Content uploaded by Arnaud Delorme
Author content
All content in this area was uploaded by Arnaud Delorme on Jun 22, 2020
Content may be subject to copyright.
The role of midfrontal theta oscillations across the development of
cognitive control in preschoolers and school-age children
Nicolas Adam1,2*, Agnès Blaye2,3, Rasa Gulbinaite 4,5, Arnaud Delorme1,2,6 &, Chloé Farrer1,2
1 Université de Toulouse, Centre de recherche Cerveau et Cognition, Toulouse, France
2 Centre National de la Recherche Scientifique, France
3 Université Aix-Marseille, Laboratoire de Psychologie Cognitive, Marseille, France
4 Université de Lyon, Centre de Recherche en Neurosciences, Lyon, France
5 Institut National de la Santé et de la Recherche Médicale U1028, Lyon, France
6 Swartz Center for Computational Neuroscience, University of California, San Diego, USA
* Corresponding author: nicolas.adam@cnrs.fr ; Orcid ID: 0000-0001-6193-1472
!!Citation of the published version of this manuscript :
Adam, N, Blaye, A, Gulbinaite, R, Delorme, A, Farrer, C. The role of midfrontal theta oscillations across the
development of cognitive control in preschoolers and school!age children. Dev
Sci. 2020; 00:e12936. https://doi.org/10.1111/desc.12936
Highlights
•!Through measures of mid-frontal theta oscillations in children, this study tests for a common mechanism
involved in different cognitive control processes across development.
•!Mid-frontal theta power was increased during the resolution of task-set and response interferences in
both preschool and school age children.
•!Mid-frontal theta is involved in the development of different cognitive control processes and
independently of the way cognitive control is structured.
•!Mid-frontal theta latency was shorter in school children than in preschool children.
Abstract
The development of cognitive control enables children to better resist acting based on distracting information
that interferes with the current action. Cognitive control improvement serves different functions that differ in
part by the type of interference to resolve. Indeed, resisting to interference at the task-set level or at the response-
preparation level is respectively associated with cognitive flexibility and inhibition. It is, however, unknown
whether the same neural mechanism underlies these two functions across development. Studies in adults have
revealed the contribution of mid-frontal theta (MFT) oscillations in interference resolution. This study
investigated whether MFT is involved in the resolution of different types of interference in two age groups
identified as corresponding to different latent structures of executive functions. Preschool (4-6 years) and school
children (6-8 years) were tested with a task involving interference at the response level and/or the task-set level
while EEG was recorded. Behaviorally, response time and accuracy were affected by task-set. Both age groups
were less accurate when the interference occurred at the task-set level and only the younger group showed
decreased accuracy when interference was presented at the response-preparation level. Furthermore, MFT power
was increased, relative to the baseline, during the resolution of both types of interference and in both age groups.
These findings suggest that MFT is involved in immature cognitive control (i.e, preschool and school-ages), by
orchestrating its different cognitive processes, irrespective of the interference to resolve and of the level of
cognitive control development (i.e. the degree of differentiation of executive functions).
Keywords: Cognitive control, early development, Theta, conflict, interference, EEG
1!Introduction
One important hallmark of cognitive development in children is the increased ability to regulate behavior
voluntarily, allowing children to better adapt their behavior to changing environmental demands for achieving
a goal (Miller & Cohen, 2001). Cognitive control develops until adolescence and relies on the progressive and
asynchronous development of cognitive processes (e.g. inhibition of prepotent response and active maintenance
of goal). For example, during development, there is an increased capacity in detecting and resolving a conflict
between task-relevant and task-irrelevant information (Zelazo & Jacques, 1997). Managing conflicts requires
cognitive control for voluntarily inhibiting the processing of information that is irrelevant to an ongoing action
and concomitantly enhancing the processing of information that is relevant for achieving the goal. In a classical
conflict task, the Simon task, the stimulus is characterized by a task-relevant dimension (i.e. color) and a task-
irrelevant dimension (i.e. spatial location) (Simon & Berbaum, 1988). The conflict arises between the irrelevant
stimulus dimension and a response dimension (handside of the correct response key) (Kornblum, Hasbroucq, &
Osman, 1990). This task relies on the prepotent tendency to press the key on the same side as the stimulus.
Therefore, compatible trials in which the irrelevant spatial location of the stimulus is congruent with the location
of the correct response key are typically associated with shorter responses times than incompatible ones.
However, in the incompatible trials, the two dimensions elicit a response interference, resulting in a conflict that
must be detected and resolved for successful task performance. Resolving this conflict typically comes at a
behavioral cost, termed the Simon effect, reflected by longer response times (RT) and lower accuracy scores
compared to the compatible trials (O’leary & Barber, 1994). The size of the Simon effect has been shown to
decrease with age, across childhood (Cao et al., 2013; Davidson, Amso, Anderson, & Diamond, 2006;
Ridderinkhof & van der Molen, 1995), reflecting improvements in inhibitory control.
The development of cognitive control also promotes behavioral flexibility, which allows switching between
different tasks (Miyake et al., 2000; St Clair-Thompson & Gathercole, 2006). Although switching and inhibition
are well-differentiated functions, they also share common processes like active maintenance of a goal and
inhibition of automatic response. Additionally, flexibility also requires to resist an interference that occurs at the
task-set level. In the task-switching paradigm, the task-set interference is due to two task-sets that are associated
with different stimulus-response mappings. Cognitive control is engaged for resolving the interference by
inhibiting the previous task-set and responding to the current one accordingly. This engagement entails
behavioral costs, termed “local switch costs” reflected by increased RT and lower accuracy scores on the switch
trials compared to the non-switch trials (Meiran, 1996; Monsell, 2003). These task switching costs particularly
decrease during the preschool and school ages, (Blaye & Chevalier, 2011; Chevalier & Blaye, 2008; Crone,
Bunge, Van Der Molen, & Richard Ridderinkhof, 2006; Davidson et al., 2006; Kray, Gaspard, Karbach, &
Blaye, 2013) and children reach adult-like performance in between the ages of 9 and 15 years (Bunge & Wright,
2007; Crone et al., 2006; Huizinga & van der Molen, 2011; Wendelken, Munakata, Baym, Souza, & Bunge,
2012).
Cognitive control drastically develops during the preschool years, and to a lesser extent during the school years,
resulting in improved inhibition and cognitive flexibility (Best, Miller, & Jones, 2009; Calderon, Jambaqué,
Bonnet, & Angeard, 2014; Chevalier, Blaye, Dufau, & Lucenet, 2010; Diamond, 2013; Lorsbach & Reimer,
2008). Furthermore, there is a change in the latent structure of cognitive control in this age range with an increase
in the differentiation of executive functions (Lee, Bull, & Ho, 2013; Monette, Bigras, & Lafrenière, 2015; Senn,
Espy, & Kaufmann, 2004; van der Ven, Kroesbergen, Boom, & Leseman, 2013; S. A. Wiebe, Espy, & Charak,
2008; S. a. Wiebe et al., 2011). Whereas a one-factor structure best accounts for 3-year-olds’ control, inhibition
and flexibility functions are differentiated from around 6-8 years old (Huizinga, Dolan, & van der Molen, 2006;
Lehto et al., 2003). Improved efficiency of these functions are associated with neuroanatomical and neuro-
functional changes in their respective fronto-parietal circuits (Bell & Wolfe, 2007; Casey et al., 2004; Ezekiel,
Bosma, & Morton, 2013), which follow asynchronous maturation (Gogtay et al., 2004). Yet, the neural
mechanisms that support cognitive control across development are still poorly known. In particular, it is not
known (a) whether a common mechanism is involved in flexibility and inhibition and (b) whether this
mechanism depends on the degree of differentiation of these executive functions.
Neural oscillatory activity within the 4-8 Hz theta frequency range has been proposed to orchestrate cognitive
processes within the neural circuits supporting cognitive control (Cohen, 2014; Nigbur, Ivanova, & Stürmer,
2011). The implementation of cognitive control requires the formation of functional networks in theta power
(reflecting the activation of neurons in this frequency range), which is generated within the mid-frontal areas
(Cavanagh, Cohen, & Allen, 2009; Cohen & Cavanagh, 2011; Nigbur, Ivanova, & Stürmer, 2011) and allows
the coordination of the neural computations underlying cognitive control within this network (Buzsaki &
Draguhn, 2004; Doesburg, Vidal, & Taylor, 2013; Von Stein & Sarnthein, 2000). Theta power in mid-frontal
areas is increased in conflict situations that require cognitive control for detecting and signaling the conflict to
other brain systems involved in the selection of an action (Cohen & Ridderinkhof, 2013; Cohen & Donner, 2013;
Gulbinaite, van Rijn, & Cohen, 2014; Nigbur et al., 2011). Furthermore, mid-frontal theta power enhancement
was observed for different types of cognitive conflicts in adults (Simon, Flanker, Stroop, Go-NoGo effects),
suggesting that theta might reflect the engagement of cognitive control in various conflict tasks that necessitate
important cognitive control demands (Jensen & Tesche, 2002; Nigbur et al., 2011; Richardson, Anderson, Reid,
& Fox, 2018; Sheridan, Kharitonova, Martin, Chatterjee, & Gabrieli, 2014). Theta oscillations are also involved
in task switching as indicated by increased theta power when the task-set changed (Cunillera et al., 2012; van
de Vijver, Richard Ridderinkhof, & Cohen, 2011; Womelsdorf, Johnston, Vinck, & Everling, 2010). Mid-frontal
theta power enhancement is considered as reflecting the engagement of cognitive control to resolve the
interference between the previous and the current task-sets. Overall, these findings suggest that mid-frontal theta
power might reflect the engagement of cognitive control in various conflict tasks and at different levels of
conflict (task-set and response-preparation).
Theta oscillations has been reported to serve different top-down cognitive functions during childhood. For
example, as early as 5 months of age, efficient gaze direction control and speech perception are reported to rely
on theta oscillatory activity (Bosseler et al., 2013; Michel et al., 2015). Theta oscillations were also found to be
associated with cognitive control (Morasch & Bell, 2011). Widespread increase in theta power relative to the
baseline were observed in tasks targeting different executive processes across different age groups: in infants
(Bell, 2002; E. V. Orekhova, Stroganova, Posikera, & Elam, 2006; Elena V. Orekhova, Stroganova, & Posikera,
1999), toddlers (Conejero, Guerra, Abundis-Gutiérrez, & Rueda, 2018), as well as kindergarteners (Bell &
Wolfe, 2007; Watson & Bell, 2013; Wolfe & Bell, 2004). Recently, Meyer and colleagues (Meyer, Endedijk,
van Ede, & Hunnius, 2019) reported that fronto-medial theta power was enhanced in preschool children when
they engaged in a cognitive task.
Based on the evidence of an involvement of theta oscillations in different executive functions in adults and of
its role in cognition across development, we hypothesized that theta oscillations might play a general role in
cognitive control across development, being involved in the implementation of different cognitive control
processes (i.e., in the resolution of distinct types of interferences) and at different levels of development of
cognitive control (i.e., at distinct developmental periods that differ in the degree of differentiation of the
executive functions). We therefore tested preschool (4 to 6 years) and school children (6 to 8 years) with an
interference task adapted from the “heart and flower” task (Davidson et al., 2006a). In this task, switch and non-
switch trials were randomly mixed with compatible and non-compatible trials, requiring the engagement of
cognitive control for resolving the interference at the response level (in the incompatible trials) and/or the task-
set level (in the switch trials). We predicted that mid-frontal theta power would increase during the resolution
of an interference at the task-set level as well as at the response level compared to a baseline condition. We also
predicted that this increase would be observed in both preschool and school children. Furthermore, because
executive functions' performance importantly improves between the preschool and school ages, we also assessed
whether MFT activity differed between preschool and school children within the interference conditions.
