Left- and right-handed school children with differing reading/writing exper
iences (unidirectional left-to-right vs bidirectional) were asked to draw 3
-cm lines from light to-left or from left-to-right with each hand. With eit
her hand, lines drawn from left to right were more accurate than those draw
n from right to left, particularly for right-handed left-to-right users: bi
directional readers showed no directional bias. Moreover, bidirectional rea
ders were more accurate than unidirectional readers. The findings support a
greater influence of directional scanning effects than handedness on the t
ask of line length estimation. (C) 2000 Academic Press.