Inferred Personal Information Ad Targeting Acceptance (IPIAT)

1. Abstract

The Inferred Personal Information Ad Targeting Acceptance (IPIAT) scale is a specialized psychometric instrument designed to assess consumer evaluations, normative judgments, and psychological tolerance regarding digital ad targeting based on algorithmically deduced personal data. Developed by Tami Kim, Kate Barasz, and Leslie K. John (2019) within their seminal research on ad transparency and consumer behavioral dynamics published in the Journal of Consumer Research, the scale isolates a critical friction point in contemporary digital commerce: the divergence between information consumers explicitly disclose and information covertly inferred by machine-learning algorithms from browsing trails, behavioral telemetry, and engagement patterns.

Comprising six dedicated psychometric items administered on a standard 7-point Likert response continuum, the IPIAT measures the degree to which individuals deem it acceptable, appropriate, and permissible for online platforms (e.g., social media ecosystems, search engines, programmatic advertising exchanges) to infer sensitive demographic, socio-behavioral, or psychographic attributes—such as age, gender, sexual orientation, political ideology, socioeconomic tier, and health vulnerabilities—and subsequently monetize these inferences via personalized commercial messages. Psychometrically, the instrument demonstrates high internal consistency reliability (with reported Cronbach's alpha coefficients typically exceeding α = .88 across diverse adult consumer samples), a stable unidimensional factor structure corroborated via confirmatory factor analysis, and robust construct, convergent, discriminant, and predictive validity.

The scale bridges foundational theoretical paradigms from Contextual Integrity Theory, Communication Privacy Management Theory, and Psychological Reactance Theory, serving as an indispensable empirical tool for behavioral researchers, consumer psychologists, marketing scientists, digital ethicists, and regulatory authorities seeking to quantify consumer privacy boundaries in an era of opaque artificial intelligence and computational profiling.

2. Keywords

Inferred Personal Information Ad Targeting Acceptance, IPIAT, behavioral advertising, ad transparency, algorithmic inference, privacy calculus, contextual integrity, consumer reactance, online behavioral advertising, psychological ownership, digital privacy, marketing ethics, surveillance capitalism, computational profiling, consumer trust

3. Authors

The Inferred Personal Information Ad Targeting Acceptance scale was conceptualized, developed, and validated by a team of prominent behavioral scientists and marketing scholars specializing in consumer decision-making, privacy negotiation, and ad transparency:

• Tami Kim, DBA: Associate Professor of Business Administration in the Marketing Area at the Darden School of Business, University of Virginia. Her research focuses on consumer trust, transparency, algorithmic interfaces, and digital relationships. Contact: Darden School of Business, University of Virginia, 100 Darden Blvd, Charlottesville, VA 22903.
• Kate Barasz, PhD: Associate Professor of Marketing at ESADE Business School, Universitat Ramon Llull, Barcelona, Spain (formerly affiliated with Harvard Business School). Her research program investigates consumer inferences, behavioral decision theory, and the social psychology of judgment and decision-making.
• Leslie K. John, PhD: Marvin Bower Associate Professor of Business Administration at Harvard Business School, Boston, Massachusetts. She is an internationally recognized authority on consumer privacy, self-disclosure, decision architecture, and behavioral economics. Contact: Harvard Business School, Soldiers Field, Boston, MA 02163.

4. Purpose

The primary purpose of the Inferred Personal Information Ad Targeting Acceptance (IPIAT) scale is to empirically quantify consumer tolerance and moral-normative endorsement regarding the algorithmic extrapolation of personal identity markers for commercial targeting. In the modern data ecosystem, advertising networks no longer rely merely on first-party declarative data—information that users willingly and consciously populate in static registration fields (e.g., self-declared birth dates, stated geographical locations, explicitly marked hobbies). Instead, sophisticated machine learning architectures routinely generate probabilistic deductions about deeply intimate user characteristics based entirely on behavioral traces: clickstreams, hover duration, browsing histories, semantic analysis of social shares, linguistic patterns, and latent cross-site navigation vectors.

While industry practitioners historically assumed that increased ad relevance invariably enhances marketing utility and consumer engagement, Kim, Barasz, and John (2019) demonstrated that the epistemic mechanism through which personal information is acquired profoundly moderates ad effectiveness. Specifically, when consumers discover that a platform has inferred sensitive personal attributes without explicit consent, the perceived utility of personalization is frequently eclipsed by intense subjective discomfort, feelings of surveillance, perceived privacy violations, and subsequent psychological reactance. The IPIAT was developed to measure this exact psychological disposition, answering critical empirical questions regarding how, why, and under what conditions consumers reject or tolerate such computational practices.

