Types of Positive Affect Scale

Abstract

The Types of Positive Affect Scale (TPAS) is an 18-item self-report psychometric instrument designed to assess the frequency and qualitative diversity of positive emotional experiences. Developed by Paul Gilbert and colleagues (2008), the instrument addresses a foundational limitation in traditional affective neuroscience and psychometrics: the historical overemphasis on high-arousal, goal-focused positive affect (e.g., excitement, enthusiasm) at the expense of low-arousal, soothing, and contentment-based emotional states. Rooted in evolutionary psychology, attachment theory, and affective neuroscience, the TPAS operationalizes positive affect across three distinct neurobiologically grounded dimensions: Activated Positive Affect (driven by dopamine-mediated incentive and resource-seeking systems), Relaxed Positive Affect (characterized by parasympathetic quiescence and low-arousal calm), and Safe/Content Positive Affect (mediated by oxytocinergic and endogenous opioid pathways associated with social safeness, affiliation, and relational security).

Respondents evaluate the degree to which 18 affective adjectives are characteristic of their general emotional experience using a 5-point Likert scale ranging from 0 (Not characteristic of me) to 4 (Very characteristic of me). Extensive psychometric investigations demonstrate that the TPAS possesses robust internal consistency (Cronbach’s α typically ranging from .73 to .89 across subscales), sound test-retest reliability, and strong factorial, convergent, and discriminant validity. Crucially, research reveals that Safe/Content Positive Affect is inversely associated with self-criticism, depression, anxiety, and stress, often exhibiting stronger negative relationships with psychopathology than activating positive affect. The TPAS serves as an essential measurement tool in Compassion-Focused Therapy (CFT), clinical psychology, behavioral medicine, and positive psychology, illuminating the mechanisms through which social safeness and soothing downregulate human threat-processing systems.

Keywords

Types of Positive Affect Scale, TPAS, Paul Gilbert, Compassion-Focused Therapy, positive affect, safe affect, soothing system, contentment, emotional regulation, evolutionary psychology, affective neuroscience, psychometrics

Authors

The Types of Positive Affect Scale was developed and validated by an interdisciplinary team of clinical psychologists and researchers specializing in evolutionary affect regulation, compassion, and mental health:

• Paul Gilbert, Ph.D., FBPsS, OBE — Mental Health Research Unit, Kingsway Hospital, Derbyshire Mental Health Services NHS Trust, and University of Derby, Derby, United Kingdom. Professor Gilbert is the founder of Compassion-Focused Therapy (CFT) and the Compassionate Mind Foundation.
• Ken McEwan — Mental Health Research Unit, Kingsway Hospital, Derbyshire Mental Health Services NHS Trust, Derby, United Kingdom.
• Rekha Mitra — Mental Health Research Unit, Kingsway Hospital, Derbyshire Mental Health Services NHS Trust, Derby, United Kingdom.
• Linda Franks — Mental Health Research Unit, Kingsway Hospital, Derbyshire Mental Health Services NHS Trust, Derby, United Kingdom.
• Alex Richter — Mental Health Research Unit, Kingsway Hospital, Derbyshire Mental Health Services NHS Trust, Derby, United Kingdom.
• Helen Rockliff — Department of Psychology, University of Bristol, and Mental Health Research Unit, Kingsway Hospital, Derby, United Kingdom.

Purpose

The primary objective of the Types of Positive Affect Scale is to capture the multidimensional nature of positive emotions by differentiating high-arousal reward seeking from low-arousal safeness and contentment. Historically, psychological assessment conceptualized positive affect as a monolithic construct. Widely used inventories, such as the Positive and Negative Affect Schedule (PANAS), heavily prioritize activated, energetic, and goal-attainment affective descriptors (e.g., active, alert, excited, attentive). Although these indicators adequately capture behavioral approach system (BAS) activation, they omit feelings of peace, quiescence, and interpersonal safety.

From an evolutionary perspective, human emotional life is governed not merely by the pursuit of external resources (food, status, mates), but equally by the capacity to experience contentment, interpersonal soothing, and safety when threats are absent and needs are satisfied. Gilbert and colleagues designed the TPAS to address several core clinical and theoretical imperatives:

• Deconstructing Positive Affect Homogeneity: Differentiating affective states that serve divergent evolutionary functions—specifically isolating resource-seeking drive states from affiliative soothing and tranquil rest.
• Identifying Affective Deficits in Psychopathology: Clinical observation indicates that individuals with elevated shame, chronic self-criticism, borderline personality patterns, and major depressive disorder do not necessarily lack the capacity for drive-related excitement; rather, they experience a selective and profound impairment in the capacity to feel safe, reassured, and content.
• Facilitating Compassion-Focused Therapy (CFT) Interventions: CFT posits that psychological health relies on balancing threat, drive, and soothing systems. The TPAS provides a reliable metric to track clinical changes in the soothing/safeness system over the course of compassion-oriented interventions.
• Investigating Mind-Body Psychophysiology: The scale enables researchers to parse the divergent autonomic and neuroendocrine profiles associated with different positive emotions, establishing empirical linkages between safe positive affect, heart rate variability (HRV), vagal tone, and parasympathetic nervous system activation.

