Social Support List – Interactions / Social Support List – Discrepancies / Social Support List – Interactions-12

1. Abstract

The Social Support List (in Dutch: Sociale Steun Lijst; comprising the Social Support List – Interactions [SSL-I], the Social Support List – Discrepancies [SSL-D], and the abbreviated 12-item version, the Social Support List – Interactions-12 [SSL-I-12]) represents one of the most rigorously developed psychometric systems designed to operationalize and evaluate received social support and support discrepancies in adults and older populations. Originally developed by Eric van Sonderen (1991) at the University of Groningen and subsequently adapted into a parsimonious brief form by L. M. van Eijk, Gertrudis I. J. M. Kempen, and F. L. P. van Sonderen (1993), the instrument framework distinguishes between the structural enactment of supportive behaviors and the cognitive-evaluative appraisal of whether such interactions fulfill the recipient’s subjective needs. The full-length SSL-I consists of 34 supportive interaction items alongside 7 negative interaction items, measuring behaviors across diverse ecological domains. The complementary SSL-D mirrors these 34 items to assess the extent to which received support diverges from individual expectations, capturing deficits or surpluses in interpersonal assistance. To minimize respondent burden in large-scale cohort studies such as the Groningen Longitudinal Aging Study (GLAS), the abbreviated SSL-I-12 was constructed. The SSL-I-12 condenses supportive transactions into 12 core items across three distinct theoretical dimensions: Everyday social support (Alledaagse steun; 4 items), Support in times of trouble (Steun bij problemen; 4 items), and Esteem support (Waarderingssteun; 4 items). Responses are captured using an authentic 4-point Likert scale ranging from 1 (“Zelden of nooit” / Rarely or never) to 4 (“Vaak” / Often). The total SSL-I-12 score ranges from 12 to 48, with higher scores reflecting greater frequencies of supportive transactions. Psychometric evaluations across diverse community-dwelling, clinical, and geriatric cohorts confirm exceptional structural validity, high internal consistency (Cronbach’s alpha typically spanning 0.76 to 0.88 for subscales and exceeding 0.85 for the total scale), and robust convergent validity with indices of psychological well-being, life satisfaction, and reduced depressive symptomatology.

2. Keywords

Social Support List, SSL-I, SSL-D, SSL-I-12, Sociale Steun Lijst, received social support, everyday social support, esteem support, support discrepancy, psychometrics, geriatric assessment, interpersonal interactions, Groningen Longitudinal Aging Study.

3. Authors

The Social Support List family of instruments was conceptualized, validated, and refined by researchers affiliated with the Northern Centre for Healthcare Research (NCH) and the Department of Health Sciences at the University of Groningen, The Netherlands:

• Eric (F. L. P.) van Sonderen, Ph.D.: Primary developer of the original Sociale Steun Lijst: Interacties (SSL-I) and Discrepanties (SSL-D) (1991). Research Methodologist and Psychometrician at the Northern Centre for Healthcare Research (NCH) and the Department of Health Sciences, University Medical Center Groningen (UMCG), University of Groningen. Dr. van Sonderen’s research focused on psychometric measurement, social network theory, longitudinal survey methodologies, and the psychological mechanisms mediating social interactions and physical health outcomes.
• L. M. van Eijk, M.Sc. / Ph.D.: Lead investigator in the empirical reduction, psychometric validation, and longitudinal modeling of the abbreviated 12-item inventory (SSL-I-12; 1993), focusing on aging research, functional status, and community health dynamics.
• Gertrudis I. J. M. (Jos) Kempen, Ph.D.: Professor of Health Services Research and Social Gerontology at Maastricht University and previously senior investigator at the University of Groningen. Dr. Kempen is recognized internationally for his extensive contributions to geriatric assessment, quality of life measurement, activities of daily living (ADL) scales, and longitudinal aging cohorts.

4. Purpose

The primary purpose of the Social Support List instrument battery—encompassing the comprehensive SSL-I, the evaluative SSL-D, and the abbreviated SSL-I-12—is to provide an empirically grounded, methodologically robust operationalization of interpersonal support behaviors. In clinical, psychological, and epidemiological settings, social support is frequently conceptualized in an ambiguous manner, often blurring the boundaries between network size, perceived support availability, behavioral enactment of support, and the recipient’s subjective satisfaction. The SSL instruments were constructed to systematically disaggregate these constructs into measurable, concrete behavioral units.

