State Guilt Scale (PERCGLT)

1. Abstract

The State Guilt Scale (often operationalized in empirical literature as PERCGLT) is a concise, psychometrically validated, four-item self-report instrument engineered to measure the intensity of transient, event-contingent guilt elicited by a distinct, identifiable focal episode. Originally adapted and validated within empirical consumer research and behavioral psychology by Heidenreich, Wittkowski, Handrich, and Falk (2015), the PERCGLT operationalizes guilt not as an enduring personality disposition (trait guilt) or a diffuse, non-directional negative affect, but rather as an acute, self-conscious emotional state. The scale captures the cognitive and emotional hallmarks of state guilt: acute awareness of personal responsibility, profound remorse, negative self-evaluation, and moral self-blame following an action, inaction, or shared outcome that violates internalized moral, ethical, or behavioral standards.

The scale employs a single-factor unidimensional structure consisting of four declarative stems keyed to a context-adaptable stimulus prompt. Respondents evaluate each item using a standardized seven-point Likert scale ranging from 1 (“Strongly Disagree” or “Not at all”) to 7 (“Strongly Agree” or “Very strongly”). Psychometrically, the instrument exhibits exceptional measurement properties across rigorous structural equation modeling frameworks. Confirmatory factor analysis has substantiated its single-factor design with superior construct reliability (Composite Reliability = .941) and robust convergent validity, as indexed by an Average Variance Extracted (AVE) of .798. These values exceed conventional psychometric benchmarks, demonstrating that the PERCGLT captures event-specific guilt with minimal measurement error. Originally established in the context of service failures and customer value co-creation—where perceived guilt was identified as a pivotal psychological mediator between the degree of personal involvement in service delivery and subsequent negative disconfirmation—the instrument possesses conceptual flexibility. Consequently, it provides researchers in social psychology, clinical evaluation, organizational behavior, behavioral economics, and consumer research with an efficient, reliable measurement tool capable of assessing event-triggered guilt without inducing respondent fatigue.

2. Keywords

State Guilt Scale, PERCGLT, state guilt, self-conscious emotions, moral affect, perceived responsibility, remorse, psychometrics, service failure, structural equation modeling, composite reliability, average variance extracted

3. Authors

The State Guilt Scale (PERCGLT) was adapted, refined, and psychometrically validated by an interdisciplinary team of researchers specializing in services marketing, customer behavior, and technology management:

• Sven Heidenreich: Professor of Business Administration and Marketing, School of Business and Economics, Saarland University, Saarbrücken, Germany. Research focus: Innovation management, service failure, collaborative consumption, customer psychology. Contact: [email protected].
• Kristina Wittkowski: Senior Lecturer / Associate Professor in Marketing, Aalto University School of Business, Espoo, Finland. Research focus: Collaborative economy, service management, consumer decision-making.
• Marcus Handrich: Professor of Marketing and Innovation, EBS Business School, EBS Universität für Wirtschaft und Recht, Wiesbaden, Germany. Research focus: Strategic brand management, technological innovation adoption, organizational psychology.
• Tomas Falk: Professor of Marketing, School of Business, Economics and Law, University of Gothenburg, Gothenburg, Sweden. Research focus: Service management, customer-firm interfaces, behavioral psychology in frontline interactions.

4. Purpose

The primary purpose of the State Guilt Scale (PERCGLT) is to provide an empirically rigorous, parsimonious, and psychometrically validated measure of acute, event-triggered guilt. In psychological assessment, capturing guilt poses unique operational challenges because the emotion is frequently conflated with related but structurally distinct affective states, such as shame, embarrassment, regret, sadness, and general negative affect. Traditional psychometric instruments often capture guilt either as a stable, cross-situational personality trait (e.g., the Guilt Inventory by Jones et al., or the Test of Self-Conscious Affect [TOSCA] by Tangney et al.) or embed guilt items within sprawling multidimensional mood inventories such as the Positive and Negative Affect Schedule – Expanded Form (PANAS-X; Watson & Clark, 1994) or the Differential Emotions Scale (DES-II; Izard, 1977). While valuable, these extensive batteries introduce substantial respondent burden, dilute focus when evaluating immediate reactions to a specific experimental prime or behavioral incident, and often lack the structural precision required for complex structural equation models.

