An alternative to classical latent class models selection methods for sparse binary data: An illustration with simulated data

C. Araya Alpizar

Within the context of a latent class model with manifest binary variables, we propose an alternative method that solves the problem of estimating empirical distribution with sparse contingency tables.
We analyze sparse binary data, where there are many response patterns with very small expected frequencies in several data sets varying in degree of sparseness from 1 to 5 defined d=n/R..
Results from the proposal presented compare the rates of Type I for traditional goodness-of-fit tests. We also show that with data density d< 5, Pearson's statistic should not be used to select latent class models using the Patterns Method, given that this has the probability of Type I error being greater than 5%.
The parametric bootstrap require both knowledge of advanced statistics. Meanwhile, the Patterns Method is presented as a rapid, simple, and labour-saving technique to provide tables of critical for diagnosis models.

Palabras clave: Sparse data; latent class; goodness-of-fit; binary data

L05.1 Grupo de Análisis Multivariante y Clasificación I
5 de septiembre de 2016  11:30
0.02 - Aula de proyectos 1