Differential Loneliness Scale – short student version

Abstract

The Differential Loneliness Scale – Short Student Version (DLS-SV) is a psychometric instrument designed to evaluate multidimensional, domain-specific experiences of subjective social isolation among post-secondary and university student populations. Originally developed by Neena Schmidt and Vello Sermat in 1983 as a 60-item comprehensive assessment and subsequently adapted into an efficient 20-item student-specific inventory, the DLS-SV addresses a fundamental limitation in psychometric research: the historical tendency to treat loneliness as an undifferentiated, unidimensional affective state. Grounded in the cognitive-discrepancy and relational deficit models of loneliness, the instrument posits that perceived relational deprivation is qualitatively distinct across different interpersonal ecosystems.

The DLS-SV assesses relational deficits across four discrete subscales: Romantic/Sexual Relationships (5 items), Friendships (5 items), Family Relationships (5 items), and Larger Community/Group Affiliations (5 items). Utilizing a forced-choice dichotomous response format (True/False) with balanced keying to mitigate acquiescence bias, the measure yields both subscale-specific deficit scores (0–5) and an aggregate differential loneliness index (0–20). Psychometric evaluations consistently demonstrate robust internal consistency, with total scale Cronbach’s alpha coefficients typically ranging from .85 to .90, and subscale coefficients spanning .72 to .86. Confirmatory factor analyses support a four-factor oblique structural model that demonstrates superior fit compared to unitary construct formulations. The scale exhibits strong convergent validity with general loneliness measures, such as the UCLA Loneliness Scale, while demonstrating exceptional discriminant validity by disentangling depressive symptomatology, social anxiety, and generalized introversion from domain-targeted relational shortfalls. Consequently, the DLS-SV serves as an indispensable diagnostic and empirical tool within higher education counselling centres, developmental psychopathology, and social psychological research.

Keywords

Differential Loneliness Scale, DLS-SV, multidimensional loneliness, relational deficits, student mental health, psychometrics, romantic loneliness, social isolation, community belonging, Schmidt and Sermat

Authors

The Differential Loneliness Scale and its specialized variants were conceptualized and validated by:

• Neena Schmidt, Ph.D. — Department of Psychology, York University, Toronto, Ontario, Canada. Dr. Schmidt led foundational research into multidimensional relationship expectations and relational deficits across developmental transitions.
• Vello Sermat, Ph.D. — Professor Emeritus of Psychology, York University, Toronto, Ontario, Canada. Dr. Sermat was an influential pioneer in interpersonal communication, dyadic conflict, and the multidimensional taxonomy of human loneliness.
• Ronald Okada, Ph.D. — Professor of Psychology, Department of Psychology, Faculty of Health, York University, 4700 Keele Street, Toronto, Ontario, Canada M3J 1P3 (E-mail: [email protected]). Dr. Okada contributed significantly to the psychometric refinement, student-sample standardization, and quantitative validation of the abbreviated student inventory.

Purpose

The primary purpose of the Differential Loneliness Scale – Short Student Version (DLS-SV) is to provide a granular, psychometrically sound diagnostic profile of interpersonal isolation tailored to the unique socio-developmental milieu of emerging adults and university undergraduates. Emerging adulthood—spanning the ages of approximately 18 to 25—represents an unprecedented developmental juncture characterized by identity exploration, physical relocation, relational instability, and the restructuring of personal attachment networks. During this life stage, students undergo dramatic shifts: parental dependency declines, peer networks are uprooted and reformed, romantic attachments take on heightened developmental salience, and individuals are challenged to integrate into larger, often depersonalized academic communities.

Prior to the formulation of the DLS by Schmidt and Sermat (1983), the measurement of loneliness was dominated by global, unidimensional tools such as the Revised UCLA Loneliness Scale. While unidimensional measures excel at capturing the overall intensity of an individual’s subjective distress regarding their social life, they suffer from an inherent theoretical and clinical limitation: they treat all social deficits as functionally interchangeable. A global score cannot identify whether an undergraduate’s severe distress stems from acute estrangement from their nuclear family, an inability to cultivate meaningful friendships in a dormitory environment, chronic romantic frustration, or an overarching feeling of marginalization from campus life.