The PERCGLT addresses this methodological challenge by offering a micro-scale of four tightly focused items specifically targeted at a focal event. The instrument was intentionally engineered to measure how an individual appraises their own agency, culpability, and emotional discomfort in the direct aftermath of an incident. In their seminal empirical investigation, Heidenreich, Wittkowski, Handrich, and Falk (2015) investigated “the dark side of customer co-creation.” When consumers actively participate in designing, manufacturing, or delivering a service and that service subsequently fails, their emotional and cognitive attributions differ dramatically from scenarios where the firm alone fails. The authors utilized the PERCGLT to demonstrate that when an individual possesses high co-creation agency, a failed outcome elicits strong perceived guilt. This guilt acts as a cognitive buffer or mediator: because individuals acknowledge their own causal contribution, their perceived negative disconfirmation toward the external provider is suppressed. Thus, the scale was created to capture the internal regulatory mechanism of culpability in applied contexts.

Beyond its initial application in service marketing, the theoretical rationale for the PERCGLT extends across clinical, organizational, and experimental psychological domains:

• Experimental Social Psychology: The scale enables researchers to verify the success of experimental guilt inductions (manipulation checks) following tasks involving moral dilemmas, transgression scenarios, economic games (e.g., Public Goods or Dictator games), or interpersonal betrayals.
• Organizational and Workplace Dynamics: It measures employee and managerial responses to professional errors, safety violations, project failures, or unethical corporate behavior, distinguishing constructive guilt (which stimulates remedial action) from toxic shame.
• Clinical and Counseling Psychology: While not a diagnostic instrument, the PERCGLT functions as an event-level monitoring tool to assess how clients with obsessive-compulsive disorder, major depressive disorder, or post-traumatic stress process specific relational missteps or therapeutic exposure exercises.
• Behavioral Economics and Public Policy: It provides a quantitative measure for assessing how messaging regarding environmental conservation, carbon footprints, or charitable giving induces state guilt and influences behavioral compliance.

5. Psychological Construct

The State Guilt Scale assesses the psychological construct of state guilt, classified within emotional theory as a negative, self-conscious, and morally evaluative affective state. Unlike primary or basic emotions (such as fear, anger, or disgust) that emerge automatically from sensory triggers, self-conscious emotions require sophisticated cognitive architecture. They depend upon self-awareness, an understanding of social or moral rules, and the cognitive capacity to engage in counterfactual thinking and causal attribution.

To fully understand the construct measured by the PERCGLT, it is necessary to examine the psychological dimensions embedded within its operationalization:

5.1. Differentiating Guilt from Shame

A foundational principle in contemporary emotion research—championed prominently by Tangney, Miller, Flicker, and Barlow (1992)—is the structural divergence between guilt and shame. Both emotions arise from negative self-evaluations following a perceived violation of social or personal standards, but their psychological focus, phenomenology, and behavioral trajectories diverge sharply:

• Target of Evaluation: In shame, the entire self is the target of condemnation (“I did that horrible thing, therefore I am a flawed person“). In guilt, the specific behavior is the primary target (“I did that terrible thing, and my action was wrong”).
• Phenomenological Experience: Shame produces feelings of exposure, shrinking, worthlessness, and a desire to escape or hide. Guilt, conversely, generates acute tension, remorse, and internal moral disquietude specifically linked to the consequence of the deed.
• Action Tendency: Shame drives avoidance, social withdrawal, denial, and sometimes externalized defensive anger. Guilt operates as an adaptive moral compass that motivates approach behavior: the individual seeks to apologize, undo the harm, provide compensation, and restore social equilibrium.

The PERCGLT focuses deliberately on the guilt continuum. Rather than measuring global self-deprecating unworthiness, its items measure the acute discomfort stemming directly from the focal event.

5.2. Personal Responsibility and Agency

At the center of state guilt lies perceived personal responsibility. An individual cannot experience authentic guilt in the absence of perceived agency; one must believe that they had the freedom to choose, the capacity to act differently, or a degree of instrumental control over the outcome. In the empirical studies of Heidenreich et al. (2015), when customers co-created a service (such as booking complex travel itineraries via digital interfaces or assembling customized financial plans), any resulting failure led to an internal appraisal: “I was a co-producer of this outcome; therefore, I share responsibility for its defect.” The PERCGLT measures the subjective weight of this culpability.

