Carroll Rating Scale for Depression (CRS)

Abstract

The Carroll Rating Scale for Depression (CRS) is a 52-item, self-administered psychometric instrument developed by Bernard J. Carroll and colleagues in 1981 to provide a direct self-report operationalization of the clinician-administered Hamilton Rating Scale for Depression (HRSD/HAM-D). While observer-rated scales such as the HRSD represent the historic gold standard for assessing depressive symptom severity in clinical trials, their reliance on skilled psychiatric interviewers imposes significant logistical, temporal, and financial burdens. The CRS was formulated to capture the exact dimensional architecture of the 17-item HRSD—including core affective symptoms, cognitive features of guilt and suicidality, vegetative and sleep architecture disturbances, anxiety manifestations, and somatic complaints—using a clear, forced-choice binary format (Yes/No). Extensive psychometric evaluations demonstrate that the CRS possesses high internal consistency (Cronbach’s alpha typically ranging between .87 and .93), exceptional convergent validity with the HRSD (Pearson’s r ranging from .71 to .87), and robust concordance with other established self-report metrics, such as the Beck Depression Inventory (BDI). Factor analytic investigations reveal a dominant general depressive factor alongside distinct subdimensions corresponding to affective-cognitive distress, psychomotor inhibition, anxiety, and neurovegetative exhaustion. With an established clinical cutoff score of 10 or 12 for mild-to-moderate depression and 14 or 16 for major depressive episodes, the CRS serves as an indispensable tool in clinical trials, outpatient screening, and epidemiological psychiatric research where rapid, standardized, and clinician-free quantification of depressive pathology is required.

Keywords

Carroll Rating Scale for Depression, CRS, Hamilton Rating Scale for Depression, HRSD, major depressive disorder, psychometrics, self-rating scale, depression assessment, melancholia, clinical psychometrics, symptom severity, diagnostic screening, internal consistency, psychomotor retardation, vegetative symptoms

Authors

The Carroll Rating Scale for Depression was designed, developed, and psychometrically validated by a multidisciplinary research team based primarily at the Mental Health Research Institute and the Department of Psychiatry at the University of Michigan Medical Center in Ann Arbor, Michigan, USA:

• Bernard J. Carroll, MD, PhD: Chief investigator, biological psychiatrist, and internationally recognized pioneer in neuroendocrinology and affective disorders. Dr. Carroll was renowned for developing the Dexamethasone Suppression Test (DST) for melancholic depression and held prominent academic chairs at the University of Michigan, Duke University Medical Center, and the Pacific Behavioral Research Foundation.
• Michael Feinberg, MD, PhD: Associate Professor of Psychiatry, University of Michigan Medical Center; co-developer of neurobiological and psychometric diagnostic protocols for affective disorders.
• Peter E. Smouse, PhD: Professor of Human Genetics and Biostatistics, University of Michigan; prominent quantitative geneticist and biostatistician who oversaw the structural equation modeling, statistical calibration, and mathematical validation of the scale.
• S. Gail Rawson, MSW: Clinical research coordinator and psychometric specialist, Clinical Psychobiology Unit, Department of Psychiatry, University of Michigan.
• John F. Greden, MD: Research psychiatrist, later Chair of the Department of Psychiatry at the University of Michigan and Founding Director of the University of Michigan Depression Center; leading authority on the neurobiology, biomarkers, and longitudinal course of major depressive disorder.

Purpose

The primary clinical and psychometric purpose of the Carroll Rating Scale for Depression (CRS) is to provide an objective, reliable, and standardized self-report quantification of depressive symptom severity that parallels the clinician-rated Hamilton Rating Scale for Depression (HRSD). In the late 1970s and early 1980s, psychopharmacological research and psychiatric epidemiological studies were heavily reliant upon the HRSD to establish treatment efficacy and document clinical outcomes. However, the HRSD requires an experienced, highly trained clinician to conduct a comprehensive semi-structured clinical interview, which typically demands 30 to 45 minutes per patient evaluation. In large-scale clinical trials, longitudinal psychiatric monitoring, and high-volume clinical settings, the repeated administration of the HRSD introduces considerable observer burden, escalating research costs, and introduces inter-rater variability that can obscure subtle biological or pharmacotherapeutic treatment effects.