The SSL-I was designed to quantify the specific behavioral frequency of received supportive interactions occurring within the respondent’s primary social network over a designated reference period (typically the preceding four weeks). Rather than asking abstract questions regarding whether the respondent feels supported, the instrument solicits concrete behavioral observations, such as whether persons in the respondent’s social ecology dropped by informally, assisted with domestic chores, or provided comforting words during distress. This behavioral orientation minimizes abstract cognitive heuristics and memory distortions, yielding an authentic reflection of enacted social transactions.

Concurrently, the development of the SSL-D served a critical theoretical imperative: the assessment of support discrepancies. Drawing on social exchange theory and cognitive discrepancy models, the SSL-D measures the psychological gap between the actual frequency of received supportive interactions and the individual’s subjective desire or expectation for such interactions. By evaluating whether received support is perceived as insufficient, just right, or excessive, the SSL-D allows clinicians and researchers to capture relational deficits (under-support) as well as intrusive or autonomy-undermining transactions (over-support), both of which are predictive of psychological distress.

The abbreviated SSL-I-12 was formulated to address practical constraints inherent to geriatric epidemiology, oncology, chronic disease management, and large population surveys. In these research contexts, extensive test batteries induce respondent fatigue, cognitive overload, and attrition. By paring down the initial pool of items to 12 psychometrically pristine indicators representing three primary transactional domains—everyday companionship, problem-focused assistance, and esteem validation—the SSL-I-12 preserves the structural validity of the parent instrument while optimizing completion feasibility. The instrument serves vital functions in identifying socially isolated older adults, monitoring the efficacy of community care interventions, quantifying informal caregiving exchanges, and evaluating the protective buffering effects of social environments against somatic decline, chronic disability, and late-life depression.

5. Psychological Construct

The psychological construct underlying the Social Support List framework is rooted in multidimensional models of interpersonal relations and functional social support. Rather than viewing social support as a monolithic global asset, the SSL framework posits that supportive interactions represent distinct behavioral transactions serving differentiated psychological, instrumental, and emotional functions. Within the abbreviated SSL-I-12, the construct is formally delineated across three distinct first-order dimensions, alongside a unified higher-order received support construct:

Everyday Social Support (Alledaagse Steun)

Everyday social support, comprising items 1 through 4, operationalizes informal, low-threshold, non-crisis social transactions embedded in daily living. This dimension reflects spontaneous interpersonal contact, informal socializing, and ongoing social companionship. Unlike problem-focused interventions that are triggered by negative life events or somatic impairment, everyday support occurs in the absence of crisis, signifying normative integration into an interpersonal network. Examples include having someone stop by casually for a chat (item 1), exhibiting unsolicited personal interest (item 2), placing an informal telephone call (item 3), or extending an invitation to engage in social outings (item 4). The psychological significance of everyday support lies in its capacity to cultivate a stable sense of belonging, reduce baseline feelings of loneliness, and establish regular relational routines that affirm an individual’s membership in a shared community.

Support in Times of Trouble (Steun bij Problemen)

Support in times of trouble, captured by items 5 through 8, encompasses responsive instrumental and emotional coping assistance mobilized in reaction to acute or chronic stressors. This dimension represents the classic buffer construct within stress-coping theory, delineating transactional assistance provided when the individual faces physical incapacity, acute illness, emotional strain, or complex decision-making. Specific behavioral manifestations include practical physical assistance with heavy household tasks or chores (item 5), emotional availability providing a safe interpersonal space for cathartic crying and disclosure (item 6), concrete instrumental care during periods of illness such as preparing meals or purchasing groceries (item 7), and informational guidance or advice when navigating intricate problems (item 8). This subscale gauges the responsiveness and functional utility of the respondent’s primary safety net during periods of heightened vulnerability.

Esteem Support (Waarderingssteun)

Esteem support (also referred to as appraisal or emotional validation support), measured across items 9 through 12, operationalizes interpersonal communications that reinforce an individual’s sense of personal worth, dignity, self-competence, and identity. Esteem support addresses core human needs for positive social regard and self-worth. Behaviors defining this construct include explicit expressions that convey genuine appreciation (item 9), social acceptance that validates the individual unconditionally as they are (item 10), spontaneous compliments regarding character, appearance, or accomplishments (item 11), and nonverbal or verbal gestures of encouragement such as a pat on the back or reassurance (“steuntje in de rug”, item 12). Esteem support serves as a potent psychological antidote to self-stigmatization, depressive demoralization, and age-related declines in perceived self-efficacy.

