ABSTRACT
Las creencias y su valoración son fundamentales a investigar para explicar la etiología y mantenimiento del trastorno obsesivo-compulsivo de acuerdo a modelos actuales. Éste estudio presenta el desarrollo y validación de una versión corta del ICO (Inventario de Creencias Obsesivas, Belloch et al. 2003) en población mexicana. Para ello se realizaron dos estudios con grupos de universitarios. Estudio 1: Análisis factorial exploratorio de la versión reducida (200 participantes no clínicos). Reveló una estructura factorial diferente al instrumento original. Estudio 2: Análisis de la estructura factorial latente de la versión reducida mediante un análisis multigrupo (200 y 202 participantes no clínicos). La mejor solución factorial posible fue de tres factores con 20 ítems, y propiedades psicométricas semejantes a una versión inglesa (OBQ, Obsessional Beliefs Questionaire, 2003). Por su consistencia interna y repetibilidad, la versión mexicana es adecuada para la evaluación de creencias disfuncionales, pero se necesitan estudios con muestras clínicas.
Recent models emphasized the importance in research of beliefs and its appraisal in the etiology and maintenance of OCD. The aim of the present study is the development and validation of a short form of the Obsessive Beliefs Questionnaire (OBQ, Belloch et al. 2003) in Mexican population. This topic was addressed through two different studies. Study 1: Exploratory factor analysis of the short version (200 non clinical subjects). The results show a different factor solution. Study 2: Testing for the equivalence of latent mean structures in multi-group analyses (200 and 202 non clinical subjects). A three-factor, 20 items, model emerged as the best factorial solution with similar psychometric properties to those of an English version (OBQ, Obsessive Beliefs Questionnaire, 2003). Given its consistency and repeatability, the Mexican version of the OBQ is a suitable instrument for the assessment of dysfunctional beliefs, although studies with clinical samples are needed.
ABSTRACT
La investigación que se informa se inserta en el marco de una tesis doctoral cuyo objetivo es evaluar el perfil cognitivo de los pacientes con Trastorno Obsesivo Compulsivo (TOC). Se describen dos estudios: una adaptación lingüística y conceptual del Cuestionario de Creencias Obsesivas (Obsessive Compulsive Cognitions Working Group, 2005) y una evaluación de las características psicométricas del instrumento desarrollado. Como resultado de ambos estudios se llegó a una versión revisada y abreviada del instrumento. Los análisis realizados arrojaron una solución factorial de cuatro componentes, los cuales conformaron cuatro subescalas denominadas: (a) Perfeccionismo y necesidad de certeza (PNC), (b) Importancia otorgada a los pensamientos intrusivos y necesidad de controlarlos (IPC), (c) Responsabilidad por daño (RD) y (d) Sobrestimación del peligro (SP). Dicha composición factorial alcanzó a explicar más del 60% de la variancia total y fue similar a la encontrada por los autores de la versión original, exceptuando el hecho de que las subescalas de SP y RD constituyeron dos factores independientes, mientras que en la versión original se agruparon en uno solo. Los índices de consistencia interna (α = .937) y la confiabilidad test-retest (r = .83) fueron excelentes. Por otro lado, los resultados de las pruebas de validez convergente y discriminante fueron muy satisfactorios. Los datos expuestos son sólo preliminares, se necesitan investigaciones posteriores que tiendan a ampliar los estudios sobre su validez.
This article is based on a doctoral thesis framework which assesses cognitive profiles in Obsessive Compulsive Disorder (OCD) patients. We have divided the process into the following two studies: a linguistic and conceptual adaptation of the Obsessive Belief Questionnaire - OBQ (Obsessive Compulsive Cognitions Working Group, 2005) and an assessment of the psychometric characteristics of the adapted questionnaire. In the first study, original OBQ items were translated into Spanish and then back into English, as is generally suggested by cross-cultural research methods. We achieved a revised and abbreviated version of the instrument consisting of 31 items. The results, based on a sample of 260 community controls (120 men and 140 women), were factor analyzed and divided into four sub-scales called: (a) Perfectionism and intolerance for uncertainty (PIC), (b) Importance and control of intrusive thoughts (ICT), (c) Responsibility (R), and (d) Overestimation of threat (OT). The above mentioned factorial categories accounted for more than 60% of total variance and were similar to those found by the authors of the original version, with the exception that OT and R sub-scales became two independent factors, while in the original version they were grouped in only one factor. Internal consistency and test/re-test reliability, assessed over a 30 day period, were found to be excellent (α = .937; r = .83, respectively). In the second study, the OBQ-31, the Beck Depression Inventory and the Yale-Brown Obsessive-Compulsive Scale were taken by 30 patients diagnosed with obsessive compulsive disorder (OCD), 40 patients with other anxiety disorders (AC), and 120 community controls (CC). We also factor analyzed data from this total sample (N = 190) and obtained identical results as in the first study. In order to examine the convergent and discriminant validity of the OBQ-31, we did a series of analyses. On the one hand, the results of a one-way ANOVA showed significant differences across groups (F = 79.853; gl = 2; p < .000) in relation to the total scores of OBQ-31. Post hoc comparisons showed that OCD group patients had significantly higher scores than both AC and CC groups. The AC group, in turn, scored significantly higher than the CC group. Furthermore, a Pearson correlation between the Y-Bocs and the OBQ-31 scores of the OCD group was calculated, resulting in a significant correlation between both instruments (r = .715; p < .004). Our next step involved using the total sample (N = 190) to explore correlations between the OBQ-31 and the BDI scores; however, our findings were low and non significant (r = .161; p = .80). Later, the coefficient was estimated again, this time considering scores obtained by both clinical groups (n = 70). The correlation coefficient was low again, although significant at a statistical level (r = .31; p < .05). Consequently, both convergent and discriminant validity analyses proved satisfactory. The obtained data and results are preliminary, and will require further research to solidify the validity of OBQ-31.