Sum scores and categories for the German AAS.

BackgroundUnintentionally pregnant individuals in Germany seeking an abortion face challenges due to legal regulations, stigma and difficult access to abortion care. Abortion attitudes of (prospective) physicians influence the care situation. To measure these attitudes, psychometrically sound instruments like the Abortion Attitude Scale (AAS) are necessary. So far, no instruments assessing attitudes toward abortions are available in German. This study aims to translate, culturally adapt and psychometrically test the AAS.MethodsThis is a secondary analysis of a cross-sectional study on abortion attitudes of medical students in Germany. The English 14-item AAS was translated into German and adapted using a team translation protocol. Comprehensibility was tested via cognitive interviews (n = 10 medical students). We analyzed acceptance (completion rate), factor structure (confirmatory factor analysis (CFA), model fit), item characteristics (response distribution, item difficulties, corrected item-total correlations, inter-item correlations), and reliability (McDonald’s omega).ResultsThe translated and adapted AAS version was comprehensible. AAS data of 305 medical students could be included in analysis. Completion rate was above 98% for all items. The CFA results confirmed a one-factor structure, but a model without item 10 and correlations between item 8 and item 13 showed the best model fits. Floor or ceiling effects could be found for 7 items, item difficulties ranged between 0.39 and 0.94, corrected item-total correlations ranged between 0.460 and 0.766 for the best model, inter-item correlations ranged between.129 to.681, and McDonald’s omega was above 0.9 for both models.ConclusionThe German AAS is a brief measure that was accepted and demonstrated adequate psychometric properties. Suboptimal performance of some items indicated the need for further validation, particularly of the 13-item version. The AAS can improve evaluation of abortion attitudes in Germany. This may support identification of barriers in healthcare provision and improve care for unintentionally pregnant individuals.