Online Trust Scale (OTS)

Abstract

The Online Trust Scale (OTS), formulated and empirically validated by David Gefen, Elena Karahanna, and Detmar W. Straub (2003), represents a foundational psychometric instrument in information systems (IS), electronic commerce, and consumer psychology. Built upon the classic tripartite conceptualization of interpersonal and organizational trust developed by Mayer, Davis, and Schoorman (1995), the scale measures consumer trust in an online vendor across three distinct cognitive dimensions: Integrity, Benevolence, and Ability (Competence). Comprising 13 items evaluated on a 7-point Likert response format (ranging from 1 = Strongly disagree to 7 = Strongly agree), the instrument captures consumer cognitive appraisals regarding whether a digital merchant is honest, cares about the consumer’s welfare, and possesses the requisite capacity and skills to execute digital transactions reliably.

In their seminal investigation, Gefen et al. (2003) integrated this multidimensional trust construct into Davis’s (1989) Technology Acceptance Model (TAM), illustrating how trust operates concurrently with perceived usefulness (PU) and perceived ease of use (PEOU) to govern consumer behavioral intentions and continued e-commerce adoption. Extensive psychometric evaluations across both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) within structural equation modeling (SEM) frameworks demonstrate exceptional internal consistency reliability (Cronbach’s α typically ranging between .87 and .94 across subdimensions), strong convergent validity with average variance extracted (AVE) values exceeding .60, and distinct discriminant validity against related technological constructs. This article delivers an exhaustive review of the scale’s theoretical origins, construct operationalization, psychometric properties, factor structure, administration parameters, and academic applications.

Keywords

Online Trust Scale, consumer trust, e-commerce adoption, Mayer Davis and Schoorman model, integrity, benevolence, ability, Technology Acceptance Model, psychometrics, structural equation modeling, vendor credibility, digital marketing

Authors

The scale was developed and psychometrically established by three prominent scholars in management information systems and organizational behavior:

• David Gefen, Ph.D. — Professor of Management Information Systems and Provost Planning Fellow at the Lebow College of Business, Drexel University. Dr. Gefen is internationally recognized for his pioneering research on electronic commerce, structural equation modeling methodologies, and trust dynamics in virtual environments.
• Elena Karahanna, Ph.D. — C. Herman and Mary Virginia Terry Chair of Business Administration and Distinguished Research Professor of MIS at the Terry College of Business, University of Georgia. Her scholarship investigates the adoption, implementation, and diffusion of information technologies, alongside consumer healthcare informatics.
• Detmar W. Straub, Ph.D. — J. Mack Robinson Distinguished Professor of Information Systems at Georgia State University (Emeritus) and Visiting Professor at Temple University. Dr. Straub is an internationally renowned methodologist who has served as Editor-in-Chief of MIS Quarterly and published extensively on research methods, cybersecurity, and global e-commerce.

Purpose

The advent of consumer-oriented World Wide Web transactions in the late 1990s and early 2000s introduced profound behavioral frictions that traditional brick-and-mortar retail theories failed to accommodate fully. Unlike conventional retail settings where tangible merchandise can be physically inspected, transactions occur face-to-face, and legal jurisdictions are localized, e-commerce inherently separates buyers and sellers across both spatial and temporal dimensions. This structural disintermediation introduces substantial transactional uncertainty, privacy risks, and potential vulnerabilities regarding financial fraud, product misrepresentation, and post-purchase neglect.

The primary purpose of the Online Trust Scale is to provide a theoretically grounded, psychometrically rigorous, and standardized measurement instrument capable of quantifying the specific cognitive beliefs consumers hold regarding an online vendor’s dependability. Specifically, the scale assesses:

• The extent to which consumers perceive that an online commercial entity will not exploit customer vulnerabilities for opportunistic gain.
• The degree of confidence buyers place in an automated web interface to fulfill transactional promises accurately.
• The mediating role of trust beliefs between website interface attributes (e.g., system navigation, structural assurances) and ultimate transactional adoption.

In empirical research, the OTS serves as a diagnostic and predictive mechanism. It allows investigators to disentangle the technological characteristics of a web interface (such as ease of use) from the social and interpersonal perceptions directed toward the merchant behind the screen. In managerial and market-research settings, the OTS enables organizations to audit consumer trust deficits systematically, discerning whether customer attrition stems from doubts about technical competence, questions regarding operational integrity, or perceptions of organizational callousness.

