Compassion for Others Scales

Abstract

The Compassion for Others Scales, commonly operationalized in psychometric literature as the Compassion Scale (CS), is a multidimensional self-report assessment developed by Elizabeth A. Pommier and Kristin D. Neff to measure an individual's dispositional capacity to extend compassion toward other human beings experiencing emotional, psychological, or physical distress. Rooted theoretically in Neff’s tripartite operational definition of compassion—originally formulated to measure self-compassion—the instrument conceptualizes compassion for others through three structurally interdependent, bipolar facets consisting of six distinct subscales: Kindness versus Indifference, Common Humanity versus Separation, and Mindfulness versus Disengagement. Comprising 24 items rated along an authentic 5-point Likert response scale ranging from 1 (Almost Never) to 5 (Almost Always), the scale balances positive, compassionate approaches with negative, uncompassionate reactions to others' suffering.

Empirical investigations across normative collegiate, community, and clinical cohorts demonstrate robust psychometric properties for the Compassion Scale. Confirmatory factor analyses (CFA) support a hierarchically organized, bi-factor or six-factor intercorrelated structure that converges cleanly on a general compassion latent factor. The total scale exhibits high internal consistency, with Cronbach’s alpha coefficients typically ranging from α = .87 to .91, and subscale reliabilities demonstrating adequate to strong homogeneity (α = .68 to .85). The instrument demonstrates convergent validity via robust positive associations with empathic concern, perspective taking, altruism, and prosocial orientation, while demonstrating divergent validity via negative correlations with personal distress, interpersonal hostility, callousness, and social cynicism. By systematically disentangling compassionate action and mindful awareness from emotional enmeshment, burnout, and detached indifference, the Compassion Scale serves as a cornerstone psychometric instrument in social psychology, clinical research, contemplative science, and healthcare professional training.

Keywords

Compassion Scale, Pommier Compassion for Others, prosocial behavior, empathy, common humanity, altruism, mindfulness, psychometrics, relational compassion, interpersonal kindness, emotional detachment, dispositional compassion.

Authors

The Compassion Scale was originally designed, developed, and empirically validated by Elizabeth A. Pommier, Ph.D., in collaboration with Kristin D. Neff, Ph.D., at the Department of Educational Psychology, University of Texas at Austin.

• Elizabeth A. Pommier, Ph.D.
  Department of Educational Psychology, The University of Texas at Austin, Austin, TX, USA.
  Email contact: [email protected]
• Kristin D. Neff, Ph.D.
  Associate Professor, Department of Educational Psychology, The University of Texas at Austin, Austin, TX, USA.
  Email contact: [email protected]
  Scholarly portal: Self-Compassion Research Lab

Purpose

The primary purpose of the Compassion for Others Scales is to provide an empirically rigorous, theory-driven psychometric instrument capable of capturing the nuanced multidimensionality of interpersonal compassion. For decades, empirical investigations into prosocial emotions suffered from conceptual ambiguities, frequently conflating compassion with related yet distinct constructs such as empathy, sympathy, pity, and generalized altruism. While empathy denotes the vicarious affective experiencing of another’s emotional state, and sympathy often implies feeling sorrow or concern without necessarily recognizing common human fallibility, compassion specifically entails: (a) an accurate cognitive and emotional awareness of another's suffering; (b) an affective response characterized by warmth, care, and kindness; (c) a universal recognition of shared human vulnerability; and (d) an active motivational drive to alleviate or ameliorate that suffering.

Prior to the establishment of the Compassion Scale by Pommier (2011), existing instruments measuring prosocial responding—such as the Interpersonal Reactivity Index (IRI; Davis, 1983)—placed heavy emphasis on affective contagion and personal distress. Such frameworks often failed to delineate between functional, sustainable compassion and dysfunctional empathic over-identification, which frequently culminates in compassion fatigue, secondary traumatic stress, and emotional burnout among caregiving professionals. Pommier and Neff sought to bridge this psychometric gap by creating a symmetrical counterpoint to Neff’s widely validated Self-Compassion Scale (SCS; Neff, 2003). The Compassion Scale operationalizes how an individual cognitively, emotionally, and behaviorally responds when witnessing the pain, failure, and suffering of other people.