Existing self-report inventories of that era, most notably the Beck Depression Inventory (BDI) and the Zung Self-Rating Depression Scale (SDS), exhibited distinct conceptual emphases that departed from the clinician-rated HRSD. The BDI was intentionally constructed to emphasize the cognitive formulations of depression—focusing primarily on negative self-attributions, pessimism, guilt, and cognitive distortions—while placing comparatively less weight on somatic, motoric, and vegetative manifestations. Conversely, the HRSD heavily weights somatic anxiety, initial, middle, and delayed insomnia, psychomotor agitation and retardation, gastrointestinal disruptions, and general physical fatigue—symptoms that are central to the biological and melancholic phenotypes of affective disease. Carroll and colleagues identified a vital methodological discrepancy: clinical research protocols were evaluating biological markers (such as hypothalamic-pituitary-adrenal axis dysregulation via the Dexamethasone Suppression Test) against clinician ratings on the HRSD, while simultaneously collecting self-report data via instruments that measured divergent psychological constructs.

The CRS was explicitly constructed to bridge this operational gap. By translating each of the 17 behavioral, affective, and somatic dimensions of the classic HRSD into sets of straightforward, patient-accessible statements, the CRS allows patients to self-report their depressive phenomenology within an exact parallel framework. The scale was designed to achieve several critical operational objectives:

• Equivalence to Observer Ratings: To generate total and subscale scores that correlate maximally with expert clinical judgments derived from the HRSD, establishing an empirical bridge between patient self-perception and objective psychiatric observation.
• Ease of Administration: To utilize a binary forced-choice format (“Yes” vs. “No”) written at a basic reading level, reducing the cognitive and decision-making burden that frequently impairs depressed patients when confronted with multi-tiered Likert scales or complex multiple-choice matrices.
• Monitoring Treatment Trajectories: To provide a sensitive, repeatable metric capable of tracking rapid symptomatic changes during pharmacotherapy, psychotherapy, electroconvulsive therapy (ECT), or hospitalization.
• Broad Clinical and Epidemiological Utility: To facilitate cost-effective, high-throughput screening in psychiatric intake units, primary care clinics, and community mental health surveys without sacrificing the clinical specificity inherent in the HRSD taxonomy.

Psychological Construct

The Carroll Rating Scale for Depression operationalizes depression not as an undifferentiated psychological state, but as a complex, multifaceted clinical syndrome characterized by simultaneous disturbances across affective, cognitive, psychomotor, autonomic, and neurovegetative domains. Because the instrument was formulated to map directly onto the 17-item Hamilton Rating Scale for Depression, its 52 items are distributed across 17 distinct, clinically defined symptomatic dimensions:

1. Depressed Mood (Items 16, 32, 34, 48)

This dimension captures the core subjective affective state of depression, characterized by profound gloom, despondency, sorrow, unprovoked weeping, an overwhelming sense of hopelessness, and a conviction of unending future misery. Keyed responses reflect pervasive negative affect and the subjective inability to experience positive emotional states.

2. Feelings of Guilt (Items 14, 20, 24, 44)

Reflecting cognitive distortions common to major affective disorders, this construct evaluates self-reproach, pathological shame, feelings of worthlessness, the conviction that current emotional distress constitutes deserved punishment for past misdeeds, and a profound loss of personal self-esteem.

3. Suicidal Ideation and Behaviors (Items 12, 17, 29, 46)

Spanning passive death wishes to explicit active suicidal contemplation, this critical dimension evaluates thoughts that dying represents the only viable solution, recurrent wishes to be dead, the cognitive loss of life’s worth, and concrete thoughts of self-harm or suicide attempts.

