Attitude Toward the Person (Relational Liking) (ATPRS)

1. Abstract

The Attitude Toward the Person (Relational Liking) scale (often abbreviated as ATPRS) is a psychometric instrument designed to capture the affective evaluation, perceived interpersonal compatibility, and relational affinity an individual holds toward a specific interaction partner within a defined social, organizational, or commercial dyad. Rooted originally in organizational psychology research on leader-member exchange and impression management (Wayne & Ferris, 1990), the operationalization was refined and adapted into a parsimonious, modular three-item instrument by Hennig-Thurau, Marchand, and Marx (2012) within the consumer research and decision-science literature. The instrument isolates interpersonal liking—a foundational dimension of social relationship quality—by evaluating direct personal liking, prospective friendship appraisal, and perceived relational harmony.

The scale employs a 7-point Likert response format ranging from 1 (Strongly disagree) to 7 (Strongly agree), characterized by a modular slot syntax that allows researchers to substitute the generic referent with context-specific dyadic partners (e.g., co-consumer, financial adviser, physician, group recommender peer, or workplace subordinate). Psychometrically, the ATPRS demonstrates robust unidimensionality across diverse collaborative settings. Confirmatory factor analyses across published replications consistently indicate high standardized factor loadings (λ ≥ .85), negligible measurement error, and strong internal consistency, with Cronbach’s alpha and composite reliability (ρ_c) coefficients typically exceeding .88 to .94. Furthermore, the measure displays sharp discriminant validity against adjacent social relational constructs such as interpersonal trust, expertise asymmetry, and relational commitment. This article presents a comprehensive psychometric review of the ATPRS, encompassing its theoretical foundations, structural and predictive validity, factor stability, scoring rubrics, and research utilities.

2. Keywords

Relational liking, interpersonal attraction, social relationship quality, dyadic interaction, Attitude Toward the Person, ATPRS, consumer decision making, automated group recommender systems, leader-member exchange, psychometrics

3. Authors

The contemporary three-item modular adaptation of the Attitude Toward the Person (Relational Liking) scale was developed and validated by:

• Thorsten Hennig-Thurau, Ph.D. — Professor of Marketing and Media Research, Department of Marketing, University of Münster, Münster, Germany. Known for foundational work on relationship marketing, electronic word-of-mouth, and media economics.
• André Marchand, Ph.D. — Professor of Marketing & Digital Environment, Institute for Marketing, Leipzig University, Leipzig, Germany. Research specializes in digital marketing, recommender algorithms, consumer-technology interfaces, and entertainment media.
• Paul Marx, Ph.D. — Postdoctoral Researcher and Data Scientist, University of Münster, Münster, Germany; currently serving in senior analytics and decision-support roles within applied industry and academic consulting.

The operational items originated as an adaptation of a four-item interpersonal liking scale formulated in organizational behavior by:

• Sandy J. Wayne, Ph.D. — Professor of Management, College of Business Administration, University of Illinois Chicago, Chicago, Illinois, USA. Renowned for scholarship on Leader-Member Exchange (LMX), organizational citizenship behavior, and impression management.
• Gerald R. Ferris, Ph.D. — Francis Eppes Professor of Management and Psychology (Emeritus), College of Business, Florida State University, Tallahassee, Florida, USA. A pioneering scholar in organizational politics, social influence processes, and workplace interpersonal dynamics.

4. Purpose

Interpersonal attitudes play an indispensable role in shaping human decision making, collaborative task performance, negotiation outcomes, and service evaluations. The Attitude Toward the Person (Relational Liking) scale was developed to provide an efficient, psychometrically sound, and cross-contextually flexible assessment of positive dyadic affect. While comprehensive batteries assessing dyadic social relationships often span dozens of items, empirical investigations in experimental economics, consumer psychology, organizational behavior, and human-computer interaction (HCI) frequently require rapid, low-burden instruments that can capture social relationship quality without inducing participant fatigue or common-method contamination.