Psychological Construct

Trust in computer-mediated commerce is fundamentally a multi-component psychological construct rooted in vulnerability and risk. Drawing from the cross-disciplinary work of Mayer, Davis, and Schoorman (1995), trust is defined as the willingness of a party (the trustor/consumer) to be vulnerable to the actions of another party (the trustee/online vendor), based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party. Gefen et al. (2003) translated this overarching construct into three distinct yet highly correlated cognitive subdimensions:

1. Integrity

Integrity refers to the consumer’s perception that the vendor adheres to a mutually acceptable set of ethical principles, moral standards, and operational commitments. It captures fundamental beliefs concerning honesty, truthfulness, promise fulfillment, and the avoidance of opportunistic deception. In an online environment, where formal legal recourse is often perceived as complex and impractical for modest consumer transactions, integrity beliefs function as an informal social contract. An online consumer evaluating integrity considers whether the vendor provides authentic product descriptions, honors stated pricing and return policies, and preserves data confidentiality without deceptive monetization. (Items 1–4 operationalize this dimension).

2. Benevolence

Benevolence represents the trustor’s perception that the trustee genuinely cares about the consumer’s welfare and possesses a sincere desire to do good to the consumer, aside from an egocentric profit motive. While business organizations inherently pursue economic profitability, a benevolent vendor is perceived as one that would not purposefully act against the customer’s best interests, even when doing so might yield short-term financial gains. In digital shopping contexts, benevolence manifests when an online merchant proactively contacts buyers regarding product recalls, provides transparent disclosures about shipping delays, or offers frictionless, empathetic resolution of disputed transactions. (Items 5–8 operationalize this dimension).

3. Ability / Competence

Ability encapsulates the group of skills, competencies, technical capabilities, and domain characteristics that enable a party to exert influence within some specific domain. Even if a vendor is impeccably honest (integrity) and genuinely well-intentioned (benevolence), transactions will fail if the vendor lacks the infrastructure, logistical acumen, and technical capacity to execute orders accurately. In e-commerce, ability reflects the merchant’s perceived proficiency in inventory management, secure payment processing, domain knowledge concerning the products sold, and reliable post-transaction customer support. (Items 9–13 operationalize this dimension).

Theoretical Framework

The conceptual architecture of the Online Trust Scale rests upon the integration of two major paradigms in human behavior and social science: The Integrative Model of Organizational Trust (Mayer, Davis, & Schoorman, 1995) and the Technology Acceptance Model (Davis, 1989), which itself stems from Fishbein and Ajzen’s Theory of Reasoned Action (TRA).

The Mayer, Davis, and Schoorman (1995) Integrative Trust Model

Prior to Mayer et al.’s formulation, trust research across organizational psychology, sociology, and economics suffered from severe conceptual confusion, often conflating the antecedents of trust, trust itself, and the behavioral outcomes of trust. Mayer et al. clarified this space by delineating:

• Factors of Perceived Trustworthiness: The cognitive appraisals made about the trustee along the dimensions of Ability, Benevolence, and Integrity.
• Trust (The State): The willingness to assume vulnerability.
• Risk Taking in Relationship (RTR): The overt behavioral manifestation (e.g., executing a monetary transaction or sharing proprietary data).

Gefen, Karahanna, and Straub (2003) operationalized the “Factors of Perceived Trustworthiness” as the fundamental cognitive core of online trust, arguing that in digital retail transactions, consumers form holistic behavioral intentions based precisely on these three pillars.

Integration with the Technology Acceptance Model (TAM)

Davis’s (1989) TAM posits that the adoption of an information system is predominantly driven by two cognitive evaluations: Perceived Usefulness (PU)—the degree to which an individual believes that using a particular system will enhance performance—and Perceived Ease of Use (PEOU)—the degree to which an individual believes that using a system will be free of effort. While TAM proved exceptionally robust in explaining internal organizational software adoption (e.g., spreadsheets, word processors), it proved insufficient for open, risk-laden environments like the internet, where users are not merely system operators, but consumers engaging in economic exchanges with independent legal entities.

Gefen et al. demonstrated that trust acts as an essential socio-cognitive antecedent and co-determinant alongside TAM constructs:

• Trust directly enhances Perceived Usefulness: When a consumer trusts an online vendor, they expect that the anticipated utility of the website (e.g., finding and receiving products) will materialize without opportunistic disruption.
• Perceived Ease of Use enhances Trust: A clear, intuitive, and error-free website design signals to the consumer that the underlying vendor possesses professional competence and organizational investment, fostering perceptions of ability and integrity.
• Trust directly drives Behavioral Intention: Trust reduces psychological complexity and perceived vulnerability, directly facilitating consumer willingness to engage in transactions.

