Consciousness and Cognition 26 (2014) 51–63

Contents lists available at ScienceDirect

Consciousness and Cognition

journal homepage: www.elsevier.com/locate/concog

Differential recruitment of executive resources during mind

wandering
Julia W.Y. Kam ⇑, Todd C. Handy
Department of Psychology, University of British Columbia, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Recent research has shown that mind wandering recruits executive resources away from
Received 13 October 2013 the external task towards inner thoughts. No studies however have determined whether
Available online 24 March 2014 executive functions are drawn away in a unitary manner during mind wandering episodes,
or whether there is variation in specific functions impacted. Accordingly, we examined
Keywords: whether mind wandering differentially modulates three core executive functions—
Mind wandering response inhibition, updating of working memory, and mental set shifting. In three exper-
Executive function
iments, participants performed one of these three executive function tasks and reported
Response inhibition
Information updating
their attentional state as either on-task or mind wandering at random intervals. We found
Working memory that mind wandering led to poorer performance in the response inhibition and working
Set-shifting memory tasks, but not the set-shifting task. These findings suggest that mind wandering
Task-related attention does not recruit executive functions in a monolithic manner. Rather, it appears to selec-
tively engage certain executive functions, which may reflect the adaptive maintenance of
ongoing task performance.
Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction

Whether fantasizing about vacation plans, worrying about an upcoming exam, or reliving an unforgettable event, we of-
ten times find ourselves engaged intensely in our own thoughts. This common experience, mind wandering, has been closely
tied to our transient emotional states and daily functioning. Importantly, research has begun to reveal its widespread impact
on the neurocognitive processing of external events. In addition to attenuated sensory and cognitive processing (Braboszcz &
Delorme, 2011; Kam et al., 2011; O’Connell et al., 2009; Smallwood, Beach, Schooler, & Handy, 2008), mind wandering has
also been associated with disrupted attentional orienting (Kam, Dao, Stanciulescu, Tildesley, & Handy, 2013) and perfor-
mance monitoring (Kam et al., 2012). Given the wide range of attenuation in sensory-motor responses, does mind wandering
also impair the higher-order processing of stimulus inputs, and in particular, the allocation of executive resources to the gi-
ven task at hand?
Aligning with this possibility, it has been proposed that when we mind wander, executive resources are drawn away from
the ongoing task to facilitate inner trains of thought (see Smallwood & Schooler, 2006 for a review). This is consistent with
the observation that internally-generated thoughts recruit the same domain-specific processes as stimulus in the external
world (e.g. Schooler et al., 2011; Smallwood, 2013; Smallwood & Schooler, 2006). Similarly, visual imagery (e.g., Kosslyn,
Thompson, & Alpert, 1997) and auditory imagery (e.g., Zatorre & Halpern, 2005) have been shown to activate many

⇑ Corresponding author. Address: Department of Psychology, 2136 West Mall, Vancouver, B.C. V6T 1Z4, Canada.
E-mail address: kamjulia@gmail.com (J.W.Y. Kam).

http://dx.doi.org/10.1016/j.concog.2014.03.002
1053-8100/Ó 2014 Elsevier Inc. All rights reserved.
52 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63

domain-specific regions of cortex that are also activated during interactions with the outside world. To the extent that task-
unrelated thoughts parallel mental imagery, then both lines of evidence indicate that externally and internally guided
thoughts can engage similar processes. Therefore, if executive resources are critical to maintaining internal thoughts, then
these resources become unavailable for processing external stimuli during periods of mind wandering.
Direct evidence in support for the executive function model of mind wandering comes from both behavioral and neuro-
imaging findings. For example, executive processing as indexed by performance on a random number generation task was
disrupted during task-independent thoughts (Teasdale et al., 1995), suggesting that mind wandering can recruit executive
resources that would normally be devoted to the external task. Moreover, mind wandering has been shown to recruit not
only the brain’s default mode network (DMN; Christoff, Gordon, Smallwood, Smith, & Schooler, 2009; Kirschner, Kam, Handy,
& Ward, 2012; Mason et al., 2007), but also the executive network as well (e.g., Christoff, 2012; Christoff et al., 2009). These
findings indicate that the facilitation of task-unrelated thoughts does indeed require executive functions.
Support for the executive function model notwithstanding, the term ‘‘executive functions’’ in previous studies of mind
wandering has been defined loosely, and it is unclear which facet of executive functions are specifically recruited during
mind wandering. To wit, executive functions are higher-level cognitive functions that regulate other thought and behavioral
processes (e.g. Alvarez & Emory, 2006; Miyake, Friedman, Emerson, Witzki, & Howerter, 2000). In particular, they are con-
sidered ‘‘general purpose control mechanisms that modulate the operation of various cognitive subprocesses and thereby
regulate the dynamics of human cognition’’ (Miyake et al., 2000, p. 50). According to the model by Miyake and colleagues
(2000), the three commonly postulated executive functions, inhibition of prepotent responses, updating of working memory
representations, and mental set-shifting, show some underlying commonality but are clearly distinct constructs.
Specifically, prepotent response inhibition concerns the deliberate inhibition of a dominant or automatic response (e.g.
Blasi et al., 2006; Mostofsky & Simmonds, 2008; Vendrell et al., 1995), and is considered to be the hallmark of executive con-
trol (e.g. Aron, 2007; Logan, 1994). This function has been examined beyond the experimental context and linked to perfor-
mance-based measures of daily functioning (Vaughan & Giovanello, 2010). In contrast, information updating pertains to
monitoring and evaluating the relevance of current stimulus inputs. It requires the active manipulation of content in working
memory to ensure that task-relevant information is enhanced while irrelevant inputs are suppressed (e.g. Baddeley, 1986;
Cohen et al., 1997; D’esposito et al., 1995; Morris & Jones, 1990; Smith & Jonides, 1997). This executive function is closely
related to working memory capacity, which has been shown to predict performance in laboratory-based and real-world cog-
nitive tasks (e.g. Engle, 2002; Kane et al., 2007; McVay & Kane, 2009). Finally, mental set-shifting refers to the ability to shift
from one task or operation to another (e.g. Monsell, 2003; Rogers & Monsell, 1995; Rubinstein, Meyer, & Evans, 2001).
Evidence suggests this involves overcoming active interference, especially when the new task relies on previously used stim-
uli (e.g. Allport & Wylie, 2000; Gilbert & Shallice, 2002). This cognitive flexibility to efficiently switch between tasks or men-
tal sets is considered to play a critical role in cognitive control (e.g. Monsell, 1996, 2003).
Given this understanding, the current study examined whether mind wandering engages all these executive functions in
a unitary or monolithic manner, or whether there can be a dissociation in which specific executive functions are actually
engaged during a mind wandering episode. In three separate experiments, participants performed one of three tasks that
required response inhibition, information updating or set-shifting. At frequent and random intervals throughout the task,
participants reported their attentional state at that moment as on-task or mind wandering. We then examined behavioral
performance as a function of whether it immediately preceded an on-task or mind wandering report. If mind wandering does
differentially recruit specific executive resources, and if these resources are presumably finite in that they cannot be utilized
to facilitate both mind wandering and task performance simultaneously, then tasks requiring deployment of these specific
executive functions should be disrupted during mind wandering.

