Attitude Toward the Brand (Spears & Singh) (Ab)

1. Abstract

The Attitude Toward the Brand (Ab) scale, formulated and rigorously validated by Nancy Spears and Surendra N. Singh (2004), is one of the most widely implemented, psychometrically established measurement instruments in consumer behavior, marketing communications, and advertising research. Developed to resolve systemic psychometric ambiguities, conceptual conflations, and measurement fragmentation within the marketing literature, the scale distills a historical corpus of over fifty disparate evaluative items down to a parsimonious, five-item, unidimensional construct. Operating via a 7-point semantic differential scale (1 to 7) anchored by bipolar evaluative adjectives (Unappealing / Appealing, Bad / Good, Unpleasant / Pleasant, Unfavorable / Favorable, and Unlikable / Likable), the instrument captures an individual’s enduring, global affective and evaluative disposition toward a specified commercial brand or referent object.

Across two rigorous, independent empirical studies involving diverse consumer cohorts, the Spears and Singh Ab scale demonstrated exceptional psychometric properties. Internal consistency reliability benchmarks were notably high, demonstrating Cronbach’s alpha coefficients of .97 in Study 1 and .95 in Study 2, alongside composite reliability indices of .97 and .94, respectively. Convergent validity was established with average variance extracted (AVE) estimates of .86 (Study 1) and .77 (Study 2), substantially exceeding canonical thresholds. Nomological and construct validity were verified using structural equation modeling (SEM), positioning brand attitude as a critical mediating variable within the classic advertising hierarchy: positive affective responses and attitude toward the advertisement ($A_{ad}$) drive brand attitude ($A_b$), which directly governs behavioral purchase intentions ($PI$). Furthermore, stringent tests of discriminant validity, including chi-square difference tests and the Fornell-Larcker criterion, demonstrated empirical distinctiveness between brand evaluations and conative intentions. The instrument remains an essential, highly adaptable diagnostic and theoretical tool across academic and commercial domains.

2. Keywords

Attitude Toward the Brand, Brand Evaluation, Semantic Differential, Spears and Singh, Consumer Psychology, Advertising Effectiveness, Psychometrics, Nomological Validity, Purchase Intentions, Hierarchy of Effects, Structural Equation Modeling, Construct Validity, Attitude Toward the Ad

3. Authors

The scale was developed and psychometrically validated by:

• Nancy Spears, Ph.D. — Professor of Marketing, Department of Marketing, G. Brint Ryan College of Business, University of North Texas, Denton, Texas, United States. Dr. Spears has published extensively on consumer behavior, brand perception, metaphorical advertising, and measurement development in marketing.
• Surendra N. Singh, Ph.D. — Southwestern Bell Chair and Professor of Marketing, School of Business, University of Kansas, Lawrence, Kansas, United States. Dr. Singh is an internationally recognized authority on consumer information processing, advertising psychology, psychometric measurement, and nonconscious brand effects.

Correspondence regarding the foundational validation study was originally addressed to the authors via the Department of Marketing, University of North Texas, and the School of Business, University of Kansas, and published within the peer-reviewed Journal of Current Issues & Research in Advertising.

4. Purpose

The primary purpose of the Spears and Singh (2004) Attitude Toward the Brand scale is to provide a theoretically grounded, psychometrically standardized, and parsimonious assessment tool for capturing an individual’s overall evaluative judgment of a brand. Prior to its formal publication, the empirical measurement of brand attitude within marketing, consumer psychology, and applied communication studies was marred by extensive methodological inconsistency. Researchers routinely devised idiosyncratic, ad hoc scales spanning between 3 and 15 items, frequently confounding purely affective appraisals (e.g., pleasantness, appeal) with cognitive dimensions (e.g., utility, quality, performance) or conflating attitude with behavioral intentions.

Spears and Singh addressed this measurement fragmentation by conducting an exhaustive synthesis of the prior literature, extracting an initial pool of 52 historical $A_b$ items, and subjecting them to rigorous purification procedures. Their goal was to operationalize a clean, unidimensional semantic differential tool that satisfies the classical psychometric criteria established by Churchill (1979) and Gerbing and Anderson (1988). The resulting 5-item scale is engineered to capture a consumer’s summary evaluation without placing excessive cognitive burden on survey respondents, rendering it ideal for laboratory experiments, longitudinal tracking, commercial brand audits, and complex structural equation models where respondent fatigue must be minimized.

