Ambivalence of Product Evaluation (APE)

1. Abstract

The Ambivalence of Product Evaluation (APE) scale is a specialized psychometric instrument developed by Stephen M. Nowlis, Barbara E. Kahn, and Ravi Dhar (2002) to quantify the subjective experience of evaluative conflict, psychological tension, and affective-cognitive contradiction encountered by individuals during consumer choice and comparative product appraisals. Grounded in structural models of attitude ambivalence, the instrument operationalizes the subjective state wherein an evaluator simultaneously holds positive and negative evaluations toward an assortment, choice set, or specific multi-attribute alternatives. Comprising three precisely framed self-report items administered on a 7-point semantic differential and rating scale, the instrument captures three core phenomenological dimensions of evaluative ambivalence: experienced mixed feelings, subjective decisional difficulty regarding valence discernment, and the structural tension between one-sidedness versus mixed reactions.

Psychometrically, the APE exhibits exceptional internal consistency reliability across varied experimental and consumer behavioral settings, consistently demonstrating Cronbach’s alpha coefficients exceeding α = .80, and reaching up to α = .87 to .91 across empirical replications. Confirmatory factor analytic investigations establish a unidimensional latent structure with standardized factor loadings typically spanning .78 to .91, accounting for substantial common variance and demonstrating strict cross-contextual metric invariance. The scale possesses robust convergent validity with physiological markers of decisional conflict, choice deferral rates, and response latency paradigms, while maintaining discriminant validity from related constructs such as generalized trait indecisiveness, task involvement, cognitive overload, and basic post-decisional dissonance. By serving as an operational bridge between social psychological theories of attitudinal duality and behavioral decision research, the APE scale provides researchers and applied practitioners with an efficient, highly sensitive diagnostic tool to assess how assortment composition, the presence or absence of neutral default options, attribute trade-offs, and preference uncertainty moderate downstream judgment, brand loyalty, and choice deferral.

2. Keywords

Ambivalence of Product Evaluation, APE scale, subjective ambivalence, consumer attitude, evaluative conflict, choice deferral, neutral option removal, attitudinal duality, behavioral decision making, psychometrics

3. Authors

The Ambivalence of Product Evaluation scale was conceptualized, operationalized, and validated by a team of prominent scholars in the fields of marketing, consumer psychology, and behavioral economics:

• Stephen M. Nowlis, Ph.D. — August A. Busch Jr. Distinguished Professor of Marketing at the Olin Business School, Washington University in St. Louis. Dr. Nowlis is an internationally recognized authority on consumer decision making, sensory marketing, choice conflict, and product evaluation dynamics. His empirical scholarship explores how contextual cues, comparative assortment framing, and choice architecture influence consumer preferences and judgment volatility.
• Barbara E. Kahn, Ph.D. — Patty and Jay H. Baker Professor of Marketing at The Wharton School of the University of Pennsylvania. Dr. Kahn is a pioneering researcher in consumer choice variety, assortment structure, brand perception, and retail merchandising. Her research investigates the psychological mechanisms underlying consumer curiosity, choice satiation, and the cognitive consequences of navigating complex product arrays.
• Ravi Dhar, Ph.D. — George Rogers Clark Professor of Management and Marketing and Director of the Center for Customer Insights at the Yale School of Management. Dr. Dhar is a world-renowned expert in behavioral decision theory, consumer motivation, heuristic processing, and choice conflict resolution. His seminal research on choice deferral, trade-off avoidance, and consumer goal pursuit has fundamentally reshaped modern behavioral marketing science.

Correspondence regarding the original development and empirical deployment of the instrument was anchored through the foundational publication in the Journal of Consumer Research (Nowlis, Kahn, & Dhar, 2002).

