Ajuste de análise fatorial exploratória com método dos mínimos resíduos a dados de testes neurocognitivos

Abstract

In this study, an exploratory factor analysis (EFA) was applied using the minimum residuals method (Minres) to data from the NeuroCognitive Performance Test (NCPT), a battery of neurocognitive tests. The central objective of the study was to investigate the applicability of the Minres method as a robust alternative to traditional extraction methods, such as principal components and maximum likelihood. The research used data from the NCPT battery 26, comprising 11 subtests that assess different cognitive processes. The analytical procedure began by verifying the suitability of the data for EFA using the Kaiser-Meyer-Olkin (KMO) index, which indicated good suitability (0.82). The determination of the number of factors to retain was conducted through parallel analysis, which indicated the existence of three factors. The extraction of factor loadings was then performed using the Minres algorithm, followed by an orthogonal varimax rotation to facilitate the interpretation of the factors. The results revealed a three-factor structure. Factor 1 showed a contrast between working memory and fluid intelligence subtests, Factor 2 was predominantly loaded by verbal memory subtests, and Factor 3 was dominated by the arithmetic reasoning subtest. The model’s fit quality was confirmed by the Root Mean Square Error of Approximation, or RMSEA (0.037), and the Tucker-Lewis Index, or TLI (0.969), indices. Despite the successful application of the method, the study has some limitations, such as the low communalities found in some subtests and the possible influence of the type of rotation used. It is concluded that the minimum residuals method proved to be fully applicable and viable for exploratory factor analysis of neurocognitive data.