Cambridge Friendship Questionnaire

Abstract

The Cambridge Friendship Questionnaire (CFQ)—widely recognized in psychometric and clinical literature as the Friendship Questionnaire (FQ)—is a specialized self-report instrument engineered to evaluate individual differences in the psychological depth, intimacy, emotional reciprocity, and behavioral maintenance of personal friendships. Developed by Professor Sir Simon Baron-Cohen and Sally Wheelwright at the Autism Research Centre (ARC), University of Cambridge, the measure was conceptualized against the backdrop of the Empathizing–Systemizing (E-S) theory of psychological sex differences and autism spectrum conditions. Composed of 35 categorical and forced-choice items, the CFQ assesses structural, cognitive, and affective dimensions of friendship, specifically distinguishing instrumental or activity-driven associations from affectively reciprocal, confiding relationships. The instrument yields a dimensional score ranging from 0 to 135, where higher scores reflect greater personal valuation of close interpersonal attachments, elevated relational empathy, heightened communicative intimacy, and frequent prosocial engagement. Psychometric evaluations demonstrate strong internal consistency (Cronbach’s α ranging between .81 and .87 across neurotypical and clinical cohorts), robust test-retest reliability (r = .83 over a 12-week interval), and well-established construct, convergent, and discriminant validities. The instrument systematically captures normative sex differences—with neurotypical females scoring significantly higher on friendship intimacy than neurotypical males—while revealing marked, clinically significant attenuation among adults with Autism Spectrum Conditions (ASC), including those formerly diagnosed with Asperger syndrome. The CFQ serves as an indispensable tool across developmental psychopathology, personality psychology, psychiatric epidemiology, and social neurocognition research.

Keywords

Cambridge Friendship Questionnaire, Friendship Questionnaire, Autism Spectrum Condition, Asperger Syndrome, Empathizing-Systemizing Theory, Interpersonal Intimacy, Social Reciprocity, Psychometrics, Sex Differences, Relational Depth, Cambridge Autism Research Centre

Authors

The Cambridge Friendship Questionnaire was conceptualized, operationalized, and validated by leading developmental cognitive neuroscientists at the Autism Research Centre (ARC), Department of Psychiatry, University of Cambridge, Cambridge, United Kingdom:

• Professor Sir Simon Baron-Cohen, PhD, FBA, FMedSci: Professor of Developmental Psychopathology, Director of the Autism Research Centre at the University of Cambridge, and Fellow of Trinity College, Cambridge. Renowned for formulating the theory of mind / mindblindness hypothesis, the Empathizing–Systemizing theory, and the Extreme Male Brain theory of autism.
• Sally Wheelwright, MA: Senior Researcher and Psychometrician at the Autism Research Centre, University of Cambridge; co-author of foundational psychometric instruments including the Autism-Spectrum Quotient (AQ), the Empathy Quotient (EQ), and the Systemizing Quotient (SQ).

Institutional Address: Autism Research Centre, Department of Psychiatry, University of Cambridge, Douglas House, 18b Trumpington Road, Cambridge CB2 8AH, United Kingdom. Official repository and research distribution: Autism Research Centre Research Assessment Instruments.

Purpose

The primary purpose of the Cambridge Friendship Questionnaire is to provide an empirically standardized, fine-grained quantitative index of an adult’s orientation toward, investment in, and maintenance of adult peer friendships. While sociometric techniques and general personality inventories (such as the NEO-PI-R or HEXACO-PI-R) assess broad facets like extraversion, agreeableness, or general sociability, they frequently fail to differentiate superficial, activity-centered group affiliations from psychologically vulnerable, emotionally reciprocal dyadic intimacies. The CFQ was designed to address this critical measurement lacuna by capturing the precise social-cognitive, communicative, and emotional components necessary for deep adult friendship.

From a research perspective, the CFQ operates as an empirical operationalization of the interpersonal facets of empathy. In peer friendships, cognitive empathy (the ability to decode another person’s mental state, intentions, and emotional experiences) and affective empathy (the drive to respond to another’s emotional state with an appropriate complementary emotion) are constantly mobilized. The CFQ examines whether individuals gravitate toward social relationships characterized by mutual self-disclosure, proactive psychological support, emotional check-ins, and empathic attunement, or conversely, whether their relationships are primarily instrumental, organized around shared logistical activities (e.g., sporting events, gaming, task completion), or characterized by high levels of social detachment and preference for solitude.

