A0329
Title: Diagnostic techniques for Heckman selection models
Authors: Marcos Oliveira - Federal University of Sao Joao Del Rei (Brazil) [presenting]
Marcos Prates - Universidade Federal de Minas Gerais (Brazil)
Christian Eduardo Galarza Morales - Escuela Superior Politecnica del Litoral (Ecuador)
Victor Hugo Lachos Davila - University of Connecticut (United States)
Abstract: Diagnostic techniques for Heckman selection models estimated using the EM algorithm are presented. The focus is on the selection-t and normal models, based on the bivariate Student-t and bivariate normal distributions, respectively. The Heckman selection model is a key econometric tool for estimating relationships while addressing selection bias. Relying on the EM-type algorithm, global and local influence analyses are developed based on the conditional expectation of the complete-data log-likelihood function, exploring four perturbation schemes for local influence analysis. The effectiveness of the proposed diagnostic measures for identifying influential observations is evaluated through a simulation study, along with a real-data application that illustrates how these techniques can effectively identify influential points. The algorithms and methodologies developed are integrated into the R package HeckmanEM.
