folkman\u2019s (1984)<\/a> Cognitive Appraisal Theory, environmental stimuli do not directly elicit behavioral outcomes; rather, individuals evaluate the significance of an encounter in relation to their personal well-being. This occurs via:<\/p>\n Acceptance or rejection of technology agency thus emerges as an active behavioral coping response derived from this complex internal cognitive calculus.<\/p>\n Grounded in the foundational work of Deci and Ryan (1985)<\/a>, Self-Determination Theory posits that psychological health and intrinsic motivation rely upon the fulfillment of three basic needs: autonomy, competence, and relatedness. Autonomous AI systems directly challenge the need for competence and personal autonomy. Concurrently, Brehm\u2019s (1966)<\/a> Psychological Reactance Theory states that whenever individuals perceive that their behavioral freedom is threatened or constrained, an intense motivational state is aroused to re-establish that threatened freedom. The model explicitly incorporates these deep self-concept dynamics through the Perceived Loss of Competence<\/em> dimension, recognizing that algorithmic interventions can induce psychological distress if consumers perceive that algorithmic decision-making reduces their personal agency.<\/p>\n The psychometric properties of the Antecedents of Consumer Acceptance Technology Agency–Model have undergone extensive empirical evaluation, demonstrating robust construct validity, convergent validity, and discriminant validity across rigorous structural equation modeling procedures.<\/p>\n To establish rigorous content validity, scale development adhered to standardized multi-stage psychometric procedures. Initial item pools were adapted from validated inventories across related disciplines\u2014including electronic commerce ethics (e.g., Roman, 2007), consumer information privacy (e.g., Xu et al., 2011), and autonomous consumer interaction paradigms (e.g., Bertrandias et al., 2021)\u2014complemented by newly formulated items reflecting the specific realities of autonomous machine agency. Expert panels specializing in consumer psychology, hospitality management, and human-computer interaction reviewed the preliminary item pool to eliminate semantic ambiguities, confirm structural relevance, and verify that items exhibited clear face validity prior to quantitative deployment.<\/p>\n Convergent validity was systematically evaluated through Confirmatory Factor Analysis (CFA) by examining standardized factor loadings, average variance extracted (AVE), and statistical significance (Hair et al., 2009). The empirical validation yielded the following benchmarks:<\/p>\n Discriminant validity was established via the classical Fornell-Larcker criterion, alongside modern multi-trait testing. For every latent construct within the model, the square root of the Average Variance Extracted exceeded the highest bivariate correlation between that construct and any other latent construct in the measurement model. Furthermore, cross-loadings were inspected; no single indicator exhibited higher factor loadings on alternative cross-constructs than on its assigned theoretical dimension, establishing that the six factors represent empirically distinct psychological phenomena.<\/p>\n Structural equation modeling demonstrated robust predictive and nomological validity. In alignment with theoretical postulates, Perceived Ethics<\/em> emerged as a powerful positive predictor of Perceived Benefits<\/em> and a negative predictor of Perceived Risks<\/em> and Unpredictability<\/em>. In turn, Perceived Benefits<\/em> exerted a strong, positive, statistically significant direct effect on Acceptance of Technology Agency<\/em>, whereas Perceived Risks<\/em>, Loss of Competence<\/em>, and Unpredictability<\/em> exerted direct and indirect inhibitory influences. This confirms that the operationalized battery functions seamlessly within the broader nomological network of consumer decision-making and modern technology adoption.<\/p>\n The reliability of the Antecedents of Consumer Acceptance Technology Agency–Model has been documented through tests of internal consistency and composite metric reliability, fulfilling and exceeding established psychometric thresholds.