Search

The purpose of this study was to analyse, using the Rasch Rating Scale Model, an instrument of obsessive-compulsive symptoms (ISOC), presented in 2008 by Rosa and Baptista, that consists of 105 items of characteristic symptoms of obsessive-compulsive disorder. Two hundred and twenty subjects (87 diagnosed with OCD and 133 without this disorder) undertook the ISOC (Rosa & Baptista, 2008) and the SCL-90-R (Derogatis, 1977). It was confirmed that the response categories were adequate, although 27 items had to be removed since they showed differential functioning between groups (with OCD/ without OCD) and/or sex. The final version of the inventory consists of 78 items with five response categories that have met the criteria set (Linacre, 2002) and shows adequate psychometric qualities. Thus, the results obtained indicate that the ISOC is a reliable, valid and useful measure, useful as a screening instrument to assess obsessive-compulsive symptoms., O objetivo deste estudo foi analisar, com o Modelo de Rasch, um inventário de sintomas obsessivo-compulsivo (ISOC), trabalhado por Rosa e Baptista, em 2008, que consiste de 105 itens de sintomas característicos da perturbação obsessivo-compulsiva (POC). Duzentos evinte pacientes (87 com diagnóstico de POC e 133 sem esta perturbação) preencheram o ISOC (Rosa & Baptista, 2008) e a SCL-90-R (Derogatis, 1977). Verificou-se que as categorias de resposta se revelaram adequadas, embora 27 itens tenham mostrado funcionamentodiferencial entre grupos (com POC/sem POC) e entre sexos, pelo que foram excluídos. A versão final do inventário ficou composta por 78 itens com cinco categorias de resposta que cumpriram os critérios estabelecidos (Linacre, 2002) e apresentaram qualidades psicométricasadequadas. Assim, os resultados evidenciaram que o ISOC é uma medida válida, confiável e útil como instrumento de screening para avaliar os sintomas obsessivo-compulsivos.

Resumen Reconocer y denominar expresiones emocionales son habilidades en cuyo estudio existe un creciente interés por parte de la psicología y las neurociencias; trastornos tales como el autismo, la esquizofrenia y la depresión se caracterizan por dificultades en reconocer las emociones ajenas. El hallazgo de diferencias a favor de las mujeres en Percepción Emocional (PE) ha sido replicado transculturalmente por lo que sirve como evidencia de validez. En esta investigación, de diseño instrumental, se construyeron 28 ítems y se pusieron a prueba mediante el modelo de Rasch (MR) en una muestra de población general española (102 varones y 102 mujeres; rango de edad=18-65). El ajuste datos-modelo fue suficiente. Las medidas de las personas muestran que el test resultó fácil (rango = -.60 a 4.14 logits). Se hallaron las diferencias sexuales predichas, d= -.36 (d con corrección de la atenuación= -.50); que ningún ítem mostrase funcionamiento diferencial (DIF) asociado al sexo y que la variable construida tenga propiedades intervalares añade confianza al valor estimado del tamaño del efecto. La construcción de una escala de intervalo en la que se localizan conjuntamente ítems y personas ofrece información cuantitativa y permite detectar patrones inesperados de respuesta que pueden ser de utilidad para el diagnóstico neuropsicológico. Abstract Recognizing and labeling emotion expressions are abilities in whose study psychology and the neurosciences are increasingly interested; disorders such as autism, schizophrenia and depression are characterized by difficulties in recognizing others' emotions. The finding of sex differences favoring women in Emotion Perception (EP) has been replicated cross-culturally, and so it can be used as validity evidence. In this instrumental design study, some 28 items were constructed and tested by means of the Rasch model (RM) in a Spanish community sample (102males and 102 females; age range=18-65). Data-model fit was good enough. Person measures show that the test was easy (range = -.60 to 4.14 logits). Even so, the predicted sex differences favoring women were found, d= -.36 (disattenuated d= -.50). That no item showed sex-related Differential Item Functioning (DIF) and that the constructed variable has interval scale properties add confidence to the effect size estimated value. The construction of an interval scale on which EP items and persons are conjointly represented offers quantitative information and allows the discovery of unexpected response patterns.

The aim of this study was to construct emotion understanding items from the Rasch Model approach by experimenting through comparison between verbal /image response format. The participants were 204 subjects from a Spanish community sample. A randomized experiment was carried out to test the effect of response format (verbal/image) and participants’ gender on the emotion understanding Rasch measurements. The effect on item difficulty of the social distance (close/far) was also contrasted. No interaction effect was found. Response format had a significant effect on measurement regardless of gender: the verbal response format was easier than the image one. There were significant gender differences on emotion understanding favoring women. Items describing situations with close receivers were significantly easier than the items showing far recipients. El objetivo fue construir ítems de comprensión emocional desde la perspectiva del Modelo de Rasch, comparando mediante un experimento aleatorizado los formatos de respuesta verbal/imagen. Participaron 204 sujetos de población general. Se contrastó el efecto del formato (verbal/imagen) y del género en las medidas Rasch de comprensión de emociones de los participantes y se analizó el efecto de la distancia social (cerca/lejos) sobre la dificultad Rasch de los ítems. El formato de respuesta tuvo un efecto significativo sobre las medidas independientemente del género: el formato verbal resultó más fácil que el de imagen. El efecto del género fue significativo, favorable a las mujeres. Los ítems que describían situaciones con receptores cercanos resultaron más fáciles que los que incluían receptores lejanos.