Depression Anxiety Stress Scales – 21 (DASS-21)

1. Abstract

The Depression Anxiety Stress Scales – 21 (DASS-21) represents a refined, quantitatively robust dimensional instrument engineered to assess three interrelated yet empirically distinct negative affective syndromes: depression, anxiety, and tension/stress. Derived by Peter F. Lovibond and Syd H. Lovibond (1995) as an abbreviated alternative to the original 42-item inventory (DASS-42), the instrument comprises 21 self-report items organized into three 7-item subscales. Items are evaluated along a 4-point severity/frequency continuum reflecting symptoms experienced over the preceding week. The fundamental psychometric objective of the DASS-21 was to address longstanding measurement challenges inherent in traditional affective screeners, which frequently conflate somatic anxiety symptoms with depressive manifestations or subsume both under an undifferentiated construct of general psychological distress. Through iterative exploratory and confirmatory psychometric procedures, the authors isolated core affective criteria: dysphoria, anhedonia, and devaluation of life for Depression; autonomic physiological arousal, skeletal musculature symptoms, and subjective apprehension for Anxiety; and chronic non-specific autonomic arousal, difficulty relaxing, nervous irritability, and hyperreactivity for Stress. The scale exhibits exceptional reliability across diverse samples, with internal consistency coefficients (Cronbach’s alpha) typically exceeding .88 for Depression, .82 for Anxiety, and .85 for Stress in both non-clinical and psychiatric cohorts. Structural equation modeling and extensive factor analyses across multicultural populations have supported both a three-factor oblique model and a bifactor structure consisting of a general distress dimension alongside orthogonal specific factors. Due to its public domain status, minimal administrative burden, and diagnostic discriminant validity, the DASS-21 has established itself as an indispensable benchmark in global psychological research and evidence-based clinical monitoring.

2. Keywords

DASS-21, Depression Anxiety Stress Scales, psychometrics, tripartite model, negative affectivity, affective assessment, factor analysis, construct validity, clinical psychology, psychological distress

3. Authors

The Depression Anxiety Stress Scales were formulated and psychometrically standardized by Peter F. Lovibond and Sydney H. Lovibond within the School of Psychology at the University of New South Wales (UNSW) in Sydney, Australia.

• Sydney H. Lovibond, PhD: Late Emeritus Professor of Psychology at the University of New South Wales. Renowned for foundational contributions to behavioral therapy, conditioning paradigms, and empirical psychometric development.
• Peter F. Lovibond, PhD: Professor of Psychology at the University of New South Wales. Specialist in associative learning, human fear conditioning, cognitive-behavioral processes, and structural conceptualizations of emotional disorders.
• Institutional Correspondence: School of Psychology, Faculty of Science, University of New South Wales, Sydney NSW 2052, Australia. Official repository and documentation are maintained via the UNSW DASS Web Resource.

4. Purpose

The primary clinical and empirical motivation for engineering the DASS-21 was to establish a conceptually pure, non-redundant, and empirically grounded instrument capable of differentiating between manifestations of depression, physical anxiety, and persistent stress. For decades, the psychometric assessment of emotional disturbances was dominated by instruments such as the Beck Depression Inventory (BDI), the Beck Anxiety Inventory (BAI), and the State-Trait Anxiety Inventory (STAI). While clinically revolutionary, these conventional self-report inventories shared substantial shared variance, typically demonstrating intercorrelations ranging from .60 to .75. Consequently, clinicians and researchers struggled to determine whether high scores indicated genuine syndromal comorbidity or merely reflected a single underlying trait of negative affectivity or demoralization.

To overcome this psychometric confound, the Lovibond research team set out to construct a dimensional measurement system based not on historical diagnostic consensus criteria—which are frequently compromised by arbitrary diagnostic committees—but rather on empirical response patterning within clinical and general populations. The specific objectives were threefold:

1. Construct Clarification: Isolating the essential features of depression (e.g., loss of self-worth, anhedonia, total absence of positive reinforcement) from the physiological activation patterns characteristic of panic and acute anxiety (e.g., palpitations, tremors, dry mouth).
2. Delineation of a Third State (Stress): Operationalizing an intermediate emotional state—colloquially termed ‘stress’ or nervous tension—characterized by sustained physiological arousal, impatience, irritability, difficulty tolerating frustration, and an inability to wind down. This state is frequently observed in work environments, primary care settings, and non-psychiatric clinical environments, yet historically lacked systematic quantification outside somatic symptom inventories.
3. Streamlined Clinical Utility: Offering clinicians an ultra-brief instrument that can be administered within three minutes, easily scored, and tracked across longitudinal treatment courses (e.g., measuring weekly session-by-session outcomes in cognitive behavioral therapy) without inducing cognitive fatigue in patients experiencing acute cognitive deceleration or severe psychomotor agitation.

