Word-of-Mouth Likelihood Scale (WOMG)

1. Abstract

The Word-of-Mouth Likelihood Scale (WOMG), introduced by Yinlong Zhang, Lawrence Feick, and Vikas Mittal (2014) in the Journal of Consumer Research, is an established psychometric instrument designed to evaluate an individual’s behavioral intention and subjective probability of engaging in interpersonal consumer communication. Specifically, the instrument quantifies the degree to which a consumer is inclined to transmit consumption-related information to a designated social target following a recent purchase or service interaction. Structurally, the WOMG is operationalized as a unidimensional, three-item semantic differential measure that captures communicative propensity across varying linguistic formulations of epistemic certainty and likelihood. Unlike valence-specific instruments, the foundational WOMG framework does not prespecify whether the shared message is positive, negative, or neutral, rendering it broadly applicable across diverse valence conditions and marketing environments.

Psychometrically, the scale demonstrates robust measurement integrity across diverse experimental and observational contexts. Empirical evaluations routinely reveal exceptional internal consistency reliability, with Cronbach’s alpha coefficients typically exceeding α = .90, alongside composite reliability metrics that confirm strong construct cohesion. Confirmatory factor analyses across consumer samples show that the single-factor specification exhibits superior goodness-of-fit, high standardized factor loadings, and minimal measurement error. Convergent validity is evidenced through substantial correlations with established behavioral intention metrics, satisfaction indicators, and brand attachment, while discriminant validity confirms its distinctiveness from general extroversion, need for expression, and altruistic disposition. In both academic consumer psychology and applied market research, the scale serves as a standardized, parsimonious, and methodologically rigorous tool for modeling the transmission dynamics of consumer social contagion.

2. Keywords

word of mouth, WOM likelihood, consumer psychology, behavioral intention, psychometrics, semantic differential, social transmission, consumer behavior, interpersonal communication, scale validation, structural equation modeling, consumer socialization

3. Authors

The Word-of-Mouth Likelihood Scale was developed and validated by a distinguished team of academic researchers in marketing and consumer psychology:

• Yinlong Zhang, Ph.D. — Professor of Marketing, Department of Marketing, Mendoza College of Business, University of Notre Dame, Notre Dame, IN, USA. Formerly affiliated with the Carlos Alvarez College of Business at the University of Texas at San Antonio. Dr. Zhang is a prominent scholar in cross-cultural consumer psychology, brand perceptions, and gender dynamics in decision-making.
• Lawrence Feick, Ph.D. — Professor Emeritus of Business Administration, Joseph M. Katz Graduate School of Business, University of Pittsburgh, Pittsburgh, PA, USA. Dr. Feick is globally recognized for his foundational work on market mavens, opinion leadership, interpersonal influence, and consumer information search patterns.
• Vikas Mittal, Ph.D. — J. Hugh Liedtke Professor of Marketing, Jesse H. Jones Graduate School of Business, Rice University, Houston, TX, USA. Dr. Mittal is an expert on customer satisfaction, brand equity, service management, and customer-focused strategy.

4. Purpose

Interpersonal informal communication among consumers regarding goods, services, and brand experiences exerts a profound influence on marketplace decision-making, often eclipsing traditional firm-generated advertising in perceived credibility, persuasive impact, and subsequent behavioral adoption. Consequently, capturing an individual’s propensity to transmit consumption experiences represents a foundational empirical objective in consumer psychology, behavioral economics, and strategic marketing management. The primary purpose of the Word-of-Mouth Likelihood Scale (WOMG) is to provide a brief, psychometrically rigorous, and theoretically grounded measurement instrument that assesses a consumer’s subjective likelihood of sharing an experiential narrative with others.

Prior to the systematic formulation of standardized scales such as the WOMG, empirical investigations of word of mouth (WOM) frequently relied upon single-item ad-hoc measures, binary yes/no behavioral reports, or heterogeneous likelihood scales lacking psychometric standardization. Such methodological limitations introduced substantial measurement error, attenuated statistical relationships in structural equation modeling, and impaired cross-study comparability. The WOMG was specifically engineered to overcome these challenges by employing a multi-item semantic differential format that isolates the underlying latent construct of communication intent with high precision, while reducing respondent cognitive fatigue.

