Item response estimates of the parameters of a brief measure of student psychological well-being in the university context

• Autor: Franco Mendoza, Josué Michael
• Formato: artículo
• Fecha de Publicación: 2021
• Institución: Universidad Nacional Mayor de San Marcos
• Repositorio: Revistas - Universidad Nacional Mayor de San Marcos
• Lenguaje: español
• OAI Identifier: oai:ojs.csi.unmsm:article/20513
• Enlace del recurso: https://revistasinvestigacion.unmsm.edu.pe/index.php/psico/article/view/20513
• Nivel de acceso: acceso abierto
• Materia: Psychological well-being; test construction; psychometrics; college students; Bienestar psicológico; construcción de pruebas; psicometría; estudiantes universitarios

The purpose of the present study was to construct a brief scale of psychological well-being in the university context, in a preliminary phase. For this purpose, 417 students from the faculty of education of a public university in Lima, of both sexes, aged between 17 and 35 years (M = 23, S.D. = 1.86), participated in the study. We worked with the topics of Mental Health (MH) and Family and Interpersonal Relations (FR). The method was instrumental, with a scale of 35 items and seven components, approved by five experts through Aiken's V statistic, with three evaluation criteria (V = .83, 1.00). The items went through different analysis processes in parameter estimations with the two-logistic parameter model (2LP) of the Item Response Theory (IRT), in contrast with indicators of the Classical Test Theory (CTT), from the Confirmatory Factor Analysis (CFA) and internal consistency measures. The results showed that the scale components adequately adjusted to the 2LP model (x2 = .58, 1.73, p > .05), fulfilling the assumptions of unidimensionality and local independence for the items for each component. Information functions of the items were also reported, which were complemented with evidence of the CFA fit for each topic (x2/gl = 1.18, 2.45, GFI = .99, 1.00, TLI = .94, .99, CFI = .95, .99, NNFI = .94, .99, RMR = .001, SRMR = .05, .06, RMSEA = .02, .06). On the other hand, the internal consistency measures ranged between .61 and .82, taking the Omega coefficient and KR-20 as indicators. It is recommended to continue revising the model in a next phase of the study, while complementing the analysis with components oriented to educational needs and the development of soft skills.