The interaction between group and difficulty only showed a trending result and was driven by the one-handers performing worse on the difficult configurations relative to controls ( t (38) = 2.38; p = 0.022; Bonferroni’s adjusted α , 0.05/3 = 0.0167).
As shown in Figure 3 C , for amputees we found a significant (above zero) correlation between both contralateral and ipsilateral ROIs relative to the typical contralateral representational structure in controls (amputees contralateral, t (15) = 6.82; p < 0.001; amputees ipsilateral, t (15) = 4.38; p < 0.001; one-sample t test; Bonferroni’s adjusted α , 0.05/2 = 0.025), whereas the correlation between one-handers and controls was approaching significance for the contralateral ROI only (one-handers contralateral, t (12) = 2.51; p = 0.027; one-handers ipsilateral, t (12) = 1.70; p = 0.114; Bonferroni’s adjusted α , 0.05/2 = 0.025).
When comparing across groups (using a one-way ANCOVA, accounting for age), we found a trend toward significance ( F (2, 38) = 2.99; p = 0.062; η p = 0.14; BF 10 = 2.05), which is also reflected in greater homotopy in amputees relative to controls ( t (38) = 2.41; p = 0.021) but not for one-handers relative to controls ( t (38) = 1.63; p = 0.112; BF 10 = 1.44; Bonferroni’s adjusted α , 0.05/2 = 0.025 for the last two comparisons).