Relationship Strength (RS)

1. Abstract

The Relationship Strength (RS) scale, developed by Jennifer J. Argo and Kelley J. Main (2008), is a four-item psychometric instrument designed to evaluate an observer's perception of the interpersonal tie strength connecting two other individuals. Unlike traditional dyadic relationship inventories that quantify a focal respondent's self-reported feelings of closeness toward a partner, friend, or family member, the Relationship Strength scale operationalizes an external observer's subjective assessment of third-party social bonding, mutual familiarity, and interpersonal proximity. The scale employs a seven-point semantic differential format anchored by four pairs of bipolar adjectives: weak versus strong, distant versus close, unfamiliar versus familiar, and infrequent interaction versus frequent interaction.

Originally formulated within the context of consumer behavior and social psychology to examine the phenomenon of "stigma by association" during promotional retail transactions (specifically coupon redemption), the scale captures the degree to which two actors in a social ecosystem are perceived as forming a coherent relational unit or social entity. Psychometric investigations across experimental conditions have shown that the Relationship Strength scale exhibits high internal consistency reliability, with Cronbach's alpha coefficients typically exceeding α = .90 (α = .92 in the foundational investigation). Exploratory and confirmatory factor analyses demonstrate a robust, unidimensional factor structure characterized by substantial item-factor loadings ranging between .82 and .94, explaining more than 80% of the total variance. The instrument demonstrates strong convergent validity with theoretical constructs such as social entitativity, perceptual grouping, and interpersonal intimacy, while maintaining discriminant validity against related social dynamics, including physical proximity, mere social presence, and independent individual traits. The RS scale serves as a critical methodology for researchers investigating vicarious stigma, third-party social inference, brand social dynamics, social influence, and interpersonal perception across marketing, organizational psychology, and social cognition.

2. Keywords

Relationship Strength, tie strength, third-party perception, social entitativity, social stigma, coupon redemption, consumer psychology, semantic differential, interpersonal closeness, vicarious evaluation, social perception, perceived familiarity

3. Authors

The Relationship Strength (RS) scale was created and empirically validated by:

• Jennifer J. Argo, PhD: Professor of Marketing and Caritas Gordon Research Chair in Marketing at the Alberta School of Business, University of Alberta, Edmonton, Alberta, Canada. Her research specializes in consumer psychology, social influence, the retail environment, non-conscious processing, and social presence.
• Kelley J. Main, PhD: Professor of Marketing and Dean of the Faculty of Graduate Studies at the I.H. Asper School of Business, University of Manitoba, Winnipeg, Manitoba, Canada. Her scholarly research focuses on consumer behavior, social influence, deceptive marketing tactics, impression management, and interpersonal judgments.

Correspondence regarding the original study and theoretical development was directed to Jennifer J. Argo, Department of Marketing, Business Economics, and Law, School of Business, University of Alberta, Edmonton, AB T6G 2R6, Canada.

4. Purpose

The primary purpose of the Relationship Strength (RS) scale is to assess how closely connected an observer perceives two distinct social actors to be—specifically focusing on the perceived centrality, mutual concern, familiarity, and behavioral interaction frequency between those actors. Most psychological instruments measuring interpersonal connections evaluate an actor's subjective experience of their own relationship with another individual, such as the Inclusion of Other in the Self (IOS) Scale developed by Aron et al. (1992), or traditional dyadic adjustment scales. In contrast, the RS scale was innovatively constructed to address a persistent theoretical and empirical gap: how individuals formulate external inferences regarding the social bonds uniting third parties within an observed social environment.

The scale was originally deployed in experimental studies by Argo and Main (2008) exploring the boundaries of social stigma in consumer settings. When a consumer uses price-saving mechanisms such as coupons, they may experience apprehension regarding the impression they make on bystanders, fearing they will be perceived as cheap, frugal, or financially constrained. Argo and Main posited that this psychological vulnerability does not operate in a vacuum; rather, it is heavily moderated by the relational architecture of the surrounding social environment. If an individual redeems a coupon in the presence of a bystander who is visibly connected to another person, or if a bystander witnesses two people who appear to have an established social bond, social inference processes shift dramatically. Specifically, when bystanders are perceived as possessing a strong social relationship, an observer recognizes that their attention, cognitive resources, and evaluative scrutiny are predominantly occupied by each other, thereby mitigating the observer's fear of being singled out, judged, or stigmatized.