Psychological Construct

The TPAS measures three interrelated yet functionally and neurochemically distinct dimensions of positive affect. Rather than treating positive valence as uniform, the scale delineates positive emotions according to evolutionary goals, behavioral actions, and physiological signatures.

1. Activated Positive Affect

Activated Positive Affect reflects energizing, vigorous, high-arousal positive emotions tied to the evolutionary drive system. Described by markers such as Active, Energetic, Excited, Lively, Enthusiastic, and Dynamic, this construct corresponds to the classical Behavioral Activation System (BAS) or Incentive Salience system. Functionally, it motivates organisms to locate, compete for, and secure crucial resources such as territory, sustenance, sexual partners, and social status. Activated positive affect is characterized by forward momentum, curiosity, reward expectancy, and physiological sympathetic arousal. Although vital for adaptation and accomplishment, an over-reliance on activated affect without counterbalancing soothing states can manifest as manic defense, hyper-competitiveness, workaholism, and affective instability.

2. Relaxed Positive Affect

Relaxed Positive Affect captures low-arousal, peaceful, and restorative affective experiences indicated by descriptors such as Relaxed, Quiet, Still, and Restful. Unlike activating affect, relaxed affect arises when an organism is not currently engaged in resource acquisition or threat avoidance. It reflects a physiological state of low motor output, reduced vigilance, and baseline parasympathetic dominance. Relaxed affect allows the organism to conserve metabolic energy, facilitate biological repair, and experience somatic ease in an unthreatened environment.

3. Safe/Content Positive Affect

Safe/Content Positive Affect is the theoretical cornerstone of the TPAS. Operationalized through adjectives such as Safe, Secure, Content, Calm, Reassured, Peaceful, Laid back, and Serene, this dimension captures the emotional state of experiencing interpersonal warmth, relational safeness, belonging, and emotional equanimity. Contentment represents a state of satisfaction with the current state of affairs, accompanied by an absence of wanting, striving, or vigilance.

Safeness is inherently relational, rooted in early attachment experiences with protective and attuned caregivers. While relaxed affect emphasizes low physiological activation and physical rest, safe affect entails a profound psychological appraisal that the self is accepted, valued, unthreatened, and worthy of care. Gilbert et al. demonstrated that it is specifically a deficit in Safe/Content Positive Affect—rather than a deficiency in activated enthusiasm—that characterizes vulnerability to depressive episodes, severe anxiety, and toxic self-criticism.

Theoretical Framework

The TPAS is anchored in Gilbert’s Evolutionary Affect Regulation Systems Model, which synthesizes evolutionary biology, attachment theory (John Bowlby), affective neuroscience (Jaak Panksepp), and neurobiology (Stephen Porges’ Polyvagal Theory).

The Tripartite Model of Affect Regulation

Gilbert posits that human emotion regulation evolved around three fundamental functional priorities, each supported by distinct brain architectures and physiological substrates:

• The Threat and Self-Protection System: Designed to detect, anticipate, and respond to perils, predators, and social rejection. It triggers emotions such as fear, anger, disgust, and shame, engaging the amygdala, the hypothalamic-pituitary-adrenal (HPA) axis, and the sympathetic branch of the autonomic nervous system.
• The Incentive, Resource-Seeking, and Drive System: Geared toward identifying, anticipating, and securing vital environmental and social assets. It produces activation, pleasure, joy, and competitive drive. Mediated substantially by central dopaminergic pathways (notably the mesolimbic and mesocortical reward circuits), it promotes exploration, goal orientation, and triumph.
• The Soothing, Affiliative, and Safeness System: Designed to facilitate social bonding, cooperative alliances, infant-parent attachment, and biological recovery. Triggered by cues of non-threat, affection, prosocial acceptance, and kindness, this system downregulates threat-related arousal and terminates drive-based strivings. Neurochemically, it is sustained by oxytocin, endogenous opioids (endorphins), and myelinated vagal parasympathetic pathways (social engagement system).

Traditional cognitive-behavioral paradigms often assumed that positive affect was universally therapeutic. However, Gilbert’s framework reveals that if a person perceives the social world as hostile, unsafe, or highly competitive, high-arousal positive affect may merely represent conditional relief or performance-contingent reward. When social safeness is impaired, activating emotions fail to soothe or downregulate threat processing. The TPAS provides the empirical architecture to measure the activation of these disparate evolutionary networks, distinguishing the dopaminergic “drive” state from the oxytocinergic-endorphinergic “soothing-safeness” system.

