Assessing the dimensionality of the CES-D using multi-dimensional multi-level Rasch models.

Area of Science:

• Psychometrics
• Mental Health Assessment
• Statistical Modeling

Background:

• The Center for Epidemiologic Studies Depression Scale (CES-D) is a common depression screening tool.
• Previous psychometric analyses primarily used factor analysis, with limited application of Rasch models, especially multidimensional ones.

Purpose of the Study:

• To apply a multidimensional Rasch model to evaluate the psychometric properties of the CES-D.
• To compare the fit of one-dimensional versus four-dimensional models of the CES-D.
• To investigate the influence of demographic variables (gender, age) on the CES-D structure.

Main Methods:

• Utilized a multidimensional Rasch model on a representative sample of 518 Austrian adults (aged 18-65).
• Compared a one-dimensional CES-D model with two four-dimensional models (one reflecting subscales, another including background variables).

Main Results:

• The four-dimensional model demonstrated a significantly better fit than the one-dimensional model.
• Reliability indices were generally satisfactory, with latent correlations ranging from 0.31 to 0.88.
• Identified specific subscale weaknesses (Interpersonal Difficulties) and unexpected results in Positive Affect, with limited gender effects.

Conclusions:

• A multidimensional approach to the CES-D provides preferable differentiation over a one-dimensional score.
• The CES-D is recommended for screening, but modifications are suggested based on identified psychometric weaknesses.
• Further refinement of the CES-D, particularly concerning specific subscales, may enhance its utility.