2!Material and methods
2.1! Participants
A total of 114 children aged 4 to 8 years were recruited from two Kindergarten French schools (N = 71) and one
Elementary French school (N=43). Sixteen participants were removed from the analyses because of very poor
performance on the task (error rate > 45%; N = 8), technical problems during EEG recording (N = 7), or because
they stopped the task before the end and did not allow the collection of enough data (N = 1). The attrition mainly
concerned preschool children. Behavioral analyses were finally conducted on 98 children, 57 preschool children
(mean age = 5 years ±6 months, 21 boys) and 41 school children (mean age= 7 years, ±10 months, 17 boys).
Children were tested individually by an experimenter in a quiet room at their school. The testing session lasted
approximatively an hour, with 35 minutes dedicated to the EEG set up. The first period of familiarization
allowed children to get accustomed to the materials and the task. They were first provided with some brief
information about the EEG material and its use, the experimenter then explained the task to the children and set-
up the EEG cap on the children’s head.
The study was approved by the local ethics committee (N° 2-15038) and was conducted in accordance with the
Declaration of Helsinki. Informed consent was obtained from parents and children gave verbal approval before
starting the experiment. Children were rewarded with a small gift and a ‘diploma’ at the end of the testing.
2.2!Apparatus and stimuli
Children were comfortably seated at a viewing distance of 0.4 m from a Mac Laptop (13.3-in, screen refresh
rate: 60Hz). The stimuli were presented with the OpenSesame software (Mathôt, Schreij, & Theeuwes, 2012).
Children were presented with a central black fixation cross (7 mm diameter, 1°) on a white background, with
either a red heart or a red flower (30 mm diameter, 4.3°) appearing on the left or the right (20 mm, 2.87°
eccentricity) of the fixation cross. Children were instructed to respond as fast as possible when a stimulus
appeared, using a button box with two buttons (for left and right index fingers) whose pressure was adapted to
children’s fingers’ strength. The stimulus remained on the screen until a response was made or a maximal RT
limitation was exceeded, the duration of which differed between school and preschool children. Time limitation
was dependent upon the age group to ensure a similar level of difficulty across children. Following Davidson
and colleagues (2006), we chose a 2500 ms time response window after the stimulus onset for the preschool
children (younger than 6.5 years old). For school children (older than 6.5 years old), we followed the
recommendations of previous researchers and extended the original time window (750ms), resulting in a time
window of 1500ms (Calderon et al., 2014; Obradović, Sulik, Finch, & Tirado-Strayer, 2018). Following a
response, children received visual and auditory feedback with a low-pitched tone of 500ms accompanied by a
surprised face for an incorrect response (same feedback was used in case of no-response trials), and a high-
pitched tone of 500ms accompanied by a happy face for a correct response.
2.3!Procedure
The task was adapted from the 'heart and flower' task of Davidson et al., (2006). In the original task, compatible
and incompatible trials were presented either in separate blocks or mixed. In the present study, we used only the
mixed-trial condition that presents randomly compatible and incompatible trials during the same block. The
choice of using only the mixed-trial condition of the original task was justified by the main objective of this
study: evaluate the role of mid-frontal theta in different cognitive control processes across development. Based
on previous studies, maximal conflict effects and age differences were observed only in the mixed-trial condition
(Calderon et al., 2014; Davidson et al., 2006; Diamond & Ling, 2016). Finally, the constraints of EEG recording
and analysis with children were also taken into account (Bell & Cuevas, 2012). EEG analysis requires a
sufficient number of trials per condition, thus we limited the experimental design to the mixed condition to keep
recording sessions relatively short, in order to minimize children’s fatigue and keep them concentrated and
engaged in the task.
If the “heart” stimulus was presented on the screen, children were instructed to press the response button that
was compatible with the location of the stimulus on the screen (i.e. press left if the heart appeared on the left).
If the “flower” stimulus appeared on the screen, children were instructed to press the opposite-side button from
that of the stimulus location (i.e. press right if the flower appeared on the left). Because the “flower” trials
involved a conflict between the location of the response and the location of the stimulus, these trials were
considered as incompatible trials, while the heart trials were considered as compatible trials. Compatible and
incompatible trials were randomly presented throughout the experiment, resulting in sequences of trials where
the rule was the same than the preceding trial (non-switch trials) and trials in which the rule changed (switch
trials).
Children first performed three training blocks to become familiarized with the task. These training blocks
consisted of compatible only, incompatible only and mixed-trial blocks with the same order of presentation for
all children. For each training block, children received and completed 4 trials without time limitation. During
practice and also during experimental trials, children received positive or negative visual and auditory feedback
after each response. If they understood the task well, they then performed a practice block of 12 trials with time
limitation. Training practice could be carried out up to three times if the task had not been fully understood. If
the child demonstrated at least 75% of correct responses during the training session and if he was able to explain
the rules to the experimenter, the experimental session was then started (Diamond, Steven Barnett, Thomas, &
Munro, 2007).
The experimental session was composed of 8 blocks of 16 trials each (132 trials in total), with the same number
of compatible and incompatible trials randomly ordered with an inter-trial interval that randomly varied between
400 and 1600ms. Rules were reminded at the beginning of each block, systematically marked by a short
interruption. Longer breaks were proposed to the children if needed (experimental paradigm is schematized in
figure 1).
2.4!EEG recording and preprocessing
EEG was recorded for 87 children (52 preschool children and 35 school children) in a quiet room of the schools.
Stimuli were displayed on a Macintosh laptop (MacBook Pro, 13 inches, 2.9 GHz Intel Core i7) and EEG
monitoring and recording was conducted with another identical Macintosh laptop. We used a Biosemi Active-
Two amplifiers system with 64-channel positioned according to the 10-20 International system (American
Electro-Encephalographic Society, 1991) with two additional online references placed at occipital sites.
The acquisition software was ActiView (a LabVIEW toolbox designed by BIOSEMI). Prior to recording,
impedance was checked to be below the 20 offset threshold using the Actiview impedance indicator. All offline
data preprocessing and analyses were done using EEGLAB toolbox (v.14.1.1) for Matlab
(sccn.ucsd.edu/eeglab/) and custom-written Matlab scripts. EEG was recorded with a 2048 Hz sampling rate
and was then down sampled to 512 Hz. Data were high-pass filtered at 1Hz, notched filtered at 50 Hz (using
CleanLine EEGLAB plugin v1.03) and re-referenced to average reference. Epochs were extracted from
continuous data from -1500 to 3000 ms relatively to the stimulus onset. This large epoch’s size allows
performing time-frequency decomposition at low frequencies. For electrode-level analysis, noisy channels were
identified by visual inspection and interpolated using spherical interpolation (function eeg_interp of EEGLAB).
First, epochs containing large ocular, motor artefacts or very noisy signal (during baseline or stimulus processing
time) were visually identified and manually rejected from the dataset to optimize the subsequent component
identification process. Furthermore, we also excluded participants for whom the previous artefact rejection
procedure led to the removal of more than 40% of the total number of correct trials. Thereafter, independent
component analysis (ICA) was performed on clean data (Delorme & Makeig, 2004). Components that did not
account for brain activity (eye-blinks, horizontal eye-movements or muscle activity) were manually identified
by experimenter and subtracted from the data (Chaumon, Bishop, & Busch, 2015). Only clearly identifiable eye
blink component was removed in order to preserve the maximum of information. This corresponded to an
average of 0.9 component per participant. Only data epochs corresponding to correct responses trials were
included in the analysis, errors, post-errors, anticipatory responses, as well as warm-up trials (the first trial of
each block that could not be classified as a sequential condition) were removed from the data (1361 trials were
rejected in total: 740 for preschool children and 641 for school children). Following the experimental design of
Davidson et al., (2006), inter-trial-interval varied between 400ms and 1500ms. In order to avoid a non-neutral
baseline period due to brain activity related to the previous trial affecting time-frequency analyses, trials with
ITI < 800ms were also excluded from the EEG analyses (N = 70 trials). Children who were excluded from the
behavioral analyses were also excluded from the EEG analyses. At this step 75 children were retained in EEG
analyses.
2.5!Time-frequencies analysis
We first conducted a channel-level analysis. For this, the surface-Laplacian, using laplacian perrinX function
(Cohen, 2014), was applied to artefact-free data prior to analysis to increase the topographical specificity and
attenuate volume conduction (Kayser & Tenke, 2015). Time-frequency decomposition was performed by
convolving stimulus-locked single-trial data from all electrodes with complex Morlet wavelets which increased
from 2 to 40 Hz in 40 logarithmically spaced steps, and wavelet cycles varied from 3 to 6 in logarithmically
spaced steps.
From the resulting complex signal, an estimate of frequency band-specific power at each time point was defined
as the squared magnitude of the result of the convolution. Single-trial power values were then averaged across
trials for each condition separately, and then normalized relative to the condition-average baseline period (-400
to -100 ms) using a decibel (dB) transform at each frequency. Conversion to a dB scale ensures that data across
all frequencies, time points, electrodes, conditions, and participants were in the same scale and thus were
comparable. The use of a short baseline window was the most appropriate approach to avoid a contamination of
the previous trial activity over subsequent trials as reported in previous studies (Nigbur, Cohen, Ridderinkhof,
& Stürmer, 2012; [-350 to -150ms] ;for Nigbur et al., 2011; [-300 to -100ms] for Cavanagh et al., 2009; Cohen
& Cavanagh, 2011; Cohen & Ridderinkhof, 2013; Cohen & Donner, 2013; Gulbinaite et al., 2014 ; [-200 to -
50ms] for Cunillera et al., 2012).
2.6!Analytical procedure
To assess the engagement of theta oscillations in the resolution of different types of interference in preschool
and school children, we took into account all the interference conditions for the analyses of both children’s
behavioral and EEG data.
The task involved different interference conditions (presented in Table 1). The cC interference condition
consisted of compatible trials (C) that were preceded by a compatible trial (c). This condition corresponded to
the null-interference condition because no interference was elicited. In iI condition, incompatible trials (I) were
preceded by incompatible trials (i), here the interference is related to the conflict generated by the two
dimensions of the stimulus (the relevant and irrelevant dimensions of the stimulus), this interference was
subsequently called repeated response interference. iC condition refers to compatible trials (C) preceded by
incompatible trials (i) and the conflict was related to the change in the stimulus-response mapping between the
previous and the current trials, this interference was subsequently called task-set interference. Finally, in the cI
condition incompatible trials (I) were preceded by compatible trials (i), resulting in two types of interference,
one related to the conflict between the two dimensions of the stimulus and the other to the change in the stimulus-
response mapping. This interference was subsequently called task-set and response interference.
