Available via license: CC BY 4.0
Content may be subject to copyright.
Error monitoring in dual-tasking 1
Adaptive rescheduling of error monitoring in multitasking
Abbreviated title: Error monitoring in multitasking
Robert Steinhauser*1, Marco Steinhauser1
1 Department of Psychology, Catholic University of Eichstätt-Ingolstadt, Ostenstraße 25,
85072 Eichstätt, Germany
Corresponding author: Robert Steinhauser, Catholic University of Eichstätt-Ingolstadt,
Ostenstraße 25, 85072 Eichstätt, Germany. Phone: +49 8421 9321120. E-mail:
robert.steinhauser@ku.de.
33 pages, 8 figures, 0 tables
Word counts: Abstract: 150; Main text: 6393; Methods: 2196
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 2
Abstract 1
The concurrent execution of temporally overlapping tasks leads to considerable 2
interference between the subtasks, which is overcome by a neural mechanism that actively 3
defers processing of the secondary task. In this study, we recorded error-related scalp 4
potentials to investigate whether such a mechanism can also be found in error monitoring. 5
Participants worked on a multitasking paradigm with partially overlapping execution of two 6
tasks (T1, T2). The error positivity, a correlate of conscious error evaluation, was reduced 7
after T1 errors but occurred after a correct T2-response instead. MVPA-based and regression-8
based single-trial analysis revealed a trial-wise trade-off between immediate and postponed 9
error processing. Two follow-up experiments confirmed this finding, showed that this result is 10
not due to credit-assignment errors in which a T1 error is falsely attributed to T2, and verified 11
the link to conscious error detection. These findings demonstrate a remarkable flexibility of 12
the human brain when adapting to multitasking situations. 13
14
(150 words) 15
16
Keywords: multitasking, dual-tasking, error awareness, Pe, ERN 17
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 3
Introduction 18
Adaptive human behavior requires an error monitoring system that constantly checks 19
whether an error has occurred and initiates adjustments to cognition and behavior 20
accordingly. Numerous studies could show that such a system exists in the human brain, 21
detecting errors fast and reliably (Ullsperger, Danielmeier, & Jocham, 2014). Most studies, 22
however, investigated error monitoring when single tasks were executed in isolation. In 23
contrast, everyday behavior is characterized by the concurrent execution of temporally 24
overlapping tasks. As dual-tasking is associated with decrements in task performance 25
(Pashler, 1994; Tombu & Jolicœur, 2003), it is plausible to assume that also error monitoring 26
suffers under these conditions. Indeed, recent evidence suggested that dual-tasking leads to 27
specific impairments to the neural correlates of error monitoring (Klawohn, Endrass, Preuss, 28
Riesel, & Kathmann, 2016; Weißbecker-Klaus, Ullsperger, Freude, & Schapkin, 2016). In the 29
present study, we investigate how the brain adapts to these dual-tasking conditions. By 30
considering error-related brain activity in event-related potentials (ERPs), we ask how 31
monitoring processes are reorganized to maintain reliable error detection when the brain is 32
confronted with two temporally overlapping tasks. 33
Close temporal succession of more than one task leads to considerable interference 34
between the subtasks, resulting in performance decrements compared to the execution of 35
single tasks (Jersild, 1927; Telford, 1931). As dual-task interference arises predominantly in 36
central, decision-related processes, the brain adapts to this interference by serializing these 37
processes, granting attentional resources to only one task representation at a time (Meyer & 38
Kieras, 1997; Pashler, 1994). Neuroscientific studies suggested that interference as well as the 39
resulting serialization is linked to a fronto-parietal network that is able to grant access to 40
consciousness to only one task representation, by synchronizing distinct brain regions in a 41
self-amplifying process (Dehaene, Kerszberg, & Changeux, 1998; Del Cul, Baillet, & 42
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 4
Dehaene, 2007; Gross et al., 2004; Hesselmann, Flandin, & Dehaene, 2011; Sergent, Baillet, 43
& Dehaene, 2005; Sergent & Dehaene, 2004). 44
A number of electrophysiological studies has shown that the serialized access to 45
consciousness manifests in the P300 component (Dell’Acqua, Jolicoeur, Vespignani, & 46
Toffanin, 2005; Hesselmann et al., 2011; Sergent & Dehaene, 2004; Sigman & Dehaene, 47
2008). Sergent and colleagues (2005) observed a significantly reduced P300 on stimuli that 48
were not perceived consciously due to the continuous processing of a preceding stimulus, 49
whereas earlier visual potentials were equally elicited by consciously perceived and 50
unperceived stimuli. Similarly, Sigman and Dehaene (2008) found in a dual-tasking paradigm 51
that neural correlates of early attentional processes were executed in parallel for both tasks, 52
whereas the P300 for the second task (T2) was delayed until response selection for the first 53
task (T1), represented by its own P300, had been completed (see also Dell’Acqua et al., 2005; 54
Hesselmann et al., 2011). Recently, Marti, King, & Dehaene (2015) provided evidence that 55
this serialization of task selection is the result of a competition between task-related processes 56
when they strive for conscious access. As only one task representation at a time can get access 57
to consciousness, neural representations of the two tasks repel each other in a way that 58
initially, only processing of T1 is enabled, and that of T2 is actively delayed. 59
It has recently been suggested that also processes involved in the detection and 60
evaluation of performance errors suffer from dual-tasking interference. Research has robustly 61
identified two ERPs associated with distinct stages of error monitoring, which are executed in 62
a fixed temporal cascade immediately following the erroneous response (Ullsperger, Fischer, 63
Nigbur, & Endrass, 2014) – the error-related negativity (Ne/ERN, Falkenstein et al., 1991; 64
Gehring et al., 1993) and the error positivity (Pe, Overbeek et al., 2005). The Ne/ERN, an 65
early fronto-central negativity on error trials, has been assumed to reflect response conflict 66
(Yeung, Botvinick, & Cohen, 2004) or prediction errors (Holroyd & Coles, 2002) that 67
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 5
accompany performance errors. The Pe, on the other hand, is a later positive deflection over 68
parietal electrodes, which is supposed to reflect conscious error detection (Boldt & Yeung, 69
2015; Nieuwenhuis, Ridderinkhof, Blom, Band, & Kok, 2001; Overbeek et al., 2005; M. 70
Steinhauser & Yeung, 2010), likely resulting from the process of accumulating evidence that 71
an error has occurred (M. Steinhauser & Yeung, 2010). It has recently been shown that the Pe 72
and conscious error awareness can occur even if the earlier Ne/ERN is impaired or absent (Di 73
Gregorio, Maier, & Steinhauser, 2018; Maier, Di Gregorio, Muricchio, & Di Pellegrino, 74
2015a). Moreover, particularly the Pe interferes with subsequent stimulus processing (Buzzell 75
et al. 2017, but see Beatty et al., 2018) or is reduced when two temporally overlapping tasks 76
have to be executed (Weißbecker-Klaus et al., 2016). In contrast, the Ne/ERN appears to be 77
sensitive to working memory load even if no concurrent task execution is required (Klawohn 78
et al., 2016; Maier & Steinhauser, 2017; Moser, Moran, Schroder, Donnellan, & Yeung, 79
2013). The observation that mainly the Pe, and thus conscious error processing, suffers from 80
dual-tasking interference is in accordance with the aforementioned studies on dual-tasking. 81
The Pe exhibits striking physiological similarities with the P300, which, as outlined above, is 82
particularly affected by dual-tasking interference. It shares the time course and scalp 83
topography of the P300 (Leuthold & Sommer, 1999; Overbeek et al., 2005) and has been 84
linked to joint activation in bilateral prefrontal and parietal brain regions (Hester, Foxe, 85
Molholm, Shpaner, & Garavan, 2005), equivalently to the synchronization of brain areas that 86
was previously suggested to form the basis of the P300 (Dehaene et al., 1998; Sergent & 87
Dehaene, 2004; Sigman & Dehaene, 2008). 88
Whereas the similarity between the Pe and P300 can explain why specifically the Pe 89
suffers from dual-tasking interference, little is known how the brain maintains the ability to 90
consciously detect errors under dual-tasking. One solution to this problem would be to 91
serialize not only task processing but also the accompanying error monitoring processes 92
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 6
(Hochman & Meiran, 2005). Processing a T1 error could lead to an additional deferment of 93
T2 processing, an effect that has indeed been demonstrated in behavioral studies (Jentzsch & 94
Dudschig, 2009; M. Steinhauser, Ernst, & Ibald, 2017). However, reduced error signaling 95
rates when an erroneous response is quickly followed by another task (Rabbitt, 2002) and the 96
observation of a reduced Pe in dual-tasking (Weißbecker-Klaus et al., 2016) suggest that such 97
a serialization cannot fully protect error monitoring from dual-task interference. 98
Consequently, it might be necessary to actively defer conscious error detection to the end of a 99
dual-task scenario. In this case, processing a T1 error would in some or all trials not occur 100
until the response to T2. Such an adaptive rescheduling of conscious error processing would 101