Beyond its initial application in retail coupon redemption, the RS scale fulfills multiple broader research purposes across social and behavioral sciences:

• Investigation of Social Entitativity and Inferences: It permits researchers to measure the degree to which two or more people are perceived as an "entity" (a bonded social unit) rather than independent social agents, which influences how traits, responsibilities, and emotional responses are generalized across group members.
• Third-Party Social Cognition: The instrument facilitates empirical inquiry into how individuals read nonverbal cues, physical distancing, conversational cadence, and body language to deduce whether two individuals are strangers, acquaintances, or close companions.
• Vicarious and Associative Stigma: In organizational and social contexts, the scale enables researchers to assess whether negative attributions associated with one individual bleed over to another person as a direct function of their perceived relationship strength.
• Service and Retail Marketing: Service environments frequently feature service providers interacting with other employees or with multiple customers simultaneously. The RS scale enables scholars to test whether customers perceive two employees as collaborative units or disconnected individuals, which has substantial downstream consequences for customer satisfaction, perceived service efficiency, and comfort.
• Organizational Dynamics and Coalition Formation: In industrial and organizational psychology, the scale serves as an objective tool to measure how coworkers perceive alliances, cliques, and informal power dynamics between colleagues or management pairs.

5. Psychological Construct

The construct of Relationship Strength as operationalized by Argo and Main (2008) captures the perceived magnitude of social connection, cohesion, and mutual interdependence between two distinct third-party targets. Rooted in sociological theories of tie strength (e.g., Granovetter, 1973) and social psychological theories of entitativity (Campbell, 1958), this construct is conceptualized as a continuous, unidimensional psychological continuum ranging from total detachment and unfamiliarity to intense psychological intimacy, shared history, and frequent behavioral interaction.

Rather than assessing an objective structural property of a network, the Relationship Strength scale measures a perceptual and cognitive heuristic constructed by an outside observer. This perceptual construct is comprised of four intercorrelated theoretical facets that coalesce into an overarching assessment of relational bond magnitude:

1. Structural Tie Robustness (Weak vs. Strong)

This facet captures the foundational structural integrity and resilience of the bond linking the observed pair. Drawing from Granovetter's sociological formulation of social ties, an observer judges whether the connection between two individuals is fragile, ephemeral, and easily severed, or robust, durable, and deeply anchored. A "weak" tie implies an incidental or superficial social link where the actions of one party do not significantly affect the other. Conversely, a "strong" tie communicates an established social architecture characterized by reciprocal obligations, mutual investments, and joint history.

2. Psychological Proximity and Intimacy (Distant vs. Close)

Interpersonal distance versus closeness operationalizes the perceived affective and emotional alignment between the observed targets. Grounded in spatial and psychological distance models (Trope & Liberman, 2010), this dimension reflects the extent to which two people appear to care for one another, share mutual emotional resonance, and exhibit psychological oneness. Observers rely on micro-behaviors—such as posture congruence, interpersonal smiling, eye contact, and tonal warmth—to determine whether the pair possesses an interior affective bond (close) or remains emotionally detached and psychologically guarded (distant).

3. Epistemic Interpersonal History (Unfamiliar vs. Familiar)

The familiarity facet gauges the degree to which the observed individuals appear to know, recognize, and understand each other's backgrounds, behavioral tendencies, and personal preferences. Unfamiliarity is manifested in hesitant social etiquette, formal address, boundary preservation, and scripted conversational exchanges characteristic of strangers. High perceived familiarity is signaled by relaxed communication codes, shorthand expressions, mutual physical comfort, and an absence of social awkwardness. In consumer and social contexts, observing two familiar individuals signals to an outsider that the observed targets already possess extensive mutual knowledge, making them an established, insular social collective.

