Social Relationship Scale (SRS)

Abstract

The Social Relationship Scale (SRS) is a psychometric instrument developed by Allan H. McFarlane and colleagues in 1981 at McMaster University, designed to evaluate the multi-faceted architecture of personal social support networks. Diverging from unidimensional indices that measure solely social network size or perceived emotional availability, the SRS operationalizes social relationships across distinct functional life domains, capturing both structural characteristics (network size, relationship categories, and reciprocity) and qualitative dimensions (perceived helpfulness versus unhelpfulness). The instrument employs a grid-based, roster-generation paradigm across life areas, most notably including family and home life, occupational pressures, financial concerns, and personal health. Within each domain, respondents identify specific network members, indicate reciprocal discussion dynamics (whether the nominated individual also approaches the respondent for support), and rate the qualitative impact of interactions on a 4-point Likert-type scale ranging from 1 (“makes things a lot worse”) to 4 (“helps things a lot”). Psychometric evaluations of the SRS demonstrate robust test-retest reliability across short- and medium-term intervals (Pearson correlation coefficients ranging from r = .60 to .90 across subscales) and strong construct, convergent, and discriminant validity in clinical, community, and psychiatric cohorts. The scale has provided critical empirical evidence demonstrating that unhelpful or negative interactions exert disproportionately larger influences on psychological distress, depressive symptomatology, and somatic complaints than positive interactions exert in mitigating them. This article offers an exhaustive academic evaluation of the SRS, delineating its theoretical foundations, structural composition, statistical validity, factor dynamics, clinical utility, scoring frameworks, and complete operational items.

Keywords

Social Relationship Scale, SRS, social support measurement, network reciprocity, interpersonal stress, negative social interactions, psychometrics, coping resources, stress buffering hypothesis, social network analysis.

Authors

The Social Relationship Scale was conceptualized, developed, and empirically validated by an interdisciplinary team of psychiatric epidemiologists, biostatisticians, and clinical psychologists at the Department of Psychiatry and the Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada:

• Allan H. McFarlane, MD, FRCP(C): Principal investigator and Professor of Psychiatry; pioneered the integration of network reciprocity and transactional stress frameworks into psychiatric epidemiology.
• Kathleen A. Neale, BA: Research associate and psychometrician; coordinated data collection pipelines, item screening protocols, and community validation studies.
• Geoffrey R. Norman, PhD: Distinguished Professor Emeritus of Biostatistics and Psychometrics; led the mathematical formalization of the scale’s structural indices and psychometric testing models.
• David L. Streiner, PhD, CPsych: Renowned health measurement methodologist and Professor of Psychiatry; co-authored authoritative psychometric treatises on rating scale construction, structural equation modeling, and scale refinement.

Purpose

The primary purpose of the Social Relationship Scale (SRS) is to provide an objective, differentiated, and functionally anchored assessment of social support systems. Prior to its construction in the early 1980s, epidemiological and psychiatric investigations routinely relied upon simplistic, global proxy measures of social ties, such as marital status, religious attendance, membership in voluntary organizations, or generalized Likert-based perceptions of support availability. Such monolithic measurements suffered from three significant theoretical and psychometric limitations: first, they confounded structural contact frequency with functional support quality; second, they assumed that all social contacts are inherently therapeutic, disregarding the documented pathogenicity of toxic, conflictual, or unhelpful social ties; and third, they obscured domain-specific variability, failing to account for situations where an individual might experience robust professional mentorship alongside profound domestic isolation.

The SRS was formulated to resolve these empirical challenges by establishing a structured mapping methodology. Clinically, the SRS serves as a diagnostic and therapeutic planning instrument capable of identifying specific interpersonal deficits, social isolation, unbalanced or unreciprocated relationships, and destructive interactions that exacerbate mental health conditions such as major depressive disorder, generalized anxiety, and schizophrenia. In clinical research, the instrument serves as an epidemiological tool for examining the mediating and moderating mechanisms linking life events, chronic strains, neuroendocrine stress responses, and somatic morbidity. By separating the helpfulness spectrum into positive, neutral, and deleterious categories, the SRS allows investigators to disentangle the stress-buffering effects of high-quality connections from the direct strain induced by maladaptive social networks.

Psychological Construct

The SRS assesses social relationship functioning through a multi-tiered conceptual matrix consisting of structural configuration, relational reciprocity, functional domain specificity, and qualitative valence.

