Attitude Confidence (ATTCON)

Abstract

The Attitude Confidence (ATTCON) scale is a brief psychometric instrument originally introduced by Richard E. Petty, Pablo Briñol, and Zakary L. Tormala (2002) within the theoretical architecture of the Self-Validation Hypothesis (SVH). Designed to capture metacognitive appraisal, the scale evaluates the subjective degree of conviction, perceived validity, and certainty an individual maintains toward their own cognitive responses or formed attitudes regarding a specific target issue or entity. Comprising four unipolar items, the ATTCON assesses core operationalizations of attitude certainty without confounding evaluative extremity with metacognitive confidence. The response format typically uses a 7-point or 9-point semantic differential or Likert-type anchor spectrum, ranging from low confidence (e.g., “Not at all confident”) to high confidence (e.g., “Extremely confident”). Psychometric evaluations consistently demonstrate that the ATTCON exhibits robust internal consistency (Cronbach’s $\alpha$ frequently exceeding $.85$ and ranging up to $.92$) and a clear unidimensional factor structure verified through confirmatory factor analysis (CFA). The instrument possesses high predictive and convergent validity: it selectively moderates the impact of cognitive responses on subsequent evaluations, such that positive or negative thoughts influence overall attitudes most strongly when confidence in those thoughts is high. The instrument has become a benchmark assessment in social psychology, consumer behavior, political psychology, and health communications, facilitating rigorous empirical investigations into attitude persistence, resistance to counter-persuasion, and attitude-behavior consistency.

Keywords

Attitude Confidence, ATTCON, Metacognition, Self-Validation Hypothesis, Attitude Certainty, Persuasion, Elaboration Likelihood Model, Cognitive Response, Attitude Strength, Psychometrics, Social Cognition

Authors

The foundational development and formal psychometric deployment of the Attitude Confidence (ATTCON) operationalization emerged from the joint scholarship of:

• Richard E. Petty, Ph.D. — Distinguished University Professor of Psychology, Department of Psychology, The Ohio State University, Columbus, OH, USA. (Specialist in social cognition, persuasion, and the Elaboration Likelihood Model).
• Pablo Briñol, Ph.D. — Professor of Social Psychology, Department of Social Psychology and Methodology, Universidad Autónoma de Madrid, Madrid, Spain. (Pioneer in metacognition, self-validation, and bodily feedback mechanisms).
• Zakary L. Tormala, Ph.D. — The Botha-Chan Professor of Marketing, Graduate School of Business, Stanford University, Stanford, CA, USA. (Expert on attitude strength, certainty, resistance to persuasion, and consumer information processing).

Purpose

The Attitude Confidence (ATTCON) measure was engineered to resolve a persistent conceptual and methodological limitation in persuasion research: the inability of primary evaluative metrics to distinguish between the valence of a thought or attitude and the metacognitive endorsement of that thought or attitude. Historically, persuasion theories—including early cognitive response models—presumed that generating favorable thoughts toward a persuasive message directly induced positive attitude change, while counterarguing caused negative attitudes. However, empirical studies demonstrated substantial unaccounted variance in whether generated thoughts translated into durable attitudes and overt actions.

The primary purpose of the ATTCON scale is to capture the secondary, metacognitive judgment of confidence that individuals place in their thoughts or overall attitude. Developed under the Self-Validation Hypothesis (Petty et al., 2002), the scale identifies when individuals view their mental contents as valid, truthful, and representative of objective reality. In research settings, it serves as a critical moderator variable: thoughts that are held with high confidence guide judgment, predict behavioral intention, and resist decay, whereas thoughts held with low confidence are discarded or discounted, attenuating their impact on downstream evaluations.

Beyond basic social psychological experimentation, the ATTCON scale provides extensive clinical, political, and communicative applications:

• Cognitive and Behavioral Interventions: In clinical settings, therapeutic progress frequently hinges not merely on altering the frequency or valence of depressive or anxiogenic automatic thoughts, but on diminishing the patient’s metacognitive confidence in the validity of those maladaptive cognitions. The ATTCON framework enables clinicians and researchers to track changes in cognitive endorsement across treatment regimens.
• Health and Risk Communication: Public health researchers use the tool to evaluate whether intervention messages (e.g., vaccine advocacy, anti-smoking campaigns) generate thoughts that recipients genuinely trust. An intervention that prompts positive cognitive responses will fail if recipients perceive their own pro-health thoughts to be uncertain or ill-founded.
• Consumer Decision-Making: Marketers apply the scale to gauge purchase intention stability. High attitude confidence regarding a brand or product acts as a protective shield against competitor advertising, whereas low confidence indicates volatility and openness to brand-switching.

