The Toronto Empathy Questionnaire

1. Abstract

The Toronto Empathy Questionnaire (TEQ) is a brief, unidimensional self-report measurement instrument developed by R. Nathan Spreng, Margaret C. McKinnon, Raymond A. Mar, and Brian Levine in 2009. Designed to address theoretical ambiguities and psychometric heterogeneity in the assessment of empathy, the instrument operationalizes empathy as a primarily emotional, affective process rooted in emotional contagion, mental state decoding, sympathetic concern, and altruistic motivation. The scale consists of 16 items evaluated on a 5-point Likert response scale ranging from 0 (“Never”) to 4 (“Always”), yielding a total composite score between 0 and 64, with higher scores reflecting greater emotional empathy.

The TEQ was constructed by factor analyzing the common variance among widely used empathy measures, successfully extracting a robust single-factor latent structure. Psychometric evaluations demonstrate strong internal consistency (Cronbach’s alpha typically ranging from .85 to .87 across normative adult cohorts), high test-retest reliability ($r = .81$ over a two-month duration), and pronounced construct validity. The scale correlates positively with established measures of interpersonal reactivity, social decoding ability (such as the Reading the Mind in the Eyes test), and prosocial behavior, while demonstrating theoretical inverse correlations with measures of autism spectrum symptomatology and psychopathic personality traits. Cross-culturally validated in numerous languages, the TEQ serves as an efficient, parsimonious assessment tool for behavioral neuroscience, clinical psychology, psychiatry, and social-personality research.

2. Keywords

Toronto Empathy Questionnaire, affective empathy, emotional contagion, psychometrics, social neuroscience, empathic concern, construct validity, factor analysis, prosocial behavior, Reading the Mind in the Eyes

3. Authors

The Toronto Empathy Questionnaire was conceptualized, developed, and empirically validated by a team of researchers associated with the Department of Psychology at the University of Toronto, the Rotman Research Institute at Baycrest Centre, and affiliated academic medical centers in Ontario, Canada:

• R. Nathan Spreng, Ph.D. — Laboratory of Brain and Cognition, Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University (formerly affiliated with the Department of Psychology, University of Toronto, and Rotman Research Institute, Baycrest Centre). Dr. Spreng’s research centers on the neurocognitive networks underlying social cognition, default mode network dynamics, and autobiographical memory across the adult lifespan.
• Margaret C. McKinnon, Ph.D., C.Psych. — Department of Psychiatry and Behavioural Neurosciences, McMaster University, and St. Joseph’s Healthcare Hamilton. Her scholarship focuses on social cognition, post-traumatic stress, and cognitive dysfunction secondary to psychological and neurological conditions.
• Raymond A. Mar, Ph.D. — Department of Psychology, York University, Toronto, Canada. Dr. Mar investigates the psychological processes underpinning narrative comprehension, imagination, empathy, and social processing through behavioural and neuroimaging methodologies.
• Brian Levine, Ph.D., ABPP — Senior Scientist, Rotman Research Institute, Baycrest Centre; Professor of Psychology and Medicine (Neurology), University of Toronto. Dr. Levine is an internationally recognized expert in episodic memory, executive functions, and structural-functional neuroimaging following focal brain injury.

4. Purpose

The foundational objective of the Toronto Empathy Questionnaire is to provide a brief, methodologically rigorous, and unidimensional self-report inventory that directly taps into empathy conceptualized as an affective and social-emotional phenomenon. Prior to the introduction of the TEQ in 2009, empirical research in social and affective neuroscience faced considerable fragmentation. Existing assessment inventories differed widely in their conceptual definitions of empathy, conflating emotional resonance with general cognitive perspective-taking, sympathy, imaginative fantasy, prosocial actions, and executive communicative competence.

