[Factor structure validity of the social capital scale used at baseline in the ELSA-Brasil study]

Cad Saude Publica. 2016 Jul 21;32(7):S0102-311X2016000704002. doi: 10.1590/0102-311X00101515.
[Article in Portuguese]

Abstract

This study aims to analyze the factor structure of the Brazilian version of the Resource Generator (RG) scale, using baseline data from the Brazilian Longitudinal Health Study in Adults (ELSA-Brasil). Cross-validation was performed in three random subsamples. Exploratory factor analysis using exploratory structural equation models was conducted in the first two subsamples to diagnose the factor structure, and confirmatory factor analysis was used in the third to corroborate the model defined by the exploratory analyses. Based on the 31 initial items, the model with the best fit included 25 items distributed across three dimensions. They all presented satisfactory convergent validity (values greater than 0.50 for the extracted variance) and precision (values greater than 0.70 for compound reliability). All factor correlations were below 0.85, indicating full discriminative factor validity. The RG scale presents acceptable psychometric properties and can be used in populations with similar characteristics.

Publication types

  • Validation Study

MeSH terms

  • Adult
  • Brazil
  • Factor Analysis, Statistical
  • Humans
  • Longitudinal Studies
  • Psychometrics
  • Social Capital*
  • Social Support
  • Surveys and Questionnaires*