In applied and academic contexts, the scale serves multiple vital functions:

• Academic Research in Consumer Psychology: Investigates the cognitive and affective determinants of consumer preference, the mediating mechanisms of marketing stimuli, and the longitudinal impact of brand repositioning campaigns.
• Advertising Testing and Copy Optimization: Quantifies pre- and post-exposure shifts in brand sentiment elicited by specific advertising executions, digital creative variants, influencer marketing, or cross-media campaigns.
• Brand Equity and Commercial Diagnostics: Serves as a key performance indicator (KPI) within brand tracking systems, benchmark testing against competitors, and post-crisis brand damage assessments.
• Nomological Theory Testing: Evaluates theoretical models linking antecedent constructs—such as brand personality, corporate social responsibility, and attitude toward the advertisement ($A_{ad}$)—to downstream behavioral outcomes such as brand loyalty, willingness to pay a price premium, and immediate purchase intention ($PI$).

5. Psychological Construct

The psychological construct assessed by this instrument is Attitude Toward the Brand ($A_b$). Rooted in social psychology and consumer behavior, an attitude is classically defined as a learned predisposition to respond in a consistently favorable or unfavorable manner with respect to a given object (Fishbein & Ajzen, 1975). Within consumer research, Spears and Singh align their conceptualization with the influential framework established by Mitchell and Olson (1981), who defined attitude toward the brand as an individual’s internal, enduring, unidimensional summary evaluation of the brand that energizes, directs, and sustains consumer behavior.

Crucially, Spears and Singh conceptualized and verified $A_b$ as a strictly unidimensional construct. While traditional attitude theories have historically debated the tripartite model (affect, cognition, and conation), empirical consensus in psychometrics emphasizes that summary brand evaluation operates as a unified evaluative continuum. Cognitive beliefs (e.g., “This brand is technologically advanced”) serve as antecedents or informational foundations, whereas conative responses (e.g., “I will buy this brand tomorrow”) represent downstream behavioral consequences. The construct of $A_b$ occupies the central, integrating evaluative node within this chain.

The five bipolar adjective pairs selected by Spears and Singh reflect essential nuances of this global evaluative space while preserving statistical unidimensionality:

• Unappealing / Appealing: Reflects the motivational and valence pull of the brand. An appealing brand stimulates intrinsic interest and curiosity, drawing the consumer toward further cognitive or behavioral interaction.
• Bad / Good: Represents the classical, overarching evaluative dimension described in Osgood, Suci, and Tannenbaum’s (1957) landmark work on semantic measurement. It captures the normative, fundamental valence of the object.
• Unpleasant / Pleasant: Taps into the experiential, affective tone associated with the brand name or logo. It assesses whether cognitive activation of the brand elicits subjective affective comfort or discomfort.
• Unfavorable / Favorable: Captures an overarching, deliberate judgment of orientation. A favorable attitude indicates overall approval, endorsement, and readiness to support the entity.
• Unlikable / Likable: Reflects personal resonance, affinity, and relational warmth. Likability has long been recognized in commercial communication as an emotional catalyst that differentiates functional parity products.

By measuring these five closely aligned facets on a standardized continuous scale, the construct captures the consumer’s aggregate affective-evaluative stance toward the focal brand.

6. Theoretical Framework

The Spears and Singh Attitude Toward the Brand scale is embedded within the Hierarchy of Effects paradigm and information-processing models of consumer decision-making. Historically, marketing communication theory posits that consumers pass through a sequence of psychological stages when exposed to commercial stimuli: cognitive (awareness, knowledge, perception), affective (liking, preference, attitude formation), and conative (conviction, purchase intention, actual behavior).