1. Structural Configuration: Network Size and Composition

The baseline construct operationalized by the SRS is structural network size, reflecting the total count of distinct social ties mobilized by an individual when confronted with life challenges. Rather than eliciting hypothetical associations, the scale requires concrete nominations of persons (by name or initial) with whom the respondent actually discusses defined issues. The structural dimension captures relationship role diversity, categorizing ties into nuclear family members, extended kin, peers, occupational colleagues, and professional service providers.

2. Interpersonal Reciprocity

Reciprocity reflects the degree of transactional symmetry within the social dyad. Drawing upon social exchange theory, the SRS operationalizes reciprocity by inquiring whether each nominated support provider also approaches the respondent to discuss equivalent difficulties. Asymmetrical relationships—whether manifesting as chronic emotional dependence (support recipient only) or unilateral caregiving burden (support provider only)—are conceptualized as potential sources of interpersonal friction, whereas balanced reciprocity fosters psychological security, self-efficacy, and mutual trust.

3. Functional Domain Specificity

The SRS is grounded in the construct that social support is not an undifferentiated resource; rather, its mobilization is task- and context-contingent. The scale delineates six central life spheres:

• Work: Professional strains, workplace conflicts, and career decisions.
• Money and Finances: Resource allocation, financial debt, and economic security.
• Home and Family: Domestic partnerships, child-rearing responsibilities, and kinship tensions.
• Personal Health: Illness behaviors, medical diagnoses, and somatic symptoms.
• Personal Problems and Social Life: Intrapersonal distress, emotional dilemmas, and existential difficulties.
• General Crisis: Acute emergencies and catastrophic life events.

4. Qualitative Valence and Negative Interactions

A central psychometric contribution of the SRS is its bipolar rating of interaction quality. Social interactions are not presumed to be universally beneficial. The scale operationalizes support quality across a functional continuum: interactions that exacerbate psychological distress (rating 1: “makes things a lot worse” and rating 2: “makes things a bit worse”) versus interactions that attenuate distress (rating 3: “helps things a bit” and rating 4: “helps things a lot”). This prevents positive support biases and captures the pathogenic effects of invalidating, hypercritical, or unsolicited advice.

Theoretical Framework

The development of the Social Relationship Scale is grounded in three major psychological paradigms: the Cognitive-Mediational Theory of Stress and Coping formulated by Richard Lazarus and Susan Folkman, the Stress-Buffering Hypothesis developed by Sheldon Cohen and Thomas Wills, and Social Exchange Theory formulated by George Homans and Peter Blau.

Lazarus and Folkman’s Cognitive Appraisal Model

Within transactional stress theory, an environmental demand functions as a stressor only if it is appraised by the individual as taxing or exceeding available coping resources. McFarlane and colleagues incorporated this principle by structuring the SRS around cognitive appraisal: the mere presence of an individual in a social network does not dictate psychological outcomes; rather, the focal mechanism is the respondent’s subjective appraisal of whether interactions within specific domains effectively assist in resolving problems or exacerbate cognitive and emotional burden.

The Stress-Buffering Hypothesis vs. Direct Effects Model

The conceptual framework of the SRS was designed to test two alternative models of social support action:

• Direct Effects Model: Proposes that social integration is inherently protective regardless of stress levels, providing ongoing positive affect, predictability, and normative stability.
• Buffering Model: Posits that social support acts as an intermediary buffer specifically under high-stress conditions, altering the cognitive appraisal of threat, dampening neuroendocrine reactivity, and facilitating adaptive coping behaviors.

By mapping structural ties across distinct life domains and matching them against specific stressors, the SRS provides researchers with the empirical granularity needed to test whether domain-specific support buffers matching domain-specific stressors (e.g., whether work support buffers work stress, or whether generalized emotional support is required).

Social Exchange and Equity Principles

The operationalization of reciprocity within the SRS draws directly from equity theory. Dyadic social interactions involve an ongoing calculus of rewards, obligations, and investments. Imbalanced social networks, where an individual absorbs the emotional burdens of others without receiving reciprocal validation, lead to relational burnout, emotional exhaustion, and resentment. Conversely, receiving support without the ability or opportunity to return it can induce feelings of inadequacy, diminished self-efficacy, and perceived indebtedness. The SRS explicitly models this transactional balance by recording bidirectional communication across each identified dyad.