4. Initial Insomnia (Items 9, 22)

Measures sleep-onset latency and difficulties initiating sleep, defined clinically as requiring prolonged periods (e.g., more than half an hour to an hour) to transition from wakefulness to sleep.

5. Middle Insomnia (Items 19, 27)

Assesses nocturnal sleep continuity disturbances, fragmented sleep architecture, repeated awakenings during the middle of the night, and subjective restlessness throughout the sleep period.

6. Delayed / Terminal Insomnia (Items 11, 35)

Quantifies early morning awakening, a classical vegetative hallmark of melancholic and endogenous depression. Items assess spontaneous early awakening significantly ahead of intended waking times without the ability to return to sleep.

7. Work and Interests (Items 3, 7, 25, 42)

Reflects anhedonia, avolition, and functional vocational impairment. This dimension evaluates the abandonment of routine leisure pursuits, social withdrawal, loss of satisfaction from activities, and subjective difficulty fulfilling vocational or occupational obligations.

8. Psychomotor Retardation (Items 21, 28, 30, 47)

Captures the central nervous system slowing characteristic of severe affective states, including subjective mental bradyphrenia, slowness of thought, reduction of vocal inflection and vitality, gross motor inertia, and severe reductions in behavioral productivity.

9. Psychomotor Agitation (Items 6, 10, 37, 43)

Evaluates motor restlessness, inner tension, fidgeting, involuntary motor output, pacing behaviors, and the inability to maintain physical or motoric stillness.

10. Psychic Anxiety (Items 8, 23, 31, 38)

Measures the psychological manifestation of anxiety, encompassing unexplainable dread, cognitive apprehension, irritability, subjective jitteriness, concentration impairments, and acute states of panic.

11. Somatic Anxiety (Items 13, 18, 33, 41)

Assesses the physiological and peripheral manifestations of autonomic arousal, including palpitations, tachycardia, postural dizziness, autonomic gastrointestinal distress (indigestion), and visible tremors of the hands.

12. Gastrointestinal Somatic Symptoms (Items 36, 50)

Focuses directly on appetite disturbance, severe hypophagia, loss of the hedonic quality of food, and the necessity of forcing oneself to ingest nutrition.

13. Items of the Scale (Questionnaire)

14. Genital Symptoms / Libido (Items 4, 15)

Evaluates the sexual consequences of depressive illness, specifically targeting loss of libido, diminished sexual interest, and general sexual dysfunction following illness onset.

15. Hypochondriasis and Somatic Preoccupation (Items 5, 39, 45, 49)

Measures excessive, pathologically amplified somatic concern, catastrophic misinterpretation of bodily sensations, obsessional worry over physical health, and, at its severe extreme, depressive somatic delusions (e.g., beliefs that one’s internal organs are decaying or rotten).

16. Loss of Weight (Items 2, 52)

Directly assesses progressive, non-dietary physical weight loss associated with the metabolic and appetitive shifts of clinical depression.

17. Loss of Insight (Items 26, 40)

Assesses the patient’s capacity to recognize their depressive condition as an internal psychiatric or psychological disorder, rather than attributing their profound incapacitation purely to external environmental factors (such as the weather) or trivializing their illness as merely needing rest.

Theoretical Framework

The theoretical architecture underpinning the Carroll Rating Scale for Depression is rooted in the empirical, biological, and clinical nosology of affective disorders that matured during the mid-20th century, championed prominently by Max Hamilton and later refined by biological psychiatrists such as Bernard J. Carroll. Unlike psychological models of depression that emerged from psychoanalysis or pure cognitive-behavioral paradigms—such as Aaron T. Beck’s cognitive theory, which posits that negative cognitive triads (distorted views of self, world, and future) serve as the primary etiologic drivers of depression—the framework informing the CRS and HRSD is fundamentally psychiatric and phenomenological.

