The Effect of the Number of Answer Choices on the Psychometric Properties of Stress Measurement in an Instrument Applied to Children

The main objective of this study was to use Item Response Theory (IRT) models to measure the effect exerted by the number of response options on the psychometric properties of a test measuring stress in children. In this study, we applied the 30-item Child Stress Perception Inventory (CSPI) scale to...

Full description

Bibliographic Details
Main Authors: González-Betanzos, Fabiola, Leenen, Iwin, Lira-Mandujano, Jennifer, Vega-Valero, Zaira
Format: Online
Language:spa
Published: Instituto de Investigaciones Psicológicas (IIPSI, Conicet-UNC) 2012
Subjects:
Online Access:https://revistas.unc.edu.ar/index.php/revaluar/article/view/4694
Description
Summary:The main objective of this study was to use Item Response Theory (IRT) models to measure the effect exerted by the number of response options on the psychometric properties of a test measuring stress in children. In this study, we applied the 30-item Child Stress Perception Inventory (CSPI) scale to 583 children; the items have different response alternatives (3, 5, or 7). We studied whether the scales measure the same trait and whether the alternatives that the same items possess are equivalent. As evidenceof validity, we present measurements that examine the internal structure of the instrument and its relationship with other variables. The result indicates that the three forms measure the same trait, but that there is no equivalency among the categories.The scale adjustment of 7 response alternatives is best; however, validity in relation to other variables is optimal for 5 response alternatives, which in addition, performs best in terms of reliability and information.