<\/p>\n Traditional psychometric assessments rely heavily on Cronbach\u2019s alpha (\u03b1)<\/a>. However, because Cronbach\u2019s alpha is vulnerable to underestimation or overestimation based on scale length and assumes tau-equivalence (equal factor loadings), the authors prioritized Composite Construct Reliability (CCR)<\/strong>, also termed composite reliability (CR), alongside coefficient alpha.<\/p>\n Recognizing that self-reported cross-sectional survey designs can be susceptible to common method variance (CMV<\/a>), the researchers implemented rigorous procedural remedies during administration (Podsakoff et al., 2003):<\/p>\n The factorial validity of the instrument was tested and refined through rigorous Confirmatory Factor Analysis (CFA) utilizing maximum likelihood estimation techniques.<\/p>\n A comprehensive measurement model was specified comprising the six core latent dimensions: Perceived Ethics<\/em> (5 indicators), Perceived Benefits<\/em> (3 indicators), Perceived Risks<\/em> (3 indicators), Perceived Loss of Competence<\/em> (3 indicators), Perceived Unpredictability<\/em> (4 indicators), and Acceptance of Technology Agency<\/em> (4 indicators). The total measurement configuration consisted of 22 observed variables.<\/p>\n Following initial diagnostic checks and standard covariance inspections, the respecified structural measurement model demonstrated exceptional model-to-data fit across all primary descriptive, relative, and absolute fit indices, meeting the rigorous standards defined by Hu and Bentler (1999) and Toh et al. (2006):<\/p>\n Standardized factor loadings were uniformly high across all constructs:<\/p>\n Test Publication Year:<\/strong> 2024<\/p>\n Copyright and Ownership:<\/strong> The Antecedents of Consumer Acceptance Technology Agency–Model was developed and published by Cristian Morosan and Asl\u0131han Dursun-Cengizci (2024). The formal academic manuscript is copyrighted by Emerald Publishing Limited in the International Journal of Contemporary Hospitality Management<\/em>.<\/p>\n Licensing and Research Permissions:<\/strong> Under standard international academic research practices, the instrument\u2019s items and operationalized scales are available for non-commercial educational, scientific, and scholarly research purposes, provided that appropriate formal academic attribution is extended to the original authors and the published source journal. Commercial implementations, proprietary software integration, diagnostic enterprise consulting, or publication within fee-bearing diagnostic batteries typically require explicit written authorization and licensing from the authors and\/or the publisher.<\/p>\n Inquiries and Contact:<\/strong> Researchers seeking collaboration, customized implementations, or cross-cultural validation protocols may contact the lead author, Dr. Cristian Morosan, via the Conrad N. Hilton College of Global Hospitality Leadership at the University of Houston, or through his registered academic ORCID identifier (0000-0001-6751-1348<\/a>).<\/p>\n2. Keywords<\/h2>\n
3. Authors<\/h2>\n
\n
4. Purpose<\/h2>\n
\n
5. Psychological Construct<\/h2>\n
1. Perceived Ethics (5 Items)<\/h3>\n
\n
2. Perceived Benefits (3 Items)<\/h3>\n
3. Perceived Risks (3 Items)<\/h3>\n
4. Perceived Loss of Competence (3 Items)<\/h3>\n
5. Perceived Unpredictability (4 Items)<\/h3>\n
6. Acceptance of Technology Agency (4 Items)<\/h3>\n
6. Theoretical Framework<\/h2>\n
Principal-Agent Theory<\/h3>\n
Cognitive Appraisal Theory<\/h3>\n
\n
Self-Determination Theory and Psychological Reactance<\/h3>\n
7. Validity<\/h2>\n
Content and Face Validity<\/h3>\n
Convergent Validity<\/h3>\n
\n
Discriminant Validity<\/h3>\n
Construct and Predictive Validity<\/h3>\n
8. Reliability<\/h2>\n
Internal Consistency Metrics<\/h3>\n
\n
Mitigation of Common Method Bias<\/h3>\n
\n
9. Factor Analysis<\/h2>\n
Confirmatory Factor Analysis (CFA) Model Specification<\/h3>\n
Goodness-of-Fit Assessment<\/h3>\n
\n
Structural Loading Distribution<\/h3>\n
\n
10. Instrument \/ Measurement Tool<\/h2>\n
\n
\n
\n
\n
11. Permissions & Fee and Test Year<\/h2>\n
12. References<\/h2>\n
\n
13. Items of the Scale (Questionnaire)<\/h2>\n