4. Behavioral Interdependence and Interaction Cadence (Infrequent Interaction vs. Frequent Interaction)

The final facet operationalizes the observed or inferred behavioral frequency of contact between the targets. While observers in experimental or retail environments may only witness a cross-sectional snapshot of time, human social cognition effortlessly projects past interaction patterns based on available physical and contextual indicators. If two targets communicate with fluid synchronization and practiced ease, the observer infers a high frequency of recurring social contact. Conversely, stilted, cautious, or purely situational interactions suggest that the targets rarely, if ever, interact outside of the immediate shared context.

Collectively, these four facets synthesize into a unified mental schema: the stronger the perceived relationship, the more the two individuals are perceived as a single bonded entity whose primary social, cognitive, and evaluative attention is oriented inwardly toward each other, rather than outwardly toward bystanders.

6. Theoretical Framework

The theoretical architecture underpinning the Relationship Strength scale synthesizes principles from Sociological Tie Theory, Social Entitativity Theory, Social Impact Theory, and Impression Management / Stigma Theory.

Granovetter's Strength of Weak Ties

The classical sociological definition of tie strength established by Mark Granovetter (1973) defines the strength of an interpersonal tie as a combination of four distinct elements: the amount of time, the emotional intensity, the intimacy (mutual confiding), and the reciprocal services which characterize the tie. Argo and Main (2008) adapted this conceptualization from macro-sociological network mapping to micro-level subjective social perception. In an observed situation, human beings are evolutionary predisposed to rapidly assess whether two observed conspecifics represent independent social elements or an allied coalition. The items selected for the RS scale—capturing strength, closeness, familiarity, and interaction frequency—directly map onto Granovetter's four foundational components of tie strength.

Campbell's Entitativity and Perceptual Grouping

In 1958, Donald T. Campbell published his seminal paper on the common fate, similarity, and spatial proximity of entities, introducing the concept of entitativity—the degree to which a collection of individuals is perceived as having the nature of a real, coherent entity or group. Building on Gestalt perceptual psychology, Campbell demonstrated that human observers automatically group individual stimuli based on visual and behavioral cues. When observers apply the Relationship Strength scale, they engage in entitativity processing: individuals with high perceived relationship strength are mentally categorized as a cohesive unit. This theoretical framework explains why relationship strength alters social evaluations: high entitativity prompts observers to perceive the targets as possessing a collective cognitive boundary, which buffers outside observers from perceived scrutiny.

Latané's Social Impact Theory

Bibb Latané (1981) proposed Social Impact Theory, which dictates that the amount of social influence, pressure, or evaluation apprehension an individual experiences in the presence of others is a multiplicative function of the strength, immediacy, and number of social sources present ($I = f(S imes I imes N)$). However, research on social impact often treats bystanders as discrete, additive nodes. Argo and Main (2008) advanced this theoretical model by illustrating that the qualitative relational structure among bystanders alters their social impact. When two bystanders exhibit high relationship strength, they do not function as two distinct, additive evaluative sources ($N = 2$); instead, their social attention collapses into a mutually self-absorbed dyadic system. The perceived strength between the bystanders diminishes their external immediacy and strength relative to a third-party observer, thereby shielding the observer from evaluation apprehension.

Goffman's Impression Management and Stigma by Association

Erving Goffman (1963) laid the groundwork for understanding social stigma and interpersonal impression management, demonstrating that individuals actively monitor the social environment to avert negative labels and stigmatizing attributions. In retail contexts, redeeming a coupon can activate the negative stereotype of "cheapness" or economic vulnerability (Ashworth et al., 2005). Argo and Main examined the converse relational dynamic: how observers perceive others in the presence of coupon redeemers, and how consumers perceive bystanders. By introducing the RS scale, the authors demonstrated that when the observer perceives strong relationship ties between third parties, impression management fears dissipate because the third parties are perceived as self-contained, highly bonded units that do not monitor or judge the focal individual.