Support Discrepancy Construct (SSL-D Dynamics)

Complementing these behavioral dimensions, the discrepancy construct measured by the SSL-D models the cognitive-evaluative interface between interpersonal input and subjective need. Social support interactions do not possess uniform psychological utility; when support interactions occur at frequencies below an individual’s internal standard, feelings of abandonment, neglect, and unmet dependence emerge. Conversely, when support transactions—particularly instrumental assistance—occur at frequencies that exceed the individual’s desire, they may induce psychological reactance, perceived loss of autonomy, and feelings of infantilization. The SSL-D explicitly models this evaluative axis, allowing investigators to distinguish absolute behavioral deficits from subjective appraisal deficits.

6. Theoretical Framework

The architecture of the SSL instruments is anchored in several prominent sociological, cognitive, and health psychology paradigms, most notably the Stress-Buffering Model articulated by Sheldon Cohen and Thomas A. Wills (1985), Social Exchange Theory (Blau, 1964; Homans, 1958), and the cognitive appraisal frameworks formulated by Richard Lazarus and Susan Folkman (1984).

The Stress-Buffering vs. Main-Effect Paradigms

A foundational theoretical distinction in social epidemiology is whether social support acts via a main-effect model (conferring direct health and psychological benefits regardless of stress level) or a buffering model (protecting individuals selectively against the pathogenic consequences of acute stress). The SSL battery was explicitly constructed to capture both mechanisms:

• Main-Effect Alignment: The Everyday social support subscale directly reflects the main-effect hypothesis. Regular, positive social integration and informal companionship sustain positive emotional states, encourage normative health-promoting behaviors, and provide a neurobiological sense of social safety and predictability, even when explicit life stressors are absent.
• Stress-Buffering Alignment: The Support in times of trouble subscale maps directly onto the stress-buffering mechanism. When an individual confronts a somatic, environmental, or psychological stressor, received instrumental and emotional aid mitigates primary appraisal (the perceived severity of the threat) and enhances secondary appraisal (the perceived availability of resources to cope with the demand), attenuating downstream neuroendocrine and emotional dysregulation.

Cognitive Appraisal and Discrepancy Theory

The inclusion of the SSL-D builds upon cognitive appraisal theory. Lazarus and Folkman emphasized that objective environmental occurrences do not dictate psychological outcomes; rather, subjective cognitive appraisals determine emotional responses. Van Sonderen integrated this principle into psychometrics by recognizing that counting behavioral interactions (SSL-I) accounts for only half of the psychological equation. By pairing enacted behavioral frequencies with subjective discrepancy evaluations (SSL-D), the measurement framework captures how an individual cognitively processes their relational environment. If a recipient views received help as a sign of personal incompetence or an unsolicited intrusion upon their autonomy, identical objective behaviors that alleviate distress in one person may exacerbate distress in another.

Social Exchange and Equity Principles

The theoretical framework also acknowledges social exchange dynamics. Human relationships are characterized by norms of reciprocity. In older age or during chronic illness, individuals may find themselves in chronically asymmetrical relationships where they receive high volumes of instrumental care but possess limited avenues to reciprocate. Such asymmetry can generate feelings of indebtedness, guilt, and reduced self-esteem. The SSL-I-12 captures esteem support alongside instrumental support, allowing researchers to evaluate whether relational partners provide psychological affirmations that preserve the recipient’s dignity despite structural dependencies in everyday functioning.

7. Validity

The psychometric validity of the SSL-I, SSL-D, and SSL-I-12 has been extensively demonstrated through classic and modern psychometric methodologies within community, clinical, and geriatric populations.

Construct and Structural Validity

During the foundational development of the SSL-I by van Sonderen (1991), construct validity was confirmed through detailed analyses of relational transactions across multiple cohort panels. The items demonstrated strong associations with network density, structural social network indices, and self-reported relational satisfaction. For the abbreviated SSL-I-12, van Eijk, Kempen, and van Sonderen (1993) demonstrated that a three-factor latent model (Everyday support, Support in times of trouble, Esteem support) accounted for the inter-item variance far superiorly to a unidimensional model, while still maintaining sufficient shared covariance to justify a higher-order composite score of overall received support.

Convergent Validity

Convergent validity has been established across numerous epidemiological investigations, notably the Groningen Longitudinal Aging Study (GLAS), which tracked over 3,000 community-dwelling older adults. SSL-I-12 subscale and total scores exhibit moderate-to-strong positive correlations with measures of life satisfaction, subjective health appraisal, and positive affect as measured by the PANAS (Pearson $r$ values typically ranging from 0.35 to 0.52, $p < 0.001$). Furthermore, the SSL-I-12 correlates robustly with alternative social support metrics, including the Pearlin Support Scale and the Cantril Ladder of life satisfaction.

