Models and Strategies for Factor Mixture Analysis: An Example Concerning the Structure Underlying Psychological...

Area of Science:

• Psychometrics
• Psychopathology Research
• Statistical Modeling

Background:

• The factor mixture model (FMM) combines categorical and continuous latent variables.
• FMMs are suitable for modeling psychopathology due to their ability to represent both diagnostic classes and severity dimensions concurrently.
• Despite conceptual explanations, FMMs are underutilized in practice due to limited guidance on application and interpretation.

Purpose of the Study:

• To explore the practical application and interpretation of the factor mixture model (FMM).
• To provide guidance on building and selecting FMMs by examining a real-world dataset on conduct disorder.
• To elucidate different FMM formulations and decision-making processes when comparing FMMs to alternative models.

Main Methods:

• Application of the factor mixture model (FMM) to a real data example concerning conduct disorder.
• Detailed explanation of various FMM formulations and the steps involved in constructing an FMM.
• Comparative analysis to differentiate FMMs from alternative statistical models.

Main Results:

• Demonstration of how FMMs can simultaneously model diagnostic class membership and dimensional severity in psychopathology.
• Illustrative example showing the practical steps for building and interpreting an FMM using conduct disorder data.
• Guidance provided on selecting an FMM over alternative models based on data characteristics and research questions.

Conclusions:

• Factor mixture models offer a robust framework for understanding the complex, dual nature of psychopathology.
• This paper provides practical insights and methodological guidance for researchers to effectively implement and interpret FMMs.
• The study highlights the utility of FMMs in advancing the analysis of psychiatric disorders, specifically conduct disorder.