imply that the accumulation of evidence for an error that underlies the Pe can be detached 102
from early error signals represented by the Ne/ERN. 103
The present study investigated this adaptive rescheduling hypothesis, which proposes 104
that the brain adapts to dual-task interference in error monitoring by deferring conscious error 105
detection to after completion of the dual task. To this end, we focused on T1 errors but 106
analyzed ERPs following both the T1-response and the T2-response. We contrasted trials with 107
a sufficiently long interval between the two task stimuli that allowed serial task execution 108
(1200 ms) with trials that required overlapping task execution due to a considerably shorter 109
interval (300 ms). In serial task execution, we expected to see the typical Ne/ERN and Pe 110
after the erroneous T1-response. Crucially, in overlapping task execution, we predicted that 111
the Pe after the T1-response would be reduced or even absent while a Pe would now be 112
obtained after the correct T2-response. Such a result would demonstrate that the decision 113
about the correctness of T1 is deferred in some or all trials, respectively, until the completion 114
of T2 and extend previous accounts on an active deferment of component processes in dual-115
tasking (Hesselmann et al., 2011; Sigman & Dehaene, 2008; R. Steinhauser & Steinhauser, 116
2018) to the area of performance monitoring. 117
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 7
Results 118
Experiment 1 119
24 healthy adult participants worked on a variant of the psychological refractory 120
period (PRP) paradigm, which is commonly used to investigate mutual interference between 121
subtasks in dual-tasking situations (Pashler, 1994; Tombu & Jolicœur, 2003). The details of 122
the experimental paradigm can be found in Figure 1. 123
124
125
126
Figure 1. Time course of two example trials. On each trial, participants had to respond 127
to two tasks that were separated by an unpredictable stimulus onset asynchrony of 300 ms 128
(overlapping task execution, A) or 1200 ms (serial task execution, B). A three-choice color 129
variant of the flanker task, in which the color of the central square had to be indicated (red, 130
yellow, blue), served as Task 1. A two-choice pitch discrimination task was presented 131
subsequently as Task 2, with an auditory sine tone stimulus of 400 Hz or 900 Hz. Participants 132
were instructed to respond as fast as possible to both tasks, in the given order. S1 = Stimulus 133
1. S2 = Stimulus 2. R1 = Response 1. R2 = Response 2. RT1 = response time to Task 1. RT2 134
= response time to Task 2. 135
136
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 8
Behavior. RTs and error rates are depicted in Figure 2. To verify that our paradigm 137
creates a dual-tasking scenario with overlapping task execution, we first examined whether 138
two typical effects of dual-tasking can be observed in this dataset: First, the so-called PRP 139
effect refers to the observation that RTs to T2 increase with a decreasing stimulus onset 140
asynchrony (SOA), and thus indicates a form of dual-task cost. This effect is typically 141
explained by the idea that, with more overlap between tasks, T2 execution is delayed (Pashler, 142
1994) or suffers from depleted resources (Tombu and Jolicœur, 2003). Second, it has recently 143
been shown that T1 errors lead to increased RTs to T2, and this post-error slowing is larger 144
for short SOAs than for long SOAs (Steinhauser et al., 2017). This phenomenon presumably 145
reflects interference between T1 error monitoring and execution of T2, which again is larger 146
with short SOAs. 147
To analyze both effects in our data, we subjected RTs for T1 and T2 separately to 148
repeated measures ANOVAs on the variables SOA (short vs. long) and T1 Correctness 149
(correct vs. error). Characteristic of PRP paradigms, no significant effects were found for RTs 150
to T1, Fs(1, 23) < 2.70, ps > .11, ηp² < .11. In contrast, RTs to T2 were considerably slower 151
on trials with short SOA than with long SOA, F(1, 23) = 83.4, p < .001, ηp² = .78, reflecting 152
the typical PRP effect. In addition, RTs to T2 were slower when the preceding T1-response 153
was an error, F(1,23) = 56.8, p < .001, ηp² = .71, and as expected, this post-error slowing on 154
T2 was increased with short SOA, F(1, 23) = 25.9, p < .01, ηp² = .53. Thus, our data replicate 155
well-known signatures of dual-task interference, which demonstrates that the short-SOA 156
condition induced overlapping task execution. In addition, an analysis of inter-response 157
intervals (IRIs, the time between the T1 response and the T2 response) in the short-SOA 158
condition ruled out that participants had grouped their responses. While IRIs were 159
considerable longer on T1 errors (587 ms) compared to corrects (381 ms) due to the above 160
mentioned post-error slowing, t(23) = 8.35, p < .001, d = 1.25, IRIs in both conditions were 161
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 9
far beyond the commonly used threshold for response grouping of 50 – 100 ms (Hommel, 162
1998; Pashler & Johnston, 1989; Welford, 1952). 163
Error rates for T1 were high enough for an analysis of error-related brain activity. A 164
mean T1 error rate of 8.31 % resulted in an average number of 98.3 trials with T1 errors but 165
correct T2 responses per participant. A repeated measures ANOVA on the variables SOA 166
(short vs. long) and Task (T1 vs. T2) yielded a significant interaction, F(1, 23) = 8.06, p = 167
.009, ηp² = .26, indicating that, whereas T1 errors were equally frequent in short-SOA and 168
long-SOA trials, T2 errors were more frequent in short-SOA trials as compared to long-SOA 169
trials, t(23) = 2.69, p = .013, d = .50. This again demonstrates increased interference in the 170
short-SOA condition. 171
172
173
Figure 2. Behavioral results of Experiment 1. RTs to Task 1 and Task 2 are depicted 174
in the left and middle panel, respectively. Error rates are presented in the right panel. Error 175
bars indicate within-subject standard errors of the mean (Cousineau, 2005; Morey, 2008). RT 176
= response time. SOA = stimulus onset asynchrony. T1 = Task 1. T2 = Task 2. 177
178
179
Ne/ERN and Pe. To investigate our rescheduling hypothesis, we analyzed error-related 180
brain activity in short-SOA trials, which induce overlapping task execution, and long-SOA 181
trials, which allow for serial task execution. Our central prediction was that, with a short 182
SOA, the Pe associated with T1 errors should partially be rescheduled to the end of the dual-183
task. This would result in a reduced Pe following incorrect T1 responses but the emergence of 184
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 10
a Pe after correct T2 responses. We had no explicit hypothesis on the Ne/ERN but 185
nevertheless analyzed this component to determine whether comparable effects can be found 186
for this early form of error processing. 187
We initially analyzed T1-response-locked ERPs to find out about immediate neural 188
correlates of error processing after T1 errors. A distinct parietal positivity, the Pe, was clearly 189
observable for both SOA conditions (Fig 3A). For an analysis of the Pe, we subjected mean 190
amplitudes at electrode POz to a repeated measures ANOVA on the variables SOA (short vs. 191
long) and T1 Correctness (correct vs. error). A Pe was evident across both SOA conditions, as 192
indicated by a main effect of T1 Correctness, F(1, 23) = 28.6, p < .001, ηp² = .55, but a 193
significant interaction revealed that this Pe was far less pronounced in trials with short SOA 194
than in trials with long SOA, F(1, 23) = 11.0, p = .003, ηp² = .32. Raw ERP waves in Figure 3 195
suggest that this interaction is mainly driven by a difference in the amplitudes of correct trials. 196
This is likely caused by the fact that in most trials of the short SOA condition, the stimulus of 197
T2 is presented and processed sometime within the time window observed here, eliciting 198
stimulus-locked ERPs that are superimposed on the raw ERP waveforms of both correct and 199
error trials in the short SOA condition (see Sigman & Dehaene, 2008). For this reason, we 200
quantify the Pe not from error trials alone but as the difference between correct and error 201
trials. To additionally rule out that the observed interaction may truly be rooted in differences 202
in correct trials, we correlated the difference wave of conditions correct long and correct short 203
with the difference wave of the interaction term of the above analysis. We found no 204
significant correlation, r = 0.14, p = .50, indicating that the Pe reduction in trials with short 205
SOA is not a mere consequence of the amplitude difference in correct trials. 206
Figure 3B shows that also a clear frontocentral Ne/ERN was observable in both SOA 207
conditions. Indeed, subjecting mean amplitudes at electrode FCz to a repeated measures 208
ANOVA of the same variables as above revealed that a significant Ne/ERN across both SOAs 209
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 11
was obtained, F(1, 23) = 22.8, p < .001, ηp² = .50. In addition, a main effect of SOA 210
demonstrated that mean amplitudes of correct as well as error trials were more negative in 211
trials with short SOA, F(1, 23) = 13.5, p = .001, ηp² = .37. However, the interaction between 212
both variables did not reach significance, F(1, 23) = 2.63, p = .12, ηp² = .10. 213
214
215
Figure 3. ERPs locked to the T1-response at posterior (A) and frontocentral (B) 216
electrodes. Difference waves are computed from the respective T1 error minus T1 correct raw 217
ERPs. Scalp topographies represent these difference waves. Gray areas indicate the time 218
