Análise de instrumentos de mensuração do estresse em gestantes: Escala de Estresse Percebido (PSS) e do Inventário de Eventos de Vida Produtores de Estresse (IEVPE)

• Ano de defesa: 2017
• Autor(a) principal: Rizzini, Marta
• Orientador(a): SANTOS, Alcione Miranda dos
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Maranhão
• Programa de Pós-Graduação: PROGRAMA DE PÓS-GRADUAÇÃO EM SAÚDE COLETIVA/CCBS
• Departamento: DEPARTAMENTO DE MEDICINA I/CCBS
• País: Brasil
• Palavras-chave em Português: Estresse; Gestante; Análise fatorial; Teoria de Resposta ao Item; Instrumentos de Aferição
• Palavras-chave em Inglês: Stress; Pregnant; Factor Analysis; Item Response Theory; Measuring instruments
• Área do conhecimento CNPq: Saúde Materno-Infantil; Saúde Coletiva
• Link de acesso: http://tedebc.ufma.br:8080/jspui/handle/tede/1579

Early identification of stress by means of measuring instruments is a simple strategy and low cost to the public sector, but it is necessary to confirm that we are actually measuring what it purports to measure for a specific population, ie, stress in women during pregnancy. The objective of this study is the analysis of two stress measurement instruments in pregnant women in two cities: stressful life events (EVE) and the perceived stress scale (PSS). A crosssectional study was conducted with pregnant women in São Luis/MA and in Ribeirão Preto/SP (BRISA cohort). The construct validity of EVE was done by Exploratory Factor Analysis and Confirmatory. The parameters of PSS items were estimated using the Gradual Response Model Samejima based on Item Response Theory (IRT). For analysis of the items Differential Functioning (DIF) were used the likelihood ratio for non-uniform DIF and differences between the weights of the regressions (Δβ1) for uniform DIF. As an estimate of the magnitude of DIF was used Pseudo-R 2 . The model with three factors then called Life Events Causing Stress (LECS) showed a good fit (RMSEA <0.05; IFC / TLI> 0.90) for the two samples. For the PSS only the 10-item scale (PSS 10) presented the prerequisite of unidimensionality. The results showed good variation of the response thresholds (bi) and discrimination (ai) from moderate to high. The analysis found items 9 and 10 with probable DIF in the PSS 10, but there was no significant effect of residence location in the responses to items, suggesting that the answers are not related to any other factor beyond trace latent perceived stress. This study found evidence that confirms the construct validity of a list of stressor events being titled Life Events Inventary Causing Stress (LEICS). By TRI and DIF, it is concluded that the PSS 10 scale was invariant when applied to samples of pregnant women in two cities.