For decades, instruments such as the Interpersonal Reactivity Index (IRI; Davis, 1983) split the domain into four separate subscales: Perspective Taking, Fantasy, Empathic Concern, and Personal Distress. While multifaceted, this division frequently led investigators to evaluate empathy via disparate indices that exhibited divergent or contradictory relationships with clinical endophenotypes. Other omnibus instruments, such as the Empathy Quotient (EQ; Baron-Cohen & Wheelwright, 2004), blended social skills, cognitive inference, emotional sensitivity, and affective reactivity into a single lengthy battery of 40 or 60 items, obscuring the primary emotional substrate that defines fundamental empathy.

Spreng and colleagues sought to distill the broad psychometric core of empathy by evaluating what the predominant, validated empathy instruments possessed in common. By identifying the central underlying latent factor connecting established paradigms, the creators engineered the TEQ to serve several specific research and clinical applications:

• Social and Affective Neuroscience Research: Brain-mapping paradigms, functional magnetic resonance imaging (fMRI), and event-related potential (ERP) investigations require brief yet psychometrically sensitive indices of trait empathy that can be administered in conjunction with demanding experimental cognitive batteries without introducing participant fatigue.
• Clinical Neuropsychological Assessment: The scale assists in characterizing socio-emotional deficits associated with neurodevelopmental conditions such as Autism Spectrum Disorder (ASD), neurodegenerative disorders including frontotemporal dementia (bvFTD), traumatic brain injury (TBI), and psychiatric conditions characterized by empathic dysfunction, such as borderline personality disorder and antisocial personality disorder.
• Behavioral Outcome Prediction: The TEQ serves as an empirical indicator for predicting altruistic behavior, moral reasoning, compassionate communication in healthcare professionals, interpersonal conflict negotiation, and socio-emotional competence across educational, organizational, and forensic environments.

5. Psychological Construct

Empathy has historically been debated as either an intellectualized capacity to infer internal mental states (Theory of Mind or cognitive empathy) or a vicarious emotional response to the perceived affective states of others (affective or emotional empathy). The Toronto Empathy Questionnaire is purposefully rooted in the conceptualization of empathy as an intrinsically emotional process. Rather than capturing detached mental perspective-taking, the TEQ indexes an individual’s propensity to perceive, decode, experience, and sympathetically respond to the affective experiences of other human beings.

Facets of the Affective Construct within the TEQ

Although the TEQ is unidimensional in its overarching factor structure, the constituent items address several neurobiologically and phenomenologically interconnected facets of emotional empathy:

• Emotional Contagion and Shared Affect: Several items (e.g., Item 1: “When someone else is feeling excited, I tend to get excited too”; Item 4: “I remain unaffected when someone close to me is happy” [reverse-scored]) measure basic, primitive affective resonance. This reflects the automatic mirroring and sharing of emotional valences, whether positive (e.g., excitement, joy) or negative, serving as the biological baseline for interpersonal connection.
• Mental State and Mood Decoding: Emotional empathy presupposes the perceptual ability to decode nonverbal affective cues in social contexts. The TEQ incorporates items measuring intuitive, nonverbal emotional sensitivity (e.g., Item 8: “I can tell when others are sad even when they do not say anything”; Item 9: “I find that I am ‘in tune’ with other people’s moods”). This represents the bridge between non-verbal social perception and subsequent emotional induction.
• Sympathetic Concern and Tenderness: Empathy extends beyond primitive mirroring into third-person compassionate concern. Items such as Item 6 (“I have tender, concerned feelings for people less fortunate than me”) and Item 2 (“Other people’s misfortunes do not disturb me a great deal” [reverse-scored]) gauge the emotional capacity to experience sympathy, care, and genuine distress on behalf of someone experiencing hardship.
• Aversion to Interpersonal Cruelty and Injustice: Empathic individuals register affective aversion when witnessing the devaluation or suffering of others. Item 3 (“It upsets me to see someone being treated disrespectfully”) and Item 14 (“When I see someone being treated unfairly, I do not feel very much pity for them” [reverse-scored]) quantify emotional sensitivity to social injustice, mistreatment, and abuse.
• Altruistic and Prosocial Impulse: True affective empathy reliably motivates prosocial intervention. Item 5 (“I enjoy making other people feel better”), Item 13 (“I get a strong urge to help when I see someone who is upset”), and Item 16 (“When I see someone being taken advantage of, I feel kind of protective towards him/her”) evaluate action-oriented manifestations of emotional empathy, where emotional vicariousness transitions directly into helping motives.
• Emotional Suppression versus Openness: Several reverse-scored items measure the avoidance or irritation triggered by another individual’s vulnerability (e.g., Item 7: “When a friend starts to talk about his/her problems, I try to steer the conversation towards something else”; Item 11: “I become irritated when someone cries”). These items identify defensive emotional disengagement or lack of affective tolerance.