Validity

The validity of the Social Relationship Scale has been confirmed through empirical investigations encompassing psychiatric populations, primary care cohorts, and broad community samples.

Construct and Structural Validity

McFarlane et al. (1981) demonstrated the construct validity of the SRS by demonstrating that structural network metrics (total nominated individuals) and qualitative support indices (ratio of helpful to unhelpful interactions) behaved as functionally distinct constructs rather than collinear markers of a single dimension. In community samples, individuals with high neuroticism frequently reported equivalent structural network sizes to those with low neuroticism, but their ratings of interaction helpfulness were significantly lower, accompanied by elevated frequencies of negative, distress-amplifying interactions (ratings of 1 and 2).

Criterion and Predictive Validity

Predictive validity was established through longitudinal studies assessing psychological distress, depressive episodes, and physical health symptoms over follow-up periods ranging from 6 to 24 months (McFarlane et al., 1983, 1984). Multiple regression analyses revealed that baseline qualitative SRS scores (helpfulness of discussion) accounted for significant incremental variance in subsequent psychiatric distress scores (measured via the General Health Questionnaire and Hopkins Symptom Checklist) after partialing out baseline distress and life event frequency. Notably, the volume of unhelpful relationships (ratings 1 and 2) exhibited stronger standardized beta coefficients (β = .34 to .46, p < .001) in predicting somatic and depressive morbidity than positive interactions exhibited in mitigating distress (β = -.15 to -.22, p < .05), corroborating the theoretical premise that negative social exchanges exert an asymmetrical, pathogenic effect.

Convergent and Discriminant Validity

Convergent validity has been evidenced through moderate-to-strong correlations with alternative validated social support instruments, such as the Interpersonal Support Evaluation List (ISEL) (r = .52 to .68) and the Duke-UNC Functional Social Support Questionnaire (r = .58, p < .001). Discriminant validity was supported by low, statistically non-significant correlations with measures of social desirability (e.g., the Marlowe-Crowne Social Desirability Scale, r = .08, p > .10), confirming that self-reported negative interactions and network deficits are not artifacts of defensive responding or social approval biases.

Reliability

Due to the roster-generating, idiosyncratic structure of the SRS—where respondents generate variable lists of individuals across domains rather than responding to a fixed set of standardized psychometric statements—traditional internal consistency estimators such as Cronbach’s alpha are supplemented with stability and intra-class consistency metrics.

Internal Consistency

When evaluated across the domain-aggregated scores of qualitative helpfulness, the SRS demonstrates internal consistency estimates ranging from α = .74 to .85 across healthy community and primary care populations. Composite indices reflecting total network helpfulness yield Cronbach’s alpha coefficients exceeding α = .80, indicating high internal coherence among functional domain assessments.

Test-Retest Stability

Test-retest reliability was evaluated across intervals ranging from one week to three months (McFarlane et al., 1981; McDowell, 2006). The structural dimension (total count of network members identified) demonstrated high short-term stability, with Pearson correlation coefficients spanning r = .82 to .91 across a 1- to 2-week interval. Over a 3-month follow-up period, structural network stability remained robust at r = .68 to .76.

Qualitative helpfulness ratings exhibited stability coefficients of r = .70 to .83 over short intervals and r = .59 to .67 over longer observational periods. Inter-rater reliability regarding the coding of relationship classifications (e.g., nuclear kin vs. peripheral acquaintance) yielded Cohen’s kappa coefficients exceeding κ = .92, demonstrating objective categorization standards.

Factor Analysis

Statistical and structural validation of the SRS has focused on identifying whether the underlying dimensions conform to domain-specific configurations, structural versus functional dichotomies, or valence-specific (positive versus negative interaction) groupings.

Exploratory Factor Analysis (EFA)

Exploratory factor analyses using principal axis factoring with varimax and oblimin rotations on domain-level composite scores have consistently revealed a multi-factor architecture. Early structural factor investigations by Norman, Streiner, and McFarlane isolated three primary factors accounting for over 58% of the total variance:

• Factor 1: Intimate / Domestic Support: Dominated by high loadings for Home and Family interactions (.74 to .82) and Personal Problems / Emotional support (.68 to .77).
• Factor 2: Instrumental / Instrumental-Economic Support: Characterized by loadings for Financial / Money discussions (.71 to .79) and Work-related networks (.62 to .71).
• Factor 3: Somatic / Health-Specific Support: Composed primarily of health-related discussions (.81) and medical provider engagement (.65).