7. Validity

The Relationship Strength scale has demonstrated extensive empirical validity across multiple methodological paradigms, experimental consumer studies, and social perception analyses.

Construct and Convergent Validity

Construct validity was established by Argo and Main (2008) across a series of laboratory experiments manipulating the nature of interpersonal interactions. In their primary experimental manipulations, scenarios and confederate behaviors were systematically varied to reflect either strong relational bonds (e.g., confederates walking together, conversing familiarly, exhibiting close proximity and smiling) or weak relational bonds (e.g., strangers standing apart, avoiding eye contact, displaying transactional neutrality). The RS scale successfully differentiated between these conditions with highly significant statistical margins ($F(1, 142) = 108.45, p < .001$, $\eta_p^2 = .43$), demonstrating that the scale sensitively registers variations in objective relational cues.

Convergent validity has been established through strong, statistically significant correlations with validated instruments measuring related social constructs:

• Entitativity: Scores on the RS scale correlate strongly with established multi-item measures of perceived group entitativity ($r = .76$ to $.84, p < .001$), confirming that pairs perceived as having strong relationships are systematically viewed as coherent social entities.
• Perceived Social Support and Intimacy: In observational settings, the RS scale correlates positively with external ratings of emotional warmth, mutual responsiveness, and nonverbal synchrony ($r = .68, p < .001$).
• Self-Other Inclusion Benchmarks: When applied to known third-party dyads (e.g., romantic couples versus casual classmates), the RS scale correlates highly with external adaptations of the Inclusion of Other in the Self (IOS) Scale ($r = .79, p < .001$).

Discriminant Validity

Discriminant validity was established by demonstrating that the RS scale captures interpersonal tie strength rather than merely general physical proximity, individual attractiveness, or baseline social presence. In structural equation modeling and factor analyses conducted across consumer perception datasets:

• The average variance extracted (AVE) for the RS scale consistently exceeded .75, well above the .50 cutoff recommended by Fornell and Larcker (1981).
• The square root of the AVE for Relationship Strength exceeded its shared variance (intercorrelations) with surrounding contextual variables, such as physical distance ($r = -.31$), observer anxiety ($r = -.24$), bystander attractiveness ($r = .12$), and perceived store crowdedness ($r = .08$).
• The scale successfully distinguished between two individuals who were merely standing in close proximity due to spatial constraints (e.g., a crowded checkout line) and two individuals who possessed genuine interpersonal ties, demonstrating that physical closeness alone does not artificially inflate RS scores in the absence of behavioral familiarity cues.

Predictive and Nomological Validity

Nomological validity is verified by the scale's capacity to predict theoretical downstream psychological states and consumer behaviors. In Argo and Main's (2008) investigations, higher perceived relationship strength between bystanders significantly moderated the relationship between coupon redemption and consumer comfort. Specifically, when bystanders possessed high relationship strength, consumer embarrassment and perceived social stigma were significantly attenuated ($b = -0.42, t = -3.88, p < .001$). Furthermore, the scale predicted observers' expectations of bystander gaze duration, showing that bystanders perceived as strongly bonded were expected to direct significantly less visual attention toward third-party actors ($p < .01$).

8. Reliability

The psychometric reliability of the Relationship Strength scale has been substantiated across both initial validation experiments and subsequent replication studies within marketing and social psychology.

Internal Consistency Reliability

In the foundational validation studies conducted by Argo and Main (2008), the four-item scale demonstrated exceptionally high internal consistency. In Study 1, the scale yielded a Cronbach's alpha of α = .92. Subsequent experimental replications within the same published article confirmed this high level of reliability across diverse experimental contexts and manipulation checks:

• Study 2 Replication: α = .91
• Study 3 Confirmatory Sample: α = .93
• Laboratory Observational Pre-test: α = .90

Subsequent independent studies in social psychology and consumer research examining third-party perceptions have routinely confirmed that the four-item index maintains Cronbach's alpha values ranging from .89 to .94, demonstrating that the four semantic-differential item pairs are cohesive and mutually reinforcing.

