Journal of Experimental Child Psychology 108 (2011) 621–637

Contents lists available at ScienceDirect

Journal of Experimental Child

Psychology
journal homepage: www.elsevier.com/locate/jecp

Development of hot and cool executive function

during the transition to adolescence
Angela Prencipe a, Amanda Kesek b, Julia Cohen b, Connie Lamm c,
Marc D. Lewis d, Philip David Zelazo b,⇑
a
Department of Psychology, University of Toronto, Toronto, Ontario, Canada M5S 3G3
b
Institute of Child Development, University of Minnesota, Minneapolis, MN 55455, USA
c
Child Development Lab, University of Maryland, College Park, MD 20742, USA
d
Department of Human Development and Applied Psychology, University of Toronto, Toronto, Ontario, Canada M5S 1V6

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

Article history: This study examined the development of executive function (EF) in
Available online 1 November 2010 a typically developing sample from middle childhood to adoles-
cence using a range of tasks varying in affective significance. A total
Keywords: of 102 participants between 8 and 15 years of age completed the
Executive function
Iowa Gambling Task, the Color Word Stroop, a Delay Discounting
Cognitive control
task, and a Digit Span task. Age-related improvements were found
Emotion
Iowa gambling task
on all tasks, but improvements on relatively cool tasks (Color Word
Color word stroop Stroop and Backward Digit Span) occurred earlier in this age range,
Delay discounting whereas improvements on relatively hot tasks (Iowa Gambling
Digit span Task and Delay Discounting) were more gradual and occurred later.
Exploratory factor analysis indicated that performance on all tasks
could be accounted for by a single-factor model. Together, these
findings indicate that although similar abilities may underlie both
hot and cool EF, hot EF develops relatively slowly, which may have
implications for the risky behavior often observed during adoles-
cence. Future work should include additional measures to charac-
terize more intensively the development of both hot and cool EF
during the transition to adolescence.
Ó 2010 Elsevier Inc. All rights reserved.

Introduction

Adolescence is a period of increasing autonomy and self-regulatory demands that is often accom-
panied by risky behavior and poor decisions that can have lifelong negative consequences (e.g., Dahl,

⇑ Corresponding author.
E-mail address: zelazo@umn.edu (P.D. Zelazo).

0022-0965/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jecp.2010.09.008
622 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