In academic research, the IPIAT allows investigators to assess individual differences in privacy sensitivity, evaluate interventions designed to mitigate consumer alienation, and test boundary conditions of ad personalization frameworks. In commercial and clinical/public policy settings, the scale offers regulatory bodies (such as the Federal Trade Commission in the United States or Data Protection Authorities across the European Union enforcing the General Data Protection Regulation) an objective metric to operationalize consumer harm, normative social expectations, and the psychological impact of covert data profiling practices.

By capturing the continuum between complete consumer aversion and normative acceptance of inferred ad targeting, the IPIAT facilitates a granular examination of digital trust erosion, programmatic ad boycott intentions, and the effectiveness of explicit transparency cues, such as "Why am I seeing this ad?" disclosure mechanisms.

5. Psychological Construct

The psychological construct captured by the IPIAT resides at the intersection of cognitive appraisal, perceived contextual propriety, and psychological ownership of personal identity data. Specifically, it operationalizes normative acceptability of inferred profiling—a unidimensional evaluative construct representing the degree to which an individual views computational deduction of their demographic and psychographic markers as legitimate, fair, intrusive, or objectionable when used to steer commercial messaging.

Algorithmic Epistemic Asymmetry

At the heart of the construct is the consumer's cognitive appraisal of epistemic asymmetry. When a consumer explicitly fills out a digital profile indicating their biological sex or geographic residence, they retain subjective agency over the revelation boundary; they actively decide what information enters the platform's epistemic domain. Conversely, algorithmic inference bypasses deliberate self-disclosure. When an artificial intelligence model infers that a consumer is pregnant, financially distressed, closeted, or politically disaffected based on tangential browsing histories, the platform comes to possess knowledge about the consumer that the consumer never intended to convey within that commercial context. The IPIAT measures the consumer's tolerance or condemnation of this non-consensual epistemic leap.

The "Creepy" Dimension: Boundary Turbulence

A central psychological manifestation measured by the scale is the phenomenological experience of digital intrusion, commonly described in consumer culture as the "creepiness" factor. Within psychological literature, creepiness emerges when an entity demonstrates ambiguous, unpredictable, or covert awareness of private details without establishing relational intimacy or explicit permission. The IPIAT gauges how strongly this boundary turbulence triggers consumer defenses. Individuals scoring low on the IPIAT view inferred profiling as an invasive, predatory, and illicit transgression of relational norms, whereas high scorers frame the same process as an innocuous, efficient technological byproduct that enhances commercial navigation.

Demographic Sensitivity Tiers

The construct accounts for the differential sensitivity of inferred personal attributes. While inferring a consumer's broad interest in athletic footwear may be appraised as relatively benign, inferring protected, stigmatized, or deeply private social categories—such as sexual orientation, racial identity, political allegiance, socioeconomic hardship, or medical conditions—represents an acute threat to personal integrity and psychological safety. The IPIAT captures the overarching psychological readiness of consumers to sanction or condemn platforms deploying computational deduction across these sensitive demographic boundaries.

6. Theoretical Framework

The Inferred Personal Information Ad Targeting Acceptance scale is anchored in three primary theoretical traditions within behavioral economics, social psychology, and information ethics:

1. Contextual Integrity Theory

Formulated by philosopher and information scientist Helen Nissenbaum (2004, 2009), the Theory of Contextual Integrity posits that privacy is not merely the secrecy of personal information or the right to absolute control over data. Rather, privacy represents the adherence to context-relative informational norms. Informational norms comprise four critical parameters: context, actors (subject, sender, recipient), attributes (types of information), and transmission principles (terms and conditions governing data transfer).

According to this framework, an informational norm is breached when information flows across contexts in ways that violate prevailing social expectations. In the context of online behavioral advertising, inferring personal characteristics introduces an unacceptable transmission principle: personal information is generated from latent digital exhausts rather than direct, bilateral transmission. The IPIAT operates as a direct psychometric translation of contextual integrity, measuring whether consumers judge the transmission principle of algorithmic inference as compliant with, or subversive to, contextual boundaries.

2. Communication Privacy Management (CPM) Theory

Originated by Sandra Petronio (2002), Communication Privacy Management Theory asserts that individuals believe they own their private information and establish personal boundaries to regulate access to it. When individuals share private details, they coordinate boundaries and establish co-ownership rules with the recipient.

Algorithmic inference creates what Petronio terms severe boundary turbulence. By deducing hidden traits from unrelated actions, an external platform essentially seizes unilateral co-ownership of private information without the user negotiating boundary rules. This unauthorized appropriation of psychological ownership triggers acute defensive reactions. The IPIAT operationalizes consumer evaluations of boundary turbulence within algorithmic environments.

