Preliminary Rasch analysis of the Multidimensional Assessment of Interoceptive Awareness in adults with stroke

Abstract

AbstractPurposeThe Multidimensional Assessment of Interoceptive Awareness (MAIA) measures interoceptive body awareness, which includes aspects such as attention regulation, self-regulation, and body listening. Our purpose was to validate the MAIA in adults with stroke using Rasch Measurement Theory.MethodsThe original MAIA has 32 items grouped into 8 separately scored subscales that measure aspects of body awareness. Using Rasch Measurement Theory we evaluated the unidimensionality of the entire scale and investigated person and item fit, person separation reliability, targeting, local item dependence, and principal components analysis of residuals.ResultsForty-one adults with chronic stroke (average 3.8 years post-stroke, 13 women, average age 57±13 years) participated in the study. Overall fit (χ2=62.26, p=0.26) and item fit were obtained after deleting 3 items and rescoring 26 items. One participant did not fit the model (2.44%). There were no floor (0.00%) or ceiling effects (0.00%). Local item dependence was found in 42 pairs. The person separation reliability was 0.91, and the person mean location was 0.06±1.12 logits.ConclusionsThe MAIA demonstrated good targeting and reliability, as well as good item and person fit in adults with chronic stroke. A study with a larger sample size is needed to validate our findings.