This framework is grounded in the European and Kraepelinian tradition of conceptualizing depression as a systemic, biological illness (frequently termed endogenous or melancholic depression). Within this paradigm, mood disturbance represents merely one facet of a comprehensive neurobiological dysregulation affecting the diencephalic and limbic systems. Consequently, major depressive disorder produces predictable disruptions in:

• Circadian and Sleep Homeostasis: Alterations in rapid eye movement (REM) latency and slow-wave sleep architecture, manifesting clinically as fragmented sleep, sleep maintenance failure, and pathognomonic terminal insomnia.
• Metabolic and Autonomic Regulation: Dysregulation of the autonomic nervous system and neuroendocrine axes (such as the hypothalamic-pituitary-adrenal axis), producing weight loss, anorexia, gastrointestinal slowing, and somatic anxiety.
• Basal Ganglia and Motor Output Systems: Disruption of motor pathways resulting in observable psychomotor retardation (bradykinesia, expressive poverty) or psychomotor agitation.

Max Hamilton intentionally designed the HRSD from extensive clinical interviews with hospitalized, severely depressed patients. His scale rejected theoretical speculation regarding unconscious conflicts or cognitive schemata, relying instead on an empirical index of clinical severity derived from symptoms that experienced psychiatrists used to gauge illness depth and treatment response. When Carroll and colleagues formulated the CRS, they embraced Hamilton’s exact phenomenological realism. Their foundational theoretical assumption was that an accurately operationalized self-report inventory must mirror the clinician’s diagnostic lens.

Furthermore, the CRS was designed alongside neurobiological investigations, particularly Dr. Carroll’s landmark work on the Dexamethasone Suppression Test (DST) as a potential biological marker for melancholic depression. Carroll hypothesized that if laboratory markers were to be meaningfully correlated with clinical phenotypes, psychometric evaluation needed to capture the neurovegetative and physical core of depressive pathology rather than exclusively psychological or existential dysphoria. Thus, the theoretical core of the CRS is a balanced, medically oriented model of depression that gives equal prominence to cognitive, affective, vegetative, and motoric manifestations.

Validity

The psychometric validity of the Carroll Rating Scale for Depression has been extensively investigated across diverse inpatient, outpatient, and community populations. Multiple validation studies have confirmed its robust construct, convergent, discriminant, and criterion-related validity.

Convergent and Criterion Validity

In the seminal validation studies conducted by Carroll et al. (1981) and Feinberg et al. (1981), the CRS was administered concurrently with the 17-item Hamilton Rating Scale for Depression to psychiatric inpatients and outpatients suffering from major depressive disorder, bipolar depression, schizoaffective disorder, and other non-depressive psychiatric conditions. The Pearson correlation coefficient between the total CRS score and the total HRSD score was exceptionally high, ranging from r = .71 to r = .87 across clinical samples. In direct comparative trials, this level of correspondence equaled or exceeded that observed between the HRSD and the Beck Depression Inventory (which typically correlates with the HRSD at r = .60 to .75 in inpatient samples), demonstrating that the CRS achieved its primary objective of mimicking observer ratings via patient self-report.

The correlation between the CRS and the Beck Depression Inventory is equally strong, frequently yielding coefficients between r = .75 and r = .86. This demonstrates strong convergent validity across independent self-report modalities while retaining superior representation of vegetative and motor symptoms. Longitudinal studies evaluating patients undergoing somatic therapies (such as tricyclic antidepressants, SSRIs, or ECT) demonstrate that change scores on the CRS track closely with change scores on the HRSD, validating the CRS as an objective indicator of therapeutic response and remission.

Discriminant Validity

The CRS effectively discriminates between clinical populations diagnosed with major depressive episodes and healthy non-clinical controls, as well as between depressed patients and those with non-affective psychiatric conditions (such as generalized anxiety disorder, personality disorders, or schizophrenia). In validation studies, healthy control subjects consistently obtain mean CRS scores well below 6 (typically M = 3.5, SD = 3.2), whereas actively depressed psychiatric inpatients consistently display mean scores exceeding 20 (M = 22.8, SD = 8.6). Receiver Operating Characteristic (ROC) analyses confirm the diagnostic accuracy of the CRS, with area under the curve (AUC) values consistently exceeding .90 when distinguishing major depression from normal control conditions.