Validity

The psychometric validity of the TPAS has been established across clinical, non-clinical, student, and community cohorts internationally.

Construct and Convergent Validity

Convergent validity is evidenced by patterns of correlation with established affective and personality scales. In the initial validation study by Gilbert et al. (2008), Activated Positive Affect correlated positively with measures of reward responsiveness and behavioral approach, aligning strongly with the PANAS Positive Affect subscale. In contrast, the Safe/Content Positive Affect subscale demonstrated substantial positive associations with social safeness, secure attachment styles, feelings of social warmth, and self-reassurance (as measured by the Forms of Self-Criticising/Attacking and Self-Reassuring Scale, FSCRS).

Discriminant and Predictive Validity

The discriminant validity of the TPAS subscales is particularly pronounced when examining relationships with psychological distress. Gilbert et al. (2008, 2009) demonstrated that Safe/Content Positive Affect displays moderate-to-strong negative correlations with measures of:

• Depression (measured via the DASS-42): r values typically range from −.45 to −.60.
• Inadequate Self and Hated Self (subscales of FSCRS): r values range from −.40 to −.55.
• Anxiety and Stress: r values range from −.35 to −.50.

Conversely, Activated Positive Affect displays significantly weaker inverse correlations with depression (often between −.15 and −.30) and frequently exhibits negligible or non-significant correlations with anxiety and stress. This critical divergence confirms that feeling energetic and active does not protect against threat-based emotional pathology in the same manner as feeling safe, reassured, and content. Subsequent structural equation modeling and regression analyses have demonstrated that Safe/Content Affect uniquely accounts for variance in depressive symptomatology even after controlling for traditional high-arousal positive affect and global negative affect.

Reliability

The TPAS exhibits robust internal consistency and temporal stability across diverse populations.

Internal Consistency

In the seminal publication by Gilbert et al. (2008), which evaluated university students and community adults, the internal consistency coefficients (Cronbach’s α) across the subscales were as follows:

• Activated Positive Affect: α = .83
• Relaxed Positive Affect: α = .83
• Safe/Contentment Positive Affect: α = .73

In follow-up and clinical validation cohorts (e.g., Gilbert et al., 2009; Kelly et al., 2012; Duarte et al., 2015), the reliability estimates remained elevated and stable:

• Activated Positive Affect α coefficients consistently span .81 to .88.
• Safe/Content Positive Affect α coefficients consistently range between .80 and .89.
• Relaxed Positive Affect α coefficients range between .78 and .85.

Composite reliability (ω) in structural equation modeling studies has corroborated these estimates, confirming that the measurement error of individual subscales is low and that the items reliably load on their respective underlying latent factors.

Test-Retest Stability

Temporal stability evaluations conducted across two-to-four-week intervals in non-clinical samples have yielded test-retest correlation coefficients (r) ranging between .71 and .82 across all three subscales, indicating that while the TPAS is sensitive to mood fluctuations and therapeutic intervention, it effectively reflects stable trait-like dispositional affect regulation tendencies over time.

Factor Analysis

The latent dimensionality of the TPAS has been scrutinized using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

In the original scale development (Gilbert et al., 2008), an initial pool of positive affect adjectives was subjected to principal components analysis (PCA) with varimax and oblimin rotations. The scree test, eigenvalues-greater-than-one criterion, and parallel analysis converged on a distinct three-factor solution explaining over 58% of the total variance:

• Factor 1 (Safe/Content Affect): Accounted for the largest proportion of common variance, with salient loadings (> .55) from items reflecting relational security, reassurance, and contentment (Secure, Content, Safe, Reassured, Calm, Peaceful).
• Factor 2 (Activated Affect): Defined by high loadings (> .60) from high-energy descriptors (Active, Energetic, Dynamic, Lively, Enthusiastic, Excited).
• Factor 3 (Relaxed Affect): Loaded predominantly by somatic quiescence items (Relaxed, Still, Quiet, Restful).

Confirmatory Factor Analysis (CFA)

Subsequent psychometric investigations have compared competing latent factor models using Confirmatory Factor Analysis:

• One-Factor Model: Conceptualizing all positive affect items as indicators of a single general positive affect construct. This model consistently yields poor fit indices (χ²/df > 5.0, CFI < .80, TLI < .77, RMSEA > .11).
• Two-Factor Model: Collapsing Safe and Relaxed affect into a single low-arousal dimension versus Activated affect. This model demonstrates acceptable but suboptimal fit.
• Three-Factor Correlated Model: Modeling Activated, Relaxed, and Safe affect as three correlated first-order latent constructs. CFA studies demonstrate superior model fit across cohorts (e.g., χ²/df < 2.50, Comparative Fit Index [CFI] = .94–.97, Tucker-Lewis Index [TLI] = .93–.96, Root Mean Square Error of Approximation [RMSEA] = .048–.062, Standardized Root Mean Square Residual [SRMR] = .042–.055).

