The Depression Anxiety Stress Scales – Long Form (DASS-42)<\/strong> is a 42-item self-report instrument developed by University of New South Wales<\/a> researchers Syd H. Lovibond and Peter F. Lovibond (1995) to assess the negative emotional states of depression<\/a>, anxiety<\/a>, and stress<\/a>. Developed using empirical construct-validation and factor-analytic procedures, the instrument was intentionally engineered to maximize discrimination between the core affective manifestations of depression (characterized primarily by anhedonia<\/a>, dysphoria, devaluation of life, and psychomotor inertia) and anxiety (dominated by physiological hyperarousal, autonomic activation, and panic sensations), while introducing an independent third dimension of chronic non-specific tension and stress (manifested as irritability, impatience, difficulty winding down, and persistent mental agitation). Each of the three subscales comprises 14 items scored on a 4-point Likert-type severity scale ranging from 0 (“Did not apply to me at all”<\/em>) to 3 (“Applied to me very much, or most of the time”<\/em>), evaluating experiences over the antecedent week.<\/p>\n Psychometric evaluations across non-clinical normative populations (N<\/em> = 2,914) and diverse clinical cohorts demonstrate robust psychometric properties. Internal consistency estimates routinely reach or exceed Cronbach’s alpha<\/a> coefficients of α = .91 for Depression, α = .84 to .88 for Anxiety, and α = .90 for Stress. Confirmatory factor analyses (CFA) consistently confirm that an oblique three-factor model or a bifactor structural specification demonstrates superior fit compared to unitary or two-factor affective models, supporting the discriminant validity of the tripartite affective model. The DASS-42 provides researchers and clinicians with a psychometrically sound, dimensional assessment system suitable for clarifying the primary locus of affective disturbance, monitoring longitudinal treatment trajectories, and investigating the empirical boundaries between overlapping emotional disorders<\/span>.<\/p>\n DASS-42, Depression Anxiety Stress Scales, psychometrics, affective disorders, anxiety assessment, depression screening, psychological stress, tripartite model, confirmatory factor analysis, internal consistency<\/p>\n The Depression Anxiety Stress Scales were conceived and developed by:<\/p>\n Institutional oversight and archival publication have been managed historically through The Psychology Foundation of Australia, based at the Mathews Building, School of Psychology, University of New South Wales, Sydney NSW 2052, Australia.<\/p>\n The primary purpose of the DASS-42 is to provide a reliable, dimensionally pure, and empirically validated measurement of the severity of the core symptoms characterizing depression, anxiety, and psychological stress. Prior to the development of the DASS, traditional self-report inventories—most notably the Beck Depression Inventory (BDI)<\/a> and the Beck Anxiety Inventory (BAI)<\/a>—suffered from high inter-scale correlations, frequently ranging between .60 and .75. This substantial overlap led many psychometricians to hypothesize that existing measures failed to separate distinct clinical syndromes, reflecting instead a broad, undifferentiated factor of general demoralization or negative affectivity.<\/p>\n Lovibond and Lovibond initiated an extensive scale development program aimed at isolating the core symptoms that uniquely define depression and anxiety, while setting aside non-discriminating symptoms such as sleep disturbance, appetite changes, fatigue, and generic somatic distress. During the initial empirical factor analyses of candidate items, a distinct third cluster of items repeatedly separated from both the classic anhedonic\/depressive factor and the autonomic\/anxiety factor. This third factor was characterized by persistent non-specific nervous tension, difficulty relaxing, impatience, irritability, and being easily startled or agitated. Consequently, the authors formalized this construct as the Stress<\/strong> subscale, establishing the DASS as a comprehensive tri-dimensional assessment platform.<\/p>\n In clinical practice, the DASS-42 fulfills several assessment functions:<\/p>\n The DASS-42 evaluates emotional distress as three distinct, interrelated dimensions rather than discrete, qualitative disease entities. This dimensional framework assumes that affective disturbances exist on a continuous spectrum ranging from subclinical, everyday emotional fluctuations to severe psychiatric disorders.