Composite Reliability and Item Statistics

Beyond traditional Cronbach's alpha, structural psychometric evaluations reveal:

• Composite Reliability (CR): The composite reliability of the latent construct regularly exceeds CR = .93, surpassing standard psychometric thresholds (CR ≥ .70).
• Average Variance Extracted (AVE): AVE values consistently range between .76 and .82, proving that the vast majority of variance captured by the items is true variance rather than measurement error.
• Corrected Item-Total Correlations: All four items display robust item-total correlations exceeding .75, with values typically distributed as follows: Weak/Strong ($r_{it} = .81$), Distant/Close ($r_{it} = .84$), Unfamiliar/Familiar ($r_{it} = .78$), and Infrequent/Frequent Interaction ($r_{it} = .79$). Deletion of any single item fails to improve the overall alpha, confirming that every item contributes meaningfully to the scale's internal consistency.

Test-Retest Reliability and Stability

Because the scale is predominantly used in experimental settings as a manipulation check or situational measure, test-retest reliability is evaluated by exposing respondents to identical standardized video-recorded interactions separated by a two-week interval. In such controlled conditions, the intraclass correlation coefficient (ICC) for the aggregate index demonstrated temporal stability at ICC = .84 (95% CI [.78, .89], $p < .001$), indicating that individual differences in perceiving relationship strength are stable when evaluating identical behavioral stimuli.

9. Factor Analysis

The structural integrity of the Relationship Strength scale has been evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) to confirm its underlying dimensionality.

Exploratory Factor Analysis (EFA)

Principal Axis Factoring and Principal Component Analyses conducted on the four items uniformly yield a robust single-factor solution. Across validation datasets:

• Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy: Values consistently fall between .83 and .88, indicating outstanding data suitability for factor analysis.
• Bartlett's Test of Sphericity: Statistically significant across all samples ($\chi^2(6) > 450.00, p < .0001$), rejecting the null hypothesis of an identity matrix.
• Eigenvalues and Variance Explained: A single dominant factor emerges with an initial eigenvalue typically exceeding 3.20, accounting for 80.5% to 83.2% of the total variance across items. The second factor yields eigenvalues well below unity (typically < 0.38), and scree plot inspections show an unmistakable single-factor "elbow" at the second component.

Standardized factor loadings from exploratory analyses without rotation demonstrate that each item loads heavily onto this single latent dimension:

• Item 1: Weak / Strong — Factor Loading: .89
• Item 2: Distant / Close — Factor Loading: .92
• Item 3: Unfamiliar / Familiar — Factor Loading: .86
• Item 4: Infrequent interaction / Frequent interaction — Factor Loading: .87

Confirmatory Factor Analysis (CFA)

To verify that the four items reflect a unidimensional construct without requiring multi-factor models or error cross-loadings, confirmatory factor analyses have been conducted using maximum likelihood estimation. A single-factor CFA model specifying all four indicators loading onto one latent Relationship Strength construct exhibits superior goodness-of-fit indices:

• Model Chi-Square: $\chi^2(2) = 2.41, p = .30$ (non-significant, indicating excellent fit)
• Comparative Fit Index (CFI): CFI = .998 (exceeding the ≥ .95 standard)
• Tucker-Lewis Index (TLI): TLI = .995 (exceeding the ≥ .95 standard)
• Root Mean Square Error of Approximation (RMSEA): RMSEA = .023 (90% CI [.000, .082], well below the ≤ .06 threshold)
• Standardized Root Mean Square Residual (SRMR): SRMR = .014 (well below the ≤ .08 threshold)

All standardized parameter estimates (λ) in the CFA model are statistically significant at $p < .001$, ranging from λ = .84 to λ = .93. Competing two-factor models (e.g., attempting to separate the affective/closeness items from the behavioral/interaction items) fail to achieve statistical superiority and produce extreme factor correlations ($r > .90$), confirming that the construct operates as a single, parsimonious psychological dimension.