In research contexts, the DASS-21 functions as an optimal dependent variable or covariate. It allows clinical researchers to demonstrate whether an experimental intervention selectively remediates autonomic hyperarousal, depressive cognitions, or generalized tension. Its clean psychometric separation minimizes suppressor effects and multicollinearity in multivariate regression equations and structural equation models.

5. Psychological Construct

The DASS-21 measures three fundamental constructs along dimensional axes, explicitly conceptualizing affective pathology on a continuum that extends seamlessly from healthy, normal fluctuations in mood to profound psychiatric debilitation. Rather than assigning categorical diagnoses, the instrument delineates quantitative symptom density.

Depression Subscale

The Depression subscale of the DASS-21 is distinguished by its primary emphasis on low positive affectivity, loss of incentive motivation, and fundamental cognitive appraisals of failure and meaninglessness. In clinical psychometrics, it captures the melancholic and core cognitive features of unipolar mood disorders while avoiding non-specific somatic items (e.g., appetite disruption, insomnia) that are readily confounded by physical illness, medication side effects, or aging. The seven operational items evaluate:

• Anhedonia and Loss of Positive Valence: Inability to experience positive emotional states, joy, or enthusiasm (e.g., Item 3: “I couldn’t seem to experience any positive feeling at all”; Item 16: “I was unable to become enthusiastic about anything”).
• Inertia and Loss of Volition: Subjective absence of drive, initiative, or executive energy necessary to initiate daily tasks (e.g., Item 5: “I found it difficult to work up the initiative to do things”).
• Hopelessness and Devaluation of Life: Pessimistic projections regarding the future and loss of existential value (e.g., Item 10: “I felt that I had nothing to look forward to”; Item 21: “I felt that life was meaningless”).
• Dysphoria and Self-Deprecation: Profound sadness coupled with diminished self-worth (e.g., Item 13: “I felt down-hearted and blue”; Item 17: “I felt I wasn’t worth much as a person”).

Anxiety Subscale

The Anxiety subscale targets the acute and episodic subjective, cognitive, and autonomic manifestations of fear and physiological activation. Unlike scales that evaluate diffuse worry or cognitive rumination, the DASS-21 anxiety dimension aligns closely with acute sympathetic nervous system excitation and somatic cues of emergency defense activation:

• Autonomic Physiological Arousal: Peripheral sympathetic discharges, such as cardiovascular reactivity and mucosal dryness (e.g., Item 2: “I was aware of dryness of my mouth”; Item 19: “I was aware of the action of my heart in the absence of physical exertion”).
• Respiratory and Skeletal Musculature Disturbances: Hyperventilation sensations and somatic tremor (e.g., Item 4: “I experienced breathing difficulty”; Item 7: “I experienced trembling”).
• Subjective Panic and Threat Apprehension: The phenomenological perception of imminent loss of control or impending catastrophic decompensation (e.g., Item 9: “I was worried about situations in which I might panic and make a fool of myself”; Item 15: “I felt I was close to panic”; Item 20: “I felt scared without any good reason”).

Stress Subscale

The Stress subscale quantifies a chronic, persistent state of non-specific autonomic nervous system hyperarousal. Whereas Anxiety captures acute panic and autonomic surges, Stress captures an ongoing, persistent state of neuromuscular tension, poor frustration tolerance, and irritable reactivity. It reflects an organism operating under a sustained, exhausting appraisal of environmental demand:

• Difficulty Relaxing and Sustained Tension: Physical inability to unwind or achieve a resting neuromuscular baseline (e.g., Item 1: “I found it hard to wind down”; Item 12: “I found it difficult to relax”).
• Hyperarousal and Nervous Energy Depletion: Pervasive feeling of excessive neuromuscular mobilization (e.g., Item 8: “I felt that I was using a lot of nervous energy”; Item 11: “I found myself getting agitated”).
• Irritability, Impatience, and Over-Reactivity: Diminished threshold for external disturbances and disproportionate emotional reactions (e.g., Item 6: “I tended to over-react to situations”; Item 14: “I was intolerant of anything that kept me from getting on with what I was doing”; Item 18: “I felt that I was rather touchy”).

