Elaboração e validação de um instrumento para avaliação de transtornos da personalidade

• Ano de defesa: 2016
• Autor(a) principal: Guimarães, Luize Anny Cardoso
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Ciências Exatas e da Saúde; Programa de Pós-Graduação em Modelos de Decisão e Saúde; UFPB
• Palavras-chave em Português: Transtornos da personalidade; Validade; Teoria de resposta ao Item; Personality disorders; Validity; Item response theory; CIENCIAS DA SAUDE::SAUDE COLETIVA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/9054

Personality disorders are defined as a persistent pattern of internal experience and behaviors that deviate markedly from the individual culture's expectations in a widespread and inflexibly way. Begins in adolescence and early adulthood, is stable over time and leads to injury. The types of personality disorders (PD) according to DSM-V are: Paranoid, Schizoid, Schizotypal, Antisocial, Borderline, Histrionic, Narcissistic, Avoidant, Dependent and Obsessive-Compulsive. The aim of this master thesis is to elaborate, validate and verify the psychometric properties of the items of an instrument to assess personality disorders. To do this we applied the Assessment Instrument of Personality Disorders (AIPD), developed and validated in this study in a sample of 470 people. The investigation verified the one-dimensionality of each item set of disorders through factor analysis for two forms of the instrument (a polytomic and a dichotomous). The factor analysis conducted by the principal-axis method and VARIMAX rotation resulted in a final matrix composed of nine factors that are organized as representative of the following personality disorders: Paranoid, Schizoid, Schizotypal, Antisocial, Borderline, Narcissistic, Avoidant, Dependent and Obsessive-Compulsive. For the Histrionic, there was no satisfactory psychometric items and in line with the literature to interpret this PD. The reliability of the measure in each disorder showed levels above 0.60, an acceptable value considering the low number of items in some factors. Regarding the instrument in dichotomized form, this did not show interpretable factors. The structure pointed to only one factor that explained 20% of the variance (considered low) and 0.955 trust index. The IRT analysis was conducted with polytomous format because it presented the best results. Three factors stood out as the amount of information offered on the latent trait, they are: Narcissistic PD, Dependent PD and Avoidant DP. Factors that have the best parameters of difficulty and discrimination were the Narcissistic and Avoidant, all between median and high. The teta values (ability) estimated from the subjects from the TRI are presented in this investigation as a possible tool for decision making about the presence or absence of a specific PD. The overall results are considered satisfactory and the validated instrument can be used as a tool for research in the evaluation of personality disorders.