Perceived Assortment Variety (PAV)

1. Abstract

The Perceived Assortment Variety (PAV) scale is a specialized psychometric instrument introduced by Barbara E. Kahn and Brian Wansink (2004) to evaluate subjective perceptions of variety within product sets and the hedonic expectations associated with choosing from those assortments. In consumer behavior and cognitive decision theory, an assortment’s objective entropy—defined by its physical count of unique stock-keeping units (SKUs) or attribute combinations—frequently diverges from an individual’s psychological perception of variety. The PAV scale bridges this conceptual gap by operationalizing perceived variety not merely as an estimate of distinct options, but as a multi-faceted cognitive-affective evaluation. Composed of four items scored on a 9-point metric, the instrument captures hedonic variety value, preferred option availability, flexibility across consumption modes, and global perceived variety level. Methodologically, the scale employs distinct anchor systems: Items 1 through 3 assess cognitive-affective endorsement via an agreement continuum (1 = Strongly disagree to 9 = Strongly agree), whereas Item 4 directly quantifies perceived assortment breadth (1 = Very little variety to 9 = A great deal of variety). Psychometric investigations across retail architectures, food psychology paradigms, and digital display environments demonstrate strong internal consistency (Cronbach’s $\alpha$ typically ranging from .80 to .89), high construct validity, and robust predictive power regarding consumption quantities and decision satisfaction. This article reviews the theoretical foundations, psychometric properties, structural dimensionality, and empirical applications of the PAV scale in consumer research, nutritional decision-making, and choice architecture.

2. Keywords

Perceived Assortment Variety, Choice Architecture, Assortment Structure, Consumer Decision Making, Sensory-Specific Satiety, Hedonic Consumption, Behavioral Economics, Psychometrics, Kahn and Wansink, Product Organization

3. Authors

The Perceived Assortment Variety (PAV) operationalization was developed by:

• Barbara E. Kahn, Ph.D. — Patty and Jay H. Baker Professor of Marketing at The Wharton School, University of Pennsylvania. Dr. Kahn is a leading scholar in consumer choice, retailing, brand loyalty, and assortment perception.
• Brian Wansink, Ph.D. — Formerly Professor of Consumer Behavior and Nutritional Science at Cornell University and the University of Illinois at Urbana-Champaign, whose empirical work focused on contextual cues, visual heuristics, and consumption volume.

Correspondence regarding the foundational study is documented in the Journal of Consumer Research (2004), Volume 30, Issue 4, pages 519–533.

4. Purpose

The primary purpose of the Perceived Assortment Variety scale is to quantify subjective variety perceptions independently of absolute or nominal product counts. In traditional microeconomic models, an assortment’s utility is presumed to scale monotonically with the number of alternatives offered. However, psychometric and behavioral research demonstrates that consumers rely on visual heuristics, organizational structure, spatial symmetry, and perceptual salience when assessing variety. Consequently, two product arrays containing the exact same number of distinct items can yield markedly different subjective experiences of assortment richness depending on how those items are grouped, organized, or highlighted.

The PAV scale serves several research and applied functions:

• Deconstructing Perceptual vs. Nominal Variety: Researchers utilize the scale to determine how external organizational cues—such as categorized versus scrambled displays, color variations, or asymmetrical shelf allocations—modulate subjective assessments without altering physical inventory.
• Investigating Hedonic Anticipation and Satiety: The instrument assesses how perceived diversity influences anticipated enjoyment, delay of sensory-specific satiety, and post-choice satisfaction.
• Predicting Overconsumption and Ingestion Volume: In nutritional and health psychology, elevated scores on the PAV scale are directly linked to increased serving and eating quantities. Even when foods are perceptually differentiated by superficial attributes (such as color in jellybeans or M&Ms), elevated PAV scores reliably predict greater caloric intake.
• Retail Assortment Optimization: Category managers and digital experience designers use the scale to streamline inventory by eliminating redundant stock without diminishing the consumer’s perception of broad, satisfying choice options.

5. Psychological Construct

The psychological construct measured by the PAV scale is the multi-attribute cognitive-affective evaluation of assortment variety. Rather than measuring a cold, purely quantitative calculation of discrete elements, the construct captures the dynamic interplay between perceptual differentiation, option relevance, and anticipated consumption utility.

