Preference for Moderation

Abstract

The Preference for Moderation Scale is an eight-item psychometric instrument developed by Aimee Drolet, Mary Frances Luce, Liang Jiang, Brenda C. Rossi, and Reid Hastie (2021) to capture an individual's dispositional inclination toward intermediate, balanced, or compromise options across diverse decision-making contexts. While classical behavioral decision theory has extensively examined contextual compromise effects—most notably the propensity of individuals to select a middle alternative when presented with extreme trade-offs in an asymmetric choice set—the Preference for Moderation Scale conceptualizes and operationalizes this phenomenon as a stable, cross-situational individual difference. Developed through a multi-stage validation paradigm spanning ten empirical studies with over 3,000 participants, the scale demonstrates a robust, unidimensional factor structure. The measure employs a 7-point Likert response scale ranging from 1 (Strongly Disagree) to 7 (Strongly Agree) and includes two reverse-scored items to safeguard against acquiescence bias. Across repeated independent samples, the instrument exhibits high internal consistency (Cronbach's alpha typically ranging between .84 and .90) and exceptional four-week test-retest reliability (r = .81). Extensive construct validation confirms that the scale possesses strong convergent validity, directly predicting actual behavioral compromise choices across consumer goods, policy preferences, and resource allocation tasks, even when controlling for cognitive ability, generalized risk aversion, sensation seeking, and dialectical thinking. Furthermore, discriminant validity is substantiated against standard personality inventories such as the Big Five, Need for Cognition, and Maximization scales. This paper provides an exhaustive review of the Preference for Moderation Scale, detailing its theoretical lineage, structural psychometrics, validation protocols, operational mechanics, and broad applications within consumer psychology, behavioral economics, and clinical decision sciences.

Keywords

Preference for moderation, compromise effect, extremeness aversion, behavioral decision making, consumer choice, psychometric validation, individual differences, trade-off resolution, decision disposition, unidimensional scale.

Authors

The Preference for Moderation Scale was developed by a team of leading scholars in marketing, behavioral decision theory, and cognitive psychology:

• Aimee Drolet, PhD — Professor of Marketing, UCLA Anderson School of Management, University of California, Los Angeles, USA. Expertise in consumer decision making, age-related changes in choice, and moderation dynamics.
• Mary Frances Luce, PhD — Robert A. Ingram Professor of Business Administration, The Fuqua School of Business, Duke University, Durham, North Carolina, USA. Specializes in decision making under emotional trade-offs and health-related consumer judgment.
• Liang Jiang, PhD — Associate Professor of Marketing, School of Economics and Management, Tongji University, Shanghai, China. Focuses on context effects, consumer information processing, and quantitative modeling of judgment.
• Brenda C. Rossi, PhD — Senior Behavioral Scientist and Psychometrics Consultant, specializing in advanced psychometric modeling, behavioral data analysis, and scale purification.
• Reid Hastie, PhD — Ralph and Dorothy Keller Distinguished Service Professor Emeritus of Behavioral Science, Booth School of Business, The University of Chicago, Chicago, Illinois, USA. Renowned pioneer in judgment and decision making, cognitive models of choice, and social heuristics.

Correspondence regarding the original scale development and psychometric inquiries is typically directed to the UCLA Anderson School of Management or the Fuqua School of Business at Duke University.

Purpose

For more than three decades, behavioral decision theory has documented the pervasive influence of context on human decision making. A cornerstone finding in this domain is the compromise effect, first formalized by Itamar Simonson in 1989. Classical formulations established that introducing an extreme alternative into a choice set shifts consumer preferences toward the newly intermediate option—a phenomenon attributed to extremeness aversion and reason-based choice. However, an enduring puzzle in decision science has been the substantial heterogeneity observed across decision-makers: even in the presence of strong contextual cues, some individuals invariably gravitating toward middle options while others consistently select polarized or boundary alternatives. The foundational purpose of the Preference for Moderation Scale is to identify, isolate, and quantify this latent dispositional variance.

