Perceived Social Support Scale (MSPSS)

Abstract

The Multidimensional Scale of Perceived Social Support (MSPSS) is a seminal, 12-item self-report psychometric instrument designed to evaluate an individual’s cognitive appraisal of the adequacy and availability of their social support networks. Originally developed by Gregory D. Zimet, Nancy W. Dahlem, Sara G. Zimet, and Gordon K. Farley in 1988, the instrument partitions the overarching construct of perceived social support into three distinct relational domains: Family (Fam), Friends (Fri), and a Significant Other (SO). Each domain is evaluated via four balance-calibrated items captured on a 7-point Likert scale, ranging from 1 (“Very Strongly Disagree”) to 7 (“Very Strongly Agree”). Over three decades of extensive cross-cultural, epidemiological, and clinical research have established the scale’s exceptional psychometric integrity. Internal consistency reliabilities across varied adult, adolescent, clinical, and community cohorts regularly exhibit Cronbach’s alpha coefficients exceeding .85 for the subscales and .88 to .93 for the global score. Structural equation modeling and confirmatory factor analyses consistently substantiate the tripartite latent structure over unidimensional models, while convergent, discriminant, and criterion validity have been rigorously confirmed against metrics of depression, psychological distress, systemic loneliness, and quality of life. The MSPSS remains one of the most widely implemented instruments in behavioral medicine, psychiatric epidemiology, and social psychology due to its brevity, ease of administration, robust theoretical grounding in the stress-buffering paradigm, and open accessibility for research and clinical diagnostic screening.

Keywords

Multidimensional Scale of Perceived Social Support, MSPSS, perceived social support, psychometrics, stress-buffering hypothesis, family support, friend support, significant other, social isolation, clinical assessment

Authors

The Multidimensional Scale of Perceived Social Support was formulated and validated through collaborative clinical research conducted primarily at the University of Colorado Health Sciences Center and subsequent academic institutions:

• Gregory D. Zimet, Ph.D.: Professor Emeritus of Pediatrics and Clinical Psychology at the Indiana University School of Medicine. Dr. Zimet has authored hundreds of peer-reviewed articles focusing on behavioral health, adolescent health psychology, and the psycho-behavioral determinants of healthcare utilization and social adaptation.
• Nancy W. Dahlem, Ph.D.: Clinical psychologist and researcher previously affiliated with the Department of Psychiatry at the University of Colorado Health Sciences Center, contributing extensively to adolescent and young adult developmental psychology.
• Sara G. Zimet, Ed.D.: Associate Clinical Professor of Psychiatry at the University of Colorado Health Sciences Center, specializing in educational psychology, child development, and family systems research.
• Gordon K. Farley, M.D.: Professor of Psychiatry and Director of Day Treatment Services at the University of Colorado Health Sciences Center, recognized for his contributions to child and adolescent psychiatry and institutional psychotherapeutic evaluation.

Correspondence regarding original scale development and psychometric inquiries is typically directed through academic archives or the primary developer’s academic portal at Indiana University School of Medicine, Indianapolis, Indiana.

Purpose

The primary objective of the Multidimensional Scale of Perceived Social Support (MSPSS) is to provide an empirically rigorous, theoretically coherent, and time-efficient metric for assessing an individual’s subjective appraisal of their social support resources. Historically, psychological research differentiated between received or enacted support (the objective receipt of assistance or resource transfers) and perceived support (the subjective cognitive confidence that adequate support is readily accessible when needed). Psychometric and epidemiological investigations have repeatedly demonstrated that perceived social support exerts a substantially more pronounced protective effect on mental and physiological health outcomes than objective network size or counted instances of received aid. The MSPSS was engineered specifically to capture this subjective appraisal across discrete, highly salient functional sources.

In clinical psychology and psychiatric epidemiology, the MSPSS serves critical diagnostic and prognostic functions. Social isolation and perceived relational neglect represent profound transdiagnostic risk factors for major depressive disorder, generalized anxiety, suicidality, post-traumatic stress pathology, and substance use disorders. By deploying the MSPSS, clinicians can immediately discern specific deficits within a client’s relational ecosystem. For instance, an individual experiencing low family support but elevated friend support presents a vastly different clinical profile and treatment trajectory than one who feels entirely alienated across all three spheres. Such granularity enables targeted interventions, such as couples or family therapy, social skills training, or engagement in community support networks.

