Title: Robust estimation of multilevel models with sufficient sample sizes
Authors: Ronrick Da-ano - University of the Philippines-Diliman (Philippines) [presenting]
Abstract: The aim is to estimate multilevel models with sufficient sample sizes with structural change using a hybrid of forward search algorithm preceding bootstrap method. The influence of the different sample sizes at the group level on the accuracy of the estimates and its standard error was examined. The simulation study exhibits a comparable predictive ability with the classical restricted maximum likelihood estimation (REML). However, the hybrid method yields estimates of the parameters with lower bias in respect to REML. The hybrid of forward search algorithm preceding bootstrap method can even provide more robust estimates of fixed and random coefficients under various levels of interclass correlation or in the presence of structural change.