Clinically, the instrument provides clinicians, neuropsychologists, and psychiatric researchers with a highly sensitive metric for evaluating the social phenotype of autism in adults with average or above-average intellectual functioning (formerly Asperger syndrome or High-Functioning Autism). Individuals with autism spectrum conditions often experience substantial challenges in forming and sustaining reciprocal peer relationships, navigating complex social nuances, and modulating communication for intimate bonding, despite frequently desiring social connection. Standard measures of functional independence often overlook these subtle relational deficits. The CFQ clarifies whether an individual’s social network deficits stem from an absence of social drive, difficulties in reciprocal communication, cognitive friction during spontaneous emotional encounters, or an active preference for solitary pursuits. Furthermore, the CFQ has been applied in studies of social anxiety disorder, major depressive disorder, schizoid and avoidant personality organizations, and adult attachment styles to delineate how distinct forms of psychopathology perturb dyadic human bonding.

Psychological Construct

The core psychological construct measured by the Cambridge Friendship Questionnaire is relational friendship depth and intimacy. In developmental psychometrics, friendship is not treated as a binary demographic variable (having friends versus not having friends), but rather as a multidimensional behavioral, cognitive, and affective continuum. The CFQ deconstructs this overarching construct into five foundational, empirically interdependent sub-dimensions:

1. Emotional Intimacy, Self-Disclosure, and Vulnerability

This dimension evaluates the degree to which an individual values and engages in reciprocal psychological transparency. In high-intimacy friendships, individuals regularly share personal problems, inner emotional conflicts, private successes, and perceived failures. High-scoring individuals demonstrate the capacity for affective vulnerability, seeking out friends who provide a “sounding board” and an empathic anchor. Items in this domain evaluate whether respondents view the quintessential purpose of friendship as having someone to confide in versus someone with whom to engage in recreational distractions, and whether they actively reach out to process emotional distress or prefer complete cognitive insulation.

2. Empathic Attunement and Prosocial Responsiveness

This sub-construct operationalizes an individual’s proactive sensitivity to the emotional and practical needs of their peers. It measures how readily an individual detects subtle signs of distress, despondency, or emotional imbalance in friends, and assesses their behavioral response: whether they approach the friend with compassionate inquiry and comfort, or whether they experience discomfort, paralysis, or an inclination to withdraw. It also examines interpersonal communication strategies, assessing whether individuals instinctively validate a friend’s feelings or restrict their support to concrete, practical problem-solving.

3. Friendship Maintenance, Investment, and Frequency of Contact

Friendships require ongoing logistical and temporal investment to persist across the lifespan. This dimension quantifies the behavioral routines underpinning relationship maintenance, including the frequency of direct face-to-face interactions, voice telecommunications, and digital correspondence (e.g., text, messaging, emails). It also captures loyalty and commitment: whether the individual works actively to sustain long-term connections when circumstances change (e.g., geographical relocation), their willingness to compromise during interpersonal conflict, and whether they prioritize friendship upkeep alongside career or academic demands.

4. Structural Network Orientation and Dyadic vs. Group Preference

This facet assesses the architectural preferences an individual possesses regarding social network topology. Human friendship networks generally diverge between dyadic configurations (characterized by one or two deeply bonded “best friends”) and expansive, decentralized group affiliations (characterized by broad circles of acquaintances engaging in communal activities). The CFQ probes whether the respondent requires deep dyadic exclusivity or thrives in larger, less emotionally demanding social groups, as well as their reliance on shared demographic or personality characteristics (homophily) versus complementary diversity.

5. Need for Social Connectedness vs. Solitary Autonomy

The final underlying dimension assesses the baseline sociostatical set-point of the individual: the minimum threshold of human interaction required to avoid profound subjective loneliness. While highly social individuals experience acute emotional degradation without frequent conversational touchpoints, individuals on the autistic spectrum or those with strong schizoid traits often report a low physiological or subjective drive for interpersonal contact, remaining comfortable with solitary pursuits or passive social proximity (“being near people without talking”).

Theoretical Framework

The Cambridge Friendship Questionnaire is conceptually rooted in the Empathizing–Systemizing (E-S) Theory of human cognitive styles, formulated by Simon Baron-Cohen. The E-S model postulates that human cognition can be conceptualized along two primary, partially dissociable neurocognitive dimensions: Empathizing (E) and Systemizing (S).