Discriminant Validity

Discriminant validity is confirmed by the scale’s empirical distinction from loneliness and general psychopathology. Although received support is inversely associated with loneliness as measured by the De Jong Gierveld Loneliness Scale, the correlation is moderate (typically between $r = -0.30$ and $-0.45$), confirming that perceived loneliness (an emotional state of subjective deficit) and received support interactions (the behavioral occurrence of social transactions) are distinct psychological phenomena. Older adults residing in supportive care environments may report high SSL-I-12 interaction frequencies while simultaneously experiencing existential loneliness, an empirical distinction that the SSL framework preserves. Additionally, correlations between the SSL-I-12 and generalized neuroticism or cognitive impairment scores are low ($r < 0.20$), indicating that the scale does not merely mirror generalized negative affectivity or cognitive status.

Predictive and Criterion Validity

Longitudinal studies demonstrate that baseline SSL-I-12 scores significantly predict psychological adjustment following negative life events, including bereavement, the onset of physical functional limitations, and acute cardiovascular incidents. High scores on Everyday social support and Esteem support prospectively buffer older adults against increases in depressive symptoms measured via the CES-D over follow-up periods spanning one to three years. In contrast, elevated SSL-D discrepancy scores (indicating unfulfilled support desires) reliably predict longitudinal declines in subjective health status and increased primary healthcare utilization.

8. Reliability

The reliability of the Social Support List instruments has been scrutinized using classical test theory metrics, item-total correlations, and test-retest stability coefficients across diverse research programs.

Internal Consistency

In the original validation of the full 34-item SSL-I, van Sonderen (1991) reported high Cronbach’s alpha coefficients, typically exceeding 0.90 for the aggregated supportive interaction scale, with domain-specific subscales ranging between 0.70 and 0.85. In the validation of the abbreviated SSL-I-12, van Eijk, Kempen, and van Sonderen (1993) documented the following internal consistency values in community samples of older adults ($N > 3,000$):

• Everyday social support (Items 1–4): Cronbach’s $\alpha = 0.76 \text{ to } 0.82$
• Support in times of trouble (Items 5–8): Cronbach’s $\alpha = 0.78 \text{ to } 0.84$
• Esteem support (Items 9–12): Cronbach’s $\alpha = 0.75 \text{ to } 0.81$
• Total SSL-I-12 Scale (12 items): Cronbach’s $\alpha = 0.85 \text{ to } 0.88$

Corrected item-total correlations across all 12 items uniformly exceed the psychometric threshold of 0.40, with most indicators falling between 0.48 and 0.65. This demonstrates that despite substantial data reduction from the original 34-item pool, each four-item subscale retains high conceptual cohesion and minimal measurement error.

Test-Retest Stability

Test-retest stability has been evaluated across varying temporal intervals. Over a stable two-to-four-week test-retest window in non-crisis community-dwelling adults, the intra-class correlation coefficients (ICC) and stability coefficients for the SSL-I-12 total score consistently fall between $r = 0.78$ and $r = 0.85$, demonstrating high temporal reproducibility. Over longer longitudinal intervals (e.g., 12 to 24 months in the GLAS cohort), stability coefficients moderate to $r \approx 0.50 – 0.60$. This moderate long-term stability is theoretically anticipated, reflecting dynamic shifts in social networks, changes in physical functional ability, and natural transitions in informal care arrangements over time.

9. Factor Analysis

The dimensional architecture of the SSL instruments has been evaluated using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) within large empirical databases.

Exploratory Factor Analyses of the Parent Scale

During the initial development of the full SSL-I (van Sonderen, 1991), principal component analyses with oblimin and varimax rotations identified six primary empirical domains: everyday affective interactions, emotional support in problems, appraisal/esteem interactions, instrumental assistance, informational assistance, and social companionship. However, high inter-factor correlations among instrumental and informational dimensions, alongside high cross-loadings in smaller sub-domains, prompted the search for a more refined, structurally stable operationalization.

