Measurement equivalence of the Birth Satisfaction Scale-Revised (BSS-R): further evidence of construct validity

Objective and background
The 10-item Birth Satisfaction Scale-Revised (BSS-R) is being increasingly used internationally, including the development of translated versions of the tool. However, to date, a direct comparison between the original version of the tool and a non-English language translated version has yet to be conducted. Recognising that measurement equivalence is critical in order to be able to meaningfully compare scores on the measure between different versions, the current study sought to evaluate the measurement invariance characteristics of the BSS-R within this context. Methods: A secondary analysis of two data sets. The study used a measurement invariance testing approach to determine the relative equivalence between the original UK English-language version and the Greek-translated version of the BSS-R. Participants were a convenience sample of UK (n = 228) and Greek (n = 162) postnatal women. Results: The BSS-R was found to offer an excellent model fit with pooled data, a robust configural model and metric-level invariance between English- and Greek-language versions. The BSS-R was also found to demonstrate partial scalar invariance, with 80% of item intercepts non-invariant between both versions. Two non-invariant items at the scalar level are likely to represent real differences between participant groups in terms of birth satisfaction as an artefact of service delivery type and relative difference in delivery mode. Conclusion: The BSS-R is both conceptually and statistically comparable between different versions of the tool suggesting the utility of the measure for international comparative studies.