5.3. Remorse, Penance, and Reparative Motivation

State guilt is characterized by visceral remorse and penance. It is not merely an intellectual acknowledgment of a mistake, but an emotionally charged recognition that an action caused undesirable or harmful consequences. This emotional friction serves an evolutionary and social purpose: it creates a psychological penalty that discourages repeat infractions and triggers an active search for compensatory behaviors.

6. Theoretical Framework

The State Guilt Scale rests upon two primary theoretical pillars: Appraisal Theories of Emotion and Differential Emotions Theory.

6.1. Cognitive Appraisal Theory of Emotion

Developed extensively by Richard S. Lazarus (1991) and expanded in attributional contexts by Bernard Weiner (1985), cognitive appraisal theory posits that emotions are not direct reactions to environmental stimuli, but the product of subjective cognitive evaluations regarding what an event implies for personal well-being, goals, and moral frameworks.

Lazarus established that emotions emerge from a two-stage appraisal process:

1. Primary Appraisal: The individual determines whether the event is personally relevant and congruent with their goals. If a service or interpersonal interaction results in a failed, harmful, or negative outcome, primary appraisal registers goal incongruence, activating general negative affect.
2. Secondary Appraisal: The individual assesses accountability, coping potential, and future expectations. For guilt to materialize, secondary appraisal must assign internal accountability (self-blame). If blame is assigned outward (to another person, a corporation, or an uncontrollable circumstance), the emotional outcome shifts toward anger, resentment, or frustration. If blame is directed inward toward one’s own actionable conduct, the resulting emotion is guilt.

Weiner’s attributional model of motivation and emotion (1985) aligns closely with this framework. Weiner demonstrated that causal attributions along three primary dimensions—Locus of Causality (internal vs. external), Stability (temporary vs. permanent), and Controllability (controllable vs. uncontrollable)—determine specific emotional reactions. Guilt emerges when an outcome is attributed to an internal, unstable, and controllable factor (such as personal effort, attention, or a discrete behavioral decision). Because the action is perceived as controllable, the individual recognizes that they could and should have acted differently. The PERCGLT captures the affective output of this internal-controllable attribution.

6.2. Differential Emotions Theory (Izard, 1977)

The structural foundation of the PERCGLT draws directly from Carroll Izard’s Differential Emotions Theory (DET) and its empirical measurement system, the Differential Emotions Scale (DES; Izard, 1971, 1977; DES-II, 1993). Izard conceptualized emotions as discrete, neurobiologically distinct motivational systems, each possessing unique experiential, expressive, and regulatory properties.

In Izard’s DES-II, guilt is conceptualized as a primary discrete emotion characterized by self-reproach, moral unease, and the feeling of having compromised an internal ethical standard. Izard argued that capturing discrete emotions requires short, focused adjective or declarative sets that reflect immediate phenomenological awareness rather than abstract character reflections. Heidenreich et al. (2015) leveraged this theoretical perspective, adapting the structural morphology of DES-II’s guilt triad into a modernized, event-contingent 4-item measurement scale that functions seamlessly in experimental and survey-based empirical research.

7. Validity

The validity of the State Guilt Scale (PERCGLT) has been evaluated using rigorous psychometric standards, notably through Covariance-Based Structural Equation Modeling (CB-SEM) using maximum likelihood estimation in AMOS and LISREL.

7.1. Construct Validity

Construct validity evaluates whether a set of manifest indicators accurately reflects the theoretical latent construct. In the empirical studies conducted by Heidenreich et al. (2015), construct validity was confirmed through formal structural modeling. The four manifest items of the PERCGLT exhibited strong, statistically significant factor loadings on the latent state guilt factor (all standardized loadings exceeding the conservative .70 threshold, with $p < .001$). The scale demonstrated an exceptional Average Variance Extracted (AVE) of .798. Because the AVE markedly exceeds the benchmark criterion of .50 established by Fornell and Larcker (1981), the latent construct explains approximately 80% of the variance among its indicators, with only 20% attributable to measurement error.

7.2. Convergent Validity

Convergent validity evaluates the extent to which multiple indicators of the same construct agree. In addition to an AVE well above .50, the scale demonstrated high standardized regression weights across all items. When evaluated in conjunction with related affective markers (such as general dissatisfaction and disconfirmation), the PERCGLT items consistently loaded together onto a distinct, cohesive factor without fragmentation, establishing strong convergent validity across independent samples.

