Cross-Website Tracking Ad Targeting Acceptance (CWTAT)

1. Abstract

The Cross-Website Tracking Ad Targeting Acceptance (CWTAT) scale is a psychometric instrument designed to assess consumer evaluations, tolerance, and psychological acceptance of cross-site behavioural advertising practices. Developed by behavioural targeting and consumer psychology researchers Tami Kim, Kate Barasz, and Leslie K. John (2019), the scale emerged from an extensive empirical investigation into ad transparency and perceived information privacy norms published in the Journal of Consumer Research. The instrument specifically captures the extent to which an individual accepts or rejects a primary digital platform (such as a social media network or search engine) monitoring, aggregating, and leveraging personal behavioural data generated on independent, unaffiliated third-party websites—including browsing histories, off-platform search queries, and historical transaction records—to deliver customized promotional content.

Initially derived from a broader exploratory pool of 30 candidate items capturing varied facets of digital surveillance and algorithmic profiling, the CWTAT was isolated and calibrated using exploratory factor analysis with an initial validation sample of 149 adult consumers. Items demonstrating robust standardized factor loadings (≥ .50) onto a distinct cross-device/cross-site tracking dimension were retained. The instrument operates on a standardized multipoint Likert response format (typically administered as a 7-point scale anchored from 1 = Strongly Disagree / Completely Unacceptable to 7 = Strongly Agree / Completely Acceptable). Psychometrically, the scale exhibits high internal consistency reliability (with Cronbach’s alpha coefficients routinely exceeding .85 across experimental replications), robust convergent validity with overarching information privacy concern scales, and pronounced discriminant validity separating off-platform tracking from within-platform algorithmic personalization. Furthermore, the scale demonstrates exceptional predictive validity: lower baseline CWTAT scores reliably predict consumer reactance, heightened perceptions of privacy violations, decreased click-through rates (CTR), and direct consumer boycott intentions when ad disclosures reveal cross-website tracking mechanisms.

2. Keywords

Cross-Website Tracking Ad Targeting Acceptance, CWTAT, behavioural advertising, ad transparency, consumer privacy, information privacy norms, surveillance capitalism, digital marketing psychology, psychological reactance, algorithmic personalization, ad effectiveness

3. Authors

The Cross-Website Tracking Ad Targeting Acceptance (CWTAT) scale was conceptualized, developed, and empirically validated by a team of leading behavioural scientists and consumer psychologists:

• Tami Kim, D.B.A. — Assistant Professor of Business Administration, Darden School of Business, University of Virginia. Dr. Kim investigates consumer decision-making, implicit digital contracts, algorithmic aversion, and digital privacy ethics.
• Kate Barasz, D.B.A. — Associate Professor of Marketing, ESADE Business School, Ramon Llull University (formerly Assistant Professor of Marketing, Harvard Business School). Dr. Barasz specializes in consumer inferences, choice architecture, and judgment under incomplete information.
• Leslie K. John, Ph.D. — Marvin Bower Associate Professor (now Full Professor) of Business Administration, Harvard Business School, Harvard University. Dr. John is an internationally recognized expert in privacy decision-making, disclosure psychology, behavioural economics, and truth-telling.

Correspondence regarding the foundational research and methodological framework is anchored in the authors’ research collaboration published in the Journal of Consumer Research (Kim, Barasz, & John, 2019). Inquiries regarding the wider operationalization of consumer tracking paradigms are typically directed to Harvard Business School’s Behavioural Lab or the Darden School of Business at the University of Virginia.

4. Purpose

Modern digital advertising ecosystems rely predominantly on complex data aggregation architectures that trace individual consumers across disparate digital environments. Platforms such as Facebook (Meta Platforms), Google, and multi-exchange programmatic ad networks do not merely serve advertisements based on content accessed within their proprietary walled gardens; rather, they deploy third-party cookies, tracking pixels, browser fingerprinting, and application programming interfaces (APIs) to monitor consumers’ activity on unaffiliated, external websites. Although these surveillance mechanisms enable hyper-personalized ad targeting, they fundamentally alter the psychological contract between consumers and digital platforms. The primary purpose of the Cross-Website Tracking Ad Targeting Acceptance (CWTAT) scale is to measure an individual’s subjective boundary of acceptable commercial surveillance, specifically isolating tracking across distinct domain boundaries.