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 12
intervals for statistical testing for the Pe (A) and the Ne/ERN (B). R1 = Response 1. T1 = 219
Task 1. SOA = stimulus onset asynchrony. 220
221
222
Having shown that conscious error detection reflected by the Pe was largely impaired 223
in the short-SOA condition, we subsequently examined neural correlates of T1 error 224
processing after completion of the whole dual-task, i.e., in T2-response-locked data (Fig. 4). 225
We investigated the possible emergence of such a deferred Pe (Fig. 4A) by means of 226
ANOVAs on the variables SOA (short vs. long) and T1 Correctness (T1 correct vs. T1 error). 227
It must be noted, though, that both conditions of the variable T1 Correctness here represent 228
trials whose T2 was answered correctly, i.e., the “error-related” brain activity reported here 229
was locked to a correct T2-response. Nonetheless, we obtained a significant interaction 230
between SOA and T1 Correctness on Pe amplitudes, F(1, 23) = 5.55, p = .027, ηp² = .19, 231
reflecting that correct T2-responses elicited a significant Pe on short-SOA trials, t(23) = 2.00, 232
p = .029, d = .27, but not on long-SOA trials. 233
The same analysis on the Ne/ERN amplitudes revealed only a marginally significant 234
main effect of T1 Correctness, F(1, 23) = 3.48, p = .075, η²part. = .13, indicating a negativity 235
after T1 errors compared to T1 correct trials, but no significant interaction, F(1, 23) = 1.73, p 236
= .20. Visual inspection of the topographies (Fig. 4B) revealed that this effect peaked earlier 237
than the typical Ne/ERN and had a more frontal distribution than the Ne/ERN in T1-response-238
locked potentials (however, the same analysis at electrode Fz showed similar results). This 239
suggests that, while there appears to be some frontocentral activity related to the T1 error in 240
T2-response-locked data, this effect lacks robustness and differs from the usually observed 241
Ne/ERN. 242
To sum up, we could provide evidence that overlapping task execution goes along 243
with a reduction of the Pe to the incorrect T1 response whereas a sizeable Pe emerges after 244
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 13
the correct T2-response. We interpret this deferred Pe as a rescheduling of conscious error 245
processing to the end of the dual-task trial. 246
247
248
Figure 4. ERPs locked to the (correct) T2-response at posterior (A) and frontocentral 249
(B) electrodes. Difference waves are computed from the respective T1 error minus T1 correct 250
raw ERPs. Scalp topographies represent these difference waves. Gray areas indicate the time 251
intervals for statistical testing for the Pe (A) and the Ne/ERN (B). R2 = Response 1. T1 = 252
Task 1. SOA = stimulus onset asynchrony. 253
254
255
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 14
Relationship between immediate and deferred error processing. Although the 256
immediate Pe that followed the T1 response was strongly reduced in the short-SOA condition, 257
it was observable there nonetheless. This suggests two possible ways of how error monitoring 258
is carried out under conditions of overlapping task execution. On the one hand, the 259
accumulation of evidence that a T1 error has occurred could be evenly distributed between 260
immediate (following T1 response) and deferred (following T2 response) error monitoring. 261
On the other hand, error monitoring could occur for some trials immediately after the 262
erroneous T1 response, whereas it is deferred until after the T2 response in other trials. To 263
distinguish between these two accounts, we examined the trial-wise relationship between 264
immediate and deferred Pe in two ways. 265
First, we investigated the inverse relationship of immediate and deferred error 266
processing by deriving single-trial estimates for the R1-locked and the R2-locked Pe from 267
mean amplitudes in raw data and comparing them in a regression-based analysis. This 268
analysis had to be limited to trials with IRI > 400 ms, however, because raw EEG data within 269
the same epoch are highly susceptible to autocorrelations in overlapping time periods. 270
Importantly, this does not affect the remaining analyses of the present study, because EEG 271
data from different, individually baseline-corrected epochs are compared. The restriction to 272
IRIs above 400 ms resulted in M = 31.04 trials per participant. Figure 5 depicts strong 273
negative beta weights in a significant cluster from 290 ms onwards in R1-locked data and 170 274
ms in R2-locked data, that is, around the time of the R1-locked and R2-locked Pe. This is 275
evidence for a trade-off of the immediate and the delayed Pe, supporting the account that error 276
processing happens immediately after some erroneous responses, whereas it is deferred on 277
other trials. 278
Second, we used multivariate pattern analysis (MVPA) to create a set of classifiers 279
over consecutive time windows that optimally distinguished between correct trials and T1 280
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 15
error trials based on T1-response-locked brain activity, which allowed us to examine all trials 281
irrespective of their IRI. Mean classifier accuracy peaked around 190 ms with an Az of .61 282
but inspection of participants’ individual classifier accuracies revealed that one participant 283
(Subject 13) exhibited a remarkably low classification accuracy of .34, which is more than 2.5 284
standard deviations below the mean classification accuracy across all participants. This 285
apparently failed attempt to compute a successful MVPA for this participant is likely due to 286
the small number of 20 trials that entered the training set (the average training set size of all 287
participants was 120.63 trials). For this reason, data from Subject 13 was excluded from 288
subsequent MVPA-based analyses. Nonetheless, all effects remain significant also when 289
including that participant (all ps < .05). Based on the peak classifier window, we computed 290
the prediction value for each trial. These prediction values represent the degree to which each 291
trial elicits error-related brain activity in this time window (Boldt & Yeung, 2015; Steinhauser 292
& Yeung, 2010), and thus represent a single-trial estimate of the T1-response-locked Pe. We 293
utilized these prediction values to split the error trials of each participant into three equally-294
sized bins that represent a small, medium, and large T1-response-locked Pe (Fig. 5A). A one-295
way repeated measures ANOVA on mean Pe amplitudes with the variable Bin (small Pe vs. 296
medium Pe vs. large Pe) showed that this MVPA-based separation process was in fact able to 297
separate trials according to Pe size, F(2, 44) = 4.72, p = .015, ηp² = .18. Contrasts revealed 298
that this effect was mainly driven by trials in the large Pe bin having a larger T1-response-299
locked Pe than trials in the small Pe bin, t(22) = 2.90, p = .008, d = .52. Crucially, an 300
equivalent analysis of T2-response-locked Pe values based on this T1-response-based bin 301
separation (Fig. 5B) shows the opposite pattern, F(2, 44) = 6.75, p = .004, ηp² = .23. Trials 302
with a large immediate Pe following the T1-response showed a particularly small deferred Pe 303
following the T2-response, t(22) = 4.10, p < .001, d = .63. This conceptually replicates the 304
above regression-based findings on a trade-off of the immediate and deferred Pe and thus 305
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 16
provides additional support for the idea that the rescheduling of conscious error processing is 306
all-or-nothing. Apparently, T1-error processing is deferred in some trials to the end of the 307
dual-task, whereas in other trials it occurs immediately after the incorrect T1 response. 308
The MVPA-based differentiation of trials with small, medium and large immediate Pe 309
eventually also allowed us to investigate how immediate error processing affects selecting and 310
executing the response to Task 2 (Fig. 5C). We compared IRIs in a one-way repeated 311
measures ANOVA (small vs. medium vs. large immediate Pe) and found a significant 312
difference between the conditions, F(2,44) = 3.42, p = .045, ηp² = .13. Contrasts confirmed 313
that trials with a large immediate Pe were followed by slower responses to T2 than trials with 314
a small immediate Pe, t(22) = 2.94, p = .001, d = .40, which in turn supports the idea that 315
immediate error processing negatively affects efficient execution of the second task. 316
317
318
319
320
Figure 5 Regression weights of the regression-based single-trial analyses for 321
Experiments 1-3, in which amplitudes of R1-locked data predict amplitudes of R2-locked 322
data. Black lines indicate the borders of significant clusters as revealed by a cluster-based 323
permutation test. Red dotted lines indicate the peak amplitude of the Pe in the respective 324
difference waves. While on the numerical level, a negative association of the R1-locked Pe 325
and R2-locked Pe becomes evident in all three experiments, statistical significance is reached 326
only in Experiments 1 and 2. 327
328
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 17
329
Figure 6. ERPs at a posterior electrode locked to the T1-response (A) and T2-response 330
(B) and inter-response intervals after T1 errors (C). T1 error trials are divided into three 331
separate conditions – small, medium, and large immediate Pe – based on a single-trial 332
estimate of the Pe. Correct trials are presented for comparison (thin line). A 20 Hz lowpass 333
filter was applied for better visibility. Gray areas indicate the time intervals for statistical 334