2. Experiment 1

The first experiment examined the extent to which mind wandering impairs response inhibition. To address this question,
participants performed the Stroop task (Jensen & Rohwer, 1966; Macleod, 1991; MacLeod & MacDonald, 2000; Stroop, 1935),
in which they responded to the ink color that word stimuli were presented in, and reported their attentional states fre-
quently throughout the task. The words in this task were one of three colors, and were printed in one of those three colors.
Given that we tend to automatically respond to the meaning of the word (e.g. red), when that particular word is presented in
blue ink, we have to inhibit our dominant response of red, and accurately report it as blue instead. Previous studies have
found significantly slower reaction time to trials in which words were incongruent with the ink color they were presented
in compared to congruent or neutral trials (e.g. Ilan & Polich, 1999; van Maanen, van Rijn, & Borst, 2009). We hypothesized
that mind wandering relative to on-task states would further slow down reaction time to incongruent trials.

2.1. Methods

2.1.1. Participants
46 undergraduate students participated in the experiment in exchange for extra course credits, however six were ex-
cluded due to insufficient reports of a particular attention state (<15%), and three were excluded due to technical errors.
In the end, of the 37 participants included in our analyses (28 females; mean age = 19.75 years, S.D. = 2.83), 31 were right
 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63 53

handed, and all had normal or corrected-to-normal vision. Participants provided written informed consent to the experimen-
tal procedure, which was approved by the UBC Behavioral Review Ethics Board.

2.1.2. Stimuli and procedure

To assess response inhibition, participants performed the Stroop task (Bettner, Jarvik, & Blum, 1971; Jensen & Rohwer,
1966; Stroop, 1935; Stuss, Floden, Alexander, Levine, & Katz, 2001), and were instructed to respond to the ink color of the
word via three separate key presses corresponding to the three colors as quickly and accurately as possible. The task para-
digm is shown in Fig. 1A. Stimuli were presented on an 18 in. color monitor, with a viewing distance of approximately 50 cm.
They consisted of three color-words (i.e. blue, red, and green) and ‘‘XXXX’’ presented in three possible colors (i.e. red, blue, or
green), Helvetica style, size 44. In this task, a fixation dot first appeared in the center of the screen for 500 ms. Each trial be-
gan with a word presented at fixation for 200 ms, followed by a 1600 ms interval during which a response was recorded.
Marking the end of a trial, the inter-trial interval lasted approximately 200 ms, and was randomly jittered between 10
and 400 ms. The task itself consisted of 540 trials, with equal numbers of incongruent, congruent, and neutral trials. Incon-
gruent trials consisted of words printed in a color not denoted by the color name, whereas congruent trials were those in
which the name of the color matched the color of the ink. Neutral trials consisted of ‘‘XXXX’’ printed in one of the three afore-
mentioned colors. Each combination of stimulus and ink color had an equal probability of occurring.
To assess attention-related errors experienced in our daily lives, participants were also asked to complete the Attention-
Related Cognitive Errors Scale (ARCES; Cheyne, Carriere, & Smilek, 2006). With this additional, secondary measure, we
wanted to determine whether errors committed outside of the laboratory differentially relate to errors committed in exper-
imental tasks tapping into three distinct types of executive functions. The ARCES is a 12-item measure of cognitive errors we
commit in our daily life as a result of not paying sufficient attention to the ongoing task. It has satisfactory psychometric
properties and internal consistency (see Cheyne et al., 2006). Each testing session lasted approximately 45 min.

2.1.3. Task-related attention

Participants were instructed to report their attentional state as either being on-task or mind wandering at the end of each
trial block (c.f. Kam et al., 2011, 2012, 2013; Kirschner et al., 2012; Smallwood et al., 2008). Before the testing session, they
were provided with definitions of these two attention states: on-task states were defined as when one’s attention was firmly
directed towards the task, whereas mind wandering states were described as when one’s attention was not focused on the
task. Participants reported their attentional state via key presses at the end of each block. We then categorized the behavioral
data based on whether they preceded on-task or mind wandering states. The block duration itself was randomly varied be-
tween 44 and 64 s in order to minimize predictability of block completion and maximize variability of attentional state at the
time of block completion. There were a total of 20 blocks.
We used a direct measure of attentional state, namely experience sampling (e.g., Schooler et al., 2011; Smallwood &
Schooler, 2006), which relies on participants’ ability to reliably report whether their thoughts were focused on the task-
at-hand or on task-irrelevant issues. By using the attention report to categorize a participant’s attentional state in the

Fig. 1. Task paradigms. The different trial types in the Stroop, 1-back and Number–Letter tasks are shown in A, B and C respectively.
54 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63

10–15 s immediately prior to the report, the methodology has been used to demonstrate reliable and replicable differences
in neurocognitive functioning between ‘‘on-task’’ and ‘‘mind wandering’’ states (e.g., Christoff et al., 2009; Kam et al., 2011;
Kirschner et al., 2012; Mason et al., 2007; Smallwood et al., 2008; Stawarczyk, Majerus, Maquet, & D’Argembeau, 2011). As
such, our approach to defining attentional states aligned with widely-accepted norms in the field of mind wandering
research.

2.1.4. Statistical analysis

Performance on a Stroop task was assessed by comparing reaction time (RT) in neutral trials and incongruent trials. Faster
RTs in neutral relative to incongruent trials indicate a semantic interference, which suggests that the meaning of the word is
the cause of the interference. Therefore, our primary measure is RT, but we report accuracy rates as well. Behavioral data for
both conditions were based on averaging together the RT and accuracy data for the 6 targets, which occurred approximately
within 12 s preceding each of the two attentional state reports (on-task and mind wandering). The number of events in-
cluded in the averages was an attempt to maximize the number of events per condition without extending the window
so far back in time as to consistently capture another attentional state or transition between states. This particular time win-
dow was based on the finding that attentional states ebb and flow approximately at the rate of 10–15 s (Sonuga-Barke &
Castellanos, 2007). Importantly, both neural and behavioral measures using this time window have been shown to system-
atically fluctuate with the reported attentional state (Christoff et al., 2009; Kam et al., 2011, 2012, 2013; Smallwood et al.,
2008), highlighting the validity of the experience sampling method and choice of time window.
To determine whether our tendency to commit cognitive-related errors in our daily lives is associated with task perfor-
mance and mind wandering frequency in an laboratory setting, we correlated ARCES with proportion of mind wandering
reports, and accuracy measures on the Stroop task.

2.2. Results

2.2.1. Attention reports

Across the 20 experimental blocks, 57.2% ended with an ‘‘on-task’’ report and 42.8% ended with a ‘‘mind wandering’’ re-
port (S.D. = 20.0).