Clinical Cutoff Scores and Diagnostic Accuracy

Clinical calibration studies have identified practical score thresholds for clinical decision-making:

• Score < 10: Normal or non-depressed range; minimal depressive symptomatology.
• Score 10 to 13: Borderline / mild depressive disturbance; warrants clinical observation.
• Score ≥ 14 (or ≥ 16 in strict inpatient contexts): Highly suggestive of a major depressive episode, exhibiting optimal diagnostic sensitivity (approx. 85%) and specificity (approx. 88%) against DSM clinical criteria and HRSD thresholds (≥ 14 on the 17-item HRSD).

Reliability

The Carroll Rating Scale for Depression exhibits high reliability across varied testing conditions, psychiatric populations, and translation adaptations.

Internal Consistency

In initial validation studies involving both depressed inpatients and normal adult populations, Carroll et al. (1981) demonstrated that the CRS displays high internal consistency. The total scale yields a Cronbach’s alpha coefficient typically between .87 and .93, reflecting exceptional item homogeneity and cohesive construct representation despite the wide clinical diversity of its 17 constituent symptom categories. Split-half reliability coefficients, calculated via the Spearman-Brown prophecy formula, consistently range from .85 to .91 across clinical samples.

Item-Total Correlations

Item-total correlation analyses show that the vast majority of the 52 items correlate significantly with the aggregate score at p < .001. Core affective and cognitive items (such as Item 16: “I am miserable or often feel like crying”, Item 34: “I think my case is hopeless”, and Item 48: “There is only misery in the future for me”) exhibit corrected item-total correlations exceeding .65. Somatic and vegetative items, which naturally display greater variance based on medical comorbidity and age, demonstrate corrected item-total correlations ranging between .35 and .55. Items tapping loss of insight (Items 26 and 40) exhibit lower item-total correlations (.20 to .35), a psychometric pattern identical to that observed on the clinician-rated HRSD, where lack of insight is generally confined to severe psychotic or delusional depression.

Test-Retest Reliability

Test-retest stability coefficients for the CRS demonstrate reliable measurement consistency over short intervals when clinical status remains stable. Over intervals of 24 to 72 hours in clinically stabilized psychiatric outpatients and healthy controls, test-retest reliability coefficients range between r = .80 and r = .89. In non-clinical longitudinal cohorts retested across intervals of two to four weeks, temporal stability remains solid (r = .74 to .82). In clinical cohorts undergoing active medical or psychological interventions, CRS scores decrease in tandem with clinician ratings, showing that the scale remains sensitive to genuine symptom change while resisting random measurement drift.

Factor Analysis

Multiple exploratory (EFA) and confirmatory factor analytic (CFA) investigations have dissected the latent dimensionality of the 52-item Carroll Rating Scale for Depression. The structural architecture of the CRS mirrors the psychometric debates surrounding the Hamilton Rating Scale for Depression, demonstrating both a dominant general depression dimension and several meaningful lower-order symptom factors.

Exploratory Factor Structure

Initial principal components and exploratory factor analyses conducted by Carroll, Feinberg, and colleagues revealed that a large, unrotated first principal component accounts for approximately 25% to 35% of the total variance across clinical samples. This prominent “general depression” factor (a g-factor of depressive severity) is characterized by high positive loadings (> .50) from items reflecting depressed mood, pessimism, anhedonia, anergia, and suicidal ideation, confirming that the scale functions effectively as a unidimensional metric of global depressive burden.

