The endpoints' task: an analysis of length reproduction in unilateral neglect

When patients with left hemineglect are shown the centre of a memorised line and are asked to mark both the endpoints, they place the left one significantly farther. This contralesional bias appears to be incoherent with respect to most current interpretations of neglect [7]. In the present study, ten patients with left hemineglect and two control groups were presented the centre of a line and one of its endpoints and had to mark the missing one at the correct distance. With this modification, subjects had not to rely on stored information about the line, but simply to reproduce a distance either to the left or to the right separately. The task was performed in three different spatial locations with respect to subjects' body midline. Similarly to the original endpoints' task, neglect patients showed a leftward bias, placing the left endpoint significantly farther than the right one. However, this was not associated with a paradoxical contralesional over-extension with respect to the distance to reproduce. Indeed, when marking the left endpoint, patients were accurate (when performing in the right hemispace) or even underestimated the distance to reproduce (when performing centrally and in the left hemispace). Instead, the leftward bias was due to severe underestimation errors when marking the right endpoint, which were independent of the stimulus location. These results demonstrate that the leftward bias in the endpoints' task is relative rather than absolute and does not reflect a paradoxical overestimation in length reproduction towards the left side.