Psychological Construct

The psychological construct assessed by the ATTCON is metacognitive attitude confidence (frequently studied interchangeably with attitude certainty or thought confidence). Within contemporary social cognitive taxonomy, attitudes are conceptualized as evaluative summaries stored in memory, ranging from positive to negative. However, attitudes also possess structural properties collectively known as attitude strength. Metacognitive confidence is one of the most vital components of attitude strength.

Metacognition refers to “cognitions about cognitions” or thoughts about thoughts. While a primary cognition reflects the immediate evaluation of an object (e.g., “Electric vehicles reduce carbon emissions”), a secondary cognition reflects an evaluative judgment concerning the primary cognition itself (e.g., “I am completely sure that this statement is accurate”). The ATTCON measures this secondary appraisal across several interrelated facets:

• Subjective Certainty: The personal feeling of conviction and clarity regarding one’s attitude or thought. It reflects the subjective absence of doubt or ambivalence. For instance, an individual might generate arguments in favor of a policy, yet experience an internal sensation of hesitation; subjective certainty captures the felt resolution of that internal conflict.
• Perceived Validity and Accuracy: The degree to which an individual believes their cognitive appraisal reflects objective truth and reality rather than subjective bias, fleeting emotion, or misleading information. Individuals with high perceived validity consider their reactions to be robustly grounded in facts.
• Cognitive Trust and Reliability: The extent to which an individual feels they can safely rely on their thoughts as a basis for action, public endorsement, and ongoing commitment. Trust reflects the perceived functional utility of the cognitive response in navigating relevant real-world decisions.

Crucially, the construct of attitude confidence is conceptually distinct from attitude extremity (how far an attitude departs from a neutral midpoint) and attitude importance (the personal significance or psychological investment assigned to an issue). Two individuals may share an identical, highly favorable rating of an issue (+3 on a -3 to +3 semantic differential), yet one person may hold this position with profound confidence (score of 9 on a 9-point ATTCON scale) while the other holds it tentatively (score of 2). The ATTCON isolates this metacognitive property from evaluative direction and intensity.

Theoretical Framework

The ATTCON scale is anchored within the Self-Validation Hypothesis (SVH; Petty, Briñol, & Tormala, 2002), which extended the Elaboration Likelihood Model (ELM) of persuasion developed by Petty and Cacioppo. The ELM posits that persuasion can occur via a central route—characterized by high cognitive elaboration, scrutiny of message arguments, and the generation of cognitive responses—or via a peripheral route dominated by heuristic cues.

Historically, the central route presumed that the primary determinants of persuasion were the number and valence of cognitive responses generated during message reception. If strong arguments generated predominantly favorable thoughts, persuasion followed; if weak arguments generated counterarguments, persuasion was undermined. The Self-Validation Hypothesis expanded this paradigm by introducing a metacognitive tier: individuals not only generate thoughts, but they also evaluate those thoughts along dimensions of confidence, validity, and certainty.

According to the SVH, thought confidence acts as a multiplicative or gating mechanism:

Under this theoretical model, if an individual generates predominantly positive thoughts regarding a persuasive appeal, high attitude confidence amplifies those positive thoughts, resulting in substantial positive persuasion. Conversely, if confidence is low, those positive thoughts are cognitively invalidated, dampening persuasion. Symmetrically, when an individual generates counterarguments toward a flawed message, high confidence strengthens resistance, whereas low confidence undermines the defensive utility of the counterarguments, paradoxical though it may seem, occasionally increasing agreement with the rejected communication.

The framework also integrates insights from the Metacognitive Model (MCM) of attitudes (Petty, Briñol, & DeMarree, 2007), which specifies that attitudes are represented as evaluative associations stored with an explicit validity tag. The ATTCON scale acts as the direct psychometric probe of this validity tag, quantifying whether an associative evaluation is tagged as true, dependable, and actionable versus false, questionable, or dubious.

Validity

The construct, convergent, discriminant, and predictive validity of the ATTCON scale have been examined across dozens of laboratory experiments and field investigations.

