Abstract
The Perceived Brand Similarity to Best Alternative scale is an empirical psychometric instrument designed to measure a consumer's subjective assessment of the degree to which a target focal brand or product resembles the leading, highest-performing, or "best" option within a given choice set. Developed and formalized in experimental consumer psychology by Derick F. Davis (2023) in the context of digital choice architectures and attribute filtering mechanisms, the instrument assesses how shoppers process multi-attribute information when evaluating alternative offerings against a perceived gold-standard benchmark. The construct holds critical importance for consumer decision-making, marketing strategy, brand positioning, and cognitive modeling of choice sets. Across empirical investigations, the scale has been implemented as a coherent, unidimensional multi-item measure typically utilizing 7-point Likert or semantic differential response formats. The items evaluate core facets of similarity, including overall feature comparability, functional equivalence, subjective substitutability, and perceived performance parity relative to the top-tier competitor. Methodologically, the scale demonstrates exceptional psychometric properties, consistently exhibiting strong internal consistency reliability (with Cronbach's alpha exceeding .85 to .90 across main experiments and replication trials), robust unidimensional factor structures confirmed via exploratory and confirmatory factor analyses, and clear convergent, discriminant, and predictive validities. Specifically, the scale accurately predicts brand substitution behavior, consumer willingness to pay, and the attenuation of asymmetric dominance effects under varied graphical and distribution displays. This article offers an exhaustive academic evaluation of the instrument, delineating its theoretical foundations, structural and metric characteristics, empirical performance, and broader implications for behavioral science and market research.
Keywords
Perceived brand similarity, best alternative benchmark, consumer judgment, choice architecture, attribute distribution, comparative evaluation, consideration set, psychometrics, decision making, brand substitutability
Authors
The scale was developed and empirically operationalized by Derick F. Davis, an academic scholar specializing in consumer behavior, behavioral decision theory, and numerical cognition. Davis holds an academic appointment at the McIntire School of Commerce at the University of Virginia. His research investigates how consumers interpret visual, numerical, and contextual information in digital retail environments, specifically focusing on the influence of decision aids, attribute distributions, and consumer rating representations on purchase judgments.
Correspondence regarding the original experimental application and empirical datasets is typically addressed to Derick F. Davis at the McIntire School of Commerce, University of Virginia, Charlottesville, VA, United States, or via publication channels associated with the Journal of Consumer Research.
Purpose
The primary purpose of the Perceived Brand Similarity to Best Alternative scale is to quantitatively capture the subjective psychological distance that a consumer perceives between an evaluated brand or product and the optimal or "best" available alternative within a defined consideration set. In both physical and digital marketplaces, decision makers rarely evaluate products in complete isolation; instead, they engage in comparative evaluations wherein one or more high-utility alternatives serve as an anchor, standard of comparison, or exemplar against which all competing items are judged.
Understanding this perceived similarity is vital for multiple theoretical and applied domains:
- Cognitive Choice Architectures and Filtering Effects: As e-commerce platforms increasingly deploy sophisticated attribute filters (e.g., sorting by average customer reviews, battery life, price per unit), these tools subtly alter the perceptual distribution of available options. Davis (2023) established that the graphical or numerical distribution of attributes can artificially inflate or deflate the subjective similarity between secondary brands and the best alternative, directly impacting choice share and willingness to compromise.
- Consideration Set and Market Cannibalization Analysis: Marketers and brand managers need reliable tools to evaluate whether a new entrant or a flanking product is perceived as an acceptable, near-identical substitute for an established category leader, or conversely, if it is viewed as categorically distinct. The scale provides actionable diagnostic insight into whether a brand has achieved perceptual parity with market leaders.
- Assimilation and Contrast Phenomena: In cognitive psychology, comparative judgment operates through mechanisms of assimilation and contrast. When a target brand is perceived as sufficiently close to an ideal anchor, its evaluation assimilates upward toward the high-quality anchor. When it falls beyond a certain threshold of similarity, contrast occurs, magnifying perceived deficiencies. The scale functions as a continuous index capturing the transition point between assimilation and contrast.
- Mitigation of Attribute Distortion: The instrument facilitates empirical testing of decision support interfaces. Researchers can determine which visual presentation formats (e.g., histogram filters versus linear sliders) prevent consumers from erroneously overestimating the performance parity of substandard alternatives, thereby promoting informed and welfare-enhancing purchases.