Validity

The psychometric validity of the Online Trust Scale has been thoroughly evaluated across multiple samples, retail domains, and analytical configurations:

1. Construct and Convergent Validity

Gefen, Karahanna, and Straub (2003) established convergent validity through both exploratory factor analysis (principal components with varimax rotation) and confirmatory structural equation modeling (utilizing both LISREL and Partial Least Squares / PLS). Across empirical tests involving university student cohorts and real-world e-commerce shoppers purchasing books and consumer goods, all 13 items demonstrated statistically significant (p < .001) loadings on their targeted latent constructs. Standardized factor loadings across studies consistently surpassed the widely accepted .70 benchmark (ranging predominantly between .72 and .92), indicating that each item accounts for at least 50% of the variance in its corresponding latent factor.

The Average Variance Extracted (AVE) for each of the three dimensions exceeded the recommended .50 threshold, routinely falling within the .62 to .78 range, providing robust statistical verification of convergent validity according to the Fornell and Larcker (1981) criteria.

2. Discriminant Validity

Discriminant validity was verified through multiple analytical approaches:

• Fornell-Larcker Criterion: The square root of the AVE for each trust construct (Integrity, Benevolence, Ability) consistently exceeded the inter-construct correlations with any other latent variable in the model (including Perceived Usefulness, Perceived Ease of Use, and Intended Use).
• Cross-Loading Inspection: In factor matrices, every item loaded considerably higher on its assigned theoretical construct than on any other construct, with cross-loadings universally below .35.
• Nested Model Chi-Square Difference Tests: In confirmatory analyses, constraining the correlation between any two trust dimensions to unity (1.00) resulted in a statistically significant degradation of model fit (Δχ², p < .001), confirming that Integrity, Benevolence, and Ability represent distinct empirical phenomena rather than a singular undifferentiated attitude.

3. Predictive and Nomological Validity

The nomological validity of the OTS is demonstrated by its structural alignment within broader theoretical frameworks. Across studies, the three trust dimensions accounted for substantial variance in online purchase intentions (R² values between .35 and .55 when combined with TAM constructs). Furthermore, trust exhibited significant structural paths to Perceived Usefulness (β values typically ranging from .30 to .45, p < .001), confirming the theoretical premise that consumer trust fundamentally reduces the friction and perceived risk of utility realization in electronic channels.

Reliability

The Online Trust Scale demonstrates high internal consistency reliability across varied empirical applications. In the original Gefen, Karahanna, and Straub (2003) validation study, the estimated reliability indices exceeded all conventional psychometric cutoffs:

• Integrity (4 items): Cronbach’s α values ranged from .88 to .92 across different study waves and vendor platforms. Composite reliability (CR) was reported above .91.
• Benevolence (4 items): Cronbach’s α values ranged from .87 to .91; Composite reliability exceeded .90.
• Ability / Competence (5 items): Cronbach’s α values ranged from .89 to .94; Composite reliability consistently surpassed .92.
• Overall Multidimensional Trust Construct: When operationalized as a higher-order construct or composite scale, the overall Cronbach’s α routinely exceeds .94.

Subsequent replications in global contexts (e.g., cross-cultural e-commerce studies across North America, Europe, and Asia) have consistently replicated these high internal consistency metrics, with subscale alphas rarely falling below .85. These findings verify that the instrument’s measurement error is minimal and that the operationalized items consistently capture their intended latent dimensions.

Factor Analysis

The underlying dimensionality of the OTS was thoroughly analyzed by Gefen et al. (2003) using both exploratory and confirmatory factor analytic procedures:

Exploratory Factor Analysis (EFA)

During initial instrument development, principal components analysis with varimax orthogonal rotation and promax oblique rotation was conducted on the 13 trust items alongside technological items from the TAM inventory. The resulting factor patterns revealed clear, clean factor solutions where:

• Items 1 through 4 cleanly loaded onto a distinct factor corresponding to Integrity (eigenvalues > 1.0, accounting for significant unique variance).
• Items 5 through 8 loaded onto a separate factor capturing Benevolence.
• Items 9 through 13 loaded cleanly onto a third factor capturing Ability.