In behavioral medicine, public health, and psychosomatic research, the scale is routinely employed to investigate the biological and psychological pathways through which social relationships buffer against physical morbidity and mortality. Lower scores on the MSPSS have been systematically associated with dysregulated hypothalamic-pituitary-adrenal (HPA) axis functioning, elevated systemic inflammatory markers (such as interleukin-6 and C-reactive protein), cardiovascular vulnerability, and diminished adherence to complex medical regimens among patients with chronic diseases (including diabetes, cancer, cardiovascular disease, and HIV/AIDS). The operational purpose of the scale is therefore multifaceted: it functions simultaneously as an epidemiological screening device, a clinical outcome measure for evaluating therapeutic interventions, and an empirical research instrument.

Psychological Construct

The psychological construct assessed by the MSPSS is perceived social support, conceptualized as a multi-tiered, subjective cognitive schema regarding the reliability, emotional warmth, and instrumental availability of relational ties. Unlike sociometric assessments that quantify network density, frequency of contact, or structural social capital, perceived social support captures an internal cognitive representation shaped by attachment histories, cognitive processing styles, and interpersonal transactions. The MSPSS operationalizes this construct across three empirically and theoretically distinct relational dimensions:

1. Significant Other Support (SO)

The Significant Other subscale (comprising Items 1, 2, 5, and 10) assesses the perceived availability of a specific, primary relational anchor. In the construct’s initial conceptualization, the authors deliberately selected the non-restrictive designation “special person” rather than “spouse,” “romantic partner,” or “mentor.” This deliberate psychometric flexibility allows respondents across varying developmental milestones, cultural backgrounds, and marital statuses to project their primary attachment figure—whether that individual is a spouse, romantic partner, fiancé, dating companion, mentor, coach, clergy member, or close professional confidant—into the assessment framework. This dimension taps into core experiences of intimacy, comfort during acute crisis (“around when I am in need”), bidirectional emotional sharing (“share my joys and sorrows”), and the perception that one’s internal affective states are valued and validated by an essential other.

2. Family Support (Fam)

The Family subscale (comprising Items 3, 4, 8, and 11) measures the subjective adequacy of support emanating from kinship structures or primary familial units. This dimension evaluates both instrumental and affective assistance, querying whether the family actively attempts to help (“really tries to help me”), delivers necessary emotional solace (“emotional help and support”), creates a safe environment for transparent disclosure of vulnerabilities (“talk about my problems”), and demonstrates collaborative engagement in critical life choices (“willing to help me make decisions”). Familial perceived support is theoretically tied to early developmental socialization, long-term security, and structural safety nets, often displaying unique predictive associations with psychological stabilization during adolescent transitions and acute familial crises.

3. Friends Support (Fri)

The Friends subscale (comprising Items 6, 7, 9, and 12) captures perceived support originating from non-kin, egalitarian peer affiliations. Items in this subscale measure active responsiveness (“really try to help me”), reliability under adverse conditions (“count on my friends when things go wrong”), communicative intimacy (“talk about my problems”), and shared emotional resonance (“share my joys and sorrows”). Friend support reflects voluntary affiliation, horizontal social validation, and reciprocal companionship. Across adolescence and emerging adulthood, this construct frequently manifests as a decisive buffer against peer-related stressors, existential loneliness, and academic or vocational exhaustion.