Construct and Convergent Validity

Construct validity has been verified by demonstrating that ATTCON scores respond predictably to theoretical antecedents of metacognitive certainty. For instance, Petty, Briñol, and Tormala (2002, Experiment 1) demonstrated that manipulating source credibility after thought generation significantly shifted ATTCON scores without altering the number or valence of thoughts generated. Participants exposed to a high-credibility source exhibited significantly higher ATTCON scores ($M = 6.84, SD = 1.28$) than participants exposed to a low-credibility source ($M = 5.62, SD = 1.54; F(1, 74) = 14.33, p < .001$), confirming that the scale is sensitive to external informational cues that validate mental contents.

Convergent validity is documented via strong positive correlations between the ATTCON and alternative indicators of attitude strength, such as subjective knowledge ($r \approx .45$ to $.60$), repeated expression of the attitude ($r \approx .38$ to $.52$), and perceived social consensus ($r \approx .35$ to $.48$).

Discriminant Validity

Crucially, the ATTCON demonstrates sharp discriminant validity from measures of primary cognitive valence and affective extremity. In factor-analytic models that include semantic differential evaluations (e.g., Good/Bad, Favorable/Unfavorable) alongside ATTCON items, confidence items load on a completely separate latent factor ($r < .20$ with primary valence under balanced experimental conditions). Furthermore, experimental evidence indicates that while thought valence accounts for the direction of persuasion, ATTCON items specifically account for the interaction variance, showing negligible direct effects on attitude valence when cognitive responses are balanced.

Predictive and Criterion-Related Validity

Predictive validity is robust across multiple domains:

• Attitude-Behavior Consistency: Higher scores on the ATTCON significantly predict whether individuals will translate their attitudes into overt behaviors. In health compliance paradigms, individuals with identical positive attitudes toward exercise or vaccination adhere significantly more often if their ATTCON scores fall in the upper quartile ($OR = 2.41, 95% CI [1.65, 3.52]$).
• Temporal Persistence: Longitudinal studies show that attitudes accompanied by high ATTCON scores exhibit minimal decay across multi-week delays compared to attitudes accompanied by low ATTCON scores.
• Resistance to Counter-Attitudinal Appeals: In competitive persuasion experiments (e.g., Tormala & Petty, 2002), participants scoring high on the ATTCON resisted subsequent attacks by rival messages, showing significantly less attitude change ($eta = -.42, p < .001$) relative to low-confidence controls.

Reliability

The ATTCON scale demonstrates consistently high internal consistency across a diverse range of study populations, experimental contexts, and translated languages.

Internal Consistency

Across the foundational studies reported by Petty, Briñol, and Tormala (2002), the four unipolar items comprising the scale yielded Cronbach’s alpha coefficients consistently above $.80$:

• In Experiment 1 (evaluating thoughts toward a new university examination policy), the scale demonstrated an internal consistency of $\alpha = .85$.
• In Experiment 2 (examining thoughts regarding a commercial product under differing ease-of-retrieval conditions), the four-item scale yielded $\alpha = .89$.
• In subsequent programmatic extensions investigating bodily feedback, emotion, and power (e.g., Briñol & Petty, 2003; Briñol, Petty, & Tormala, 2004), internal consistency estimates regularly fell between $\alpha = .86$ and $\alpha = .93$.

McDonald’s omega ($\omega_t$), computed in modern re-analyses of the scale, similarly demonstrates strong composite reliability ($\omega > .88$), establishing that the scale captures its target construct with minimal measurement error.

Test-Retest Stability

Because the ATTCON is commonly deployed in experimental contexts to gauge momentary metacognitive evaluations of newly formed thoughts, baseline test-retest reliability reflects the stability of the underlying attitude. When administered in stable survey designs without persuasive interventions, the scale demonstrates adequate test-retest correlation across a two-week interval ($r_{tt} = .76$ to $.82$), indicating that chronic attitude certainty remains stable over time in the absence of external perturbations.

Factor Analysis

Both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) substantiate the unidimensionality of the ATTCON scale.

Exploratory Factor Analysis (EFA)

Principal Axis Factoring and Maximum Likelihood extractions routinely yield a single primary factor with an eigenvalue far exceeding Kaiser’s criterion of 1.0 (typically ranging from 2.80 to 3.35), accounting for $70%$ to $84%$ of the total variance across the four items. The scree plot shows a steep descent from the first factor to negligible subsequent eigenvalues (second factor eigenvalues consistently remain below $0.40$).