testing for the Pe. R1 = Response 1. R2 = Response 2. T1 = Task 1. T2 = Task 2. 335
336
337
338
339
340
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 18
Experiment 2 341
Individual trials that elicit a deferred Pe with simultaneous omission of an immediate 342
Pe could also be rooted in the problem of credit assignment (Fu & Anderson, 2008; Sutton & 343
Barto, 1998; Walsh & Anderson, 2011). It is conceivable that trials with a deferred Pe 344
actually reflect that error monitoring occasionally misattributed internal error signals to T2, 345
and therefore falsely detected a T2 error. Rather than a rescheduling of T1-error processing, 346
such a Pe would represent immediate T2-error processing. In a follow-up experiment with 24 347
different participants (out of which one participant had to be excluded due to technical 348
difficulties during data acquisition), we did not only want to replicate our initial findings but 349
also investigated whether this alternative explanation could account for our findings. 350
In Experiment 2, participants worked on the same PRP paradigm but reported after 351
each trial by key press, whether they had committed an error in T1, T2, or both. This allowed 352
us to distinguish between correctly assigned T1 errors (T1 hits) and T1 errors that were 353
mistakenly reported as T2 errors (T2 false alarms). Rates of averaged error reports are 354
depicted in Figure 7A. In fact, only 4.9% of all T1 errors were T2 false alarms (on average 0.7 355
trials per participant) and thus occurred very rarely. As only 17.0% of T1 errors were reported 356
to be correct (T1 misses; on average 6.78 trials per participant), we could base our further 357
ERP analyses on a solid subset of 73.8% of T1 hits (on average 31.0 trials per participant). 358
359
360
361
362
363
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 19
364
Figure 7. Error reports (A) and ERP results (B) of Experiment 2. Error bars in Panel A 365
indicate within-subject standard errors of the mean (Cousineau, 2005; Morey, 2008). Panel B 366
represents ERPs locked to the T2-response at posterior (left) and frontocentral (right) 367
electrodes. Difference waves (dotted lines) are computed from T1 error minus correct raw 368
ERPs. Scalp topographies represent these difference waves. Gray areas indicate the time 369
intervals for statistical testing for the Pe (left) and the Ne/ERN (right). R1 = Response 1. R2 = 370
Response 2. T1 = Task 1. T2 = Task 2.371
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 20
To compensate for the increased trial duration due to trial-wise error reporting,
Experiment 2 was restricted to the short SOA condition. Only T2-response-locked data are
reported here as no immediate Ne/ERN and Pe could be compared between SOA conditions.
Although only T1 hits were included in the analysis, a clear deferred Pe could be observed
after T2-response-locked data (Fig. 6B), t(22) = 2.86, p = .009, d = .57.
As in Experiment 1, the analysis of the Ne/ERN in T2-response-locked data revealed
only a marginally significant negativity for T1 errors relative to correct trials at electrode FCz,
t(22) = 1.87, p = .075, d = .51. However, inspection of the topographies (Fig. 7B) showed that
the peak of the Ne/ERN was again at more frontal electrodes. Correspondingly, additional
testing was conducted at electrode Fz and now revealed a significant deferred Ne/ERN, t(22)
= 2.26, p = .034, d = .62.
The regression-based approach on raw data with IRI > 400ms yielded a pattern with
close similarity to that in Experiment 1, albeit somewhat earlier in R2-locked data (Figure 6).
Again, strong negative beta weights form a significant cluster from 290 ms onwards in R1-
locked data and between 125 ms and 230 ms in R2-locked data. We tentatively suggest that
this negative correlation between the R1-locked Pe and the R2-locked Pe is reduced in the
R2-locked time window beyond 250 ms due to the strong influence of a pronounced
overlaying Contingent Negative Variation in anticipation of the error report (see Figure 7B).
However, a robust MVPA-based bin separation of trials with large vs. medium vs. small
immediate Pe as in Experiment 1 was not possible for Experiment 2 because classifier
training did not yield a classification accuracy above the 5% significance threshold as
established by a permutation test, likely due to the smaller size of the training data set.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 21
Taken together, we could replicate the deferred Pe in Experiment 2. As this analysis
was restricted to those error trials that were correctly reported as T1 errors, we can rule out
that this deferred Pe originates in occasional T1-error trials that are misattributed as T2-errors.
Experiment 3
While a link between the Pe and conscious error awareness has been established in
previous studies (Boldt & Yeung, 2015; Maier, Di Gregorio, Muricchio, & Di Pellegrino,
2015b; Nieuwenhuis et al., 2001; Overbeek et al., 2005; M. Steinhauser & Yeung, 2010),
direct empirical evidence for this assumption within the scope of the present paradigm is still
missing. Although Experiment 2 included behavioral error reports, a comparison of
consciously aware and unaware errors was impossible due to the small number of T1 error
trials that were reported as correct. We consequently conducted a variant of Experiment 2 as
an additional experiment with a larger number of 40 new subjects (of which 3 subjects had to
be excluded due to technical problems during data acquisition), which allowed us to conduct
analyses on sub-pools of subjects with a minimal number of 6 trials in all relevant conditions.
This was previously suggested as a minimum threshold to obtain reliable measures of error
processing in EEG data (Olvet & Hajcak, 2009). To additionally address the question if the
objective delay of the Pe in multitasking directly relates to a subjective feeling of later error
detection, we added a second question after every trial. Here, participants were asked if they
had the subjective feeling of error awareness at an early or late point of time (see also Di
Gregorio, Maier, & Steinhauser, 2020). As a result, Experiment 3 provided two variants of
behavioral measures for conscious error awareness.
Error reports in Experiment 3 closely resemble those in Experiment 2. 72.34 % of the
T1 errors were correctly reported, whereas 24.77 % were not reported at all and only 2.9 %
were falsely reported as T2 errors. Of the correctly reported T1 errors, 61.94 % were reported
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 22
with a subjective feeling of immediate error awareness and 35.85 % with a feeling of delayed
error awareness, whereas only 2.22% were not classified in this regard at all, replicating the
pattern found by Di Gregorio et al. (2020) in a single-task scenario. The larger number of
participants in Experiment 3 allowed us to create a sub-pool of 21/22 participants for an ERP
analysis of reported vs. missed T1 errors (1 subject fell below the threshold of 6 trials only in
R1-locked analysis as artefact rejection was conducted separately on the two datasets; see
Methods section), and a sub-pool of 25 participants for an ERP analysis of reported T1 errors
with a feeling of immediate vs. delayed error awareness.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 23
Figure 8. ERP results of the two main ERP analyses of Experiment 3. Error reports
(A) and ERP results (B) of Experiment 2. Gray areas indicate the time intervals for statistical
testing for the Pe in R1-locked data (left) and R2-locked data (right). R1 = Response 1. R2 =
Response 2. T1 = Task 1. T2 = Task 2.
Figure 8 displays a pronounced Pe both in R1-locked data, F(2,40) = 7.03, p = .004,
ηp² = .26, and in R2-locked data, F(2,42) = 6.86, p = .004, ηp² = .25, once again replicating our
initial finding of a deferred Pe after execution of the correct second task. Also a considerable
reduction of the Pe in misses vs hits can be seen in both analyses. While in R1-locked data,
this difference only crosses the significance threshold in a somewhat later time window of
400-600 ms, t(20) = 2.64, p = .016, d = .51, in R2-locked data, misses and hits exhibit a
significant difference also in the a-priori time window of 200-400 ms, t(21) = 2.23, p = .037,
d = .48. These findings suggest a direct link between conscious error reports and both the
immediate and the deferred Pe.
In contrast to this previous analysis, differentiating T1 error trials with the subjective
feeling of immediate vs. deferred error awareness, however, did not yield any significant
differences in Pe amplitudes, all ts(24) < .72, all ps(24) > .47.
And equivalently to Experiment 2, an MVPA-based bin separation was not possible
here, either, because classifier training did not yield a classification accuracy beyond the 5.0
% significance threshold. The regression-based approach on trials with IRI > 400ms,
however, though not yielding a significant cluster, exhibits a descriptive pattern similar to that
in Experiment 1, with a peak in negative beta weights around 325 ms in R1-locked data and
315 ms in R2-locked data, temporally close to the peak amplitudes of the R1-locked and R2-
locked Pe.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 24
Discussion
In three ERP experiments, we investigated neural correlates of error monitoring in
dual-tasking. Previous studies suggested that increased interference due to temporally
overlapping tasks is countered by an active deferment of task components (Hesselmann et al.,
2011; Meyer & Kieras, 1997; Sigman & Dehaene, 2008). Based on this idea, we hypothesized
that because conscious processing of T1 errors can interfere with the execution of T2 (M.
