Fears of Compassion Scales

Abstract

The Fears of Compassion Scales (FCS) represent a pivotal psychometric battery developed by Professor Paul Gilbert and colleagues (2011) to operationalize, quantify, and clinically assess psychological resistance, ambivalence, and explicit dread associated with prosocial affiliation and compassion. Rooted deeply in Compassion Focused Therapy (CFT) and evolutionary affective neuroscience, the instrument conceptualizes compassion across three distinct directional flows: compassion felt and expressed toward other individuals, compassion received from external social agents, and kindness directed inward toward oneself. The complete assessment comprises three separate self-report subscales totaling 38 items: (a) Fear of Compassion for Others (10 items), (b) Fear of Compassion from Others (13 items), and (c) Fear of Compassion for Self (15 items). Respondents evaluate statements on a 5-point Likert scale ranging from 0 (Don’t agree at all) to 4 (Completely agree). Psychometric investigations consistently indicate strong internal consistency, with initial validation studies reporting Cronbach’s alpha values of .84 to .76 for compassion for others, .85 for compassion from others, and .92 to .86 for compassion for self across student and therapist samples. Confirmatory factor analyses support the tripartite structural integrity of the battery while illuminating significant associations with psychopathology, including major depression, shame, self-criticism, and insecure attachment styles. The FCS serves as an indispensable diagnostic and empirical tool for identifying affective processing blocks that undermine conventional psychological interventions, providing clinicians with actionable data to calibrate therapeutic pacing and address fears of warmth, vulnerability, and affiliative bonding.

Keywords

Fears of Compassion Scales, Compassion Focused Therapy, Self-Compassion, Affect Regulation, Social Safeness, Psychometrics, Attachment Theory, Shame, Self-Criticism, Evolutionary Psychology

Authors

The Fears of Compassion Scales were conceived, developed, and empirically validated by a team of clinical psychologists and researchers specializing in evolutionary psychopathology and compassion science:

• Paul Gilbert, PhD, FBPsS, OBE: Mental Health Research Unit, Kingsway Hospital, Derby, and Centre for Psychological Research, University of Derby, United Kingdom. Professor Gilbert is the founder of Compassion Focused Therapy and has pioneered the study of social mentalities, shame, and self-criticism.
• Kirsten McEwan, PhD: Centre for Health and Clinical Research, College of Health and Social Care, University of Derby, United Kingdom. Dr. McEwan’s research focuses on the physiological correlates of affiliative emotions, heart rate variability, and psychological interventions.
• Marcelo Matos, PhD: CINEICC – Cognitive and Behavioural Centre for Research and Intervention, Faculty of Psychology and Educational Sciences, University of Coimbra, Portugal. Dr. Matos specializes in the longitudinal trajectories of traumatic memories, shame, and compassion-based clinical protocols.
• Alison Rivis, PhD: Department of Psychology, University of Derby, United Kingdom. Dr. Rivis is an academic psychologist with extensive expertise in health psychology, social cognition, and quantitative psychometrics.

Purpose

The primary purpose of the Fears of Compassion Scales is to quantitatively map and delineate the explicit psychological blocks, dread, and defensive avoidance strategies that individuals manifest when encountering warm, prosocial, and affiliative relational interactions. In Western psychopathology and psychotherapy research, an implicit assumption historically prevailed suggesting that prosocial feelings, kindness, and self-directed care are universally perceived as positive, soothing, and desirable by clinical and non-clinical populations alike. However, contemporary clinical practice—particularly within cognitive behavioral therapy (CBT) and dialectical behavior therapy (DBT)—consistently observed that patients with extensive histories of early developmental trauma, relational abuse, emotional neglect, or severe internal shame frequently react to warmth, kindness, and empathetic care with elevated distress, panic, grief, cynicism, or emotional dysregulation.

Gilbert and colleagues devised the FCS to bridge a critical empirical void: the lack of an objective, standardized metric that measures compassion not merely as a trait skill, capacity, or behavioral deficit, but as a potential perceived threat. From an evolutionary perspective, affiliative care and social safeness require the neurological downregulation of the threat-defense system and the activation of the mammalian soothing-contentment system mediated by the oxytocin and endogenous opioid networks. In individuals where early attachment figures served as sources of terror, humiliation, exploitation, or emotional abandonment, the physiological substrates associated with warmth become conditioned cues for interpersonal danger. Consequently, when genuine kindness is encountered, it triggers conditioned defensive mechanisms, social vigilance, and fight-flight-freeze responses.