Theoretical Framework

The Multidimensional Scale of Perceived Social Support is firmly situated within several foundational paradigms of health psychology, cognitive appraisal theory, and interpersonal psychology:

The Stress-Buffering Hypothesis vs. Main-Effect Model

The primary theoretical foundation of the MSPSS originates from the seminal models formulated by Sheldon Cohen and Thomas Ashby Wills (1985). Cohen and Wills posited two competing, yet complementary, theoretical models explaining how social relationships protect health: the main-effect model and the stress-buffering model. The main-effect model asserts that regular social ties provide positive affect, stability, a sense of meaning, and normative influences toward health-promoting behaviors, yielding continuous benefits regardless of the individual’s prevailing stress levels. Conversely, the stress-buffering model stipulates that social support operates primarily as an affective and functional buffer during periods of high environmental stress. Under the buffering paradigm, the subjective belief that others will mobilize resources attenuates cognitive appraisals of potential threats (primary appraisal) and enhances perceived coping capabilities (secondary appraisal), thereby dampening downstream physiological reactivity and depressive cascades. The MSPSS explicitly assesses the cognitive substrate underlying this buffering process: the psychological certainty that relational resources are standing by to mitigate adversity.

Cognitive Appraisal and Relational Schema Theory

The instrument draws conceptually on Richard Lazarus and Susan Folkman’s Transactional Model of Stress and Coping (1984). According to Lazarus and Folkman, stress is not an environmental event per se, but an ongoing relational dynamic wherein environmental demands tax or exceed an individual’s appraised resources. Perceived social support represents a generalized cognitive schema. Individuals with high perceived support approach relational exchanges with optimistic perceptual filters, interpreting ambiguous interpersonal signals as supportive and displaying higher psychological resilience. The tripartite division of the MSPSS operationalizes the notion that individuals do not assess “support” in an undifferentiated vacuum; rather, they hold separate, cognitive schemas regarding distinct relational domains.

Attachment Theory

From an attachment theory perspective (Bowlby, 1969/1982), perceived social support serves as an adult manifestation of internal working models of self and others. Secure attachment orientations foster expectations of others as trustworthy, accessible, and responsive, directly mapping onto the high-scoring end of the MSPSS. Conversely, anxious and avoidant attachment styles correlate with systematic decrements across the Family, Friends, and Significant Other subscales, reflecting persistent distrust, fear of intimacy, or hyperactivating/deactivating coping strategies.

Validity

Extensive empirical investigations spanning more than thirty years have systematically validated the MSPSS across clinical populations, psychiatric inpatients, community-dwelling elderly cohorts, military personnel, and culturally diverse international samples.

Construct and Factorial Validity

Construct validity was initially established by Zimet et al. (1988) among university undergraduates and was subsequently replicated across diverse clinical populations, including adolescent psychiatric patients (Zimet et al., 1990) and urban, low-income adolescents (Canty-Mitchell & Zimet, 2000). Across studies, construct validity has been affirmed by demonstrating that the three theoretical sources (Family, Friends, Significant Other) emerge as psychometrically distinguishable latent factors rather than a single monolithic construct. When tested via structural equation modeling, the three-factor model consistently out-indexes single-factor or two-factor permutations across goodness-of-fit parameters.

Convergent and Concurrent Validity

Convergent validity is documented through robust, statistically significant correlations with established constructs of affective well-being and distress:

• Depression and Mood Pathology: MSPSS scores demonstrate strong inverse correlations with standardized depression metrics, including the Beck Depression Inventory (BDI) (typically ranging between r = -.35 to -.55, p < .001) and the Center for Epidemiologic Studies Depression Scale (CES-D).
• Anxiety and General Distress: Negative correlations are consistently documented with the State-Trait Anxiety Inventory (STAI) and the General Health Questionnaire (GHQ-28).
• Loneliness: Strong, negative convergent correlations are observed with the UCLA Loneliness Scale (r values frequently between -.50 and -.70), demonstrating that higher perceived support directly counters subjective social isolation.
• Self-Esteem and Psychological Well-Being: The MSPSS correlates positively and robustly with the Rosenberg Self-Esteem Scale (r = .40 to .55) and subjective happiness scales.

Discriminant Validity

The scale effectively discriminates between perceived support and distinct constructs such as social desirability, external locus of control, or general neuroticism. When partialing out trait negative affectivity, the protective associations of the MSPSS on psychological symptomatology remain statistically significant, confirming that the instrument does not merely index low neuroticism or general optimism, but captures distinct relational variance.