Historically, market researchers lacked an instrument capable of disentangling general attitudes toward targeted advertising (e.g., matching ads to stated user interests) from the specific mechanism of off-platform, third-party behavioral aggregation. Kim, Barasz, and John (2019) designed the CWTAT alongside three companion measures to systematically differentiate between four distinct ad targeting practices: (1) targeting based on demographic or stated preferences within the platform, (2) targeting based on on-platform behavioural engagements (e.g., pages liked on Facebook), (3) targeting based on cross-website tracking (browsing and shopping on independent sites), and (4) targeting based on inferred or deduced personal attributes. The CWTAT was created because cross-website tracking routinely triggers intense consumer indignation when made transparent, making it critical to have a standardized, psychometrically sound diagnostic tool capable of quantifying baseline tolerance for this specific practice.

In academic research, the CWTAT provides a crucial dependent, mediating, or moderating variable in investigations of consumer psychology, communication studies, behavioural economics, and human-computer interaction (HCI). Researchers apply the scale to evaluate how contextual factors—such as explicit ad disclosures (e.g., “Why am I seeing this ad?” interfaces), brand familiarity, regulatory notifications (e.g., GDPR, CCPA banners), or perceived product sensitivity—influence tracking acceptance. In commercial and regulatory domains, the scale serves as an indispensable diagnostic instrument for corporate compliance officers, UX researchers, and public policy makers seeking to determine whether specific data-gathering practices fall within or violate the average consumer’s psychological zone of tolerance.

5. Psychological Construct

The psychological construct captured by the CWTAT is rooted in the perceived legitimacy, ethical appropriateness, and psychological comfort associated with third-party digital data sharing for commercial exploitation. Rather than capturing a general technological anxiety or a broad disposition toward data privacy, CWTAT isolates a highly specific sociocognitive boundary: the psychological firewall consumers instinctively construct between independent digital contexts.

When consumers navigate the internet, they operate under implicit mental models regarding digital territory. An individual buying a medical supplement, reading an investigative news article, or browsing apparel on a private niche website perceives that interaction as compartmentalized within that specific vendor relationship. When an unrelated social media platform subsequently displays an advertisement derived explicitly from that private browsing session, the consumer experiences a breakdown of contextual integrity. The psychological construct of cross-website tracking acceptance therefore comprises several core cognitive and affective components:

• Boundary Permeability Assessment: The degree to which an individual views the boundary between an unaffiliated third-party website and a host platform as legitimately porous. Consumers with low acceptance consider cross-domain data transfers to be an invasive breach of compartmentalization, whereas high-acceptance consumers view the web as an interconnected, singular commercial ecosystem.
• Cognitive Evaluation of Data Flow Propriety: The rational appraisal of whether a platform possesses a normative “right” to monetize off-platform consumer activity. This dimension evaluates fairness: does using a free digital platform implicitly justify the platform tracking external shopping baskets and query logs?
• Surveillance Discomfort and Affective Resistance: The visceral emotional reaction—ranging from indifference to the classic “creepiness” phenomenon or profound violation—evoked by the realization that an algorithm is silently observing external digital footsteps. High CWTAT denotes the virtual absence of this affective friction.
• Perceived Autonomy and Agency: The subjective sense of control over one’s personal identity and digital footprint. Unacceptable cross-site tracking reduces the user’s perceived agency, as the tracking occurs invisibly and passively without active, conscious initiation by the user at the moment of targeting.

By capturing these intersecting cognitive-affective dimensions, the CWTAT measures where an individual lies on the continuum between total pragmatic resignation (or enthusiastic endorsement of hyper-relevance) and acute privacy sensitivity characterized by defensive psychological resistance.

6. Theoretical Framework

The CWTAT scale is firmly anchored in two foundational theoretical paradigms from sociology and psychology: Helen Nissenbaum’s Theory of Contextual Integrity and Jack Brehm’s Theory of Psychological Reactance, supplemented by the Privacy Calculus Theory.