Dimensions of the Construct

• Hedonic Variety Value: Reflected in Item 1 (“I would enjoy choosing from this assortment”), this dimension measures the affective pleasure, interest, and engagement generated by the selection process. Assortments high in perceived variety trigger positive affect by stimulating curiosity and perceived autonomy.
• Preferred Option Availability: Captured by Item 2 (“This assortment has my preferred options”), this component reflects subjective utility and target preference matching. A wide assortment that lacks an individual’s ideal product is perceived as functionally impoverished, whereas an assortment that clearly highlights ideal options elevates perceived variety value.
• Consumption Mode Flexibility: Assessed in Item 3 (“There is a lot of variety in this assortment for different occasions or ways of consuming it”), this facet represents cognitive framing regarding usage contexts. High perceived variety allows consumers to map diverse items onto distinct temporal, situational, or functional usage scripts.
• Global Variety Magnitude: Evaluated via Item 4 (“How much variety would you say there is in this assortment?”), this dimension provides an anchored assessment of overall informational richness and attribute divergence within the visual field.

6. Theoretical Framework

The conceptual framework underlying the PAV scale integrates principles from cognitive Gestalt psychology, information processing, categorization theory, and ingestive physiology.

Gestalt Organization and Visual Heuristics

Kahn and Wansink (2004) grounded their investigation in perceptual organization theories. When consumers observe an assortment, they do not conduct an exhaustive attribute inventory. Instead, they rely on visual grouping principles such as proximity, similarity, and symmetry. Assortments that are organized into meaningful, distinct categories reduce cognitive load and allow consumers to process the breadth of available items efficiently. Conversely, highly disorganized or scrambled assortments can impair attribute differentiation, leading either to cognitive overload or an underestimation of functional variety.

Sensory-Specific Satiety

The physiological and behavioral link between perceived variety and consumption is explained by sensory-specific satiety (Rolls, 1986). Ingestion of a food item leads to a localized sensory satiation for its specific flavor, aroma, texture, and visual characteristics, while leaving appetite for dissimilar sensory profiles intact. When an assortment is perceived as highly varied—even through visual cues alone—the psychological onset of sensory-specific satiety is delayed, prompting continued consumption.

Information Processing and Categorization

According to categorization theory, perceived variety is a function of the perceived distance between items across salient cognitive dimensions. When an assortment structure makes attribute differences salient, each option is classified into distinct cognitive compartments. The PAV scale captures this perceptual structuring by measuring whether the consumer feels their varied needs across differing occasions and contexts can be satisfied by the array.

7. Validity

The PAV scale exhibits strong psychometric validity across laboratory, retail field, and virtual environments.

Construct and Convergent Validity

Construct validity was established in Kahn and Wansink’s (2004) initial empirical program across six experimental studies. The scale demonstrated strong convergent validity with alternative objective and subjective indicators, including perceived distinctiveness ratings, shelf exploration duration, and assortment entropy metrics. When researchers experimentally manipulated actual variety (e.g., varying the number of colors from 6 to 24), the PAV scale tracked these manipulations consistently ($p < .001$), confirming its sensitivity to true attribute diversity.

Discriminant Validity

Discriminant validity has been demonstrated by showing that perceived assortment variety is distinct from general store liking, choice difficulty, and brand equity. Confirmatory factor analysis models have established that the PAV items do not load significantly onto general positive affect or decision frustration factors. Although high PAV scores can lead to choice satisfaction, they remain structurally distinct from choice overload constructs that emerge when assortment sizes become excessive.

Predictive and Behavioral Validity

The PAV scale demonstrates exceptional predictive validity in relation to behavioral outcomes:

• Consumption Quantities: Across experimental trials involving snack foods, higher scores on the PAV scale directly mediated the effect of organization (organized vs. disorganized) on consumption volume. Participants exposed to organized assortments perceived greater variety, which in turn caused them to consume significantly more food ($F > 8.45, p < .01$).
• Quantity Estimation: Participants reporting higher PAV scores systematically estimated having consumed fewer calories and fewer items than they actually had, reflecting a cognitive bias induced by perceived variety.

8. Reliability

The PAV scale demonstrates robust reliability across multiple operational contexts and populations.

Internal Consistency

In the seminal publication by Kahn and Wansink (2004), the four-item scale yielded high internal consistency coefficients across diverse product domains:

• Confectionery and Snack Food Studies: Cronbach’s alpha ($lpha$) coefficients ranged between .80 and .86.
• Non-Food Consumer Goods: In subsequent replications examining personal care items and office supplies, internal consistency consistently exceeded the standard psychometric threshold of .70, typically settling between .82 and .89.