In applied clinical contexts, social psychology experiments, and public health settings, the Compassion Scale serves diverse functions:

• Clinical and Psychotherapeutic Assessment: It quantifies baseline levels of interpersonal warmth, detachment, and perspective-taking among individuals undergoing psychotherapy, particularly relational therapies, Compassion-Focused Therapy (CFT), and mindfulness-based interventions.
• Caregiver and Healthcare Professional Training: It provides a reliable benchmark to measure the protective and prosocial benefits of compassion training (e.g., Compassion Cultivation Training [CCT], Cognitively-Based Compassion Training [CBCT]) among physicians, nurses, and clinical social workers, distinguishing true compassion from burnout-inducing over-identification.
• Organizational and Educational Research: It evaluates organizational climates, prosocial leadership behaviors, workplace incivility, and peer support networks in corporate, academic, and community ecosystems.

Psychological Construct

The psychological construct measured by the Compassion Scale is operationalized through a tripartite architecture comprising three core dimensions, each represented by a dialectical pair of positive (compassionate) and negative (uncompassionate) orientations. This yields six distinct subscales:

1. Kindness versus Indifference

Kindness represents an active emotional and behavioral posture of care, benevolence, patience, and warmth directed toward individuals experiencing hardship. Rather than remaining indifferent or treating others' predicaments with cold neutrality, high scorers proactively offer comfort, express genuine concern, and demonstrate a dispositional desire to support others in distress. An illustrative item reflecting Kindness includes: “If I see someone going through a difficult time, I try to be caring toward that person.”

Indifference denotes an emotional detachment, coldness, neglect, or active apathy toward the struggles and suffering of others. Individuals scoring high on Indifference disengage emotionally, disregard the problems of others, and often adopt a dismissive posture when confronted with someone else's distress. An illustrative item representing Indifference is: “Sometimes when people talk about their problems, I feel like I don't care.”

2. Common Humanity versus Separation

Common Humanity captures the cognitive and philosophical understanding that suffering, imperfection, failure, and vulnerability are universal facets of the human condition, rather than anomalies that isolate specific individuals. Individuals characterized by high Common Humanity recognize that all people share inherent vulnerabilities and emotional distress, mitigating feelings of superiority, condescension, or alienating pity. A representative item includes: “Everyone feels down sometimes, it is part of being human.”

Separation reflects an alienating, othering stance wherein the individual perceives the suffering or flaws of others as signs of fundamental weakness, divergence, or defective character. This dimension creates an emotional and psychological chasm between the self and the suffering person, leading to disconnection and inability to relate to others' distress. A representative item reflecting Separation is: “When I see someone feeling down, I feel like I can't relate to them.”

3. Mindfulness versus Disengagement

Mindfulness involves an open, clear, receptive, and balanced awareness of another person’s distress. It requires paying careful, non-judgmental attention to verbal and non-verbal cues of distress without turning away or becoming defensively reactive. High mindfulness enables an individual to accurately witness pain without distorting the situation or losing equilibrium. An illustrative item is: “I notice when people are upset, even if they don't say anything.”

Disengagement refers to cognitive avoidance, tuning out, deliberate distraction, or emotional numbing when exposed to the pain and adversity of others. Instead of maintaining mindful presence, individuals exhibiting high Disengagement mentally dissociate, ignore communicative bids for support, or physically avoid those who are suffering to protect themselves from emotional discomfort. A representative item includes: “I often tune out when people tell me about their troubles.”

Theoretical Framework

The foundational framework of the Compassion Scale originates within Buddhist contemplative psychology, translated into Western empirical psychological science through the theoretical synthesis of Kristin Neff (2003) and elaborated by Paul Gilbert (2005, 2010), C. Daniel Batson (1991), and Richard Davidson. In classical Buddhist ontology (specifically within Theravada and Mahayana traditions), compassion (karuṇā) is defined as the trembling of the heart in response to suffering and the active aspiration to free beings from that suffering.