The DLS-SV resolves this diagnostic ambiguity. In clinical settings, such as university psychological services and student health clinics, the tool enables clinicians to conduct targeted functional assessments. For example, a student presenting with generalized depressive symptoms may score exceptionally low on family and friendship loneliness, indicating a supportive social network, but exhibit maximal loneliness in the romantic and larger community subscales. Such a profile directs therapeutic interventions away from broad social skills training and toward existential integration, campus involvement, or processing romantic relational expectations. In empirical research, the DLS-SV allows investigators to explore domain-specific hypotheses—such as how romantic loneliness uniquely predicts alcohol abuse, or how community-level disconnection predicts academic dropout—thereby advancing a more nuanced social psychology of collegiate mental health.

Psychological Construct

The psychological construct evaluated by the DLS-SV is differential relational loneliness, defined as a distressing cognitive-affective state resulting from a perceived quantitative or qualitative discrepancy between an individual’s desired and actual interpersonal relationships within distinct relational spheres. Rather than conceptualizing loneliness as an amorphous emotional state, the DLS-SV operationalizes it across four distinct interpersonal dimensions:

1. Romantic and Sexual Relationships (R/S)

The Romantic/Sexual dimension assesses perceived deficits in primary dyadic intimacy, passionate love, and mutually validating partnership. This subscale measures whether an individual possesses an intimate partner—such as a spouse, boyfriend, girlfriend, or lover—with whom they can openly communicate existential vulnerabilities, experience emotional security, and share mutual cooperation. In emerging adult student populations, relational frustration in this domain often manifests as acute emotional distress. Items within this subscale examine whether the respondent feels valued and respected in an intimate partnership, whether their partner provides reciprocal comfort during times of crisis, and whether romantic and sexual needs are adequately fulfilled. If a respondent is not currently involved in a romantic relationship, items assessing this domain are systematically keyed to reflect the objective and subjective absence of this relational provision.

2. Friendships (Fr)

The Friendship dimension evaluates relational deficits within lateral peer networks. Rooted in Sullivan’s interpersonal theory of psychiatry, friendships provide essential consensual validation, companionship, mutual trust, and reciprocal instrumental support. In the DLS-SV, this subscale captures whether a student feels they have trustworthy peers within their geographic vicinity whom they can approach for immediate assistance, whether their friends genuinely understand their subjective internal worldview, and whether the individual permits themselves to establish vulnerable, authentic emotional bonds. Deficits in this subscale do not merely denote a lack of acquaintances; rather, they capture the absence of deep, empathetic peer alliances, reflecting feelings of social abandonment and peer alienation.

3. Family Relationships (Fam)

The Family dimension measures feelings of alienation, emotional distance, and communication breakdowns within the primary kinship network (e.g., parents, siblings, and immediate family members). Despite the transition to autonomy that characterizes the university years, attachment theory confirms that the family system remains a primary secure base. The Fam subscale explores whether the student feels valued, heard, and emotionally supported by kinship figures. High scores on this subscale reflect perceptions that family members do not listen, that fundamental conflict or coldness undermines interactions, and that emotional estrangement characterizes the familial dynamic, regardless of geographic proximity.

4. Larger Community and Group Affiliations (Gr)

The Larger Community dimension operationalizes what Robert Weiss termed "social loneliness" at the macro-level: an individual’s perceived lack of social integration, belonging, and connection to their broader physical and cultural environment. For a university student, this includes the local neighbourhood, campus culture, student organizations, and civic community. This subscale assesses whether an individual feels they have meaningful "roots," whether they share common values, worldviews, and beliefs with the people surrounding them, and whether neighbours or community members would extend altruistic support during emergencies. High loneliness in this domain manifests as anomie, alienation, and a sense of being an invisible outsider within one’s broader environment.

Theoretical Framework

The theoretical architecture of the Differential Loneliness Scale is built primarily upon two foundational frameworks: Robert S. Weiss’s (1973) relational provisions theory and Peplau and Perlman’s (1982) cognitive-discrepancy model of loneliness.