2.2.2. Reaction time

The RT and accuracy data are shown in Fig. 2. RTs below 200 ms or exceeding three standard deviations from the mean
were excluded from further analysis. Moreover, only correct trials were included in the RT analysis. We first compared the
RTs in neutral and incongruent trials as a function of task-related attention. An ANOVA with trial type (incongruent vs. neu-
tral) and task-related attention (on-task vs. mind wandering) as within-subject factors yielded an interaction that only ap-
proached significance (F(36) = 2.93, p = .095).
To directly test our hypothesis that mind wandering relative to on-task states would lead to slower RT in incongruent
trials, we conducted a priori paired-samples t-tests separately for these conditions. We observed that RTs were slower when
in mind wandering states relative to on-task states, but only for incongruent trials. This was confirmed statistically by a sig-
nificant main effect of task-related attention for incongruent trials (t(36) = 4.09; p < .001; effect size r = .183), with faster
RTs during on-task vs. mind wandering states, but not for neutral trials (t(36) = .70; p = .487; effect size r = .044).

2.2.3. Accuracy
We then examined differences in accuracy rates for incongruent and neutral trials preceding an on-task vs. mind wander-
ing report. We found a trend for the effect of condition (F(1, 36) = 2.94; p = .095), and a significant main effect of task-related
attention (F(1, 36) = 17.26; p < .001), with higher accuracy rates during on-task states. However, the condition by task-related
attention interaction was not significant (F(1, 36) = .21; p = .654).

2.2.4. Daily cognitive errors

The mean ARCES score in this sample correlated significantly with accuracy rates in incongruent trials immediately pre-
ceding mind wandering reports (r(37) = .45, p = .005), suggesting higher levels of cognitive-related errors experienced in
daily lives were associated with lower accuracy rates in the more difficult trials of the Stroop task during mind wandering.
The mean ARCES score did not correlate significantly with any other accuracy measures on the Stroop task (p’s > .600), nor
did it correlate with proportion of mind wandering reports (r(37) = .16, p = .353).

2.3. Discussion

We found that RTs for incongruent trials were quicker during on-task relative to mind wandering states, a difference that
was absent for neutral trials. That this distinction was only observed in RT, but not accuracy, is consistent with the obser-
vation that RT is a more sensitive measure in the Stroop task. These results indicate that mind wandering indeed disrupts
performance on a response inhibition task, but exclusively on trials that actually require engagement of response inhibition.
Two issues arise in turn.
 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63 55

Fig. 2. Behavioral performance on Stroop task in Experiment 1. (A) Reaction times and (B) accuracy rates for neutral and incongruent conditions are shown
as a function of whether they preceded on-task (OT) and mind wandering (MW) reports. p < .05.

First, the lack of significant difference between attention states in the neutral trials may explain similar findings previ-
ously observed in different tasks. For example, no difference in RT was observed between on-task and mind wandering states
in the sustained-attention-to-response task (SART; Kam et al., 2011). A modified version of the Go–NoGo task, the SART re-
quires one to respond to a frequently occurring numbers via button press, and withhold response to an infrequently occur-
ring letter ‘‘X’’. While successful performance in this task requires sustained attention, it is generally not considered as
cognitively demanding especially in the context of executive function and can quickly become repetitive. Notably, all the
trials included in the averaged RT preceding attentional state reports were frequent stimuli. To the extent that the frequently
occurring trials in the SART are comparable to the neutral trials in the Stroop in terms of their low levels of cognitive de-
mands, then the lack of attention effect would be corroborated using different task paradigms.
Second, the finding that mind wandering impairs response inhibition is in line with a previous report (Smallwood, McSp-
adden, & Schooler, 2007). Specifically, they found poor response inhibition in a modified Go/NoGo task only when mind wan-
dering occurred without awareness, but not when participants were aware of their mind wandering state (Smallwood et al.,
2007). Given that the current study did not measure participants’ awareness of their attentional state however, we cannot
definitively determine the role of meta-awareness in our findings. What we can conclude based on this experiment is that
mind wandering does appear to lead to poorer response inhibition. Together, these findings support the executive function
theory of mind wandering (Smallwood & Schooler, 2006) in that executive resources, and in this case the capacity for re-
sponse inhibition, are directed away from external environment during mind wandering.

3. Experiment 2

Another facet of executive function, is updating information in working memory, and is closely associated with working
memory capacity, which has been examined in the context of mind wandering (e.g. Kane & McVay, 2012; Kane et al., 2007;
McVay & Kane, 2009). According to the executive control failure model, mind wandering as a form of distraction from the
task-at-hand reflects a failure in working memory to restrict attention to task-relevant stimulus (Kane & McVay, 2012;
McVay & Kane, 2009). Interestingly, this model focuses primarily on the extent to which working memory capacity influ-
ences the frequency of mind wandering, and its impact on task performance. It does not make any assumptions about
56 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63

the working memory resources necessary to maintain internal trains of thought, nor the extent to which mind wandering
can engage these resources which can then impact external tasks requiring working memory.
In contrast, the decoupling model of mind wandering proposed that executive control processes are recruited when the
mind is focused on inner thoughts (Smallwood, 2013). Specifically, in line with the executive function model of mind wan-
dering (Smallwood & Schooler, 2006), executive control, and in particular, working memory resources are generally thought
to be engaged in order to maintain attentional focus on the relevant stream of information, whether that stream is internal or
external. In the context of mind wandering then, working memory resources may be used to suppress irrelevant, external
information in order to sustain the internal trains of thought. This theory is supported by recent empirical evidence suggest-
ing that working memory facilitates the continuity of mind wandering (Levinson, Smallwood, & Dvaidson, 2012). As such,
the decoupling hypothesis would thus predict that working memory resources would be recruited during mind wandering,
thereby leading to impaired performance on an external working memory task.
In fact, several recent studies provided evidence that supports this prediction. For example, Mrazek and colleagues (2012)
reported an association between higher levels of mind wandering and impaired performance on the operation span (OSPAN)
task, a relationship that occurred regardless of the difficulty of task conditions. Similarly, a later study found that mind wan-
dering mediated the beneficial effects of mindfulness training on OSPAN task performance among those individuals who
were prone to mind wandering (Mrazek, Franklin, Phillips, Baird, & Schooler, 2013). While both studies reported effects
of mind wandering on working memory, successful performance on the specific task used in these studies (i.e.: OSPAN) re-
quired individuals to maintain a list of items in their working memory over a period of time after which they were asked to
recall the items in the presented order. This capacity to maintain information over a relatively prolonged time period in
working memory stands in contrast to the trial-by-trial updating of information in working memory as described by Miyake
and colleagues (2000). Accordingly, the direct effect of mind wandering on such constant updating of working memory rep-
resentations, as measured using the N-back task (Kirchner, 1958), has yet to be tested.
In Experiment 2, we thus examined the behavioral response to trials in the 0-back and 1-back conditions of the N-back
task as a function of whether or not participants were in a mind wandering state. The varying difficulty levels between the
conditions are manifested in accuracy rates, such that higher n-back conditions were associated with lower accuracy (e.g.
Carlson et al., 1998; McElree, 2001). Therefore our primary measure is accuracy rate, but we report RT as well. If working
memory is in fact engaged during mind wandering then it is predicted that mind wandering should be associated with lower
accuracy rates.