Neff translated this philosophical paradigm into empirical personality and clinical psychology by postulating that compassion—whether directed inward toward oneself or outward toward others—is governed by three dynamically interacting psychological components:

1. Affective-Motivational Component: The presence of emotional benevolence and warmth (Kindness) versus cold disregard or neglect (Indifference).
2. Cognitive-Attributive Component: The contextualization of distress within universal experiential interconnectedness (Common Humanity) versus fragmented alienation (Separation).
3. Attentional-Perceptual Component: Equanimous, clear-sighted awareness of suffering (Mindfulness) versus cognitive suppression, denial, or tuning out (Disengagement).

Importantly, Pommier and Neff emphasized that genuine compassion differs substantially from pure empathic distress. Under Batson’s Empathy-Altruism Hypothesis, empathy can sometimes lead to acute personal distress, which prompts an individual to retreat from the suffering individual to reduce their own vicarious autonomic arousal. In contrast, the theoretical architecture of the Compassion Scale posits that the combination of Mindfulness and Common Humanity serves as a protective cognitive buffer. When an observer views another’s suffering through the lens of common humanity, they do not pathologize the sufferer, and when observing mindfully with emotional balance, they prevent personal over-identification and affective overwhelm.

Furthermore, this framework connects directly with Gilbert’s Social Mentality Theory and Evolutionary Model of Affect Regulation. Compassion operates via the mammalian caregiving and soothing-contentment system (mediated neurobiologically by oxytocin, vasopressin, and the parasympathetic vagal complex), actively down-regulating the threat-defense system (fight/flight/freeze/submit) and inhibiting detached indifference. Thus, the Compassion Scale measures an integrated, adaptive psychological capacity that balances affective warmth with stable cognitive framing and perceptual clarity.

Validity

Extensive psychometric investigations have established robust construct, convergent, discriminant, and predictive validity for the Compassion Scale across multiple independent studies (Pommier, 2011; Neff & Pommier, 2013; Pommier et al., 2020).

Construct and Structural Validity

Construct validity has been supported through structural equation modeling (SEM) and confirmatory factor analysis across diverse populations, including university undergraduates, non-student community adults, and long-term Buddhist practitioners. The scale demonstrates invariance across gender, although women consistently report moderately higher scores on Kindness and Mindfulness subscales compared to men, consistent with global trends in prosocial responding.

Convergent Validity

The Compassion Scale displays significant, theoretically consistent correlations with established instruments measuring interpersonal functioning and prosociality:

• Empathic Concern and Perspective Taking: Scores on the Compassion Scale correlate positively and strongly with the Empathic Concern (r ≈ .50 to .62, p < .001) and Perspective Taking (r ≈ .38 to .48, p < .001) subscales of Davis’s Interpersonal Reactivity Index (IRI).
• Altruism and Prosocial Behaviors: Moderate to high positive associations are routinely observed with self-reported altruistic actions, volunteerism frequency, and validated measures of prosocial value orientation (r = .40 to .55).
• Social Connectedness: Compassion for others significantly predicts subjective feelings of social connectedness (r = .35 to .46) and relational secure attachment styles.
• Agreeableness: Within the Five-Factor Model of personality, the Compassion Scale correlates most robustly with Agreeableness (r = .45 to .58, p < .001).

Discriminant and Incremental Validity

Crucially, the Compassion Scale distinguishes compassion from unhelpful personal distress and pathological enmeshment:

• Personal Distress: Correlations between the total Compassion Scale score and the Personal Distress subscale of the IRI are non-significant or weakly negative (r = -.08 to -.18), demonstrating that high compassion for others does not reflect debilitating affective contagion or neurotic distress.
• Callous-Unemotional Traits and Hostility: Negative dimensions of the scale (Indifference, Separation, Disengagement) correlate strongly and positively with measures of Machiavellianism, interpersonal cynicism, narcissism, and hostility, confirming clear separation from anti-social personality traits.
• Relationship to Self-Compassion: Counterintuitively, early research by Neff and Pommier (2013) demonstrated that in general community and student populations, compassion for others is only weakly to moderately correlated with self-compassion (r ≈ .15 to .26). This divergence demonstrates that being compassionate toward others does not automatically translate into self-compassionate responding, establishing strong discriminant validity between the CS and the SCS.

