POLYCHORIC AND TETRACHORIC CORRELATIONS IN EXPLORATORY AND CONFIRMATORY FACTORIAL STUDIES

Scientific advances and software development have increased the amount of exploratory and confirmatory studies in psychometric research. Regularly these studies use Pearson´s correlation coefficient, which was originally conceived to be used with continuous variables, and later was extended to categ...

Descripción completa

Detalles Bibliográficos
Autores: Freiberg Hoffmann, Agustín, Stover, Juliana Beatriz, de la Iglesia, Guadalupe, Fernández Liporace, Mercedes
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2013
País:Uruguay
Institución:Universidad Católica del Uruguay
Repositorio:LIBERI
Idioma:español
OAI Identifier:oai:liberi.ucu.edu.uy:10895/5346
Acceso en línea:https://revistas.ucu.edu.uy/index.php/cienciaspsicologicas/article/view/1057
https://hdl.handle.net/10895/5346
Access Level:acceso abierto
Palabra clave:correlaciones tetracóricas
correlaciones policóricas
análisis factorial
variables categóricas
ítems ordinales
ítems dicotómicos
tetrachoric correlations
polychoric correlations
factor analysis
categorical variables
ordinal items
dichotomous items
Descripción
Sumario:Scientific advances and software development have increased the amount of exploratory and confirmatory studies in psychometric research. Regularly these studies use Pearson´s correlation coefficient, which was originally conceived to be used with continuous variables, and later was extended to categorical items (dichotomous or polytomous). Currently, improved statistical packages allow scientists to carry on robust procedures, designed specifically for categorical variables, among which stand out tetrachoric and polychoric correlations. Since most of psychometric scales consist in dichotomous and polytomous items (mainly Likert formats), the analysis of such correlations becomes methodologically relevant. This paper presents some peculiarities about the use of these statistics, different software packages to facilitate their implementation, as well as the usual problems associated with its employment, and the possible solutions are discussed.