Steinhauser et al. 2017), also performance monitoring, more precisely the conscious
evaluation of T1 errors, is adaptively rescheduled to after the completion of the whole dual-
task. To this end, we analyzed the Pe, an ERP component that can be linked to conscious
error awareness based on previous studies (Boldt & Yeung, 2015; Di Gregorio et al., 2018,
2020; Maier et al., 2015a; Nieuwenhuis et al., 2001; Overbeek et al., 2005; M. Steinhauser &
Yeung, 2010) as well as based on an empirical confirmation from behavioral error reports in
Experiment 3. In all three experiments, we found a result pattern that confirms our prediction.
When the execution of two tasks overlaps, errors in the first task lead to a reduced immediate
Pe whereas a deferred Pe appears after the response to the second task.
Our results are in line with the idea that conscious error processing as reflected by the
Pe interferes with the execution of a subsequent task (Buzzell et al., 2017; Weißbecker-Klaus
et al., 2016). In our study, only the Pe but not the Ne/ERN after T1 errors was reduced when
task execution was overlapping. This conforms with recent findings by Weißbecker-Klaus et
al. (2016), who found a similar reduction of the Pe when a flanker task was executed
concurrently with a semantic task. In both studies, this impairment to immediate conscious
error detection likely originates from interference by T2, which requires cognitive resources
for T2-response selection at the same time as T1-error processing would occur. The decision
process for selecting the correct response to T2 was previously suggested to rely on similar
resources as conscious error processing (Hochman & Meiran, 2005). On the neural level, this
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 25
can well be explained by recent electrophysiological studies that view also conscious error
processing as a decision process (Wessel, 2012; Wessel, Danielmeier, & Ullsperger, 2011)
and studies that highlight the physiological similarities of the P300 and the Pe (Leuthold &
Sommer, 1999; Overbeek et al., 2005). Given that interference between central decision
processes is seen as the main origin of dual-task costs (Pashler 1994; Tombu and Jolicœur,
2003), this explains why particularly the Pe as a neural correlate of conscious error detection
(Overbeek et al., 2005) is impaired in dual-tasking.
In addition to a reduced immediate Pe after the erroneous T1 response, we found a
deferred Pe after completion of the dual-task even though the second response itself was
correct. MVPA-based single-trial estimates of the Pe as well as a regression-based approach
on raw data suggest that the deferred Pe occurred predominantly on trials with little
immediate conscious error processing. This temporal detachment of conscious error
processing from the early error signals that are represented by the Ne/ERN – on average by
more than 580 ms – goes beyond a mere serialization as it was previously found for the P300
in response selection (Hesselmann et al., 2011; Marti et al., 2015; Sigman & Dehaene, 2008).
In the present study, conscious error detection appears to be suspended until a whole
additional task has been executed, which is potentially implemented on the neural level as the
result of error evidence accumulation being stored in a buffer because the fronto-parietal
network that represents conscious access is still engaged in stimulus processing of Task 2
(Meyer & Kieras, 1997). Analysis of IRIs furthermore confirmed that the deferment of error
processing is associated with faster responses to T2, suggesting reduced dual-tasking
interference on such trials, in line with Buzzell et al.’s (2017) finding that a smaller Pe is also
linked to improved sensory processing in the subsequent trial. This is strong support for the
adaptive rescheduling account of error detection in dual-tasking, which suggests such a
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 26
mechanism to reduce interference between the subtask representations and hence to maintain
the ability to detect and evaluate errors.
Analyzing brain activity in a PRP paradigm bears the risk that activity related to T1 is
accidentally attributed to T2 (Sigman & Dehaene, 2008). For instance, what appears to be a
deferred Pe after the T2-response could simply be a carry-over of the immediate Pe after the
T1-response. However, there are several reasons why this cannot account for the present
results. First, the deferred Pe shows a distinct time course, emerging about 200 ms after the
T2 response, which is comparable to the latency of the immediate Pe as well as of the Pe in
other studies (Overbeek et al., 2005). Second, even with a short SOA of 300 ms, the average
interval between the T1-response and T2-response on error trials was 587 ms. That is, the
onset of the immediate Pe is clearly prior to the T2-response, and any sustained differences
between T1 errors and correct trials emerging prior to the T2-response are controlled by the
pre-response baseline. Finally, and most importantly, our results indicate an inverse
correlation between immediate Pe and deferred Pe. If the deferred Pe reflected a carry-over of
activity from the immediate Pe, both effects should be positively correlated.
In Experiments 2 and 3, we could also rule out that the deferred Pe results from
internal misattribution of occasional T1 errors as T2 errors, which would imply a credit
assignment problem (Fu & Anderson, 2008; Sutton & Barto, 1998; Walsh & Anderson, 2011)
as the origin of our findings. Trial-wise error-reports in Experiments 2 and 3 revealed that
credit-assignment errors occurred extremely rarely in our tasks. Moreover, we obtained a
deferred Pe even on the subset of correctly assigned T1 errors. This suggests that even when
conscious processing of the T1 error is postponed, these errors are still detected and correctly
assigned to their corresponding task. One could speculate that the proposed scheduling
mechanism even contributes to solving the credit assignment problem. On the one hand,
rescheduling counteracts interference between T1 error processing and T2 response selection
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 27
which otherwise would increase the risk of confusing error signals from T1 and T2 responses.
On the other hand, when conscious error evaluation occurs after both responses were
executed, the involved decision process could utilize cues from both tasks to correctly assign
error signals (like post-response conflict) to their corresponding tasks.
Experiment 3 was an important addition to this study, because only by being able to
analyze a considerably large number of participants with regard to reported vs. missed T1
errors, we could empirically confirm that the Pe represents conscious error awareness also in
the present dual-tasking paradigm. Replicating a core finding of previous single-task studies
(Boldt & Yeung, 2015; Di Gregorio, Steinhauser, & Maier, 2016; Maier et al., 2015a;
Nieuwenhuis et al., 2001; Overbeek et al., 2005; M. Steinhauser & Yeung, 2010), the Pe was
reduced immediately after an erroneous T1 response when participants did not report that
error. Interestingly, an equivalent reduction was also found on missed T1 errors for the
deferred Pe following the T2 response. This substantiates that also this later ERP component
is directly linked to conscious error awareness. The analysis with regard to a subjective
feeling of early vs. late awareness, on the other hand, did not yield any significant results. The
immediate and the deferred Pe showed similar amplitudes in both conditions. This is,
however, in line with Di Gregorio et al.’s (2020) argumentation that the subjective feeling of
early vs. late error awareness may well be a metacognitive illusion that is created at a later
point in time. It is hence plausible that the trial-wise objective delay of the Pe is independent
from subjective deferments of error awareness.
Though far less robust than the deferred Pe, we also obtained frontocentral activity
related to T1 errors following the T2 response. This could reflect a deferred Ne/ERN,
although this activity had a slightly different time course and spatial distribution as the
immediate Ne/ERN after T1-responses. However, this deferred Ne/ERN differed in a crucial
aspect from the deferred Pe as it was not associated with a reduced immediate Ne/ERN.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 28
Without this trade-off between immediate and deferred error processing, this effect cannot
reflect a scheduling mechanism. It is also implausible that the deferred Ne/ERN reflects post-
response conflict induced by a corrective response tendency (Yeung et al., 2004) or a reward
prediction error (Holroyd & Coles, 2002) as both these mechanisms are temporally linked to
the erroneous response. Instead, the deferred Ne/ERN could represent a negative affective
signal to the T2-response. Several studies could show that the Ne/ERN involves affective
processing of an error (Aarts, Houwer, & Pourtois, 2013; Maier, Scarpazza, Starita, Filogamo,
& Ladavas, 2016). The deferred Ne/ERN hence could reflect that committing an error in one
task leads to a devaluation also of the second task, indicating that the whole dual-task trial is
processed as erroneous.
Taken together, our results suggest that reliable error detection in dual-tasking is
maintained by a mechanism that adaptively reschedules the accumulation of evidence
underlying conscious error processing to after the completion of the whole dual-task. Our
findings indicate that the flexible reorganization of componential processes under dual-
tasking is not only a viable strategy to prevent interference between decision processes
involved in task execution (Hesselmann et al., 2011; Marti et al., 2015; Meyer & Kieras,
1997; Sigman & Dehaene, 2008), but also serves to reduce interference between task
execution and performance monitoring.