6. Theoretical Framework

The construction of the DASS-21 is deeply rooted in contemporary cognitive and affective science, particularly integrating the Tripartite Model of Anxiety and Depression advanced by Clark and Watson (1991). The tripartite formulation emerged to explain the immense structural overlap between affective states. Clark and Watson posited that emotional disorders could be decomposed into three primary components:

1. General Distress / Negative Affectivity (NA): A non-specific, transdiagnostic factor shared across both depression and anxiety disorders, manifesting as general demoralization, insomnia, emotional instability, and subjective misery.
2. Physiological Hyperarousal (PH): A specific somatic signature unique to panic, somatic anxiety, and acute phobic reactions, encompassing autonomic arousal, tachycardia, tremor, and motor restlessness.
3. Low Positive Affectivity (LPA): A specific motivational and affective signature unique to depressive illness, manifesting as anhedonia, apathy, lack of reward responsiveness, and fatigue.

The Lovibond research initiative sought to instantiate and refine this conceptual model through empirical psychometric selection. During item development, candidate items were administered to broad non-clinical and psychiatric samples. Items that demonstrated substantial cross-loadings across clinical categories were systematically eliminated. Consequently, the DASS Depression subscale represents a direct psychometric operationalization of Clark and Watson’s Low Positive Affectivity coupled with cognitive helplessness, while the DASS Anxiety subscale serves as an empirical operationalization of Physiological Hyperarousal.

Critically, Lovibond and Lovibond (1995) extended beyond the original tripartite formulation by isolating the third independent dimension: Stress. While Clark and Watson conceived Negative Affectivity as a broad, diffuse higher-order trait, the Lovibonds established that once depressive dysphoria and somatic fear were removed, there remained a coherent, clinically salient cluster of symptoms reflecting persistent, non-episodic cognitive and behavioral strain. This conceptualization aligns closely with Hans Selye’s general adaptation syndrome and Richard Lazarus’s transactional model of stress. Under this cognitive-relational framework, stress manifests when environmental demands exceed subjective coping resources, resulting in cognitive impatience, vigilance, and sustained readiness for action without the acute autonomic surge typical of acute danger appraisals.

7. Validity

The construct, convergent, discriminant, and predictive validity of the DASS-21 has been rigorously corroborated across clinical, occupational, university, and community cohorts worldwide.

Convergent and Discriminant Validity

Empirical studies consistently show strong convergent validity between DASS-21 subscales and corresponding standard criterion measures. In a seminal validation study on large clinical and non-clinical cohorts, Antony et al. (1998) examined the psychometric functioning of the DASS-21 relative to legacy inventories:

• The DASS-21 Depression subscale demonstrated high correlation with the Beck Depression Inventory (r = .79 to .85) and the depression scale of the Hospital Anxiety and Depression Scale (HADS-D; r = .74), while showing substantially lower associations with somatic anxiety measures.
• The DASS-21 Anxiety subscale correlated strongly with the Beck Anxiety Inventory (r = .81 to .84) and the State-Trait Anxiety Inventory-State subscale (STAI-S; r = .71), demonstrating sensitivity to somatic hyperreactivity.
• The DASS-21 Stress subscale displayed moderate-to-high correlations with the Perceived Stress Scale (PSS-10; r = .68 to .73) and the STAI-Trait subscale (r = .69), capturing general emotional tension and frustration intolerance that standard anxiety measures fail to distinguish.

Discriminant validity is evidenced by the lower cross-construct correlations. For instance, the correlation between DASS-21 Depression and BAI (typically around .50 to .58) is significantly lower than that observed between BDI and BAI (which frequently exceeds .70), demonstrating that the DASS-21 successfully reduces redundant somatic overlap.

Predictive and Diagnostic Criterion Validity

Receiver Operating Characteristic (ROC) analyses demonstrate that the DASS-21 exhibits excellent sensitivity and specificity for identifying formal DSM and ICD affective and anxiety disorders. Research evaluating clinical diagnostic efficiency shows an Area Under the Curve (AUC) ranging between .86 and .93 for the Depression scale in identifying Major Depressive Episodes, and an AUC between .84 and .91 for the Anxiety scale in identifying Panic Disorder, Social Anxiety Disorder, and Agoraphobia. Longitudinal prospective studies have confirmed its predictive validity, showing that baseline DASS-21 Stress scores reliably predict subsequent workplace absenteeism, burnout syndromes, and cardiovascular markers of autonomic strain over 12-month follow-up periods.

8. Reliability

The psychometric reliability of the DASS-21 has been thoroughly evaluated across multiple languages, cultures, and sampling methodologies. Both internal consistency and temporal stability metrics consistently exceed conventional psychometric benchmarks across both clinical and non-clinical cohorts.