For each child, two different behavioral measures were obtained: RT (in ms) and accuracy (percentage of correct
responses). Incorrect and anticipatory responses (below 200ms), omitted and post-errors trials were removed
from the RT data analyses. RT greater than 3 standard deviations of the mean were considered as outliers and
were rejected. The first trial of each block was considered as a warm-up trial and was also removed prior to the
analyses. RT and accuracy were analyzed using a general linear mixed model approach allowing for
specification of random effects. We used this statistical model because it allows accounting for the trial-to-trial
variability, increasing the ability to detect the effect of fixed parameters. Moreover, the number of trials was
slightly different between conditions, and GLM accommodates an unbalanced experimental design. GLMer
analyses were conducted using the function glmer (Bates, Mächler, Bolker, & Walker, 2015) of the R package
lme4 (R Core Team, 2014). GLMer with Gamma family (with log link) was performed on RTs whereas
Binomial family (with logit link) was used for accuracy. Normality of residuals was checked via qq plot and the
histogram of the residuals, a plot of standardized residuals against predicted values, was used to check for
deviation from homoscedasticity. We included as fixed effects an Interference factor with 4 levels (cC, iI, cI,
iC), a Group factor (preschool and school children) and the interaction between group and interference factors;
participants (model the intra individual variability) and the interaction between participant and condition (model
the intra individual variability across conditions) were added as random factors. We reported the regression
coefficients relative to the reference condition (e.g. the null-interference condition in preschool children), with
their corresponding 95% confidence intervals (indicated in brackets), as well as an estimation of the effects
expressed in ms for RTs and in percentage for accuracy. T-values and their corresponding p-values were also
reported. When necessary, post hoc comparisons with fdr adjustments (correction for multiple comparisons)
were also conducted with the Welch-Satterthwaite method. These tests were conducted to compare performance
between conditions for preschool and school children independently as well as between preschool and school
children’s mean performance. P-value was computed from the Wald statistic (alias the z-value, corresponding
to the estimate of the coefficient divided by its standard error).
Our statistical analyses (performed on the entire sample of EEG data) involved two main steps.
First, to assess whether mid-frontal theta oscillatory activity was sensitive to the implementation of cognitive
control during the task, we compared children’s theta power during pre-stimulus baseline period (-400ms to -
100ms) to the period where children were engaged in the task (individual data averaged across interference
conditions). Previous studies showed that mid-frontal sites around FCz channel are active during
interference tasks that require cognitive control (in adults: Cohen & Donner, 2013; Gulbinaite & Johnson, 2014;
Nigbur et al., 2011 ; in children: Lamm, Zelazo, & Lewis, 2006; Papenberg, Hämmerer, Müller, Lindenberger,
& Li, 2013 ; for review see Crone & Steinbeis, 2017). Surrogate statistical testing (using permutation analyses)
was performed (Delorme, Westerfield, & Makeig, 2007). We considered data averaged on interference
conditions and subjects and computed the surrogate distribution at each frequency using baseline values. We,
therefore, obtained one surrogate distribution per frequency and then tested if post-stimulus power amplitude
values point lied in the 2.5 or 97.5% tail of the surrogate distribution at a given frequency. If it did, the specific
time-frequency point was considered significant at p"<"0.05 (a total of 1000 permutations at each frequency was
used to assess significance and Bonferroni correction for multiple comparison was applied). This procedure was
applied with the dataset including all the channels (surrogate distribution computed at each frequency by
permuting baseline values across time and channels) to appraise the topographical dynamics of the task-related
theta increase. These tests were performed on the whole sample and separately for each group, on the data
averaged across interference conditions and also on each interference condition.
Second, to examine changes in children’s theta activity as a function of interference condition, we compared
children’s MFT power amplitude between interference conditions (within and between age groups). This
approach belongs to the class of “guided” as opposed to “blind” (e.g. independent component analysis) source-
separation methods (Cheveigné & Arzounian, 2015; Nikulin, Nolte, & Curio, 2011). Because the data were
firstly pooled over interference conditions the window selection was orthogonal to the effects that we were
interested in (Cohen & Gulbinaite, 2017). The spatial filter was designed in a way that was blind to the
interference condition of any given trial. For each participant, a single component with the typical mid-frontal
spatial peak and theta-band temporal dynamics (phasic theta-power increase following stimulus onset) was
identified and selected for further analyses (more details about the MFT component extraction are provided in
the supplemental materials). In cases when no single MFT component could be identified, participant was
excluded from further analysis. Among the 75 participants with analyzable EEG data, 6 did not show MFT
component (five preschool children and 1 school child).
Task-related brain activity was extracted on an individual basis during the time window where cognitive control
was implemented (stimulus-locked analysis) (Cavanagh, Zambrano-Vazquez, & Allen, 2012; Gulbinaite &
Johnson, 2014). Considering age differences in our groups, we used a wide temporal window for all children.
We first identified individual theta power peak on data averaged across conditions within a large, 200ms to
1500ms, time window (an extended window compared to the window identified on the whole sample at the
channel level with the surrogate tests). Then, for each participant a shorter time window was determined around
the theta power amplitude peak in both frequency (± 1Hz) and time (± 150ms). Theta power within this specific
window was averaged for each condition to compute the individual MFT (see an illustration of the procedure in
the supplemental figure S1). These measures were normalized to perform single-trial analyses. For average
analyses the theta measure was computed by averaging baseline mean-corrected power values of mid-frontal
theta component into the participant-specific time window. Finally, 69 participants were included to the final
sample for EEG analyses (36 preschool children and 33 school children). This attrition corresponded to a loss
of approximately 21% of the participants that is quite classic in studies with children, principally the youngest
(Bell & Cuevas, 2012; Wolfe & Bell, 2004, 2007). There were no drastic differences between each condition in
the final sample of EEG trials (N = 5821 trials) with, for preschool-children between 19 and 21 trials per
condition and per children (with a total of 693 cC, 763 cI, 769 iC, 687 iI), and for school children between 19
and 23 trials per condition and per children (with a total of 682 cC, 753 cI, 737 iC, 637 iI).
Then, EEG analyses were conducted on the latency of the MFT peak (detected on the data averaged on
conditions for each participant) and the power of the MFT peak (estimated for each condition of each
participant). A GLM model (Gamma distribution) was conducted with the log-transformed latency of the peak
as the dependent variable and group factor (preschool and school children) as a fixed effect.
The normally distributed MFT power was analyzed with a lmer model with MFT power measure as the
dependent variable, the Interference factor, the Group factor as fixed effects and the inter-participant variability
as a random effect. Finally, previous results have shown that theta power predicted behavioral performances in
cognitive conflict task in both adults (Cohen & Donner, 2013; Gulbinaite & Johnson, 2014; Töllner et al., 2017)
and children (Liu, Woltering, & Lewis, 2014) and that age-related changes in theta power explained the age-
related changes in cognitive control performance (Liu et al., 2014). We therefore assessed whether variation in
children’s RTs was partially predicted by variation in MFT power. GLMer models were performed on the single
trial Log-transformed RT as dependent variable with as fixed effects the Group factor, the Interference factor,
the z-transformed single trial MFT power as well as the interactions between all these factors. Inter-participant
variability as well as intra-condition variability were added as random effects.
3!Results
3.1!Behavioral results
3.1.1!Response Times
Overall, school children were faster than preschool children (z=11.66, p<.0001), with an average RT of 878 ±
21ms and 1259 ± 25ms for school and preschool children respectively. Children’s RT also varied as a function
of the interference condition (see the supplemental materials for the results of all the models). In particular,
Preschool children were slower in all the conditions involving an interference compared with the null-
interference condition (cC) as shown by significant positive betas for all conditions relative to the reference
condition (CC condition in preschool children). Post-hoc comparisons further revealed that preschool children
were faster at resolving an interference between two response alternatives than when the interference occurred
at the task-set level (between the previous and the current stimulus-response mappings). They, however, were
equally fast in the two interference conditions involving a change in the task-set (cC < iI < cI = iC, see all the
statistics in the table 2 and the figure 2a).
School children were equally fast in the null interference condition than in the condition involving an
interference between two response alternatives. But they were faster in this interference condition than in the
conditions involving a change in the task-set. They were also faster at resolving an interference occurring at
both the task-set level and the response level than when the interference only occurred at the task-set level (cC
= iI < cI < iC, see the table 2 for all post-hoc comparison, full the statistics are provided in the supplemental
materials).
3.1.2!Accuracy
A main effect of age was found, school children were more accurate than preschool children (z=-3.29, p=.001)
with an average accuracy of 95.1 ± 0.82% and 90.3 ± 1.2% for preschool and school children respectively.
While they were as accurate in the null-interference condition as in the condition involving an interference
between response alternatives, accuracy scores of both preschool and school children decreased in the
interference conditions involving a change in the task-set compared to the null-interference condition. The two
groups were also more accurate at resolving an interference at the response level than at the task-set level and
when the interference occurred at both the task-set and the response levels than when it only occurred at the
task-set level (for both preschool and school groups: cC = iI < cI < iC, see the table 2 for all post-hoc comparison,
full the statistics are provided in the supplemental materials and visualization of the group specific interferences
effect on figure 2b).
Although we have no hypothesis with regards to the differences between the intermediate interference
conditions, behavioral results indicated that the response interference was less difficult to resolve in the two
task-set interferences. Indeed, even if preschool children’s RT were slower in this condition, they had the same
accuracy scores than in the null interference condition suggesting that they might have decreased their RT in
order to respond accurately.
3.2!EEG Results
Surrogate testing done at the electrode-level revealed an increase in theta power amplitude compared to the
baseline period (-400ms to -100ms) at the FCz location, in both groups (p<.05, Figure 3a). Permutation z-tests
on the 64 channels in the theta frequency band during the task-related period showed a widespread increase in
MFT power amplitude (all children, p< .01). The figure 3b shows the topographical distribution of theta-band
activity over time in steps of 300ms for group specific data, averaged across conditions. Task-related increase
in MFT related to baseline was found for the average condition and for each interference condition compared to
the baseline in both groups (see figure 3c for preschool children and 3d for school children). The significant area
(bounded by a black line) did contain the FCz location. However, with regards to the width of the selected time
window (because of inter-participant differences), more diffused mid-frontal theta power was observed in the
plot.
However, given the difference in brain maturation and the topographical differences in mid-frontal theta across
participants and groups, traditional single FCz electrode analysis approach would have been suboptimal for a
subset of participants. Therefore, we applied a guided multivariate spatiotemporal filtering technique (Cohen,
2017; Duprez, Gulbinaite, & Cohen, 2018; Gulbinaite, van Viegen, Wieling, Cohen, & VanRullen, 2017), which
allowed us to combine information from all recorded electrodes and to isolate mid-frontal theta activity (for
details, see Methods section). All the following analysis results are reported for MFT component.
3.2.1!Analyzes of average mid frontal theta peak latency
Comparing the MFT peak latencies between groups we found that the peak occurred earlier in school children
compared to preschool children (t(67)= -4.538; p<.0001), with a difference estimated at -368.43ms [203.26
533.61]. While school children’s peak occurred around 587.12 ± 45.52ms the peak was observed at around
955.56 ±70.93ms in preschool children (MFT power peak latencies by group are represented in figure 4a).
3.2.1.1!Relation between response times and theta peak latency
Children’s RT were explained by the latencies of mid frontal theta with longer RT associated with longer
latencies (β= 0.0002 [0.0001, 0.0002]; t=3.125; p<.01). This relationship did not differ between preschool and
school children (β= -0.0001 [-0.0002, 0.0001]; t=-0.925; p=.359).