Materials and methods
Participants
We conducted three separate, consecutive experiments with different groups of
participants. 24 healthy students (2 male; age: M = 22.3 years; SD = 4.5 years; 2 left-handed)
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 29
participated in Experiment 1, 24 healthy students (2 male; age: M = 24.0 years; SD = 3.9; 2
left-handed) participated in Experiment 2 and 40 healthy students participated in Experiment
3 (4 male; age: M = 21.7 years; SD = 2.5; 2 left-handed). With an assumed correlation of .5
among cell means, a sample size of N = 24 should allow for detecting medium-sized effects (f
= .25) in the present repeated-measures design with a statistical power of .82. All participants
were recruited from the Catholic University of Eichstätt-Ingolstadt and received course credit
or payment (8€ per hour). Informed consent was provided by all participants and the study
was approved by the ethics committee of the Catholic University of Eichstätt-Ingolstadt. One
subject in Experiment 2 and 3 subjects in Experiment 3 had to be excluded from further
analysis due to technical problems during data acquisition.
Task and procedure
Experiment 1. Adopting the PRP paradigm from Steinhauser, Ernst, and Ibald (2017),
we combined an error-prone three-choice color flanker task and a two-choice pitch
discrimination task. The flanker task stimulus (see Fig. 1) consisted of three horizontally
arranged squares of .82° edge length. The central target square and the two flanker squares
were either red, yellow or blue. While both flanker squares had the same color, this color was
always different from that of the target. As three colors were used, it was possible to
exclusively present these more error-prone incongruent stimuli without enabling participants
to infer the target color from the flankers. Stimuli for the secondary pitch task were low (400
Hz) and high (900 Hz) sine tones.
The trial procedure is depicted in Figure 1. Participants first had to respond to the
flanker task (T1) and then to the pitch discrimination task (T2). Each trial started with the
presentation of a fixation cross for 500 ms. Following this, the flanker task was presented by
initially displaying the two flankers for 60 ms and then the target together with the flankers
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 30
for 200 ms. The pitch discrimination task started after an SOA of either 300 ms (short) or
1200 ms (long), which was chosen randomly within the blocks. The stimulus for the pitch
task was presented for 150 ms. After responses to both tasks were given, the fixation cross of
the next trial was presented 500 ms after the response. The short SOA was supposed to create
a dual-tasking situation with overlapping task execution while at the same time preventing
excessive response grouping, which was observed in a piloting phase with even shorter SOAs.
In contrast, the long SOA was intended to serve as a baseline that allows for serial task
execution.
For the flanker task, participants had to indicate the color of the central target square
by pressing the “Y”, “X”, or “C” button on a German QWERTZ keyboard (in which Y and Z
are exchanged) with their left hand (the T1-response). For the pitch discrimination task,
participants entered their response with “arrow down” for a low pitch and “arrow up” for a
high pitch (the T2-response). The mapping of colors to keys was counterbalanced across
participants, and for one half of the participants, the task-to-hand assignment was reversed
(resulting in “,”, “.”, “-” for the flanker task and “A” and “Y” for the pitch task). If the T2-
response was given before the onset of the pitch stimulus, written feedback (“AUFGABE 2
ZU FRÜH!”, engl. task 2 too early) was given immediately and the trial was excluded from
data analysis.
The experiment started with a series of practice blocks that ensured that participants
had thoroughly learned the experimental paradigm and the mappings of colors/pitches to
keys. First, in three blocks of 24 trials the flanker task was presented alone. Then, one block
of 16 trials served to practice the pitch task alone, and in one subsequent block of 36 trials,
both tasks were responded to together. Actual testing consisted of 10 blocks of 108 trials
each, resulting in a total number of 1080 trials. Oral feedback to respond faster was given
after blocks, if the error rate for any of the tasks was below 10%.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 31
Experiment 2. Experiment 2 adopted the experimental method of Experiment 1 but
added a trial-wise report of errors. To this end, response keys to the two tasks were changed
(“A”, “S”, “D” and “L”, “P” for one half of participants, “L”, “Ö”, “Ä” and “A”, “Q” for the
other half) and two additional keys, “alt” and “alt gr”, served to report errors. In each trial,
600 ms after the response to Task 2, an additional instruction (“FEHLER?”, engl. error?)
appeared for 1000 ms in the center of the screen. During this time, participants should
indicate if they had just committed an error in the task of their left hand (“ALT”) or their right
hand (“ALT GR”), or both. No response was required if the participants considered both their
responses correct. To account for the increased trial duration, Experiment 2 was restricted to
only the short SOA of 300 ms. With 10 blocks of 84 trials each, this resulted in 840 trials for
the condition with overlapping task execution.
Experiment 3. Experiment 3 was a replication of Experiment 2 with the following two
changes. First, a third response possibility was included in the error report. By pressing the
“SPACE” key with one of their thumbs, participants should also indicate when no error had
occurred. This was added to reduce the effect of the frontal Readiness Potential that we
observed in Experiment 2 after the response only on error trials because only there, a key
press would soon have to be made to report the error (Brunia, van Boxtel, & Böcker, 2012).
500 ms after the error report, a second instruction (“FRÜH ODER SPÄT?”, engl. early or
late?) required participants to decide whether they had a subjective feeling of “early” or “late”
conscious error awareness. Following the basic rationale of Di Gregorio et al. (2020),
participants were told in the initial instruction of the experiment that some errors are detected
already at or before the time of the response button press, whereas others are detected later.
Participants hence should distinguish between such early and late detected errors. Responses
to this second question were given with the same keys as for the first one (left hand: “early”;
right hand: “late”; space key: “I don’t know”).
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 32
Data acquisition
The EEG was recorded from 64 electrodes at a sample rate of 512 Hz with a BioSemi
Active-Two system (BioSemi, Amsterdam, The Netherlands; channels Fp1, AF7, AF3, F1,
F3, F5, F7, FT7, FC5, FC3, FC1, C1, C3, C5, T7, TP7, CP5, CP3, CP1, P1, P3, P5, P7, P9,
PO7, PO3, O1, Iz, Oz, POz, Pz, CPz, Fpz, Fp2, AF8, AF4, AFz, Fz, F2, F4, F6, F8, FT8,
FC6, FC4, FC2, FCz, Cz, C2, C4, C6, T8, TP8, CP6, CP4, CP2, P2, P4, P6, P8, P10, PO8,
PO4, O2 as well as the left and right mastoid). The CMS (Common Mode Sense) and DRL
(Driven Right Leg) electrodes were used as reference and ground electrodes. Vertical and
horizontal electrooculogram (EOG) was recorded from electrodes above and below the right
eye and on the outer canthi of both eyes. All electrodes were off-line re-referenced to
averaged mastoids.
Experimental design and statistical analysis
Design. Each trial was assigned to a condition based on the SOA (short, long) and T1
Correctness (correct, error). T1 Correctness relied on the post-hoc classification of whether
the response in the flanker task (T1) was correct or not. Trials on which an error in the pitch
discrimination task (T2) occurred were removed from all analyses of response times (RTs)
and ERP data.
Data analysis. For the analysis of RTs, trials were excluded whose RT deviated more
than three standard deviations from the RT mean of each condition and participant. Error rates
were arcsine-transformed prior to statistical testing (Winer, Brown, & Michels, 1991). All
analyses were performed using custom MATLAB v8.2 (The Mathworks, Natick, MA) scripts
and EEGLAB v12.0 (Delorme & Makeig, 2004) functions. All data and analysis scripts are
publicly available in an online repository hosted by the Open Science Framework
(https://osf.io/5ub8z/). For Experiment 1, continuous EEG data was initially band-pass
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 33
filtered to exclude frequencies below 0.1 Hz and above 40 Hz. Then, epochs were created for
two separate analyses, first from -500 ms before to 1000 ms after the T1-response (T1-
response-locked dataset) and, second, from -500 ms before to 1000 ms after the T2-response
(T2-response-locked dataset). Epochs in both analyses were baseline-corrected by subtracting
mean activity between -150 ms and -50 ms before the response, as neural correlates of
performance monitoring were previously found to emerge slightly before the response button
press. Separately for each dataset, electrodes were interpolated using spherical spline
interpolation if they met the joint probability criterion (threshold 5) as well as the kurtosis
criterion (threshold 5) in EEGLAB’s channel rejection routine (pop_rejchan.m). Epochs were
removed that contained activity exceeding +/-300 μV in any channel except AF1, Fp1, Fpz,
Fp2, AF8 (to prevent exclusion of blink artifacts, which were corrected at a later stage) and
whose joint probability deviated more than 5 standard deviations from the epoch mean. To
correct for eye blinks and muscular artefacts, an infomax-based ICA (Bell & Sejnowski,
1995) was computed and components with time courses and topographies typical for these
artefacts were removed after visual inspection.
