These pages use javascript to create fly outs and drop down navigation elements.

Symptom Checklist 90 Revised (SCL-90-R)

Please note that this section is an archive (last updated in June 2006). [disclaimer]

Sections:   Overview | Instrument Reviews | Construct Overviews | Book Compendium Reviews | Internet Site Reviews

Created 2003 April 3
Jump To A Section

Practical Information | Research Contacts | Annotated Bibliography | Factors & Norms | Reliability Evidence | Validity Evidence | Comments | Updates | Feedback

Practical Information

Instrument Name:

Symptom Checklist 90 Revised (SCL-90-R)

Instrument Description:

The SCL-90-R is a measure of psychological distress, appropriate for both adolescents and adults. It is a revised form (two items were replaced and seven were modified) of the SCL-90, an unnormed version. Other precursors include the Hopkins Symptom Check List-58 (HSCL-58) and the HSCL-90. (Ref: 1) It consists of 90 items total, with 83 items representing nine subscales: somatization (SOM), obsessive-compulsive, interpersonal sensitivity (I-S), depression (DEP), anxiety (ANX), hostility (HOS), phobic anxiety (PHOB), paranoid ideation (PAR) and psychoticism (PSY). In addition to these nine symptoms, the SCL-90-R also contains seven items which relate to appetite and sleep disturbances. It utilizes three global distress indices: Global Severity Index (GSI), Positive Symptom Distress Index (PSDI), Positive Symptom Total (PST). It may be used in screening or as a measure of patient progress.

Price:

Price varies by package. Order through Pearson Assessments

Administration Time:

12-15 minutes

Publication Year:

1975

Item Readability:

Flesch-Kincaid 7th grade level score. Phrases are short and easy to comprehend for someone with at least a 7th grade level reading ability.

Scale Format:

5-point Likert, ranging from “not at all distressing” (0) to “extremely distressing” (4).

Administration Technique:

Self-administered, computer-administered, audiocassette, or online assessment available.

Scoring and Interpretation:

Raw scores are converted to T-scores and plotted on a profile showing the centile equivalent for each subscale. Subscale scores are found by summing item scores for the entire instrument and dividing by the total number of items on each subscale. The PST score is found by adding all items not scored zero. PSDI is found by dividing the grand total by the PST score. The GSI is found by summing the scores for the 90 items and dividing by 90. (Ref: 2) The author suggests that the GSI is the best indicator of the current degree of distress when one simple summary score is needed.

Forms:

Spanish and French (Canada) language versions are available.

Research Contacts

Instrument Developers:

Leonard R. Derogatis
This instrument may be purchased from Pearson Assessments online at http://www.pearsonassessments.com/tests/scl90r.htm.

Instrument Development Location:

Johns Hopkins University School of Medicine
Baltimore, Maryland

Instrument Developer Email:

No information found.

Instrument Developer Website:

www.derogatis-tests.com

Annotated Bibliography

1. Cyr JJ, McKenna-Foley JM, Peacock E. Factor structure of the SCL-90-R: Is there one? J Pers Assess 1985 Dec;49(6):571-8. [PMID: 4093836]
Purpose: To review the literature on the factor structure of the SCL-90-R and other versions.
Methods: The instruments are evaluated for criteria including factor stability, factor loadings, and proportion of variance.
Implications: The SCL-90-R may be a better measure of general distress than of nine separate dimensions of symptomatology.

2. Kiger JA, Murphy SA. A reliability assessment of the symptom Checklist-90-R. Suitability for postdisaster-bereaved and nonloss study samples. West J Nurs Res. 1987 Nov;9(4):572-88. [PMID: 3433743]
Purpose: To report on test-retest and internal consistency reliability of the SCL-90-R as used with a high-risk bereaved population.
Sample: Total N=83. 69 bereaved subjects participated in the 11-month follow-up, and 49 of these persons participated in the 35-month follow-up. The control group consisted of 50 persons at 11-month follow-up, and 34 persons were involved in 35-month follow-up data. The groups were similar in age, gender, occupation, and geographic location: 1) Bereaved Group---74% female, mean age=30, 87% self-employed, skilled laborers, and professionals; 2) Control Group---65% female, mean age=37, 94% skilled laborers and professionals.
Methods: A longitudinal design was utilized. A disaster-bereaved group was compared with a control group 1 and 3 years postdisaster (volcanic eruption at Mt. St. Helens). Subjects were given the SCL-90-R.
Implications: The authors found that the SCL-90-R was a useful measure for disaster-bereaved subjects. The Global Severity Index is a reliable measure of overall mental distress.