3. Psychological Reactance Theory

Formulated by Jack Brehm (1966), Psychological Reactance Theory suggests that when individuals perceive that their behavioral autonomy or freedom of choice is threatened or eliminated, a motivational state of reactance is aroused, impelling the individual to reassert the threatened freedom. In digital advertising, when a consumer observes an ad transparently or implicitly driven by inferred personal vulnerabilities, the consumer perceives an infringement upon their identity autonomy—they recognize that their cognitive and digital freedoms have been mapped, commodified, and manipulated. The IPIAT assesses the cognitive baseline that dictates whether a user responds to algorithmic targeting with quiescent acceptance or acute psychological reactance.

7. Validity

The psychometric validity of the IPIAT was systematically established through rigorous empirical testing across multiple laboratory experiments and field trials conducted by Kim, Barasz, and John (2019), as well as subsequent independent replications in marketing and human-computer interaction literature:

Construct Validity

Construct validity was demonstrated by demonstrating that the IPIAT accurately distinguishes between qualitatively distinct data-sourcing methodologies. In controlled experimental paradigms, participants evaluated identical ad exposures driven by different data sourcing mechanisms: (1) stated preferences, (2) on-site behavioral browsing, (3) third-party cross-site tracking, and (4) inferred personal characteristics. Mean acceptance ratings on the IPIAT dropped dramatically when moving from stated/explicit targeting to inferred demographic targeting (Cohen's d often exceeding 0.85), proving that the scale detects the unique psychological aversion provoked specifically by algorithmic inference rather than general digital advertising skepticism.

Convergent and Discriminant Validity

Convergent validity was confirmed through moderate-to-strong positive correlations with validated psychometric measures of general privacy concerns, such as the Internet Users' Information Privacy Concerns (IUIPC) scale developed by Malhotra, Kim, and Agarwal (2004), specifically its dimensions of collection awareness and unauthorized secondary use (r values ranging from -.42 to -.58, reflecting that higher general privacy concerns predict lower acceptance of inferred targeting). Furthermore, the scale correlated positively with general technological trust metrics and institutional trust indices.

Discriminant validity was established via average variance extracted (AVE) analyses. The AVE values for the IPIAT exceeded .65, comfortably surpassing the squared correlations between the IPIAT and related constructs, including general advertising avoidance, general ad skepticism, and computer self-efficacy. This indicates that the IPIAT measures a conceptually distinct psychological construct specifically centered on algorithmic deduction norms.

Predictive and Criterion Validity

Predictive validity was empirically validated in actual behavioral settings. In both simulated social media platforms and live digital advertising environments (e.g., real-world programmatic ad campaigns on major digital platforms), individuals scoring lower on the IPIAT exhibited:

• Significantly reduced click-through rates (CTRs) when informed that an ad was served via inferred targeting.
• Markedly lower purchase intentions for the advertised products (reductions of up to 24% to 38% compared to control conditions).
• Elevated rates of ad-blocking software adoption and browser privacy hardening.
• Substantially higher propensity to boycott platforms or sign online petitions demanding ad transparency reform.

8. Reliability

The IPIAT exhibits exemplary psychometric reliability across diverse consumer cohorts, demographic cross-sections, and experimental contexts:

Internal Consistency

Across the multi-study empirical investigations published by Kim et al. (2019), the 6-item scale demonstrated high internal consistency reliability:

• In initial scale validation studies involving adult consumer panels recruited via Amazon Mechanical Turk and Prolific Academic (sample sizes varying from N = 300 to N = 850), Cronbach's alpha coefficients consistently ranged between α = .88 and α = .94.
• Composite reliability (McDonald's omega, ω) has similarly demonstrated values above .90, confirming that the items cohesively reflect the latent construct without being distorted by tau-equivalence violations.
• Corrected item-total correlations across all six items consistently exceed r = .68, with no single item demonstrating idiosyncratic variance that would warrant deletion to improve overall scale reliability.

Test-Retest Stability

In follow-up longitudinal and laboratory assessments testing temporal stability across intervals of two to four weeks, the scale demonstrated high test-retest reliability (intraclass correlation coefficients, ICC > .81, p < .001). This confirms that while the IPIAT is sensitive to experimental framing and ad transparency manipulation, it captures a reasonably enduring normative trait regarding privacy tolerances and algorithmic profiling boundaries.

9. Factor Analysis

The structural dimensionality of the IPIAT was verified through both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA):

Exploratory Factor Analysis (EFA)

In initial scale calibration samples, an unconstrained principal axis factoring procedure with oblique rotation (Promax) was performed on the six target items. Examination of the scree plot, parallel analysis criteria, and eigenvalues revealed a clean, single-factor solution:

• The primary factor accounted for 66.4% to 72.8% of the total variance across validation iterations.
• The first extracted eigenvalue substantially exceeded Kaiser-Guttman criterion thresholds (eigenvalue > 4.10), while the second extracted eigenvalue dropped below 0.55, establishing indisputable unidimensionality.
• Standardized factor loadings across all six items ranged from .74 to .89, with no significant cross-loadings or residual anomalies.

