ABSTRACT
A validation study the Iranian Modified Barthel Index (MBI) in hospitalized acute stroke elderly by classical test theory approach and investigate Rasch analysis for both Iranian version MBI and BI and compare the hierarchical item difficulty of them. Face-to-face interview with 100 geriatric stroke inpatients 60+ or their caregivers was done in a cross-sectional study. First, construct validity of MBI analyzed by the classical test theory, then Rasch analysis were done for BI and MBI. The reliability of the Iranian MBI was significant at 0.955. One factor achieved by the variance of 83.2%. In Rasch analysis for MBI, the most difficult item was stair climbing, whereas the simplest items were bowel and bladder control. In BI, the most difficult items were toilet use and ambulation. The Iranian MBI is very accurate and reliable; therefore the use of MBI to measure better outcomes in stroke elderly inpatients is recommended comparing with BI.