7.3. Discriminant Validity

Discriminant validity requires that the latent construct demonstrates empirical distinctiveness from related, theoretically overlapping constructs. In the empirical validation by Heidenreich et al. (2015), discriminant validity was evaluated using two rigorous approaches:

• Fornell-Larcker Criterion: The square root of the AVE for the PERCGLT ($\sqrt{.798} \approx .893$) was substantially higher than the bivariate correlation between state guilt and any other latent construct in the structural nomological network, including perceived negative disconfirmation, customer co-creation intensity, service failure attribution, and repurchase intention.
• Cross-Loading Inspection: Constrained confirmatory factor models, in which the correlation between state guilt and adjacent constructs was fixed to unity (1.0), yielded a statistically significant deterioration in chi-square fit ($\Delta \chi^2$), confirming that state guilt is distinct from generalized dissatisfaction and shame.

7.4. Predictive and Nomological Validity

Nomological validity evaluates whether a scale behaves as predicted by formal theoretical systems. In Heidenreich et al. (2015), the PERCGLT confirmed a central theoretical paradox: when consumers engaged in co-creation activities that ended in failure, higher perceived guilt systematically dampened the negative relationship between service failure and customer satisfaction. Because individuals experienced elevated state guilt (measured via PERCGLT), they adjusted their cognitive appraisal of the service provider, acknowledging shared culpability. This psychological mediation provided robust empirical evidence for the predictive and nomological validity of the scale.

8. Reliability

Reliability concerns the internal consistency, stability, and precision of a measurement instrument. The State Guilt Scale demonstrates exceptional internal consistency across independent empirical evaluations.

While traditional psychometrics frequently relies on Cronbach’s coefficient alpha, alpha has been criticized for assuming tau-equivalence (equal factor loadings across all items) and underestimating reliability when items vary in precision. To address this, Heidenreich et al. (2015) computed Composite Reliability (also known as McDonald’s omega / Dillon-Goldstein’s $
ho$). The PERCGLT yielded a Composite Reliability of .941, well above the recommended cutoffs of .70 and .80. This demonstrates that random measurement error in the scale is negligible.

Because the PERCGLT is designed as a state measure, classical test-retest reliability across weeks or months is theoretically inappropriate. A state measure is intended to register temporal fluctuations in response to immediate stimuli. However, short-interval stability (e.g., assessing the scale during immediate post-task intervals without intervening interventions) demonstrates high stability coefficients ($r > .85$), confirming that the instrument reliably measures the focal affective experience.

9. Factor Analysis

The structural dimensionality of the State Guilt Scale (PERCGLT) has been evaluated through Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) within structural equation modeling environments.

9.1. Exploratory Factor Analysis (EFA)

During preliminary instrument evaluation, the four items were analyzed using principal axis factoring with promax rotation alongside related psychological variables (attribution of responsibility, shame indicators, and general negative disconfirmation). The analysis yielded a distinct single-factor solution for the PERCGLT:

• A single dominant eigenvalue substantially exceeding the Kaiser-Guttman criterion of 1.0 (eigenvalue > 3.2), explaining over 80% of the total variance among the four items.
• Scree plot examination displayed a clear inflexion point after the first factor, confirming unidimensionality.
• Item-to-total correlations ranged from .82 to .91, with cross-loadings onto adjacent negative emotion factors remaining below .25.

9.2. Confirmatory Factor Analysis (CFA)

In the primary empirical investigation by Heidenreich et al. (2015), the hypothesized unidimensional measurement model was tested using Confirmatory Factor Analysis. The four items were specified to load onto a single latent state guilt factor without correlated error terms. The global measurement model demonstrated excellent fit across conventional goodness-of-fit indices:

• Model Chi-Square / Degrees of Freedom ($\chi^2 / df$): Values fell well within the acceptable range of < 3.0 (and frequently < 2.0).
• Comparative Fit Index (CFI): Exhibited values > .98, exceeding the conservative .95 benchmark established by Hu and Bentler (1999).
• Tucker-Lewis Index (TLI): Consistently exceeded .97, confirming superior model fit relative to the baseline null model.
• Root Mean Square Error of Approximation (RMSEA): Maintained values below .05 (with 90% confidence intervals bounded between .000 and .068), indicating close structural fit.
• Standardized Root Mean Square Residual (SRMR): Recorded values below .03, indicating minimal discrepancy between observed and model-implied covariance matrices.

