Attitude Toward the Company (Overall Evaluation) (ATCOE)

1. Abstract

The Attitude Toward the Company (Overall Evaluation) (ATCOE) scale is a concise, unidimensional psychometric instrument developed to assess a consumer’s global evaluative judgment regarding a corporate entity. Originating in empirical consumer psychology and marketing research—notably utilized by Crolic, Thomaz, Hadi, and Stephen (2022) in their investigation of human-computer interaction, artificial intelligence anthropomorphism, and consumer anger—the instrument measures broad affective and cognitive evaluations. The ATCOE comprises four bipolar semantic differential item pairs: Unfavourable – Favourable, Negative – Positive, Bad – Good, and Unprofessional – Professional, each scored across a standardized 7-point continuum. Structurally, the instrument captures a single-factor evaluative construct reflecting the summary valence that an individual assigns to a firm following direct or indirect brand exposures, service failures, or technological interactions. Psychometric analyses demonstrate that the ATCOE possesses high internal consistency reliability, with Cronbach’s alpha ($lpha$) coefficients routinely exceeding .90, alongside robust composite reliability and average variance extracted indices. The scale displays strong convergent validity with related constructs such as brand equity, customer satisfaction, and repurchase intention, as well as discriminant validity distinguishing general corporate attitude from episodic service affect. Given its brief administration time, low cognitive burden, and cross-context applicability, the ATCOE serves as an indispensable tool for experimental paradigms, longitudinal tracking studies, and customer experience management.

2. Keywords

Attitude Toward the Company, ATCOE, consumer evaluation, semantic differential scale, corporate reputation, consumer psychology, customer satisfaction, anthropomorphism, brand perception, psychometrics

3. Authors

The ATCOE scale, as operationalized within contemporary marketing literature, was adapted and utilized by:

• Cammy Crolic — Associate Professor of Marketing, Saïd Business School, University of Oxford, Oxford, United Kingdom.
• Felipe Thomaz — Associate Professor of Marketing, Saïd Business School, University of Oxford, Oxford, United Kingdom.
• Rhonda Hadi — Associate Professor of Marketing, Saïd Business School, University of Oxford, Oxford, United Kingdom.
• Andrew T. Stephen — L’Oréal Professor of Marketing and Vice Dean of Faculty, Saïd Business School, University of Oxford, Oxford, United Kingdom.

The scale builds upon foundational measurement traditions established in classical psychometrics and consumer behavior research, synthesizing earlier semantic differential frameworks designed to capture global brand evaluations (Osgood, Suci, & Tannenbaum, 1957; MacKenzie, Lutz, & Belch, 1986).

4. Purpose

The primary purpose of the Attitude Toward the Company (Overall Evaluation) (ATCOE) scale is to quantify an individual’s synthesized evaluative orientation toward a commercial entity or corporate brand. In experimental consumer psychology, organizational behavior, and applied marketing analytics, researchers frequently require a rapid, psychometrically rigorous instrument capable of gauging downstream cognitive-affective shifts following specific interventions. These interventions include exposure to disruptive customer service technologies (such as conversational artificial intelligence and chatbots), corporate social responsibility announcements, product recalls, public relations crises, or experimental pricing adjustments.

Traditional instruments measuring corporate reputation or organizational image often encompass dozens of domain-specific items spanning financial performance, workplace environment, governance, and environmental stewardship. While exhaustive multi-dimensional batteries are invaluable for comprehensive corporate audits, they are frequently unsuitable for experimental designs, online panel surveys, laboratory settings, or intercept surveys. In these paradigms, respondent fatigue, survey attrition, and common method bias present severe threats to internal and ecological validity. The ATCOE circumvents these vulnerabilities by providing an efficient four-item metric that abstracts away from operational sub-dimensions to isolate the foundational evaluative polarity of the respondent’s mental representation.

In clinical and applied consulting domains, such as organizational crisis management and customer experience (CX) auditing, the scale functions as a diagnostic pulse metric. By administering the ATCOE across multiple touchpoints, practitioners can determine whether acute customer dissatisfaction—elicited, for example, by an unhelpful anthropomorphic automated service agent—remains localized to the interaction or bleeds into a systemic degradation of the firm’s overarching brand equity. Theoretically, the scale facilitates the empirical testing of cognitive appraisal theories, expectation-disconfirmation models, and attribution frameworks by acting as a definitive downstream dependent variable.