Internal Consistency

In a large-scale non-clinical normative investigation of 1,794 British participants, Henry and Crawford (2005) established high internal consistency coefficients for all three subscales:

• Depression Subscale: Cronbach’s α = .88 (McDonald’s ω = .89)
• Anxiety Subscale: Cronbach’s α = .82 (McDonald’s ω = .84)
• Stress Subscale: Cronbach’s α = .90 (McDonald’s ω = .91)
• Total Scale: Cronbach’s α = .93

In clinical psychiatric populations, internal consistency is even more robust. Studies conducted across outpatient treatment clinics (e.g., Antony et al., 1998; Brown et al., 1997) routinely report alphas between .91 and .94 for Depression, .84 and .87 for Anxiety, and .88 and .91 for Stress. Corrected item-total correlations for each scale universally exceed .50, confirming strong item homogeneity within each intended target construct.

Test-Retest Reliability

Temporal stability assessments over intervals ranging from two weeks to three months reveal excellent test-retest reliability under conditions where participants remain untreated. Over a two-week interval, intra-class correlation coefficients (ICCs) or Pearson coefficients typically range from .78 to .86 for Depression, .74 to .81 for Anxiety, and .80 to .87 for Stress. Crucially, in clinical clinical intervention trials, the DASS-21 exhibits high treatment sensitivity: scores drop significantly in response to evidence-based psychotherapy or pharmacotherapy, demonstrating that the instrument reliably captures state changes while maintaining structural integrity over time.

9. Factor Analysis

The latent structure of the DASS-21 has been extensively evaluated using Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA), and advanced bifactor structural modeling. Decades of structural research have addressed whether the instrument is best understood as three distinct oblique dimensions, a single overarching distress factor, or a hierarchical bifactor structure.

Confirmatory Factor Analysis Models

Numerous large-scale CFA studies across diverse international populations have tested competing structural models:

1. One-Factor Model: Assumes all 21 items load onto a single general psychological distress factor. This model routinely demonstrates poor fit to empirical data (e.g., Comparative Fit Index [CFI] < .80; Root Mean Square Error of Approximation [RMSEA] > .10).
2. Two-Factor Model: Collapses Anxiety and Stress onto a single dimension contrasted with Depression. While superior to the one-factor model, this specification exhibits sub-optimal fit across global samples.
3. Three-Factor Correlated Model: Freely estimates correlations between the three separate factors (Depression, Anxiety, Stress). This model typically yields strong goodness-of-fit indices (e.g., CFI ≥ .94, Tucker-Lewis Index [TLI] ≥ .93, RMSEA ≤ .055, Standardized Root Mean Square Residual [SRMR] ≤ .045), confirming the empirical utility of the three subscales.

Bifactor and Hierarchical Formulations

Recent psychometric research utilizing bifactor structural equation modeling (e.g., Norton, 2007) suggests that the optimal latent representation is a bifactor model. In this architecture, a general factor (representing non-specific negative affectivity or general psychological distress) accounts for the substantial common variance shared across all 21 items, while three orthogonal group factors capture the unique, specific variance attributable to Depression (anhedonia, low positive affect), Anxiety (somatic activation), and Stress (nervous tension, agitation). The bifactor model consistently yields superior model fit (CFI > .97, RMSEA < .04), supporting both the interpretation of individual subscale scores and the calculation of a composite Total Score as an overall indicator of distress severity.

Factor loadings for individual items onto their respective target factors are uniformly high. For the Depression scale, standardized loadings predominantly range between .65 and .84; for Anxiety, between .55 and .80; and for Stress, between .62 and .79, demonstrating robust construct coverage and minimal cross-loadings across cultural settings.

