Bayesian inference for multidimensional item response models under heavy tail skewed latent trait distributions and link functions

• Grant number: 13/26336-5
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: April 01, 2014
• End date: March 31, 2018
• Field of knowledge: Physical Sciences and Mathematics - Probability and Statistics - Statistics
• Principal Investigator: Caio Lucidius Naberezny Azevedo
• Grantee: Juan Leonardo Padilla Gomez
• Host Institution: Instituto de Matemática, Estatística e Computação Científica (IMECC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 12/21788-2 - Regression models and applications, AP.TEM
• Associated scholarship(s): 17/05326-2 - Asymmetric multidimensional Item response theory models: estimation and diagnostic methods under a Bayesian perspective, BE.EP.DR

Abstract

This project aims to extend the classes of multidimensional Item Response Theory models already considered in the literature. We will consider the scenario where subjects, belonging to different groups, answer multidimensional tests (in a broad sense of the word) with dichotomous items (or dichotomized). Such tests can be different from each other but they must present some structure of common items. Additionally, it will be considered available collateral information (associated factors), in the modeling of the latent traits, such as gender or intended course, as occurs at admissional tests, for example. The modeling will correspond to: the using of a version the multivariate centered skew Student t distribution (to be developed and studied) to model the latent traits, a link function based on the univariate centered skew normal distribution, adding a level in the whole modeling by using a regression structure in order to relate the latent traits and their associated factors. It will be developed bayesian mechanisms for both estimation and model validation. Also, it will be considered simulation studies and real data analysis.

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

GOMEZ, Juan Leonardo Padilla. Inferência Bayesiana para modelos de resposta ao item multidimensionais sob distribuições para os traços latentes e funções de ligação assimétricas de caudas pesadas. 2018. Doctoral Thesis - Universidade Estadual de Campinas (UNICAMP). Instituto de Matemática, Estatística e Computação Científica Campinas, SP.