3.1. Methods

3.1.1. Participants
42 undergraduate students participated in the experiment in exchange for extra course credits, however four were ex-
cluded due to insufficient reports of a particular attention state (<15%). A total of 38 participants were included in our anal-
yses (20 females; mean age = 19.55 years, S.D. = 1.78); all were right handed, and had normal or corrected-to-normal vision.
Participants provided written informed consent to the experimental procedure, which was approved by the UBC Behavioral
Review Ethics Board.

3.1.2. Stimuli and procedure

As with the two conditions in Experiment 1, we implemented two conditions of the N-back task: one that required updat-
ing of content in working memory (i.e. 1-back) and another one that did not (i.e. 0-back) as a source of comparison. Specif-
ically, the 0-back task requires participants to press one key for a particular letter ‘‘Z’’, and another key for all other letters.
While participants have to maintain the target letter in working memory throughout this task to perform accurately, the tar-
get letter need not be updated on a trial-by-trial basis. In contrast, the1-back task requires participants to evaluate whether
the current letter matches the last letter presented. They were instructed to press one key for a match response and another
key for a non-match response. To excel in this latter task requires one to continuously update working memory of the letter
last presented.
Participants performed two conditions of the N-back task (Kirchner, 1958). In both the 0-back and 1-back conditions, each
letter had an equal probability of occurring, with the constraint that target ‘‘Z’’ occurred 33% of the time in the 0-back con-
dition. The response mappings for both tasks were counter-balanced across subjects. Due to the nature of this task, the two
conditions have to be presented in a block design. There were 12 blocks per condition, and the order of presentation of each
condition was counter-balanced across subjects.
The task paradigm is shown in Fig. 1B. Stimuli were presented on an 18 in. color monitor, with a viewing distance of
approximately 50 cm. They were letters of the alphabet (excluding vowels and X), presented in black, Helvetica style, size
96. Each trial started with a letter presented at fixation for 300 ms, followed by a 1500 ms interval during which a response
was recorded. Following this was the inter-trial interval that lasted approximately 200 ms, and was randomly jittered be-
tween 10 and 400 ms. Participants completed a total of 360 trials per condition. Following this task, participants were asked
to complete the ARCES (Cheyne et al., 2006). Each testing session lasted approximately 45 min.
 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63 57

3.1.3. Task-related attention and statistical analysis

Our measure of task-related attention is identical to the methods used in Experiment 1, with the following exceptions.
Each block consisted of 25–35 trials, which lasted approximately 50–70 s. The duration of each block varied across experi-
ments as a result of slight differences in trial durations and total number of trials. We conducted an omnibus ANOVA that
included trial type (match vs. non-match), condition (0-back vs. 1-back) and task-related attention (on-task vs. mind wan-
dering) as within-subject factors. The behavioral data for both conditions were based on averaging together the RT and accu-
racy data for the 6 targets, which occurred approximately within 12 s, preceding each of the two attentional state reports.
We correlated ARCES with proportion of mind wandering reports, and accuracy measures on the N-back task.

3.2. Results

3.2.1. Attention reports

In the 0-back condition, 46% ended with an ‘‘on-task’’ report and 54% ended with a ‘‘mind wandering’’ report (S.D. = 20.6).
In the 1-back condition, 53% ended with an ‘‘on-task’’ report and 47% ended with a ‘‘mind wandering’’ report (S.D. = 20.0).
The proportion of on-task reports was significantly greater in the 1-back condition (t(37) = 2.18, p = .036), validating the
notion that this condition required more cognitive effort thereby leading to fewer mind wandering reports.

3.2.2. Reaction time

RT and accuracy data for Experiment 2 are shown in Fig. 3. RTs below 200 ms or exceeding three standard deviations from
the mean were excluded from further analysis. Only correct trials were included in the RT analysis. We found a significant
main effect of condition (F(1, 37) = 14.60; p < .001), with faster RT in 0-back compared to 1-back condition, as well as a sig-
nificant effect of task-related attention (F(1, 37) = 9.98; p = .003), with faster RT during on-task states relative to mind wan-
dering states. The condition by task-related attention interaction was however not significant (F(1, 37) = .65; p = .424).

3.2.3. Accuracy
We then examined whether accuracy rates for 0-back and 1-back trials were differentially modulated by attention states.
As with the RT data, there were significant main effects of condition (F(1, 37) = 13.56; p = .001), and task-related attention
(F(1, 37) = 62.61; p < .001). Importantly, there was a significant interaction between condition and task-related attention

Fig. 3. Behavioral performance on N-back task in Experiment 2. (A) Reaction times and (B) accuracy rates for 0-back and 1-back conditions are shown as a
function of on-task (OT) and mind wandering (MW) states. p < .05.
58 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63

(F(1, 37) = 5.67; p = .023). Follow-up analyses revealed that accuracy rates were significantly lower during mind wandering
states in both the 0-back condition (F(1, 37) = 27.27; p < .001) and 1-back condition (F(1, 37) = 43.07; p < .001). What appears
to contribute to the significant interaction is the magnitude of the effect of task-related attention between the two condi-
tions. Specifically, the effect size in the 0-back condition (effect size r = .48) is smaller than that in the 1-back condition (ef-
fect size r = .56).
In light of these results, we conducted a second pair of comparisons as an attempt to further elucidate the attention by
condition interaction. We found that accuracy rates were significantly higher in the 0-back condition relative to 1-back con-
dition during mind wandering periods (F(1, 37) = 11.66; p = .002). However, this difference was only approached significance
during on-task periods (F(1, 37) = 3.37; p = .074).

3.2.4. Daily cognitive errors

The average ARCES score did not correlate with proportions of mind wandering reports nor any accuracy measures.

3.3. Discussion

Confirming our hypothesis, our findings from Experiment 2 revealed that mind wandering does in fact impair working
memory updating. Specifically, we found lower accuracy rates during mind wandering relative to on-task states, and this
effect was more pronounced for the more difficult condition (i.e. 1-back). These findings indicate that mind wandering dis-
rupts task performance when the particular task requires updating of information in working memory.
Of relevance, this study adds to the ongoing debate regarding working memory and mind wandering (e.g. Kane & McVay,
2012; McVay & Kane, 2009; Smallwood, 2013; Smallwood & Schooler, 2006). In particular, we found mind wandering was
associated with disrupted N-back performance, suggesting that working memory may be recruited during mind wandering.
These results are therefore in line with the recent finding that maintaining an internal train of thoughts requires working
memory resources (Levinson et al., 2012). While our findings appear to be inconsistent with the view that mind wandering
maintenance is resource-free (e.g. McVay & Kane, 2009), these authors’ individual differences approach deserves recognition
as it can provide important insight into current issues in mind wandering. As an example, it can allow us to determine
whether there are variations across individuals in the severity of impairments in stimulus processing during mind wander-
ing. For now, averaging across participants in our study showed that working memory resources were presumably recruited
away from the external task in order to facilitate the maintenance of ongoing thoughts. Critically, trials that were not as cog-
nitively demanding and presumably did not require working memory resources were less susceptible to modulations by
mind wandering.