No item exhibited substantial cross-loadings above the established cutoff of .40 on non-target factors, demonstrating clean primary structural clustering.

Confirmatory Factor Analysis (CFA)

To confirm the tripartite dimensional structure, CFA was performed using maximum likelihood estimation in LISREL. The three-factor first-order measurement model demonstrated acceptable to superior goodness-of-fit indices:

• Chi-Square to Degrees of Freedom Ratio (χ²/df): Ranged between 1.50 and 2.40 across various validation subsamples, well within the conventional ≤ 3.0 standard.
• Comparative Fit Index (CFI): Consistently exceeded .95 (ranging from .95 to .98).
• Non-Normed Fit Index (NNFI / TLI): Consistently exceeded .94.
• Root Mean Square Error of Approximation (RMSEA): Ranged between .042 and .061, with 90% confidence intervals staying below the critical .08 threshold.
• Standardized Root Mean Square Residual (SRMR): Maintained below .05.

Comparative structural modeling revealed that the three-factor model provided a significantly superior fit over both a one-factor unconstrained model (Δχ² test significant at p < .0001) and various two-factor models (e.g., combining Integrity and Benevolence into a single factor). These results confirm that while consumers perceive the three dimensions as related components of an overarching trust posture, they process vendor integrity, vendor benevolence, and vendor operational ability as distinct cognitive attributes.

Instrument / Measurement Tool

• Construct Measured: Consumer cognitive trust in an online vendor across three dimensions (Integrity, Benevolence, Ability/Competence).
• Theoretical Basis: Mayer, Davis, and Schoorman (1995) Integrative Model of Organizational Trust; Davis (1989) Technology Acceptance Model.
• Administration Format: Self-administered paper-and-pencil or digital/online survey questionnaire.
• Item Count: 13 items total:
  • Integrity Subscale: 4 items (Items 1–4)
  • Benevolence Subscale: 4 items (Items 5–8)
  • Ability / Competence Subscale: 5 items (Items 9–13)
• Target Population: Consumers, internet users, and organizational buyers engaging in or considering commercial transactions on web-based platforms.
• Administration Time: Approximately 3 to 5 minutes.
• Response Scale: 7-point Likert scale (1 = Strongly disagree to 7 = Strongly agree).
• Scoring and Aggregation Rules:
  • All items are positively keyed (no reverse-scored items).
  • Subscale scores are computed by calculating the arithmetic mean of the items comprising each subscale: Integrity (Items 1–4), Benevolence (Items 5–8), and Ability/Competence (Items 9–13).
  • An overall Online Trust Composite Score can be derived by computing the grand mean across all 13 items, where higher composite scores represent greater consumer trust in the vendor.
• Interpretation Guidelines:
  • Scores from 1.00 to 2.99 indicate low trust, high skepticism, and significant perceived vulnerability.
  • Scores from 3.00 to 4.99 represent moderate or ambiguous trust postures, where consumer adoption may depend heavily on structural guarantees and third-party assurances.
  • Scores from 5.00 to 7.00 indicate strong positive trust perceptions, supporting continuous usage and transactional execution.

Permissions & Fee and Test Year

The Online Trust Scale items were published in the peer-reviewed scholarly literature by David Gefen, Elena Karahanna, and Detmar W. Straub in 2003 in MIS Quarterly (Volume 27, Issue 1, pp. 51–90). The copyright for the academic article is held by the Management Information Systems Research Center (MISRC) at the University of Minnesota.

The instrument is widely available for non-commercial academic research, pedagogical use, and scholarly investigation under standard academic fair-use guidelines, provided appropriate theoretical attribution and formal bibliographic citation are given to Gefen et al. (2003). Organizations seeking to implement the scale for proprietary commercial evaluation, software integration, or commercial consulting should review institutional permissions and fair-use stipulations governed by the original publisher.

References

• Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T
• Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS Quarterly, 27(1), 51–90. https://doi.org/10.2307/30036519
• Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi.org/10.2307/258792
• McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). Developing and validating trust measures for e-commerce: An integrative typology. Information Systems Research, 13(3), 334–359. https://doi.org/10.1287/isre.13.3.334.81
• Pavlou, P. A. (2003). Consumer acceptance of electronic commerce: Integrating trust and risk with the technology acceptance model. International Journal of Electronic Commerce, 7(3), 101–134. https://doi.org/10.1080/10864415.2003.11044275
• Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903. https://doi.org/10.1037/0021-9010.88.5.879

Items of the Scale