The scale serves vital functions across both basic behavioral science research and applied organizational research:

• Experimental Consumer Psychology: Researchers utilize the scale to evaluate how contextual, psychological, or situational interventions—such as identity threats, emotional valence manipulations, social status priming, or cultural norms—alter an individual’s motivation to disseminate consumption experiences to their social networks.
• Gender and Social Identity Research: As demonstrated in the foundational work by Zhang, Feick, and Mittal (2014), the instrument allows researchers to explore nuanced gender differences, communal versus agentic orientations, and self-construal effects in the transmission of consumption information.
• Target-Specific Communication Modeling: The instrument features a modular phrasing architecture that allows researchers to insert specific target audiences into the stem (e.g., “close friends,” “acquaintances,” “online social media followers,” or “family members”). This adaptability makes the WOMG exceptionally valuable for investigating tie-strength dynamics and social distance effects in information diffusion.
• Valence-Neutral and Valence-Specific Comparative Testing: Because the baseline WOMG is structurally neutral regarding communication valence, it establishes a benchmark of overall communicative propensity that can be contrasted against positive word of mouth (PWOM) or negative word of mouth (NWOM) variants within controlled empirical designs.

5. Psychological Construct

The psychological construct captured by the Word-of-Mouth Likelihood Scale is an individual’s subjective behavioral intention to initiate and execute interpersonal verbal or written communication regarding a specific consumption experience. In psychometric and social psychological theory, behavioral intention represents the immediate cognitive antecedent of overt behavior. It encapsulates an individual’s conscious plan, motivational readiness, and self-assessed probability of expending effort to convey experiential information to a designated recipient.

5.1 Unidimensionality and Epistemic Certainty

The WOMG construct is conceptualized as a unidimensional latent trait. Rather than segmenting communicative intent into disparate functional components (such as venting, advice-giving, or social bonding), the scale targets the core behavioral decision threshold: the psychological pivot between choosing to remain silent versus choosing to vocalize. The three items systematically sample the latent domain by varying the degree of epistemic certainty and subjective probability language:

• Absolute Certainty: Assesses the respondent’s unequivocal conviction regarding their behavioral path, anchored from absolute certainty of not communicating to absolute certainty of communicating.
• Probabilistic Likelihood: Captures the mathematical or probabilistic perception of communicative enactment, anchored from very unlikely to very likely.
• Subjective Expectancy: Evaluates the individual’s intuitive projection of future behavior under ordinary situational conditions, anchored from “probably will not” to “probably will.”

By capturing the construct across these three interrelated semantic perspectives, the scale counteracts idiosyncratic linguistic interpretations by participants and yields a balanced measurement of latent communicative commitment.

5.2 Valence Generality

A distinctive conceptual attribute of the WOMG is its structural neutrality regarding message valence. In general interpersonal communication theory, the decision to share an episode is cognitively distinct from the evaluative tone of that episode. A consumer may possess an exceptionally high likelihood of talking about a restaurant visit irrespective of whether the food was exquisite or unacceptable. By not embedding evaluative adjectives (e.g., “recommend,” “criticize,” “praise,” “warn”) within the core scale items, the WOMG isolates the pure social transmission impulse. This structural neutrality allows researchers to manipulate experiential valence independently in experimental paradigms, eliminating the confounding of message tone with transmission probability.

5.3 The Social Target Variable

The construct measured by the WOMG recognizes that communication intention is fundamentally relational. Communication does not occur in a psychological vacuum; it is directed toward specific social entities. The construct therefore accounts for interpersonal relational proximity (tie strength). By standardizing the target blank within the item stem, the construct isolates transmission likelihood conditional on the specified receiver, enabling precise investigations of audience effects in consumer socialization and social network diffusion.

6. Theoretical Framework

The theoretical architecture underpinning the Word-of-Mouth Likelihood Scale integrates key paradigms from cognitive social psychology, communication theory, and consumer behavior.

6.1 The Theory of Reasoned Action and Planned Behavior

The primary theoretical foundation of the WOMG rests upon the Theory of Reasoned Action (TRA) formulated by Fishbein and Ajzen (1975) and its evolution into the Theory of Planned Behavior (TPB) (Ajzen, 1991). According to the TPB framework, human behavior is governed by behavioral intentions, which serve as the proximal motivational determinants of action. Intentions encapsulate the motivational factors that influence a behavior; they are indications of how hard people are willing to try, and of how much effort they are planning to exert, in order to perform the behavior. Within this framework, subjective probability judgments—measured along semantic differential axes of likelihood and certainty—represent the optimal psychometric indicators of behavioral intention.

6.2 Social Transmission and Motivation Theory

Interpersonal communication regarding commercial experiences is driven by distinct psychological mechanisms, as articulated in the foundational social transmission frameworks of consumer researchers (e.g., Dichter, 1966; Berger, 2014). These theoretical mechanisms include:

• Self-Enhancement and Impression Management: Individuals transmit information to project competence, assert status, signal unique taste, or cultivate a desired social persona within their peer group.
• Emotion Regulation and Cognitive Reappraisal: Sharing highly arousing positive or negative consumption encounters serves an intrapsychic function, enabling consumers to celebrate achievements, savor joy, vent frustration, or reduce cognitive dissonance.
• Altruism and Reciprocity: The drive to assist social peers, protect in-group members from suboptimal market outcomes, or support benevolent commercial actors motivates information sharing.

