Pseudoneglect: a review and meta-analysis of performance factors in line bisection tasks.

An exhaustive qualitative (vote-counting) review is conducted of the literature concerning visual and non-visual line bisection in neurologically normal subject populations. Although most of these studies report a leftward bisection error (i.e., pseudoneglect), considerable between-study variability and inconsistency characterize this literature. A meta-analysis of this same literature is performed in which the total quantitative data set, comprising 73 studies (or sub-studies) and 2191 subjects, is analyzed with respect to 26 performance factors. The meta-analytic results indicate a significant leftward bisection error in neurologically normal subjects, with an overall effect size of between -0.37 and -0.44 (depending on integration method), which is significantly modulated to varying degrees by a number of additional task or subject variables. For example, visual bisection tasks, midsagittal-pointing tasks and tactile bisection tasks all lead to leftward errors, while kinesthetic tasks result in rightward errors. Tachistoscopic forced-choice testing methods reveal much greater estimates of bisection error (effect size = -1.32) than do manual method-of-adjustment procedures (effect size= -0.40). Subject age significantly modulates line bisection performance such that older subjects err significantly rightward compared to younger subjects, and to veridical line midpoint. Male subjects make slightly larger leftward errors than do female subjects. Handedness has a small effect on bisection errors, with dextrals erring slightly further to the left than sinistral subjects. The hand used to perform manual bisection tasks modulated performance, where use of the left hand lead to greater leftward errors than those obtained using the right hand. One of the most significant factors modulating bisection error is the direction in which subjects initiate motor scanning (with either eye or hand), where a left-to-right scan pattern leads to large leftward errors while a right-to-left scan pattern leads to rightward errors.