The instrument directly operationalizes the central mediating component of the Dual Mediation Hypothesis (DMH) and the classic $A_{ad}
ightarrow A_b
ightarrow PI$ model introduced by Mitchell and Olson (1981) and expanded by MacKenzie, Lutz, and Belch (1986). According to this framework, an individual’s affective reaction to an advertisement ($A_{ad}$) exerts a direct, powerful influence on their attitude toward the advertised brand ($A_b$), both directly through classical conditioning/affect transfer and indirectly via brand cognitions. In turn, brand attitude acts as the immediate psychological antecedent that drives behavioral purchase intentions ($PI$).

Furthermore, the scale’s theoretical foundation intersects with the Expectancy-Value Model (Fishbein, 1963) and the Elaboration Likelihood Model (ELM) developed by Petty and Cacioppo (1986). Under both high-elaboration (central route) and low-elaboration (peripheral route) conditions, the ultimate output of consumer persuasion is an updated, consolidated summary attitude toward the focal brand. Whether formed through deliberate evaluation of central product arguments or via peripheral heuristic cues (such as celebrity endorsements or aesthetic background music), the resultant construct is captured by the five bipolar indicators standardized by Spears and Singh.

Spears and Singh specifically designed their research to disentangle $A_b$ from purchase intentions ($PI$). Theoretically, an individual can harbor an exceptionally favorable attitude toward a luxury brand (e.g., Rolls-Royce or Rolex) while maintaining zero intention to purchase due to budgetary constraints, situational irrelevance, or social norms. Conversely, a consumer may display high purchase intentions toward a mundane utility brand (e.g., an inexpensive municipal utility or basic commodity) despite a neutral or lukewarm brand attitude. By proving that $A_b$ and $PI$ load onto distinct, albeit correlated, latent factors, Spears and Singh substantiated the theoretical boundary separating affective-evaluative states from behavioral action readiness.

7. Validity

Spears and Singh (2004) subjected their 5-item scale to extensive, multifaceted validity evaluations across two independent empirical studies involving undergraduate business students exposed to real, full-color print advertisements. The empirical findings demonstrated outstanding validity across multiple dimensions:

Construct and Convergent Validity

Convergent validity assesses the extent to which the five operational indicators share a high proportion of common variance representing the underlying latent brand attitude. In confirmatory factor analysis (CFA), convergent validity is evidenced by high, statistically significant factor loadings and an Average Variance Extracted (AVE) exceeding the conventional threshold of .50 (Fornell & Larcker, 1981).

• Study 1: Standardized factor loadings across the 5 items ranged from .90 to .95 ($p < .001$). The AVE for the $A_b$ construct was calculated at .86, indicating that 86% of the variance in the items was directly accounted for by the underlying brand attitude construct.
• Study 2: In a separate validation sample using different product categories, standardized loadings remained exceptionally high, ranging from .82 to .91 ($p < .001$). The AVE was .77, once again markedly outperforming standard psychometric criteria.

Discriminant Validity

A critical contribution of the Spears and Singh study was establishing rigorous discriminant validity between Attitude Toward the Brand ($A_b$) and Purchase Intentions ($PI$). Spears and Singh implemented three formal methodological tests:

1. Confidence Interval Test: The correlation between $A_b$ and $PI$ was examined to ensure that the 95% confidence interval ($\Phi \pm 2 imes ext{SE}$) did not include 1.0. In both studies, the confidence intervals remained well below unity (Study 1: $r = .64$; Study 2: $r = .66$).
2. Chi-Square Difference Test ($\Delta\chi^2$): CFA models where the correlation between $A_b$ and $PI$ was constrained to unity ($Phi = 1.0$) were compared against unconstrained models ($Phi$ freely estimated). In both studies, the unconstrained models yielded a statistically significant improvement in fit ($\Delta\chi^2 > 100, p < .001$), confirming that the two constructs are non-redundant.
3. Fornell-Larcker Criterion: The AVE for $A_b$ (.86 and .77) substantially exceeded the squared correlation between $A_b$ and $PI$ ($r^2 = .41$ and $.44$), providing robust proof of discriminant validity.

