McGill Friendship Questionnaire

1. Abstract

The McGill Friendship Questionnaire (MFQ), developed by Canadian developmental psychologists Morton J. Mendelson and Frances E. Aboud (1992, 1999), represents one of the most widely used and psychometrically validated assessment suites designed to evaluate the multi-faceted nature of dyadic friendship quality across late childhood, adolescence, and emerging adulthood. The instrument consists of two distinct yet complementary self-report inventories: the McGill Friendship Questionnaire–Respondent’s Affection (MFQ-RA) and the McGill Friendship Questionnaire–Friend’s Functions (MFQ-FF). The MFQ-RA evaluates the respondent’s personal subjective evaluation of and emotional investment in a specific friendship dyad, encompassing 16 items distributed across two primary subscales: Positive Feelings towards the friend and Satisfaction with the friendship relationship (with an earlier iteration also assessing global Evaluation). The MFQ-RA employs a 9-point bipolar Likert-type scale ranging from −4 (“Very Much Disagree”) through 0 (“Neutral”) to +4 (“Very Much Agree”). Conversely, the MFQ-FF measures the respondent’s perception of the degree to which the target friend fulfills vital socio-developmental and psychological provisions, consisting of a full 48-item version and an abbreviated 30-item short form structured across six theoretically grounded dimensions: Stimulating Companionship, Help, Intimacy, Reliable Alliance, Self-Validation, and Emotional Security. The MFQ-FF utilizes a 9-point frequency rating scale anchored from 0 or 1 (“Never”) to 8 (“Always”).

Psychometric evaluations demonstrate exceptional internal consistency across all subscales, with Cronbach’s alpha coefficients routinely exceeding α = .88 and reaching up to α = .96 for Positive Feelings, Intimacy, and Reliable Alliance in diverse developmental samples. Confirmatory factor analytic investigations consistently support the multidimensional architecture of friendship functions, showing that while the six provisions are positively intercorrelated, they differentiate key functional contributions to emotional well-being, buffer against depressive symptomatology and social anxiety, and predict dyadic longevity. By disaggregating subjective affection from specific functional interpersonal behaviors, the McGill Friendship Questionnaire provides clinical, educational, and developmental researchers with an indispensable methodology for diagnosing relational deficits, evaluating social skills interventions, and mapping the developmental trajectory of close interpersonal relationships.

2. Keywords

McGill Friendship Questionnaire, MFQ, MFQ-FF, MFQ-RA, friendship quality, close relationships, peer relations, adolescent social development, intimacy, reliable alliance, self-validation, emotional security, psychometrics

3. Authors

The McGill Friendship Questionnaire suite was conceptualized, designed, and psychometrically validated by researchers based in the Department of Psychology at McGill University in Montreal, Quebec, Canada:

• Morton J. Mendelson, Ph.D.: Professor Emeritus of Psychology and former Deputy Provost (Student Life and Learning) at McGill University. Dr. Mendelson’s empirical scholarship focuses on child and adolescent social development, peer relationships, friendship dynamics, emotional well-being, and psychometric measurement. Contact: [email protected].
• Frances E. Aboud, Ph.D.: Professor Emerita of Psychology at McGill University. Dr. Aboud is an internationally renowned developmental psychologist whose research encompasses peer relationships, child social cognition, prejudice and intergroup attitudes, and global child health and cognitive development interventions.
• Subsequent Contributors: Dr. Richard P. Lanthier (George Washington University) and Dr. Aaron C. Kay (Duke University), who collaborated on developmental extensions, kindergarten sociometric adaptations, and dyadic reciprocity dynamics of the scale.

4. Purpose

Friendships represent fundamental interpersonal contexts that promote social competence, emotional regulation, identity consolidation, and resilience across the lifespan. However, prior to the conceptual formulation of the McGill Friendship Questionnaires, empirical research on peer relationships was heavily dominated by sociometric popularity techniques (e.g., peer nominations and social acceptance matrices) or unidimensional measures of general relationship satisfaction. Such traditional paradigms often conflated being widely liked by a peer group with having high-quality, mutually supportive dyadic friendships. Furthermore, existing dyadic instruments often blurred the boundary between an individual’s personal emotional attachment to a peer and the objective or perceived socio-emotional provisions supplied by that peer.