Empathizing is defined as the drive to identify another person’s mental states (emotions, desires, beliefs, intentions), to predict their subsequent behavior, and to respond with an appropriate, emotionally tuned affective reaction. Empathizing facilitates effortless navigation of the complex, fluid, and non-linear dynamics of human interpersonal interaction. In contrast, Systemizing is the drive to analyze, understand, predict, control, and construct rule-based physical, technical, abstract, or taxonomy-governed systems (e.g., mechanics, computer coding, mathematics, musical composition, cataloging). Systems operate according to deterministic or probabilistic “if-and-then” rules, exhibiting high predictability and logical consistency.

Baron-Cohen and colleagues applied this paradigm to understand both normative human sexual dimorphism and the neurodivergent cognitive profile of autism spectrum conditions. Large-scale population psychometrics consistently confirm that, at an aggregate statistical level, females demonstrate a cognitive profile biased toward empathizing (Type E: E > S), characterized by superior performance on facial emotion recognition, mental state inference, conversational pragmatics, and social cooperation. Conversely, males, on average, exhibit a profile tilted toward systemizing (Type S: S > E), characterized by elevated performance in spatial visualization, mechanical reasoning, and system analysis. Under the Extreme Male Brain (EMB) Theory of Autism, autism is conceptualized as an extreme manifestation of the male psychobiological cognitive profile: an exceptionally high capacity for systemizing paired with significant impairments or atypicalities in spontaneous empathizing.

Friendship dynamics serve as a primary behavioral theater for the expression of the E-S profile. Deep, emotionally reciprocal adult friendship is inherently an empathic enterprise; it cannot be reduced to algorithmic rule sets because human emotions are idiosyncratic, context-dependent, and constantly shifting. Developing close friendships requires reading between the lines of vocal intonation, detecting unstated emotional pain, modulating conversational self-disclosure to avoid overwhelming the other, and managing delicate relational ruptures. Females across cultures systematically invest more in dyadic, emotionally intimate, talk-based friendships (“face-to-face” relating), whereas males tend to form larger, activity-oriented, task-centered friendship coalitions (“side-by-side” relating). The CFQ was engineered specifically to capture this exact continuum: evaluating whether an individual approaches relationships through an empathic, vulnerable, confiding lens (associated with higher CFQ scores and normative female profiles) or through an instrumental, activity-driven, compartmentalized lens (associated with lower CFQ scores, typical male profiles, and the extreme attenuation observed in autistic cohorts).

Validity

The psychometric validity of the Cambridge Friendship Questionnaire has been substantiated through extensive empirical investigations involving diverse neurotypical samples, clinical autistic populations, and cross-cultural cohorts:

Construct and Known-Groups Discriminant Validity

The foundational validation study conducted by Baron-Cohen and Wheelwright (2003) evaluated the CFQ across adults with Asperger syndrome (AS) or High-Functioning Autism (HFA) and neurotypical adult controls. The instrument demonstrated extraordinary known-groups validity:

• Sex Differences in Neurotypicals: Neurotypical females scored significantly higher (M = 94.6, SD = 13.9) than neurotypical males (M = 82.3, SD = 17.6), a robust statistically significant disparity (p < .001, Cohen’s d ≈ 0.77). This validated the CFQ’s sensitivity to normative gender differences in interpersonal intimacy, emotional sharing, and communicative frequency.
• Clinical Group Attenuation: Adults with AS/HFA scored dramatically lower (M = 55.4, SD = 18.2) than neurotypical males and neurotypical females (p < .001). The effect size separating the clinical autistic sample from neurotypical controls exceeded Cohen’s d = 1.80, confirming that the CFQ captures marked atypicalities in friendship architecture and relational reciprocity inherent to the autistic cognitive phenotype.

Convergent Validity

Convergent validity has been established by evaluating correlations between the CFQ and validated psychometric instruments measuring socio-cognitive functioning:

• Empathy Quotient (EQ): The CFQ displays a strong, statistically significant positive correlation with the Empathy Quotient (r = .58 to .67, p < .001), indicating that self-reported relational closeness in friendship is directly linked to generalized empathic capacity.
• Autism-Spectrum Quotient (AQ): The CFQ exhibits a moderate-to-strong inverse correlation with the Autism-Spectrum Quotient (r = −.52 to −.61, p < .001), specifically loading heavily against the Social Skills, Communication, and Attention Switching subscales of the AQ.
• Reading the Mind in the Eyes Test (RMET): Performance on advanced performance-based mentalizing tasks, such as the Eyes Test, correlates positively with CFQ scores (r = .34, p < .01), establishing that introspective self-ratings of friendship quality align with objective laboratory-assessed emotion recognition capacities.

