Use of Latent Class Analysis to define groups based on validity, cognition, and emotional functioning

Area of Science:

• Neuropsychology
• Psychometrics
• Statistical Modeling

Background:

• Neuropsychological evaluations often involve heterogeneous samples.
• Understanding distinct functioning patterns is crucial for accurate assessment.
• Existing methods may not fully capture complex subgroupings.

Purpose of the Study:

• To classify a heterogeneous sample of neuropsychological data using Latent Class Analysis (LCA).
• To identify latent groups based on performance validity, symptom validity, cognition, and emotional functioning.
• To describe the characteristics of these latent groups.

Main Methods:

• Latent Class Analysis (LCA) was applied to a dataset of 680 neuropsychological evaluation protocols.
• Data originated from clinical evaluations conducted at an urban medical center.
• Measures included performance validity, symptom validity, cognitive, and emotional functioning.

Main Results:

• A four-class model provided the best fit for the data.
• Classes were distinguished by performance and symptom validity measures.
• Performance invalidity correlated with poorer cognitive performance; symptom invalidity correlated with greater emotional distress.

Conclusions:

• Latent Class Analysis is effective for identifying distinct performance patterns in heterogeneous neuropsychological samples.
• The findings support the orthogonal nature of performance and symptom validities.
• LCA aids in understanding complex patient profiles in clinical neuropsychology.