Event Title
Comparison of Three Different Methods for Missing Data Estimation in Growth Curve Model: FIML, Multiple Imputation, MIPCA
Start Date
3-2-2021 4:10 PM
End Date
3-2-2021 5:40 PM
Date of Publication
2-3-2021
Document Type
Presentation
Abstract
The present study aims to evaluate the performance of three different modern missing data estimation methods available for growth curve modeling ¬– full information maximum likelihood and two different multiple imputation strategies. As missing data estimation in growth curve model requires missing at (completely) random assumption, the result of three methods from violated data (i.e., missing not at random) would be investigated as well. The result of a Monte Carlo simulation study will provide methodological guidelines for practice of GCM under diverse missing data environments.
Keywords
FIML, Multiple Imputation, MIPCA
Description
Methodology, Discussant: Grant Morgan
Persistent Identifier
http://hdl.handle.net/10950/2846
Comparison of Three Different Methods for Missing Data Estimation in Growth Curve Model: FIML, Multiple Imputation, MIPCA
The present study aims to evaluate the performance of three different modern missing data estimation methods available for growth curve modeling ¬– full information maximum likelihood and two different multiple imputation strategies. As missing data estimation in growth curve model requires missing at (completely) random assumption, the result of three methods from violated data (i.e., missing not at random) would be investigated as well. The result of a Monte Carlo simulation study will provide methodological guidelines for practice of GCM under diverse missing data environments.