3.2.2!Analyses of average mid frontal theta power
In preschool children MFT power was increased in the interference conditions involving a change in the task-
set compared to the null-interference condition (βcI = 0.562 [0.111, 1.012], t(201)=2.441, p=.015; β iC = 0.554
[0.103, 1.005], t(201)=2.408, p=.017). However, MFT power did not vary when the interference occurred at the
response level compared to the null interference condition (βiI= 0.321 [-0.129, 0.772], t(201)=1.397, p=.163).
In school children MFT power did not differ between all the interference conditions including the null-
interference condition (cC - cI: t(197.68) = 0.923, p=.5601; cC - iC: t(197.68) = -0.776, p=.560; cC - iI:
t(197.68)=0.729, p=.5601).
Overall, MFT power did not differ between preschool and school children (Δ = -0.172 [-0.58 0.236]; t(197.68)
= -0.831, p=.407 with Type II error = 0.24) (MFT power averaged across interference conditions for preschool
and school children , see figure 4b). However, some differences were observed between groups for some
interference conditions (see figure 4c). In particular, MFT power was higher in school children than in preschool
children in the null-interference condition (β= 0.584 [0.015, 1.153], t(197)=2.011, p=.045).
3.2.3!Relation between response times and theta power
Examining the relation between children’s MFT Power and RT we found that variations in RT were not
explained by variations in MFT in any of the interference conditions (see the table in the supplemental materials
for all statistics). However, we noticed that for preschool children there was a near significant difference between
the regression slope of the null interference condition and that of the response interference condition (Power x
iI: β= -0.027 [-0.056, 0.002], t = -1.821, p = .069).
4!Discussion
Mid-frontal theta (MFT) has been proposed to serve as a mechanism used for implementing cognitive control
(Cavanagh et al., 2009; Cavanagh & Frank, 2014; Cohen & Cavanagh, 2011). In particular, MFT is involved in
the resolution of different types of interference occurring at distinct processing stages (Nigbur et al., 2011). Yet
it is not known whether such a mechanism plays a general role in cognitive control across development. Here
we show involvement of MFT in different interference situations in both school and preschool children,
suggesting a role for MFT in different cognitive control processes and at different stages of their development.
The resolution of interference involves different cognitive processes like conflict detection (between task-sets
or response alternatives), signaling or monitoring of the need for cognitive control, selection and preparation of
the correct response and inhibition of the incorrect one. MFT activity is known to be involved in both conflict
detection and resolution processes (Cavanagh et al., 2012; Cohen & Cavanagh, 2011). However, the uncertainty
about the timing of such processes (Huster, Enriquez-Geppert, Lavallee, Falkenstein, & Herrmann, 2013;
Kropotov, Ponomarev, Hollup, & Mueller, 2011) as well as the selection, in the present study, of an extended
time window for the detection of the MFT peak, prevent us from unequivocally associate MFT power to some
specific processes of cognitive control. In the following discussion, we, therefore, interpret MFT power as a
generic indicator of cognitive control implementation.
MFT was increased in all the interference conditions compared to the baseline in both preschool and school
children. Consistent with previous findings in adults showing enhanced theta power during interference trials in
a Simon task (Cohen & Donner, 2013; Nigbur et al., 2011) and a Stroop task (Hanslmayr et al., 2008), MFT
power was increased in the interference condition involving a conflict between the response alternatives elicited
by the irrelevant dimension of the stimulus (the spatial location of the stimulus) and the relevant dimension of
the stimulus (the shape of the stimulus). MFT power was also enhanced when the interference occurred at the
task-set level. This result is consistent with the observation of theta power changes during switch tasks in adults
(Cunillera et al., 2012; Sauseng et al., 2006). To the best of our knowledge, this study is the first to show that
MFT is also involved in the resolution of interference at the task-set level during development. These results
extend previous results showing a link between frontal theta and cognitive control processes in preschool
children (Bell & Wolfe, 2007) and in infants (Bazhenova, Stroganova, Doussard-Roosevelt, Posikera, & Porges,
2006; Orekhova et al., 1999) by further showing its involvement for different cognitive control processes.
Cognitive control implementation change over childhood, in particular through an increased functional
specialization of brain systems that are originally relatively undifferentiated (Johnson & Munakata, 2005). Those
changes occurs with respect to the degree of unity of the executive functions with well-differentiated working
memory, shifting and inhibitory processes, occurring at the school ages (Huizinga, Dolan, & van der Molen,
2006; Lehto et al., 2003). Of particular relevance to the present study is the observation that executive functions
are less differentiated in preschool-age children compared with school-age children (Hughes & Ensor, 2011).
Yet, in the present study MFT power was increased during the engagement of cognitive control in both preschool
and school groups. This, suggest that MFT is involved in the implementation of immature cognitive control at
different levels of its development.
Consistent with previous findings we observed age differences in MFT latency, with school children engaging
theta oscillations more rapidly than preschool children. Faster latencies were indeed observed for some event
related components in adolescents and adults compared to preschool and school children. In inhibition tasks,
faster latencies were observed for both the N200 component (Lamm, Zelazo, & Lewis, 2006), which reflects
conflict monitoring (Botvinick, Nystrom, Fissell, Carter, & Cohen, 1999) and in particular the detection of an
interference and the need of cognitive control (Hämmerer, Müller, & Li, 2014; Nieurwenhuis, Yeung, Den
Wildenberg, & Ridderinkhof, 2003) as well as the P300 component (Davis, Bruce, Snyder, & Nelson, 2003;
Ridderinkhof & van der Molen, 1995). Faster brain responses may result from brain structural changes occurring
during childhood, like axonal myelination or pruning, resulting in enhanced connectivity and faster neural
communication for more mature children (Brown & Jernigan, 2012) as well as more developed connections of
the brain networks supporting cognitive control (Gogtay et al., 2004; Hämmerer et al., 2014; Hwang, Velanova,
& Luna, 2010; Sowell et al., 2003). Faster neural communication allows to increase the speed of information
processing and explains age-related improvements in cognitive control. Consistent with this interpretation we
found faster RT in school children as well as positive correlations between MFT latencies and RT in both groups.
These results suggest that children’s improved performance as reflected by faster RT were explained by a faster
implementation of MFT oscillatory activity.
We did not observe any differences between preschool and school children in MFT power averaged across
interference conditions. At first sight this result seems contradictory with the observation that the development
of cognitive control is accompanied by increased activity in mid-frontal areas as reported with EEG studies (Bell
& Wolfe, 2007; Liu et al., 2014) as well as fMRI studies in the anterior cingulate cortex and the medial prefrontal
cortex (Ordaz, Foran, Velanova, & Luna, 2013; Rubia, Smith, Taylor, & Brammer, 2007) and the inferior
prefrontal cortex (Rubia et al., 2007; Tamm, Menon, & Reiss, 2002). However, age-related decreases in brain
activity within the frontal areas have also been reported with inhibition tasks (Booth et al., 2003; Tamm et al.,
2002). Therefore, this literature shows that there is no clear understanding of the changes in brain activity related
to the development of cognitive control. Furthermore, to the best of our knowledge, no study has assessed these
changes in children aged four to eight years old. It is therefore unknown whether brain activity related to
cognitive control shows any substantial changes over this age range. The absence of any differences in MFT
power between preschool and school children could also be explained by the very small age difference between
our two groups (four to six years old and six to eight years old) that might not have been enough to observe
some changes in the activity of these brain networks.
The level of MFT power reflects the engagement of cognitive control with greater power associated with greater
cognitive control demands (Cavanagh & Shackman, 2015; Hanslmayr et al., 2008; Wascher et al., 2014). MFT
power was enhanced in some conditions of interference that engaged more cognitive control compared to other
interference conditions. In particular, in preschool children, MFT power increased in the task-set interference
conditions compared to the null interference condition. Resolving an interference at the task-set level is indeed
very costly in terms of cognitive control resources as shown by slower and less accurate responses compared to
all other conditions in both preschool and school children. This finding is consistent with developmental studies
showing that the lowest performance is obtained with set-shifting tasks and that the resolution of a task-set
presents the longer developmental changes with respect to other executive functions (Davidson, Amso,
Anderson, & Diamond, 2006). However, in preschool children the resolution of response interference did not
elicit an increase in theta power when it followed another response interference trial. Furthermore, preschool
children were equally accurate in this condition than in the null interference condition. These findings can be
interpreted in terms of a facilitatory repetition effect in which the activation of cognitive control for the
resolution of the interference in the preceding trial is activated over the subsequent trial. It has indeed been
shown that preschool children’s performance can benefit from such sequential effects (Ambrosi, Lemaire, &
Blaye, 2016; Iani, Stella, & Rubichi, 2014). In school children, theta power was maintained equal in all
interference conditions including the null-interference condition. These results partly diverge from previous
findings in adults showing enhanced theta power consecutively to interference generated by the incompatibility
between task-relevant and irrelevant dimensions in both a Simon task (Cohen & Donner, 2013; Nigbur et al.,
2011) and a Stroop task (Hanslmayr et al., 2008). The absence of modulation of theta power across conditions,
and in particular between the conditions involving interference and the baseline condition, could suggest that
school children equally engaged cognitive control across all the conditions. We indeed observed that theta power
was higher in the null interference condition in school children compared to preschool children. School
children’s engagement of cognitive control throughout the present task can be explained by the mixed nature of
the task where trials with different types of interference were randomly mixed, eliciting important uncertainty
about the upcoming trial with respect to the presence of interference at the task-set and/or the response-tendency
level. Increased theta power was indeed found to be linked to the need of dynamical adjustments of cognitive
control in mixed tasks (van de Vijver et al., 2011; Womelsdorf et al., 2010).
The development of cognitive control is mirrored by improvements in the efficiency of executive functions like
cognitive flexibility and inhibition. These functions present some differences in their developmental time
courses, partly due to differences in the maturational changes of their respective parieto-prefrontal networks
(Bunge & Zelazo, 2006). However, beyond these differences we show that MFT is involved in the
implementation of immature cognitive control processes that support both inhibitory and flexibility functions.
We propose that MFT serve as a general neural mechanism to coordinate the cognitive processes involved in
cognitive control across development. These findings could have important implications in the field of cognitive
interventions (Ursache, Blair, & Raver, 2012) in children by helping to better understand the effects of such
interventions on the development of the brain networks supporting cognitive control.
Authors contributions
NA, CF and AB designed the experiment. NA and CF carried out the data collection. RG shared her expertise
on EEG data processing and analysis and co-wrote with NA the EEG analysis code. AD helped with data
analyses. NA performed the analyses and NA and CF drafted the manuscript. CF, AB and RG participated in
the interpretation of the results and provided critical revisions. All authors approved the final version of the
manuscript.
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships
that could be construed as a potential conflict of interest.
Acknowledgements
We thank Sasskia Brüers and Jean-Michel Hupé for helpful comments on the data analysis. We thank all the
children who participated in the study as well as their parents who gave their consents. We thank all the
employees of the three schools where this research was conducted, Michelet, Villebourbon and Leo Ferré
schools. Special thanks to Jerome De La Osa, Nathalie Delpoux and Fanny Haiart.
Funding
The present research was funded by a private donation of a philanthropic not-for-profit foundation to CF (grant
number R142068).
5!References
Ambrosi, S., Lemaire, P., & Blaye, A. (2016). Do young children modulate their cognitive control?:
Sequential congruency effects across three conflict tasks in 5-to-6 year-olds. Experimental Psychology,
63(2), 117–126. https://doi.org/10.1027/1618-3169/a000320
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models using lme4. JSS
Journal of Statistical Software, 67(1). https://doi.org/10.18637/jss.v067.i01
Bazhenova, O. V., Stroganova, T. A., Doussard-Roosevelt, J. A., Posikera, I. A., & Porges, S. W. (2007).