4. Purpose

The primary purpose of the Ambivalence of Product Evaluation (APE) scale is to capture and quantify the immediate, consciously experienced state of evaluative conflict that emerges when consumers evaluate product offerings characterized by competing, contradictory, or non-compensatory attributes. In traditional expected utility frameworks and classical compensatory models of attitude formation (such as the Fishbein multi-attribute model), attitudes were routinely conceptualized as unidimensional bipolar continua ranging from purely negative to purely positive. Such paradigms implicitly assumed that positive and negative beliefs cancel one another out algebraically, yielding a single neutral or moderate score if a consumer perceives both pros and cons. The APE scale was specifically formulated to dismantle this measurement artifact by isolating true subjective ambivalence—the active, conflicted tension of possessing strong co-occurring positive and negative reactions—from mere evaluative indifference, neutrality, or unformed preferences.

In experimental consumer research, the APE scale was introduced to resolve critical theoretical questions regarding how choice architecture and assortment framing manipulate cognitive-affective equilibrium. A central empirical purpose highlighted in Nowlis, Kahn, and Dhar (2002) is the examination of what occurs when consumers are compelled to make trade-offs under conditions where an easy default or neutral middle alternative is systematically withdrawn. Under standard choice environments, ambivalent consumers frequently gravitate toward compromise options, middle-ground products, or completely defer the decision. When these coping strategies are blocked (e.g., in forced binary choices), the latent evaluative conflict intensifies dramatically. The APE instrument enables researchers to capture this heightened psychological friction as a mediating or moderating mechanism, directly explaining why forcing choices in ambivalent states causes subsequent polarization, post-choice regret, or systematic shifts in attitude certainty.

Beyond academic laboratory investigations, the APE scale serves vital applied functions across market research, user experience (UX) testing, and new product development (NPD). Organizations introducing innovative, disruptive, or multi-attribute hybrid products (e.g., electric vehicles with high technological appeal but range limitations; ultra-healthy functional foods with novel taste profiles) routinely confront ambivalent consumer reactions. Applying the APE scale allows brand strategists to diagnose whether sluggish sales or high cart abandonment rates stem from lack of interest (indifference) or from high-intensity friction between polarized attributes (ambivalence). Because the scale consists of only three targeted questions, it incurs minimal respondent burden, making it exceptionally feasible for inclusion in high-throughput digital shopping experiments, field surveys, and fast-paced eye-tracking or neuro-marketing protocols.

5. Psychological Construct

The psychological construct assessed by the APE scale is subjective (or felt) evaluative ambivalence within the domain of comparative product evaluation. Contemporary social cognition and psychometric literature distinguish rigorously between two operational manifestations of ambivalence: objective (structural) ambivalence and subjective (felt) ambivalence (Priester & Petty, 1996; Thompson et al., 1995). Objective ambivalence refers to the independent, mathematical co-existence of separate positive and negative cognitions or ratings (often calculated via formulas such as the Griffin formula or Kaplan’s separate bivariate scales). In contrast, subjective ambivalence represents the holistic, phenomenological meta-cognitive appraisal that one’s internal psychological state is characterized by internal strife, indecisiveness, and affective confusion. The APE scale directly targets this subjective meta-cognitive reality.

The construct measured by the APE comprises three tightly interconnected facet manifestations that collectively reflect the unified latent construct of evaluative ambivalence:

1. Experienced Mixed Feelings (Affective and Cognitive Co-activation)

This facet captures the direct phenomenological awareness of affective duality. Rather than feeling a harmonious, single-toned reaction toward the target products, the evaluator reports simultaneous attraction and repulsion, or simultaneous satisfaction and disappointment with different facets of the assortment. For example, a consumer assessing two competing laptop computers may experience strong excitement regarding the premium aesthetic and ultra-fast processing speed of Option A, coupled with acute distress regarding its exorbitant price and lack of essential connectivity ports. The co-activation of these polarized affective responses produces an unmistakable conscious sense of having “mixed feelings.”