Confirmatory Factor Modeling of the SSL-I-12

When deriving the SSL-I-12, van Eijk, Kempen, and van Sonderen (1993) subjected candidate items from the primary domains to rigorous CFA procedures using maximum likelihood estimation. Several competitive structural models were tested against the empirical data:

• Single-Factor Model: A global model where all 12 items loaded onto an undifferentiated ‘received support’ factor. This model demonstrated poor-to-mediocre fit ($\chi^2 / df > 8.0$, RMSEA $\approx 0.092$, CFI $\approx 0.84$), indicating that received support cannot be treated as a purely unidimensional behavioral entity.
• Uncorrelated Three-Factor Model: A model specifying three orthogonal dimensions (Everyday, Problem, Esteem). Fit indices were unacceptable, rejecting the premise that these supportive interaction forms operate independently in personal life.
• Correlated Three-Factor Model: A model specifying three distinct, correlated latent factors: Everyday Social Support (items 1–4), Support in Times of Trouble (items 5–8), and Esteem Support (items 9–12). This model demonstrated superior fit across diverse age cohorts: Comparative Fit Index ($ ext{CFI}) ge 0.95$, Tucker-Lewis Index ($ ext{TLI}) ge 0.94$, Root Mean Square Error of Approximation ($ ext{RMSEA}) le 0.048$ (90% CI [0.042, 0.054]), and Standardized Root Mean Square Residual ($ ext{SRMR}) le 0.040$.
• Higher-Order Hierarchical Model: A second-order model where the three first-order latent factors loaded onto a general second-order factor of “Total Received Social Support” yielded virtually identical fit to the correlated three-factor model. This empirical finding provides statistical justification for researchers calculating both the three specific subscale scores and a single composite summary score.

Factor Loadings

Standardized factor loadings for the items onto their respective latent factors in community-dwelling older samples are uniformly high:

• Everyday Support: Item 1 ($lambda = 0.68$), Item 2 ($lambda = 0.74$), Item 3 ($lambda = 0.71$), Item 4 ($lambda = 0.65$).
• Support in Times of Trouble: Item 5 ($lambda = 0.66$), Item 6 ($lambda = 0.72$), Item 7 ($lambda = 0.78$), Item 8 ($lambda = 0.69$).
• Esteem Support: Item 9 ($lambda = 0.75$), Item 10 ($lambda = 0.77$), Item 11 ($lambda = 0.70$), Item 12 ($lambda = 0.67$).

Inter-factor correlations between the three latent dimensions typically range between $r = 0.45$ and $r = 0.65$, confirming that while these domains co-occur within supportive relationships, they preserve sufficient distinct variance to warrant discrete measurement and subscale profiling.

10. Instrument / Measurement Tool

Below is the technical specification and administrative profile of the Social Support List instruments, with a primary focus on the standardized 12-item inventory (SSL-I-12):

• Test Name: Social Support List – Interactions-12 (Dutch: Sociale Steun Lijst – Interacties-12; Acronym: SSL-I-12). Closely associated with the full 34-item Social Support List – Interactions (SSL-I) and the 34-item Social Support List – Discrepancies (SSL-D).
• Target Population: Community-dwelling adults, older adults (geriatric populations), individuals with chronic somatic diseases, psychiatric outpatients, and informal care recipients.
• Instrument Type: Standardized self-report rating inventory (can also be administered as a structured face-to-face or telephone interview for individuals with visual or motor limitations).
• Item Count:
  • SSL-I-12: 12 behavioral interaction items.
  • SSL-I (Full): 34 supportive interaction items + 7 negative interaction items (41 items total).
  • SSL-D (Full): 34 discrepancy-evaluative items.
• Subscales & Dimensional Structure (SSL-I-12):
  • Everyday social support (Alledaagse steun): Items 1, 2, 3, 4 (4 items). Evaluates casual contact, spontaneous interest, calls, and invitations.
  • Support in times of trouble (Steun bij problemen): Items 5, 6, 7, 8 (4 items). Evaluates instrumental help with chores, emotional comfort when crying, sick care, and practical advice.
  • Esteem support (Waarderingssteun): Items 9, 10, 11, 12 (4 items). Evaluates perceived appreciation, unconditional acceptance, compliments, and behavioral encouragement.
• Response Scale (SSL-I-12 Mandatory Format):
  A 4-point Likert scale assessing behavioral frequency over the past four weeks:
  • 1 = Zelden of nooit (Rarely or never)
  • 2 = Soms (Sometimes)
  • 3 = Regelmatig (Regularly)
  • 4 = Vaak (Often)
• Scoring Rules:
  • Item Scoring: All items are scored directly on the 1–4 integer continuum. There is no reverse scoring on the SSL-I-12; all items are positively keyed toward supportive interaction frequency.
  • Subscale Scores: Calculated by summing the 4 constituent items for each dimension:
    • Everyday support: Sum of Items 1, 2, 3, 4 (Score range: 4 to 16).
    • Support in times of trouble: Sum of Items 5, 6, 7, 8 (Score range: 4 to 16).
    • Esteem support: Sum of Items 9, 10, 11, 12 (Score range: 4 to 16).
  • Overall Social Support Score: Calculated by summing all 12 items (Score range: 12 to 48).
  • Score Interpretation: Higher scores denote a higher frequency of received supportive interactions within the respondent’s social environment. Clinical or epidemiological cutoffs typically utilize population-based quintiles or standard deviation bands (e.g., scores below 1 SD from the age-stratified mean indicate marked relational vulnerability or social isolation).
  • Missing Data Handling: If a respondent omits a single item within a 4-item subscale, mean imputation based on the remaining 3 items of that subscale is methodologically permissible. If more than one item in a subscale is missing, that subscale score should be treated as missing.
• Administration Time: Approximately 3 to 5 minutes for the SSL-I-12; approximately 15 to 20 minutes for the full SSL-I / SSL-D battery.