Physiological responses of 5-month-old infants to smiling and blank faces. International Journal of
Psychophysiology, 63(1), 64–76. https://doi.org/10.1016/j.ijpsycho.2006.08.008
Bell, M. A. (2002). Power changes in infant EEG frequency bands during a spatial working memory task.
Psychophysiology, 39(4), 450–458. https://doi.org/10.1017/S0048577201393174
Bell, M. A., & Cuevas, K. (2012). Using EEG to Study Cognitive Development: Issues and Practices. Journal
of Cognition and Development. https://doi.org/10.1080/15248372.2012.691143
Bell, M. A., & Wolfe, C. D. (2007). Changes in brain functioning from infancy to early childhood: evidence
from EEG power and coherence working memory tasks. Developmental Neuropsychology, 31(1), 21–38.
https://doi.org/10.1207/s15326942dn3101_2
Best, J. R., Miller, P. H., & Jones, L. L. (2009). Executive functions after age 5: Changes and correlates.
Developmental Review. https://doi.org/10.1016/j.dr.2009.05.002
Blaye, A., & Chevalier, N. (2011). The role of goal representation in preschoolers’ flexibility and inhibition.
Journal of Experimental Child Psychology, 108(3), 469–483. https://doi.org/10.1016/j.jecp.2010.09.006
Booth, J. R., Burman, D. D., Meyer, J. R., Lei, Z., Trommer, B. L., Davenport, N. D., … Mesulam, M. M.
(2003). Neural development of selective attention and response inhibition. NeuroImage, 20(2), 737–751.
https://doi.org/10.1016/S1053-8119(03)00404-X
Bosseler, A. N., Taulu, S., Pihko, E., Mäkelä, J. P., Imada, T., Ahonen, A., & Kuhl, P. K. (2013). Theta brain
rhythms index perceptual narrowing in infant speech perception. Frontiers in Psychology, 4(October), 1–
12. https://doi.org/10.3389/fpsyg.2013.00690
Botvinick, M., Nystrom, L. E., Fissell, K., Carter, C. S., & Cohen, J. D. (1999). Conflict monitoring versus
selection for-action in anterior cingulate cortex. Nature, 402(6758), 179–181.
https://doi.org/10.1038/46035
Brown, T. T., & Jernigan, T. L. (2012). Brain development during the preschool years. Neuropsychology
Review. https://doi.org/10.1007/s11065-012-9214-1
Bunge, S. A., & Wright, S. B. (2007). Neurodevelopmental changes in working memory and cognitive control.
Current Opinion in Neurobiology. https://doi.org/10.1016/j.conb.2007.02.005
Bunge, S. A., & Zelazo, P. D. (2006). A Brain-Based Account of the Development of Rule Use in Childhood.
Current Directions in Psychological Science, 15(3), 118–121.
Buzsaki, G., & Draguhn, A. (2004). Neuronal Oscillations in Cortical Networks. Science, 304(5679), 1926–
1938. Retrieved from http://science.sciencemag.org/
Calderon, J., Jambaqué, I., Bonnet, D., & Angeard, N. (2014). Executive functions development in 5- to 7-
year-old children with transposition of the great arteries: Longitudinal study. Developmental
Neuropsychology, 39(5), 365–384. https://doi.org/10.1080/87565641.2014.916709
Cao, J., Wang, S., Ren, Y., Zhang, Y., Cai, J., Tu, W., … Xia, Y. (2013). Interference control in 6-11 year-old
children with and without ADHD: Behavioral and ERP study. International Journal of Developmental
Neuroscience, 31(5), 342–349. https://doi.org/10.1016/j.ijdevneu.2013.04.005
Casey, B. J., Davidson, M. C., Hara, Y., Thomas, K. M., Martinez, A., Galvan, A., … Tottenham, N. (2004).
Early development of subcortical regions involved in non-cued attention switching. Developmental
Science, 7(5), 534–542. https://doi.org/10.1111/j.1467-7687.2004.00377.x
Cavanagh, J. F., Cohen, M. X., & Allen, J. J. B. (2009). Prelude to and Resolution of an Error: EEG Phase
Synchrony Reveals Cognitive Control Dynamics during Action Monitoring. Journal of Neuroscience,
29(1), 98–105. https://doi.org/10.1523/jneurosci.4137-08.2009
Cavanagh, James F., & Frank, M. J. (2014). Frontal theta as a mechanism for cognitive control. Trends in
Cognitive Sciences. https://doi.org/10.1016/j.tics.2014.04.012
Cavanagh, James F., & Shackman, A. J. (2015). Frontal midline theta reflects anxiety and cognitive control:
Meta-analytic evidence. Journal of Physiology Paris, 109(1–3), 3–15.
https://doi.org/10.1016/j.jphysparis.2014.04.003
Cavanagh, James F., Zambrano-Vazquez, L., & Allen, J. J. B. (2012). Theta lingua franca: A common mid-
frontal substrate for action monitoring processes. Psychophysiology, 49(2), 220–238.
https://doi.org/10.1111/j.1469-8986.2011.01293.x
Chaumon, M., Bishop, D. V. M., & Busch, N. A. (2015). A practical guide to the selection of independent
components of the electroencephalogram for artifact correction. Journal of Neuroscience Methods, 250,
47–63. https://doi.org/10.1016/j.jneumeth.2015.02.025
Chevalier, N., & Blaye, A. (2008). Cognitive flexibility in preschoolers: The role of representation activation
and maintenance: Paper. Developmental Science, 11(3), 339–353. https://doi.org/10.1111/j.1467-
7687.2008.00679.x
Chevalier, N., Blaye, A., Dufau, S., & Lucenet, J. (2010). What Visual Information Do Children and Adults
Consider While Switching Between Tasks? Eye-Tracking Investigation of Cognitive Flexibility
Development. Developmental Psychology, 46(4), 955–972. https://doi.org/10.1037/a0019674
Cheveigné, A. De, & Arzounian, D. (2015). Scanning for oscillations. Journal of Neural Engineering, 12(6).
https://doi.org/10.1088/1741-2560/12/6/066020
Cohen, Michael X. (2014). A neural microcircuit for cognitive conflict detection and signaling. Trends in
Neurosciences, 37(9), 480–490. https://doi.org/10.1016/j.tins.2014.06.004
Cohen, Michael X. (2017). Comparison of linear spatial filters for identifying oscillatory activity in
multichannel data. Journal of Neuroscience Methods, 278, 1–12.
https://doi.org/10.1016/j.jneumeth.2016.12.016
Cohen, Michael X., & Cavanagh, J. F. (2011). Single-trial regression elucidates the role of prefrontal theta
oscillations in response conflict. Frontiers in Psychology, 2(FEB).
https://doi.org/10.3389/fpsyg.2011.00030
Cohen, Michael X., & Gulbinaite, R. (2017). Rhythmic entrainment source separation: Optimizing analyses of
neural responses to rhythmic sensory stimulation. NeuroImage, 147(November 2016), 43–56.
https://doi.org/10.1016/j.neuroimage.2016.11.036
Cohen, Michael X., & Ridderinkhof, K. R. (2013). EEG Source Reconstruction Reveals Frontal-Parietal
Dynamics of Spatial Conflict Processing. PLoS ONE, 8(2). https://doi.org/10.1371/journal.pone.0057293
Cohen, Michael X, & Donner, T. H. (2013a). Midfrontal conflict-related theta-band power reflects neural
oscillations that predict behavior. Journal of Neurophysiology, 110(12), 2752–2763.
https://doi.org/10.1152/jn.00479.2013
Cohen, Michael X, & Donner, T. H. (2013b). Midfrontal conflict-related theta-band power reflects neural
oscillations that predict behavior. Journal of Neurophysiology, 110(12), 2752–2763.
https://doi.org/10.1152/jn.00479.2013
Cohen, Mike X. (2014). Analyzing neural time series data: theory and practice. MIT press.
Conejero, Á., Guerra, S., Abundis-Gutiérrez, A., & Rueda, M. R. (2018). Frontal theta activation associated
with error detection in toddlers: influence of familial socioeconomic status. Developmental Science,
21(1). https://doi.org/10.1111/desc.12494
Crone, E. A., Bunge, S. A., Van Der Molen, M. W., & Richard Ridderinkhof, K. (2006). Switching between
tasks and responses: A developmental study. Developmental Science, 9(3), 278–287.
https://doi.org/10.1111/j.1467-7687.2006.00490.x
Crone, E. A., & Steinbeis, N. (2017). Neural Perspectives on Cognitive Control Development during
Childhood and Adolescence. Trends in Cognitive Sciences, 21(3), 205–215.
https://doi.org/10.1016/j.tics.2017.01.003
Cunillera, T., Fuentemilla, L., Periañez, J., Marco-Pallarès, J., Krämer, U. M., Càmara, E., … Antoni, R. F.
(2012). Brain oscillatory activity associated with task switching and feedback processing. Cognitive,
Affective and Behavioral Neuroscience, 12(1), 16–33. https://doi.org/10.3758/s13415-011-0075-5
Davidson, M. C., Amso, D., Anderson, L. C., & Diamond, A. (2006a). Development of cognitive control and
executive functions from 4 to 13 years: evidence from manipulations of memory, inhibition, and task
switching. Neuropsychologia, 44(11), 2037–2078.
https://doi.org/10.1016/j.neuropsychologia.2006.02.006
Davidson, M. C., Amso, D., Anderson, L. C., & Diamond, A. (2006b). Development of cognitive control and
executive functions from 4 to 13 years: Evidence from manipulations of memory, inhibition, and task
switching. Neuropsychologia, 44(11), 2037–2078.
https://doi.org/10.1016/j.neuropsychologia.2006.02.006
Davis, E. P., Bruce, J., Snyder, K., & Nelson, C. A. (2003). The X-Trials: Neural Correlates of an Inhibitory
Control Task in Children and Adults. Journal of Cognitive Neuroscience, 15(3), 432–443.
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(1), 9–21.
https://doi.org/10.1016/j.jneumeth.2003.10.009
Delorme, A., Westerfield, M., & Makeig, S. (2007). Medial prefrontal theta bursts precede rapid motor
responses during visual selective attention. Journal of Neuroscience, 27(44), 11949–11959.
https://doi.org/10.1523/JNEUROSCI.3477-07.2007
Diamond, A. (2013). Executive Function. Annual Review of Psychology, 64, 135:168.
https://doi.org/10.1016/B978-0-12-385157-4.01147-7
Diamond, A., & Ling, D. S. (2016). Conclusions about interventions, programs, and approaches for improving
executive functions that appear justified and those that, despite much hype, do not. Developmental
Cognitive Neuroscience, 18, 34–48. https://doi.org/10.1016/j.dcn.2015.11.005
Diamond, A., Steven Barnett, W., Thomas, J., & Munro, S. (2007). Preschool Program Improves Cognitive
Control NIH Public Access. Science , 318(5855), 1387–1388. Retrieved from
www.sciencemag.org/cgi/content/full/318/5855/1387/DC1
Doesburg, S. M., Vidal, J., & Taylor, M. J. (2013). Reduced Theta Connectivity during Set-Shifting in
Children with Autism. Frontiers in Human Neuroscience, 7(November), 785.
https://doi.org/10.3389/fnhum.2013.00785
Duprez, J., Gulbinaite, R., & Cohen, M. X. (2018). Midfrontal theta phase coordinates behaviorally relevant
brain computations during response conflict. BioRxiv, 502716. https://doi.org/10.1101/502716
Ezekiel, F., Bosma, R., & Morton, J. B. (2013). Dimensional Change Card Sort performance associated with
age-related differences in functional connectivity of lateral prefrontal cortex. Developmental Cognitive
Neuroscience, 5, 40–50. https://doi.org/10.1016/j.dcn.2012.12.001
Gogtay, N., Nugent, T. F., Thompson, P. M., Hayashi, K. M., Giedd, J. N., Vaituzis, A. C., … Toga, A. W.