Subsequent orthogonal (Varimax) and oblique (Promax) rotations have consistently extracted distinct multi-factorial solutions, typically ranging between four and six primary factors:

• Factor 1: Cognitive-Affective Depressive Core: Explains the largest proportion of common variance; characterized by heavy loadings from items reflecting sadness, crying, worthlessness, guilt, pessimism, and suicidal contemplation (e.g., Items 12, 14, 16, 20, 34, 48).
• Factor 2: Vegetative Sleep Disturbance / Insomnia: A distinct biological factor consisting of the six sleep items spanning initial insomnia (Items 9, 22), middle insomnia (Items 19, 27), and terminal early morning awakening (Items 11, 35).
• Factor 3: Somatic and Psychic Anxiety: Captures subjective apprehension, nervous tension, tremors, tachycardia, and visceral distress (e.g., Items 6, 10, 13, 23, 33, 38, 41).
• Factor 4: Psychomotor Retardation and Anergia: Marked by items tapping mental bradyphrenia, speech changes, physical sluggishness, severe fatigue, and anhedonic withdrawal from social and vocational roles (e.g., Items 1, 3, 7, 21, 28, 30, 47, 51).
• Factor 5: Gastrointestinal and Weight Disturbance: Encompasses loss of appetite, forced eating, and non-voluntary weight loss (Items 2, 36, 50, 52).
• Factor 6: Somatic Preoccupation and Hypochondriasis: Marked by amplified focus on bodily health, organic disease fears, and somatic distortions (Items 5, 39, 45, 49).

Confirmatory Factor Analysis and Model Fit

Modern structural equation modeling and CFA evaluations demonstrate that while a strict unidimensional model exhibits marginal goodness-of-fit due to the clinical heterogeneity of vegetative and somatic symptoms, a hierarchical bifactor model—comprising a general depression factor alongside orthogonal specific symptom group factors—demonstrates superior fit to empirical data. Goodness-of-fit indices for the bifactor specification typically achieve acceptable standards in clinical cohorts (Comparative Fit Index [CFI] > .90; Tucker-Lewis Index [TLI] > .90; Root Mean Square Error of Approximation [RMSEA] ≤ .05 to .06). These factor-analytic findings confirm that the internal architecture of the CRS closely parallels the multidimensional structure observed across international studies of the clinician-rated HRSD.

Instrument / Measurement Tool

• Test Type: Self-administered psychological assessment / clinical rating scale.
• Format: Paper-and-pencil or computer-administered questionnaire.
• Number of Items: 52 items.
• Response Scale: Forced-choice binary format (Yes / No).
• Administration Time: Approximately 5 to 10 minutes.
• Target Population: Adolescents and adults (ages 16+) presenting with suspected or diagnosed affective disorders, as well as individuals in general medical or community settings.
• Reading Level: Approximately 6th-grade reading level; accessible language designed to minimize cognitive strain.
• Scoring Rules & Keying Mechanism:
  • Each item is assigned a score of either 0 or 1.
  • The instrument includes 40 keyed-positive items where answering “Yes” indicates depressive pathology and earns 1 point, while “No” earns 0 points (Items: 2, 3, 4, 5, 6, 9, 10, 11, 12, 13, 14, 16, 17, 18, 19, 20, 21, 22, 23, 24, 26, 27, 30, 31, 33, 34, 35, 37, 38, 39, 40, 41, 45, 46, 47, 48, 49, 50, 51, 52).
  • The instrument includes 12 reverse-keyed items where answering “No” indicates depressive pathology and earns 1 point, while “Yes” earns 0 points (Items marked with asterisk on No: 1, 7, 8, 15, 25, 28, 29, 32, 36, 42, 43, 44).
  • The total CRS score is calculated by summing all keyed responses, yielding an overall score ranging from 0 to 52.
• Subscale Breakdown (17 Dimensions):
  • 1. Depression: Items 16, 32, 34, 48 (Range: 0–4)
  • 2. Guilt: Items 14, 20, 24, 44 (Range: 0–4)
  • 3. Suicide: Items 12, 17, 29, 46 (Range: 0–4)
  • 4. Initial Insomnia: Items 9, 22 (Range: 0–2)
  • 5. Middle Insomnia: Items 19, 27 (Range: 0–2)
  • 6. Delayed Insomnia: Items 11, 35 (Range: 0–2)
  • 7. Work and Interests: Items 3, 7, 25, 42 (Range: 0–4)
  • 8. Retardation: Items 21, 28, 30, 47 (Range: 0–4)
  • 9. Agitation: Items 6, 10, 37, 43 (Range: 0–4)
  • 10. Psychic Anxiety: Items 8, 23, 31, 38 (Range: 0–4)
  • 11. Somatic Anxiety: Items 13, 18, 33, 41 (Range: 0–4)
  • 12. Gastrointestinal Somatic Symptoms: Items 36, 50 (Range: 0–2)
  • 13. General Somatic Symptoms: Items 1, 51 (Range: 0–2)
  • 14. Libido: Items 4, 15 (Range: 0–2)
  • 15. Hypochondriasis: Items 5, 39, 45, 49 (Range: 0–4)
  • 16. Loss of Weight: Items 2, 52 (Range: 0–2)
  • 17. Loss of Insight: Items 26, 40 (Range: 0–2)
• Clinical Severity Interpretation Guidelines:
  • 0–9: Within normal limits / Absence of depression.
  • 10–13: Mild / Borderline depressive symptomatology.
  • 14–25: Moderate depressive episode; clinically significant depression.
  • 26–52: Severe to extremely severe major depression (frequently accompanied by melancholic, vegetative, or psychomotor features).