In clinical settings, the FCS functions as a nuanced diagnostic battery administered during initial psychotherapeutic conceptualization. It enables clinicians to identify why a client may be resistant to traditional cognitive restructuring, behavioral self-care tasks, or therapeutic empathy. In empirical research, the FCS enables investigators to parse the complex mediating and moderating mechanisms linking adverse childhood experiences (ACEs) to adult psychopathology, non-suicidal self-injury, personality disorders, and treatment refractoriness. By elucidating how fear operates differently across interpersonal giving, interpersonal receiving, and intrapersonal soothing, the scale provides granular insight into the architecture of relational vulnerability.

Psychological Construct

The theoretical construct underlying the FCS is multidimensional, organized around what Compassion Focused Therapy terms the “three orientations or flows of compassion.” Compassion is defined fundamentally as a deep sensitivity to the suffering of self and others, accompanied by an authentic motivation and commitment to alleviate and prevent that suffering. Each flow encompasses specific cognitive distortions, evolutionary fears, conditioned affective reactions, and defensive coping behaviors.

1. Fear of Compassion for Others

This dimension assesses an individual’s conscious hesitation, anxiety, and moral or behavioral aversion toward showing kindness, forgiveness, and empathetic concern toward other people. Rather than stemming from a psychopathic absence of empathy, this fear is rooted in functional social defensiveness and instrumental resource preservation. Individuals scoring high on this dimension express chronic worries that displaying compassion makes them look vulnerable, naive, or weak, inviting interpersonal exploitation and boundary violations. Core cognitive appraisals reflect beliefs such as “People will take advantage of me if they see me as too compassionate” or “Being compassionate towards people who have done bad things is letting them off the hook.” This subscale captures expectations that compassion drains emotional resources, fosters maladaptive dependency in others, or fails to uphold necessary societal discipline and retribution.

2. Fear of Compassion from Others

This subscale measures the dread, hypervigilance, and emotional distress evoked when external social agents direct genuine care, warmth, validation, and compassion toward the individual. For individuals raised in insecure or traumatic attachment environments, expressions of external kindness are met with profound suspicion. Affiliative overtures are interpreted through cognitive filters of manipulation (e.g., “I worry that people are only kind and compassionate if they want something from me”), vulnerability to devastating rejection (e.g., “If people are friendly and kind I worry they will find out something bad about me that will change their mind”), or impending abandonment. Crucially, warmth from others can stimulate intense sadness, emotional emptiness, or grief by activating a latent longing for the protective care they were historically denied. Behavioral manifestations include interpersonal distancing, emotional barrier erection, and pre-emptive social withdrawal.

3. Fear of Compassion for Self

The third dimension evaluates an individual’s explicit reluctance, apprehension, or absolute terror regarding cultivating kindness, gentleness, self-acceptance, and soothing toward oneself, especially in moments of failure, distress, or personal inadequacy. High scorers frequently equate self-compassion with moral failure, laziness, self-indulgence, or loss of competitive edge. They hold strong metacognitive beliefs that relentless self-criticism and punitive self-discipline are necessary mechanisms to sustain personal standards and avoid social expulsion (e.g., “I fear that if I become kinder and less self-critical to myself then my standards will drop”; “I fear that if I am more self-compassionate I will become a weak person”). Moreover, self-directed warmth frequently triggers what Gilbert terms the “grief of compassion”: when a person attempts to validate their own suffering, the soothing system downregulates threat vigilance, releasing a profound, unprocessed reservoir of childhood pain, loneliness, and unmourned loss (e.g., “I fear that if I start to feel compassion and warmth for myself, I will feel overcome with a sense of loss/grief”).