(2004). Dynamic mapping of human cortical development during childhood through early adulthood.
Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.0402680101
Gulbinaite, R., & Johnson, A. (2014). Working memory capacity predicts conflict-task performance.
Quarterly Journal of Experimental Psychology (2006), 67(7), 1383–1400.
https://doi.org/10.1080/17470218.2013.863374
Gulbinaite, R., van Rijn, H., & Cohen, M. X. (2014). Fronto-parietal network oscillations reveal relationship
between working memory capacity and cognitive control. Frontiers in Human Neuroscience, 8.
https://doi.org/10.3389/fnhum.2014.00761
Gulbinaite, R., van Viegen, T., Wieling, M., Cohen, M. X., & VanRullen, R. (2017). Individual Alpha Peak
Frequency Predicts 10 Hz Flicker Effects on Selective Attention. The Journal of Neuroscience, 37(42),
10173–10184. https://doi.org/10.1523/jneurosci.1163-17.2017
Hämmerer, D., Müller, V., & Li, S. C. (2014). Performance monitoring across the lifespan: Still maturing
post-conflict regulation in children and declining task-set monitoring in older adults. Neuroscience and
Biobehavioral Reviews. https://doi.org/10.1016/j.neubiorev.2014.06.008
Hanslmayr, S., Pastötter, B., Bäuml, K. H., Gruber, S., Wimber, M., & Klimesch, W. (2008). The
electrophysiological dynamics of interference during the stroop task. Journal of Cognitive Neuroscience,
20(2), 215–225. https://doi.org/10.1162/jocn.2008.20020
Hughes, C., & Ensor, R. (2011). Individual differences in growth in executive function across the transition to
school predict externalizing and internalizing behaviors and self-perceived academic success at 6 years
of age. Journal of Experimental Child Psychology, 108(3), 663–676.
https://doi.org/10.1016/j.jecp.2010.06.005
Huizinga, M., Dolan, C. V., & van der Molen, M. W. (2006). Age-related change in executive function:
Developmental trends and a latent variable analysis. Neuropsychologia, 44(11), 2017–2036.
https://doi.org/10.1016/j.neuropsychologia.2006.01.010
Huizinga, M., & van der Molen, M. W. (2011). Task switching and shifting between stopping and going:
Developmental change in between-trial control adjustments. Journal of Experimental Child Psychology,
108(3), 484–503. https://doi.org/10.1016/j.jecp.2010.10.004
Huster, R. J., Enriquez-Geppert, S., Lavallee, C. F., Falkenstein, M., & Herrmann, C. S. (2013).
Electroencephalography of response inhibition tasks: Functional networks and cognitive contributions.
International Journal of Psychophysiology, 87(3), 217–233.
https://doi.org/10.1016/j.ijpsycho.2012.08.001
Hwang, K., Velanova, K., & Luna, B. (2010). Strengthening of Top-Down Frontal Cognitive Control
Networks Underlying the Development of Inhibitory Control: A Functional Magnetic Resonance
Imaging Effective Connectivity Study. Journal of Neuroscience, 30(46), 15535–15545.
https://doi.org/10.1523/JNEUROSCI.2825-10.2010
Iani, C., Stella, G., & Rubichi, S. (2014). Response inhibition and adaptations to response conflict in 6- to 8-
year-old children: Evidence from the Simon effect. Attention, Perception, and Psychophysics, 76(4),
1234–1241. https://doi.org/10.3758/s13414-014-0656-9
Jensen, O., & Tesche, C. D. (2002). Frontal theta activity in humans increases with memory load in a working
memory task. European Journal of Neuroscience, 15(8), 1395–1399. https://doi.org/10.1046/j.1460-
9568.2002.01975.x
Johnson, M. H., & Munakata, Y. (2005). Processes of change in brain and cognitive development. In Trends in
Cognitive Sciences (Vol. 9, pp. 152–158). https://doi.org/10.1016/j.tics.2005.01.009
Kayser, J., & Tenke, C. E. (2015). On the benefits of using surface Laplacian (current source density)
methodology in electrophysiology. International Journal of Psychophysiology, 97(3), 171–173.
https://doi.org/10.1016/j.ijpsycho.2015.06.001
Kornblum, S., Hasbroucq, T., & Osman, A. (1990). Dimensional Overlap: Cognitive Basis for Stimulus-
Response Compatibility-A Model and Taxonomy. Psychological Review, 97(2), 253–270.
https://doi.org/10.1037/0033-295X.97.2.253
Kray, J., Gaspard, H., Karbach, J., & Blaye, A. (2013). Developmental changes in using verbal self-cueing in
task-switching situations: The impact of task practice and task-sequencing demands. Frontiers in
Psychology. https://doi.org/10.3389/fpsyg.2013.00940
Kropotov, J. D., Ponomarev, V. A., Hollup, S., & Mueller, A. (2011). Dissociating action inhibition, conflict
monitoring and sensory mismatch into independent components of event related potentials in GO/NOGO
task. NeuroImage, 57(2), 565–575. https://doi.org/10.1016/j.neuroimage.2011.04.060
Lamm, C., Zelazo, P. D., & Lewis, M. D. (2006a). Neural correlates of cognitive control in childhood and
adolescence: Disentangling the contributions of age and executive function. Neuropsychologia, 44(11),
2139–2148. https://doi.org/10.1016/j.neuropsychologia.2005.10.013
Lamm, C., Zelazo, P. D., & Lewis, M. D. (2006b). Neural correlates of cognitive control in childhood and
adolescence: Disentangling the contributions of age and executive function. Neuropsychologia, 44(11),
2139–2148. https://doi.org/10.1016/j.neuropsychologia.2005.10.013
Lee, K., Bull, R., & Ho, R. M. H. (2013). Developmental changes in executive functioning. Child
Development, 84(6), 1933–1953. https://doi.org/10.1111/cdev.12096
Lehto, J. E., Juujärvi, P., Kooistra, L., & Pulkkinen, L. (2003). Dimensions of executive functioning: Evidence
from children. British Journal of Developmental Psychology, 21(1), 59–80.
https://doi.org/10.1348/026151003321164627
Liu, Z. X., Woltering, S., & Lewis, M. D. (2014). Developmental change in EEG theta activity in the medial
prefrontal cortex during response control. NeuroImage, 85, 873–887.
https://doi.org/10.1016/j.neuroimage.2013.08.054
Lorsbach, T. C., & Reimer, J. F. (2008). Context Processing and Cognitive Control in Children and Young
Adults. The Journal of Genetic Psychology, 169(1), 34–50. Retrieved from http://lama.univ-
amu.fr/login?url=http://search.proquest.com/docview/228460038?accountid=15337
Mathôt, S., Schreij, D., & Theeuwes, J. (2012). OpenSesame: An open-source, graphical experiment builder
for the social sciences. Behavior Research Methods, 44(2), 314–324. https://doi.org/10.3758/s13428-
011-0168-7
Meiran, N. (1996). Reconfiguration of processing mode prior to task performance. Journal of Experimental
Psychology: Learning Memory and Cognition, 22(6), 1423–1442. https://doi.org/10.1037/0278-
7393.22.6.1423
Meyer, M., Endedijk, H. M., van Ede, F., & Hunnius, S. (2019). Theta oscillations in 4-year-olds are sensitive
to task engagement and task demands. Scientific Reports, 9(1), 1–11. https://doi.org/10.1038/s41598-
019-42615-x
Michel, C., Stets, M., Parise, E., Reid, V. M., Striano, T., & Hoehl, S. (2015). Theta- and alpha-band EEG
activity in response to eye gaze cues in early infancy. NeuroImage, 118, 576–583.
https://doi.org/10.1016/j.neuroimage.2015.06.042
Miller, E. K., & Cohen, J. D. (2001). An Integrative Theory of Prefrontal Cortex Function. Annual Review of
Neuroscience, 24(1), 167–202. https://doi.org/10.1146/annurev.neuro.24.1.167
Miyake, a, Friedman, N. P., Emerson, M. J., Witzki, a H., Howerter, a, & Wager, T. D. (2000). The unity
and diversity of executive functions and their contributions to complex “Frontal Lobe” tasks: a latent
variable analysis. Cognitive Psychology, 41(1), 49–100. https://doi.org/10.1006/cogp.1999.0734
Monette, S., Bigras, M., & Lafrenière, M.-A. (2015). Structure of executive functions in typically developing
kindergarteners. Journal of Experimental Child Psychology, 140, 120–139.
https://doi.org/10.1016/j.jecp.2015.07.005
Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7(3), 134–140.
https://doi.org/10.1016/S1364-6613(03)00028-7
Morasch, K. C., & Bell, M. A. (2011). The role of inhibitory control in behavioral and physiological
expressions of toddler executive function. Journal of Experimental Child Psychology, 108(3), 593–606.
https://doi.org/10.1016/j.jecp.2010.07.003
Nieurwenhuis, S., Yeung, N., Den Wildenberg, W., & Ridderinkhof, K. R. (2003). Electrophysiological
correlates of anterior cingulate function in a gonogo task. Cognitive, Affective & Behavioral
Neuroscience, 3(1), 17–26.
Nigbur, R., Cohen, M. X., Ridderinkhof, K. R., & Stürmer, B. (2012). Theta Dynamics Reveal Domain-
specific Control over Stimulus and Response Conflict. Journal of Cognitive Neuroscience, 24(5), 1264–
1274.
Nigbur, R., Ivanova, G., & Stürmer, B. (2011). Theta power as a marker for cognitive interference. Clinical
Neurophysiology, 122(11), 2185–2194. https://doi.org/10.1016/j.clinph.2011.03.030
Nikulin, V. V., Nolte, G., & Curio, G. (2011). A novel method for reliable and fast extraction of neuronal
EEG/MEG oscillations on the basis of spatio-spectral decomposition. NeuroImage, 55(4), 1528–1535.
https://doi.org/10.1016/j.neuroimage.2011.01.057
O’leary, M. J., & Barber, P. J. (1994). Stimulus congruence and the Simon effect. Psychol Res (Vol. 56).