The primary purpose of the McGill Friendship Questionnaire was to resolve these conceptual and methodological limitations by establishing a dual-inventory measurement battery that systematically distinguishes between:

1. The respondent’s internal subjective feelings, attachment, and satisfaction directed toward the friend (captured by the MFQ-RA).
2. The specific interpersonal, developmental, and supportive functions enacted by the friend within the dyad (captured by the MFQ-FF).

From a theoretical perspective, the MFQ was designed to operationalize the friendship provisions initially posited by developmental and clinical theorists such as Harry Stack Sullivan and Robert S. Weiss. Sullivan argued that close interpersonal relationships during preadolescence and adolescence serve essential functions that protect individuals from loneliness, cultivate self-worth, and foster empathic capabilities. The MFQ translates these theoretical postulations into concrete, psychometrically testable behavioral dimensions that allow researchers to examine which specific interpersonal mechanisms contribute to psychological adjustment.

In clinical and educational settings, the MFQ has wide applicability. It enables clinicians to assess children, adolescents, and young adults who experience social alienation, depression, victimization, or social anxiety. By administering the MFQ, practitioners can determine whether an individual’s relational vulnerabilities stem from an inability to experience positive affect and relationship satisfaction, a deficit in establishing intimate self-disclosure, or an absence of reliable, trustworthy companions. In school psychology and developmental research, the MFQ provides a robust criterion measure for evaluating anti-bullying initiatives, social-emotional learning (SEL) curricula, and peer-support interventions designed to enhance dyadic relational competence.

5. Psychological Construct

The McGill Friendship Questionnaire operationalizes friendship quality as a multidimensional construct comprising both affective-evaluative attitudes and functional behavioral provisions. Rather than viewing friendship quality as a single monolithic entity, Mendelson and Aboud (1999) established eight distinct psychological dimensions organized across the two inventories.

Dimensions of the Respondent’s Affection (MFQ-RA)

• Positive Feelings (P): Reflects the intensity of the respondent’s personal affection, emotional attachment, liking, and dedication toward the friend. It measures expressions of personal closeness, happiness in the friend’s company, longing during periods of separation, and the desire to maintain the relationship over an extended temporal horizon (e.g., “I care about ___”, “I feel close to ___”).
• Satisfaction (S): Captures the cognitive and affective appraisal of the friendship dyad’s overall health, strength, and fulfillment. It reflects the degree to which the relationship meets the respondent’s interpersonal standards and needs (e.g., “I am satisfied with my friendship with ___”, “I think my friendship with ___ is strong”).
• Evaluation (E; Original Version): In the original 1992 formulation of the MFQ-RA, three items assessed general positive regard for the friend’s intrinsic personal qualities (e.g., “I think ___ is a good person”). In the revised MFQ-RA (1999), these evaluative items were replaced with direct indices of contentment and pleasure with the friendship itself to ensure pure measurement of relational satisfaction.

Dimensions of the Friend’s Functions (MFQ-FF)

The MFQ-FF measures six distinct socio-developmental functions that close friends provide for one another:

• Stimulating Companionship (S): The extent to which the friend provides engaging, entertaining, enjoyable, and lively interactions. This dimension reflects playful engagement, shared humor, creative activity generation, and the alleviation of boredom (e.g., “___ is fun to do things with”, “___ makes me laugh”, “___ has good ideas about entertaining things to do”).
• Help (H): The provision of instrumental, informational, and material assistance. This encompasses guidance, lending personal belongings, offering objective advice, and assisting with tasks or problem resolution (e.g., “___ helps me when I need it”, “___ lends me things that I need”, “___ shows me how to do things better”).
• Intimacy (I): The degree of reciprocal emotional openness, mutual disclosure of sensitive private thoughts, secrets, and vulnerabilities, as well as perceptive sensitivity to the other’s emotional states (e.g., “___ is someone I can tell private things to”, “___ knows when I’m upset”, “___ understands my feelings”).
• Reliable Alliance (R): The perception of steadfast loyalty, mutual dependability, ongoing commitment, and unconditional acceptance that survives external pressures, arguments, physical separation, and social disapproval (e.g., “___ would still want to be my friend even if we had a fight”, “___ would stay my friend even if other people criticized me”).
• Self-Validation (SV): The degree to which the friend bolsters the respondent’s self-esteem, self-efficacy, and personal worth through encouragement, genuine compliments, reassurance of competence, and unconditional positive regard even following failures (e.g., “___ makes me feel important”, “___ compliments me when I do something well”, “___ makes me feel that I can do things well”).
• Emotional Security (E): The sense of safety, comfort, and anxiety reduction provided by the friend in threatening, unfamiliar, frightening, or stressful situations (e.g., “___ would make me feel comfortable in a new situation”, “___ would make me feel calmer if I were nervous”).

