A associação entre apoio social e desfechos relacionados à saúde tem sido objeto de interesse de pesquisadores. O objetivo desse artigo foi avaliar a confiabilidade e estrutura de concordância da escala de apoio social. A confiabilidade teste-reteste foi investigada num grupo de gestantes de maternidade pública (n = 65) por intermédio do coeficiente de correlação intraclasse (CCIC) e kappa ponderado (kw2). Avaliou-se a estrutura de concordância pelos modelos log-lineares. O CCIC da escala foi 0,90 e kw² variou entre 0,23 e 0,70. Modelos com melhores ajustes foram o de concordância diagonal mais associação linear por linear e de semi-independência. Considerou-se a escala um instrumento capaz de medir, de maneira confiável, o apoio social entre gestantes de baixa renda.

There is a growing interest in research on the association between social support and health outcomes. The objective of this study was to evaluate the reliability and structure of agreement of a social support scale. Test-retest reliability was measured in a group of pregnant women (n = 65) in a public maternity ward. Intraclass correlation coefficient (ICC) and quadractically weighted kappa (kw²) were used as agreement measures. Log-linear statistical models were fitted to describe patterns of agreement. ICC for social support score was 0.90. The kw² ranged from 0.23 to 0.70. Log-linear models that provided the best fit to the data were diagonal agreement plus linear-by-linear association and quasi-independence models. The scale was considered a reliable instrument to measure social support scale in low-income pregnant women.