Acceso usuarios
Diego Blum, Facundo Juan Pablo Abal, María Silvia Galibert, Horacio Félix Attorresi
Se presentan las caracteristicas psicometricas de una Prueba de Analogias Figurales desde el marco de la Teoria Clasica de Tests. Se construyeron matrices de figuras de 2x2. Se utilizaron las reglas de relacion de rotacion, traslacion, reflejo, distorsion del tamano, distorsion de la forma, adicion y sustraccion para la elaboracion de reactivos con una regla, dos reglas y tres reglas de relacion. Participaron 475 estudiantes de Psicologia de la Universidad de Buenos Aires, de los cuales 96 eran varones y 379 mujeres. El analisis de la consistencia interna brinda un indice altamente satisfactorio (�¿ = .91), las correlaciones item-test corregidas puntuan todas por encima de .30 y se corrobora la unidimensionalidad de la Prueba por medio del analisis exploratorio. Tambien se detecta una diferencia intergenero significativa en el puntaje total, a favor de los varones (p < .001). Entre otras cuestiones, se discuten las condiciones para graduar la dificultad del item y limitaciones potenciales.
This paper presents the psychometric characteristics of a Test of Figural Analogies from the Classical Test Theory perspective. 2x2 figural matrices were constructed. The rules of relation of rotation, displacement, reflection, size distortion, shape distortion, addition and subtraction were implemented to elaborate one rule-based, two rule-based and tree rule-based items. 475 Psychology students from the University of Buenos Aires participated, 96 of whom were males and 379 females. The internal consistency analysis provides a highly satisfactory score (á = .91), all item-test corrected correlations are above .30 and the Test�s unidimensionality was confirmed through the exploratory analysis. Student t test shows a significant between-gender difference in the total score, favouring men (p < .001). Among other issues, conditions for increasing items� difficulty level are discussed as well as potential limitations.