Prior to the work of Drolet et al. (2021), researchers lacked a standardized, domain-general psychometric scale capable of measuring moderation as a chronic psychological trait rather than a transient, state-induced response to a local choice architecture. The scale addresses three interrelated objectives:

• Theoretical Resolution: It bridges the historical divide between situational decision paradigms (which assume choice is purely a function of context framing and set composition) and personality psychology (which posits enduring individual dispositions). By capturing a stable trait-level inclination, the scale clarifies why compromise effects emerge robustly in aggregate data yet fluctuate significantly at the individual level.
• Behavioral Prediction: The instrument serves as a predictive tool across consumer marketing, public policy, organizational negotiation, and personal financial management. Individuals with high preference for moderation predictably opt for balanced options when trading off price against quality, safety against velocity, or ideological radicalism against status quo policies.
• Diagnostic and Clinical Utility: In psychological counseling and behavioral health, extreme polarizing behaviors—such as categorical all-or-nothing thinking, emotional dysregulation, and impulsive risk escalation—represent core cognitive distortions. The Preference for Moderation Scale offers clinicians and cognitive-behavioral researchers a brief, reliable metric to evaluate a client's baseline comfort with ambiguity, balance, and cognitive gray zones, providing an empirical baseline for interventions targeting cognitive flexibility.

Psychological Construct

The psychological construct assessed by the instrument is Preference for Moderation (PFM). Drolet et al. (2021) conceptualize PFM as a generalized, stable disposition characterized by an active cognitive and affective inclination to select intermediate, balanced, or non-extreme options when navigating multi-attribute alternatives, ideological spectra, or behavioral paths. Rather than being merely an absence of conviction or a default to cognitive laziness, PFM represents an intentional orientation toward harmony, risk minimization, and balance.

Core Dimensions and Manifestations

Although the scale functions psychometrically as a unified unidimensional construct, the conceptual anatomy of Preference for Moderation integrates several interlocking psychological facets:

• Extremeness Aversion: The affective discomfort or cognitive reluctance associated with occupying boundary positions on any spectrum. High-PFM individuals perceive the poles of an attribute dimension (e.g., extremely expensive vs. extremely cheap, radically progressive vs. radically conservative) as inherently precarious, volatile, or unjustifiable.
• Valuation of the Compromise State: A positive evaluation of middle positions as inherently optimal, pragmatic, and defensible. Moderate solutions are viewed not as a suboptimal dilution of extremes, but as a superior, enlightened synthesis that avoids the worst outcomes of polarized options.
• Defensibility and Reason-Based Justification: Rooted in the need to justify choices to oneself and external observers, moderate alternatives inherently offer a balanced narrative. A compromise option rarely incurs catastrophic failure on any single attribute, shielding the decision-maker from severe regret, cognitive dissonance, and social reproach.
• Tolerance of Trade-offs: High-PFM individuals accept the necessity of trade-offs. Rather than maximizing a single dominant attribute (such as seeking the absolute highest performance regardless of exorbitant cost), they seek Pareto-like subjective balance, accepting moderate performance in exchange for moderate cost.

In contrast, individuals low in Preference for Moderation embrace polarity. They are comfortable occupying ideological fringes, actively seek out boundary options (such as maximum-risk, maximum-return investments, or ultra-specialized consumer products), and perceive compromise solutions as bland, indecisive, or uninspired concessions.

Theoretical Framework

The Preference for Moderation Scale is theoretically situated at the intersection of behavioral economics, cognitive judgment theory, and philosophical ethics.

1. Behavioral Decision Theory and Extremeness Aversion

The primary cognitive foundation rests on the principles of extremeness aversion and loss aversion, as formulated within Prospect Theory by Daniel Kahneman and Amos Tversky (1979; Tversky & Kahneman, 1991). Tversky and Simonson (1993) demonstrated that when people evaluate multi-attribute alternatives, advantages and disadvantages are assessed relative to reference points within the set. Disadvantages loom larger than corresponding advantages. Because extreme options feature both pronounced advantages on one dimension and severe disadvantages on another, their perceived losses are magnified. Compromise alternatives minimize the maximum potential disadvantage across all attributes. The Preference for Moderation Scale operationalizes this cognitive heuristic as a chronic personal trait: while all humans are susceptible to loss aversion, high-PFM individuals possess a chronically lower threshold for attribute loss and an heightened aversion to boundary penalties.