Theoretical Framework

The Fears of Compassion Scales are grounded in an integrative theoretical synthesis incorporating evolutionary psychology, attachment theory, affective neuroscience, and social mentality theory. The primary theoretical architecture draws directly from Paul Gilbert’s model of affect regulation systems, which posits that human emotional processing and behavior are organized by three major evolved, neurochemically mediated systems:

• The Threat-Defense System: Centered in the amygdala, hypothalamic-pituitary-adrenal (HPA) axis, and sympathetic nervous system. It detects danger, regulates negative affect (fear, anger, disgust), and initiates survival strategies (fight, flight, freeze, submissive appeasement).
• The Incentive/Drive-Resource Seeking System: Governed by dopaminergic pathways, motivating organisms to seek out resources, territory, status, mates, and success.
• The Soothing-Contentment and Affiliation System: Facilitated by oxytocin and parasympathetic vagal tone. It promotes social safeness, mutual trust, affiliative bonding, restorative rest, and psychological security.

Under optimal developmental circumstances, sensitive and attuned caregiving stimulates and calibrates the child’s soothing-contentment system. The child internalizes an enduring neurobiological blueprint of social safeness: receiving warmth signals safety, which de-escalates threat processing. However, when early caregivers are unpredictable, hostile, neglectful, or abusive, the mammalian caregiving system becomes structurally intertwined with the threat-defense system. According to John Bowlby’s attachment theory (1969, 1973, 1980), infants construct “internal working models” of self and others based on these early transactions.

When an individual with an insecure or disorganized attachment history encounters compassion in adulthood, the cue that should activate calming parasympathetic tone instead acts as a conditioned stimulus for relational trauma. The theoretical mechanism can be conceptualized through Pavlovian associative learning: Affiliation + Trauma/Betrayal = Fear/Distrust. In psychodynamic and evolutionary terms (e.g., Arieti & Bemporad, 1980), self-criticism emerges as a defensive “inward-directed appeasement” strategy aimed at pre-empting social attacks and retaining membership in a critical group. Abandoning self-criticism in favor of self-compassion feels, to the primitive threat architecture of the brain, like dropping one’s shield on an active battlefield.

Validity

Extensive psychometric investigations across international clinical and non-clinical populations have established rigorous evidence for the construct, convergent, discriminant, and predictive validity of the Fears of Compassion Scales.

Construct and Factorial Validity

Construct validity was initially confirmed by Gilbert et al. (2011) via item-to-total correlations and structural assessments confirming that the three directional scales represent distinct yet interrelated domains of affective threat appraisal. Subsequent exploratory and confirmatory factor analyses worldwide have confirmed that a three-factor latent structure provides superior model fit over unidimensional or orthogonal configurations, reflecting the unique psychological realities of expressing, receiving, and self-directing affiliative care.

Convergent and Divergent Validity

The convergent validity of the FCS has been substantiated through robust correlational profiles with well-established psychopathological metrics. The Fear of Compassion for Self and Fear of Compassion from Others scales demonstrate strong, positive correlations with measures of:

• Depression, anxiety, and stress on the Depression Anxiety Stress Scales (DASS-42; rs ranging from .40 to .65)
• Internalized shame on the Other As Shamer Scale (OAS) and Experience of Shame Scale (ESS; rs = .50 to .71)
• Hated-self and inadequate-self self-criticism on the Forms of Self-Criticising/Attacking & Self-Reassuring Scale (FSCRS; rs = .55 to .75)
• Insecure (anxious and avoidant) attachment dimensions on the Experiences in Close Relationships Scale (ECR)

Conversely, discriminant and divergent validity are evidenced by substantial, inverse correlations with measures of social safeness, positive affect, and self-compassion. The FCS subscales correlate negatively with the Social Safeness and Pleasure Scale (SSPS; rs ranging from -.50 to -.70) and with Kristin Neff’s Self-Compassion Scale (SCS; rs typically between -.45 and -.68). Notably, the Fear of Compassion for Others scale exhibits lower to moderate correlations with internalized distress metrics, instead demonstrating unique associations with submissive behavior, burnout, cynicism, and specific personality constructs related to antagonism and Machiavellianism.

Predictive and Clinical Validity

Predictive validity is demonstrated by the instrument’s sensitivity to clinical status and intervention outcomes. Clinical cohorts presenting with borderline personality disorder, major depressive disorder, eating disorders, and complex PTSD score significantly higher across all three FCS scales than non-clinical community samples. Longitudinal and outcome studies show that baseline levels of fear of self-compassion predict poorer response to standard cognitive-behavioral therapies. Moreover, clinical reductions in FCS scores across the course of Compassion Focused Therapy directly mediate significant reductions in shame, self-harming impulses, and global psychological distress.