6. Theoretical Framework

The architecture of the McGill Friendship Questionnaire is anchored in developmental relationship theory, drawing heavily upon the interpersonal psychiatry of Harry Stack Sullivan (1953), the social provisions framework of Robert S. Weiss (1974), and contemporary peer-relations models articulated by Willard W. Hartup (1996) and Wyndol Furman and Duane Buhrmester (1985).

Sullivan’s Interpersonal Theory of Psychiatry

Sullivan posited that human personality develops primarily within interpersonal matrices. Specifically, during the “chumship” phase of preadolescence (roughly ages 9 to 12), children experience a dramatic shift from egocentric social needs to an intense need for interpersonal intimacy. A chum is an intimate peer of equal status with whom one shares mutual validation, collaborative problem-solving, and private disclosures. Sullivan argued that a high-functioning chumship serves as an emotional crucible that can actively remediate psychological injuries, insecure attachment patterns, and low self-esteem sustained within earlier adverse family environments. The MFQ-FF dimensions of Intimacy, Self-Validation, and Reliable Alliance directly capture Sullivan’s concept of collaborative chumship.

Weiss’s Social Provisions Model

In The Experience of Loneliness (1974), Robert S. Weiss identified six relational provisions essential for adult psychological stability: attachment, social integration, opportunity for nurturance, reassurance of worth, reliable alliance, and guidance. Weiss argued that the absence of specific provisions leads to distinct psychological distress profiles; for example, a lack of emotional attachment generates emotional loneliness, while a deficit in social integration generates social loneliness. Mendelson and Aboud adapted this model specifically to peer friendships, translating abstract provisions into concrete peer-interaction functions such as Stimulating Companionship (social integration), Help (guidance), Reliable Alliance (dependable alliance), and Self-Validation (reassurance of worth).

Developmental Trajectory and Relational Differentiation

Developmental psychologists (Hartup, 1996; Newcomb & Bagwell, 1995) emphasize that friendship expectations evolve systematically across childhood, adolescence, and emerging adulthood. Young children define friends primarily in terms of common activities, propinquity, and concrete exchanges (corresponding to MFQ’s Stimulating Companionship and Help). As individuals transition into middle adolescence and emerging adulthood, psychological concerns turn toward identity validation, emotional safety, and loyalty (corresponding to Intimacy, Self-Validation, and Emotional Security). Mendelson and Aboud designed the MFQ to maintain measurement invariance across these developmental transitions, allowing researchers to observe how the relative predictive power of each function shifts over the life course.

7. Validity

The validity of both the MFQ-RA and MFQ-FF has been thoroughly demonstrated through extensive construct, convergent, discriminant, and predictive validation studies conducted in diverse populations.

Construct and Convergent Validity

Mendelson and Aboud (1999) established convergent validity by correlating the MFQ dimensions with established interpersonal and psychological adjustment measures. Both the MFQ-RA subscales (Positive Feelings and Satisfaction) and the MFQ-FF functional scales exhibit robust positive correlations with:

• The Network of Relationships Inventory (NRI) developed by Furman and Buhrmester (1985), with convergent correlations ranging from r = .62 to r = .81 across matching subscales such as companionship, intimacy, and instrumental aid.
• The Friendship Quality Questionnaire (FQQ) developed by Parker and Asher (1993), demonstrating high convergent coefficients on validation, help, and conflict resolution provisions (r > .65).
• Self-esteem indices, notably the Rosenberg Self-Esteem Scale, which correlates particularly strongly with the Self-Validation (r = .41, p < .001) and Emotional Security (r = .38, p < .001) subscales.

Discriminant Validity

Discriminant validity is supported by the scale’s ability to differentiate between close mutual friendships, casual acquaintances, and antagonistic peer relationships. In comparative trials, participants report significantly higher scores on all MFQ-FF subscales for nominated “best friends” compared to “regular friends” or “acquaintances” (paired t-tests yielding effect sizes ranging from Cohen’s d = 0.85 to 1.42). Crucially, the MFQ subscales correlate weakly or non-significantly with measures of social desirability (e.g., Marlowe-Crowne Social Desirability Scale, r < .15), verifying that high scores reflect genuine relationship quality rather than acquiescence or impression management.