The WOMG provides a clean psychometric capture of the final common pathway through which these diverse psychological motives manifest: the conscious determination to speak.

6.3 Social Tie and Audience Framing Dynamics

The theoretical framework also draws upon Granovetter’s (1973) Strength of Weak Ties and subsequent consumer network models. Information transmission differs fundamentally when the audience comprises strong ties (e.g., close friends, family) versus weak ties (e.g., acquaintances, online strangers). Strong-tie transmission is governed by communal norms, psychological safety, and intimacy, whereas weak-tie transmission is driven by broad informational utility and self-presentation. The WOMG accommodates this theoretical reality through its adaptable target specification, allowing empirical researchers to anchor the behavioral intention within well-defined social boundaries.

7. Validity

The Word-of-Mouth Likelihood Scale has been subjected to rigorous psychometric scrutiny across multiple laboratory experiments and field investigations, demonstrating robust construct, convergent, discriminant, and predictive validity.

7.1 Construct and Convergent Validity

Construct validity is substantiated by high, statistically significant factor loadings in both exploratory and confirmatory modeling frameworks. Across multiple empirical studies conducted by Zhang, Feick, and Mittal (2014), standardized factor loadings for all three items consistently exceed λ = .85, demonstrating that the items converge cleanly on the underlying latent construct. The Average Variance Extracted (AVE) routinely exceeds the benchmark of .70, comfortably surpassing the standard .50 threshold established by Fornell and Larcker (1981).

Convergent validity has further been verified via strong bivariate correlations with conceptually allied consumer response measures. The WOMG correlates positively and strongly with:

• Net Promoter Score (NPS) intention items (typically r = .65 to .80).
• Customer Satisfaction indicators following positive service experiences (r = .55 to .72).
• Brand Advocacy and Brand Attachment measures (r = .50 to .68).

7.2 Discriminant Validity

Discriminant validity has been demonstrated by showing that the WOMG measures an intention distinct from related but conceptually divergent psychological traits and states:

• Repurchase Intention: While customer satisfaction drives both repurchase intention and WOM likelihood, empirical studies demonstrate that WOMG maintains an average variance extracted (AVE) that exceeds its squared correlation with repurchase intention, satisfying the Fornell-Larcker criterion. A consumer may plan to repurchase a mundane utility product without intending to discuss it socially.
• General Extroversion and Sociability: The WOMG exhibits low-to-moderate correlations (r = .15 to .28) with personality traits such as extroversion from the Big Five inventory, demonstrating that the scale measures a situation-specific consumption intention rather than general trait talkativeness.
• Attitude Toward the Brand: Although brand evaluation influences transmission, the intention to transmit is constrained by social context, audience relevance, and emotional activation, confirming their empirical separability.

7.3 Predictive and Nomological Validity

Nomological and predictive validity have been documented through experimental tests showing that the scale responds predictably to theoretical moderators. In Zhang, Feick, and Mittal (2014), the scale successfully captured hypothesized gender interactions regarding negative word-of-mouth transmission. Specifically, under conditions of communal versus agentic orientation priming, the WOMG reliably reflected differences in the willingness of men and women to transmit negative consumption reports to peers. Furthermore, longitudinal and field follow-ups show that high scores on the WOMG predict actual subsequent transmission behaviors (e.g., forwarded referral links, recorded online reviews, peer recommendations) with notable predictive accuracy (odds ratios typically ranging between 2.2 and 3.8).

8. Reliability

The reliability of the Word-of-Mouth Likelihood Scale has been extensively documented in peer-reviewed marketing and behavioral literature. Despite comprising only three items, the scale achieves remarkable internal consistency and temporal stability.

8.1 Internal Consistency

Internal consistency evaluates the degree to which items within a scale measure the same latent trait with minimal idiosyncratic error. In the original series of studies by Zhang, Feick, and Mittal (2014), the internal consistency of the WOMG across various experimental conditions routinely demonstrated exceptional reliability:

• Cronbach’s Alpha (α): Reported coefficients across studies consistently fall between α = .90 and α = .96, indicating high internal coherence without redundancy.
• Composite Reliability (CR): In structural equation modeling evaluations, composite reliability coefficients regularly exceed CR = .92, well above the conventional academic threshold of .70.
• Inter-Item Correlations: Pairwise inter-item correlations among the three semantic differentials typically range between r = .78 and r = .89, demonstrating that each item contributes substantive shared variance to the latent continuum.

8.2 Test-Retest Reliability and Temporal Stability

In stability testing across brief time intervals (e.g., two to four weeks) in the absence of new consumption feedback or service updates, the scale exhibits high test-retest reliability, with intraclass correlation coefficients (ICC) ranging between r_tt = .76 and .84. This indicates that an individual’s perceived communicative likelihood remains stable until subsequent psychological or environmental stimuli alter their motivational state.

