Generalizability studies of the Global Assessment of Functioning-Split version

Area of Science:

• Psychometrics
• Clinical Psychology
• Psychiatric Assessment

Background:

• The Global Assessment of Functioning (GAF) is a widely used scale for assessing symptom and functioning levels in psychiatric patients.
• Investigating the reliability and precision of assessment tools is crucial for accurate clinical evaluation and research.

Purpose of the Study:

• To apply Generalizability Theory to evaluate the reliability and precision of the split-version Global Assessment of Functioning (GAF) scale.
• To identify optimal measurement designs for GAF assessments.

Main Methods:

• Employed Generalizability (G) studies and decision studies on GAF symptom and function scores.
• Utilized data from 6 case vignettes assessed by two independent samples of experienced raters (n=19 and n=58) from diverse treatment settings.
• Analyzed variance components to quantify error and assess reliability and precision.

Main Results:

• Both GAF symptom and function scores demonstrated high generalizability.
• A measurement design involving 2 raters per subject was identified as the most efficient in balancing reliability, precision, and resource utilization.
• Error variance components were successfully disentangled, providing insights into score reliability.

Conclusions:

• GAF symptom and function scores exhibit high consistency among experienced raters.
• The findings support the reliability and precision of the GAF scale when administered with appropriate measurement designs.
• A two-rater design offers an efficient approach for reliable GAF assessments.