Cross-cultural adaptations (e.g., Portuguese, Spanish, Italian, and Japanese translations) have replicated the three-factor structure via multigroup CFA, confirming metric and scalar measurement invariance across gender and age groups.

Instrument / Measurement Tool

• Full Instrument Name: Types of Positive Affect Scale (TPAS)
• Alternative Acronym: TPAS
• Constructs Evaluated:
  • Activated Positive Affect (high-arousal, drive/incentive-seeking emotion)
  • Relaxed Positive Affect (low-arousal, somatic quiescence)
  • Safe/Content Positive Affect (low-arousal, affiliative, soothing, and contentment emotion)
• Item Count: 18 emotion-word items
• Administration Format: Self-report paper-and-pencil or digital computer-administered questionnaire
• Completion Time: Approximately 3 to 5 minutes
• Target Population: Adults and adolescents (ages 16+) in clinical, research, and organizational settings
• Response Format: 5-point Likert scale:
  • 0 = Not characteristic of me
  • 1 = [unlabelled]
  • 2 = Fairly characteristic of me
  • 3 = [unlabelled]
  • 4 = Very characteristic of me
• Subscale Item Allocation:
  • Safe/Content Affect: Secure (Item 1), Calm (Item 2), Laid back (Item 4), Peaceful (Item 6), Content (Item 8), Reassured (Item 11), Safe (Item 13), Serene (Item 14) [Note: In the original 2008 paper, items were grouped across a 3-factor structure; subsequent 2009 refinements grouped 8 safe/content items, 6 activated items, and 4 relaxed items].
  • Activated Positive Affect: Active (Item 3), Energetic (Item 5), Excited (Item 7), Lively (Item 9), Enthusiastic (Item 12), Dynamic (Item 17).
  • Relaxed Positive Affect: Relaxed (Item 10), Quiet (Item 15), Still (Item 16), Restful (Item 18).
• Scoring Procedures:
  • There are no reverse-scored items.
  • Subscale scores are obtained by calculating the sum of the items corresponding to each dimension.
  • Alternatively, mean subscale scores (ranging from 0 to 4) can be computed to facilitate direct comparison across dimensions with unequal item lengths.
  • Higher scores reflect greater frequency and self-identification with that specific affective state.

Permissions & Fee and Test Year

The Types of Positive Affect Scale was published in 2008 by Paul Gilbert, Ken McEwan, Rekha Mitra, Linda Franks, Alex Richter, and Helen Rockliff. The scale is copyrighted by Paul Gilbert and the Mental Health Research Unit (2008).

Licensing and Accessibility: The TPAS is considered an open-access psychometric instrument for non-commercial academic research, pedagogical purposes, and clinical diagnostic monitoring. Clinicians and researchers may use and reproduce the scale free of financial charge, provided proper academic attribution is maintained and the copyright notice remains intact. Commercial distribution, translation rights for published commercial software, or integration into fee-charging digital platforms requires explicit formal authorization from the author via The Compassionate Mind Foundation.

References

• Gilbert, P., McEwan, K., Mitra, R., Franks, L., Richter, A., & Rockliff, H. (2008). Feeling safe and content: A specific affect regulation system? Relationship to depression, anxiety, stress, and self-criticism. The Journal of Positive Psychology, 3(3), 182–191. https://doi.org/10.1080/17439760801999461
• Gilbert, P., McEwan, K., Catarino, F., Baião, R., & Palmeira, L. (2014). Fears of negative emotion in relation to fears of happiness and compassion. Clinical Psychology & Psychotherapy, 21(5), 442–452. https://doi.org/10.1002/cpp.1848
• Gilbert, P., McEwan, K., Bellew, R., Mills, A., & Gale, C. (2009). The dark side of competition: How competitive behaviour and striving to avoid inferiority are associated with narcissism, social anxiety, depression and self-criticism. Psychology and Psychotherapy: Theory, Research and Practice, 82(2), 123–136. https://doi.org/10.1348/147608308X379806
• Kelly, A. C., Zuroff, D. C., Leybman, M. J., & Gilbert, P. (2012). Social safeness, received social support, and psychological well-being: A dynamic relationship. Journal of Social and Clinical Psychology, 31(8), 813–845. https://doi.org/10.1521/jscp.2012.31.8.813
• Panksepp, J. (1998). Affective Neuroscience: The Foundations of Human and Animal Emotions. Oxford University Press.
• Porges, S. W. (2007). The polyvagal perspective. Biological Psychology, 74(2), 116–143. https://doi.org/10.1016/j.biopsycho.2006.06.009
• Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. https://doi.org/10.1037/0022-3514.54.6.1063

Items of the Scale