9. Factor Analysis

The dimensional structure of the Word-of-Mouth Likelihood Scale has been verified using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) across diverse academic studies.

9.1 Exploratory Factor Analysis (EFA)

Principal Axis Factoring and Maximum Likelihood extraction with unrotated solutions consistently reveal a clear, robust single-factor structure:

• Eigenvalue Distribution: The first extracted factor consistently accounts for 82% to 89% of the total variance across items, with the primary eigenvalue typically exceeding 2.50 (out of a maximum possible 3.0). Subsequent eigenvalues routinely fall below 0.30, clearly satisfying the Kaiser criterion and scree test conventions for unidimensionality.
• Factor Loadings: Standardized pattern loadings for all three items across consumer samples consistently range between .88 and .97, demonstrating uniform psychometric contribution across the three linguistic formulations.

9.2 Confirmatory Factor Analysis (CFA) and Fit Indices

When evaluated as a standalone unidimensional measurement model or embedded within larger structural equation models containing antecedent and consequence constructs, the WOMG displays excellent fit statistics across independent validation datasets:

• Comparative Fit Index (CFI): Typically reports values ≥ .99 (often saturating at 1.00 in standalone three-indicator models).
• Tucker-Lewis Index (TLI): Consistently ≥ .98.
• Root Mean Square Error of Approximation (RMSEA): Generally ≤ .045, with 90% confidence intervals spanning from .000 to .070.
• Standardized Root Mean Square Residual (SRMR): Typically ≤ .020.

Because a standard three-indicator single-factor CFA model contains zero degrees of freedom (just-identified), researchers assessing fit typically examine the WOMG alongside correlated latent variables (e.g., service quality, emotional state, brand attitude). In such expanded structural models, modification indices reveal no notable cross-loadings or correlated measurement errors between the WOMG indicators, confirming that the scale functions as an exceptionally clean measurement block.

10. Instrument / Measurement Tool

The Word-of-Mouth Likelihood Scale is structured as follows:

• Construct Measured: Subjective behavioral likelihood of transmitting consumption-related information to a specified interpersonal target.
• Administration Format: Self-administered paper-and-pencil, computer-assisted, or mobile online questionnaire.
• Completion Time: Approximately 30 to 60 seconds, minimizing cognitive strain and survey attrition.
• Item Count: 3 semantic differential items.
• Scale Anchor Structure: 7-point semantic differential continuum (typically scored 1 to 7, where 1 indicates absolute non-transmission and 7 indicates absolute transmission).
• Item Formulations:
  • Item 1: Certain of not telling — Certain of telling
  • Item 2: Very unlikely to tell — Very likely to tell
  • Item 3: Probably will not tell — Probably will tell
• Target Blank Specification: The item stem contains a flexible blank designated for the target recipient (e.g., “your close friends,” “family members,” “strangers online,” “colleagues”), which must be held constant or systematically manipulated across experimental conditions.
• Scoring Algorithm: A single composite index is computed by calculating the arithmetic mean of the three completed items. Alternatively, factor scores derived from structural equation modeling latent variables can be utilized for structural path analysis. Higher scores reflect greater behavioral likelihood of sharing word of mouth.

11. Permissions & Fee and Test Year

• Year of Publication: 2014.
• Original Publication Venue: Journal of Consumer Research, Vol. 40, No. 6, pp. 1097–1108.
• Copyright Status: The conceptual methodology and empirical findings are copyrighted by the Journal of Consumer Research, Inc., and published by Oxford University Press.
• Academic Usage and Fees: The scale items are publicly documented in the original academic paper. In accordance with standard academic practice, the scale may be used free of charge by academic researchers and students for non-commercial scientific research, provided appropriate attribution and formal citation are given to Zhang, Feick, and Mittal (2014).
• Commercial Applications: Commercial organizations, market research firms, or practitioners seeking to integrate the scale into proprietary commercial diagnostic software or monetize its application should verify licensing compliance through the Copyright Clearance Center or contact Oxford University Press / the original authors for guidance.

12. References

• Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T
• Berger, J. (2014). Word of mouth and interpersonal communication: A review and directions for future research. Journal of Consumer Psychology, 24(4), 586–607. https://doi.org/10.1016/j.jcps.2014.05.002
• Dichter, E. (1966). How word-of-mouth advertising works. Harvard Business Review, 44(6), 147–166.
• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.
• 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
• Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360–1380. https://doi.org/10.1086/225469
• Zhang, Y., Feick, L., & Mittal, V. (2014). How males and females differ in their likelihood of transmitting negative word of mouth. Journal of Consumer Research, 40(6), 1097–1108. https://doi.org/10.1086/674211

13. Items of the Scale