marginally significantp < 0.01
A repeated-measures ANOVA on the training-session RTs with context (old, new) and epoch (1–4) as factors revealed a significant main effect of epoch, though only a marginally significant effect of context: epoch, F (1.71, 25.63) = 18.74, p < 0.01, η p 2 = 0.56 (RTs were 287 ms faster in epoch 4 than in epoch 1); context, F (1,15) = 4.22, p = 0.058, η p 2 = 0.20 (RTs were 106 ms faster to old than to new displays); the context × epoch interaction was not significant, F (1.43, 21.48) = 0.78, p = 0.51, η p 2 = 0.05.