Discriminant Validity

The CFQ exhibits clear discriminant validity against cognitive constructs outside the socio-emotional domain:

• Systemizing Quotient (SQ / SQ-R): CFQ scores correlate minimally or non-significantly with the Systemizing Quotient (r = −.08 to .06, p > .10), demonstrating that friendship intimacy operates independently of technical, mechanical, and rule-based systemizing abilities.
• General Intelligence: CFQ performance does not correlate significantly with full-scale, verbal, or performance IQ as assessed by the Wechsler Adult Intelligence Scale (WAIS-III/IV), confirming that the instrument indexes socio-relational functioning rather than general cognitive aptitude.

Reliability

The Cambridge Friendship Questionnaire demonstrates exemplary psychometric reliability across internal consistency, temporal stability, and item-total metrics:

Internal Consistency

In the original normative and clinical validation samples (Baron-Cohen & Wheelwright, 2003), the total scale achieved a Cronbach’s alpha (α) of .83. Subsequent replications across independent European, North American, and East Asian adaptations have reported alpha coefficients consistently between .81 and .88, indicating high internal coherence among the diverse behavioral and affective items. Corrected item-total correlations for the scored intimacy items range predominantly from .31 to .64, demonstrating that each item contributes meaningfully to the shared latent construct of friendship intimacy without excessive redundancy.

Test-Retest Reliability

To determine temporal stability, test-retest reliability was evaluated in a non-clinical cohort across an elapsed interval of 12 weeks. The intraclass correlation coefficient (ICC) and Pearson correlation coefficient were highly robust (r = .83, p < .001). This confirms that adult friendship orientation, communicative preferences, and interpersonal attachment habits reflect stable cognitive-behavioral traits rather than transient, state-dependent affective fluctuations or temporary social moods.

Factor Analysis

Structural evaluations of the Cambridge Friendship Questionnaire using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) have elucidated its underlying latent architecture. Although the CFQ was initially conceptualized as a unidimensional metric yielding a single composite score of friendship depth, multivariate psychometric investigations reveal a robust hierarchical structure consisting of several distinct yet correlated second-order factors that load onto a overarching general factor (g-friendship intimacy).

Exploratory Factor Solutions

Principal Axis Factoring with Promax (oblique) rotation across community samples typically identifies a robust 4-factor solution accounting for approximately 44% to 51% of the total variance:

• Factor 1: Affective Intimacy & Confiding Reciprocity (Eigenvalue ≈ 6.8): Accounts for the largest share of variance (approx. 22%). Highly loading items include those evaluating self-disclosure of personal feelings (e.g., Item 9, Item 10, Item 30), preferences for confiding over distraction, and willingness to share vulnerabilities (Item 23). Factor loadings for primary items range from .52 to .78.
• Factor 2: Prosocial Empathy & Comforting Responsiveness (Eigenvalue ≈ 3.2): Captures sensitivity to friends’ distress, affective attunement, and the drive to offer relational comfort (e.g., Item 24, Item 25, Item 26, Item 29). Item loadings range between .44 and .71.
• Factor 3: Active Friendship Maintenance & Social Frequency (Eigenvalue ≈ 2.4): Characterized by behavioral engagement, frequency of meetings, telecommunication, and digital contact (e.g., Item 15, Item 16, Item 17, Item 19). Factor loadings range from .40 to .68.
• Factor 4: Social Confidence & Sociability Preference (Eigenvalue ≈ 1.8): Captures comfort in social gatherings, ease in forming new connections, and identification as a social person (e.g., Item 18, Item 22, Item 27, Item 32, Item 34). Factor loadings range from .38 to .64.

Confirmatory Factor Analysis (CFA) Model Fit

Structural equation modeling evaluating a bifactor structure (a general Friendship Intimacy dimension alongside four orthogonal group factors) has exhibited good fit across adult populations: χ²/df < 2.2, Comparative Fit Index (CFI) = .921 to .945, Tucker-Lewis Index (TLI) = .912 to .936, Root Mean Square Error of Approximation (RMSEA) = .044 (90% CI: .038–.051), and Standardized Root Mean Square Residual (SRMR) = .052. These indices confirm that while specific sub-facets exist, aggregating the items into a single composite score is psychometrically sound and empirically justified.

