Figure 6

Changes relative to trial types of interest in ACC and VTA 4 Hz power and coherence. Trial position 0 in these plots indicates the normalised residual vertex mean of all trials of each type for each dependent measure, respectively. The data represented at trial position 1 are the vertex means of all trials immediately after each trial that was used in trial position 0. Other trial position data were calculated in a similar fashion. Consistent with the linearized results, vertex ACC and VTA 4 Hz power and coherence were greatest during incorrect choice trials as compared to during instruction and correct choice trials. ACC and VTA 4 Hz power and coherence were elevated in the vertex region as the animal made a post-error correct choice. These data from phase A1 are presented as mean ± SEM of all normalised residual data where the normalisation was relative to the vertex. ^**Indicates that correct choices were significantly different from incorrect choices at the p < 0.005 significance level via multiple comparisons tests; && indicates that incorrect and instruction trials were significantly different at the p < 0.005 significance level via multiple comparisons tests; ^##Indicates that correct choices and instruction trials were significantly different at the p < 0.005 significance level via multiple comparisons tests. See also Fig. S4 for data processed without the regression procedure.