Development and validation of the Rasch-based Depression Screening (DESC) using Rasch analysis and structural...

Area of Science:

• Psychometric assessment
• Clinical psychology
• Medical statistics

Background:

• Traditional depression assessment tools can be improved with modern psychometric techniques.
• Item-response theory and structural equation modeling offer advanced methods for test construction.

Purpose of the Study:

• To develop and validate two parallel 10-item depression screening questionnaires (DESC-I and DESC-II).
• To ensure the scales are based on a Rasch homogeneous item bank and demonstrate unidimensionality.
• To establish cut-off scores for sensitivity and specificity in depression screening.

Main Methods:

• Combined Rasch analysis and structural equation modeling.
• Utilized patient samples with mental and somatic diseases.
• Employed Receiver Operating Characteristic (ROC) analyses to determine cut-off scores.

Main Results:

• Developed two parallel, Rasch-based depression screening questionnaires (DESC-I and DESC-II).
• Both scales demonstrated unidimensionality and good model fit.
• Established cut-off scores with high sensitivity and specificity.

Conclusions:

• The DESC questionnaires are suitable for screening depression in diverse patient populations.
• These tools can be beneficial for monitoring depression status through repeated measurements.
• Modern test construction methods enhance the reliability and utility of depression assessment.