Nomological and Predictive Validity

Nomological validity was demonstrated through full structural equation modeling testing the theoretical sequence: $ ext{Affective Responses}
ightarrow A_{ad}
ightarrow A_b
ightarrow PI$. The path coefficient from$A_b$ to Purchase Intentions was positive, strong, and highly significant ($\gamma = .64$ in Study 1 and $\gamma = .66$ in Study 2, $p < .001$). Furthermore, $A_b$ fully mediated the influence of advertising attitudes on final conative intentions, affirming its theoretical role within consumer decision hierarchies.

8. Reliability

The Spears and Singh Attitude Toward the Brand instrument exhibits remarkable internal consistency reliability across varied testing conditions. In psychometric research, internal consistency evaluates how well the individual items within an instrument measure the same unified latent phenomenon.

Spears and Singh reported the following statistical reliability indices across their investigations:

• Cronbach’s Alpha ($lpha$):
  • Study 1 ($N = 194$): $lpha = \mathbf{.97}$
  • Study 2 ($N = 259$): $lpha = \mathbf{.95}$

  Both values exceed the stringent .90 threshold recommended by Nunnally and Bernstein (1994) for high-stakes, applied, and basic research settings, indicating near-negligible measurement error.

• Composite Reliability (CR):
  • Study 1: $ ext{CR} = mathbf{.97}$
  • Study 2: $ ext{CR} = mathbf{.94}$

  Composite reliability calculations (which do not assume tau-equivalence, unlike Cronbach’s alpha) confirmed that the weighted linear combinations of the five indicators exhibit outstanding internal cohesion.

Subsequent replications across independent laboratories, consumer demographics, and varied media contexts (including digital banner ads, video commercials, and social media promotions) have routinely yielded alpha coefficients ranging between .91 and .98. The instrument possesses virtually no item redundancy while securing robust score stability across different experimental conditions.

9. Factor Analysis

The scale development procedure executed by Spears and Singh involved a rigorous two-phase analytical process utilizing both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA) Phase

The authors began with an extensive pool of 52 historical $A_b$ items drawn from 30 published advertising and consumer studies. In the initial purification stage, items were administered across multiple print advertisements. Principal Axis Factoring with oblique (Promax) rotation revealed that the vast majority of items clustered heavily along a dominant evaluative factor. Items exhibiting cross-loadings greater than .30, communalities below .50, or semantic redundancy were iteratively eliminated. This systematic reduction isolated the 5 cleanest, highest-loading bipolar adjective pairs, which explained over 80% of the total variance with no secondary eigenvalues exceeding unity.

Confirmatory Factor Analysis (CFA) Phase

In Study 1 and Study 2, the five-item structure was formally evaluated via CFA using Maximum Likelihood estimation within LISREL/AMOS software suites. The measurement model specified a single latent variable ($A_b$) with five continuous observed indicators and independent, uncorrelated error terms ($\delta_1 – \delta_5$).

Model fit indices across both validation studies were exemplary, demonstrating exceptional fit between the proposed unidimensional model and the observed covariance matrices:

• Study 1 CFA Results:
  • Chi-Square to Degrees of Freedom Ratio ($\chi^2 / df$): $1.82$ ($\chi^2(5) = 9.10, p > .10$)
  • Goodness-of-Fit Index (GFI): $.98$
  • Comparative Fit Index (CFI): $.99$
  • Non-Normed Fit Index (NNFI / TLI): $.99$
  • Root Mean Square Error of Approximation (RMSEA): $.042$ (with 90% CI $[.000, .085]$)
  • Standardized Factor Loadings ($lambda$): All five items exhibited loadings $ge .90$.
• Study 2 CFA Results:
  • Chi-Square to Degrees of Freedom Ratio ($\chi^2 / df$): $2.14$ ($\chi^2(5) = 10.70, p < .06$)
  • Goodness-of-Fit Index (GFI): $.97$
  • Comparative Fit Index (CFI): $.99$
  • Non-Normed Fit Index (NNFI / TLI): $.98$
  • Root Mean Square Error of Approximation (RMSEA): $.051$ (with 90% CI $[.010, .092]$)
  • Standardized Factor Loadings ($lambda$): Loadings ranged between $.82$ and $.91$.