4. Experiment 3

In Experiment 3, we focused on the third facet of executive function, mental set-shifting, and examined the extent to
which it is disrupted by mind wandering. Mental set-shifting has been closely tied to cognitive control, which involves
the representation of two sets of rules and switching between them that specify task-relevant stimulus–response associa-
tions (e.g. Monsell, 1996, 2003). Interestingly, when shifting between tasks, reaction times tend to increase, presumably
due to the need to re-configure one’s task set (e.g. Monsell, 2003; Rogers & Monsell, 1995; Rubinstein et al., 2001). Based
on previous literature and our two experiments, we predicted that mind wandering should exclusively modulate the reac-
tion time performance in trials requiring switching between task rules.
There were two conditions of the number–letter task: one that required switching between task rules (i.e. switch) and
one that did not (i.e. non-switch). The switch trials required participants to shift from evaluating letters to evaluating num-
bers (e.g. letter–number), or vice versa. On the contrary, during the non-switch trials, participants made the same evaluation
as the last trial (e.g. letter–letter). As such, while other cognitive processes may be involved in accurately responding to the
number or letter, the sole difference between the two conditions is in whether the switching of task rules is required.

4.1. Methods

4.1.1. Participants
42 undergraduate students participated in the experiment in exchange for extra course credits, but four were excluded
due to insufficient reports of a particular attention state (<15%). In the end, of the 38 participants included in our analyses (32
females; mean age = 20.74 years, S.D. = 2.56), 35 were right handed, and all had normal or corrected-to-normal vision. Par-
ticipants provided written informed consent to the experimental procedure, which was approved by the UBC Behavioral Re-
view Ethics Board.

4.1.2. Stimuli and procedure

Participants performed the number–letter task (Rogers & Monsell, 1995), where they were instructed to respond accord-
ing to two sets of rules: (1) determine whether the letter is a vowel (i.e. A, I, U, E) or a consonant (i.e. K, R, M, G), and (2)
determine whether the number is even (i.e. 2, 4, 6, 8) or odd (i.e. 1, 3, 5, 7). Stimuli were presented on an 18 in. color monitor,
 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63 59

with a viewing distance of approximately 50 cm. Each stimulus was made up of a symbol (i.e. $,%, &,?) paired with a number
or letter, presented in black, Helvetica style, size 54. Only letters occurred in the top left or right quadrants, and only num-
bers occurred in the bottom left or right quadrants. All possible letter–symbol combinations (e.g. A%, $K) and number–sym-
bol combinations (e.g. $3, &8) had an equal probability of occurring in the top and bottom quadrants respectively. The task
paradigm is shown in Fig. 1C.
In this task, a fixation dot first appeared in the center of the screen for 500 ms. Each trial began with a stimulus presented
at fixation for 400 ms. Participants had up to 1300 ms to respond, after which the trial would time out. Immediately follow-
ing the response was the inter-trial interval, which lasted approximately 200 ms and was randomly jittered between 100 and
300 ms. Unlike the Stroop and N-back tasks, the duration of each trial was therefore mainly dependent upon the response
time of the participant.
The task consisted of 1536 trials, with one pair of stimulus presented at a time, in a clock-wise rotating manner. Trials in
the top quadrants required participants to evaluate whether the letter was a vowel or consonant, whereas trials in the bot-
tom quadrants required participants to evaluate whether the number was even or odd. Accordingly, trials occurring in the
top right and bottom left quadrants did not require a switch in response. In contrast, stimulus presented in the top left quad-
rant required participants to switch response from numbers to letters, and stimulus in the bottom right quadrant required
switching from letters to numbers. There were equal numbers of switch and non-switch trials, which were presented in
alternating manner in clockwise rotation. After the task, participants were asked to complete ARCES (Cheyne et al., 2006).
Each testing session lasted approximately 45 min.

4.1.3. Task-related Attention and Statistical Analysis

Our measure of task-related attention is identical to the methods used in Experiment 1 and 2 with the following excep-
tions. There were 26 experimental blocks; each block consisted of 20–60 trials, which lasted approximately 24–72 s. We con-
ducted an omnibus ANOVA that included condition (non-switch vs. switch) and task-related attention (on-task vs. mind
wandering) as within-subject factors. The behavioral data for both conditions were based on averaging together the RT
and accuracy data for the 10 targets, which occurred approximately within 12 s preceding each of the two attentional state
reports (on-task and mind wandering).

Fig. 4. Behavioral performance on Number–Letter task in Experiment 3. (A) Reaction times and (B) accuracy rates for non-switch and switch conditions are
shown as a function of whether they preceded on-task (OT) and mind wandering (MW) reports.
60 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63

4.2. Results

4.2.1. Attention reports

Across the experimental blocks, 48% ended with an on-task report and 52% ended with a mind wandering report
(S.D. = 14.6).

4.2.2. Reaction time

The RT and accuracy data for Experiment 3 are shown in Fig. 4. RTs below 200 ms or exceeding three standard deviations
from the mean were excluded from further analysis. Only correct trials were included in the RT analysis. We examined
whether mind wandering differentially modulated RT on switch vs. no-switch trials. While the main effect of condition
was significant (F(1, 37) = 93.89, p < .001), with faster RT in non-switch trials relative to switch trials, the task-related atten-
tion effect was not significant (F(1, 37) = .36, p = .554). The condition by task-related attention interaction was also not sig-
nificant (F(1, 37) = 1.57, p = .219).

4.2.3. Accuracy
With respect to accuracy rates, we found a significant main effect of condition (F(1, 37) = 21.08, p < .001), with higher
accuracy rates in non-switch trials, as well as a significant effect of task-related attention (F(1, 37) = 47.80, p < .001), with
higher accuracy rates during on-task states. However, the interaction between condition and task-related attention was
not significant (F(1, 37) = .22, p = .642).

4.2.4. Daily cognitive errors

The mean ARCES score was significantly positively correlated with percentage of mind wandering reports (r(38) = .35,
p = .029), suggesting those who reported mind wandering more frequently during the task also committed more cogni-
tive-related errors in their daily lives.

4.2.5. Control experiment

In both behavioral measures of set-shifting, there was no interaction between task-related attention and condition. Thus,
contrary to our hypothesis, we did not observe faster reaction times during switch trials while mind wandering relative to
on-task states. A potential explanation for this lack of finding is the difficulty of the number–letter task. The trials progressed
as soon as a response was made, and this increased the speed of the task compared to the Stroop or N-back tasks. The swift-
ness of going from switch trials to non-switch trials back to switch trials may have increased the overall difficulty of the task.
This alternating pattern between switch and non-switch trials requires a shift in and of itself. In fact, previous studies have
considered both of these trial types to belong to ‘switch’ trials.
As a result, we ran a control experiment wherein the task started with only non-switch trials appearing in the first half of
the experiment, and switch trials occurring in the last half of the experiment. We implemented this block design to ensure a
purer measure of non-switch trials. The procedure of this control experiment was identical to the previous number–letter
task, with a few exceptions. The first half of the experiment only consisted of non-switch trials, with the first 25% of stimuli
presented only in the top quadrants, and the next 25% of stimuli presented only in the bottom quadrants. The second half of
the experiment consisted of switch-trials, wherein stimuli were presented in clockwise rotation across all 4 quadrants. Fur-
ther, each of the 26 experimental blocks consisted of 45–75 trials, which lasted approximately 45–75 s.
Across the experimental blocks, 49% ended with an on-task report and 51% ended with a mind wandering report
(S.D. = 18.3). We conducted the same ANOVAs examining behavioral performance as a function of task-related attention.
Notably, we still did not observe an interaction between attention and condition in reaction time (F(1, 21) = .50, p = .486)
or accuracy (F(1, 21) = .23, p = .639), suggesting that the lack of interaction in the original version of the task was not simply
due to the contamination of switch trials in non-switch trials.