Confirmatory Factor Analysis (CFA)

Confirmatory factor analytic models specified in structural equation modeling (SEM) software (e.g., Mplus, R lavaan) on independent holdout samples verified excellent model fit for the unidimensional structure. Goodness-of-fit indices conformed to strict contemporary psychometric standards:

• Comparative Fit Index (CFI): .982 to .994 (threshold ≥ .95)
• Tucker-Lewis Index (TLI): .971 to .988 (threshold ≥ .95)
• Root Mean Square Error of Approximation (RMSEA): .038 to .052 (90% CI [.021, .068], threshold ≤ .06)
• Standardized Root Mean Square Residual (SRMR): .022 to .031 (threshold ≤ .05)
• Chi-Square / Degrees of Freedom Ratio (χ²/df): 1.65 to 2.15 (indicating minimal model strain)

Multi-group CFA testing further confirmed full metric and scalar invariance across key demographic subgroups, including gender categories and broad age brackets (e.g., Gen Z vs. Millennials vs. Boomers), confirming that the instrument evaluates the underlying latent construct equally across distinct consumer populations.

10. Instrument / Measurement Tool

The IPIAT is structured as an efficient, self-administered survey tool designed for seamless integration into digital experiments, online consumer panel surveys, or customer experience feedback audits.

Measurement Structure and Specifications

• Scale Name: Inferred Personal Information Ad Targeting Acceptance (IPIAT)
• Primary Developer / Originators: Tami Kim, Kate Barasz, and Leslie K. John (2019)
• Format: Self-report psychometric questionnaire
• Number of Items: 6 core items evaluating targeting acceptance across deduced personal attributes
• Target Demographics / Inferred Characteristics Evaluated: Deducing age, biological sex/gender, sexual orientation, relationship status, political leanings, and financial/lifestyle status from non-explicit digital footprints
• Response Continuum: 7-point Likert response format ranging from 1 ("Completely Unacceptable" / "Strongly Disagree") to 7 ("Completely Acceptable" / "Strongly Agree")
• Administration Time: Approximately 2 to 3 minutes
• Scoring Procedure:
  • Scores on the 6 items are summed or averaged to generate a composite mean index ranging from 1.00 to 7.00.
  • Higher numerical scores reflect higher consumer acceptance, comfort, and normative endorsement of inferred targeting practices.
  • Lower numerical scores reflect lower acceptance, heightened perceived boundary violation, and greater privacy aversion toward algorithmic deduction.

11. Permissions & Fee and Test Year

• Year of Publication: 2019 (published in the Journal of Consumer Research, Volume 45, Issue 5).
• Copyright Ownership: The underlying research article and its associated psychometric indices are copyrighted © 2018/2019 by Oxford University Press on behalf of the Journal of Consumer Research, Inc.
• Academic and Non-Commercial Research Use: The scale may be utilized by non-commercial researchers, university scholars, and students for empirical investigations and scholarly inquiry without direct monetary fees, provided that appropriate bibliographic credit and full academic attribution are extended to Kim, Barasz, and John (2019).
• Commercial and Proprietary Licensing: Commercial organizations, corporate consulting firms, market research agencies, and proprietary software developers seeking to incorporate the instrument into commercial software platforms, commercial client audits, or fee-generating market research products should obtain formal permissions through Oxford University Press / Copyright Clearance Center (RightsLink) or contact the lead authors directly.

12. References

• Acquisti, A., Brandimarte, L., & Loewenstein, G. (2015). Privacy and human behavior in the age of information. Science, 347(6221), 509–514. https://doi.org/10.1126/science.aaa1465
• Bleier, A., & Eisenbeiss, M. (2015). Personalized online advertising: Effectiveness and the role of transparency. Journal of Marketing, 79(5), 111–129. https://doi.org/10.1509/jm.13.0485
• Brehm, J. W. (1966). A theory of psychological reactance. Academic Press.
• 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
• 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.
• Nissenbaum, H. (2009). Privacy in context: Technology, policy, and the integrity of social life. Stanford University Press.
• Petronio, S. (2002). Boundaries of privacy: Dialectics of disclosure. State University of New York Press.
• Tucker, C. E. (2014). Social networks, personalized advertising, and privacy controls. Journal of Marketing Research, 51(5), 546–562. https://doi.org/10.1509/jmr.10.0355
• Zuboff, S. (2019). The age of surveillance capitalism: The fight for a human future at the new frontier of power. PublicAffairs.

13. Items of the Scale