In both experiments and for each dataset (T1-response-locked and T2-response-
locked), following a considerable number of previous studies that feature a Pe peaking over
parieto-occipital electrodes (Beatty et al., 2018; Endrass, Reuter, & Kathmann, 2007; Shalgi,
Barkan, & Deouell, 2009; M. Steinhauser & Yeung, 2010, 2012), the Pe was quantified by
comparing mean amplitudes from 200 ms to 400 ms after the respective response at electrode
POz between T1 error trials and correct trials. The Ne/ERN was quantified by comparing
mean amplitudes from 0 ms to 100 ms after the respective response at electrode FCz between
T1-error trials and correct trials. In Experiment 2, these analyses were restricted to T1-error
trials that were reported as T1 errors, and correct trials that were reported as correct trials.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 34
As we found that T1 errors elicited both an immediate Pe in T1-response-locked data
and a deferred Pe in T2-response-locked data, we aimed to investigate the relationship
between both types of Pe on a single-trial level. Our approach was to categorize trials
according to the size of the immediate Pe and then analyze differences between these trials
with respect to the deferred Pe. To acquire a robust single-trial estimate of the T1-response-
locked Pe, we used a multivariate pattern analysis (MVPA) based on the linear integration
method introduced by Parra et al. (2002, 2005). This method has previously been used to
quantify error-related brain activity on a single-trial level (Boldt & Yeung, 2015; M.
Steinhauser & Yeung, 2010). Here, we provide only a brief description of this method while
details can be found elsewhere (e.g., Steinhauser & Yeung 2010). In a first step, we computed
a set of classifiers on T1-response-locked data that discriminated optimally between correct
trials and T1 errors. Classifiers were constructed for consecutive, partially overlapping time
windows from 0 ms to 400 ms after the T1-response (width 50 ms, step size 10 ms). All
classifiers were trained, separately for each participant, on T1 error trials and the same
number of randomly drawn correct trials. Then, the classifier was selected that featured the
highest discrimination sensitivity, as indicated by the Az score. To prevent overfitting, Az
was computed using leave-one-out cross-validation. Using this classifier, we calculated
prediction values for each error trial. These prediction values represent single-trial estimates
of error-related brain activity in the respective classifier window. Based on these estimates,
we assigned each error trial to one of three equally sized bins: small Pe, medium Pe, and large
Pe in T1-response-locked data. By analyzing Pe amplitudes in T2-response-locked data
according to these bins, we were able to investigate the relationship between immediate and
deferred Pe elicited by T1 errors.
As a second way of investigating a possible trade-off between the R1-locked Pe and
the R2-locked Pe, we computed an analysis based on linear regression on raw data. To get a
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 35
sufficiently robust measure of the Pe in single trials, we computed mean amplitudes of trial-
wise EEG data in a 100 ms window on a cluster of posterior electrodes (Pz, P1, P2, POz, P03,
P04). This was done in steps of 10 ms on time windows centering from 0 ms to 400 ms after
both responses. For each such combination of single-trial Pe values in R1-locked data and R2-
locked data, the following linear regression model was fitted on the data of every participant:
= + +
where EEGR2 represents z-scored posterior response 2-locked EEG activity, EEGR1 represents
z-scored posterior response 1-locked EEG activity, IRI represents the z-scored inter-response
interval of the current trial. All z-scores were computed within-subject (i.e., using subject-
specific across-trial mean and standard deviations) and βi represent within-subject regression
coefficients. Participants’ regression coefficients were subsequently standardized by their SDs
and tested against zero by means of Student’s t-tests. Due to the large number of the resulting
t-tests, we corrected for multiple comparisons by means of a cluster-based permutation test
with 100.000 permutations, a cluster inclusion threshold of p = .01 and an output threshold of
p = .05, utilizing the Mass Univariate ERP Toolbox (Groppe, Urbach, & Kutas, 2011).
Acknowledgements
This work was supported by a grant within the Priority Program, SPP 1772 from the German
Research Foundation (Deutsche Forschungsgemeinschaft, DFG, grant number STE 1708/4-
1). The open access publication of this article was supported by the Open Access Fund of the
Catholic University Eichstätt-Ingolstadt.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 36
References
Aarts, K., Houwer, J. De, & Pourtois, G. (2013). Erroneous and correct actions have a
different affective valence: Evidence from ERPs. Emotion, 13, 960–973.
Beatty, P. J., Buzzell, G. A., Roberts, D. M., & McDonald, C. G. (2018). Speeded response
errors and the error-related negativity modulate early sensory processing. NeuroImage,
183, 112–120.
Bell, A. J., & Sejnowski, T. J. (1995). An information-maximization approach to blind
separation and blind deconvolution. Neural Computation, 7, 1129–1159.
Boldt, A., & Yeung, N. (2015). Shared neural markers of decision confidence and error
detection. Journal of Neuroscience, 35, 3478–3484.
Brunia, C. H. M., van Boxtel, G. J. M., & Böcker, K. B. E. (2012). Negative slow waves as
indices of anticipation: the Bereitschaftspotential, the Contingent Negative Variation,
and the Stimulus-Preceding Negativity. The Oxford Handbook of Event-Related
Potential Components.
Buzzell, G. A., Beatty, P., Paquette, N., Roberts, D., & McDonald, C. (2017). Error-induced
blindness: Error detection leads to impaired sensory processing and lower accuracy at
short response-stimulus intervals. Journal of Neuroscience.
Cousineau, D. (2005). Confidence intervals in within-subject designs: a simpler solution to
Loftus and Masson’s method. Tutorials in Quantitative Methods for Psychology, 1, 42–
45.
Dehaene, S., Kerszberg, M., & Changeux, J.-P. (1998). A neuronal model of a global
workspace in effortful cognitive tasks. Proceedings of the National Academy of Sciences,
95, 14529–14534.
Del Cul, A., Baillet, S., & Dehaene, S. (2007). Brain dynamics underlying the nonlinear
threshold for access to consciousness. PLoS Biology, 5, 2408–2423.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 37
Dell’Acqua, R., Jolicoeur, P., Vespignani, F., & Toffanin, P. (2005). Central processing
overlap modulates P3 latency. Experimental Brain Research, 165, 54–68.
Delorme, A., & Makeig, S. (2004). EEGLAB: an open source toolbox for analysis of single-
trial EEG dynamics including independent component analysis. Journal of Neuroscience
Methods, 134, 9–21.
Di Gregorio, F., Maier, M. E., & Steinhauser, M. (2018). Errors can elicit an error positivity
in the absence of an error negativity: Evidence for independent systems of human error
monitoring. NeuroImage.
Di Gregorio, F., Maier, M. E., & Steinhauser, M. (2020). Are errors detected before they
occur? Early error sensations revealed by metacognitive judgments on the timing of error
awareness. Consciousness and Cognition, 77, 102857.
Di Gregorio, F., Steinhauser, M., & Maier, M. E. (2016). Error-related brain activity and error
awareness in an error classification paradigm. NeuroImage, 139, 202–210.
Endrass, T., Reuter, B., & Kathmann, N. (2007). ERP correlates of conscious error
recognition: Aware and unaware errors in an antisaccade task. European Journal of
Neuroscience, 26, 1714–1720.
Falkenstein, M., Hohnsbein, J., Hoormann, J., & Blanke, L. (1991). Effects of crossmodal
divided attention on late ERP components. II. Error processing in choice reaction tasks.
Electroencephalography and Clinical Neurophysiology, 78, 447–455.
Fu, W. T., & Anderson, J. R. (2008). Solving the credit assignment problem: Explicit and
implicit learning of action sequences with probabilistic outcomes. Psychological
Research, 72, 321–330.
Gehring, W. J., Goss, B., & Coles, M. G. H. (1993). A neural system for error detection and
compensation. Psychological Science, 4, 385–390.
Groppe, D. M., Urbach, T. P., & Kutas, M. (2011). Mass Univariate Analysis of Event-
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 38
Related Brain Potentials/ Fields II: Simulation Studies. Psychophysiology, 48, 1726–
1737.
Gross, J., Schmitz, F., Schnitzler, I., Kessler, K., Shapiro, K., Hommel, B., & Schnitzler, A.
(2004). Modulation of long-range neural synchrony reflects temporal limitations of
visual attention in humans. Proceedings of the National Academy of Sciences, 101,
13050–13055. Retrieved from papers3://publication/uuid/D192C02A-B83A-4F02-
AEA5-BC0E51C10D82
Hesselmann, G., Flandin, G., & Dehaene, S. (2011). NeuroImage Probing the cortical
network underlying the psychological refractory period : A combined EEG – fMRI
study. NeuroImage, 56, 1608–1621.
