Fig. 3.

Error vs. correct theta isolated using TF-PCA. PC-weighted TF representations of response-locked average power for (A) the error-correct difference, (B) Error-nogo trials, and (C) Error-go trials. Note that the timing of the error-related theta component identified is > 100 ms later than similar effects observed in healthy adults; without employing a TF-PCA approach, this effect of interest might have been missed.

Figure reprinted from Kim et al. (2020).