Criterion and Predictive Validity

Prospective longitudinal designs have repeatedly shown that baseline MSPSS scores prospectively predict lower postpartum depression rates, improved recovery trajectories following acute coronary events, enhanced treatment adherence among oncology patients, and lower attrition in substance abuse rehabilitation programs. These longitudinal outcomes firmly corroborate the predictive utility of the instrument in medical and psychotherapeutic environments.

Reliability

The psychometric reliability of the MSPSS has been confirmed across hundreds of independent empirical investigations worldwide, exhibiting high internal consistency and exceptional temporal stability.

Internal Consistency

In the original validation study conducted by Zimet et al. (1988), Cronbach’s alpha coefficients reflected high internal consistency across all subscales and the aggregate measure:

• Total Scale: α = .88
• Significant Other Subscale: α = .91
• Family Subscale: α = .87
• Friends Subscale: α = .85

In their follow-up validation study involving adolescent psychiatric inpatients, Zimet, Powell, Farley, Werkman, and Berkoff (1990) documented comparable reliability indices: Total Scale α = .86; Significant Other α = .83; Family α = .87; and Friends α = .89. Similarly, Canty-Mitchell and Zimet (2000) reported alphas of .93 for the global scale, .91 for Significant Other, .87 for Family, and .89 for Friends in a multi-ethnic urban adolescent cohort. International adaptations (including Spanish, Turkish, Chinese, Arabic, French, and Urdu translations) systematically report Cronbach’s alphas and McDonald’s omegas ranging from .82 to .94, confirming that the scale’s structural cohesion is preserved across linguistic and cultural contexts.

Test-Retest Stability

Test-retest reliability was originally evaluated across a 2- to 3-month latency period among undergraduate students (Zimet et al., 1988). The resulting Pearson product-moment stability coefficients demonstrated strong temporal endurance:

• Significant Other Subscale: r = .72
• Family Subscale: r = .85
• Friends Subscale: r = .75
• Total Score: r = .85

These findings illustrate that while perceived social support fluctuates sensibly in response to major interpersonal ruptures or therapeutic interventions, it operates as a remarkably stable cognitive-evaluative trait in non-transitional daily living.

Factor Analysis

The internal architecture of the MSPSS has been evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA), serving as an exemplary benchmark in structural equation modeling literature.

Exploratory Factor Analysis (EFA)

In the seminal 1988 study, Zimet and colleagues subjected the 12 items to an exploratory principal components analysis followed by orthogonal (Varimax) and oblique (Promax) rotations. Three clear factors emerged with eigenvalues substantially exceeding 1.0 (specifically: 4.88 for Factor 1, 2.70 for Factor 2, and 1.39 for Factor 3), collectively accounting for 74.8% to 80% of the total variance. Factor loadings were clean and pronounced:

• Factor 1 (Significant Other): Items 1, 2, 5, and 10 loaded between .78 and .91 onto their primary construct, with negligible cross-loadings (all < .25).
• Factor 2 (Family): Items 3, 4, 8, and 11 loaded between .74 and .90 onto the familial construct.
• Factor 3 (Friends): Items 6, 7, 9, and 12 loaded between .71 and .87 onto the friendship construct.

Confirmatory Factor Analysis (CFA)

Subsequent contemporary structural validation studies have extensively contrasted multiple competing structural models via Confirmatory Factor Analysis (CFA):

• Model 1: Unidimensional Model. Assumes all 12 items reflect a single, homogeneous latent support factor. Across all published empirical CFA studies, this model exhibits extremely poor fit (e.g., χ²/df > 8.0, RMSEA > .15, CFI < .80), confirming that perceived social support cannot be reduced to a single monolithic dimension.
• Model 2: Two-Factor Model. Combines Friends and Significant Other into an “extrafamilial” factor contrasted against Family. While fitting better than the unidimensional model, it remains structurally inferior.
• Model 3: Tripartite Correlated First-Order Model (Original Model). Specifies three distinct, correlated latent factors (Family, Friends, Significant Other). This model consistently exhibits superior, exceptional fit across cultural settings: Root Mean Square Error of Approximation (RMSEA) < .05, Comparative Fit Index (CFI) > .97, Tucker-Lewis Index (TLI) > .96, and Standardized Root Mean Square Residual (SRMR) < .04.
• Model 4: Hierarchical / Second-Order Model. Specifies three first-order factors that load onto a overarching global Perceived Social Support factor. This model also demonstrates acceptable fit, providing psychometric justification for calculating both individual subscale means and an omnibus global mean score.

