As a preliminary step in developing the physical-functioning measure of the Late-Life Function and Disability Instrument (LLFDI), the authors compared its items with the physical-functioning items (PF-10) on the SF-36 Health Survey. They compared the item coverage, hierarchy, and scale-separation properties of the PF-10 items with those of the physical-functioning items of the LLFDI. Both questionnaires were administered to 50 community-dwelling older adults. A partial-credit, 1-parameter, item-response-theory model was used to scale the items. The LLFDI improved the range of ability of daily activities that was encompassed by the PF-10 items by 46%. By sequentially deleting new items with poor fit to the overall scale and items with redundant content, the authors developed a scale more capable of accurately assessing low-functioning activities. The LLFDI function component incorporates a broader content range and better person and item separation than the PF-10 items. It appears to have potential as a comprehensive functional-activity assessment for community-dwelling older adults.
Haley and Kooyoomjian are with the Roybal Center for the Enhancement of Late Life Function, Sargent College of Health and Rehabilitation Sciences, Boston University, Boston, MA 02215-1605. Ludlow is with the Lynch School of Education, Boston College, Chestnut Hill, MA 02467.