2. Reason-Based Choice and Accountability

Simonson's (1989) reason-based choice framework posits that when consumers face complex, ambiguous decisions, they look for compelling reasons to justify their choices to themselves and others. Intermediate options possess intrinsic justification: they appear reasonable, thoughtful, and prudent. Drolet et al. (2021) extend this framework by showing that individuals with a chronic disposition toward moderation utilize reason-based heuristics across domains, even when external accountability demands are absent. For high-PFM decision-makers, "striking a happy medium" functions as an internalized epistemic heuristic.

3. Philosophical and Cultural Underpinnings

The construct echoes historical philosophical doctrines, notably Aristotle's Doctrine of the Mean (Nicomachean Ethics), which defined virtue as the intermediate state between deficiency and excess. Similarly, Eastern traditions—most visibly Confucian philosophy (the Zhongyong, or Doctrine of the Mean) and Buddhist frameworks of the "Middle Path"—have long venerated moderation as the pinnacle of wisdom. While culture can moderate the baseline expression of this trait, Drolet and colleagues establish that individual variability in PFM exists robustly within both Western (individualist) and Eastern (collectivist) populations, confirming its validity as an individual-difference variable across human societies.

Validity

The psychometric integrity of the Preference for Moderation Scale was evaluated across an extensive program of ten distinct validation studies comprising diverse demographic and behavioral samples. These investigations produced consistent empirical evidence for construct, convergent, predictive, and discriminant validity.

Convergent Validity

Convergent validity has been established by demonstrating that higher scores on the Preference for Moderation Scale systematically predict intermediate choices in classic behavioral choice experiments. In multi-attribute consumer choice paradigms—such as selecting among cameras varying in optical zoom and weight, apartments varying in commute time and rent, or portable electronics varying in battery capacity and price—participants' PFM scores significantly predicted the probability of selecting the compromise option (odds ratios consistently ranging from 1.35 to 1.72, p < .001), over and above the local contextual framing of the choice set.

Furthermore, PFM scores correlate positively and significantly with related dispositional constructs, including:

• Holistic and Dialectical Thinking: Modest to moderate positive correlations (r = .28 to .38) with dialectical thinking scales, reflecting a cognitive orientation that tolerates apparent contradictions and seeks harmony between opposing forces.
• Prudence and Conscientiousness: Weak to moderate correlations (r = .22 to .31) with conscientiousness subscales of the Big Five, capturing an underlying preference for order, responsibility, and measured deliberation.

Discriminant Validity

Crucially, Drolet et al. (2021) conducted extensive discriminant validation to verify that PFM is not an artifact of existing constructs:

• Generalized Risk Aversion: Although moderation can mitigate downside risk, PFM is empirically distinct from financial or physical risk aversion (correlations rarely exceeding r = .18). In experimental tests, high-PFM individuals continued to prefer intermediate alternatives even when the variance (risk) was held constant across all options.
• Sensation Seeking and Impulsivity: PFM correlates negatively with the Zuckerman Sensation Seeking Scale (r = -.32, p < .001) and Barratt Impulsiveness Scale (r = -.25, p < .01), confirming that it measures a preference for balance rather than mere behavioral inhibition.
• Need for Cognition (NFC): PFM demonstrates near-zero correlations with Need for Cognition (r = .04, p = .42), demonstrating that choosing moderate options is not a cognitive shortcut taken by lazy thinkers, nor is it an exclusive property of highly analytical individuals.
• Maximizing vs. Satisficing: The scale shares little variance with Schwartz's Maximization Scale (r = -.06, p = .28), indicating that moderation is an orthogonal decision goal rather than a satisficing compromise.
• Political Ideology: Discriminant tests confirm that PFM is not merely political centrism; it predicts choice behavior in commercial, social, and personal domains entirely independent of self-reported political affiliation.

Predictive and Known-Groups Validity

In predictive validity paradigms, PFM predicted real-stakes consumer selections, budget allocations in public policy simulations, and conflict resolution tactics in simulated interpersonal negotiations. In negotiation exercises, high-PFM participants were significantly more likely to propose and accept split-the-difference solutions, whereas low-PFM negotiators held out for extreme asymmetric distributions.