Rationale and Contextual Scope

Hennig-Thurau, Marchand, and Marx (2012) adapted this scale specifically to investigate how pre-existing social relational dynamics condition the effectiveness of automated group recommender systems. In group consumption environments (such as joint media selection, collaborative financial planning, dining decisions, or shared travel booking), individuals rarely deliberate in social vacuums; rather, group dynamics are fundamentally anchored by how much participants like, enjoy, and relate to one another. The ATPRS measures this personal affinity independently of task-oriented competencies, cognitive assessments of capability, or institutional obligations.

Research Applications

The ATPRS is engineered for diverse empirical domains:

• Group Recommender and Decision Support Systems: Examining whether algorithmically generated recommendations for groups (e.g., Collaborative Filtering, consensus models) enhance hedonic satisfaction when group members exhibit varying degrees of interpersonal liking versus stranger dynamics.
• Service Encounters and Professional Dyads: Quantifying client-adviser rapport, patient-physician relational warmth, and retail frontline interpersonal affinity.
• Collaborative Consumption and Sharing Economy: Measuring social trust and peer affinity in peer-to-peer (P2P) platforms (e.g., Airbnb hosts and guests, ride-sharing dyads).
• Workplace Dyadic Research: Assessing subordinate-supervisor compatibility, peer-to-peer team liking, and cross-functional project group integration, following the original lineage established by Wayne and Ferris (1990).

Clinical, Counseling, and Organizational Diagnostic Utility

Beyond academic laboratory and field experiments, the scale serves as a rapid diagnostic screening index in team mediation, organizational coaching, and relational counseling. It allows practitioners to assess baseline mutual interpersonal warmth before implementing structural interventions, distinguishing whether interpersonal friction stems from a sheer deficit in personal liking or from misaligned procedural expectations.

5. Psychological Construct

The construct captured by the ATPRS is Relational Liking (also termed interpersonal affect, social affinity, or personal attraction to the other). In structural relationship paradigms, social relationship quality is not a monolithic entity; rather, it decomposes into differentiated cognitive, behavioral, and affective pillars. Relational liking constitutes the primary affective pillar.

Conceptual Deconstruction

Relational liking denotes an individual’s positive, subjective evaluation of another specific person, characterized by sentiments of warmth, pleasantness during mutual interaction, and perceived psychological proximity. It answers the fundamental intuitive query: “Does this specific individual evoke positive subjective valence in me?” The ATPRS operationalizes this construct across three tightly interlinked conceptual facets:

1. Direct Global Liking: The baseline affective valence directed toward the individual (operationalized by: “I like my [partner] very much”). This facet captures instantaneous, unreflective positive sentiment and gut-level interpersonal affinity.
2. Sociable Friendship Potential: The appraisal of the partner as an individual with whom one could sustain an voluntary, communal bond (operationalized by: “I think my [partner] would make a good friend”). This dimension operationalizes psychological openness, sociometric attraction, and voluntary relational approach behaviors.
3. Perceived Interactional Harmony: The subjective absence of friction and the subjective presence of communicative ease during dyadic coordination (operationalized by: “I get along well with my [partner]”). This facet represents the behavioral manifestation of smooth interpersonal coordination.

Construct Distinctions and Boundary Conditions

To avoid theoretical conflation, psychometricians distinguish relational liking from adjacent relational constructs:

• Liking vs. Cognitive Trust: Trust entails an individual’s willingness to be vulnerable based on expectations of the partner’s competence, integrity, and predictability (Mayer, Davis, & Schoorman, 1995). One may possess immense cognitive trust in an aloof surgeon or an abrasive accountant while registering near-zero relational liking. Conversely, high relational liking can emerge toward an endearing friend whose competence in technical domains is acknowledged to be low.
• Liking vs. Relational Commitment: Commitment represents an enduring, behavioral intention to maintain a relationship over time, often sustained by structural switching costs, legal contracts, or normative duties. Liking, by contrast, is an intrinsic affective state that can flourish or fade independently of structural retention mechanisms.
• Liking vs. Perceived Homophily: While perceived similarity (homophily) is an antecedent to liking according to Byrne’s classical frameworks, homophily describes the cognitive realization that “we share similar traits, tastes, or values,” whereas relational liking is the affective consequence “I enjoy being connected with you.”

