APLICACIÓN DE LA TEORÍA DE RESPUESTA AL ÍTEM A LA ESCALA NEUROTICISMO DEL BIG FIVE INVENTORY
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Ajayu
Abstract
El impacto extendido que tiene el Neuroticismo en el sistema de salud ha motivado su medición para ser usada en tareas de prevención. Por su brevedad y su reconocimiento internacional, la escala de Neuroticismo del Big Five Inventory presenta características óptimas para estas aplicaciones. En este trabajo se propone profundizar el estudio de las propiedades psicométricas de esta escala utilizando Teoría de Respuesta al ítem (TRI). Colaboraron 742 adultos de Buenos Aires, Argentina (55.3% mujeres). Un primer análisis con TRI reveló la necesidad de eliminar uno de los ítems originales por presentar funcionamiento inadecuado. Para los 7 ítems restantes se verificaron los supuestos de unidimensionalidad e independencia local. Todos mostraron un ajuste aceptable al Modelo de Respuesta Graduada y una capacidad discriminativa media-alta. Tampoco se observó funcionamiento diferencial en función del género. Se hallaron correlaciones moderadas con las escalas del Inventario SCL-90-R. La Función de Información del Test mostró que la precisión es aceptable aun para los valores extremos del rasgo, los cuales son de especial interés para las aplicaciones del ámbito clínico-epidemiológico.
The widespread impact that Neuroticism has on the health system has motivated its measurement to be used in prevention tasks. Due to its brevity and international recognition, the Neuroticism scale of the Big Five Inventory presents optimal characteristics for these applications. In this work it is proposed to deepen the study of the psychometric properties of this scale using Item Response Theory (IRT). There were 742 adults from Buenos Aires, Argentina (55.3% women). At first IRT analysis revealed the need to eliminate one of the original items due to inadequate functioning. For the remaining 7 items, the assumptions of unidimensionality and local independence were verified. All showed an acceptable fit to the Graded Response Model and a medium-high discriminatory capacity. Gender differential functioning was not observed. Moderate correlations were found with the scales of the SCL-90-R Inventory. The Test Information Function showed that precision is acceptable even for extreme trait values, which are of special interest for clinical-epidemiological applications.
The widespread impact that Neuroticism has on the health system has motivated its measurement to be used in prevention tasks. Due to its brevity and international recognition, the Neuroticism scale of the Big Five Inventory presents optimal characteristics for these applications. In this work it is proposed to deepen the study of the psychometric properties of this scale using Item Response Theory (IRT). There were 742 adults from Buenos Aires, Argentina (55.3% women). At first IRT analysis revealed the need to eliminate one of the original items due to inadequate functioning. For the remaining 7 items, the assumptions of unidimensionality and local independence were verified. All showed an acceptable fit to the Graded Response Model and a medium-high discriminatory capacity. Gender differential functioning was not observed. Moderate correlations were found with the scales of the SCL-90-R Inventory. The Test Information Function showed that precision is acceptable even for extreme trait values, which are of special interest for clinical-epidemiological applications.
Description
Vol. 17, No. 2