A validation study of the psychometric properties of the Groningen Reflection Ability Scale

Abstract

Background: Reflection, the ability to examine critically one's own learning and functioning, is considered important for 'the good doctor'. The Groningen Reflection Ability Scale (GRAS) is an instrument measuring student reflection, which has not yet been validated beyond the original Dutch study. The aim of this study was to adapt GRAS for use in a Danish setting and to investigate the psychometric properties of GRAS-DK.

Methods: We performed a cross-cultural adaptation of GRAS from Dutch to Danish. Next, we collected primary data online, performed a retest, analysed data descriptively, estimated measurement error, performed an exploratory and a confirmatory factor analysis to test the proposed three-factor structure.

Results: 361 (69%) of 523 invited students completed GRAS-DK. Their mean score was 88 (SD = 11.42; scale maximum 115). Scores were approximately normally distributed. Measurement error and test-retest score differences were acceptable, apart from a few extreme outliers. However, the confirmatory factor analysis did not replicate the original three-factor model and neither could a one-dimensional structure be confirmed.

Conclusions: GRAS is already in use, however we advise that use of GRAS-DK for effect measurements and group comparison awaits further review and validation studies. Our negative finding might be explained by a weak conceptualisation of personal reflection.

Figure 1

Process flow chart. The flow chart shows the translation and cultural adaptation process leading to GRAS-DK following the stages suggested by Beaton and colleagues [26].

Figure 2

Overall validation process. The flow chart shows the overall validation process from sampling to the statistical analyses.

Figure 3

Bland-Altman plot. The Bland-Altman plot shows the average test score plotted against the difference between the test and retest average scores.