Skip to main content
. 2020 Dec 15;3(1):fcaa202. doi: 10.1093/braincomms/fcaa202

Figure 3.

Figure 3

Classifier performance. For each unique subject, receiver operating characteristic curves show the performance of a logistic-regression classifier tested on held-out sessions of the delayed free-recall task (grey lines), with the group average performance shown in bold (TBI group, left: mean AUC = 0.624, n =37, SEM = 0.011; non-TBI group, right: mean AUC = 0.620, n =111, SEM = 0.008). The distributions of the two groups were statistically equivalent [t(142) = 0.29, P =0.77]