highly significantp ≤ 0.0001
By applying a paired t test, the differences between the WN and RNN model to the TAtt model were shown to be both highly significant with p ≤ 0.0001 across subjects (Cohen’s d = 10.68 and d = 10.66, respectively).
By applying a paired t test, the differences between the WN and RNN model to the TAtt model were shown to be both highly significant with p ≤ 0.0001 across subjects (Cohen’s d = 10.68 and d = 10.66, respectively).