[The factor structure of direct and indirect methods for measuring change in medical rehabilitation--analyses on item level].

AIM OF THE STUDY: The aim of the study was to examine the direct and the indirect change measurement; previous studies comparing the 2 methods demonstrated that they agree only to a limited extent. Analyses of the factorial structure have shown that indicators of direct change measurement were assigned regardless of the content of a dimension common to them. This paper examined whether previous results obtained on scale level were also valid on item level.
METHODS: Empirical data were obtained from 3 rehabilitation patient samples (n=466, n=194, n=610). The factorial structures of the direct and indirect measurements of change were assessed and compared using principal components analysis with oblique rotation. Additionally, analyses were performed stratified across clinical indications and only for items of the direct change measurement.
RESULTS: The factorial structure of 3 data sets showed that the previously published scale-level results were also valid on item level. Analyses demonstrated only low to medium agreement between the 2 methods of change measurement. The indicators of direct change measurement loaded regardless of their content on one or a few common dimensions, which did not agree with those of the indirect change measurement.
CONCLUSION: In considering the 2 methods of measuring change, it is apparent that the standard method of change measurement (indirect change measurement) is suitable to reflect a comprehensive and multi-dimensional representation of changes within the field of medical rehabilitation evaluation. © Georg Thieme Verlag KG Stuttgart · New York.