4.3. Discussion

In the original and modified versions of the number–letter task, we found no significant interaction between task-related
attention and condition in both behavioral measures of set-shifting. Thus, contrary to our hypothesis, we did not observe
faster reaction times during switch trials while mind wandering relative to on-task states.
A potential explanation for this lack of finding is the difficulty of the number–letter task. Given that we implemented a
blocked design in the control task, this version would presumably be easier as switch and non-switch trials were performed
separately. Importantly, the purer measure of non-switch trials here would more closely parallel the easier conditions in the
Stroop and N-back tasks, such that they do not require the executive function of interest. Yet, we also did not observe any
interaction effects involving task-related attention in this easier version. It is still possible that this null effect of attention is
specific to the number–letter task, and other set-shifting tasks may show attentional modulations. Based on our findings
however, it appears that mind wandering does not engage mental set-shifting as measured by the number–letter task, at
least not to the same extent as was found for response inhibition and working memory updating.
 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63 61

5. General discussion

The present study examined whether three facets of executive functions were differentially engaged during internally ori-
ented thoughts. In three experiments, we found that mind wandering disrupted behavioral performance on response inhi-
bition and updating of working memory tasks, but not the task that required mental set-shifting. Interestingly, mind
wandering led to poorer performance exclusively in trials that required response inhibition and information updating,
but not less demanding trials in the same task. These findings suggest that mind wandering does not monolithically recruit
all forms of executive functions as measured by the Stroop, N-back and Number–Letter tasks, highlighting the dissociation in
executive functions engaged during mind wandering. We discuss the broader implications of all three experiments below.
Why might we observe disruptive effects of mind wandering on response inhibition and information updating, but not
set-shifting tasks? One possibility may concern the relationship between these three facets and the extent to which each
of them represent ‘‘executive functions’’ as denoted in other experiments. In particular, set-shifting as a latent variable
showed the weakest correlations with other latent variables reflecting response inhibition and information updating in
working memory (Miyake et al., 2000). This raises the possibility that these latter two executive functions, relative to set-
shifting, may be more representative of executive functions as generally referred to in other studies. In support of this spec-
ulation, mind wandering has been shown to interrupt performance on a random number generation task, which required
participants to produce sequences of random numbers (Teasdale et al., 1995). Interestingly, both response inhibition and
working memory, but not set-shifting, have been shown to be independently related to performance on the random number
generation task (e.g. Miyake et al., 2000).
Another reason may be that mental set shifting is often studied in the context of switching from one task or rule to an-
other, both of which are usually relevant to the ongoing task-at-hand. However, switching from an external task to a task-
irrelevant internal train of thought may be a distinct form of shifting. This particular type of shifting relates to the functions
of the Frontal Parietal Network (FPN), which was recently proposed to be responsible for switching between externally and
internally oriented networks (Smallwood, 2013; Spreng, Stevens, Chamberlain, Gilmore, & Schacter, 2010). The FPN was sug-
gested to play a super-ordinate role that oversees and modulates the focus of attention (e.g. Shomstein, Kravitz, & Behrmann,
2012; Smallwood, Brown, Baird, & Schooler, 2012). To the extent that mental set shifting extends beyond two task-relevant
externally-oriented rules to this higher level switching between different streams of inputs, then perhaps its functioning
supersedes modulations by task-related attention.
An alternative explanation for this dissociation in mind wandering’s effect on executive functions rests in the difficulty
level of these tasks, rather than the types of executive resource they recruit. Specifically, it is possible that set-shifting is less
difficult or resource-demanding relative to response inhibition and working memory updating, allowing one to maintain an
internal stream of thought while simultaneously performing the task. However, two points argue against this interpretation.
First, in terms of behavioral performance, the overall highest accuracy rates and fastest reaction times in our data were not
found in the set-shifting task but in the working memory task instead, suggesting that the set-shifting task was not the eas-
iest. Second, the Dynamica Filtering Theory of Prefrontal Functioning (Shimamura, 2000) would suggest the opposite con-
clusion – that set-shifting is more demanding. Specifically, in his theory, executive functioning encompasses four basic
cognitive processes, including selecting, maintaining, updating and re-routing, which maps onto response inhibition, updat-
ing and set-shifting in Miyake’s model (2000). Shimamura proposed that these four basic cognitive processes are arranged by
complexity level, such that re-routing (or set-shifting) is the most resource-demanding process (2000). Taken together, the
relative difficulty level of our three tasks appears to be an unlikely explanation for the null effect of mind wandering on the
set-shifting task.
These three studies together provide partial evidence supporting the executive function model of mind wandering. This
model suggests that when mind wandering, our executive functions are directed away from the primary task towards our
internal thoughts (Smallwood, 2013; Smallwood & Schooler, 2006). In the present study, it would appear not all executive
functions are recruited away, but instead response inhibition and working memory are more subject to modulation by mind
wandering states compared to set-shifting. Thus far, accumulating evidence has shown that mind wandering disrupts a wide
range of neurocognitive processes, including sensory processing, attentional orienting and performance monitoring (Bra-
boszcz & Delorme, 2011; Kam et al., 2011, 2012, 2013; O’Connell et al., 2009). Given these processes are regulated by exec-
utive functions, our study extends the existing literature and reveals mind wandering’s differential impact on executive
functions. As few studies have directly examined the effects of mind wandering on executive functions however, more work
will be needed to determine whether our findings were task-dependent. Importantly however, the dissociation among re-
sponse inhibition, information updating, and set-shifting observed in our study supports Miyake’s (2000) theory and findings
that they represent three distinguishable facets of executive functions.
Thus far, we discussed mind wandering’s impact on executive functions; however, our findings suggest that the relation-
ship between attentional states and executive functions may be bidirectional. On one hand, mind wandering appears to dis-
rupt performance on specific executive function tasks. This interpretation is consistent with the robust finding that mind
wandering systematically leads to changes in sensory and cognitive level processing (e.g. Braboszcz & Delorme, 2011;
Kam et al., 2011; O’Connell et al., 2009; Smallwood et al., 2008). On the other hand, the difficulty level of these tasks appears
to have an impact on how frequently one engages in mind wandering. For example, in Experiment 2, participants reported
lower levels of mind wandering in the more difficult condition, as would be predicted by the executive function model
62 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63