Reliability

The Fears of Compassion Scales demonstrate robust and replicable internal consistency reliability across diverse cultural, linguistic, and demographic cohorts. In the seminal psychometric validation study conducted by Gilbert, McEwan, Matos, and Rivis (2011), the scales were administered to two primary cohorts: an undergraduate university sample (N = 222) and a professional clinical therapist sample (N = 53).

Internal Consistency Across Validation Samples

• Student Sample (N = 222):
  • Fear of Compassion for Self (15 items): Cronbach’s α = .92
  • Fear of Compassion from Others (13 items): Cronbach’s α = .85
  • Fear of Compassion for Others (10 items): Cronbach’s α = .84
• Therapist Sample (N = 53):
  • Fear of Compassion for Self (15 items): Cronbach’s α = .86
  • Fear of Compassion from Others (13 items): Cronbach’s α = .85
  • Fear of Compassion for Others (10 items): Cronbach’s α = .76

Subsequent psychometric replications across diverse worldwide populations have mirrored these initial findings. Validation studies of the Portuguese, Spanish, French, Italian, Turkish, and Japanese translations report Cronbach’s alphas routinely exceeding .85 for the Self subscale, .80 to .88 for the From Others subscale, and .75 to .85 for the For Others subscale. Composite reliability (ρ_c) and McDonald’s omega (ω) coefficients typically range from .82 to .93, confirming that the scales maintain high measurement precision with minimal random error across demographic groups.

Test-Retest Stability

Temporal stability assessments demonstrate acceptable to strong test-retest reliability across intervals ranging from two to six weeks. Intra-class correlation coefficients (ICCs) and Pearson correlation coefficients consistently fall between r = .78 and r = .89 in stable non-clinical populations, indicating that while fears of compassion are dynamic cognitive-affective schema malleable to targeted therapeutic interventions, they function as stable, trait-like dispositions in the absence of treatment.

Factor Analysis

The structural dimensionality of the Fears of Compassion Scales has been thoroughly scrutinized using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analyses

During the initial item generation phase, 60 candidate items (20 per hypothesized domain) were formulated based on clinical observations, attachment literature, and psychotherapeutic records. After rigorous expert review for face validity, the item pool was reduced to 38 items across three distinct theoretical inventories. Exploratory factor analyses with principal axis factoring and oblique rotation (promax/oblimin, allowing factors to correlate) demonstrated that each of the three scales exhibited a robust primary dominant dimension:

• Fear of Compassion for Others: Item loadings ranged from .42 to .78, accounting for a substantial proportion of unique variance, characterized primarily by themes of vulnerability to exploitation, fear of emotional exhaustion, and retributive beliefs.
• Fear of Compassion from Others: Item loadings ranged from .45 to .79, coalescing around concerns regarding interpersonal deceit, fear of emotional dependency, and shame triggered by proximity.
• Fear of Compassion for Self: Demonstrates the highest degree of structural homogeneity, with factor loadings ranging from .52 to .83, characterized by fears of performance deterioration, fear of emotional collapse/grief, and self-worth deficits.

Confirmatory Factor Analyses & Model Fit

Confirmatory factor analytical studies evaluated alternative structural architectures, including a single-factor global model, a three-factor orthogonal model, and a three-factor oblique model. The oblique three-factor model consistently demonstrates superior goodness-of-fit indices across populations, satisfying standard psychometric criteria:

• Comparative Fit Index (CFI) ≥ .92 to .96
• Tucker-Lewis Index (TLI) ≥ .91 to .95
• Root Mean Square Error of Approximation (RMSEA) ≤ .048 to .065 (with 90% confidence intervals within acceptable bounds)
• Standardized Root Mean Square Residual (SRMR) ≤ .055

Inter-factor correlations in CFA models reveal that Fear of Compassion for Self correlates strongly with Fear of Compassion from Others (typically r = .60 to .75), indicating shared variance related to personal vulnerability and social mistrust. In contrast, Fear of Compassion for Others demonstrates weaker, more moderate correlations with the other two factors (r = .30 to .45), highlighting its distinct evolutionary function as an outward defensive and instrumental boundary-maintenance mechanism.

