CRN amplitudes for all correct trial types correlated with the number of days abstinent in alcoholics (r = 0.562, p = 0.045, N = 13], indicating that the CRN amplitude was less negative (closer to control values) with more days abstinent. Accuracy correlated with ERN amplitudes in alcoholics (r = −0.668, p = 0.012, N = 13). By contrast, in controls accuracy correlated with the amplitude of the CRN (r = 0.670, p = 0.009, N = 14). CRN amplitudes in alcoholics were in turn correlated with RTs for error (r = −0.611, p = 0.026, N = 13), correct (r = −0.644, p = 0.018, N = 13), post-error (r = −0.593, p = 0.033, N = 13), and post-correct trials (r = −0.657, p = 0.015, N = 13). The ERN did not correlate with any RT measures in alcoholics, and the RTs did not correlate with the CRN or ERN amplitudes in the control group.