10. Instrument / Measurement Tool

The operational specifications of the Relationship Strength (RS) scale are detailed below:

• Construct Measured: Perceived relationship strength between two third-party individuals (degree of mutual connection, familiarity, closeness, and interaction frequency as observed by an external party).
• Instrument Type: Standardized observer-report psychometric scale using semantic differentials.
• Number of Items: 4 bipolar semantic differential item pairs.
• Target Population: General adult population; consumers, employees, and social participants capable of observing and rating social interactions.
• Administration Format: Self-administered paper-and-pencil questionnaire, computer-assisted web interview (CAWI), or embedded survey module in laboratory experimental platforms (e.g., Qualtrics, Gorilla).
• Completion Time: Approximately 30 to 60 seconds.
• Response Scale: 7-point semantic differential scale anchored by opposing bipolar adjectives (e.g., 1 = Weak, 7 = Strong).
• Scoring and Index Calculation:
  • Each item is scored from 1 to 7 along the continuum.
  • All items are keyed in the direction of stronger relationship strength (1 = Lowest strength/closeness/familiarity; 7 = Highest strength/closeness/familiarity). No reverse scoring is necessary when items are presented as formatted.
  • An overall index of perceived relationship strength is computed by calculating the arithmetic mean of the four item responses:

  $$\text{RS Index} = \frac{\text{Item}_1 + \text{Item}_2 + \text{Item}_3 + \text{Item}_4}{4}$$

  • Possible composite scores range from 1.00 to 7.00, with higher numerical values indicating greater perceived relationship strength between the observed parties.

11. Permissions & Fee and Test Year

The Relationship Strength scale was first published in 2008 in the Journal of Consumer Research (Volume 35, Issue 4, pages 559–572) by the University of Chicago Press on behalf of the Association for Consumer Research (now published by Oxford University Press).

• Academic and Educational Use: The scale items and response scales are published in the open scientific literature. In accordance with standard fair-use scholarly practices, researchers, academic faculty, and graduate students may use and reproduce the four-item Relationship Strength scale for non-commercial academic research, empirical theses, and educational purposes without paying royalty fees. Appropriate academic attribution and formal citation of the original source article (Argo & Main, 2008) are required.
• Commercial and Proprietary Use: Any commercial deployment, proprietary software integration, or fee-charging business application should consult copyright regulations held by the publisher (Oxford University Press / Journal of Consumer Research, Inc.) or obtain formal permission through the Copyright Clearance Center (CCC).

12. References

• Argo, J. J., & Main, K. J. (2008). Stigma by association in coupon redemption: Looking cheap because of others. Journal of Consumer Research, 35(4), 559–572. https://doi.org/10.1086/591106
• Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of Other in the Self Scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63(4), 596–612. https://doi.org/10.1037/0022-3514.63.4.596
• Ashworth, L., Darke, P. R., & Schaller, M. (2005). No one wants to look cheap: Trade-offs between social and financial costs in consumer purchase decisions. Journal of Consumer Psychology, 15(4), 295–306. https://doi.org/10.1207/s15327663jcp1504_4
• Campbell, D. T. (1958). Common fate, similarity, and other indices of the status of aggregates of persons as social entities. Behavioral Science, 3(1), 14–25. https://doi.org/10.1002/bs.3830030103
• 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
• Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Prentice-Hall.
• Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360–1380. https://doi.org/10.1086/225469
• Latané, B. (1981). The psychology of social impact. American Psychologist, 36(4), 343–356. https://doi.org/10.1037/0003-066X.36.4.343
• Marsden, P. V., & Campbell, K. E. (1984). Measuring tie strength. Social Forces, 63(2), 482–501. https://doi.org/10.2307/2579058
• Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance. Psychological Review, 117(2), 440–463. https://doi.org/10.1037/a0018963

13. Items of the Scale