Is self-report a valid measure of unimanual object-based task performance?

Hand preference is commonly measured via self-report using instruments known as handedness inventories. These instruments query which hand subjects use to perform unimanual object-based tasks and also the consistency of usage. Scores reveal which hand people report using most (i.e., which hand is preferred). Some people report highly consistent usage of their preferred hand while others report relatively inconsistent usage. These reports are often assumed to be accurate descriptions of people's actual behavior, but empirical evidence concerning their validity is surprisingly scant. In the present research, subjects completed a handedness inventory on which higher scores indicated more consistent usage of the preferred hand. After a brief filled interval, subjects performed tasks listed on the inventory, spontaneously choosing which hand to use for each task. Subjects who scored higher on the inventory used the preferred hand more often. Subjects were then timed while performing the tasks as fast as they could with each hand. Subjects were faster when using their preferred hand and the magnitude of the preferred hand advantage was positively correlated with self-reported consistency. These results support the validity of inventory scores by showing that they are related to spontaneous hand choice and manual fluency when performing inventory tasks.