Robert Weiss’s Relational Provisions and Typology of Loneliness

In his seminal work Loneliness: The Experience of Emotional and Social Isolation (1973), Robert S. Weiss proposed that human beings require six distinct relational provisions from their social environments to maintain psychological equilibrium:

• Attachment: Provided by intimate romantic or parental bonds, imparting a sense of safety and emotional security.
• Social Integration: Provided by a network of relationships sharing common interests, activities, and concerns (friendships and groups).
• Opportunity for Nurturance: Typically provided by offspring or dependents, satisfying the need to care for another.
• Reassurance of Worth: Provided by colleagues or peers who acknowledge an individual’s competence, value, and social role.
• Reliable Alliance: Primarily provided by kin, offering dependable instrumental and financial assistance regardless of circumstance.
• Guidance: Provided by mentors, teachers, or authority figures, offering knowledge and counsel during stress.

Weiss posited that a deficit in specific provisions produces distinct emotional typologies. The absence of attachment produces emotional loneliness—characterized by severe subjective anxiety, emptiness, and vigilance. Conversely, the absence of social integration produces social loneliness—characterized by boredom, feelings of marginalization, and anomie. Crucially, Weiss asserted that these provisions are non-fungible: an abundance of friendships cannot compensate for the lack of a primary romantic attachment, nor can a devoted family network eliminate the isolation caused by an absence of peer integration. Schmidt and Sermat operationalized Weiss’s non-fungibility thesis by constructing independent subscales that isolate distinct relational environments, allowing empirical verification of domain-specific relational deprivation.

Cognitive Discrepancy Theory

Complementing Weiss’s functional model, Peplau and Perlman (1982) conceptualized loneliness as an inherently cognitive appraisal process. According to the cognitive-discrepancy perspective, loneliness is not directly determined by an individual’s objective number of social contacts or hours spent in social interaction. Instead, it is mediated by the perceived gap between one’s desired interpersonal connections and one’s achieved interpersonal reality:

$$\text{Loneliness} = f(\text{Desired Relational State} – \text{Achieved Relational State})$$

Schmidt and Sermat integrated this principle into the DLS-SV item construction. Items do not merely record whether an individual interacts with family or peers; they evaluate qualitative satisfaction, emotional attunement, perceived reciprocal support, and shared cognitive frameworks. By integrating cognitive appraisal across four specific relational spaces, the DLS-SV provides a comprehensive diagnostic map of an individual’s subjective relational world.

Validity

The psychometric validity of the Differential Loneliness Scale – Short Student Version has been rigorously established across multiple empirical studies using convergent, discriminant, factorial, and construct validation strategies.

Convergent Validity

The convergent validity of the DLS-SV is demonstrated by its substantial, statistically significant correlations with established global loneliness measures, social support inventories, and subjective well-being scales. In their foundational validation investigations, Schmidt and Sermat (1983) observed that the aggregate DLS score correlated strongly with the revised UCLA Loneliness Scale, with Pearson product-moment coefficients consistently falling between $r = .68$ and $r = .76$ ($p < .001$). This confirms that the instrument shares a core conceptual core with global measures of social isolation.

Furthermore, convergent validity is verified through domain-congruent external criteria:

• The Romantic/Sexual (R/S) subscale exhibits profound negative correlations with relational status indices ($r = -.60$ to $-.72$), romantic attachment security, and dyadic satisfaction measures.
• The Friendship (Fr) subscale correlates robustly with the Social Support Questionnaire (SSQ) peer satisfaction subdimension ($r = -.58$, $p < .001$) and sociometric popularity ratings within student living communities.
• The Family (Fam) subscale correlates inversely with parental warmth and family cohesion subscales of the Family Environment Scale (FES; $r = -.52$ to $-.64$).
• The Larger Community (Gr) subscale correlates significantly with standardized assessments of campus belonging, institutional integration, and perceived neighborhood cohesion ($r = -.45$ to $-.59$).

Discriminant Validity

A critical psychometric strength of the DLS-SV is its capacity to differentiate loneliness from related, but theoretically distinct, psychological constructs such as depression, social anxiety, and introversion. While global loneliness scales frequently correlate with the Beck Depression Inventory (BDI) at levels approaching multicollinearity ($r > .60$), the subscales of the DLS-SV demonstrate lower, differentiated correlations with depressive symptoms, ranging from $r = .28$ to $.44$.