6. Theoretical Framework

The ATPRS rests upon several intersecting foundational theories in social psychology, sociology, and micro-organizational behavior.

Byrne’s Interpersonal Attraction Theory and the Law of Attraction

The principal conceptual architecture derives from Interpersonal Attraction Theory, pioneered by Donn Byrne (1971). Byrne posited that interpersonal attraction is a classical conditioning phenomenon wherein individuals develop positive attitudes toward social stimuli associated with rewarding interactions. Under the Law of Attraction, positive affective evaluations correlate directly with the proportion of reinforcing stimuli exchanged between two actors. The ATPRS directly captures this affective reinforcement state.

Heider’s Balance Theory

Fritz Heider’s (1958) Balance Theory provides the mechanistic foundation for why relational liking alters decision making in triads (Person–Other–Object, or P-O-X dynamics). When Person (P) likes Other (O), the system achieves psychological balance if P adopts an attitude toward an object (X)—such as an automated recommendation or a commercial choice—that harmonizes with O’s preferences. Hennig-Thurau et al. (2012) capitalized on Balance Theory to demonstrate that automated group recommenders succeed or fail partially because relational liking modulates whether group members willingly assimilate each other’s preferred consumption alternatives.

Social Exchange and Relational Cohesion Theories

According to Social Exchange Theory (Blau, 1964; Thibaut & Kelley, 1959) and Lawler’s Relational Cohesion Theory, repeated, mutually beneficial exchanges produce positive emotional arousal. Over time, actors misattribute some of this non-specific positive arousal directly to the dyadic partner, transforming instrumental transactions into expressive, socially valued relationships. Relational liking serves as the affective metric quantifying how far a relationship has evolved from a purely calculative transactional state to an emotionally bonded communal alliance.

Leader-Member Exchange (LMX) and Social Influence

In Wayne and Ferris’s (1990) foundational work, interpersonal affect was recognized as an essential determinant of subjective appraisal. Supervisors do not evaluate subordinates purely via objective performance rubrics; supervisor-subordinate liking systematically colors subjective evaluations, sponsorship, and task delegation. Adapting Wayne and Ferris’s items, Hennig-Thurau et al. demonstrated that identical social-affective dynamics govern consumer peer dyads and collaborative group choice environments.

7. Validity

The Attitude Toward the Person (Relational Liking) scale has undergone rigorous empirical validation across multiple experimental and field methodologies, demonstrating strong construct, convergent, discriminant, and predictive validity.

Construct and Convergent Validity

Construct validity is substantiated by consistently high factor saturation across independent empirical administrations. In Hennig-Thurau et al. (2012), where the scale was tested across hundreds of consumer dyads evaluating group entertainment products, average variance extracted (AVE) values exceeded the standard .50 psychometric benchmark, typically registering between .75 and .86. Standardized factor loadings across all three indicators consistently exceed .85, demonstrating that the items converge tightly on a singular latent construct representing relational liking.

Discriminant Validity

To establish that the ATPRS does not merely replicate general satisfaction or generalized positive mood, researchers systematically evaluate discriminant validity using the Fornell and Larcker (1981) criterion and the Heterotrait-Monotrait ratio of correlations (HTMT). When assessed alongside correlated relational constructs, the square root of the AVE for the ATPRS reliably exceeds its bivariate correlations with:

• Interpersonal Trust: While liking and trust correlate positively (typically r ≈ .50 to .65), HTMT ratios remain below the conservative .85 threshold, demonstrating empirical separation.
• Perceived Partner Expertise: Correlations between relational liking and partner domain competence typically remain moderate to low (r ≈ .15 to .35), confirming that personal liking is distinct from perceptions of cognitive expertise.
• General Positive Affect (PANAS): State positive affect exhibits low shared variance with the ATPRS, confirming that the scale captures target-specific interpersonal affinity rather than transient subjective mood states.