2. Valence Decisional Difficulty

This facet evaluates the cognitive impedance associated with assigning an overall summary evaluation to the product alternatives. Evaluative judgments require individuals to integrate disparate attribute values into an overall liking or disliking metric. When products embody acute trade-offs (e.g., superior quality paired with abysmal customer service, or high sustainability paired with inconvenient functionality), the mental algebra required to synthesize these inputs breaks down. Decisional difficulty in this context does not simply mean mental exhaustion; it signifies the specific inability to comfortably declare whether one overall “likes” or “dislikes” the focal items.

3. Oppositional Tension (One-Sidedness vs. Mixed Reactions)

This facet reflects the structural polarity perceived by the judge regarding their internal state. In univalent attitudes, evaluators experience their perspectives as unequivocal, clear-cut, and unidirectional (“one-sided”). In contrast, evaluative ambivalence is defined by non-convergence—the explicit perception that one’s reactions are multi-directional, fractured, and resisting resolution. The respondent recognizes that their internal assessment is not a cohesive stance, but rather a clash between competing, un-reconciled evaluative poles.

The APE construct must be psychometrically demarcated from several contiguous constructs. It is distinct from trait indecisiveness, which represents a stable, cross-situational personality predisposition toward procrastination and decision avoidance; the APE captures a state-based reaction elicited by specific stimulus characteristics and choice contexts. It is likewise distinct from post-decisional dissonance (Festinger, 1957), which occurs retrospectively following an irrevocable commitment to a course of action as an attempt to reduce psychological discomfort; the APE scale measures the ongoing, concurrent evaluative conflict experienced during the evaluation and deliberation phase prior to or directly concurrent with judgment formulation.

6. Theoretical Framework

The APE scale is theoretically anchored at the intersection of three foundational paradigms in cognitive psychology and consumer decision making: the Model of Evaluative Space (MES), the Conflict-Aversion Framework of Decision Making, and Attitude Ambivalence Theory.

The Model of Evaluative Space

Classical psychometrics frequently utilized bipolar evaluative scales ranging from “bad” (-3) to “good” (+3), assuming mutual exclusivity of positive and negative affect. However, Cacioppo and Berntson’s (1994) Model of Evaluative Space established that physical and psychological evaluation is fundamentally bivariate. The underlying biological substrates responsible for appetitive (approach) and aversive (avoidance) motivations are functionally distinct and capable of independent activation. When a stimulus or an array of stimuli provides strong inputs to both appetitive and defensive systems simultaneously, high positive activation and high negative activation occur concurrently. This simultaneous neuro-cognitive activation results in a state of high tension termed subjective ambivalence. The APE scale operationalizes the psychological readout of this bi-directional activation within consumer comparative settings.

Attitude Ambivalence Theory and Coping Mechanics

Drawing on the theoretical advancements of Priester and Petty (1996, 2001) and Thompson, Zanna, and Griffin (1995), attitude ambivalence is recognized as an unstable and aversive psychological state. Individuals naturally pursue cognitive consistency; holding competing evaluations toward the same focal target induces cognitive dissonance and emotional distress. Consequently, individuals are motivated to utilize coping mechanisms to reduce or manage this tension. In consumer behavior, Nowlis, Kahn, and Dhar (2002) integrated these theoretical assumptions to hypothesize that consumers routinely utilize neutral options (such as “neither agree nor disagree,” choosing a middle-of-the-road compromise product, or maintaining the status quo) as a primary psychological coping mechanism. When choice architects withdraw or eliminate the neutral response category, the individual can no longer employ a compromise heuristic to buffer their internal conflict. The theoretical framework posits that removing the neutral buffer forces the latent ambivalence to surface into conscious awareness, thereby elevating scores on the APE scale and moderating downstream preference reversals.

Trade-Off Difficulty and Choice Deferral

The scale also builds upon behavioral decision theory models developed by Dhar (1997), Luce (1998), and Tversky and Shafir (1992). When consumers are forced to trade off values across competing attributes that are personally important (e.g., safety vs. price, or environmental ethics vs. aesthetic pleasure), the decision process ceases to be purely computational and becomes deeply emotionally taxing. High-trade-off assortments generate evaluative ambivalence because each alternative possesses non-dominated, compelling positive features coupled with severe negative trade-offs. The APE provides the empirical mechanism needed to verify that trade-off manipulation successfully provoked internal evaluative conflict, thereby serving as an indispensable manipulation check and process mediator in behavioral decision experiments.