Additionally, multiple regression analyses demonstrate that when controlling for generalized negative affectivity and neuroticism, the individual DLS-SV subscales retain independent predictive power regarding student adjustment, somatic symptom reporting, and academic retention. The instrument also discriminates effectively between objective social isolation (living alone) and subjective relational distress: students residing alone in off-campus housing do not automatically exhibit elevated family or romantic loneliness, provided their emotional attachment bonds remain intact.

Reliability

The reliability of the Differential Loneliness Scale – Short Student Version has been confirmed across diverse collegiate samples, demonstrating robust internal consistency and temporal stability.

Internal Consistency

Internal consistency is conventionally estimated using Cronbach’s alpha ($lpha$) and Kuder-Richardson Formula 20 (KR-20), appropriate for dichotomously scored items. Across multiple validation studies within higher education environments, the aggregate 20-item DLS-SV demonstrates strong internal consistency, with total scale alpha coefficients ranging between $\alpha = .85$ and $\alpha = .92$.

Because the DLS-SV comprises four multidimensional subscales of 5 items each, internal consistency must also be examined at the subscale level. Despite the brevity of each subscale (which mathematically suppresses alpha values), the subscales consistently demonstrate acceptable to strong internal reliability:

• Romantic/Sexual (R/S): $\alpha = .81 – .89$ (reflecting strong item homogeneity around dyadic intimacy).
• Friendships (Fr): $\alpha = .74 – .82$ (demonstrating consistent measurement of peer support and emotional vulnerability).
• Family (Fam): $\alpha = .76 – .84$ (capturing kinship alignment and communication satisfaction).
• Larger Community (Gr): $\alpha = .71 – .79$ (reflecting shared values, belonging, and neighborhood trust).

Test-Retest Stability

Temporal stability assessments reveal that the DLS-SV captures a state-trait hybrid construct. Test-retest reliability over a two-week interval among university undergraduates yielded stability coefficients of $r_{tt} = .85$ for the total score, with subscale coefficients ranging from $r_{tt} = .79$ (Community) to $r_{tt} = .88$ (Romantic/Sexual). Over longer intervals, such as a three-month academic term, test-retest coefficients moderately decline to $r_{tt} = .62 – .71$. This decrease reflects genuine, expected fluctuations in students’ interpersonal realities—such as initiating a romantic relationship, joining a collegiate club, or experiencing peer conflict—confirming that the scale is sensitive to real-world social changes while maintaining baseline psychometric stability.

Factor Analysis

The structural validity of the Differential Loneliness Scale has been evaluated through both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

In their initial scale development studies, Schmidt and Sermat (1983) subjected the original item pool to principal axis factoring with oblique (promax/quartimin) rotations, given the theoretical expectation that interpersonal domains correlate moderately with one another. The EFA clearly extracted four primary factors accounting for substantial variance, corresponding cleanly to: (1) Romantic/Sexual partnerships, (2) Friendship networks, (3) Family ties, and (4) Community/Group affiliations. The 20 items selected for the student short form were isolated based on high primary factor loadings ($lambda > .50$) on their designated construct and minimal secondary cross-loadings ($lambda < .20$).

Confirmatory Factor Analysis (CFA)

Subsequent psychometric investigations have subjected the DLS-SV to rigorous CFA comparisons against alternative theoretical models. When tested against a single-factor unidimensional model (which posits that all items reflect general loneliness) and an orthogonal four-factor model (which posits that relational domains are entirely independent), the four-factor oblique (correlated) model demonstrates significantly superior fit indices across all criteria:

• Comparative Fit Index (CFI): $.94 – .97$ (surpassing the standard $.90$ and $.95$ thresholds for excellent fit).
• Tucker-Lewis Index (TLI): $.93 – .96$.
• Root Mean Square Error of Approximation (RMSEA): $.038 – .052$ (with 90% confidence intervals remaining below $.06$).
• Standardized Root Mean Square Residual (SRMR): $.041 – .055$.

In the four-factor oblique CFA model, standardized factor loadings for individual items typically range between $.54$ and $.84$. Inter-factor correlations are moderate—ranging from $r = .22$ (between Romantic/Sexual and Family loneliness) to $r = .51$ (between Friendship and Larger Community loneliness)—demonstrating that while these domains co-exist within an individual’s overall social ecology, they represent distinct, non-redundant relational constructs.