6. Theoretical Framework

The theoretical framework guiding the Toronto Empathy Questionnaire draws extensively from developmental psychology, cognitive neuroscience, and evolutionary biology. Most prominently, the TEQ aligns with the Perception-Action Model (PAM) of empathy articulated by Frans de Waal and Stephanie Preston (2002). Under this evolutionary framework, empathy is not a modern cognitive luxury, but an evolutionary adaptation rooted in mammalian parental care and social coordination.

The Perception-Action Mechanism and Shared Neural Representations

The Perception-Action Model posits that perceiving another individual’s behavioral, facial, vocal, or postural state automatically activates the observer’s own neural representations of that state. Neurobiologically, this shared representation is sustained by the mirror neuron system (including the premotor cortex and inferior parietal lobule) linked to the limbic and paralimbic systems via the insular cortex. When an observer watches someone in distress, the anterior insula and anterior midcingulate cortex (the “pain matrix”) activate, evoking an isomorphic, vicarious affective experience.

The TEQ operates under the hypothesis that this shared neural mapping serves as the foundation of interpersonal connection. Without a baseline emotional resonance mechanism, higher-order cognitive processing (e.g., deliberate perspective taking) remains cold, analytical, and devoid of genuine care.

Decety and Jackson’s Multicomponent Model of Empathy

The construction of the TEQ reflects Jean Decety and Philip L. Jackson’s (2004) neurodevelopmental model of human empathy, which identifies three interconnected, dynamically regulated components:

1. Affective Sharing: An automatic, bottom-up process driven by perception-action mechanisms that generate visceral emotional resonance between observer and target.
2. Self-Other Awareness: The ability to recognize that the felt emotional state originates in another person, preventing personal emotional flooding and preserving agency.
3. Emotion Regulation and Cognitive Modulation: Top-down executive processes orchestrated by the prefrontal cortex that modulate empathic concern versus personal distress and channel feelings into prosocial interventions.

By extracting a single common factor across disparate empathy inventories, Spreng et al. demonstrated that items reflecting affective sharing, sympathetic concern, and direct prosocial motivations share substantial latent variance. The TEQ measures the emotional core of this system, demonstrating that individuals who readily process and experience shared feelings are consistently more responsive to the welfare of others.

7. Validity

The Toronto Empathy Questionnaire has undergone extensive empirical validation across multiple independent investigations, cultural settings, and clinical cohorts. The empirical findings firmly establish its construct, convergent, discriminant, and criterion-related validity.

Convergent Validity

During its initial validation by Spreng et al. (2009), the TEQ demonstrated strong positive correlations with leading empathy batteries, supporting the assertion that the questionnaire reliably assesses the shared core of the empathy construct:

• Empathy Quotient (EQ): The TEQ correlated strongly with the EQ ($r = .80, p < .001$), confirming that the 16-item scale captures nearly identical core variance as the 40-item Baron-Cohen scale.
• Interpersonal Reactivity Index (IRI): The TEQ exhibited pronounced positive correlations with the Empathic Concern subscale of the IRI ($r = .68, p < .001$), which gauges other-oriented feelings of sympathy and compassion. It correlated moderately with the Perspective Taking subscale ($r = .35, p < .001$) and the Fantasy subscale ($r = .31, p < .01$), while demonstrating a weak or non-significant correlation with the Personal Distress subscale ($r = -.07, p = .38$), confirming that the TEQ measures constructive empathy rather than self-focused emotional dysregulation.
• Hogan Empathy Scale (HES) and Mehrabian’s Balanced Emotional Empathy Scale (BEES): The TEQ exhibited significant positive associations with the HES ($r = .51$) and the BEES ($r = .72$), corroborating its alignment with established traditions of affective assessment.

