Multi-Level Narrative Evaluation Outperforms Lexical Features for Mental Health

2026-04-30

AI Summary

This study addresses the fragmentation in current computational approaches to mental health text analysis, where dictionary-based methods overlook discourse structure and embedding models conflate local coherence with global narrative organization. To bridge this gap, the authors propose the first three-tiered computational framework that aligns with cognitive levels of narrative generation, integrating micro-level lexical features, meso-level semantic embeddings, and macro-level narrative assessments derived from large language models. Applied to 830 Chinese psychotherapy transcripts, the framework predicts depression, anxiety, and trauma symptoms by incorporating structured narrative features grounded in Labov’s narrative syntax, Rhetorical Structure Theory, and propositional compositionality. Results demonstrate that macro-level narrative structure itself carries critical clinical signals, significantly outperforming conventional lexical and embedding-based methods. Moreover, these structural features exhibit independent explanatory power, challenging the dominant paradigm reliant on word counts and offering testable hypotheses for clinical intervention and longitudinal research.

Abstract

How people narrate their experiences offers a window into how the mind organizes them. Computational approaches to therapeutic writing have evolved from lexical counting to neural methods, yet remain fragmented: dictionary tools miss discourse structure, while embeddings conflate local coherence with global organization. No existing framework maps these techniques onto the hierarchical processes through which narratives are constructed. Here we introduce a three-level framework - micro-level lexical features, meso-level semantic embeddings, and macro-level LLM narrative evaluation - and show, across 830 Chinese therapeutic texts spanning depression, anxiety, and trauma, that macro-level evaluation substantially outperforms lexical and embedding features for mental health prediction. This challenges the field's emphasis on word-counting: formal structural features (Labov's story grammar, RST coherence, propositional composition) demonstrate that narrative organization per se carries predictive signal, while clinically-grounded narrative dimensions capture how psychological states are expressed through discourse. Semantic embeddings add minimal independent value but yield incremental gains in multi-level classification. By grounding computational levels in discourse processing theory, this framework identifies macro-structural organization as the primary locus of clinical signal and generates testable hypotheses for intervention design and longitudinal research.

Problem

• narrative evaluation
• mental health prediction
• discourse structure
• computational framework
• therapeutic writing

Innovation

• multi-level narrative evaluation
• macro-structural organization
• therapeutic writing
• mental health prediction
• discourse processing theory