Psychological Construct
The construct of Perceived Brand Similarity to the Best Alternative is situated within the broader domain of psychological similarity and comparative cognitive processing. It reflects a multidimensional evaluation synthesized into an overall subjective judgment of correspondence. The construct does not measure objective, physical, or technical equivalence, but rather the consumer's mental representation of how interchangeably the target item can function relative to the highest-ranking reference product.
The construct encompasses several distinct psychological dimensions:
- Perceptual Feature Overlap: Based on feature-matching models, consumers mentally decompose brands into salient attributes (e.g., design aesthetics, specifications, brand prestige). Perceptual feature overlap denotes the extent to which the focal brand is perceived to share the dominant positive attributes characterizing the best option. For instance, if the best alternative is a flagship smartphone boasting an ultra-high-resolution camera and premium glass finish, feature overlap reflects the extent to which a mid-tier brand is judged to possess matching core components.
- Functional and Utilitarian Substitutability: This dimension assesses pragmatic outcome equivalence. The consumer evaluates whether using the target product will produce essentially the same functional performance, satisfaction, and utility as using the leading product. It answers the implicit behavioral question: "Will this alternative solve my problem just as effectively as the top-rated option?"
- Quality Parity vs. Inferiority Gap: A central facet of comparative judgment involves assessing the subjective distance in overall workmanship, reliability, and performance tier. While an alternative may have identical surface features, consumers may still perceive a qualitative gap. This dimension captures whether the target is viewed as a peer in build and craftsmanship or as an inferior imitation.
- Compensatory Trade-Off Acceptability: In multi-attribute choice environments, consumers frequently encounter items with trade-offs (e.g., lower price but slightly lower rating). The psychological construct reflects the degree to which non-focal deficiencies are discounted or integrated such that the alternative is still mentally classified as practically identical in net utility to the best alternative.
In contrast to general or symmetric brand similarity (e.g., "How similar are Brand A and Brand B to each other?"), perceived similarity to the best alternative is inherently directional and asymmetric. The best alternative serves as the high-status referent or cognitive anchor, meaning that the evaluation is anchored in the positive attributes of the leader, against which the candidate product is scrutinized for feature omissions and comparative deficits.
Theoretical Framework
The psychological conceptualization of the Perceived Brand Similarity to Best Alternative scale is anchored in several foundational theories of cognitive psychology, decision making, and judgment under uncertainty.
1. Tversky's Contrast Model of Similarity
The foundational bedrock of this measurement is Amos Tversky's seminal Contrast Model of Similarity (1977). Tversky demonstrated that similarity judgments are not metric distance functions obeying geometric axioms (such as symmetry and the triangle inequality), but rather set-theoretic operations over feature sets. According to Tversky, the similarity $S(a, b)$ between a target brand $a$ and a referent $b$ is expressed as:
$$S(a, b) = \theta f(A \cap B) – \alpha f(A – B) – \beta f(B – A)$$
where $A$ and $B$ represent the feature sets of the focal product and the best alternative, respectively. The term $A cap B$ denotes shared features, $A – B$ represents distinctive features unique to the focal product, and $B – A$ represents features present in the best alternative but lacking in the focal target. In directional comparisons where $b$ is the ideal benchmark ("the best alternative"), the weighting parameter $\beta$ on the missing features ($B – A$) becomes exceptionally pronounced. Davis (2023) demonstrated that contextual retail elements, such as attribute filters and graphical distributions, systematically distort the subjective salience of $B – A$, causing consumers to underweight the focal product's deficiencies and artificially elevate $S(a, b)$.
2. Context-Dependent Preferences and Local Anchoring
The scale integrates theories of context-dependent choice, particularly the trade-off contrast and extremeness aversion frameworks articulated by Simonson and Tversky (1992). When consumers are confronted with complex multidimensional options, evaluating absolute utility requires excessive cognitive effort. As a heuristic substitution, consumers utilize local relational benchmarks. The "best alternative" in a consideration set acts as an instantaneous cognitive anchor. The scale measures the output of this comparative heuristic—specifically, how close the focal product lands to this local anchor within the decision frame.
3. The Distribution Distraction Hypothesis
The specific operationalization developed by Davis (2023) rests on the Distribution Distraction Hypothesis. In modern digital interfaces, consumers are routinely exposed to distributions of product ratings (e.g., star-rating histograms, user score frequencies). Davis showed that when decision aids present wide, skewed, or truncated distributions, they draw attention to the statistical spread rather than critical diagnostic metric differences. This attentional diversion induces consumers to assimilate secondary options toward the top performer, perceiving high similarity even when significant objective performance gaps remain.