3. Brophy CJ, Norvell NK, Kiluk DJ. An examination of the factor structure and convergent and discriminant validity of the SCL-90R in an outpatient clinic population. J Pers Assess. 1988 Summer;52(2):334-40. [PMID: 3404394]
Purpose: To investigate the factor structure of the SCL-90-R.
Sample: 368 adults at an outpatient psychology clinic. 61.6% female, 60% single, mean age=27.1 years, mean years of education=14.3.
Methods: The study replicated a previous factor structure analysis. Subjects were given the SCL-90-R and the Beck Depression Inventory at a pretreatment screening. 230 subjects also completed the MMPI at intake screening.
Implications: Six (6) homogeneous and stable factors emerged; however, the first factor accounted for a large amount of the variance, suggesting that the SCL-90-R is a measure of general psychopathology. Significant correlations were found between the SCL-90-R, MMPI scales and BDI. The correlation with the BDI suggests that the SCL-90-R may measure dysphoria rather than distinct symptoms.

4. McGough J, Curry JF. Utility of the SCL-90-R with depressed and conduct-disordered adolescent inpatients. J Pers Assess. 1992 Dec;59(3):552-63. [PMID: 1487808]
Purpose:
Sample: Part 1: 79 inpatients (39 boys and 40 girls), mean age=15.68 years, mean WAIS FSIQ=103.91, 87% white; Part 2: 50 inpatients from the initial 79 (27 boys and 23 girls), mean age=15.69 years.
Methods: In Part 1 of the study, convergent and discriminant validity was found by comparing subjects’ SCL-90-R scores to two adolescent self-report inventories (Children’s Depression Inventory and Jesness Inventory). In Part 2, diagnostic utility was found by comparing SCL-90-R scale scores and clinical diagnoses made during an interview, using the Schedule for Affective Disorders and Schizophrenia for School Aged Children-Epidemiological version (K-SADS-E) and a DSM-III symptom checklist.
Implications: The SCL-90-R identified a primary dimension of dysphoria and a secondary dimension of anger and mistrust in adolescents. Adolescents diagnosed with major depression showed high elevations on the depression, anxiety and obsessive-compulsive subscales.

5. Margo GM, Dewan MJ, Fisher S, Greenberg RP. Comparison of three depression rating scales. Percept Mot Skills 1992 Aug;75(1):144-6. [PMID: 1528663]
Purpose: To compare scores on three depression scales: Beck Depression Inventory (BDI), SCL-90-R, and Brief Psychiatric Rating Scale (BPRS).
Sample: 71 inpatients (44 females and 27 males): mean age=29.3 years, had at least 11th grade education, did not have serious medical or organic mental disorder. 50 patients were diagnosed with major depression. Other diagnoses were schizophrenia, eating disorder and anxiety.
Methods: Patients completed the BDI and SCL-90-R. A psychiatrist administered the BPRS.
Implications: The SCL-90-R assessed depression as well as the BDI, but also gives a little more insight into psychopathology.

6. Schauenburg H, Strack M. Measuring psychotherapeutic change with the symptom checklist SCL 90 R. Psychother Psychosom. 1999;68(4):199-206. [PMID: 10396011]
Purpose: To propose conventions and cut-off scores for psychological symptoms and change after therapy using the SCL-90-R in a German population.
Sample: Several aggregated groups: 1)German psychotherapy inpatient samples, hospitalized for 8-12 weeks with neurotic and personality disorders, 2) Two outpatient samples, 3) Private psychoanalytic practice. [Demographic information not available].
Methods: Data was collected and analyzed for several aggregated samples.
Implications: The SCL-90-R did not detect enough difference to say that any group made a clinically significant change. This data would suggest a GSI difference of 0.65 for a clinically significant change.

7. Schmitz N, Kruse J, Heckrath C, Alberti L, Tress W. Diagnosing mental disorders in primary care: the General Health Questionnaire (GHQ) and the Symptom Check List (SCL-90-R) as screening instruments. Soc Psychiatry Psychiatr Epidemiol. 1999 Jul;34(7):360-66. [PMID: 10477956]
Purpose: To investigate the screening properties of the SCL-90-R and GHQ-12 in a German primary care setting and compare the criterion validity of the two.
Sample: 572 randomly selected adult outpatients (31.3% males and 68.7% females) from 18 primary care facilities in Duesseldorf (analysis is performed on 408 patients who completed both inventories). Mean age=42.7 years.
Methods: Patients were given the German versions of the scales. They were then interviewed using the Structured Clinical Interview (SCID) and Impairment Score (IS), which were used as gold standard tests for comparison to the SCL-90-R and GHQ-12.
Implications: Both instruments were able to detect cases without much difference in performance. ROC analysis showed that the SCL-90-R Anxiety and Depression subscales showed acceptable concurrent validity for those DSM-III-R diagnostic groups.