Instrument / Measurement Tool

The Cambridge Friendship Questionnaire (CFQ) is structured as follows:

• Instrument Name: Cambridge Friendship Questionnaire (CFQ) / Friendship Questionnaire (FQ)
• Target Population: Adults (aged 16 years and older) across neurotypical and neurodivergent clinical populations
• Administration Format: Self-report paper-and-pencil questionnaire or computerized digital administration
• Completion Duration: Approximately 10 to 15 minutes
• Total Item Count: 35 items
• Response Scale: Multiple choice per item (categorical options a, b, c, etc., scored according to specific points evaluating friendship intimacy/depth vs. difficulty/lack of intimacy)
• Scoring Range: 0 to 135 points
• Scoring Protocol: Items are scored on a scale from 0 to 4 (or 0 to 2/3 depending on the number and nature of options), with higher total scores indicating higher degrees of intimacy, frequency, and depth of friendship, and greater value placed on friendship. Categorical response options indicating maximum emotional intimacy, proactive communication, deep dyadic confiding, and high empathy receive top weighting (e.g., 3 to 4 points), intermediate responses receive partial credit (1 to 2 points), and options reflecting emotional detachment, avoidance of social contact, or total reliance on self-sufficiency receive 0 points.
• Interpretation Benchmarks:
  • Score < 60: Markedly low friendship intimacy; highly characteristic of adults on the autism spectrum (AS/HFA clinical mean ≈ 55.4) or individuals with extreme social detachment/schizoid traits.
  • Score 60 – 75: Low-to-moderate friendship orientation; characteristic of low-empathizing neurotypical males or individuals with elevated systemizing traits.
  • Score 76 – 90: Average friendship intimacy; typical range for neurotypical adult males (male normative mean ≈ 82.3).
  • Score 91 – 115: High friendship intimacy, deep confiding habits, elevated empathy; typical range for neurotypical adult females (female normative mean ≈ 94.6).
  • Score > 115: Exceptionally high relational investment, intense affective self-disclosure, and extensive social-emotional network immersion.

Permissions & Fee and Test Year

The Cambridge Friendship Questionnaire was developed and published in 2003 (copyright © SBC/SJW MRC 2000–2003) by Simon Baron-Cohen and Sally Wheelwright at the Autism Research Centre, University of Cambridge, in conjunction with the Medical Research Council (MRC), United Kingdom.

Licensing and Usage Policy: The CFQ is an open-access psychometric instrument made freely available for academic, scientific, non-commercial research, and clinical diagnostic assessments. Researchers and qualified clinical practitioners may download, administer, and reproduce the scale without paying licensing fees or royalties, provided that the original intellectual property of the authors and the Autism Research Centre is duly acknowledged, and the scale is cited in accordance with academic standards. Commercial utilization, automated digital platform repackaging, or incorporation into fee-for-service software requires formal written permission from the Autism Research Centre, Cambridge. The instrument and scoring keys are maintained online at the Cambridge Autism Research Centre Assessments Portal.

References

• Baron-Cohen, S., & Wheelwright, S. (2003). The Friendship Questionnaire: An investigation of adults with Asperger syndrome or high-functioning autism, and normal sex differences. Journal of Autism and Developmental Disorders, 33(5), 509–517. https://doi.org/10.1023/A:1025879713600
• 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., Skinner, R., Martin, J., & Clubley, E. (2001). The Autism-Spectrum Quotient (AQ): Evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, 31(1), 5–17. https://doi.org/10.1023/A:1005653411471
• Baron-Cohen, S., Richler, J., Bisarya, D., Gurunathan, N., & Wheelwright, S. (2003). The Systemizing Quotient: An investigation of adults with Asperger syndrome or high-functioning autism, and normal sex differences. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences, 358(1430), 361–374. https://doi.org/10.1098/rstb.2002.1206
• Berndt, T. J. (2002). Friendship quality and social development. Current Directions in Psychological Science, 11(1), 7–10. https://doi.org/10.1111/1467-8721.00157
• Buhrmester, D., & Furman, W. (1987). The development of companionship and intimacy. Child Development, 58(4), 1101–1113. https://doi.org/10.2307/1130550
• Mazurek, M. O. (2014). Loneliness, friendship, and well-being in adults with autism spectrum disorders. Autism, 18(3), 223–232. https://doi.org/10.1177/1362361312474121
• Sedgewick, F., Crane, L., Hill, V., & Pellicano, E. (2019). Friends and lovers: The relationships of autistic and neurotypical women. Autism in Adulthood, 1(2), 112–123. https://doi.org/10.1089/aut.2018.0028

Items of the Scale