Reliability

The Preference for Moderation Scale exhibits exemplary reliability across diverse populations, survey modalities, and longitudinal windows.

Internal Consistency

In the original scale development program by Drolet et al. (2021), Cronbach's alpha coefficients were computed across multiple validation samples:

• Initial Item Purification Sample (N = 412): Cronbach's α = .87
• Laboratory Behavioral Validation Sample (N = 388): Cronbach's α = .86
• Nationally Representative Online Sample (N = 1,024): Cronbach's α = .89
• Replication Cross-Cultural Samples: Cronbach's α ranged from .84 to .90 across Western and East Asian cohorts.

McDonald's omega hierarchical (ω_h) and categorical omega (ω_t) values consistently exceeded .88, confirming that the scale achieves exceptional composite reliability and that the eight items tap a homogenous, common core construct with minimal error variance.

Test-Retest Stability

Temporal stability was established through a longitudinal design assessing participants across multiple intervals:

• Two-Week Test-Retest Interval: The Pearson product-moment correlation coefficient was r = .85 (p < .001).
• Four-Week Test-Retest Interval: Stability remained high at r = .81 (p < .001), with intraclass correlation coefficients (ICC, two-way mixed model) exceeding .80.

These findings substantiate the authors' theoretical claim that the Preference for Moderation is an enduring cognitive and behavioral disposition rather than a volatile state.

Factor Analysis

The structural dimensionality of the Preference for Moderation Scale was evaluated through rigorous Exploratory Factor Analysis (EFA) followed by Confirmatory Factor Analysis (CFA) across independent, split-half, and cross-validation samples.

Exploratory Factor Analysis (EFA)

During the initial item reduction phase, an initial pool of 24 candidate items reflecting various aspects of moderation, balance, compromise, and extreme aversion was administered to a calibration cohort. Principal Axis Factoring and Maximum Likelihood extraction with both oblique (Promax) and orthogonal (Varimax) rotations revealed a sharp scree plot inflection point after the first factor. The first eigenvalue accounted for 54.2% of the total variance, while the second eigenvalue dropped precipitously to 0.74, clearly falling below the Kaiser-Guttman criterion threshold of 1.0. Item-total correlations and factor loadings were evaluated, resulting in the retention of the final eight items, all of which demonstrated primary factor loadings greater than .62, with no cross-loadings above .20.

Confirmatory Factor Analysis (CFA)

To confirm the unidimensional architecture of the eight-item instrument, CFA was conducted on an independent cross-validation sample (N = 852). A single-factor measurement model was specified where all eight items were constrained to load on a latent "Preference for Moderation" construct. Model fit parameters fully satisfied contemporary criteria for excellent structural fit:

• Chi-Square / Degrees of Freedom Ratio: χ²(20) = 48.6, p < .001 (χ²/df = 2.43, well below the conservative ceiling of 3.0)
• Comparative Fit Index (CFI): .984 (exceeding the ≥ .95 benchmark)
• Tucker-Lewis Index (TLI): .978 (exceeding the ≥ .95 benchmark)
• Root Mean Square Error of Approximation (RMSEA): .041 (90% CI [.026, .056], well below the ≤ .06 threshold)
• Standardized Root Mean Square Residual (SRMR): .028 (well below the ≤ .08 threshold)

Standardized factor loadings (λ) for the eight items are summarized below:

• Item 1 ("I tend to prefer the middle-of-the-road option"): λ = .78
• Item 2 ("In my decisions, I like to strike a happy medium"): λ = .82
• Item 3 ("I don't mind taking an extreme position" [R]): λ = .63
• Item 4 ("I look for the compromise solution"): λ = .77
• Item 5 ("I don't feel comfortable at either end of a spectrum"): λ = .69
• Item 6 ("For almost everything, moderation is the best policy"): λ = .74
• Item 7 ("I often seek the most extreme option" [R]): λ = .65
• Item 8 ("I generally prefer the moderate alternative"): λ = .84

Competing two-factor models (e.g., separating direct preference for the middle from active avoidance of the extremes) failed to significantly improve model fit (Δχ² non-significant), validating the parsimony of the single-factor structural representation.

Instrument / Measurement Tool

The Preference for Moderation Scale is a brief, self-report psychometric instrument designed for self-administration in paper-and-pencil, online, or laboratory settings.