Instrument / Measurement Tool

• Name of Instrument: Fears of Compassion Scales (FCS)
• Alternative Titles: Gilbert’s Fears of Compassion Scales; Compassion Threat Battery
• Target Construct: Psychological fear, resistance, avoidance, and dread associated with giving compassion to others, receiving compassion from others, and directing compassion toward oneself.
• Format / Administration: Self-administered psychological assessment; available in paper-and-pencil or secure digital/interactive online formats.
• Total Number of Items: 38 items organized into three separate directional scales:
  • Scale 1: Expressing compassion for others (10 items)
  • Scale 2: Responding to the expression of compassion from others (13 items)
  • Scale 3: Expressing kindness and compassion towards yourself (15 items)
• Response Scale: 5-point Likert-type agreement scale:
  • 0 = Don’t agree at all
  • 1 = [Slightly agree]
  • 2 = Somewhat agree
  • 3 = [Mostly agree]
  • 4 = Completely agree
• Scoring Rules & Protocols:
  • Each of the three subscales is scored independently by summing the numeric responses of its constituent items. There is no combined single composite score; the scales represent separate directional flows.
  • Scale 1 (Compassion for Others): Raw score range = 0 to 40.
  • Scale 2 (Compassion from Others): Raw score range = 0 to 52.
  • Scale 3 (Compassion for Self): Raw score range = 0 to 60.
  • Reverse Scoring: None. All items are positively keyed toward the fear/resistance construct.
  • Score Interpretation: Higher scores on each subscale denote greater fear, threat appraisal, suspicion, and defensive resistance toward that specific flow of compassion. Elevated scores on Scales 2 and 3 are clinical markers for heightened vulnerability to depression, shame, relational isolation, and therapeutic impasse.

Permissions & Fee and Test Year

The Fears of Compassion Scales were published in final form in 2011 (following formal submission and dissemination via academic channels in 2010–2011). The instrument was developed under the auspices of the Mental Health Research Unit at Kingsway Hospital, Derby, and the Compassionate Mind Foundation, founded by Professor Paul Gilbert.

The assessment is placed in the public domain for research and non-commercial clinical assessment purposes. It is made freely accessible by the authors to foster scientific inquiry into evolutionary psychopathology, affective regulation, and psychotherapy integration. Researchers, clinical psychologists, and students may download, administer, and reproduce the scales without purchasing commercial licenses or paying royalty fees, provided that appropriate academic attribution is maintained in all publications and presentations. Commercial republishing, digital software redistribution behind paywalls, or integration into proprietary diagnostic platforms requires formal written permission from Professor Paul Gilbert and the Compassionate Mind Foundation (www.compassionatemind.co.uk).

References

The following peer-reviewed publications document the empirical development, theoretical lineage, psychometric validation, and clinical applications of the Fears of Compassion Scales:

• Arieti, S., & Bemporad, J. (1980). Severe and mild depression: The psychotherapeutic approach. Basic Books.
• Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books.
• Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. Basic Books.
• Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression. Basic Books.
• Gilbert, P. (2009). Introducing compassion-focused therapy. Advances in Psychiatric Treatment, 15(3), 199–208. https://doi.org/10.1192/apt.bp.107.005264
• Gilbert, P. (2010). Compassion focused therapy: Distinctive features. Routledge. https://doi.org/10.4324/9780203851197
• Gilbert, P., McEwan, K., Matos, M., & Rivis, A. (2011). Fears of compassion: Development of three self-report measures. Psychology and Psychotherapy: Theory, Research and Practice, 84(3), 239–255. https://doi.org/10.1348/147608310X526511
• Kirby, J. N., Day, J., & Sagar, V. (2019). The ‘fears’ of compassion: A systematic review and meta-analysis of associations with psychopathology, shame, and self-criticism. Clinical Psychology Review, 72, 101755. https://doi.org/10.1016/j.cpr.2019.101755
• Matos, M., Duarte, C., & Pinto-Gouveia, J. (2017). The origins of fears of compassion: Shame and lack of safeness with early caregivers. Clinical Psychology & Psychotherapy, 24(4), 894–904. https://doi.org/10.1002/cpp.2054

13. Items of the Scale (Questionnaire)