Obradović, J., Sulik, M. J., Finch, J. E., & Tirado-Strayer, N. (2018). Assessing students’ executive functions
in the classroom: Validating a scalable group-based procedure. Journal of Applied Developmental
Psychology, 55, 4–13. https://doi.org/10.1016/j.appdev.2017.03.003
Ordaz, S. J., Foran, W., Velanova, K., & Luna, B. (2013). Longitudinal Growth Curves of Brain Function
Underlying Inhibitory Control through Adolescence. Journal of Neuroscience, 33(46), 18109–18124.
https://doi.org/10.1523/jneurosci.1741-13.2013
Orekhova, E. V., Stroganova, T. A., Posikera, I. N., & Elam, M. (2006). EEG theta rhythm in infants and
preschool children. Clinical Neurophysiology, 117(5), 1047–1062.
https://doi.org/10.1016/j.clinph.2005.12.027
Orekhova, Elena V., Stroganova, T. A., & Posikera, I. N. (1999). Theta synchronization during sustained
anticipatory attention in infants over the second half of the first year of life. International Journal of
Psychophysiology, 32(2), 151–172. https://doi.org/10.1016/S0167-8760(99)00011-2
Papenberg, G., Hämmerer, D., Müller, V., Lindenberger, U., & Li, S. C. (2013). Lower theta inter-trial phase
coherence during performance monitoring is related to higher reaction time variability: A lifespan study.
NeuroImage, 83, 912–920. https://doi.org/10.1016/j.neuroimage.2013.07.032
R Core Team. (2014). R: A Language and Environment for Statistical Computing. Vienna, Austria. Retrieved
from http://www.r-project.org/
Richardson, C., Anderson, M., Reid, C. L., & Fox, A. M. (2018). Development of inhibition and switching: A
longitudinal study of the maturation of interference suppression and reversal processes during childhood.
Developmental Cognitive Neuroscience, 34, 92–100. https://doi.org/10.1016/j.dcn.2018.03.002
Ridderinkhof, K. R., & van der Molen, M. W. (1995). A Psychophysiological Analysis of Developmental
Differences in the Ability to Resist Interference. Child Development, 66(4), 1040–1056.
https://doi.org/10.1111/j.1467-8624.1995.tb00921.x
Rubia, K., Smith, A. B., Taylor, E., & Brammer, M. (2007). Linear age-correlated functional development of
right inferior fronto-striato-cerebellar networks during response inhibition and anterior cingulate during
error-related processes. Human Brain Mapping, 28(11), 1163–1177. https://doi.org/10.1002/hbm.20347
Sauseng, P., Klimesch, W., Freunberger, R., Pecherstorfer, T., Hanslmayr, S., & Doppelmayr, M. (2006).
Relevance of EEG alpha and theta oscillations during task switching. Experimental Brain Research,
170(3), 295–301. https://doi.org/10.1007/s00221-005-0211-y
Senn, T. E., Espy, K. A., & Kaufmann, P. M. (2004). Using path analysis to understand executive function
organization in preschool children. Developmental Neuropsychology, 26(1), 445–464.
https://doi.org/10.1207/s15326942dn2601_5
Sheridan, M., Kharitonova, M., Martin, R. E., Chatterjee, A., & Gabrieli, J. D. E. (2014). Neural Substrates of
the Development of Cognitive Control in Children Ages 5-10 Years. Journal of Cognitive Neuroscience,
26(8), 1840–1850. https://doi.org/10.1162/jocn_a_00597
Simon, J. R., & Berbaum, K. (1988). Effect of irrelevant information on retrieval time for relevant
information. Acta Psychologica, 67(1), 33–57. https://doi.org/10.1016/0001-6918(88)90023-6
Sowell, E. R., Peterson, B. S., Thompson, P. M., Welcome, S. E., Henkenius, A. L., & Toga, A. W. (2003).
Mapping cortical change across the human life span. Nature Neuroscience, 6(3), 309–315.
https://doi.org/10.1038/nn1008
St Clair-Thompson, H. L., & Gathercole, S. E. (2006). Executive functions and achievements in school:
Shifting, updating, inhibition, and working memory. Quarterly Journal of Experimental Psychology,
59(4), 745–759. https://doi.org/10.1080/17470210500162854
Tamm, L., Menon, V., & Reiss, A. L. (2002). Maturation of Brain Function Associated With Response
Inhibition. Journal of the American Academy of Child & Adolescent Psychiatry, 41(10), 1231–1238.
https://doi.org/10.1097/01.CHI.0000020272.43550.5E
Töllner, T., Wang, Y., Makeig, S., Müller, H. J., Jung, T.-P., & Gramann, K. (2017). Two Independent Frontal
Midline Theta Oscillations during Conflict Detection and Adaptation in a Simon-Type Manual Reaching
Task. The Journal of Neuroscience, 37(9), 2504–2515. https://doi.org/10.1523/jneurosci.1752-16.2017
Ursache, A., Blair, C., & Raver, C. C. (2012). The Promotion of Self-Regulation as a Means of Enhancing
School Readiness and Early Achievement in Children at Risk for School Failure. Child Development
Perspectives. https://doi.org/10.1111/j.1750-8606.2011.00209.x
van de Vijver, I., Richard Ridderinkhof, K., & Cohen, M. X. (2011). Frontal oscillatory dynamics predict
feedback learning and action adjustment. Journal of Cognitive Neuroscience, 23(12), 4106–4121.
https://doi.org/10.1162/jocn_a_00110
van der Ven, S. H. G., Kroesbergen, E. H., Boom, J., & Leseman, P. P. M. (2013). The structure of executive
functions in children: A closer examination of inhibition, shifting, and updating. British Journal of
Developmental Psychology, 31(1), 70–87. https://doi.org/10.1111/j.2044-835X.2012.02079.x
Von Stein, A., & Sarnthein, J. (2000). Different frequencies for different scales of cortical integration: From
local gamma to long range alpha/theta synchronization. International Journal of Psychophysiology,
38(3), 301–313. https://doi.org/10.1016/S0167-8760(00)00172-0
Wascher, E., Rasch, B., Sänger, J., Hoffmann, S., Schneider, D., Rinkenauer, G., … Gutberlet, I. (2014).
Frontal theta activity reflects distinct aspects of mental fatigue. Biological Psychology, 96(1), 57–65.
https://doi.org/10.1016/j.biopsycho.2013.11.010
Watson, A. J., & Bell, M. A. (2013). Individual differences in inhibitory control skills at three years of age.
Developmental Neuropsychology, 38(1), 1–21. https://doi.org/10.1080/87565641.2012.718818
Wendelken, C., Munakata, Y., Baym, C., Souza, M., & Bunge, S. A. (2012). Flexible rule use: Common
neural substrates in children and adults. Developmental Cognitive Neuroscience, 2(3), 329–339.
https://doi.org/10.1016/j.dcn.2012.02.001
Wiebe, S. A., Espy, K. A., & Charak, D. (2008). Using Confirmatory Factor Analysis to Understand Executive
Control in Preschool Children: I. Latent Structure. Developmental Psychology, 44(2), 575–587.
https://doi.org/10.1037/0012-1649.44.2.575
Wiebe, S. a., Sheffield, T., Nelson, J. M., Clark, C. a C., Chevalier, N., & Espy, K. A. (2011). The structure of
executive function in 3-year-olds. Journal of Experimental Child Psychology, 108(3), 436–452.
https://doi.org/10.1016/j.jecp.2010.08.008
Wolfe, C. D., & Bell, M. A. (2004). Working Memory and Inhibitory Control in Early Childhood:
Contributions from Physiology, Temperament, and Language. Developmental Psychobiology, 44(1), 68–
83. https://doi.org/10.1002/dev.10152
Wolfe, C. D., & Bell, M. A. (2007). The integration of cognition and emotion during infancy and early
childhood: Regulatory processes associated with the development of working memory. Brain and
Cognition, 65(1), 3–13. https://doi.org/10.1016/j.bandc.2006.01.009
Womelsdorf, T., Johnston, K., Vinck, M., & Everling, S. (2010). Theta-activity in anterior cingulate cortex
predicts task rules and their adjustments following errors. Proceedings of the National Academy of
Sciences, 107(11), 5248–5253. https://doi.org/10.1073/pnas.0906194107
Zelazo, P., & Jacques, S. (1997). Children’s rule use: Representation, reflection, and cognitive control.
Annal’s of Child Development, 12, 119–176. Retrieved from
https://www.researchgate.net/publication/284649375
6!Figures & Tables
Table 1: representation of the different interference conditions of the mixed task
Task-set interference
Yes
No
Response interference
Yes
cI
iI
No
iC
cC
Figure 1. Experimental protocol and stimuli. When a heart appeared on the screen children were instructed to
press on the same side as the stimulus (i.e. press left if the heart appeared on the left). When a flower stimulus
appeared on the screen, children were instructed to press on the opposite side of the stimulus (i.e. press right if
the flower appeared on the left).
Figure 2 Regression coefficients (with their corresponding 95% confidence intervals) for A) Response Time
and B) Accuracy scores in the different interference conditions (iI, cI, iC) compared to the reference condition
(i.e. the null-interference condition of each group; models were run independently in each group to visually
appreciate the difference between interference conditions within groups). For recall, if a confidence interval
crossed the dash line (representing the reference condition), it means that accuracy scores in the interference
condition do not differ from that of the reference condition.
Table 2. Post-hoc comparisons were conducted for comparing response times between interference conditions
in preschool and school children independently. Mean differences are reported as delta values with their
corresponding standard errors. The associated z- and p-values are also reported (significant differences are
indicated by grey cells).
N = 98
RT (ms)
Accuracy (%)
contrasts
Preschool children
School children
Preschool children
School children
cC / iC
(z=-6.786, p<.0001)
(z=-5.519, p<.0001)
(z=6.114, p<.0001)
(z=5.738, p<.0001)
Δ = -146.64 ±22ms
Δ = -99.83 ± 18ms
Δ = 8.34 ±1.36%
Δ = 7.88 ±1.37%
cC / cI
(z=-5.597, p<.0001)
(z=-3.198, p=.0021)
(z=4.060, p=.0001)
(z=3.376, p=.0009)
Δ = -141.19 ±25ms
Δ = -67.06 ± 21ms
Δ = 5.31 ±1.31%
Δ = 3. 44 ±1.02
cC / iI
(z=-3.018, p=.0038)
(z=-0.008, p=.9934)
(z=0.575, p=.565)
(z=-0.975, p=.329)
Δ = -79.23 ±26ms
Δ = -0.18 ± 22ms
Δ = 0.61 ±1.06%
Δ = -0.65 ± 0.67%
iC / cI
(z=0.288, p=.7734)
(z=2.112, p=.0416)
(z=-2.181, p=.035)
(z=-3.366, p=.0009)
Δ = 5.45 ± 19ms
Δ = 32.77 ±16ms
Δ = -3.04 ± 1.39%
Δ = -4.44 ±1.32%
iC / iI
(z=2.703, p=.0083)
(z=4.849, p<.0001)
(z=-5.430, p<.0001)
(z=-5.908, p<.0001)
Δ = 67.42 ±25ms
Δ = 99.66 ± 21ms
Δ = -7.73 ±1.42%
Δ = -8.53 ±1.44%
cI / iI
(z=3.239, p=.0024)
(z=4.417, p<.0001)
(z=-3.886, p=.0002)
(z=-4.161, p=.0001)
Δ = 61.97 ± 19ms
Δ = 66.89 ± 15ms
Δ = -4.69 ± 1.21%
Δ = -4.09 ±0.98%
Figure 3 A) Time frequency representation of data, averaged on groups and conditions, showing an increase
in theta power amplitude compared to the baseline at the FCz location. The topographical distribution of theta
power amplitude [4-8Hz] between 200 and 1500ms is also presented for the condition and group averaged data
(white dot indicates the FCz location). On the right, surrogate tests (with 1000 permutations) was used to
identify significant change in theta power amplitude relative to the baseline at the FCz location. The rectangle
corresponds to the spatio-temporal window used for topographic plots. ; B) Topographic distribution of theta
power amplitude [4-8Hz], averaged on condition for each group over 300ms period from -300 to 1500ms
(white dot indicates the FCz site); C) Surrogate tests (with 1000 permutations) in preschool children showing
significant change in theta power amplitude relative to the baseline. The significant area, surrounded by a
black line, did contain the FCz location. The right upper plot corresponds to the topographical distribution of
theta power amplitude [4-8Hz] between 200 to 1500ms for the data averaged on all interference conditions for
preschool children (white dot indicates the FCz location), whereas the four topoplots at the bottom show the
topographical distributions of theta amplitude in the same time window for each condition. D) Same as C for
school children.