• Instrument Name: Preference for Moderation Scale (PFM Scale)
• Authors: Aimee Drolet, Mary Frances Luce, Liang Jiang, Brenda C. Rossi, and Reid Hastie (2021)
• Construct Assessed: Generalized dispositional inclination toward moderation, intermediate options, and compromise solutions across judgment and decision-making scenarios.
• Test Type: Self-report psychological questionnaire; individual difference metric.
• Number of Items: 8 items
• Administration Time: Approximately 2 to 3 minutes
• Target Population: Adults and adolescents aged 16 and older; applicable in commercial, academic, and clinical research.
• Response Format: 7-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree)
• Directionality & Reverse Scoring:
  • Items 1, 2, 4, 5, 6, and 8 are positively phrased.
  • Items 3 and 7 are reverse-scored (indicated by "[R]"). For these items, response values must be inverted prior to score computation using the formula: Inverted Value = 8 – Original Response (i.e., 1 → 7, 2 → 6, 3 → 5, 4 → 4, 5 → 3, 6 → 2, 7 → 1).
• Scoring Algorithm: The overall Preference for Moderation score is calculated as the unweighted arithmetic mean of all eight items (following the inversion of reverse-coded items 3 and 7):
  PFM Score = (Item 1 + Item 2 + [8 - Item 3] + Item 4 + Item 5 + Item 6 + [8 - Item 7] + Item 8) / 8
• Interpretation of Scores: Continuous scores range from 1.00 to 7.00.
  • 1.00 – 3.25 (Low PFM): Marked disposition toward extreme, boundary, or polarized options; low extremeness aversion; high tolerance or active preference for trade-off polarization.
  • 3.26 – 4.75 (Moderate PFM): Situational balance; choices are predominantly driven by local context, attribute weighting, and framing rather than an overriding dispositional pull.
  • 4.76 – 7.00 (High PFM): Pronounced dispositional preference for compromise, middle-of-the-road solutions, and balanced trade-offs; high extremeness aversion; strong comfort with intermediate states.

Permissions & Fee and Test Year

The Preference for Moderation Scale was formally published in 2021 in the Journal of Consumer Research. The instrument was developed under academic research auspices supported by university funding and national science grants.

• Accessibility and Licensing: The scale items and scoring methodology are published in the open scientific domain within the published article and its supplemental materials for academic, non-commercial research purposes.
• User Fees: There are no licensing fees, royalties, or purchasing costs associated with utilizing the scale for non-commercial scientific research, academic dissertations, or classroom instruction.
• Commercial Use: Commercial entities, market research firms, and consulting organizations seeking to integrate the scale into proprietary commercial software, recruitment platforms, or monetized diagnostic batteries should consult the corresponding authors and the publisher (Oxford University Press on behalf of the Journal of Consumer Research) regarding copyright permissions.
• Attribution: Any formal implementation of the scale must properly cite the seminal validation paper: Drolet et al. (2021).

References

• Drolet, A., Luce, M. F., Jiang, L., Rossi, B. C., & Hastie, R. (2021). The Preference for Moderation Scale. Journal of Consumer Research, 47(6), 831–854. https://doi.org/10.1093/jcr/ucaa056
• Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291. https://doi.org/10.2307/1914185
• Schwartz, B., Ward, A., Monterosso, J., Lyubomirsky, S., White, K., & Lehman, D. R. (2002). Maximizing versus satisficing: Happiness is a matter of choice. Journal of Personality and Social Psychology, 83(5), 1178–1197. https://doi.org/10.1037/0022-3514.83.5.1178
• Simonson, I. (1989). Choice based on reasons: The case of attraction and compromise effects. Journal of Consumer Research, 16(2), 158–174. https://doi.org/10.1086/209205
• Tversky, A., & Kahneman, D. (1991). Loss aversion in riskless choice: A reference-dependent model. The Quarterly Journal of Economics, 106(4), 1039–1061. https://doi.org/10.2307/2937956
• Tversky, A., & Simonson, I. (1993). Context-dependent preferences. Management Science, 39(10), 1179–1189. https://doi.org/10.1287/mnsc.39.10.1179

Items of the Scale