2004; Ernst & Hardin, 2010; Steinberg, 2005; Steinberg et al., 2009; Van Leijenhorst & Crone, 2009).
Common risks taken by teens include reckless or intoxicated driving, experimenting with drugs and
alcohol, and unprotected sex. One factor contributing to the high incidence of such behaviors may
be the continued immaturity of executive function (EF), a neuropsychological term that refers to
the higher order cognitive control of thought, action, and emotion (e.g., Zelazo, Carlson, & Kesek,
2008).
Impairments in EF are associated with a wide variety of psychiatric disorders with childhood onset,
including attention deficit/hyperactivity disorder, conduct disorder, and autism (e.g., Willcutt, Doyle,
Nigg, Faraone, & Pennington, 2005), as well as with specific problem behaviors such as physical
aggression and substance abuse (e.g., Séguin & Zelazo, 2005). Even in healthy children, however, indi-
vidual differences in EF are associated with important developmental outcomes, including school
readiness during early childhood (Blair & Razza, 2007; Diamond, Barnett, Thomas, & Munro, 2007),
and social and cognitive function during young adulthood (e.g., Duckworth & Seligman, 2005; Eigsti
et al., 2006).
An important aspect of EF is the ability to respond flexibly and adaptively in situations that prime
maladaptive and/or prepotent responses, leading to impulsive acts and errors in judgment. In these
cases, highly salient information often introduces an element of emotion that can cloud otherwise
good judgment (but see also Damasio, 1999, for a discussion of the contribution of emotion to adap-
tive decision making). Adolescence may mark a period of particular vulnerability to such errors in part
because the development of EF continues throughout this period, lagging behind the development of
other cognitive skills (e.g., Diamond, 2002; Luciana & Nelson, 1998, 2002; Luna, Garver, Urban, Lazar, &
Sweeney, 2004; Welsh, Pennington, & Groisser, 1991). Moreover, the transition to adolescence is often
accompanied by a new set of challenging emotional experiences that may further undermine emerg-
ing cognitive control. Research on adolescents’ EF under more and less emotion-laden circumstances
may help us to understand discrepancies between adolescents’ theoretical understanding of the po-
tential negative consequences of their behavior and their real-life choices in emotion-laden situations
(e.g., in the face of peer pressure), and it may also inform us more generally about the role of emotion
in EF across development.
EF is increasingly believed to involve a number of related but distinct subfunctions, including
working memory, inhibitory control, and cognitive flexibility (Miyake et al., 2000), that together allow
conscious, goal-directed thought and behavior; that is, they allow the deliberate use of one’s knowl-
edge in the service of one’s goals. Research on the development of EF suggests that it emerges early
(around the end of infancy) and shows marked changes during the preschool years, but also that it
continues to develop at least through adolescence, in parallel with the protracted development of pre-
frontal cortex (Zelazo et al., 2008). Traditionally, EF has been examined using relatively abstract
decontextualized tasks that lack a significant affective or motivational component. More recent char-
acterizations of EF have suggested that EF tasks vary in motivational significance, with motivationally
and emotionally significant tasks described as ‘‘hot’’ and more abstract tasks described as ‘‘cool’’ (Zel-
azo & Cunningham, 2007; Zelazo & Müller, 2002). Hot tasks are thought to involve stimuli, decisions,
and outcomes that are motivationally salient. Considering the development of EF in more affectively
relevant hot situations extends the construct of EF to everyday decision making, which is rarely con-
ducted in the absence of motivational and emotional influences.
Rather than considering EF in hot and cool situations in a dichotomous fashion, this characteriza-
tion recognizes the interplay between relatively hot aspects of EF and relatively cool aspects of EF
(Cunningham & Zelazo, 2007). According to this model, information is processed hierarchically, with
relatively quick evaluative reactions followed by the generation of approach–avoidance-oriented re-
sponses. These relatively simple responses may suffice, or, if necessary, further processing and repro-
cessing of information may ensue that allows reflection on context and future consequences, and that
supports the top-down control of behavior. The extent to which a reflective response is likely to be
generated depends on a number of factors, including time, motivation, and neural and cognitive matu-
rity. Given that neural development also proceeds in a generally hierarchical fashion, with areas of the
brain associated with more complex processing developing later than areas of the brain associated
with more automatic processing (e.g., Bunge & Zelazo, 2006), children and adolescents may be ex-
pected to be less reflective than adults. Moreover, difficulty in generating reflective responses is likely
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 623

