RESUMO
Objective: The Social Problem-Solving Inventory-Revised (SPSI-R) is a widely used instrument to assess problem-solving ability. This study examined the factor structure of the 52-, 25-, and 10-item versions of the SPSI-R and assessed factorial invariance across English- and Spanish-speaking participants. In addition, the internal consistency, test-retest reliability and sensitivity to detect change in problem-solving skills over time were assessed across the three different versions of the SPSI-R. Methods: Data from three randomized controlled trials, in which caregivers of children with cancer (N = 1,069) were assigned to either a problem-solving skills intervention (N = 728) or a control condition (N = 341), were combined. The SPSI-R was administered at baseline (T1) and immediately post intervention (T2). Reliability and multigroup analyses were performed with confirmatory factor analysis (CFA). Sensitivity to change analyses were performed using repeated measures ANOVA. Results: Confirmatory factor analysis at T1 showed good fit statistics and internal consistency for the 52- and the 25-item versions, but not for the 10-item version. Factorial invariance was demonstrated across time (T1-T2) and language (Spanish-English) for both the 52- and 25-item versions. Adequate sensitivity to change over time was shown. Conclusion: The 52- and 25-item versions of the SPSI-R appear reliable and valid for assessment of problem-solving skills in English- and Spanish-speaking caregivers of children with newly diagnosed cancer. The 25-item SPSI-R can be used as a short version measuring problem-solving ability; the 10-item version cannot be considered a reliable measure for this population.