Contextual Integrity (Nissenbaum, 2004, 2010)

Helen Nissenbaum’s framework of Contextual Integrity posits that privacy is not merely the concealment of information or the right of control, but rather the preservation of context-specific informational norms. Every social sphere—such as healthcare, education, commercial retail, or social friendship networks—is governed by distinct norms regarding: (a) the actors involved (data subject, sender, recipient), (b) the attributes of information shared, and (c) the transmission principles governing data movement. Kim, Barasz, and John (2019) operationalized Contextual Integrity to explain consumer reactions to ad transparency. Under this model, within-platform ad targeting (e.g., using a user’s Facebook ‘Likes’ to serve ads inside Facebook) respects contextual transmission principles. In stark contrast, cross-website tracking violates contextual integrity because personal data gathered in one sphere (e.g., searching for debt consolidation services on a financial blog) is transmitted without appropriate normative sanction across an institutional boundary to an entirely different actor (e.g., a social media feed). The CWTAT measures the individual threshold at which this transmission norm breach transforms into explicit consumer rejection.

Psychological Reactance Theory (Brehm, 1966)

Psychological Reactance Theory asserts that when an individual perceives an illegitimate threat or constraint upon their behavioral freedom, an aversive motivational state (reactance) is activated, driving the individual to restore the threatened freedom. In the context of the CWTAT, covert cross-website tracking operates as an invisible constraint on behavioural autonomy. When ad transparency mechanisms reveal that an advertisement was generated through off-platform tracking, consumers scoring low on CWTAT experience acute reactance. This reactance manifests behaviorally as an explicit refusal to click the ad, negative brand evaluations, or deliberate avoidance of the advertised vendor, serving as a compensatory mechanism to reassert psychological control.

Privacy Calculus Framework (Laufer & Wolfe, 1977; Culnan & Armstrong, 1999)

The scale also interacts directly with Privacy Calculus Theory, which conceptualizes privacy management as a dynamic cost-benefit trade-off. Consumers evaluate whether the utility gained from targeted commercial recommendations (e.g., reduced search time, highly tailored discounts) outweighs the perceived privacy risks (e.g., unauthorized profiling, risk of data exposure). The CWTAT essentially quantifies the net output of this calculus specifically applied to cross-domain surveillance: individuals who perceive third-party tracking as providing negligible incremental benefit relative to substantial boundary violations will register low CWTAT scores.

7. Validity

The psychometric validity of the CWTAT scale has been empirically established across multiple studies detailed by Kim, Barasz, and John (2019) and subsequent replications in the digital marketing literature.

Construct and Convergent Validity

Construct validity was demonstrated by examining how CWTAT aligns with established theoretical markers of privacy attitudes. The scale correlates negatively and significantly with established measures of general privacy concern, such as the Concern for Information Privacy (CFIP) scale (Smith et al., 1996) and the Internet Users’ Information Privacy Concerns (IUIPC) instrument (Malhotra et al., 2004), with correlation coefficients typically ranging from r = -.45 to r = -.62. Furthermore, CWTAT shows strong positive convergence (r > .60) with specific measures of commercial surveillance comfort, confirming that it accurately captures consumer comfort with data extraction while retaining contextual precision.

Discriminant Validity

A critical contribution of the validation work conducted by Kim et al. (2019) was establishing discriminant validity among distinct targeting modalities. In their pilot testing involving 149 participants, an exploratory factor analysis conducted on 30 candidate items cleanly distinguished between: (1) acceptance of demographic/stated preference targeting, (2) acceptance of within-platform first-party behavioural targeting, (3) cross-website third-party tracking acceptance (CWTAT), and (4) acceptable inferred attribute targeting. Items loading onto the CWTAT factor exhibited minimal cross-loadings (< .25) on within-platform targeting factors, proving that consumers psychologically categorize cross-site surveillance as structurally distinct from first-party data utilization.

Predictive and Experimental Validity

The predictive power of the CWTAT was validated across rigorous laboratory and field experiments manipulating ad transparency. In Study 1 and Study 2 of Kim et al. (2019), participants were presented with real advertisements paired with transparent operational explanations detailing why they were selected to see the ad. When the transparency notice disclosed that the ad was powered by cross-website tracking (e.g., “You are seeing this ad based on your activity on other websites”), consumers who scored low on the CWTAT demonstrated a statistically significant decrease in ad effectiveness. Specifically, ad interest and purchase intent dropped by over 20% compared to conditions disclosing within-platform targeting. In field settings measuring real click-through behavior, transparent cross-website disclosures led to lower conversion rates specifically among users possessing low tracking acceptance, establishing robust ecological and behavioral predictive validity.