to be most evident in emotionally charged situations that have meaningful personal consequences; in
these situations, emerging cognitive control must be implemented in the context of potential interfer-
ence from relatively automatic responses to potential rewards and punishments.
It is clear from the literature on EF during early childhood that an important consequence of cog-
nitive development is the ability to dampen the emotional salience of distracting stimuli. This point is
illustrated by two studies on the development of young children’s affective decision making. First, in a
study by Carlson, Davis, and Leach (2005), children were presented with a reverse contingency task
that required them to point to a smaller reward in order to receive a larger reward; they were told
that the chosen array would be given away and they would keep the other one. Under these condi-
tions, 3-year-olds, but not 4-year-olds, had difficulty in pointing to the smaller reward rather than
the larger desired reward. However, when the motivational salience of the task was decreased by
replacing the treats with symbols, 3-year-olds’ performance improved. In addition, greater symbolic
distance from the appetitive properties of the reward was related to greater improvement in perfor-
mance; the greatest improvement was evident when a mouse and an elephant were used to symbolize
the small and large piles of treats, respectively. These results suggest that the presence of the rewards
interfered with the younger children’s ability to implement the top-down control required to inhibit
their tendency to point to the preferred reward, and graded differences in the abstract nature of the
stimuli were associated with graded changes in performance.
In the second study, by Prencipe and Zelazo (2005), 3- and 4-year-olds were presented with a delay
of gratification task that required them to choose between a single treat (candy, sticker, or penny)
immediately or more (two, four, or six) treats at the end of the game. Half of the children were asked
to make this decision for themselves, and half were asked to make the decision on behalf of the exper-
imenter (i.e., advise her on what they thought she should do). Under these conditions, 3-year-olds
were less likely to delay rewards compared with 4-year-olds. However, 3-year-olds were more likely
to delay, and performed more like 4-year-olds, when they were asked to make the decision for the
experimenter versus choosing for themselves. The results of this study suggest that the personal sal-
ience of the rewards interfered with the younger children’s ability to implement the top-down control
required to delay gratification and that decreasing personal relevance increased the likelihood that
younger children would make a reflective decision even in the presence of the hot appetitive proper-
ties of the rewards. Although these particular tasks are readily accomplished by older children, emo-
tional salience may similarly undermine good decision making in more challenging situations across
the lifespan (e.g., Crone, Bullens, van der Plas, Kijkuit, & Zelazo, 2008).
Compared with the large body of research on EF during the preschool years, relatively little is
known about the development of EF during the school-age years and around the transition to adoles-
cence (see Olson & Luciana, 2008, for review). In particular, there has been little work examining sys-
tematic changes in EF during later childhood and into adolescence using tasks designed to assess both
relatively hot and relatively cool EF. A related issue is the nature of the relation between hot and cool
EF and how this relation may change with development. Neuropsychological research with adult pa-
tients suggests that hot and cool EF are dissociable; impairments in hot EF can occur in the absence of
impairments of cool EF and vice versa (e.g., Bechara, Damasio, Tranel, & Anderson, 1998). It remains an
open question, however, whether these aspects of EF are also dissociable in children or whether they
become differentiated with development (cf. Johnson, in press).
Studies of brain maturation during adolescence suggest that there is continued development of
neural regions involved in both relatively hot and relatively cool EF during this period. Research using
structural magnetic resonance imaging (MRI) has revealed ongoing structural development in pre-
frontal cortex throughout the teen years (Giedd et al., 1999; Gogtay et al., 2004; Sowell, Thompson,
Holmes, Jernigan, & Toga, 1999; Sowell, Trauner, Gamst, & Jernigan, 2002). Increasing myelination
within the prefrontal cortex and between prefrontal cortex and subcortical regions, such as the amyg-
dala (Liston et al., 2006), suggests an increased efficiency of signaling in both corticocortical and cor-
ticolimbic neural circuits that subserve EF. Converging evidence from functional MRI (fMRI) studies in
which adolescents were scanned while performing cool EF tasks is consistent with a shift from more
diffuse to more focal functional activation in prefrontal regions (Casey et al., 1997; Durston et al.,
2006). This focalization of activity is often interpreted as an increased efficiency in neural circuits sub-
serving EF.
624 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