7. Validity

The psychometric validity of the APE scale has been demonstrated across multiple empirical investigations involving diverse consumer product categories, experimental assortment structures, and choice conditions.

Construct and Convergent Validity

Construct validity is evidenced by the scale’s robust responsiveness to experimental manipulations designed to induce evaluative conflict. In the validation studies conducted by Nowlis, Kahn, and Dhar (2002), participants exposed to balanced, trade-off-heavy product pairs (e.g., Product A excelling on Attribute 1 but failing on Attribute 2, while Product B exhibited the inverse profile) reported significantly higher APE scores than participants exposed to dominant assortments where one alternative clearly outperformed the other across all dimensions. Convergent validity is evidenced by strong positive correlations with established measures of subjective conflict, including Priester and Petty’s (1996) Subjective Ambivalence Scale (typically displaying correlations of r = .68 to r = .79, p < .001). Furthermore, APE scores correlate positively with behavioral indicators of decision conflict, such as prolonged response latencies during choice tasks (r = .34 to .48) and higher frequencies of choice deferral (seeking no-choice options) when deferral is made available (Dhar, 1997).

Discriminant Validity

Discriminant validity has been demonstrated against related cognitive and motivational constructs. In empirical structural equation modeling (SEM) analyses, the APE scale successfully demonstrates discriminant validity from general product involvement (measured via the Personal Involvement Inventory), cognitive need (Need for Cognition scale), and general mood valence (PANAS). While general negative affect captures a broad emotional state of distress, the APE specifically isolates conflict directed toward the evaluative object set, exhibiting low correlations with baseline negative affect (r < .22). Most critically, research confirms discriminant validity between the APE scale and neutral attitude ratings: respondents scoring near the midpoint (e.g., 4 on a 1-to-7 scale) on a standard bipolar attitude scale frequently exhibit drastically divergent scores on the APE scale, separating genuinely indifferent consumers (low APE scores) from deeply conflicted, ambivalent consumers (high APE scores).

Predictive and Nomological Validity

The nomological validity of the APE scale is demonstrated by its effectiveness as a statistical mediator and moderator in choice experiments. In the primary studies of Nowlis et al. (2002), APE scores mediated the relationship between assortment structure and attitude change following forced choice. Specifically, when forced to choose between products without a neutral option, respondents who scored higher on the APE scale exhibited greater subsequent attitude polarization and lower confidence in their chosen alternative. Conversely, under conditions where compromise or neutral options were available, APE scores were suppressed, reflecting effective conflict mitigation. Replications in sustainability marketing (e.g., evaluating eco-friendly products with premium prices) consistently confirm that the APE scale predicts consumer hesitation, purchase abandonment, and vulnerability to counter-persuasion.

8. Reliability

The APE scale exhibits exceptional reliability metrics across empirical consumer behavior literature, despite comprising only three items. The brief nature of the instrument makes its high internal consistency particularly noteworthy, as traditional reliability metrics like Cronbach’s alpha are mathematically sensitive to scale length.

Internal Consistency Reliability

In the original experimental investigations reported by Nowlis, Kahn, and Dhar (2002), the three-item instrument demonstrated high internal consistency across multiple independent consumer cohorts and product categories:

• In Study 1 (evaluating consumer choice sets across consumer goods categories), the three-item scale demonstrated a Cronbach’s alpha of α = .84.
• In subsequent experimental replications and follow-up studies exploring assortment trade-offs and neutral option removal, the reported Cronbach’s alpha coefficients ranged reliably between α = .81 and α = .87.
• Subsequent independent replications in consumer decision research evaluating durable goods, digital subscription services, and food products have reported Cronbach’s alphas routinely spanning α = .80 to α = .91, with composite reliability (CR) values in structural equation modeling exceeding .85.
• The average inter-item correlation across validation datasets typically ranges from r = .58 to r = .74, demonstrating that the three items share substantial common variance without redundant duplication.