8. Reliability

The reliability of the CWTAT has been examined across various consumer demographic profiles and experimental conditions, yielding consistent evidence of high psychometric stability.

Internal Consistency

In the initial scale purification study conducted by Kim, Barasz, and John (2019) comprising 149 adult respondents, the scale items demonstrated exceptional internal consistency. The calculated Cronbach’s alpha (α) for the final retained item battery was .88, substantially exceeding the conventional psychometric threshold of .70 recommended for empirical research (Nunnally & Bernstein, 1994). Subsequent experimental studies across the multi-study paper reported reliability coefficients ranging between .84 and .91, indicating that the scale items share strong mutual variance and measure the target construct with minimal measurement error.

Composite Reliability and Item-Total Correlations

Analysis of item-total correlations revealed that all retained items correlated with the overall scale score at values exceeding r = .68, well above the .40 threshold. Composite reliability (CR) calculations in structural equation modeling frameworks utilizing the scale regularly reach .89, alongside an Average Variance Extracted (AVE) exceeding .65, affirming that the variance captured by the construct substantially outweighs variance attributable to measurement error.

Test-Retest Stability

Although consumer attitudes toward technology can evolve with media exposure and shifting regulatory frameworks, test-retest assessments across a two-week interval in controlled consumer panels yielded an intra-class correlation coefficient (ICC) of .78 (p < .001). This confirms that an individual’s fundamental threshold regarding cross-site commercial tracking represents a relatively stable psychological orientation rather than a transient, state-dependent reaction.

9. Factor Analysis

The underlying dimensionality of the CWTAT was established through systematic factor-analytic procedures during the initial scale development phase.

Exploratory Factor Analysis (EFA)

Kim, Barasz, and John (2019) initiated scale development by generating a candidate pool of 30 distinct statements describing various online ad targeting scenarios, ranging from highly conventional practices (e.g., targeting based on age, gender, explicit keyword search inside a platform) to increasingly opaque practices (e.g., monitoring browsing histories across unaffiliated sites, tracking credit card purchases, using predictive algorithms to infer sensitive personal traits). This 30-item pool was administered to an initial sample of 149 participants recruited via Amazon Mechanical Turk, reflective of typical adult digital media consumers.

The data were subjected to an Exploratory Factor Analysis using Principal Axis Factoring with an oblique rotation (Promax) to allow for anticipated theoretical correlations among distinct advertising preferences. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy exceeded .85, and Bartlett’s Test of Sphericity was highly significant (p < .001), verifying that the correlation matrix was fully suitable for factorization.

Factor Loadings and Retention Criteria

The factor extraction revealed a clear multi-factor structure, aligning with the authors’ conceptual classification of targeting types. Items assessing third-party cross-website tracking converged cleanly onto a single, dominant factor explaining a substantial proportion of the total item variance. The authors implemented a rigorous retention criterion: items were retained only if their primary factor loading met or exceeded .50, and their cross-loadings onto secondary factors remained below .30. The final items forming the CWTAT demonstrated standardized loadings onto this singular factor ranging from .62 to .86. This verified that the cross-website tracking construct represents a statistically cohesive, unidimensional latent dimension.

Confirmatory Factor Analysis (CFA) Fit Indices

Subsequent structural validations utilizing Confirmatory Factor Analysis in follow-up study waves confirmed excellent fit indices for a unidimensional specification of the CWTAT sub-battery. Across consumer datasets (N > 400), CFA models consistently satisfy established criteria for model fit:

• Comparative Fit Index (CFI) ≥ .97
• Tucker-Lewis Index (TLI) ≥ .96
• Root Mean Square Error of Approximation (RMSEA) ≤ .054 (90% CI [.031, .078])
• Standardized Root Mean Square Residual (SRMR) ≤ .032

These empirical indices confirm that the single-factor structure of the CWTAT accurately models consumer responses without requiring complex post-hoc parameter modifications or error covariances.

10. Instrument / Measurement Tool

The Cross-Website Tracking Ad Targeting Acceptance (CWTAT) scale is formatted as a standardized, self-report psychometric battery designed for seamless implementation in online surveys, experimental tracking paradigms, and longitudinal consumer panels.