8. Schmitz N, Hartkamp N, Franke GH. Assessing clinically significant change: application to the SCL-90-R. Psychol Rep 2000 Feb;86(1):263-74. [PMID: 10778279]
Purpose: To illustrate the utility of the SCL-90-R for assessing reliable change and clinical significance.
Sample: Three normative samples: 1) functional/healthy sample (505 females and 501 males, mean age=34 years); 2) outpatients in psychotherapy/moderately symptomatic (107 males and 167 females, mean age=36.9 years); and 3) inpatients in psychotherapy/severely symptomatic (111 males and 143 females, mean age=36.9 years).
Methods: Nonparametric Kernel Densities were fitted for the samples and two cut-scores were estimated for each SCL-90-R subscale to discriminate between functional, moderately symptomatic and severely symptomatic.
Implications: Based on results, the authors suggest three normative samples (for different diagnostic groups) to measure clinically significant change.

9. Schmitz N, Hartkamp N, Brinschwitz C, Michalek S, Tress W. Comparison of the standard and the computerized versions of the Symptom Check List (SCL-90-R): a randomized trial. Acta Psychiatr Scand 2000 Aug;102(2):147-52. [PMID: 10937788]
Purpose: To examine whether computer and paper-and-pencil versions of the SCL-90-R are equivalent.
Sample: 282 randomly assigned psychosomatic outpatients at a university clinic in Duesseldorf, Germany. The computer-administered group consisted of 57 men and 86 women, mean age=38.7 years. The control group consisted of 52 men and 89 women, mean age=38.9 years. The main ICD-10 diagnoses were mood disorders and neurotic, stress-related and somatoform disorders.
Methods: Subjects completed either the paper or computer administered test, based on a computer-generated random placement number. The computer group then took the Inventory of Interpersonal Problems on screen as well. Subjects also completed the Giessen Symptom List, and clinician’s diagnosis was based on the Impairment Score.
Implications: The computerized version makes administration and scoring more efficient. Differences between the two modes were noticeable but small: a statistically significant difference was found for the Obsessive-compulsive and Anger-hostility subscales between modes. There is evidence of an age and gender effect.

10. McGuire BE, Shores EA. Simulated pain on the Symptom Checklist 90-Revised. J Clin Psychol. 2001 Dec;57(12):1589-96. [PMID: 11745600]
Purpose: To investigate whether a clinical sample of patients with pain and a sample of pain simulators would have different profiles on the SCL-90-R.
Sample: 50 pain patients (31 male and 19 female, mean age=37.9 years, main pain sites were back, hands, neck, shoulder, arm, leg and multiple locations) and 20 pain simulators (6 male and 14 female, undergraduate psychology students, mean age=21.1 years, mean years of education=14).
Methods: All patients completed a questionnaire on demographics, pain level and history of pain. Pain patients completed the instrument under normal instructions, and pain simulators were told to feign a pain disorder but to avoid detection.
Implications: Pain simulators scored higher on all subscales, with the highest scores on Depression, Obsessive-compulsive and Somatization. They overestimated the degree of psychological distress. The recommended Positive Symptom Total cut-score for “faking bad” had poor specificity.

top

Factors and Norms

Factor Analysis Work:

In the development of the scale, nine factors were found by varimax rotation. However, only four of the ten items deemed to represent “psychoticism” loaded on this factor. It was reported that those items would be adjusted, however, they were not. (Ref: 1) Principal components analysis produced a first factor that accounted for 27% of the total variance, while the second factor accounted for only 5.1% of the variance. Varimax rotation was performed and a six-factor solution was chosen: 1) Depression, 2) Somatization, 3) Anger-Hostility, 4) Paranoid-Psychoticism, 5) Phobic Anxiety, and 6) Obsessive-Compulsive. These included 73 items of sufficient loading (>=0.40). Correlations among the 6 factors ranged from 0.48 to 0.70, while correlations with the total score ranged from 0.74 to 0.91. (Ref: 3) Two significant canonical correlations were found for the SCL-90-R in a sample of adolescents. All SCL-90-R scales had correlations >=0.69 with their first canonical variable, except for Phobic Anxiety and Paranoid Ideation. The highest correlations were for Depression (r=0.96), Anxiety (r=0.92) and Interpersonal Sensitivity (r=0.89). (Ref: 4)

Normative Information Availability:

Normative information is gathered from the SCL-90-R publisher’s website: www.ncs.com. Normative information is available for samples of:

974 nonpatient males (n=494) and females (n=480) with mean age=46, 85% white and 60% married.
1,002 adult psychiatric outpatient males (425) and females (577), mean age=31.2, 67% white and 46% single.
423 adult psychiatric inpatient males (158) and females (265), mean age=33.1, 56% sample white and 44% single.
2,408 adolescent males (1,601) and females (807), mean age = 15.8, 58% white, data gathered at six schools in two states.