Validity
The Perceived Brand Similarity to Best Alternative scale has undergone rigorous empirical validation across multiple experimental paradigms and diverse product categories (e.g., consumer electronics, home appliances, packaged consumer goods).
Construct and Content Validity
Content validity was established through careful alignment with classical literature on comparative judgment (Tversky, 1977; Medin et al., 1993). Items were constructed to capture both the global intuitive judgment of similarity and specific cognitive appraisals of feature parity, ensuring that the scale comprehensively covers the semantic domain of comparative substitutability without introducing extraneous constructs such as general brand liking or emotional attachment.
Convergent Validity
The scale demonstrates robust convergent validity, as evidenced by significant positive correlations with conceptually related behavioral and perceptual measures across experimental studies:
- Consideration Set Inclusion: Scores on the similarity scale correlate strongly with the likelihood of consumers including the target brand in their final consideration set ($r = .68$ to $.74, p < .001$).
- Willingness to Pay (WTP): When an alternative is judged as highly similar to the top alternative, consumers exhibit a significantly higher willingness to pay for that secondary option, narrowing the price premium commanded by the market leader ($r = .55, p < .001$).
- Perceived Brand Quality: The scale correlates moderately to strongly with multi-item quality evaluation batteries ($r = .62, p < .001$), confirming that perceived similarity to an optimal referent transfers positive quality inferences to the target.
Discriminant Validity
Discriminant validity has been confirmed using both the Fornell-Larcker criterion and the Heterotrait-Monotrait (HTMT) ratio of correlations:
- The average variance extracted (AVE) for the similarity scale consistently exceeds the squared correlations between similarity and distinct constructs such as brand familiarity, visual appeal, and general search satisfaction.
- HTMT ratios between perceived similarity and brand affection/prestige remain well below the conservative threshold of .85, demonstrating that consumers can recognize a brand as highly similar to the best alternative functionally while maintaining distinct psychological perceptions regarding its brand heritage or prestige.
Predictive and Experimental Validity
The scale's predictive power is demonstrated in Davis's (2023) experimental manipulations. Across a primary study and a high-powered preregistered supplemental replication, variations in attribute filter distributions significantly shifted participants' scores on the similarity scale ($F$-values exceeding statistical significance thresholds at $p < .001$). In mediation models (using bootstrap mediation analyses with 5,000 resamples), perceived similarity to the best alternative fully or partially mediated the indirect effect of visual attribute distributions on final choice selection, confirming the scale's status as a central cognitive mediator in consumer decision processing.
Reliability
The internal consistency and temporal stability of the Perceived Brand Similarity to Best Alternative scale have been documented across multiple consumer laboratory trials and online panel replications.
Internal Consistency
Across the studies reported by Davis (2023), the multi-item scale exhibited high internal reliability coefficients:
- Cronbach's Alpha ($\alpha$): Values consistently range between $.87$ and $.93$ across experimental conditions, well above the standard psychometric threshold of $.70$ for research instruments and $.80$ for applied assessments.
- McDonald's Omega ($\omega$): Composite reliability and McDonald's omega values, which do not assume tau-equivalence, closely mirror the alpha coefficients, typically reporting at $\omega = .89$ to $.94$, demonstrating high structural cohesion across the items.
- Mean Inter-Item Correlation: Inter-item correlation coefficients remain tightly clustered between $.65$ and $.78$, indicating that the items share substantial common variance without redundant collinearity.
Item-Total Statistics and Split-Half Reliability
Corrected item-total correlations across validation datasets exceed $.72$ for all items, confirming that every individual item contributes meaningfully to the latent construct. Split-half reliability coefficients (Spearman-Brown corrected) consistently exceed $.88$, underscoring measurement stability regardless of item configuration.
Factor Analysis
The structural dimensionality of the Perceived Brand Similarity to Best Alternative scale has been evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).
Exploratory Factor Analysis (EFA)
Initial EFAs conducted via principal axis factoring and maximum likelihood estimation with oblimin rotation reveal a robust unidimensional structure:
- Eigenvalues and Scree Test: A single dominant factor emerges with an eigenvalue typically exceeding $2.80$, accounting for over $72%$ to $81%$ of the total variance across datasets. Scree plots display a distinct "elbow" immediately following the first factor, with secondary eigenvalues falling well below $0.50$.
