Measuring trait mindfulness: how to improve the precision of the Mindful Attention Awareness Scale using a Rasch model

The Mindful Attention Awareness Scale (MAAS) is the most widely used mindfulness scale to date, with validation studies indicating acceptable reliability and convergent validity. However, recent evidence indicates that the ability of the MAAS items to precisely discriminate between mindfulness levels is compromised. To improve item functioning and precision of the MAAS, responses of 250 participants to the scale were subjected to Rasch analysis. To improve disordered thresholds, items were re-scored, and each item was tested for Differential Item Functioning. Where misfit to Rasch model expectations was identified, items were removed and the effect on individual item-fit estimates was tested. Uniform rescoring of all items was the best solution to order thresholds of all items and to improve overall goodness of fit to the Rasch model. Satisfactory model fit was achieved after removing the misfitting items 6 and 15 and combining the locally dependent items 7 and 8. Functioning of MAAS items can be improved substantially by several minor modifications to scoring algorithms without the need to modify the current response format. Precision of the instrument can be improved further by using the ordinal-to-interval conversion tables presented here.