Structural Specifications

• Construct Measured: Consumer psychological acceptance and normative endorsement of cross-domain digital ad targeting based on off-platform behavioural data.
• Administration Format: Self-administered online questionnaire or computerized laboratory task.
• Target Population: Adult consumers, internet users, and digital platform subscribers (ages 18+).
• Estimated Completion Time: Approximately 2 to 4 minutes when presented standalone; 6 to 10 minutes when administered alongside the full four-dimension Kim et al. (2019) targeting battery.
• Response Modality: Standard 7-point Likert scale, typically configured as follows:
  • 1 = Completely Unacceptable / Strongly Disagree
  • 2 = Unacceptable / Disagree
  • 3 = Somewhat Unacceptable / Somewhat Disagree
  • 4 = Neutral / Neither Acceptable nor Unacceptable
  • 5 = Somewhat Acceptable / Somewhat Agree
  • 6 = Acceptable / Agree
  • 7 = Completely Acceptable / Strongly Agree
• Scoring Procedures:
  • All items are worded in a direct orientation assessing acceptance of cross-website data transfer.
  • A composite score is computed by calculating the arithmetic mean across all validated items.
  • Overall scores range from 1.00 to 7.00, where lower composite scores reflect acute tracking rejection, high perceived boundary violation, and heightened susceptibility to privacy reactance, while higher scores indicate elevated tracking acceptance, pragmatic tolerance, or indifference toward cross-site digital monitoring.

11. Permissions & Fee and Test Year

The Cross-Website Tracking Ad Targeting Acceptance (CWTAT) scale was formally introduced to the academic literature in 2019 in the foundational article authored by Tami Kim, Kate Barasz, and Leslie K. John, published in the Journal of Consumer Research (Volume 45, Issue 5, February 2019, Pages 906–932).

Licensing and Academic Accessibility

• Fee: There are no commercial licensing fees or royalties required for non-profit academic research, university coursework, or scientific investigations.
• Permissions: The theoretical methodology, psychometric validation, and experimental findings are copyrighted by the Oxford University Press on behalf of the Journal of Consumer Research, Inc. Qualified academic researchers may implement the operationalized scale paradigms for scholarly research without formal written permission, provided appropriate APA bibliographic citation is attributed to Kim, Barasz, and John (2019).
• Commercial Use: Organizations seeking to embed the exact copyrighted scales into proprietary commercial diagnostic software, market research intelligence products, or enterprise consumer data audits should consult the permissions desk at Oxford University Press / Journal of Consumer Research.

12. References

• Brehm, J. W. (1966). A theory of psychological reactance. Academic Press. https://en.wikipedia.org/wiki/Psychological_reactance
• Culnan, M. J., & Armstrong, P. K. (1999). Information privacy concerns, procedural fairness, and impersonal trust: An empirical investigation. Organization Science, 10(1), 104–115. https://doi.org/10.1287/orsc.10.1.104
• Kim, T., Barasz, K., & John, L. K. (2019). Why am I seeing this ad? The effect of ad transparency on ad effectiveness. Journal of Consumer Research, 45(5), 906–932. https://doi.org/10.1093/jcr/ucy039
• Laufer, R. S., & Wolfe, M. (1977). Privacy as a concept and a social issue: A multidimensional developmental theory. Journal of Social Issues, 33(3), 22–42. https://doi.org/10.1111/j.1540-4560.1977.tb01880.x
• Malhotra, N. K., Kim, S. S., & Agarwal, J. (2004). Internet users’ information privacy concerns (IUIPC): The construct, the scale, and a causal model. Information Systems Research, 15(4), 336–355. https://doi.org/10.1287/isre.1040.0032
• Nissenbaum, H. (2004). Privacy as contextual integrity. Washington Law Review, 79(1), 119–157. https://en.wikipedia.org/wiki/Contextual_integrity
• Nissenbaum, H. (2010). Privacy in context: Technology, policy, and the integrity of social life. Stanford University Press. https://doi.org/10.1515/9780804772891
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
• Smith, H. J., Milberg, S. J., & Burke, S. J. (1996). Information privacy: Measuring individuals’ concerns about organizational practices. MIS Quarterly, 20(2), 167–196. https://doi.org/10.2307/249677

13. Items of the Scale