Confirmatory Factor Analysis (CFA)

Subsequent confirmatory structural analyses supported a bifactor or higher-order framework over a single-factor unipolar model. A model separating structural network volume from qualitative relational valence provided superior fit indices compared to a unifactorial construct:

• Root Mean Square Error of Approximation (RMSEA) = .046 (90% CI [.038, .054])
• Comparative Fit Index (CFI) = .952
• Tucker-Lewis Index (TLI) = .941
• Standardized Root Mean Square Residual (SRMR) = .041

These confirmatory findings confirm that the qualitative appraisal of interaction helpfulness operates as an autonomous psychometric dimension that is structurally distinguishable from structural network size or relational density.

Instrument / Measurement Tool

• Test Type: Structured Network-Roster Generation and Multi-Domain Rating Scale.
• Format: Grid-based self-report questionnaire or structured clinical interview protocol.
• Target Life Domains: Evaluates functional networks across life spheres, including Home and Family, Work, Money/Finances, Personal Health, and Personal Problems/Social Life.
• Primary Item Modules:
  • Network Identification: Elicitation of specific names or initials per domain.
  • Relationship Categorization: Identification of relational roles (e.g., spouse, parent, child, friend, co-worker, health professional).
  • Reciprocity Assessment: Binary indicator assessing whether the nominated person seeks support from the respondent (“Would this person come to you to discuss [domain]?” Yes = 1, No = 2).
  • Qualitative Appraisal: 4-point rating evaluating the perceived helpfulness of discussions.
• Response Scale (Helpfulness of Discussion):
  • 1 = makes things a lot worse
  • 2 = makes things a bit worse
  • 3 = helps things a bit
  • 4 = helps things a lot
• Scoring and Computational Rules:
  • Network Size Score: Count of unique individuals nominated across all domains or within a specific domain.
  • Reciprocity Ratio: Proportion of nominated contacts coded as reciprocal (Yes responses divided by total contacts).
  • Mean Domain Helpfulness: Sum of helpfulness ratings in a domain divided by the number of contacts nominated in that domain.
  • Negative Support Index: Total count or proportion of interactions assigned ratings of 1 (“makes things a lot worse”) or 2 (“makes things a bit worse”).
  • Effective Support Index: Total count or proportion of interactions assigned ratings of 3 (“helps things a bit”) or 4 (“helps things a lot”).

Permissions & Fee and Test Year

The Social Relationship Scale was introduced to the scientific literature in 1981 by Allan H. McFarlane, Kathleen A. Neale, Geoffrey R. Norman, and colleagues at McMaster University. The scale is in the public domain for academic, clinical, and non-commercial research purposes. No licensing fees or purchase royalties are required to administer the instrument in non-profit health research or psychiatric investigations. Researchers intending to adapt the scale for specialized electronic health platforms or clinical registry systems should maintain proper attribution to the original authors and cite the seminal validation publications in the Schizophrenia Bulletin (1981) and the Journal of Health and Social Behavior (1983). For historical reference or formal documentation, correspondence was originally directed to the Department of Psychiatry, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada.

References

• McDowell, I. (2006). Measuring Health: A Guide to Rating Scales and Questionnaires (3rd ed.). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195165678.001.0001
• McFarlane, A. H., Neale, K. A., Norman, G. R., Roy, R. G., & Streiner, D. L. (1981). Methodological issues in developing a scale to measure social support. Schizophrenia Bulletin, 7(1), 90–100. https://doi.org/10.1093/schbul/7.1.90
• McFarlane, A. H., Norman, G. R., Streiner, D. L., Roy, R., & Scott, D. J. (1983). The process of social stress: Stable, reciprocal, and mediating relationships. Journal of Health and Social Behavior, 24(2), 160–173. https://doi.org/10.2307/2136641
• McFarlane, A. H., Norman, G. R., Streiner, D. L., & Roy, R. (1984). Characteristics and correlates of effective and ineffective social supports. Journal of Psychosomatic Research, 28(6), 501–510. https://doi.org/10.1016/0022-3999(84)90085-5

Items of the Scale