These empirical metrics firmly confirmed that a single-factor, five-indicator measurement model represents the most parsimonious and psychometrically stable representation of consumer brand attitude.

10. Instrument / Measurement Tool

• Instrument Name: Attitude Toward the Brand Scale (Ab)
• Authors: Nancy Spears and Surendra N. Singh (2004)
• Construct Measured: Consumer’s global, relatively enduring, unidimensional affective and evaluative disposition toward a specified brand
• Format / Administration: Self-administered paper-and-pencil questionnaire, web-based survey, or mobile testing panel
• Number of Items: 5 bipolar adjective pairs
• Response Scale: 7-point semantic differential scale (1 to 7)
• Item Configuration:
  • Item 1: Unappealing (1) vs. Appealing (7)
  • Item 2: Bad (1) vs. Good (7)
  • Item 3: Unpleasant (1) vs. Pleasant (7)
  • Item 4: Unfavorable (1) vs. Favorable (7)
  • Item 5: Unlikable (1) vs. Likable (7)
• Scoring Rules: All five items are scored from 1 (most negative evaluation) to 7 (most positive evaluation). An overall Attitude Toward the Brand index is calculated by computing the arithmetic mean (or sum) across all 5 completed items. Total composite mean scores range from 1.00 to 7.00 (or sum scores from 5 to 35), with higher numeric values indicating a more favorable, positive attitude toward the brand.
• Adaptability: Highly adaptable across referent entities; researchers routinely evaluate corporate brands, consumer packaged goods, service providers, retail stores, advertisements, or digital platforms by substituting the focal brand name within the questionnaire instructional stem while keeping the bipolar adjectives identical.

11. Permissions & Fee and Test Year

The scale was developed, validated, and published by Nancy Spears and Surendra N. Singh in the year 2004 within the Journal of Current Issues & Research in Advertising (Volume 26, Issue 2, pages 53–66), published by Taylor & Francis.

Licensing and Usage Permissions:

• Academic Research: The scale is considered an open-access psychometric instrument for non-commercial academic research, pedagogical use, and scholarly scientific investigations. Researchers may administer the items without explicit written permission, provided that appropriate scholarly attribution and standard academic citations (Spears & Singh, 2004) are provided.
• Commercial and Proprietary Applications: While the 5 bipolar pairs represent general semantic terms, commercial enterprises, market research agencies, or consulting organizations incorporating the scale into proprietary commercial software, diagnostic benchmarking platforms, or published industry reports should review institutional access guidelines and Taylor & Francis copyright policies to maintain intellectual compliance.
• Fees: There are no administrative fees or testing royalties associated with utilizing the scale for basic academic scientific inquiry.

12. References

• Churchill, G. A., Jr. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(1), 64–73. https://doi.org/10.1177/002224377901600110
• Fishbein, M. (1963). An investigation of the relationships between beliefs about an object and the attitude toward that object. Human Relations, 16(3), 233–239. https://doi.org/10.1177/001872676301600302
• 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
• Gerbing, D. W., & Anderson, J. C. (1988). An updated paradigm for scale development incorporating unidimensionality and its assessment. Journal of Marketing Research, 25(2), 186–192. https://doi.org/10.1177/002224378802500207
• MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of Marketing Research, 23(2), 130–143. https://doi.org/10.1177/002224378602300205
• Mitchell, A. A., & Olson, J. C. (1981). Are product attribute beliefs the only mediator of advertising effects on brand attitude? Journal of Marketing Research, 18(3), 318–332. https://doi.org/10.1177/002224378101800306
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
• Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University of Illinois Press.
• Petty, R. E., & Cacioppo, J. T. (1986). The Elaboration Likelihood Model of persuasion. Advances in Experimental Social Psychology, 19, 123–205. https://doi.org/10.1016/S0065-2601(08)60214-2
• Spears, N., & Singh, S. N. (2004). Measuring attitude toward the brand and purchase intentions. Journal of Current Issues & Research in Advertising, 26(2), 53–66. https://doi.org/10.1080/10641734.2004.10505164

13. Items of the Scale (Questionnaire)