Instrument / Measurement Tool

The Multidimensional Scale of Perceived Social Support is structured as follows:

• Instrument Designation: Multidimensional Scale of Perceived Social Support (MSPSS).
• Construct Assessed: Subjective cognitive appraisal of social support availability and adequacy across three interpersonal sources.
• Administration Modality: Self-administered pencil-and-paper questionnaire, interactive digital assessment, or structured clinical interview.
• Item Inventory: 12 concise declarative statements.
• Subscale Breakdown:
  • Significant Other Subscale: Items 1, 2, 5, and 10.
  • Family Subscale: Items 3, 4, 8, and 11.
  • Friends Subscale: Items 6, 7, 9, and 12.
• Authentic Response Scale: 7-point Likert scale formatted with the following anchors:
  • 1 = Very Strongly Disagree
  • 2 = Strongly Disagree
  • 3 = Mildly Disagree
  • 4 = Neutral
  • 5 = Mildly Agree
  • 6 = Strongly Agree
  • 7 = Very Strongly Agree
• Completion Latency: Approximately 3 to 5 minutes.
• Target Population: Adolescents (ages 12 and older) and adults across clinical, medical, and community demographics.
• Scoring and Computational Rules:
  • There are no reverse-scored items on the MSPSS; all items are keyed in a positive direction.
  • Family Subscale Mean: Sum scores for items 3, 4, 8, and 11, then divide by 4.
  • Friends Subscale Mean: Sum scores for items 6, 7, 9, and 12, then divide by 4.
  • Significant Other Subscale Mean: Sum scores for items 1, 2, 5, and 10, then divide by 4.
  • Total Perceived Support Mean: Sum scores for all 12 items, then divide by 12.
• Score Interpretation Guidelines: While Zimet and colleagues emphasize continuous scaling in inferential statistics, empirical research commonly interprets mean scores as follows:
  • Mean 1.0 to 2.9: Low perceived social support (indicates perceived interpersonal neglect, alienation, or severe social vulnerability).
  • Mean 3.0 to 5.0: Moderate perceived social support.
  • Mean 5.1 to 7.0: High perceived social support (reflects substantial subjective interpersonal integration and cognitive reassurance of relational availability).

Permissions & Fee and Test Year

The Multidimensional Scale of Perceived Social Support was initially published in 1988 in the Journal of Personality Assessment. In keeping with the scholarly philosophy of Dr. Gregory D. Zimet and his co-authors, the MSPSS is placed in the public academic domain for non-commercial educational, clinical, and scholarly research purposes. No royalties, licensing fees, or formal prior permissions are required for researchers or clinicians to utilize, reproduce, adapt into digital survey software, or translate the instrument, provided that standard academic attribution is preserved and appropriate citations are made to the original 1988 validation publication. Investigators utilizing the scale in formal investigations are expected to reference Zimet et al. (1988) in all resulting reports, manuscripts, and presentations.

References

• Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). Basic Books. (Original work published 1969).
• Canty-Mitchell, J., & Zimet, G. D. (2000). Psychometric properties of the Multidimensional Scale of Perceived Social Support in urban adolescents. American Journal of Community Psychology, 28(3), 391–400. https://doi.org/10.1023/A:1005109522457
• Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357. https://doi.org/10.1037/0033-2909.98.2.310
• Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer Publishing Company.
• Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 52(1), 30–41. https://doi.org/10.1207/s15327752jpa5201_2
• Zimet, G. D., Powell, S. S., Farley, G. K., Werkman, S., & Berkoff, K. A. (1990). Psychometric characteristics of the Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 55(3-4), 610–617. https://doi.org/10.1080/00223891.1990.9674100

13. Items of the Scale (Questionnaire)