Predictive and Criterion Validity

Criterion-related validity has been demonstrated across several behavioral and evaluative benchmarks:

• Group Recommender Choice Quality: Hennig-Thurau et al. (2012) showed that pre-existing relational liking moderates the psychological efficacy of automated recommendation algorithms. Dyads with elevated relational liking exhibited significantly higher hedonic product choice quality and post-consumption satisfaction when supported by group recommenders than dyads with low relational liking.
• Compromise and Consensus Formation: In group negotiation tasks, high scores on the ATPRS predict faster consensus convergence, diminished post-negotiation regret, and greater willingness to concede personal utility for the partner’s benefit.
• Subjective Supervisory Appraisals: In Wayne and Ferris’s (1990) baseline studies, supervisory relational liking strongly predicted subordinate performance ratings, even after controlling for objective productivity indicators.

8. Reliability

The ATPRS exhibits exemplary internal consistency and scale reliability across varied sampling frames, languages, and dyadic permutations.

Internal Consistency Metrics

Empirical studies consistently report high internal reliability estimates:

• Cronbach’s Alpha (α): In Hennig-Thurau et al. (2012), the three-item instrument yielded an internal consistency coefficient of α = .92. Wayne and Ferris’s (1990) original four-item workplace formulation achieved an α of .89. Replications across social collaborative contexts routinely report alpha coefficients ranging between .88 and .95.
• Composite Reliability (ρ_c / McDonald’s ω): Because Cronbach’s alpha assumes tau-equivalence (equal factor loadings), psychometricians prefer composite reliability and McDonald’s omega. The ATPRS routinely generates composite reliability values > .90, substantially exceeding Nunnally and Bernstein’s (1994) recommended cutoff of .70 for research instruments and .90 for individual-level diagnostic instruments.
• Average Variance Extracted (AVE): AVE metrics for the three-item construct consistently hover between .74 and .85, indicating that the latent construct explains upwards of 74% of the variance in its observed manifest indicators, leaving minimal residual error variance.

Test-Retest Stability and Cross-Dyad Invariance

While the ATPRS is sensitive to deliberate relational manipulations (such as interpersonal conflict induction or cooperative bonding exercises), test-retest assessments across short, unmanipulated intervals (e.g., two-week test-retest in stable peer dyads) demonstrate temporal stability coefficients exceeding r = .82. Furthermore, measurement invariance tests (configural, metric, and scalar invariance) confirm that the instrument measures the construct identically across male-male, female-female, and mixed-gender dyads.

9. Factor Analysis

The structural dimensionality of the Attitude Toward the Person (Relational Liking) scale has been rigorously tested using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

During initial exploratory validation, principal axis factoring and principal component analysis with varimax/promax rotations consistently identify a single dominant eigenvalue accounting for over 75% to 85% of total item variance. The scree plot unequivocally exhibits a sharp drop after the first component, with no secondary factor achieving an eigenvalue above 0.45. Item-to-total correlations consistently range from .78 to .89, indicating that each item shares profound commonality with the underlying latent continuum.

Confirmatory Factor Analysis (CFA)

In structural equation modeling frameworks, the three-item unidimensional specification is just-identified (zero degrees of freedom) when estimated in isolation. However, when embedded in multidimensional measurement models alongside adjacent constructs (such as trust, commitment, and satisfaction), the ATPRS exhibits excellent parameter estimates and model fit metrics:

Overall structural measurement models incorporating the ATPRS routinely satisfy accepted structural equation modeling goodness-of-fit thresholds:

• Chi-Square / Degrees of Freedom Ratio (χ²/df): Typically between 1.2 and 2.4, denoting satisfactory model parsimony.
• Comparative Fit Index (CFI): Reliably ≥ .98 to 1.00.
• Tucker-Lewis Index (TLI): Reliably ≥ .97 to 1.00.
• Root Mean Square Error of Approximation (RMSEA): Consistently ≤ .045 (90% confidence interval spanning .000 to .065).
• Standardized Root Mean Square Residual (SRMR): Consistently ≤ .025.