Discriminant Validity

Discriminant validity has been consistently evidenced through negative or null associations with psychopathological constructs characterized by empathic blunting or egocentrism:

• Autism Spectrum Traits: Spreng et al. (2009) administered the Autism-Spectrum Quotient (AQ) and found a significant negative correlation with the TEQ ($r = -.31, p < .001$). Individuals exhibiting elevated autistic traits endorsed substantially lower levels of affective empathy on the TEQ.
• Psychopathy and Machiavellianism: Subsequent psychometric evaluations (e.g., Brook & Kosson, 2013) demonstrated robust negative correlations between the TEQ and the affective-interpersonal facets of the Psychopathic Personality Inventory (PPI) and the Dark Triad scales ($r$ values ranging from $-.38$ to $-.52$), illustrating that the TEQ effectively identifies callous-unemotional traits.

Criterion and Predictive Validity

The TEQ demonstrates strong predictive validity with behavioral and objective performance metrics:

• Social Decoding Performance: Spreng et al. (2009) administered the “Reading the Mind in the Eyes” test (RMET; Baron-Cohen et al., 2001), an objective performance task requiring participants to infer complex emotional states from photographs of the eye region. Total TEQ scores significantly and positively predicted RMET performance ($r = .35, p < .001$), demonstrating that high TEQ scorers possess superior social-perceptual acuity.
• Prosocial Behavior: Experimental protocols examining resource allocation (e.g., dictator games and public goods games) demonstrate that individuals scoring higher on the TEQ exhibit significantly greater monetary donations, volunteer rates, and interpersonal helping behaviors.

8. Reliability

The psychometric integrity of the Toronto Empathy Questionnaire is evidenced by consistent internal consistency estimates and temporal stability across diverse international and clinical populations.

Internal Consistency

In the initial scale development study across three diverse undergraduate samples ($N = 388$, $N = 227$, and $N = 100$), Spreng et al. (2009) reported high internal consistency:

• Sample 1: Cronbach’s $\alpha = .87$
• Sample 2: Cronbach’s $\alpha = .85$
• Sample 3: Cronbach’s $\alpha = .85$

Subsequent psychometric cross-validations globally have documented comparable internal reliability coefficients:

• Italian Validation (Falcone et al., 2013): $\alpha = .81$ in a community sample of 855 adults.
• Turkish Adaptation (Totan et al., 2012): $\alpha = .82$ among university cohorts.
• Chinese Validation (Zhu et al., 2018): $\alpha = .83$, with McDonald’s $\omega = .84$, affirming strong scale composite reliability.
• Clinical Samples: In patients with traumatic brain injury and schizophrenia, alpha coefficients have remained stable, typically ranging between .80 and .86, demonstrating that the 16 items maintain strong inter-item covariance even within neurologically and psychiatrically impaired cohorts.

Test-Retest Reliability

Temporal stability was established by Spreng et al. (2009) by readministering the TEQ over a two-month follow-up period to a subset of participants ($n = 43$). The test-retest reliability coefficient was $r = .81$ ($p < .001$), demonstrating that the instrument assesses a stable trait-like disposition rather than transient, state-dependent emotional reactivity. A subsequent four-week test-retest evaluation conducted by Kourmousi et al. (2017) in a Greek cohort confirmed excellent stability with an intraclass correlation coefficient (ICC) of .84.