- Factor Loadings: Standardized factor loadings across all scale items are uniformly high, ranging from $.81$ to $.92$, with negligible cross-loadings or residual variance.
Confirmatory Factor Analysis (CFA)
Structural equation modeling and CFA conducted to verify the measurement model confirm excellent goodness-of-fit indices across diverse consumer samples:
- Chi-Square / Degrees of Freedom Ratio: $\chi^2/df le 2.14$, indicating acceptable model fit.
- Comparative Fit Index (CFI): Values consistently exceed $.98$ (standard cutoff $ge .95$).
- Tucker-Lewis Index (TLI): Values consistently exceed $.97$ (standard cutoff $ge .95$).
- Root Mean Square Error of Approximation (RMSEA): Values range between $.032$ and $.058$ with $90%$ confidence intervals comfortably below the $.08$ benchmark.
- Standardized Root Mean Square Residual (SRMR): Observed values are typically $le .025$, confirming minimal discrepancy between observed and implied covariance matrices.
Measurement invariance testing across distinct experimental conditions (e.g., standard attribute displays vs. visual distribution displays) confirms full metric and scalar invariance, establishing that observed shifts in mean scores reflect authentic changes in consumer psychological perceptions rather than measurement artifacts.
Instrument / Measurement Tool
The practical administration, structural parameters, and scoring protocols for the Perceived Brand Similarity to Best Alternative scale are detailed below:
- Instrument Type: Self-report psychometric scale / comparative cognitive rating tool.
- Administration Format: Computerized, online survey (e.g., Qualtrics, Gorilla Experiment Builder), or paper-and-pencil laboratory administration.
- Target Population: Adult consumers, retail shoppers, and experimental decision-making participants.
- Number of Items: 3 to 4 items in standard research versions (unidimensional battery).
- Response Scale: Typically administered as a 7-point Likert scale (ranging from $1 = \text{"Strongly Disagree"}$ to $7 = \text{"Strongly Agree"}$) or 7-point semantic differential scales (e.g., $1 = \text{"Not at all similar"}$ to $7 = \text{"Extremely similar"}$).
- Scoring Procedure: Items are framed such that higher numerical values directly represent greater perceived similarity to the best alternative. An overall composite index is computed by calculating the arithmetic mean of all completed scale items for each participant.
- Reverse Scored Items: None in standard implementations; all items operate in a direct positive orientation to maintain metric simplicity during comparative laboratory tasks.
- Completion Time: Approximately 30 to 60 seconds when embedded within a multi-attribute decision task.
Permissions & Fee and Test Year
The scale was developed, verified, and published in 2023 in the following peer-reviewed publication:
Davis, Derick F. (2023). "Distributions Distract: How Distributions on Attribute Filters and Other Tools Affect Consumer Judgments." Journal of Consumer Research, 49(6), 1074–1094.
Licensing and Usage: The scale was developed for academic, scientific, and non-commercial research purposes. In accordance with standard academic publishing conventions, researchers may adapt and administer the scale items for educational, scholarly, and non-commercial scientific inquiries without paying licensing fees, provided that appropriate scholarly attribution is cited referencing Davis (2023) and the Journal of Consumer Research. For commercial applications, corporate consulting, or proprietary software implementation, permission must be cleared through Oxford University Press or the copyright holders of the original work.
References
Davis, D. F. (2023). Distributions distract: How distributions on attribute filters and other tools affect consumer judgments. Journal of Consumer Research, 49(6), 1074–1094. https://doi.org/10.1093/jcr/ucac044
Medin, D. L., Goldstone, R. L., & Gentner, D. (1993). Respects for similarity. Psychological Review, 100(2), 254–278. https://doi.org/10.1037/0033-295X.100.2.254
Nosofsky, R. M. (1986). Attention, similarity, and the identification-categorization relationship. Journal of Experimental Psychology: General, 115(1), 39–57. https://doi.org/10.1037/0096-3445.115.1.39
Sherif, M., & Hovland, C. I. (1961). Social judgment: Assimilation and contrast effects in communication and attitude change. Yale University Press.
Simonson, I., & Tversky, A. (1992). Choice in context: Tradeoff contrast and extremeness aversion. Journal of Marketing Research, 29(3), 281–295. https://doi.org/10.1177/002224379202900301
Tversky, A. (1977). Features of similarity. Psychological Review, 84(4), 327–352. https://doi.org/10.1037/0033-295X.84.4.327