Using a gambling task, an fMRI study by Eshel, Nelson, Blair, Pine, and Ernst (2007) looked specif-
ically at the neural correlates of hot EF. Compared with adolescents (9–17 years of age), adults (20–
40 years of age) showed significantly more activity in orbitofrontal cortex and the dorsal anterior cin-
gulate cortex—areas thought to support emotion regulation and response monitoring—during the
selection of a relatively high-risk, high-reward option over less risky choices. Furthermore, reduced
activity in these areas was associated with more risky choices across all participants. This pattern
of results suggests that adults may more readily engage brain regions involved in processing risk
and reward. In another fMRI study of reward processing during adolescence, Galvan and colleagues
(2006) compared brain activation among children (7–11 years of age), adolescents (13–17 years of
age), and adults (23–29 years of age) during a task that parametrically manipulated reward. During
this task, the adolescent group showed adult-like activity levels in nucleus accumbens, a subcortical
area implicated in reward processing. However, brain activity in prefrontal cortex in the adolescents
was more similar to that of the children, suggesting that the development of prefrontal regulatory
areas associated with reflective processing lags behind maturation in subcortical brain regions asso-
ciated with reward. Bjork, Smith, Danube, and Hommer (2007) also examined the neural correlates
of risky decision making by varying the penalty (none, low, or high) in a monetary version of the game
‘‘Chicken’’. These authors found that performance in the low-penalty condition increased activation in
the posterior medial prefrontal cortex in adults but not in adolescents. This age difference was atten-
uated when the penalty was more substantial, suggesting that adolescents may have particular diffi-
culty in situations that involve ambiguous or marginal levels of punishment. Together, the results of
these studies suggest earlier maturation of subcortical reward circuitry relative to prefrontal regula-
tory areas associated with hot and cool EF. Enhanced activation in reward circuits relative to prefron-
tal regulatory areas may interfere with cognitive control by creating an imbalance between these
systems, and they may contribute to risk taking during adolescence (see also Crone, 2009, for a
review).
Recent behavioral studies have also begun to fill in knowledge gaps regarding continued EF
development during adolescence (e.g., Crone, Bunge, Latenstein, & van der Molen, 2005; Crone &
van der Molen, 2004; Crone et al., 2008; Hooper, Luciana, Conklin, & Yarger, 2004; Huizinga, Dolan,
& van der Molen, 2006; Lamm, Zelazo, & Lewis, 2006; Steinberg et al., 2009; see Bunge & Crone,
2009, for review). For example, Steinberg and colleagues (2009) administered a Delay Discounting
task in which participants needed to decide between a smaller immediate reward and a larger de-
layed reward and found that adolescents younger than 16 years of age tended to discount future re-
wards more steeply than either older adolescents or adults. In young adolescents, the prospect of an
immediate reward may interfere with the more prudent rational decision making usually observed
in adults.
One study that compared hot and cool measures of EF in children and adolescents was conducted
by Hooper and colleagues (2004), who tested a sample of 145 healthy participants between 9 and
17 years of age on a measure of hot EF, the Iowa Gambling Task (Bechara, Damasio, Damasio, & Ander-
son, 1994), and two measures of cool EF, Digit Span and a Go/No-Go task. The results revealed age-re-
lated improvements in performance on all three tasks, but whereas improvements on Digit Span and
Go/No-Go were seen between the two youngest age groups (9–11 and 11–13 years), only the oldest
adolescents (14–17 years) performed well on the Iowa Gambling Task. In addition, only marginally
significant (p < .10) correlations were observed between performance on the Iowa Gambling Task
and performance on the two measures of cool EF. Together, these results suggest that hot EF may fol-
low a different, perhaps delayed, trajectory of development relative to cool EF, and they encourage fur-
ther exploration of the relation between hot and cool EF in this age range.
Currently, there is only limited behavioral research examining changes in performance on tasks
with and without motivationally salient information across the transition to adolescence. Thus, the
current study was designed to characterize the developmental trajectories of, as well as the interre-
lations among, four measures of hot and cool EF during late childhood and early adolescence. In a
group of participants ranging from 8 to 15 years of age, performance was measured on two tasks
designed to measure relatively cool EF, the Color Word Stroop (Stroop, 1935) and the Digit Span
subtest of the Wechsler Intelligence Scale for Children—Third Edition (WISC-III) (Wechsler, 1991),
and two tasks designed to measure relatively hot EF, the Iowa Gambling Task (Bechara et al.,
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 625

1994) and Delay Discounting (Richards, Zhang, Mitchell, & de Wit, 1999). An important question
that arises from the preceding discussion is whether hot and cool EF rely on different underlying
processes even during development.
The Color Word Stroop and Digit Span are well-validated and widely used tasks with both child and
adult populations. Although both are considered to be prototypical cool tasks, they assess different as-
pects of EF. The selection of these two tasks allowed an examination of the developmental course of
performance on EF tasks that varied in terms of cognitive conflict. Optimal performance on the Color
Word Stroop depends on the ability to suppress interference from an automatic response (i.e., reading)
in order to perform a conflicting goal-related activity (i.e., color naming). Digit Span requires the abil-
ity to maintain and manipulate an increasingly long series of digits in working memory.
In contrast to the Color Word Stroop and Digit Span, the Iowa Gambling Task and Delay Discount-
ing assess decision making in the context of rewards. In the Iowa Gambling Task, decks of cards that
initially appear to be rewarding must be reappraised in light of their associated losses. In Delay Dis-
counting, the attractiveness of receiving an immediate reward of a particular magnitude must be
reevaluated relative to that of a future reward of greater magnitude. Both of these tasks have been
widely used to assess hot EF, and versions of the tasks have been administered to participants across
the lifespan, including both adolescents and younger children (e.g., Carlson, Zayas, & Guthormsen,
2009; Crone & van der Molen, 2004; Green, Fry, & Myerson, 1994; Hooper et al., 2004; Kerr & Zelazo,
2004; Overman et al., 2004; Steinberg et al., 2009).
Given evidence from both neural and behavioral research that emotionally significant stimuli can
interfere with cognitive control during adolescence, it was hypothesized that significant improve-
ments in performance on the hot tasks would emerge later in development than improvements in per-
formance on the cool tasks. Specifically, performance on Digit Span and the Color Word Stroop was
predicted to improve most dramatically in the youngest children, with substantial gains between 8
and 10 years of age followed by more modest improvements during early adolescence. In contrast,
improvement on the Iowa Gambling Task and Delay Discounting was expected to emerge later, with
only older adolescents demonstrating substantial improvement. Given the neuropsychological evi-
dence that hot and cool EF are dissociable (at least in adults; e.g., Bechara et al., 1998), we expected
that a factor analysis of performance on the four tasks might distinguish between hot and cool tasks.
At the same time, however, we took seriously the suggestion that hot and cool EF may depend on the
same underlying processes (e.g., reflection and rule use [Zelazo & Cunningham, 2007]), and we consid-
ered the possibility that hot and cool EF may be dissociable only later in development as a function of
neural differentiation and specialization (e.g., Johnson, in press).