<\/p>\n The Depression subscale focuses on the core features of unipolar depressive disorders, with particular emphasis on dysphoric<\/a> mood and low positive affect. Rather than assessing secondary vegetative or somatic symptoms (e.g., changes in appetite, sleep architecture, or weight) that often overlap with physical medical illnesses, this dimension measures the cognitive, existential, and motivational facets of depression:<\/p>\n The Anxiety scale measures the subjective, somatic, and behavioral correlates of acute autonomic arousal and fear-based reactions. This construct closely aligns with the physiological hyperarousal domain described in modern psychiatric literature:<\/p>\n The Stress subscale captures chronic, non-specific physiological and cognitive tension. Unlike the acute physiological episodes seen in the Anxiety scale or the blunted affect of the Depression scale, the Stress scale assesses ongoing strain, low frustration tolerance, and persistent difficulty achieving equilibrium:<\/p>\n The theoretical framework of the DASS is grounded in the dimensional perspective of psychopathology and connects directly to the Tripartite Model of Anxiety and Depression<\/a> formulated by Lee Anna Clark and David Watson (1991). The tripartite model addressed the long-standing clinical challenge of high comorbidity and diagnostic overlap between affective and anxiety syndromes by conceptualizing them around three core components:<\/p>\n During the item selection and purification phases, Syd and Peter Lovibond tested whether empirical item sets could differentiate pure anxiety from pure depression. Their empirical findings corroborated the central tenets of Clark and Watson’s model: items reflecting pure autonomic activation and fear-based physical arousal separated from items indexing pure anhedonia, self-deprecation, and existential despair.<\/p>\n However, Lovibond and Lovibond made a pivotal empirical contribution that extended beyond the traditional tripartite formulation: they discovered that a distinct third factor emerged alongside the depression and anxiety factors, which they termed Stress<\/strong>. This factor absorbed the chronic, non-specific tension, irritability, impatience, and difficulty relaxing that typically accounted for cross-loadings between earlier anxiety and depression inventories. Rather than treating this tension as an undifferentiated “general negative affectivity” wastebasket, the DASS operationalized it as a coherent, measurable affective state characterized by persistent mobilization of physiological and mental coping resources.<\/p>\n Furthermore, the DASS rests upon a continuum assumption: the quantitative difference between normal individuals and clinically diagnosed patients is primarily one of severity and duration, rather than a categorical or qualitative boundary. Individuals without psychiatric conditions frequently experience transient states of autonomic arousal, sadness, or frustration. By measuring these states continuously, the DASS supports research across both non-clinical community samples and acute clinical cohorts.<\/p>\n Extensive psychometric investigations provide substantial support for the construct, convergent, discriminant, and criterion-related validity of the DASS-42 across clinical and non-clinical populations.<\/p>\n Initial construct validation was conducted on a non-clinical normative sample of 2,914 Australian adults (Lovibond & Lovibond, 1995), spanning university students, corporate employees, blue-collar workers, and healthcare professionals. Confirmatory factor analysis (CFA) evaluated several alternative structural models. An oblique three-factor model consistently demonstrated superior fit compared to single-factor (general distress) models and two-factor (depression vs. anxiety) models. Subsequent clinical validations by Antony et al. (1998) and Brown et al. (1997) confirmed this three-factor structure in psychiatric outpatients with major depressive disorder<\/a>, panic disorder<\/a>, social anxiety disorder<\/a>, and generalized anxiety disorder (GAD)<\/a>.