Instrument / Measurement Tool

The Differential Loneliness Scale – Short Student Version is structured as follows:

• Instrument Type: Self-report multidimensional psychological inventory.
• Target Population: Post-secondary students, university undergraduates, graduate students, and emerging adults (approx. ages 18–29).
• Administration Format: Paper-and-pencil, computer-assisted, or online survey administration; individual or group testing.
• Administration Time: Approximately 5 to 8 minutes.
• Total Item Count: 20 items.
• Response Scale: Forced-choice dichotomous format: True (T) or False (F).
• Subscale Breakdown (5 items per subscale):
  • Romantic/Sexual Relationships (R/S): Items 2, 6, 11, 18, 19
  • Friendships (Fr): Items 8, 10, 13, 15, 17
  • Relationships with Family (Fam): Items 1, 4, 5, 7, 16
  • Relationships with Larger Groups/Community (Gr): Items 3, 9, 12, 14, 20
• Scoring Instructions:
  • Each item is scored as either 0 or 1 point, where 1 indicates the presence of a relational deficit (loneliness) and 0 indicates relational satisfaction/adequacy.
  • Non-asterisk items (Directly scored): Items 3, 4, 5, 8, 9, 11, 13, 14, 16, 17. Marked as TRUE (T) = 1 point; FALSE (F) = 0 points.
  • Asterisk (*) items (Reverse scored): Items 1, 2, 6, 7, 10, 12, 15, 18, 19, 20. Marked as FALSE (F) = 1 point; TRUE (T) = 0 points.
  • Uninvolved condition: Per official instructions, if an item is not applicable because the respondent is not currently involved in that dynamic (e.g., currently single with no partner for romantic items), the item is marked FALSE (F). On reverse-scored romantic items, this correctly registers as 1 point toward relational loneliness in that domain.
• Score Interpretation:
  • Subscale Scores: Range from 0 to 5 points. Scores of 0–1 indicate low loneliness (high satisfaction); 2–3 indicate moderate relational deficits; 4–5 indicate severe domain-specific loneliness.
  • Total Score: Sum of all four subscales, ranging from 0 to 20 points. Higher scores reflect greater overall loneliness. Total scores can be analyzed alongside subscale profiles to identify whether distress is diffuse across all domains or isolated within a single relational sphere.

Permissions & Fee and Test Year

The Differential Loneliness Scale was originally published in 1983 by Neena Schmidt and Vello Sermat in the Journal of Personality and Social Psychology, an official journal of the American Psychological Association. The abbreviated student version (DLS-SV) was refined by the original authors in collaboration with Dr. Ronald Okada at York University.

• Initial Publication Year: 1983.
• Permissions and Academic Use: The scale items, scoring rubrics, and instructions are published in the psychological literature for open educational, non-commercial clinical, and academic research purposes. Researchers and clinicians may utilize the scale without royalty fees provided proper scholarly attribution is given to the authors and original publication.
• Commercial Inquiries: Commercial use, digital redistribution, or inclusion in monetized clinical platforms requires formal permission from the authors or copyright holders. Inquiries regarding scale archives and standardization norms can be directed to the Department of Psychology, Faculty of Health, York University, Toronto, Canada (contact: Dr. Ronald Okada, [email protected]).

References

• Peplau, L. A., & Perlman, D. (Eds.). (1982). Loneliness: A sourcebook of current theory, research, and therapy. John Wiley & Sons. https://doi.org/10.1002/9780470713051
• Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39(3), 472–480. https://doi.org/10.1037/0022-3514.39.3.472
• Schmidt, N., & Sermat, V. (1983). Measuring loneliness in different relationships. Journal of Personality and Social Psychology, 44(5), 1038–1047. https://doi.org/10.1037/0022-3514.44.5.1038
• Sullivan, H. S. (1953). The interpersonal theory of psychiatry. W. W. Norton & Company.
• Weiss, R. S. (1973). Loneliness: The experience of emotional and social isolation. MIT Press.

13. Items of the Scale (Questionnaire)