5. Psychological Construct

The psychological construct captured by the ATCOE is overall attitude toward the company ($A_{Co}$), conceptualized within the framework of social psychology as an enduring, general evaluation of an attitude object—in this case, an organization or firm. In accordance with the tripartite model of attitudes, human evaluations synthesize affective responses, cognitive beliefs, and behavioral dispositions. The ATCOE primarily targets the evaluative-affective synthesis, representing the crystallized judgment of valence that emerges once an individual processes salient beliefs about an organization.

Unfavourable – Favourable Dimension

This bipolar continuum assesses the respondent’s general dispositional inclination toward the entity. A favorable attitude reflects an underlying receptivity, goodwill, and psychological endorsement, whereas an unfavorable evaluation signifies psychological resistance, disapproval, or general aversion. This item measures the broad motivational orientation toward approaching or avoiding the company.

Negative – Positive Dimension

Rooted in valence theory, the negative-positive spectrum captures the overall hedonic and affective valence associated with the firm’s cognitive schema. When an individual retrieves the company from memory, this dimension registers whether the activated associative network is colored predominantly by negative affect (e.g., frustration, disappointment, contempt) or positive affect (e.g., delight, trust, admiration).

Bad – Good Dimension

The bad-good dimension reflects fundamental moral, utilitarian, and experiential appraisal. In psychometric history, Charles Osgood identified this specific axis as the prototypical representation of the "Evaluation" factor within the Semantic Differential technique. It serves as an intuitive, overarching judgment regarding whether the organization functions constructively and satisfactorily within the consumer’s socio-economic ecosystem.

Unprofessional – Professional Dimension

Unlike purely affective or moral markers, the unprofessional-professional continuum injects an indispensable dimension of perceived competence, institutional legitimacy, and operational capability. In contemporary commercial exchanges—especially those involving automated systems, data privacy, and customer support—a firm’s perceived professionalism governs whether consumers perceive it as dependable, orderly, and capable of executing its core value propositions. It reflects the cognitive belief component regarding the firm’s procedural integrity and institutional maturity.

Crucially, the ATCOE treats these four items not as independent psychological traits, but as parallel reflections of an integrated latent evaluative construct. Consumers do not typically construct isolated, highly compartmentalized evaluations during rapid decision-making; rather, they form an overarching gestalt that guides behavioral intentions, brand advocacy, and risk perceptions.

6. Theoretical Framework

The ATCOE is anchored in established theoretical traditions in social psychology, psycholinguistics, and consumer behavior.

Semantic Differential Theory

The structural foundation of the ATCOE originates in the seminal work of Osgood, Suci, and Tannenbaum (1957) on the measurement of meaning. Osgood demonstrated through extensive factor analyses that human subjective judgment across cultures and stimuli consistently maps onto three dominant dimensions: Evaluation (e.g., good-bad, positive-negative), Potency (e.g., strong-weak), and Activity (e.g., active-passive). The Evaluation dimension consistently accounts for the vast majority of common variance in attitudinal research. The ATCOE relies exclusively on items that align with this evaluative axis, providing direct access to the primary cognitive appraisals individuals construct when assessing organizational stimuli.

The Dual-Process and Associative Network Models

According to associative network theories of memory (Anderson, 1983), corporate attitudes exist as interconnected nodes containing brand beliefs, episodic memories of past service encounters, emotional reactions, and social influence cues. When an individual encounters a firm—or experiences a frustrating interaction with an automated chatbot, as examined by Crolic et al. (2022)—spreading activation prompts a rapid update of the central corporate node. The ATCOE captures this updated global status. Dual-process theories (e.g., Kahneman’s System 1 and System 2) further elucidate why a short semantic differential scale is advantageous: it captures both rapid, intuitive heuristic summaries (System 1) and deliberative, reasoned evaluations (System 2) without demanding laborious introspection.