<\/p>\n The convergent validity of the DASS-42 subscales is supported by strong correlations with established reference instruments:<\/p>\n Studies examining diagnostic predictive validity confirm that clinical groups score substantially higher across all three subscales compared to non-clinical controls. Antony et al. (1998) found that patients with panic disorder scored highest on the DASS-Anxiety subscale, patients with major depressive disorder scored highest on the DASS-Depression subscale, and patients with generalized anxiety disorder scored highest on the DASS-Stress subscale. These differential patterns confirm the diagnostic utility of the instrument.<\/p>\n The DASS-42 demonstrates high internal consistency, test-retest stability, and low measurement error across diverse linguistic adaptations and demographic groups.<\/p>\n In the primary normative standardization sample of 2,914 Australian adults (Lovibond & Lovibond, 1995), Cronbach’s alpha<\/a> coefficients revealed strong internal reliability:<\/p>\n In a large independent British community cohort study (N<\/em> = 1,794) conducted by Crawford and Henry (2003), the internal reliabilities were replicated with remarkable precision: α = .95 for Depression, α = .90 for Anxiety, and α = .93 for Stress, yielding an overall scale internal consistency of α = .97. Clinical psychiatric validation studies (Antony et al., 1998; Clara et al., 2001) report comparable estimates (α = .96, .89, and .93 for Depression, Anxiety, and Stress, respectively), confirming that item-total intercorrelations remain stable regardless of psychiatric symptom severity<\/span>.<\/p>\n Given that the DASS instructions evaluate emotional states over the preceding week, test-retest coefficients over short intervals reflect temporal stability, while longer intervals capture expected symptom fluctuations resulting from life circumstances or clinical interventions. Over a two-week retest interval among non-clinical participants, Lovibond and Lovibond reported stability coefficients of r<\/em> = .71 to .81 across the three subscales. Extended longitudinal studies (e.g., 6 to 12 weeks) during treatment trials show decreases in test-retest correlations (r<\/em> ≈ .45 to .60), reflecting genuine symptomatic improvements captured by the scales<\/span>.<\/p>\n The Standard Error of Measurement (SEM) for the DASS-42 subscales remains consistently low. In normative studies, the SEM is approximately 2.10 for Depression, 1.95 for Anxiety, and 2.50 for Stress. These small margins support clinical confidence when calculating reliable change indices (RCI) during psychotherapeutic follow-ups.<\/p>\n The underlying factor structure of the DASS-42 has undergone extensive scrutiny through both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA).<\/p>\n During scale development, Syd and Peter Lovibond tested initial pools of candidate symptom descriptors on university student samples (N<\/em> = 950, followed by a trial of 504 sets of responses). Principal component and maximum likelihood factor extractions with oblique (promax\/oblimin) rotations consistently identified three distinct factors with eigenvalues greater than 1.0. Items exhibiting cross-loadings (e.g., items loading > .35 on multiple factors) or weak primary loadings (< .40) were iteratively deleted. This empirical purification yielded 42 final items (14 per subscale) that loaded strongly onto their primary emotional constructs<\/span>.<\/p>\n Confirmatory factor analyses in diverse populations have evaluated competitive structural configurations, primarily comparing:<\/p>\n Crawford and Henry (2003) analyzed these models using structural equation modeling. The empirical fit indices decisively favored the tripartite models:<\/p>\nKeywords<\/h2>\n
Authors<\/h2>\n
\n
Purpose<\/h2>\n
\n
Psychological Construct<\/h2>\n
1. The Depression Dimension (Items 3, 5, 10, 13, 16, 17, 21, 24, 26, 31, 34, 37, 38, 42)<\/h3>\n
\n
2. The Anxiety Dimension (Items 2, 4, 7, 9, 15, 19, 20, 23, 25, 28, 30, 36, 40, 41)<\/h3>\n
\n
3. The Stress Dimension (Items 1, 6, 8, 11, 12, 14, 18, 22, 27, 29, 32, 33, 35, 39)<\/h3>\n
\n
Theoretical Framework<\/h2>\n
\n
Validity<\/h2>\n
Construct and Structural Validity<\/h3>\n
Convergent and Discriminant Validity<\/h3>\n
\n
Criterion and Diagnostic Predictive Validity<\/h3>\n
Reliability<\/h2>\n
Internal Consistency Reliability<\/h3>\n
\n
Test-Retest Stability<\/h3>\n
Standard Error of Measurement<\/h3>\n
Factor Analysis<\/h2>\n
Exploratory Factor Analytic Derivation<\/h3>\n
Confirmatory Factor Analysis (CFA) Fit Comparisons<\/h3>\n
\n