highly significantP = 0.001
Comparing follow-up with pre-training performances using a Benjamini–Hochberg-corrected two-sided paired t -test, the improvements were still highly significant for the Token test [ t (8) = 4.208, P = 0.001], written language [ t (8) = 3.957, P = 0.001], naming [ t (8) = 4.318, P = 0.006] and significant for comprehension [ t (8) = 2.731, P = 0.03].