11. Permissions & Fee and Test Year

The original conceptualization and empirical validation of the Social Support List battery were published across two foundational milestones:

• 1991: Publication of the comprehensive Sociale Steun Lijst: Interacties (SSL-I) en Discrepanties (SSL-D) research manual and dissertation by Eric van Sonderen at the Northern Centre for Healthcare Research (NCH), University of Groningen.
• 1993: Development and validation of the abbreviated 12-item version (SSL-I-12) by L. M. van Eijk, Gertrudis I. J. M. Kempen, and F. L. P. van Sonderen, designed primarily for the Groningen Longitudinal Aging Study (GLAS).

Licensing and Accessibility: The SSL instruments (SSL-I, SSL-D, and SSL-I-12) reside in the public academic domain for non-commercial scientific research and clinical health assessment. No royalty fees or formal commercial test user licenses are required to administer the scale for academic, clinical trial, or non-profit health monitoring purposes. Researchers and practitioners are expected to cite the foundational validation publications (van Sonderen, 1991; van Eijk et al., 1993; Kempen & van Eijk, 1995) in all dissemination materials. Commercial distribution, proprietary digital repackaging, or inclusion within fee-for-service enterprise health software typically requires written consultation with the copyright holders and the Department of Health Sciences at the University Medical Center Groningen.

12. References

• 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
• De Jong Gierveld, J., & Kamphuis, F. (1985). The development of a Rasch-type loneliness scale. Applied Psychological Measurement, 9(3), 289–299. https://doi.org/10.1177/014662168500900307
• Kempen, G. I. J. M., & van Eijk, L. M. (1995). The psychometric properties of the SSL12-I: A short scale for measuring social support in the elderly. Social Indicators Research, 35(3), 303–312. https://doi.org/10.1007/BF01079025
• Lazarus, R. S., & Folkman, S. (1984). Stress, Appraisal, and Coping. Springer Publishing Company.
• Ormel, J., Kempen, G. I. J. M., Deeg, D. J. H., Brilman, E. I., van Sonderen, E., & Relyveld, J. (1998). Quality of life and social production functions: A framework for understanding health-related quality of life in elderly people. Social Science & Medicine, 47(5), 613–626. https://doi.org/10.1016/S0277-9536(98)00124-9
• van Eijk, L. M., Kempen, G. I. J. M., & van Sonderen, F. L. P. (1993). Een korte schaal voor het meten van sociale steun bij ouderen: De SSL-I-12. Tijdschrift voor Gerontologie en Geriatrie, 24(5), 199–206.
• van Eijk, L. M., Kempen, G. I. J. M., & van Sonderen, F. L. P. (1994). The measurement of received social support in the elderly: Construct validity and stability of the SSL-I-12. In G. I. J. M. Kempen (Ed.), Health and Well-being in the Elderly: Psychosocial Aspects and Interventions (pp. 89–104). Van Gorcum.
• van Sonderen, E. (1991). Het meten van sociale steun: Handleiding bij de Sociale Steun Lijst – Interacties (SSL-I) en Sociale Steun Lijst – Discrepanties (SSL-D). Noordelijk Centrum voor Gezondheidsvraagstukken (NCH), Rijksuniversiteit Groningen.
• van Sonderen, E. (1993). Measuring social support: A manual for the Social Support List – Interactions (SSL-I) and Social Support List – Discrepancies (SSL-D). Northern Centre for Healthcare Research, University of Groningen.

13. Items of the Scale