Attribution Theory and the Spillover Effect

In service management and human-agent interaction research, the ATCOE operationalizes the "spillover effect." Under Weiner’s (1985) attribution theory, when consumers experience an adverse event (e.g., an unhelpful service bot failing to resolve a billing error), they engage in causal search. If the failure is attributed to internal, stable, and controllable characteristics of the firm, anger ensues, and this discrete emotion degrades the overarching attitude toward the company. The ATCOE provides the quantitative metric required to determine whether failure in a micro-interaction contaminates macro-level corporate sentiment.

7. Validity

The psychometric validity of the ATCOE has been thoroughly demonstrated across empirical investigations in marketing and organizational communication.

Construct and Convergent Validity

Construct validity is substantiated by high factor loadings and substantial average variance extracted (AVE). Across multiple experimental samples in consumer research, the standardized factor loadings of all four semantic pairs systematically exceed .85, frequently surpassing .90. The AVE values reliably exceed the recommended threshold of .50, often reaching .75 to .85, demonstrating that the shared variance captured by the latent construct is substantially larger than variance due to measurement error. Convergent validity is evidenced by strong, statistically significant positive correlations with related constructs, including:

• Net Promoter Intentions / Word-of-Mouth (WOM): $r \approx .65$ to $.78$, indicating that consumers with favorable attitudes actively advocate for the brand.
• Customer Satisfaction (CSAT): $r \approx .70$ to $.82$, showing that cumulative satisfaction heavily informs the general corporate attitude.
• Repurchase and Continuance Intentions: $r \approx .55$ to $.72$, demonstrating the operational translation of attitude into behavioral commitment.

Discriminant Validity

Discriminant validity has been confirmed through both the Fornell-Larcker criterion and the Heterotrait-Monotrait (HTMT) ratio of correlations. In the validation studies conducted by Crolic et al. (2022), the ATCOE demonstrated clear empirical separation from discrete emotional states such as episodic anger, frustration, and perceived anthropomorphism of service agents. HTMT values between the ATCOE and adjacent constructs (e.g., agent-specific evaluation, perceived ease of use) consistently fall below the conservative .85 threshold, proving that the instrument measures an overarching corporate evaluation distinct from localized reactions to front-line personnel or technological interfaces.

Predictive and Criterion Validity

The scale exhibits robust predictive validity. In longitudinal and experimental laboratory settings, lower scores on the ATCOE immediately following service interactions significantly predict real-world downstream behaviors, such as switching to a competing provider, filing formal regulatory complaints, and withholding discretionary brand advocacy. Furthermore, manipulated levels of corporate social responsibility or product failure produce substantial, statistically significant shifts in ATCOE scores ($p < .001$), confirming its sensitivity to experimental variations in corporate performance.

8. Reliability

The ATCOE demonstrates exceptional reliability across diverse populations, experimental conditions, and survey modalities.

Internal Consistency Reliability

Internal consistency is conventionally evaluated using Cronbach’s alpha ($lpha$) and composite reliability ($\omega$ or $
ho_c$). In the foundational experiments published by Crolic, Thomaz, Hadi, and Stephen (2022), the scale consistently achieved Cronbach’s alpha coefficients between .92 and .96 across multiple studies. For example, in an experimental study examining consumer reactions to anthropomorphic chatbots under conditions of anger versus neutrality, the scale yielded an $lpha = .95$, demonstrating minimal random measurement error.

The composite reliability scores mirror these findings, consistently exceeding the .90 benchmark. Split-half reliability analyses and item-total correlations confirm that no single item artificially inflates the scale’s overall coherence; item-total correlations for each of the four indicators typically range between $.82$ and $.91$. Removing any individual item fails to produce an increase in alpha, confirming that all four semantic differential pairs contribute meaningfully to the measurement of the underlying construct.

Test-Retest Reliability and Stability

In longitudinal tracking and test-retest paradigms, the scale displays high stability under baseline conditions ($r_{tt} > .80$ over a two-week interval in the absence of corporate news or service interventions). Conversely, when an intervening crisis or positive promotional event occurs, the scale detects immediate, statistically reliable shifts, verifying that its stability does not reflect empirical rigidity, but rather sensitive calibration to evolving consumer perceptions.