Permissions & Fee and Test Year

The Carroll Rating Scale for Depression was introduced into the scientific literature in 1981 through a series of foundational validation articles published in the British Journal of Psychiatry. The instrument was developed within an academic medical research institution (University of Michigan) to advance clinical research, psychiatric care, and clinical trials.

For independent non-commercial research, academic inquiry, graduate dissertations, and routine clinical application, the CRS has historically been made available without licensing fees, provided that appropriate scholarly attribution is accorded to the original authors and the British Journal of Psychiatry. When adapted for commercial applications, proprietary software systems, or sponsored clinical trials conducted by commercial entities, researchers are advised to consult formal institutional copyright holders or the publishers of the foundational papers to secure explicit administrative clearance. No fee is required for individual clinicians administering the scale for clinical evaluation.

References

The development, psychometric calibration, clinical utility, and structural validity of the Carroll Rating Scale for Depression are supported by the following foundational references:

• Carroll, B. J., Feinberg, M., Smouse, P. E., Rawson, S. G., & Greden, J. F. (1981). The Carroll Rating Scale for Depression. I. Development, reliability and validation. British Journal of Psychiatry, 138(3), 194–200. https://doi.org/10.1192/bjp.138.3.194
• Feinberg, M., Carroll, B. J., Smouse, P. E., & Rawson, S. G. (1981). The Carroll Rating Scale for Depression. II. Factor analyses. British Journal of Psychiatry, 138(3), 201–204. https://doi.org/10.1192/bjp.138.3.201
• Feinberg, M., Carroll, B. J., Smouse, P. E., & Rawson, S. G. (1981). The Carroll Rating Scale for Depression. III. Comparison with other rating instruments. British Journal of Psychiatry, 138(3), 205–209. https://doi.org/10.1192/bjp.138.3.205
• Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, and Psychiatry, 23(1), 56–62. https://doi.org/10.1136/jnnp.23.1.56
• Hamilton, M. (1967). Development of a rating scale for primary depressive illness. British Journal of Social and Clinical Psychology, 6(4), 278–296. https://doi.org/10.1111/j.2044-8260.1967.tb00530.x
• McDowell, I. (2006). Measuring Health: A Guide to Rating Scales and Questionnaires (3rd ed.). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195165678.001.0001
• Smouse, P. E., Feinberg, M., Carroll, B. J., Park, M. H., & Rawson, S. G. (1981). The Carroll Rating Scale for Depression. IV. High-risk screening and diagnosis. British Journal of Psychiatry, 138(3), 210–214. https://doi.org/10.1192/bjp.138.3.210

Items of the Scale