How can computational tools help to understand language patterns in mental suffering considering social diversity

Area of Science:

• Computational linguistics
• Psychiatry
• Sociolinguistics

Background:

• Human behavior is influenced by complex biological and social factors.
• Language provides insights into cognitive processes.
• Previous computational methods for analyzing language in mental health have overlooked social determinants like education.

Purpose of the Study:

• To develop and evaluate a computational strategy for analyzing language structure.
• To investigate the impact of both psychiatric symptoms and formal education on language patterns.
• To address the need for incorporating social diversity into mental health computational strategies.

Main Methods:

• Utilized graph theory to measure the structural aspects of narrative mental organization in spontaneous oral reports.
• Analyzed narrative graph connectedness in relation to schizophrenia diagnosis and negative symptom severity.
• Compared language structure markers with typical cognitive and academic development, considering formal education's influence.

Main Results:

• A decrease in narrative graph connectedness was associated with schizophrenia diagnosis and negative symptom severity.
• The same connectedness markers increased with typical cognitive and academic development, suggesting an educational impact.
• Identified potential confounding effects of formal education on language structure markers previously linked to psychiatric conditions.

Conclusions:

• Computational language analysis in mental health must account for social factors like education.
• Integrating social diversity into computational strategies is crucial for accurate mental health assessments.
• Future research should explore the interplay between socioeconomic factors, language, and mental health conditions.