In reply
Wright and colleagues suggest in their letter that the statistical methods used in our meta-analysis1 may have led to the underestimation of potential laterality differences in our patient sample. We agree that the use of the variance term 1/(N−3) is not appropriate for the computation of the significance of Cohen q for measurements (which are highly correlated), and may thus lead to an increased risk of type II errors. Therefore, we have recomputed all values of Cohen q using an updated2 version of the formula originally given by Dunn and Clark.3