(Smallwood & Schooler, 2006). Notably, while a bidirectional relationship may exist, the manner in which mind wandering
affects task performance is different from the manner in which the nature of the task affects mind wandering frequency.
Another issue concerns the differential relationships between our measure of attention-related errors in the real world
and performance on the three executive function tasks. Specifically, while ARCES scores correlated with accuracy rates dur-
ing mind wandering in the response inhibition task, they were associated with frequency of mind wandering reports in the
set-shifting task, and not associated with either measures in the working memory task. This pattern of correlations is to
some extent consistent with our main finding of the dissociation in executive functions engaged during mind wandering.
Nonetheless, the relatively small sample size in our experiments may have yielded less stable estimates of these relation-
ships. Additional work is therefore necessary to definitively determine the mechanisms underlying the attention-related er-
rors experienced in the real world.
The dissociation in the effects of mind wandering on different facets of executive functions raises additional questions for
future investigations. For one, what is the temporal dynamics of mind wandering’s impact on stimulus processing? Mind
wandering is considered to be part of slower fluctuations in attentional states over longer periods of time. However, our re-
sults revealed that it can differentially modulate depth of stimulus processing depending on the cognitive demands of the
individual trial in a transient manner. That is, its impact on depth of processing changes on a trial-by-trial basis. It is thus
currently unclear as to how mind wandering interacts with processing that occurs at a different timescale. Electrophysiology
may help shed light on the temporal characteristics of mind wandering and its impact on stimulus processing.
A second question concerns whether meta-awareness in our current focus of attention may differentially impact task per-
formance. In our series of experiments, we found that mind wandering differentially disrupted trials that engaged an exec-
utive function compared to trials that did not engage an executive function. If lack of awareness of one’s mind wandering
state is considered as a deeper engagement in mind wandering as has been proposed (e.g. Schooler, 2002; Smallwood
et al., 2007), then fewer resources would remain to process external information. Accordingly, mind wandering without
awareness would presumably lead to greater disruption in performance on tasks requiring executive functions, whereas
mind wandering with awareness may allow for shifting of resources back to the task on cognitively demanding trials. On
the other hand, awareness of attentional state may allow one to allocate resources to task when the task demands so. Future
studies are necessary to elucidate the role of meta-awareness in the effects of mind wandering on different types of exec-
utive processing. These future investigations would extend and complement the current findings of mind wandering’s dif-
ferential recruitment of the three core executive functions.

Acknowledgments

We want to sincerely thank Megan MacPherson, Jerry Lee, and Spencer Murch for their help with data collection, as well
as Matt Dixon for his comments on earlier versions of the manuscript.

References

Allport, A., & Wylie, G. (2000). Task switching, stimulus-response bindings, and negative priming. In S. Monsell & J. Driver (Eds.), Control of cognitive
processes: Attention and performance. Cambridge, MA: MIT Press. XVIII.
Alvarez, J. A., & Emory, E. (2006). Executive function and the frontal lobes: A meta-analytic review. Neuropsychology Review, 16, 17–42.
Aron, A. R. (2007). The neural basis of inhibition in cognitive control. Neuroscientist, 13, 214–228.
Baddeley, A. D. (1986). Working memory. Oxford: Oxford University Press.
Bettner, L. G., Jarvik, L. F., & Blum, J. E. (1971). Stroop color-word test, non-psychotic organic brain syndrome, and chromosome loss in aged twins. Journal of
Gerontology, 26, 458–469.
Blasi, G., Goldberg, T. E., Weickert, T., Das, S., Kohn, P., Zoltick, B., et al (2006). Brain regions underlying response inhibition and interference monitoring and
suppression. European Journal of Neuroscience, 23, 1658–1664.
Braboszcz, C., & Delorme, A. (2011). Lost in thoughts: Neural markers of low alertness during mind wandering. Neuroimage, 54, 3040–3047.
Carlson, S., Martinkauppi, S., Rama, P., Salli, E., Korvenoja, A., & Aronen, H. J. (1998). Distribution of cortical activation during visuospatial n-back tasks as
revealed by functional magnetic resonance imaging. Cerebral Cortex, 8, 743–752.
Cheyne, J. A., Carriere, J. S. A., & Smilek, D. (2006). Absent-mindedness: Lapses of conscious awareness and everyday cognitive failures. Consciousness and
Cognition, 15, 578–592.
Christoff, K. (2012). Undirected thought: Neural determinants and correlates. Brain Research, 1428, 51–59.
Christoff, K., Gordon, A. M., Smallwood, J., Smith, R., & Schooler, J. W. (2009). Experience sampling during fMRI reveals default network and executive system
contributions. Proceedings of the National Academy of Sciences, 106(21), 8719–8724.
Cohen, J. D., Perlstein, W. M., Braver, T. S., Nystrom, L. E., Noll, D. C., Jonides, J., et al (1997). Temporal dynamics of brain activation during a working memory
task. Nature, 386, 604–608.
D’esposito, M., Detre, J. A., Alsop, D. C., Shin, R. K., Atlas, S., & Grossman, M. (1995). Neural basis of the central executive system of working memory. Nature,
378, 279–281.
Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological Science, 11, 19–23.
Gilbert, S. J., & Shallice, T. (2002). Task switching: A PDP model. Cognitive Psychology, 44, 297–337.
Ilan, A. B., & Polich, J. (1999). P300 and response time from a manual Stroop task. Clinical Neurophysiology, 110, 367–373.
Jensen, A. R., & Rohwer, W. D. (1966). The Stroop color-word test: A review. Acta Psychologia, 25, 36–93.
Kam, J. W. Y., Dao, E., Blinn, P., Krigolson, O. E., Boyd, L. A., & Handy, T. C. (2012). Mind wandering and motor control: Off-task thinking disrupts the online
adjustment of behavior. Frontiers in Human Neuroscience, 6, 329. http://dx.doi.org/10.3389/fnhum.2012.00329.
Kam, J. W. Y., Dao, E., Farley, J., Fitzpatrick, K., Smallwood, J., et al (2011). Slow fluctuations in attentional control of sensory cortex. Journal of Cognitive
Neuroscience, 23, 460–470.
Kam, J. W. Y., Dao, E., Stanciulescu, M., Tildesley, H., & Handy, T. C. (2013). Mind wandering and the adaptive control of attentional resources. Journal of
Cognitive Neuroscience, 25, 952–960.
 J.W.Y. Kam, T.C. Handy / Consciousness and Cognition 26 (2014) 51–63 63

