Date of Award
Doctor of Education
Counseling and Personnel
Dr. Robert L. Betz
Dr. Robert Hopkins
Dr. Frederick Gault
The study evaluated the concurrent validity of the Global Assessment Scale (GAS), a rating scale which is used to measure overall severity of psychiatric disturbance in program evaluation systems and in research studies. Validity was evaluated by comparing GAS ratings of 60 community mental health outpatients to scores on the MMPI.
The literature review indicated that the GAS has been used in over 90 studies. Thirty-one inter-rater reliability coefficients cited indicated an acceptable median reliability of .80, but studies relevant to the GAS's validity were sparse and marginally supportive. The GAS was utilized primarily to evaluate change due to psychotherapy or medication, or to compare severity of disturbance between experimental groups. Few studies originated from program evaluation settings, but a questionnaire sent to several state departments of mental health indicated that the GAS is used in statewide evaluation systems in five states.
GAS and MMPI data were collected at admission from adult outpatients at a representative community mental health center in a medium-sized city. GAS ratings made by therapists were correlated with four overall severity indexes from the MMPI, as well as with the individual MMPI scales. Correlations with the overall indexes ranged from .29 to .36, indicating minimal shared variance and questioning the GAS's validity as a measure of severity of disturbance. Correlations with the MMPI clinical scales were also under .36 and indicated that GAS scores are more related to psychotic symptoms than to neurotic ones. Both correlation data and a factor analysis demonstrated that the GAS was most highly related to MMPI scales measuring defensiveness and guardedness. This finding questioned the appropriateness of completing GAS ratings after a 1-hour intake session, and suggested that ratings made later in treatment may be more valid.
It was concluded that the GAS is a questionable measure of overall severity of disturbance because of its low concurrent correlations. Mean GAS scores of patient and normal groups in the literature and in the present study, however, appeared to have accuracy, suggesting that the GAS may have validity for comparing group mean scores. Using GAS scores as cutoffs or criterion measures for individuals, however, is not supported by the study's data and is discouraged.
Dekker, Daniel J., "A Study of the Validity of the Global Assessment Scale" (1983). Dissertations. 2428.