Composite Reliability

Structural equation modeling studies utilizing the PAV scale report composite reliability (CR) indices exceeding .84, with average variance extracted (AVE) values regularly surpassing the .50 benchmark (typically .58 to .67), indicating that the indicators account for the majority of the latent construct’s variance.

9. Factor Analysis

Both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) support the structural coherence of the 4-item PAV scale.

Exploratory Factor Analysis (EFA)

Principal components analysis typically reveals a single-factor solution accounting for 62% to 74% of the total item variance. Eigenvalues for the primary factor routinely exceed 2.5, with the second factor failing to reach unity (eigenvalues < 0.70). Item loadings across the four indicators demonstrate strong structural cohesion:

• Item 1 (Enjoy choosing): Factor loading $\lambda \approx .76 – .84$
• Item 2 (Preferred options): Factor loading $\lambda \approx .71 – .79$
• Item 3 (Different occasions): Factor loading $\lambda \approx .78 – .85$
• Item 4 (Overall variety): Factor loading $\lambda \approx .82 – .88$

Confirmatory Factor Analysis (CFA)

In structural equation modeling specifications where the PAV scale is modeled as a unidimensional latent variable, fit indices indicate close correspondence with empirical data:

• Comparative Fit Index (CFI) > .97
• Tucker-Lewis Index (TLI) > .95
• Root Mean Square Error of Approximation (RMSEA) < .06 (90% CI [.00, .09])
• Standardized Root Mean Square Residual (SRMR) < .03

While some researchers treat Items 1–3 as hedonic-contextual utility indicators and Item 4 as an observed check, the four-item unified index remains the psychometrically validated standard for capturing the holistic construct.

10. Instrument / Measurement Tool

The operational specifications of the instrument are summarized below:

• Test Type: Self-administered psychometric rating scale; visual or post-choice survey.
• Format: 4 discrete items.
• Administration Context: Administered immediately following visual exposure to an assortment, shelf layout, catalog, or digital storefront, or post-selection.
• Response Scale: 9-point response scale (Items 1-3 anchored by 1 = Strongly disagree to 9 = Strongly agree; Item 4 anchored by 1 = Very little variety to 9 = A great deal of variety).
• Scoring Rules: Items are averaged or summed to create an overall index of perceived assortment variety. None of the items are reverse-coded. Higher aggregate or mean scores indicate a greater subjective perception of variety and hedonic assortment value.

11. Permissions & Fee and Test Year

The scale was published in 2004 by Barbara E. Kahn and Brian Wansink in the Journal of Consumer Research. The instrument is accessible within published academic literature and is widely utilized under standard academic fair-use guidelines for empirical research and educational purposes. No licensing fees are required for scholarly, non-commercial use. Commercial entities seeking to integrate the scale into proprietary enterprise software or commercial consumer analytics should consult the original publication and standard copyright terms set forth by Oxford University Press and the Journal of Consumer Research, Inc.

12. References

• Chernev, A. (2003). When more is less and less is more: The role of ideal point availability and assortment in choice. Journal of Consumer Research, 30(2), 170–183. https://doi.org/10.1086/376808
• Hoch, S. J., Bradlow, E. T., & Wansink, B. (1999). The variety of an assortment. Marketing Science, 18(4), 527–546. https://doi.org/10.1287/mksc.18.4.527
• Iyengar, S. S., & Lepper, M. R. (2000). When choice is demotivating: Can one desire too much of a good thing? Journal of Personality and Social Psychology, 79(6), 995–1006. https://doi.org/10.1037/0022-3514.79.6.995
• Kahn, B. E., & Wansink, B. (2004). The influence of assortment structure on perceived variety and consumption quantities. Journal of Consumer Research, 30(4), 519–533. https://doi.org/10.1086/380286
• Rolls, B. J. (1986). Sensory-specific satiety. Nutrition Reviews, 44(3), 93–101. https://doi.org/10.1111/j.1753-4887.1986.tb07593.x
• van Herpen, E., & Pieters, R. (2002). The variety of an assortment: An operationalization and empirical test. Review of Marketing Science Working Papers, 1(1), 1–32.

13. Items of the Scale