|
Activity Number:
|
443
- Latent Variables, Causal Inference, Machine Learning and Other Topics in Mental Health Statistics
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 12, 2021 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
WNAR
|
|
Abstract #317816
|
|
|
Title:
|
Latent Trait Shared-Parameter Mixed-Models for Missing Ordinal Ecological Momentary Assessment Data
|
|
Author(s):
|
John Cursio* and Donald Hedeker
|
|
Companies:
|
University of Chicago and University of Chicago
|
|
Keywords:
|
Item response theory;
latent trait;
shared parameter;
ordinal;
bivariate;
longitudinal
|
|
Abstract:
|
A shared parameter model for longitudinal ordinal data collected by ecological momentary assessment (EMA) with missing responses is presented. Mood outcomes were reported by high school students over a period of one week. In this approach, a latent trait representing the responsiveness of subjects is estimated in an item response theory (IRT) sub-model and also used as a covariate in a bivariate sub-model for mood outcomes. Both likelihood-based and Bayesian approaches are shown using statistical software. In the full shared-parameter model, the latent trait of responsiveness is a significant predictor of two mood outcomes and was associated with improved moods. The model offers an advantage over missing at random approaches previously used with longitudinal ordinal EMA data with missing outcomes.
|
Authors who are presenting talks have a * after their name.