Reliability Evidence

Test-retest:

For a postdisaster bereaved group, test retest coefficient alphas ranged from 0.47 to 0.69 and were significant at p<0.001. In the control group, the Interpersonal Sensitivity subscale test-retest coefficient alpha was significant at the p<0.001 level (r=0.50), while the Obsessive-Compulsive subscale (r=0.43), the Global Severity Index (r=0.40) and the Positive Symptom Total were significant at the p<0.01 level. (Ref: 2)

Inter-rater:

No information found.

Internal Consistency:

For a postdisaster bereaved group at 11- (T1) and 35-month (T2) follow-ups, all subscale coefficient alphas ranged from 0.74 to 0.91, except for the Psychoticism scale, which was 0.67 at T2. For the control group, all subscale coefficient alphas ranged from 0.72 to 0.91, except for Phobic Anxiety (0.62 at T1 and 0.44 at T2), Psychoticism (0.68 at T1), and Paranoid Ideation (0.65 at T2). The Global Severity Index coefficient alpha=0.97 across both groups at T1 and T2. Both groups at both times had identical coefficient alphas for the Depression, Anxiety, and Interpersonal Sensitivity subscales (all >=0.86). (Ref: 2)

In another study, split-half reliability for the entire test was 0.94. (Ref: 3) Coefficient alphas for the scales and the Global Severity Index range from 0.78 to 0.97 (Ref: 7,9), while coefficient alphas for a computer-administered version were similar at 0.74-0.97. (Ref: 9)

Alternate Forms:

No information found.

Validity Evidence

Construct/ Convergent/ Discriminant:

In one study, discriminant validity was not supported: all nine dimensions of the SCL-90-R correlated significantly with the Beck Depression Inventory, ranging from 0.46 to 0.73 (p<0.0001). The Depression subscale had the highest correlation with r=0.73. (Ref: 3) Convergent validity, however, was supported: the SCL-90-R subscales were each correlated with a similar MMPI clinical scale, except for the Masculine-feminine interests scale and the Mania scale. The Hypochondriasis scale of the MMPI and Somatization scale of the SCL-90-R showed the highest correlation at r=0.64. (Ref: 3) Also, each dimension of the SCL-90-R correlated with several MMPI scales (the Somatization and Depression dimensions were significantly correlated with nine scales of the MMPI). (Ref: 3)

The Depression, Interpersonal Sensitivity, Anxiety, and Global Severity Index dimensions of the SCL-90-R are correlated significantly with the Children’s Depression Inventory at 0.70-0.74. Other dimensions correlated at 0.40-0.64. (Ref: 4)

The SCL-90-R correlates at 0.45 with the Brief Psychiatric Rating Scale (Ref: 5) and 0.64 with the General Health Questionnaire-12. (Ref: 7) It has a correlation of 0.33 with the Impairment Scale (IS), used for making clinical diagnoses. (Ref: 9) The SCL-90-R correlates with the Beck Depression Inventory at 0.66 in a group of depressed inpatients and 0.69 in a group of inpatients with depression and other disorders. (Ref: 5)

Criterion-related/ Concurrent/ Predictive:

In a ROC analysis, the Anxiety and Depression subscales showed acceptable concurrent validity for those two diagnostic groups (AUC=0.86 and 0.81) as determined by DSM-III-R (ROC curve far above the diagonal). (Ref: 7) Further ROC analysis produced the following values: (Ref: 10)

Scale AUC
Anxiety 0.84
Depression 0.77
GSI 0.83
Hostility 0.67
Interpersonal Sensitivity 0.75
Obsessive-Compulsive 0.74
Paranoid Ideation 0.71
Phobic Anxiety 0.83
PST 0.85
Psychoticism 0.77
Somatization 0.80

Content:

No information found.

Responsiveness Evidence:

Sensitivity is reported for various cut-off scores: (Ref: 7)

Cut-off score Sensitivity
0.4 0.75
0.5 0.64
0.6 0.57
0.7 0.52
0.8 0.46
0.9 0.39

Scale Application in VA Populations:

No information found.

Scale Application in non-VA Populations:

Yes. (Ref: 2-10)

Comments


This article has been abstracted by trained METRIC staff and is awaiting final review from a METRIC psychometrician for comments and recommendations. If you find that the above information contains any errors or if you have any questions about our abstraction, please use the form below to bring them to our attention.



Updates

No information found.