Transfer accuracy
A first measure of transfer is to ask whether accuracy was the same in Blocks 14 and 15. No difference implies perfect transfer. The accuracy differences are shown in
Table 2 (Block 15-14). Paired-sample t tests were trained on categorization declined when they switched to the same–different task [all ts(14)-2.94, p.05]. For the participants trained with the same–different task, the accuracy did not change when the task switched to direct classification (see Table 2).