10. Instrument / Measurement Tool

The ATPRS is a self-administered, modular psychometric scale optimized for laboratory, online, and field-based dyadic research designs.

Instrument Specifications

• Test Type: Self-report rating inventory; dyadic interpersonal perception measure.
• Format: Paper-and-pencil, computer-assisted personal interviewing (CAPI), or computerized web survey formats.
• Item Count: 3 manifest indicators.
• Modular Architecture: Contains an adaptable referent slot — [relationship partner] — allowing researchers to insert specific social roles (e.g., “neighbour,” “financial adviser,” “physician,” “co-consumer,” “partner,” “team member”).
• Response Format: 7-point Likert scale:
  • 1 = Strongly disagree
  • 2 = Disagree
  • 3 = Somewhat disagree
  • 4 = Neither agree nor disagree (Neutral midpoint)
  • 5 = Somewhat agree
  • 6 = Agree
  • 7 = Strongly agree
• Administration Time: Approximately 30 to 45 seconds per assessed partner.
• Scoring Rules: All three items are positively worded; there are no reverse-scored items. Individual item responses are numerically coded from 1 to 7. A composite Relational Liking score is calculated by taking the arithmetic mean of all three items:

  $$\text{Composite Relational Liking} = \frac{\text{Item } 1 + \text{Item } 2 + \text{Item } 3}{3}$$
  
  Higher scores (approaching 7.0) represent elevated interpersonal liking, affective rapport, and social compatibility; lower scores (approaching 1.0) reflect interpersonal coolness, social distance, or active disaffection. In structural equation modeling, the construct is represented as a reflective latent variable with the three items as direct continuous indicators.

11. Permissions & Fee and Test Year

The three-item modular adaptation of the Attitude Toward the Person (Relational Liking) scale was published by Hennig-Thurau, Marchand, and Marx in 2012 in the Journal of Marketing. Its predecessor four-item workplace liking instrument was published by Wayne and Ferris in 1990 in the Journal of Applied Psychology.

Licensing, Permissions, and Access

The scale items are published in the open scientific domain within standard academic literature. In accordance with academic fair use and standard psychometric dissemination conventions:

• Academic Research: Free of charge for scholarly, educational, non-commercial research and scientific investigations. Formal written copyright permission from the authors or the American Marketing Association (AMA) is generally not required for non-commercial academic research use, provided appropriate bibliographic citation is accorded to Hennig-Thurau et al. (2012) and Wayne and Ferris (1990).
• Commercial and Diagnostic Deployment: Organizations planning to incorporate the instrument into proprietary diagnostic software, commercial assessment engines, or revenue-generating consulting products should evaluate publisher copyright guidelines (American Marketing Association / American Psychological Association) or consult the primary authors.

12. References

The following peer-reviewed publications provide theoretical, methodological, and empirical foundations for the Attitude Toward the Person (Relational Liking) scale and its related constructs:

• Blau, P. M. (1964). Exchange and power in social life. John Wiley & Sons. https://doi.org/10.2307/2089716
• Byrne, D. (1971). The attraction paradigm. Academic Press. https://psycnet.apa.org/record/1972-03126-000
• 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
• Heider, F. (1958). The psychology of interpersonal relations. John Wiley & Sons. https://doi.org/10.1037/10628-000
• Hennig-Thurau, T., Marchand, A., & Marx, P. (2012). Can automated group recommender systems help consumers make better choices? Journal of Marketing, 76(5), 89–109. https://doi.org/10.1509/jm.10.0465
• 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/10.5465/amr.1995.9508080332
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
• Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups. John Wiley & Sons.
• Wayne, S. J., & Ferris, G. R. (1990). Influence tactics, affect, and exchange quality in supervisor-subordinate interactions: A laboratory experiment and field study. Journal of Applied Psychology, 75(5), 487–499. https://doi.org/10.1037/0021-9010.75.5.487

13. Items of the Scale (Questionnaire)