Reliability

Empirical evaluations of the Compassion Scale across normative and clinical samples confirm high internal consistency reliability and temporal stability:

Internal Consistency

The overall 24-item composite score exhibits high internal consistency across multiple studies:

• Overall Compassion Scale: Cronbach’s alpha (α) typically ranges from .87 to .91 across student and community samples (Pommier, 2011; Pommier et al., 2020), with McDonald’s omega (ω) values similarly exceeding .90.
• Kindness Subscale (4 items): α = .77 to .84
• Indifference Subscale (4 items, reverse-scored): α = .70 to .78
• Common Humanity Subscale (4 items): α = .68 to .76
• Separation Subscale (4 items, reverse-scored): α = .71 to .79
• Mindfulness Subscale (4 items): α = .69 to .75
• Disengagement Subscale (4 items, reverse-scored): α = .72 to .79

Subscale reliability metrics remain adequate despite their brief 4-item lengths. Item-total correlations across all 24 items routinely exceed the accepted psychometric cutoff of r > .30, with most items falling between .42 and .68.

Test-Retest Reliability

Temporal stability assessments conducted over intervening periods ranging from two to six weeks show stable intraclass correlation coefficients (ICC) and Pearson coefficients:

• Total Score Test-Retest Coefficient: r = .81 to .86
• Subscale Test-Retest Coefficients: Range from r = .72 (Common Humanity) to r = .83 (Kindness)

These findings demonstrate that the instrument reliably captures a stable dispositional trait while remaining sensitive to targeted longitudinal therapeutic or contemplative interventions (such as Loving-Kindness Meditation or Mindful Self-Compassion training).

Factor Analysis

The underlying dimensionality of the Compassion Scale has been extensively evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA), as well as modern Exploratory Structural Equation Modeling (ESEM) and bi-factor modeling (Pommier, 2011; Pommier, Neff, & Tóth-Király, 2020).

Exploratory Factor Analysis (EFA)

During initial instrument development by Pommier (2011), principal axis factoring with promax (oblique) rotation was applied to candidate items. The analyses consistently extracted six first-order factors with eigenvalues greater than 1.0, corresponding exactly to the theoretical constructs: Kindness, Indifference, Common Humanity, Separation, Mindfulness, and Disengagement. Standardized factor loadings for target items onto their respective latent factors exceeded λ = .50 (ranging from .52 to .84), with minimal cross-loadings onto non-target factors (λ < .20).

Confirmatory Factor Analysis (CFA) and Model Fit

Subsequent psychometric investigations tested several competing structural representations:

1. Single-factor model: All 24 items loading onto a solitary unidimensional compassion latent variable (χ²/df > 4.8, CFI < .80, RMSEA > .09), yielding unacceptable fit.
2. Two-factor model: Positive compassionate approach items loading onto one factor, and negative uncompassionate items loading onto a separate factor (χ²/df ≈ 3.6, CFI ≈ .84, RMSEA ≈ .078), which demonstrated inadequate structural fit.
3. Six intercorrelated first-order factors model: Kindness, Indifference, Common Humanity, Separation, Mindfulness, and Disengagement modeled as six distinct, correlated latent variables (χ² = 512.44, df = 237, CFI = .941, TLI = .931, RMSEA = .048 [90% CI: .042, .054], SRMR = .045), demonstrating excellent model fit.
4. Hierarchical higher-order and Bi-factor models: A bi-factor model specifying one general compassion factor alongside six specific orthogonal group factors provided superior empirical fit (χ²/df ≈ 1.82, CFI = .962, TLI = .951, RMSEA = .041). The general compassion factor explained the majority of the common variance (explained common variance [ECV] > .60), fully justifying both the calculation of a single overall compassion composite score and the independent interpretation of the six distinct subscale profiles.