Predictive and Longitudinal Validity

Empirical investigations demonstrate that the MFQ reliably predicts relational and psychological outcomes:

• Friendship Stability and Longevity: Longitudinal follow-up assessments over 6-month and 12-month periods show that baseline scores on Reliable Alliance and Satisfaction predict dyadic survival and conflict resolution success (odds ratio = 1.68, p < .01).
• Buffering Psychopathology: High scores on Emotional Security and Intimacy moderate the relationship between life stress and adolescent depressive symptoms, confirming the stress-buffering hypothesis of peer support.
• Dyadic Asymmetry: Mendelson and Kay (2000) utilized dyadic analyses to demonstrate that asymmetries in Positive Feelings between dyad members predict friendship dissolution and decreased relational well-being over time.

8. Reliability

The McGill Friendship Questionnaire consistently demonstrates exemplary internal consistency reliability and temporal stability across developmental age brackets, clinical cohorts, and cultural contexts.

Internal Consistency Reliability

Across validation studies conducted by Mendelson and Aboud (1992, 1994, 1999), Cronbach’s alpha coefficients for the full battery range from good to excellent:

• MFQ-RA Positive Feelings: α = .96
• MFQ-RA Satisfaction: α = .92
• MFQ-FF Stimulating Companionship: α = .91 (Full 48-item version); α = .87 (30-item short form)
• MFQ-FF Help: α = .89 (Full version); α = .84 (Short form)
• MFQ-FF Intimacy: α = .94 (Full version); α = .90 (Short form)
• MFQ-FF Reliable Alliance: α = .95 (Full version); α = .91 (Short form)
• MFQ-FF Self-Validation: α = .91 (Full version); α = .86 (Short form)
• MFQ-FF Emotional Security: α = .92 (Full version); α = .88 (Short form)
• MFQ-FF Total Scale Score: α = .98 across all 48 items

Test-Retest Stability

Temporal stability assessments conducted over intervals spanning 3 to 6 weeks yield test-retest intraclass correlation coefficients (ICCs) between r = .79 and r = .88 among late adolescents and university undergraduates. When friendships remain intact and unperturbed by major relational crises, individual functional ratings remain remarkably stable, demonstrating that the MFQ measures enduring relational attributes rather than transient situational mood states.

9. Factor Analysis

The factor structure of the McGill Friendship Questionnaires has been rigorously scrutinized using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Factor Architecture of the MFQ-RA

Early exploratory factor analyses with varimax and oblimin rotations of the original 16-item MFQ-RA yielded two strong, correlated factors accounting for over 68% of the total variance: Positive Feelings (explaining ~52% of variance) and Satisfaction (explaining ~16% of variance). Items evaluating the friend’s general character loaded weakly across both factors, leading Mendelson and Aboud (1999) to revise the scale by replacing the general evaluative items with focused friendship satisfaction items. A confirmatory two-factor model demonstrated an excellent fit to the data:

• χ²/df = 1.84
• Comparative Fit Index (CFI) = .97
• Tucker-Lewis Index (TLI) = .96
• Root Mean Square Error of Approximation (RMSEA) = .048 (90% CI [.036, .061])
• Standardized Root Mean Square Residual (SRMR) = .039

Factor Architecture of the MFQ-FF

For the MFQ-FF, theoretical conceptualization specified six distinct first-order factors corresponding to the six developmental provisions. In confirmatory factor analyses comparing competing models, Mendelson and Aboud (1999) evaluated:

1. A unidimensional general friendship quality model (where all 48 items load onto a single global factor).
2. A six-factor orthogonal model (assuming complete independence among functions).
3. A six-factor correlated model (reflecting differentiated yet interrelated friendship provisions).
4. A hierarchical second-order model (where a overarching higher-order Friendship Provision factor accounts for the six specific first-order functions).

The empirical findings decisively rejected the unidimensional model (χ²/df > 5.2, CFI = .74, RMSEA = .112) and the orthogonal model. The six-factor correlated model achieved superior fit indices (χ²/df = 2.01, CFI = .95, TLI = .94, RMSEA = .051, SRMR = .044), confirming that while friendship functions are interrelated (inter-factor correlations ranging between r = .51 and r = .78), each provision captures non-redundant variance. All item factor loadings onto their designated functional domains exceed .60, with the vast majority falling between .72 and .88.