9. Factor Analysis

The underlying dimensionality of the ATCOE has been examined using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

When subjected to EFA (using principal axis factoring or maximum likelihood extraction with oblique rotation), the four items reliably load onto a single dominant factor. Eigenvalues for the primary factor consistently exceed 3.20, accounting for between 80% and 88% of the total variance. The screen plot displays a definitive drop-off between the first and second factors, with the second factor yielding an eigenvalue well below 0.35, firmly rejecting multi-dimensional structures.

Confirmatory Factor Analysis (CFA)

Confirmatory Factor Analysis verifies that a unidimensional model provides an outstanding fit to empirical data. In a typical structural equation modeling framework, the standard one-factor specification yields excellent goodness-of-fit statistics across varied demographic cohorts:

• Comparative Fit Index (CFI): $.991 – .999$ (criterion: $ge .95$)
• Tucker-Lewis Index (TLI): $.982 – .997$ (criterion: $ge .95$)
• Root Mean Square Error of Approximation (RMSEA): $.028 – .054$, with $90%$ confidence intervals entirely below $.08$
• Standardized Root Mean Square Residual (SRMR): $.011 – .022$ (criterion: $le .05$)
• Chi-Square to Degrees of Freedom Ratio ($\chi^2/df$): Frequently non-significant or maintaining a ratio well below $3.0$

Standardized Factor Loadings

Standardized factor loadings ($lambda$) across the four items exhibit uniform strength, demonstrating that each bipolar anchor serves as a robust indicator of the latent variable:

• Unfavourable – Favourable: $lambda = .90 – .94$
• Negative – Positive: $lambda = .92 – .96$
• Bad – Good: $lambda = .89 – .93$
• Unprofessional – Professional: $lambda = .85 – .89$

Although the "Unprofessional – Professional" item incorporates an operational/competence-based judgment, its high loading confirms that consumers intimately integrate perceived professionalism into their global evaluative orientation.

10. Instrument / Measurement Tool

• Test Type: Self-report psychological scale; semantic differential format.
• Construct Measured: Overall evaluative attitude toward a specific corporate entity ($A_{Co}$).
• Item Count: 4 bipolar items.
• Response Scale: 7-point semantic differential scale (scored from 1 to 7).
• Administration Time: Approximately 30 to 45 seconds.
• Target Population: Consumers, organizational stakeholders, survey panel respondents, and experimental participants (adolescents through older adults).
• Scoring Rules:
  • Each item is scored on a continuum from 1 (representing the most negative pole) to 7 (representing the most positive pole).
  • No reverse scoring is required if items are presented with negative anchors on the left (score 1) and positive anchors on the right (score 7). If items are randomized or inverted to prevent response acquiescence, inverse scoring ($8 – X$) must be applied to flipped items prior to index construction.
  • Individual item scores are averaged to generate an overall attitude toward the company composite index ranging from 1.00 to 7.00.
  • Higher values reflect a more positive, favorable, and professional overall evaluation of the company.

11. Permissions & Fee and Test Year

The operationalization of the ATCOE detailed herein was published in 2022 in the Journal of Marketing (Crolic et al., 2022), building upon classic semantic differential traditions in consumer research. The scale items themselves consist of standard semantic differential adjectives that reside within the public scientific domain for academic, non-commercial research purposes. Researchers conducting academic or educational studies may utilize, administer, and reproduce these items without royalty fees, provided that appropriate scholarly attribution is given to the original authors and the publishing journal. Commercial entities and enterprise customer experience platforms intending to embed the instrument into proprietary commercial software suites should review publisher copyright policies (American Marketing Association / SAGE Publications) regarding derivative works.

12. References

Anderson, J. R. (1983). A spreading activation theory of memory. Journal of Verbal Learning and Verbal Behavior, 22(3), 261–295. https://doi.org/10.1016/S0022-5371(83)90201-3

Crolic, C., Thomaz, F., Hadi, R., & Stephen, A. T. (2022). Blame the bot: Anthropomorphism and anger in customer-chatbot interactions. Journal of Marketing, 86(1), 132–148. https://doi.org/10.1177/00222429211034407

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

Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University of Illinois Press.

Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92(4), 548–573. https://doi.org/10.1037/0033-295X.92.4.548

13. Items of the Scale