Abstract
This dissertation aims to develop an analytical framework that bridges the long-standing divide between nomothetic and idiographic approaches to defining the structure of personality and other individual differences. The nomothetic tradition seeks to explain and predict behavior through stable, universal dimensions of personality, while the idiographic approach emphasizes the fine-grained, dynamic processes unique to individuals. Although theoretical advances have begun to reconcile these perspectives, analytical methods capable of integrating them remain underdeveloped. To address this gap, I introduce the Dynamic Individualized Latent Trait (DILT) model, a framework that decomposes ordinal item responses from intensive longitudinal data into three components: 1) the nomothetic trait structure shared across individuals, 2) person-specific trait organizations of state manifestations that capture how each individual's personality items uniquely covary over time, and 3) latent dynamic processes that govern the temporal persistence and situational responsiveness of daily personality fluctuations. Model are trained in a two-phase procedure to ensure parameters' identifiability and comparability across individuals. I conduct a simulation study validates structural recovery across varying sample sizes and signal strengths. Applied to experience sampling data from 93 individuals assessed on 45 personality items, the empirical study reveals that while individuals share a common nomothetic trait structure, they differ substantially in how their personality states organize and unfold in daily life. In doing so, this framework offers a unified and scalable solution for studying personality as both a stable taxonomy and a dynamic process, meeting the growing demand for integrated person-centered and generalizable models.
Committee Chair
Joshua Jackson
Committee Members
Emily Willroth; Jacob Montgomery; John Rauthmann; Patrick Hill
Degree
Doctor of Philosophy (PhD)
Author's Department
Psychology
Document Type
Dissertation
Date of Award
4-27-2026
Language
English (en)
DOI
https://doi.org/10.7936/saeh-f413
Author's ORCID
https://orcid.org/0000-0003-3939-9568
Recommended Citation
Xi, Muchen, "Dynamic Individualized Latent Trait (DILT) Model: Bridging Nomothetic and Idiographic Psychometrics" (2026). Arts & Sciences Graduate Student Theses and Dissertations. 3785.
The definitive version is available at https://doi.org/10.7936/saeh-f413