Confirmatory Factor Analysis (CFA)

Single-factor CFA models demonstrate excellent goodness-of-fit indices across diverse samples. Standard structural equation modeling benchmarks confirm that a unidimensional model provides an optimal fit to the empirical data:

• Comparative Fit Index (CFI): $.985$ to $.999$ (indicating exemplary fit above the standard $.95$ threshold).
• Tucker-Lewis Index (TLI): $.978$ to $.997$.
• Root Mean Square Error of Approximation (RMSEA): $.032$ to $.048$ ($90% CI [0.000, 0.072]$), well within the acceptable cutoff of $le .06$.
• Standardized Root Mean Square Residual (SRMR): $.014$ to $.025$.

Standardized factor loadings for all four items are uniformly high and statistically significant ($p < .001$), typically loading between $lambda = .76$ and $lambda = .92$ on the single latent “Attitude Confidence” dimension. Inter-item correlations are consistently homogeneous (ranging from $r = .62$ to $.81$), indicating that the four items reflect a common underlying metacognitive continuum without substantive residual collinearity or multidimensional contamination.

Instrument / Measurement Tool

• Test Name: Attitude Confidence Scale (ATTCON)
• Authors: Richard E. Petty, Pablo Briñol, and Zakary L. Tormala (2002)
• Construct Assessed: Metacognitive certainty, validity, and subjective confidence in thoughts and attitudes
• Format: Paper-and-pencil questionnaire, web-based survey, or laboratory computerized task
• Number of Items: 4 unipolar items
• Response Scale: Typically formatted as a 7-point or 9-point unipolar semantic differential scale (e.g., $1 = \text{“Not at all confident / Not at all certain”}$ to $7$ or $9 = \text{“Extremely confident / Extremely certain”}$)
• Administration Time: Approximately 1 to 2 minutes
• Scoring Procedure:
  • All 4 items are positively keyed (unipolar format, indicating ascending confidence).
  • Individual scores can be calculated as the arithmetic mean (ranging from 1 to 7 or 1 to 9) or sum total (e.g., 4 to 36 on a 9-point scale).
  • Higher scores represent greater metacognitive confidence, perceived validity, and conviction.

Permissions & Fee and Test Year

The Attitude Confidence (ATTCON) scale was published in 2002 by the American Psychological Association (APA) within the Journal of Personality and Social Psychology. As an academic psychometric measure published in peer-reviewed scientific literature, the instrument is widely accessible for educational, clinical, and scientific research purposes without commercial licensing fees under fair-use provisions. Researchers intending to incorporate the scale in scholarly investigations should cite the primary seminal paper (Petty, Briñol, & Tormala, 2002). For proprietary commercial use, corporate product benchmarking, or commercial software integration, permissions should be coordinated in compliance with American Psychological Association guidelines or through the corresponding authors.

References

• Briñol, P., & Petty, R. E. (2003). Overt head movements and persuasion: A self-validation analysis. Journal of Personality and Social Psychology, 84(6), 1123–1139. https://doi.org/10.1037/0022-3514.84.6.1123
• Briñol, P., Petty, R. E., & Tormala, Z. L. (2004). The self-validation of cognitive responses: The role of source credibility. Journal of Experimental Social Psychology, 40(5), 559–581. https://doi.org/10.1016/j.jesp.2003.11.001
• Petty, R. E., Briñol, P., & DeMarree, K. G. (2007). The Meta-Cognitive Model (MCM) of attitudes: Implications for attitude measurement, change, and strength. Social Cognition, 25(5), 657–686. https://doi.org/10.1521/soco.2007.25.5.657
• Petty, R. E., Briñol, P., & Tormala, Z. L. (2002). Thought confidence as a determinant of persuasion: The self-validation hypothesis. Journal of Personality and Social Psychology, 82(5), 722–741. https://doi.org/10.1037/0022-3514.82.5.722
• Petty, R. E., & Cacioppo, J. T. (1986). The Elaboration Likelihood Model of persuasion. In Advances in Experimental Social Psychology (Vol. 19, pp. 123–205). Academic Press. https://doi.org/10.1016/S0065-2601(08)60214-2
• Tormala, Z. L., & Petty, R. E. (2002). What doesn’t kill me makes me stronger: The effects of resisting persuasion on attitude certainty. Journal of Personality and Social Psychology, 83(6), 1298–1313. https://doi.org/10.1037/0022-3514.83.6.1298

Items of the Scale