Immediate error correction process following sleep deprivation


  • 1

    The mask served to reduce the visibility of the target letter; so to increase the difficulty of identifying the target letter and to increase the flanker interference. The mask had the same visual angle as a single letter and was made of the four letters printed on top of each other.

  • 2

    There are two reasons for us to choose the time window of 150 to 50 ms prior to the response as the baseline for adjusting the EEG epochs. First, this is the same time window as in our previous study (Tsai et al., 2005), which enables us to make a fair comparison across the two studies. Second, we hoped to avoid a time window too closely to the range of P300, which has been shown to be reduced by sleepiness (e.g. Lee et al., 2004).

  • 3

    The time window of 0–150 ms for ERN and of 300–500 ms for Pe was determined based on the conventional definition in the literature (c.f. Falkenstein et al., 2000; Hajcak et al., 2003; Nieuwenhuis et al., 2005; Ridderinkhof et al., 2002; Tsai et al., 2005).

  • 4

    There was no Ne/ERN evident on error trials without correction as correct trials, so we did not include these trials into analyses.

  • 5

    Scheffers et al. (1999) used memory and visual search tasks which are not typical Ne/ERN paradigm.

Shulan Hsieh, Department of Psychology, National Chung-Cheng University, 168 University Road, Min-Hsiung, Chia-Yi 621, Taiwan. Tel.: +886-5-2720411x32200; fax: +886-6-2894836; e-mail:


Previous studies have suggested that one night of sleep deprivation decreases frontal lobe metabolic activity, particularly in the anterior cingulated cortex (ACC), resulting in decreased performance in various executive function tasks. This study thus attempted to address whether sleep deprivation impaired the executive function of error detection and error correction. Sixteen young healthy college students (seven women, nine men, with ages ranging from 18 to 23 years) participated in this study. Participants performed a modified letter flanker task and were instructed to make immediate error corrections on detecting performance errors. Event-related potentials (ERPs) during the flanker task were obtained using a within-subject, repeated-measure design. The error negativity or error-related negativity (Ne/ERN) and the error positivity (Pe) seen immediately after errors were analyzed. The results show that the amplitude of the Ne/ERN was reduced significantly following sleep deprivation. Reduction also occurred for error trials with subsequent correction, indicating that sleep deprivation influenced error correction ability. This study further demonstrated that the impairment in immediate error correction following sleep deprivation was confined to specific stimulus types, with both Ne/ERN and behavioral correction rates being reduced only for trials in which flanker stimuli were incongruent with the target stimulus, while the response to the target was compatible with that of the flanker stimuli following sleep deprivation. The results thus warrant future systematic investigation of the interaction between stimulus type and error correction following sleep deprivation.