DOI: 10.1080/16501970600632594

Objective: To investigate aspects of validity and stability of Performance Analysis of Driving Ability (P-Drive), for people with stroke when used in a driving simulator. Design: A cross-sectional observational study. Subjects: The study included a consecutive series of 101 participants with stroke referred for evaluation or selected from a stroke registry. Methods: P-Drive was used to observe driving performance in order to assess driving ability. P-Drive comprises 20 items assessing the quality of the participant's driving performance. Aspects of validity and reliability in P-Drive were evaluated using Rasch statistics. Results: The items (95%) and participants (97%) demonstrated acceptable goodness-of-fit and met statistical expectations according to the Rasch model. The results support internal scale validity and person response validity. P-Drive could separate the participants with different driving abilities and the standard errors were within reasonable criteria for drivers with a moderate-to-low ability to drive. Conclusion: The findings from this study indicated that P-Drive is an assessment tool with properties of internal scale validity, person response validity, and which also contains aspects of reliability in relation to precision of the estimates and separation. P-Drive seems to be a valid and stable assessment tool for assessing the driving ability in a simulator of people with stroke.