9. Factor Analysis

The empirical development and validation of the Toronto Empathy Questionnaire relied heavily on exploratory and confirmatory factor analysis (EFA and CFA) to address the multi-instrument fragmentation of empathy assessment.

Exploratory Factor Analysis and Scale Derivation

In Study 1 of Spreng et al. (2009), 388 participants completed a wide battery of empathy instruments, including the Interpersonal Reactivity Index (IRI), the Empathy Quotient (EQ), the Hogan Empathy Scale (HES), the Balanced Emotional Empathy Scale (BEES), and related social cognitive measures. A principal components analysis (PCA) with unrotated extraction was conducted on the full pool of items to identify common variance across all established inventories.

The analysis revealed a prominent first unrotated factor that accounted for the largest proportion of total variance. Items with factor loadings greater than $|.40|$ on this common factor were scrutinized for conceptual relevance, linguistic clarity, and construct coverage. A total of 16 items were retained, deliberately balanced between 8 positively phrased items (indicative of empathic responsiveness) and 8 negatively phrased items (indicative of empathic absence or aversion) to control for acquiescence response bias.

Confirmatory Factor Analysis and Structural Fit

In subsequent independent samples, Spreng et al. (2009) evaluated the fit of a single-factor latent structure using Confirmatory Factor Analysis (CFA). In an independent sample of 227 participants, the single-factor model exhibited adequate fit to the empirical data:

• Chi-Square / Degrees of Freedom: $\chi^2(104) = 179.35, p < .001$ ($\chi^2/df = 1.72$, well below the conservative threshold of 2.0).
• Comparative Fit Index (CFI): .91
• Incremental Fit Index (IFI): .91
• Root Mean Square Error of Approximation (RMSEA): .056 (90% Confidence Interval: .042 – .070), indicating good model fit.
• Standardized Root Mean Square Residual (SRMR): .051

Standardized item-factor loadings ($lambda$) across the 16 items were uniformly robust, ranging from .38 to .73 (all $p < .001$). Below is a representative summary of standardized item loadings derived from initial and cross-validation studies:

While some subsequent researchers (e.g., in translated versions) have posited potential two-factor models separating positively and negatively worded items (a well-documented methodological artifact known as the “method effect” of reverse-coded items), multigroup invariance analyses and bifactor modeling confirm that the single general factor accounts for over 80% of the common variance, justifying the calculation of a single composite empathy score.

10. Instrument / Measurement Tool

• Test Name: The Toronto Empathy Questionnaire (TEQ)
• Original Authors: R. Nathan Spreng, Margaret C. McKinnon, Raymond A. Mar, & Brian Levine
• Publication Year: 2009
• Construct Assessed: Affective Empathy (unidimensional construct encompassing emotional contagion, empathic concern, social-emotional decoding, and prosocial motivation)
• Administration Format: Self-administered paper-and-pencil or computerized self-report questionnaire
• Completion Time: Approximately 3 to 5 minutes
• Item Count: 16 items
• Response Scale: 5-point Likert scale (0 = Never, 1 = Rarely, 2 = Sometimes, 3 = Often, 4 = Always)
• Scoring System:
  • Positively Phrased Items (Scored directly 0 to 4): Items 1, 3, 5, 6, 8, 9, 13, and 16.
    
    Never = 0; Rarely = 1; Sometimes = 2; Often = 3; Always = 4
  • Negatively Phrased Items (Reverse-scored 4 to 0): Items 2, 4, 7, 10, 11, 12, 14, and 15.
    
    Never = 4; Rarely = 3; Sometimes = 2; Often = 1; Always = 0
  • Score Derivation: Total score is calculated by summing the scores of all 16 items.
  • Score Range: Minimum score = 0; Maximum score = 64.
• Interpretation Guidelines:
  • Higher Scores (> 45): Reflect high levels of emotional empathy, robust emotional resonance, prosocial orientation, and keen sensitivity to others’ affective states. Normative samples of healthy young adults typically average approximately 44 to 48 (with females generally scoring 3 to 5 points higher than males).
  • Average Scores (36 – 44): Indicate typical empathic responsiveness consistent with general population community norms.
  • Below-Average Scores (< 35): Suggest reduced affective sensitivity, difficulties with emotional contagion, or avoidance of interpersonal distress. Scores below 30 are frequently observed in clinical populations characterized by socio-affective impairments, such as Autism Spectrum Conditions, severe personality pathology, or acquired fronto-limbic brain trauma.

