|
Activity Number:
|
340
- SPEED: SPAAC SESSION III
|
|
Type:
|
Topic-Contributed
|
|
Date/Time:
|
Thursday, August 12, 2021 : 10:00 AM to 11:50 AM
|
|
Sponsor:
|
ENAR
|
|
Abstract #318033
|
|
|
Title:
|
Mediation Analysis for Zero-Inflated Mediators with a Lognormal Distribution
|
|
Author(s):
|
Meilin Jiang* and Zhigang Li
|
|
Companies:
|
University of Florida and University of Florida
|
|
Keywords:
|
Mediation;
Zero-inflated mediators
|
|
Abstract:
|
Mediation analysis is an important statistical technique for investigating intermediate variables, mediators, in a causal pathway. There is an unmet need for mediation analysis methods to deal with mediators with zero-inflated distributions. We propose a new method to estimate and test mediation effects for mediators with zero-inflated lognormal distributions that have two-part density functions. The mediation effects can be decomposed into two components. The first component is the mediation effect through the binary change of the mediator from zero to a non-zero status and the second component is the mediation effect through the numeric change of the mediator on its continuous scale. The excessive zero counts could be a mixture of true zeros and false zeros since non-zero values may be observed as zeros due to technical errors or limit of detection. The chance of observing false zeros is modelled by a probabilistic mechanism and all parameters are estimated with MLE. A comprehensive simulation study is conducted to evaluate the model performance along with existing methods. The model is also applied to a neuroscience dataset in comparison with standard mediation analysis methods.
|
Authors who are presenting talks have a * after their name.