Figure 4 A) Boxplot of the mean latency of the MFT peak showing an earlier peak in school compared to
preschool children; B) Time frequency plot showing the mean Mid Frontal Theta (MFT) power, calculated
over all interference conditions for preschool and school children with associated topographical MFT power
distribution (the white spot indicates the FCz location); C) Mean Mid Frontal Theta power for each
interference condition in preschool and school children (the bar indicates the 95 % confidence interval).
7!Supplementary material
Additional methodological information
Time-frequency decomposition was performed by convolving stimulus-locked single-trial data from
all electrodes with complex Morlet wavelets:
where t is time, f is frequency, which increased from 2 to 40 Hz in 40 logarithmically spaced steps,
and σ defines the width of each frequency band set according to n/(2πf), where n is a number of
wavelet cycles that varied from 3 to 6 in logarithmically spaced steps.
From the resulting complex signal, an estimate of frequency band-specific power at each time point
was defined as the squared magnitude of the result of the convolution:
Single-trial power values were then averaged across trials for each condition separately, and then
normalized relative to the condition-average baseline period (-400 to -100 ms) using a decibel (dB)
transform at each frequency:
For each participant, we identified a single MFT component using the following steps. First, single-
trial data were filtered in the theta band using a narrow-band Gaussian filter (peak 6 Hz, full-width
at half-maximum (FWHM) = 3 Hz). Second, for each subject temporally filtered single-trial data
was used to compute the covariance matrix S (“signal” matrix) and broadband unfiltered data was
used to compute the covariance matrix R (“reference” matrix). Third, generalized eigenvalue
decomposition was performed on the covariance matrices S and R (MATLAB function eig) to
obtain a set of vectors or spatial filters (W), which maximally separate matrix S from matrix R:
where W is the matrix of eigenvectors and Λ is a diagonal matrix of eigenvalues.
The first 6 vectors (spatial filters) associated with the highest eigenvalues (diagonal elements in
matrix Λ) were applied to unfiltered data resulting in 6 component time series. Topographical
representation of the spatial filters (activation patterns; (Haufe et al., 2014)) and time-frequency
representation of trial-average component time series were visually inspected and a single component
was identified per subject.
We identified individual theta power peak in the condition-averaged data within the window 200 to
1500 ms, then a smaller subject specific window was determined around the peak frequency (±
1Hz) and time (± 150ms). The theta power within this specific window was average for each
condition to compute the individual mid frontal theta power (MFT). See figure S1 for visualization
of the procedure on a single subject.
Figure S1: On the left, Spatio frequential representation of the baseline corrected MFT component power amplitude for
the condition averaged data of one randomly selected subject. The black star show the peak in amplitude within the theta
range window between 200 and 1500ms. The black rectangle show the extended window that correspond to the peak ±
1Hz and ±150ms. On the right, the previously determinate window is used to extract the condition specific MFT theta
power. The activity in the window is averaged to obtain a single measure of MFT power amplitude by condition by
subject.
Appendix A
βCI t-value p-value βCI t-value p-value βCI t-value p-value βCI t-value p-value
Intercept 7.063 (7.017, 7.109) 299.610 <.0001 2.640 (2.279, 3.001) 14.337 <.0001 7.052 (6.993, 7.112) 233.346 <.0001 2.031 (1.637, 2.425) 10.113 <.0001
Group -0.333 (-0.405, -0.260) -9.023 <.0001 0.785 (0.228, 1.341) 2.764 0.006 -0.335 (-0.421, -0.249) -7.625 <.0001 0.584 (0.015, 1.153) 2.011 0.045
cI 0.114 (0.074, 0.154) 5.575 <.0001 -0.644 (-0.945, -0.344) -4.200 <.0001 0.113 (0.060, 0.166) 4.147 <.0001 0.562 (0.111, 1.012) 2.441 0.015
iC 0.118 (0.084, 0.152) 6.831 <.0001 -0.906 (-1.185, -0.626) -6.353 <.0001 0.109 (0.065, 0.152) 4.920 <.0001 0.554 (0.103, 1.005) 2.408 0.017
iI 0.066 (0.023, 0.108) 3.009 0.003 -0.095 (-0.415, 0.226) -0.577 0.564 0.073 (0.018, 0.129) 2.580 0.010 0.321 (-0.129, 0.772) 1.397 0.163
Group : cI -0.037 (-0.099, 0.025) -1.170 0.242 -0.130 (-0.591, 0.332) -0.551 0.582 -0.033 (-0.111, 0.044) -0.846 0.398 -0.783 (-1.435, -0.131) -2.355 0.019
Group : iC -0.006 (-0.058, 0.047) -0.212 0.833 -0.432 (-0.853, -0.011) -2.012 0.045 0.003 (-0.060, 0.065) 0.082 0.935 -0.367 (-1.019, 0.284) -1.105 0.270
Group : iI -0.065 (-0.132, 0.001) -1.936 0.053 0.332 (-0.166, 0.830) 1.306 0.192 -0.067 (-0.147, 0.013) -1.635 0.103 -0.496 (-1.148, 0.155) -1.493 0.136
MFT power - - - - - - - - 0.002 (-0.019, 0.022) 0.154 0.878 - - - -
MFT power : cI - - - - - - - - 0.002 (-0.026, 0.030) 0.121 0.904 - - - -
MFT power : iC - - - - - - - - 0.012 (-0.017, 0.040) 0.819 0.413 - - - -
MFT power : iI - - - - - - - - -0.027 (-0.056, 0.002) -1.821 0.069 - - - -
MFT power : Group - - - - - - - - 0.002 (-0.027, 0.031) 0.112 0.911 - - - -
MFT power : Group : cI - - - - - - - - - 0.017 (-0.057, 0.022) -0.846 0.398 - - - -
MFT power : Group : iC - - - - - - - - -0.018 (-0.059, 0.022) -0.901 0.368 - - - -
MFT power : Group : iI - - - - - - - - 0.019 (- 0.021, 0.060) 0.939 0.348 - - - -
Intercept
cI
iC
iI
Residuals
N subject
N obs
K-children cC
K-children cI
K-children iC
K-children iI
P-children cC
P-children cI
P-children iC
P-children iI
Log Likelihood
AIC
Model's family
Link
Diagnostic tables for models runned via lme4 package on R with description of fixed and random effects as well as parameters of the models
* Response time ~Group * interference condition
** Response time ~Group * Interference condition * MFT Power
Response time*
Accuracy
Response time**
Random effects
9395
Variance
SD
0,006
0,08
0,006
0,089
0,004
Variance
SD
1,17
0,24
0,14
0,13
1,08
0,49
0,07
0,08
129920.400
-64873.310
7478.7
-3721.3
logit
log
Binomial
Gamma
0,38
0,36
98
11676
-
-
98
0,007
0,06
0,24
log
Parameters
763
769
687
682
753
737
0,09
0,06
0,23
69
5821
693
0,008
1074
1045
1566
1818
1822
1556
1159
1288
1305
MFT power
GLMer
Variance
SD
0,5
0,71
737
-40058.69
80171.38
Variance
SD
0,007
0,08
0,007
0,09
0,004
0,07
1162
1262
1356
1324
1246
1005
1083
-425.774
871.548
Gaussian (lmer)
identity
Fixed effects
-
-
-
-
-
-
0,95
0,98
69
276
-
-
-
-
-
-
-
-
Gamma
Appendix B
βCI t-value p-value βCI t-value p-value
Intercept 6.999 (6.902, 7.096) 141.449 <.0001 6.862 (6.717, 7.008) 92.453 <.0001
Group -0.267 (-0.398, -0.136) -4.003 0.0002 -0,487 (-0,697, -0,277) -4.538 <.0001
cI - - - - - - - -
iC - - - - - - - -
iI - - - - - - - -
Group : cI - - - - - - - -
Group : iC - - - - - - - -
Group : iI - - - - - - - -
Latency 0.0002 (0.0001, 0.0002) 3.125 0.003 - - - -
Grp : Latency -0.0001 (-0.0002, 0.0001) -0.925 0.359 - - - -
N subject
N obs
Log Likelihood
AIC
Residual deviance
Null Deviance
Model's family
Link
Diagnostic tables for GLM models runned on R with description of fixed and parameters of the models
** Response time ~ Group * Latency
15.240 (df = 67)
19.257 (df = 68)
Gamma
log
GLM
Latency
69
69
-497.879
999.757
69
69
Parameters
Fixed effects
Response time**
Gamma
log
0.838 (df = 65)
3.302 (df = 68)
-427.057
862.114
Appendix C
Supplemental table Post-hoc comparisons for response times and accuracy between interference
conditions in Preschool and School children independently for the 69 children involved in EEG
analysis. Mean differences are reported as delta values with their corresponding standard errors. The
associated z- and p-values are also reported (significant differences are indicated by grey cells).
N = 69
RT (ms)
Accuracy (%)
contrasts
Preschool children
School children
Preschool children
School children
cC / iC
(z=-5,549, p<.0001)
(z=-4.899, p<.0001)
(z=4.650, p<.0001)
(z=4.658, p<.0001)
Δ = -151,62 ±27ms
Δ = -99.63 ± 20ms
Δ = 8.55 ±1.84%
Δ = 7.07 ±1.52%
cC / cI
(z=-4,481, p<.0001)
(z=-2.803, p=.0076)
(z=3.441, p=.0009)
(z=2.999, p=.0041)
Δ = -147,54 ±33ms
Δ = -68.12 ± 24ms
Δ = 5.99 ±1.74%
Δ = 3. 59 ±1.19
cC / iI
(z=-2,487, p=.0193)
(z=-0.050, p=.9598)
(z=0.212, p=.8319)
(z=-1.285, p=.1988)
Δ = -84,11 ±34ms
Δ = -1.24 ± 25ms
Δ = 0.29 ±1.38%
Δ = -0.95 ± 0.74%
iC / cI
(z=0,175, p=.8613)
(z=1.799, p=.0865)
(z=-1.376, p=.2026)
(z=-2.305, p=.0254)
Δ = 4,08 ± 23ms
Δ = 30.51 ±17ms
Δ = -2.55 ± 1.85%
Δ = -3.481 ±1.51%
iC / iI
(z=2.197, p=.034)
(z=4.374, p<.0001)
(z=-4.565, p<.0001)
(z=-5.176, p<.0001)
Δ = 67.52 ±31ms
Δ = 97.39 ± 22ms
Δ = -8.26 ±1.81%
Δ = -8.02 ±1.55%
cI / iI
(z=2.954, p=.006)
(z=4.532, p<.0001)
(z=-3.660, p=.0005)
(z=-4.071, p=.0001)
Δ = 63.43 ± 21ms
Δ = 66.88 ± 15ms
Δ = -5.70 ± 1.56%
Δ = -4.54 ±1.12%