10. Instrument / Measurement Tool

• Test Designation: Depression Anxiety Stress Scales – 21 (DASS-21)
• Authors: Peter F. Lovibond and Sydney H. Lovibond (1995)
• Instrument Nature: Quantitative self-report psychometric rating scale
• Constructs Evaluated: Depression (dysphoria, anhedonia, meaninglessness), Anxiety (autonomic arousal, panic apprehension), and Stress (nervous tension, agitation, irritability)
• Total Number of Items: 21 items (7 items per subscale)
• Administration Modality: Paper-and-pencil questionnaire, computerized administration, or mobile ecological momentary assessment (EMA)
• Target Population: Adolescents (ages 14+) and adults across general, community, and clinical cohorts
• Completion Time: Approximately 2 to 4 minutes
• Response Scale (Verbatim): 4-point severity/frequency scale:
  • 0 = Did not apply to me at all
  • 1 = Applied to me to some degree, or some of the time
  • 2 = Applied to me to a considerable degree, or a good part of time
  • 3 = Applied to me very much, or most of the time
• Temporal Recall Frame: Symptoms experienced over the past week (“over the past week”)
• Subscale Item Composition:
  • Depression: Items 3, 5, 10, 13, 16, 17, 21
  • Anxiety: Items 2, 4, 7, 9, 15, 19, 20
  • Stress: Items 1, 6, 8, 11, 12, 14, 18
• Reverse Scoring Rules: No items are reverse scored. All items are positively framed toward distress.
• Scoring and Transformation Rules: Raw subscale scores are obtained by summing the 7 respective item scores (range 0 to 21 per subscale). Crucial Note: Because the DASS-21 represents the abbreviated version of the full 42-item scale, the subscale raw sum is multiplied by 2 to compare scores directly with the established normative classifications and clinical cutoffs of the DASS-42.
• Severity Stratifications (Multiplied Subscale Scores):
  • Depression: Normal (0–9), Mild (10–13), Moderate (14–20), Severe (21–27), Extremely Severe (28+)
  • Anxiety: Normal (0–7), Mild (8–9), Moderate (10–14), Severe (15–19), Extremely Severe (20+)
  • Stress: Normal (0–14), Mild (15–18), Moderate (19–25), Severe (26–33), Extremely Severe (34+)

11. Permissions & Fee and Test Year

Publication Year: The 21-item short form (DASS-21) and the original 42-item inventory (DASS) were formally published in 1995 within the journal Behaviour Research and Therapy.

Permissions and Copyright: The DASS and DASS-21 instruments are held under copyright by Peter F. Lovibond and Sydney H. Lovibond. However, in accordance with the authors’ commitment to advancing clinical psychology and open scientific research, the scale is explicitly placed in the public domain for clinical, educational, and research use. No formal licensing fees, royalties, or written permissions from the authors are required to reproduce, administer, score, or electronically implement the scale for non-commercial research or clinical practice.

Conditions of Use: Researchers and practitioners are required to preserve the wording and scoring format of the items without arbitrary modification, and to cite the primary source literature (Lovibond & Lovibond, 1995) in all formal reports and scientific publications. Commercial distributors wishing to incorporate the instrument into proprietary, fee-based diagnostic packages or software systems must verify terms directly with the authors through the official UNSW DASS portal.

12. References

Antony, M. M., Bieling, P. J., Cox, B. J., Enns, M. W., & Swinson, R. P. (1998). Psychometric properties of the 42-item and 21-item versions of the Depression Anxiety Stress Scales in clinical groups and a community sample. Psychological Assessment, 10(2), 176–181. https://doi.org/10.1037/1040-3590.10.2.176

Brown, T. A., Chorpita, B. F., Korotitsch, W., & Barlow, D. H. (1997). Psychometric properties of the Depression Anxiety Stress Scales (DASS) in clinical samples. Behaviour Research and Therapy, 35(1), 79–89. https://doi.org/10.1016/S0005-7967(96)00068-X

Clark, L. A., & Watson, D. (1991). Tripartite model of anxiety and depression: Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100(3), 316–336. https://doi.org/10.1037/0021-843X.100.3.316

Henry, J. D., & Crawford, J. R. (2005). The short-form version of the Depression Anxiety Stress Scales (DASS-21): Construct validity and normative data in a large non-clinical sample. British Journal of Clinical Psychology, 44(2), 227–239. https://doi.org/10.1348/014466505X29657

Lovibond, P. F., & Lovibond, S. H. (1995). The structure of negative emotional states: Comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories. Behaviour Research and Therapy, 33(3), 335–343. https://doi.org/10.1016/0005-7967(94)00075-U

Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety Stress Scales (2nd ed.). Psychology Foundation of Australia.

Norton, P. J. (2007). Depression Anxiety and Stress Scales (DASS-21): Psychometric analysis across four racial groups. Anxiety, Stress, and Coping, 20(3), 253–265. https://doi.org/10.1080/10615800701309279

Osman, A., Wong, J. L., Bagge, C. L., Freedenthal, S., Gutierrez, P. M., & Lozano, G. (2012). The Depression Anxiety Stress Scales-21 (DASS-21): Further examination of dimensions, scale reliability, and correlates. Journal of Clinical Psychology, 68(12), 1322–1338. https://doi.org/10.1002/jclp.21908

13. Items of the Scale (Questionnaire)