in two experiments, normal adults divided a horizontal line segment and an
equal spatial interval that did not contain a line into eight equal-appeari
ng segments bq; means of successive bisections. In the first experiment, su
bjects' average initial bisections erred to the left of objective center fo
r both stimuli. Their subsequent bisections produced similar errors for the
line-present stimulus. as the bisection of each progressively smaller line
segment was placed to the left of true center. However. this pattern did n
ot occur when bisecting the empty interval. The finding that the presence o
f a line influences bisection errors implicates an 'object-based' mechanism
in the genesis of line bisection errors and suggests that this mechanism v
aries in its operation with visual field location. In the second experiment
. subjects successively bisected longer line and interval stimuli which wer
e presented either centered on the subjects' midlines or displaced to the r
ight or left. Bisections tended to be placed farther to the left for the le
ft stimuli and farther to the right for the right stimuli. with little or n
o bias for the centrally located stimuli. Repeated measures with the centra
lly located stimulus demonstrated strong individual differences in bisectio
n biases. Errors were also found to be correlated for the line-present and
line-absent stimuli in both experiments. suggesting the additional contribu
tion of a mechanism that is not object-based. (C) 2001 Elsevier Science Ltd
. All rights reserved.