Hester, R., Foxe, J. J., Molholm, S., Shpaner, M., & Garavan, H. (2005). Neural mechanisms
involved in error processing: A comparison of errors made with and without awareness.
NeuroImage, 27, 602–608.
Hochman, E. Y., & Meiran, N. (2005). Central interference in error processing. Memory &
Cognition, 33, 635–643.
Holroyd, C. B., & Coles, M. G. H. (2002). The neural basis of human error processing:
reinforcement learning, dopamine, and the error-related negativity. Psychological
Review, 109, 679–709.
Hommel, B. (1998). Automatic Stimulus-Response Translation in Dual-Task Performance.
Journal of Experimental Psychology: Human Perception and Performance, 24, 1368–
1384.
Jentzsch, I., & Dudschig, C. (2009). Why do we slow down after an error? Mechanisms
underlying the effects of posterror slowing. The Quarterly Journal of Experimental
Psychology, 62, 209–218.
Jersild, A. T. (1927). Mental set and shift. Archives of Psychology, 14, 81.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 39
Klawohn, J., Endrass, T., Preuss, J., Riesel, A., & Kathmann, N. (2016). Modulation of
hyperactive error signals in obsessive–compulsive disorder by dual-task demands.
Journal of Abnormal Psychology, 125, 292–298.
Leuthold, H., & Sommer, W. (1999). ERP correlates of error processing in spatial S-R
compatibility tasks. Clinical Neurophysiology, 110, 342–357.
Maier, M. E., Di Gregorio, F., Muricchio, T., & Di Pellegrino, G. (2015a). Impaired rapid
error monitoring but intact error signaling following rostral anterior cingulate cortex
lesions in humans. Frontiers in Human Neuroscience, 9, 1–15.
Maier, M. E., Di Gregorio, F., Muricchio, T., & Di Pellegrino, G. (2015b). Impaired rapid
error monitoring but intact error signaling following rostral anterior cingulate cortex
lesions in humans. Frontiers in Human Neuroscience, 9, 339.
Maier, M. E., Scarpazza, C., Starita, F., Filogamo, R., & Ladavas, E. (2016). Error monitoring
is related to processing internal affective states. Cognitive, Affective and Behavioral
Neuroscience, 16, 1050–1062.
Maier, M. E., & Steinhauser, M. (2017). Working memory load impairs the evaluation of
behavioral errors in the medial frontal cortex. Psychophysiology, 54, 1472–1482.
Marti, S., King, J.-R., & Dehaene, S. (2015). Time-Resolved Decoding of Two Processing
Chains during Dual-Task Interference. Neuron.
Meyer, D. E., & Kieras, D. E. (1997). A computational theory of executive cognitive
processes and multiple-task performance: Part 1. Basic mechanisms. Psychological
Review, 104, 3.
Morey, R. D. (2008). Confidence Intervals from Normalized Data: A correction to Cousineau
(2005). Tutorials in Quantitative Methods for Psychology, 4, 61–64.
Moser, J. S., Moran, T. P., Schroder, H. S., Donnellan, M. B., & Yeung, N. (2013). On the
relationship between anxiety and error monitoring: A meta-analysis and conceptual
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 40
framework. Frontiers in Human Neuroscience, 7, 466.
Nieuwenhuis, S., Ridderinkhof, K. R., Blom, J., Band, G. P. H., & Kok, A. (2001). Error-
related brain potentials are differentially related to awareness of response errors:
evidence from an antisaccade task. Psychophysiology, 38, 752–760.
Olvet, D. M., & Hajcak, G. (2009). The stability of error-related brain activity with increasing
trials. Psychophysiology, 46, 957–961.
Overbeek, T. J. M., Nieuwenhuis, S., & Ridderinkhof, K. R. (2005). Dissociable components
of error processing: On the functional significance of the Pe vis-à-vis the ERN/Ne.
Journal of Psychophysiology, 19, 319–329.
Parra, L. C., Alvino, C., Tang, A., Pearlmutter, B., Yeung, N., Osman, A., & Sajda, P. (2002).
Linear Spatial Integration for Single-Trial Detection in Encephalography. NeuroImage,
17, 223–230.
Parra, L. C., Spence, C. D., Gerson, A. D., & Sajda, P. (2005). Recipes for the linear analysis
of EEG. NeuroImage, 28, 326–341.
Pashler, H. (1994). Dual-task interference in simple tasks: data and theory. Psychological
Bulletin, 116, 220–244.
Pashler, H., & Johnston, J. C. (1989). Chronometric evidence for central postponement in
temporally overlapping tasks. The Quarterly Journal of Experimental Psychology
Section A, 41, 19–45.
Rabbitt, P. (2002). Consciousness is slower than you think Consciousness is slower than you
think. The Quaterly Journal of Experimental Psychology: Section A, 55, 1081–1092.
Sergent, C., Baillet, S., & Dehaene, S. (2005). Timing of the brain events underlying access to
consciousness during the attentional blink. Nature Neuroscience, 8, 1391–1400.
Sergent, C., & Dehaene, S. (2004). Neural processes underlying conscious perception:
Experimental findings and a global neuronal workspace framework. Journal of
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in dual-tasking 41
Physiology Paris, 98, 374–384.
Shalgi, S., Barkan, I., & Deouell, L. Y. (2009). On the positive side of error processing:
Error-awareness positivity revisited. European Journal of Neuroscience, 29, 1522–1532.
Sigman, M., & Dehaene, S. (2008). Brain mechanisms of serial and parallel processing during
dual-task performance. The Journal of Neuroscience : The Official Journal of the Society
for Neuroscience, 28, 7585–7598.
Steinhauser, M., Ernst, B., & Ibald, K. W. (2017). Isolating component processes of posterror
slowing with the psychological refractory period paradigm. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 43, 653–659.
Steinhauser, M., & Yeung, N. Decision processes in human performance monitoring, 30
Journal of Neuroscience § (2010).
Steinhauser, M., & Yeung, N. (2012). Error awareness as evidence accumulation: effects of
speed-accuracy trade-off on error signaling. Frontiers in Human Neuroscience, 6, 1–12.
Steinhauser, R., & Steinhauser, M. (2018). Preparatory brain activity in dual-tasking.
Neuropsychologia, 114, 32–40.
Sutton, R. S., & Barto, a G. (1998). Reinforcement learning: an introduction. IEEE
Transactions on Neural Networks / a Publication of the IEEE Neural Networks Council,
9, 1054.
Telford, C. W. (1931). The refractory phase of voluntary and associative responses. Journal
of Experimental Psychology, 14, 1–36.
Tombu, M., & Jolicœur, P. (2003). A central capacity sharing model of dual-task
performance. Journal of Experimental Psychology: Human Perception and
Performance, 29, 3–18.
Ullsperger, M., Danielmeier, C., & Jocham, G. (2014). NEUROPHYSIOLOGY OF
PERFORMANCE, 35–79.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
Error monitoring in multitasking 42
Ullsperger, M., Fischer, A. G., Nigbur, R., & Endrass, T. (2014). Neural mechanisms and
temporal dynamics of performance monitoring. Trends in Cognitive Sciences, 18, 259–
267.
Walsh, M. M., & Anderson, J. R. (2011). Modulation of the feedback-related negativity by
instruction and experience. Proceedings of the National Academy of Sciences of the
United States of America, 108, 19048–19053.
Weißbecker-Klaus, X., Ullsperger, P., Freude, G., & Schapkin, S. A. (2016). Impaired Error
Processing and Semantic Processing During Multitasking. Journal of Psychophysiology,
1–12.
Welford, A. T. (1952). The “psychological refractory period” and the timing of high-speed
performance - a review and a theory. The British Journal of Psychology General Section,
43, 2–19.
Wessel, J. R. (2012). Error awareness and the error-related negativity: evaluating the first
decade of evidence, 1–16.
Wessel, J. R., Danielmeier, C., & Ullsperger, M. (2011). Error awareness revisited:
accumulation of multimodal evidence from central and autonomic nervous systems.
Journal of Cognitive Neuroscience, 23, 3021–3036.
Winer, B. J., Brown, D. R., & Michels, K. M. (1991). Statistical Principles in Experimental
Design (3rd ed.). McGraw-Hill.
Yeung, N., Botvinick, M. M., & Cohen, J. D. (2004). The Neural Basis of Error Detection:
Conflict Monitoring and The Error Related Nagativity. Psychological Review, 11, 931–
959.
.CC-BY 4.0 International licensethe author/funder. It is made available under a
The copyright holder for this preprint (which was not peer-reviewed) is. https://doi.org/10.1101/2020.04.15.043059doi: bioRxiv preprint