10. Instrument / Measurement Tool

The McGill Friendship Questionnaire is a standardized, self-report psychometric battery designed for individual or group administration. The tool is organized as follows:

• Target Respondent Group: Children (aged 9+ with reading support), adolescents, university students, and adults.
• Administration Time:
  • MFQ-RA (16 items): 3–5 minutes
  • MFQ-FF Full Version (48 items): 10–15 minutes
  • MFQ-FF Short Form (30 items): 6–8 minutes
• Target Dyad Specification: Prior to completing the questionnaire, respondents enter the first name or initials of a specific friend into a designated prompt line (e.g., “___”), anchoring every evaluation to a single real-world relationship. Researchers can administer the scale multiple times to examine different relationships (e.g., best friend vs. casual friend).
• Response Scales:
  • MFQ-RA: 9-point bipolar scale ranging from −4 (“Very Much Disagree”) to +4 (“Very Much Agree”), with 0 representing neutral agreement.
  • MFQ-FF: 9-point frequency scale ranging from 1 (“Never”) to 8/9 (“Always”) indicating how often the friend exhibits each behavior.
• Scoring Procedures:
  • All items are positively keyed; no reverse scoring is required.
  • Subscale scores are calculated as the mean (or sum) of the items comprising each subscale. Using mean scores preserves the original 9-point metric, facilitating intuitive comparison across subscales.
  • A Total Friendship Functions score on the MFQ-FF can be derived by averaging all 48 (or 30) items to reflect global perceived friendship quality.

11. Permissions & Fee and Test Year

The McGill Friendship Questionnaires were originally formulated and presented in 1992 (Aboud & Mendelson) and formally published in peer-reviewed psychometric literature in 1999 (Mendelson & Aboud). The instruments are copyrighted by Morton J. Mendelson and Frances E. Aboud.

The authors have placed the McGill Friendship Questionnaires in the public academic domain for non-commercial scientific, educational, and clinical research purposes. No licensing fees or royalty payments are required for academic investigators, graduate theses, or educational institutions. Researchers wishing to use the questionnaires are requested to cite the foundational validation publications (Mendelson & Aboud, 1999) in all resulting manuscripts and reports. Inquiries regarding commercial use, adaptations, or non-English translations should be directed to Dr. Morton J. Mendelson at [email protected].

12. References

• Aboud, F. E., & Mendelson, M. J. (1992, November). Measurement and determinants of friendship quality. Invited address to the Conference on Friendship in Childhood and Adolescence, Montreal, Quebec, Canada.
• Aboud, F. E., & Mendelson, M. J. (1996). Determinants of friendship selection and quality: Developmental perspectives. In W. M. Bukowski, A. F. Newcomb, & W. W. Hartup (Eds.), The company they keep: Friendship in childhood and adolescence (pp. 87–112). Cambridge University Press. https://doi.org/10.1017/CBO9780511571121.006
• Furman, W., & Buhrmester, D. (1985). Children’s perceptions of the personal relationships in their social networks. Developmental Psychology, 21(6), 1016–1024. https://doi.org/10.1037/0012-1649.21.6.1016
• Hartup, W. W. (1996). The company they keep: Friendships and their developmental significance. Child Development, 67(1), 1–13. https://doi.org/10.2307/1131681
• Mendelson, M. J., & Aboud, F. E. (1999). Measuring friendship quality in late adolescents and young adults: McGill Friendship Questionnaires. Canadian Journal of Behavioural Science / Revue canadienne des sciences du comportement, 31(2), 130–132. https://doi.org/10.1037/h0087080
• Mendelson, M. J., Aboud, F. E., & Lanthier, R. P. (1994). Personality predictors of friendship and popularity in kindergarten. Journal of Applied Developmental Psychology, 15(3), 413–435. https://doi.org/10.1016/0193-3973(94)90022-4
• Mendelson, M. J., & Kay, A. C. (2000, August). Positive feelings in friendship: Does imbalance in the relationship matter? Poster presented at the 108th Annual Convention of the American Psychological Association, Washington, DC.
• Parker, J. G., & Asher, S. R. (1993). Friendship and friendship quality in middle childhood: Links with peer group acceptance and feelings of loneliness and social dissatisfaction. Developmental Psychology, 29(4), 611–621. https://doi.org/10.1037/0012-1649.29.4.611
• Sullivan, H. S. (1953). The interpersonal theory of psychiatry. W. W. Norton & Company.
• Weiss, R. S. (1974). Loneliness: The experience of emotional and social isolation. The MIT Press.

13. Items of the Scale (Questionnaire)