Purpose: To investigate the internal consistency, convergent and known-groups construct validity of the Figure-of-Eight Walk Test with two novel conditions in persons with lower limb amputation, and to examine differences in walking performance between the three conditions within a group of persons with transtibial amputation and transfemoral amputation/knee disarticulation. Materials and methods: 50 adults with unilateral amputation participated, 28 of whom had undergone a transtibial amputation and 22 a transfemoral amputation/knee disarticulation. Three Figure-of-Eight Walk Test Conditions were investigated: 1) walking at a self-selected walking speed, 2) walking while carrying a tray with two cups of water and 3) walking on uneven terrain. Internal consistency was evaluated using Cronbach's alpha. Convergent construct validity was examined by analysing the relationship between the Figure-of-Eight Walk Test parameters and performance-based parameters (Amputee Mobility Predictor, Ten-Meter Walk Test, Six-Minute Walk Test) and self-report measures (Prosthetic Limb Users Survey of Mobility, Activities-specific Balance Confidence Scale) using Spearman's rankorder correlations. Known-groups construct validity was assessed by comparing the Figureof-Eight Walk Test parameters based on anatomical level of amputation. Friedman's test and posthoc analysis was used to examine differences between the walking conditions within each group. Results: Cronbach's alpha coefficients of the Figure-of-Eight Walk Test parameters for all three conditions ranged from .89 to .99. The Figure-of-Eight Walk Test time and steps parameters demonstrated moderate to good correlation (r = -.50 to -.77) for performance-

No abstract