only marginally significantp = 0.082
Following up on these two-way interactions, we found that blind participants responded more accurately than sighted participants following stimulus pairs that were anatomically incongruent under anatomical instructions and externally incongruent under external instructions (all p < 0.05, bold font in Table 4 ; as one exception, this difference was only marginally significant after Holm correction for external instructions with differentially oriented hands, corrected p = 0.082).