Kane, M. J., Brown, L. H., McVay, J. C., Silvia, P. J., Myin-Germeys, I., & Kwapil, T. R. (2007). For whom the mind wanders, and when. Psychological Science, 18,
614–621.
Kane, M. J., & McVay, J. C. (2012). What mind wandering reveals about executive-control abilities and failures. Psychological Science, 21, 348–354.
Kirchner, W. K. (1958). Age differences in short-term retention of rapidly changing information. Journal of Experimental Psychology, 55, 352–358.
Kirschner, A., Kam, J. W. Y., Handy, T. C., & Ward, L. M. (2012). Differential synchronization in default and task-specific networks of the human brain.
Frontiers in Human Neuroscience, 6, 139. http://dx.doi.org/10.3389/fnhum.2012.00139.
Kosslyn, S. M., Thompson, W. L., & Alpert, N. M. (1997). Neural systems shared by visual imagery and visual perception: A positron emission tomography
study. Neuroimage, 6, 320–334.
Levinson, D. B., Smallwood, J., & Dvaidson, R. J. (2012). The persistence of thought: Evidence for a role of working memory in the maintenance of task-
unrelated thinking. Psychological Science, 23, 375–380.
Logan, G. D. (1994). Spatial attention and the apprehension of spatial relations. Journal of Experimental Psychology: Human Perception and Performance, 20,
1015–1036.
MacLeod, C. M. (1991). Half a century of research on the Stroop effect: An integrative review. Psychological Bulletin, 109, 163–203.
MacLeod, C. M., & MacDonald, P. A. (2000). Interdimensional interference in the Stroop effect: Uncovering the cognitive and neural anatomy of attention.
Trends in Cognitive Sciences, 4, 383–391.
Mason, M. F., Norton, M. I., Van Horn, J. D., Wegner, D. M., Grafton, S. T., & Macrae, C. N. (2007). Wandering Minds: The default network and stimulus-
independent thought. Science, 315, 393–395.
McElree, B. (2001). Working memory and focal attention. Journal of Experimental Psychology: Learning, Memory and Cognition, 27, 817–835.
McVay, J. C., & Kane, M. J. (2009). Conducting the train of thought: Working memory capacity, goal neglect, and mind wandering in an executive-control
task. Journal of Experimental Psychology: Learning, Memory and Cognition, 35, 196–204.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, A. (2000). The unity and diversity of executive functions and their contributions to
complex ‘‘frontal lobe’’ tasks: A latent variable analysis. Cognitive Psychology, 41, 49–100.
Monsell, S. (1996). Control of mental processes. In V. Bruce (Ed.), Unsolved mysteries of the mind: Tutorial essays in cognition. Hove, UK: Erlbaum.
Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7, 134–140.
Morris, N., & Jones, D. M. (1990). Memory updating in working memory: The role of thecentral executive. British Journal of Psychology, 81, 111–121.
Mostofsky, S. H., & Simmonds, D. J. (2008). Response inhibition and response selection: Two sides of the same coin. Journal of Cognitive Neuroscience, 20,
751–761.
Mrazek, M. D., Franklin, M. S., Phillips, D. T., Baird, B., & Schooler, J. W. (2013). Mindfulness training improves working memory capacity & GRE performance
while reducing mind-wandering. Psychological Science, 24, 776–781.
Mrazek, M. D., Smallwood, J., Franklin, M. S., Chin, J. M., Baird, B., & Schooler, J. W. (2012). The role of mind-wandering in measurements of general aptitude.
Journal of Experimental Psychology: General, 141, 788–798.
O’Connell, R. G., Dockree, P. M., Robertson, I. H., Bellgrove, M. A., Foxe, J. J., et al (2009). Uncovering the neural signature of lapsing attention:
Electrophysiological signals predicts errors up to 20 s before they occur. Journal of Neuroscience, 29, 8604–8611.
Rogers, R. D., & Monsell, S. (1995). Costs of a predictable switch between simple cognitive tasks. Journal of Experimental Psychology: General, 124, 207–231.
Rubinstein, J. S., Meyer, D. E., & Evans, J. E. (2001). Executive control of cognitive processes in task switching. Journal of Experimental Psychology: Human
Perception and Performance, 27, 763–797.
Schooler, J. W. (2002). Re-representing consciousness: Dissociations between experience and meta-consciousness. Trends in Cognitive Science, 6, 339–344.
Schooler, J. W., Smallwood, J., Christoff, K., Handy, T., Reichle, E., & Sayette, M. A. (2011). Meta-awareness, perceptual decoupling and the wandering mind.
Trends in Cognitive Science, 15, 319–326.
Shimamura, A. (2000). The role of prefrontal cortex in dynamic filtering. Psychobiology, 28, 207–218.
Shomstein, S., Kravitz, D. J., & Behrmann, M. (2012). Attentional control: Temporal relationships within the fronto-parietal network. Neuropsychologia, 50,
1202–1210.
Smallwood, J. (2013). Distinguishing how from why the mind wanders: A process-occurrence framework for self-generated mental activity. Psychological
Bulletin, 139, 519–535.
Smallwood, J., Beach, E., Schooler, J. W., & Handy, T. C. (2008). Going AWOL in the brain: Mind wandering reduces cortical analysis of external events. Journal
of Cognitive Neuroscience, 20, 458–469.
Smallwood, J., Brown, K., Baird, B., & Schooler, J. W. (2012). Cooperation between the default mode network and frontal-parietal network in the production
of an internal train of thought. Brain Research, 1428, 60–70.
Smallwood, J., McSpadden, M., & Schooler, J. S. (2007). The lights are on but no one’s home: Meta-awareness and the decoupling of attention when the mind
wanders. Psyconomic Bulletin & Review, 14, 527–533.
Smallwood, J., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132, 946–958.
Smith, E. E., & Jonides, J. (1997). Working memory: A view from neuroimaging. Cognitive Psychology, 33, 5–42.
Sonuga-Barke, E. J. S., & Castellanos, F. X. (2007). Spontaneous attentional fluctuations in impaired states and pathological conditions: A neurobiological
hypothesis. Neuroscience & Biobehavioral Reviews, 31, 977–986.
Spreng, R. N., Stevens, W. D., Chamberlain, J. P., Gilmore, A. W., & Schacter, D. L. (2010). Default network activity, coupled with the frontoparietal control
network, supports goal-directed cognition. Neuroimage, 53, 303–317.
Stawarczyk, D., Majerus, S., Maquet, P., & D’Argembeau, A. (2011). Neural correlates of ongoing conscious experience: Both task-unrelatedness and
stimulus-independence are related to default network activity. PLoS ONE, 6, e16997. http://dx.doi.org/10.1371/journal.pone.0016997.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643–662.
Stuss, D. T., Floden, D., Alexander, M. P., Levine, B., & Katz, D. (2001). Stroop performance in focal lesion patients: Dissociation of processes and frontal lobe
lesion location. Neuropsychologia, 39, 771–786.
Teasdale, J. D., Dritschel, B. H., Taylor, M. J., Proctor, L., Lloyd, C. A., Nimmo-Smith, I., et al (1995). Stimulus independent thought depends on central
executive resources. Memory and Cognition, 23, 551–559.
van Maanen, L., van Rijn, H., & Borst, J. P. (2009). Stroop and picture-word interference are two sides of the same coin. Psychonomic Bulletin Review, 16,
987–999.
Vaughan, L., & Giovanello, K. (2010). Executive function in daily life: Age-related influences of executive processes on instrumental activities of daily living.
Psychology and Aging, 25, 343–355.
Vendrell, P., Junque, C., Pujol, J., Jurado, M. A., Molet, J., & Grafman, J. (1995). The role of prefrontal regions in the stroop task. Neuropsychologia, 33, 341–352.
Zatorre, R. J., & Halpern, A. R. (2005). Mental concerts: Musical imagery and auditory cortex. Neuron, 47(1), 9–12.