Test-Retest Stability

Because the APE scale is designed primarily as a state measure of immediate evaluative conflict elicited by a specific set of stimuli, test-retest reliability across long time intervals is theoretically expected to vary in accordance with changes in stimulus exposure or choice availability. However, in short-interval test-retest paradigms (e.g., 30 to 60-minute intervals without intervening information or choice execution), the instrument exhibits substantial temporal stability (intra-class correlation coefficients, ICC > .78), indicating that the scale reliably reflects the prevailing evaluative state of the judge during active deliberation.

9. Factor Analysis

Both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) robustly confirm that the APE scale is characterized by a unidimensional factor structure.

Exploratory Factor Analysis (EFA)

When the three items of the APE scale are subjected to unconstrained exploratory factor analysis (Principal Axis Factoring or Maximum Likelihood extraction with Promax or Varimax rotation), a single-factor solution consistently emerges across empirical studies:

• Eigenvalues and Variance Explained: The first extracted factor universally exhibits an initial eigenvalue well in excess of 2.0 (typically spanning 2.20 to 2.45), while the second factor exhibits eigenvalues dramatically below 0.50 (failing the Kaiser criterion of > 1.0). The single factor accounts for approximately 72% to 82% of the total variance across the three items.
• Scree Plot Criterion: Visual inspection of the scree plot uniformly demonstrates an unambiguous elbow immediately following the first latent factor, supporting strict unidimensionality.
• Factor Loadings: All three items exhibit strong, statistically significant factor loadings on the single latent dimension, typically ranging from .80 to .92, with minimal residual variance.

Confirmatory Factor Analysis (CFA)

In structural equation modeling frameworks, a one-factor CFA model has been tested across diverse consumer research samples. Because a standard three-item single-factor model has zero degrees of freedom ($df = 0$, exactly identified / saturated model), researchers evaluate measurement quality through standardized factor loadings, composite reliability, and average variance extracted (AVE), or through multi-group and multi-construct models where the APE factor is embedded within a broader structural system:

• Standardized Factor Loadings (λ): Loadings for Item 1 (“mixed feelings”) typically hover around λ = .83 to .89; Item 2 (“difficulty deciding liking/disliking”) ranges around λ = .78 to .85; and Item 3 (“one-sided vs. mixed reactions”) yields loadings between λ = .84 and .91 (all p < .001).
• Average Variance Extracted (AVE): The AVE consistently exceeds .65 (frequently surpassing .70), comfortably exceeding the recognized convergent validity threshold of .50 established by Fornell and Larcker (1981).
• Model Fit in Multi-Construct Systems: When modeled alongside related latent constructs (e.g., choice confidence, decision satisfaction, post-choice regret), the unidimensional APE measurement model achieves excellent overall goodness-of-fit indices: Comparative Fit Index ($CFI > .97$), Tucker-Lewis Index ($TLI > .96$), Root Mean Square Error of Approximation ($RMSEA < .05$), and Standardized Root Mean Square Residual ($SRMR < .04$).

10. Instrument / Measurement Tool

The Ambivalence of Product Evaluation scale is structured as an efficient, self-administered, multi-item psychometric rating scale designed for computerized laboratory environments, online research panels, or traditional paper-and-pencil surveys. Below is the operational profile of the instrument:

• Instrument Name: Ambivalence of Product Evaluation (APE) Scale
• Primary Conceptual Reference: Nowlis, Stephen M., Barbara E. Kahn, and Ravi Dhar (2002)
• Construct Measured: Subjective / felt evaluative ambivalence during comparative product appraisal
• Test Format: Standardized self-report questionnaire administered in question format
• Number of Items: 3 items
• Response Scale: 7-point semantic differential / rating scale anchored with specific verbal endpoints tailored to each question (ranging from 1 to 7)
• Target Population: Adult consumers, experimental study participants, and organizational decision-makers evaluating multi-attribute product alternatives
• Administration Time: Approximately 30 to 60 seconds
• Scoring Procedure: Responses are scored directly from 1 to 7 for each item. There are no reverse-coded items. The overall APE index is calculated by computing the arithmetic mean across the three items:$$\text{APE Score} = \frac{\text{Item}_1 + \text{Item}_2 + \text{Item}_3}{3}$$
• Score Interpretation:
  • 1.00 to 2.50: Low ambivalence / Univalent evaluation. The respondent experiences high evaluative clarity, perceives minimal conflict, and maintains a one-sided appraisal of the alternatives.
  • 2.51 to 4.50: Moderate ambivalence. The respondent experiences minor trade-off friction or balanced attributes, but not severe cognitive paralysis.
  • 4.51 to 7.00: High ambivalence. The respondent experiences acute evaluative conflict, strong mixed feelings, and substantial difficulty synthesizing positive and negative dimensions into a cohesive judgment.

11. Permissions & Fee and Test Year

The Ambivalence of Product Evaluation (APE) scale was originally published in 2002 in the Journal of Consumer Research. As an academic psychometric instrument developed and published within peer-reviewed scientific literature, the scale items are accessible for non-commercial scholarly research, instructional use, and academic replications without payment of licensing royalties or formal permissions fees, provided that standard academic attribution is accorded to the original authors (Nowlis, Kahn, & Dhar, 2002).

Researchers intending to utilize the scale in commercial market research platforms, proprietary consumer testing batteries, or for-profit analytics software should consult the copyright policies of the Journal of Consumer Research and Oxford University Press (or the original publishing consortium) to ascertain whether commercial licensing clearance is necessary. For standard university-sponsored empirical research, experimental dissertations, and academic studies, no formal application or fee is required.

12. References

The theoretical foundations, validation methodologies, and empirical applications of the APE scale are supported by the following scholarly literature:

• Cacioppo, J. T., & Berntson, G. G. (1994). Relationship between attitudes and evaluative space: A critical review, with emphasis on the separability of positive and negative substrates. Psychological Bulletin, 115(3), 401–423. https://doi.org/10.1037/0033-2909.115.3.401
• Dhar, R. (1997). Consumer preference for a no-choice option. Journal of Consumer Research, 24(2), 215–231. https://doi.org/10.1086/209506
• Festinger, L. (1957). A theory of cognitive dissonance. Stanford University Press.
• 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
• Luce, M. F. (1998). Choosing to avoid: Coping with negatively emotion-laden consumer decisions. Journal of Consumer Research, 24(4), 409–433. https://doi.org/10.1086/209518
• Nowlis, S. M., Kahn, B. E., & Dhar, R. (2002). Coping with ambivalence: The effect of removing a neutral option on consumer attitude and preference judgments. Journal of Consumer Research, 29(3), 319–334. https://doi.org/10.1086/344428
• Priester, J. R., & Petty, R. E. (1996). The gradual threshold model of ambivalence: Relating the positive and negative bases of attitudes to subjective ambivalence. Journal of Personality and Social Psychology, 71(3), 431–449. https://doi.org/10.1037/0022-3514.71.3.431
• Priester, J. R., & Petty, R. E. (2001). Extending the bases of subjective ambivalence: Interpersonal and intrapersonal antecedents of evaluative tension. Journal of Personality and Social Psychology, 80(1), 19–34. https://doi.org/10.1037/0022-3514.80.1.19
• Thompson, M. M., Zanna, M. P., & Griffin, D. W. (1995). Let’s not be indifferent about (attitudinal) ambivalence. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (pp. 361–386). Lawrence Erlbaum Associates.
• Tversky, A., & Shafir, E. (1992). Choice under conflict: The dynamics of deferred decision. Psychological Science, 3(6), 358–361. https://doi.org/10.1111/j.1467-9280.1992.tb00047.x

13. Items of the Scale