11. Permissions & Fee and Test Year

The Toronto Empathy Questionnaire was published in 2009 in the Journal of Personality Assessment (Volume 91, Issue 1). In accordance with standard academic dissemination practices and the intentions of the scale’s creators, the TEQ is considered an open-access, non-commercial psychological assessment instrument.

Researchers, educators, and clinicians are permitted to use, administer, reproduce, and translate the questionnaire for non-commercial research, academic inquiry, and educational purposes without paying royalty fees, provided that appropriate scholarly attribution is formally granted to Spreng, McKinnon, Mar, and Levine (2009). The scale items are published in full within the original validation article. Clinical licensing or institutional integration within commercial, fee-for-service software platforms or consulting environments requires explicit written permission from the corresponding author, Dr. R. Nathan Spreng, and the publisher (Taylor & Francis).

12. References

• Baron-Cohen, S., & Wheelwright, S. (2004). The empathy quotient: An investigation of adults with Asperger syndrome or high functioning autism, and normal sex differences. Journal of Autism and Developmental Disorders, 34(2), 163–175. https://doi.org/10.1023/B:JADD.0000022607.19833.00
• Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & Plumb, I. (2001). The “Reading the Mind in the Eyes” test revised version: A study with normal adults, and adults with Asperger syndrome or high-functioning autism. Journal of Child Psychology and Psychiatry, 42(2), 241–251. https://doi.org/10.1111/1469-7610.00715
• Brook, M., & Kosson, D. S. (2013). Impairments in facial affect recognition associated with psychopathy: An examination of specific affective signals and their cognitive correlates. Criminal Justice and Behavior, 40(12), 1479–1496. https://doi.org/10.1177/0093854813500958
• 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
• Decety, J., & Jackson, P. L. (2004). The functional architecture of human empathy. Behavioral and Cognitive Neuroscience Reviews, 3(2), 71–100. https://doi.org/10.1177/1534582304267187
• Falcone, G., Copertaro, A., Ferretti, F., & Bracco, F. (2013). Italian validation of the Toronto Empathy Questionnaire. Bollettino di Psicologia Applicata, 268, 47–56.
• Kourmousi, N., Amanaki, E., Tzavara, C., Merakou, K., Barbouni, A., & Koutras, V. (2017). The Toronto Empathy Questionnaire: Reliability and validity in a nation-wide sample of Greek teachers. Social Sciences, 6(2), Article 62. https://doi.org/10.3390/socsci6020062
• Preston, S. D., & de Waal, F. B. (2002). Empathy: Its ultimate and proximate bases. Behavioral and Brain Sciences, 25(1), 1–20. https://doi.org/10.1017/S0140525X02000018
• Spreng, R. N., McKinnon, M. C., Mar, R. A., & Levine, B. (2009). The Toronto Empathy Questionnaire: Scale development and initial validation of a factor-analytic solution to multiple empathy measures. Journal of Personality Assessment, 91(1), 62–71. https://doi.org/10.1080/00223890802484381
• Totan, T., Doğan, T., & Sapmaz, F. (2012). The Toronto Empathy Questionnaire: Evaluation of psychometric properties among Turkish university students. Eurasian Journal of Educational Research, 46, 179–198.
• Zhu, H., Shen, C., & Spreng, R. N. (2018). Psychometric evaluation and factor structure of the Chinese version of the Toronto Empathy Questionnaire. Frontiers in Psychology, 9, Article 2486. https://doi.org/10.3389/fpsyg.2018.02486

13. Items of the Scale