Instrument / Measurement Tool

• Instrument Name: Compassion Scale (CS) / Compassion for Others Scales
• Primary Authors: Elizabeth A. Pommier, Ph.D., and Kristin D. Neff, Ph.D. (2011)
• Target Population: Adults and adolescents (ages 14 and older); applicable across normative, occupational, and clinical cohorts.
• Administration Method: Self-administered paper-and-pencil or online digital survey; suitable for individual or large-group testing.
• Total Item Count: 24 items
• Factor Structure / Subscales (4 items each):
  • Kindness: Items 6, 8, 16, 24
  • Indifference: Items 2, 12, 14, 18 (Reverse-Scored)
  • Common Humanity: Items 11, 15, 17, 20
  • Separation: Items 3, 5, 10, 22 (Reverse-Scored)
  • Mindfulness: Items 4, 9, 13, 21
  • Disengagement: Items 1, 7, 19, 23 (Reverse-Scored)
• Response Scale: 5-point Likert-type scale:
  • 1 = Almost Never
  • 2 = (unlabeled intermediate anchor)
  • 3 = Sometimes
  • 4 = (unlabeled intermediate anchor)
  • 5 = Almost Always
• Scoring Instructions:
  1. Reverse Coding: Reverse score the 12 negative items belonging to Indifference (Items 2, 12, 14, 18), Separation (Items 3, 5, 10, 22), and Disengagement (Items 1, 7, 19, 23) such that: 1 → 5, 2 → 4, 3 → 3, 4 → 2, and 5 → 1.
  2. Subscale Scores: Compute the mean score for each subscale by summing the corresponding 4 items (incorporating reverse-coded scores) and dividing by 4.
  3. Total Compassion Score: Compute the overall mean across all 24 items (after reversing negative items). Scores range from 1.00 to 5.00, where higher scores reflect greater levels of dispositional compassion toward others.

Permissions & Fee and Test Year

• Year of Development: 2011 (Doctoral dissertation publication; expanded empirical validation published in 2013 and 2020).
• Copyright and Access Policy: The Compassion Scale is copyrighted by Elizabeth A. Pommier and Kristin D. Neff. It is made freely accessible for academic, research, non-commercial clinical, and non-profit educational purposes.
• Commercial Use: Any commercial implementation, integration into fee-for-service software platforms, corporate assessments, or for-profit consulting programs requires formal written authorization from the copyright holders.
• Instrument Repository: The official scale text, scoring documentation, and academic guidelines are curated at Dr. Kristin Neff’s research clearinghouse: https://self-compassion.org/scales-for-researchers/

References

• Batson, C. D. (1991). The Altruism Question: Toward a Social-Psychological Answer. Lawrence Erlbaum Associates.
• Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44(1), 113–126. https://doi.org/10.1037/0022-3514.44.1.113
• Gilbert, P. (Ed.). (2005). Compassion: Conceptualisations, Research and Use in Psychotherapy. Routledge. https://doi.org/10.4324/9780203003459
• Gilbert, P. (2010). Compassion Focused Therapy: Distinctive Features. Routledge. https://doi.org/10.4324/9780203851197
• Neff, K. D. (2003). The development and validation of a scale to measure self-compassion. Self and Identity, 2(3), 223–250. https://doi.org/10.1080/15298860309027
• Neff, K. D., & Pommier, E. (2013). The relationship between self-compassion and other-focused concern among college undergraduates, community adults, and practicing meditators. Self and Identity, 12(2), 160–176. https://doi.org/10.1080/15298868.2011.649546
• Pommier, E. A. (2011). The Compassion Scale (Doctoral dissertation). Dissertation Abstracts International Section A: Humanities and Social Sciences, 72, 1174.
• Pommier, E., Neff, K. D., & Tóth-Király, I. (2020). The development and psychometric evaluation of the Compassion Scale. Assessment, 27(1), 21–39. https://doi.org/10.1177/1073191119874876

Items of the Scale