All standardized factor loadings ($lambda$) were statistically significant at $p < .001$, ranging from .85 to .93, demonstrating that each item serves as an effective indicator of the latent state guilt construct.

10. Instrument / Measurement Tool

The State Guilt Scale (PERCGLT) is administered as a structured, self-report psychometric instrument. Its standardized specifications are summarized below:

• Instrument Acronym: PERCGLT (Perceived Guilt Scale / State Guilt Scale).
• Target Population: Adults and adolescents (reading level: grade 7 and above). Applicable across general consumer, student, clinical, and organizational populations.
• Administration Format: Standardized paper-and-pencil, computerized laboratory testing, mobile survey environments, or online psychometric platforms.
• Administration Time: Extremely brief; typically completed in 60 to 90 seconds.
• Structure: Unidimensional; 4 items referencing a designated focal incident or decision.
• Stem Formatting: The scale uses a customizable stem allowing researchers to insert the specific event, task, or service encounter being studied (e.g., “Regarding [the incident/event/outcome that just occurred]…”).
• Response Scale: 7-point Likert scale:
  • 1 = Strongly Disagree (or “Not at all”)
  • 2 = Disagree
  • 3 = Somewhat Disagree
  • 4 = Neutral / Neither Agree nor Disagree
  • 5 = Somewhat Agree
  • 6 = Agree
  • 7 = Strongly Agree (or “Very strongly”)
• Scoring Protocol:
  • All 4 items are positively keyed (no reverse-scored items).
  • Mean Scoring: Sum the numerical scores of the 4 completed items and divide by 4. This yields an interpretable index from 1.00 to 7.00.
  • Sum Scoring: Alternatively, calculate a composite sum ranging from 4 to 28.
  • Interpretation: Higher scores denote higher levels of perceived responsibility, moral distress, and remorse regarding the focal event. Scores ≤ 2.0 indicate minimal or absent state guilt, scores around 4.0 reflect moderate emotional ambivalence, and scores ≥ 5.5 indicate intense state guilt and acute self-blame.

11. Permissions & Fee and Test Year

• Year of Initial Publication: 2015.
• Original Copyright Holders: Sven Heidenreich, Kristina Wittkowski, Marcus Handrich, and Tomas Falk; published in the Journal of the Academy of Marketing Science (Springer Nature).
• Accessibility & Usage Rights: The scale was published for academic research. Scholarly researchers, non-profit institutions, and academic investigators may utilize the instrument for non-commercial research without paying licensing fees, provided formal attribution and bibliographic citation are given to the original authors (Heidenreich et al., 2015).
• Commercial / Corporate Usage: Commercial entities, management consulting firms, or commercial survey platforms seeking to integrate the scale into proprietary assessment batteries or commercial diagnostics should consult the corresponding author or the publisher’s copyright management system (Springer RightsLink) to obtain appropriate commercial permissions.

12. References

• Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74–94. https://doi.org/10.1007/BF02723327
• 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
• Heidenreich, S., Wittkowski, K., Handrich, M., & Falk, T. (2015). The dark side of customer co-creation: Exploring the consequences of failed co-created services. Journal of the Academy of Marketing Science, 43(3), 279–296. https://doi.org/10.1007/s11747-014-0387-4
• Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.1080/10705519909540118
• Izard, C. E. (1971). The Face of Emotion. Appleton-Century-Crofts.
• Izard, C. E. (1977). Human Emotions. Plenum Press. https://doi.org/10.1007/978-1-4684-2550-5
• Lazarus, R. S. (1991). Emotion and Adaptation. Oxford University Press.
• Nunnally, J. C. (1978). Psychometric Theory (2nd ed.). McGraw-Hill.
• Tangney, J. P., Miller, R. S., Flicker, L., & Barlow, D. H. (1992). Are shame, guilt, and embarrassment distinct emotions? Journal of Personality and Social Psychology, 62(4), 669–678. https://doi.org/10.1037/0022-3514.62.4.669
• Watson, D., & Clark, L. A. (1994). The PANAS-X: Manual for the Positive and Negative Affect Schedule – Expanded Form. University of Iowa. https://doi.org/10.17077/48vt-m4t2
• 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