Method

Participants

A total of 102 children and adolescents between 8 and 15 years of age participated in the current
study. Participants were divided into four age groups: 8- and 9-year-olds (n = 26, 10 boys and 16 girls,
mean age = 8.90 years, range = 7.98–9.83), 10- and 11-year-olds (n = 28, 17 boys and 11 girls, mean
age = 10.90 years, range = 10.06–11.95), 12- and 13-year-olds (n = 21, 10 boys and 11 girls, mean
age = 12.69 years, range = 12.07–13.72), and 14- and 15-year-olds (n = 27, 13 boys and 14 girls, mean
age = 15.04 years, range = 14.12–15.88).
Participants were recruited from a large metropolitan area through an advertisement placed in a
local newspaper as part of a larger study as well as through a database of families who had expressed
interest in participating in research. The sample was composed of Canadian children who were mostly
of European descent (50%), followed by African and Caribbean descent (13%), Asian descent (12%),
South Asian descent (11%), and other ethnicities (14%). Families received monetary compensation
($10 per child), and an additional $10 was presented to each child as part of his or her performance
on one of the tasks. Participants were screened for clinical disorders and uncorrected visual impair-
ments; none of the participants reported learning disabilities or behavioral problems, nor were any
on medication at the time of testing. Thus, the participants in this study were considered to represent
a typically developing sample. Informed parental consent was obtained for all participants.
626 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

General procedure

Each participant was tested individually in a quiet area of his or her home with no distracting ele-
ments. For each session, the participant and experimenter were seated side by side at a table, with the
participant seated comfortably in front of a computer monitor. To ensure that the child understood the
tasks, the experimenter clarified the requirements if needed. Each session lasted approximately
30 min. The following tasks were administered in a fixed order, alternating between hot and cool mea-
sures: Iowa Gambling Task, Color Word Stroop, Delay Discounting, and Forward/Backward Digit Span.

Measures

Iowa Gambling Task

Participants were administered a computerized version of the Iowa Gambling Task (Bechara et al.,
1994). In this task, participants were shown four decks of cards, labeled A, B, C, and D, which were dis-
played in a linear sequence across the screen. Participants were told that they could select one card at
a time from any deck by clicking on that deck with the computer mouse and that they could continue
selecting cards until the experimenter stopped the game. When selected, each card revealed a combi-
nation of gains and losses (measured in play money). Participants were given a stake of $2000 and
were asked to win as much money as possible by choosing cards from any of the four decks (one card
per trial), treating the play money as if it were real. They were not told how many trials there would be
(100), but they were told that some of the decks were better than the others. In fact, the task was de-
signed so that choosing consistently from two of the decks (C and D) would result in a net gain,
whereas choosing consistently from the other two decks (A and B) would result in a net loss. Decks
A and B delivered large immediate rewards ($100) but were disadvantageous in the long run because
they also delivered larger losses than rewards. Decks C and D delivered smaller immediate rewards
($50) than the other two decks but were advantageous in the long run because they delivered even
smaller losses. In contrast to rewards, the amounts of which were fixed, losses from all decks were
variable and unpredictable; in particular, two of the decks (A and C) delivered more frequent losses
(50% of cards), whereas the other two decks (B and D) delivered less frequent losses (10% of cards).
To ensure that participants understood that outcomes were contingent on their selections, the exper-
imenter guided them on the first few trials by stating the amounts won and lost. Using this protocol,
all participants appeared to understand the link between choices and outcomes. After 100 selections,
the task stopped automatically.
As in previous work with this task (e.g., Bechara et al., 1994), performance was computed for five
consecutive blocks of 20 trials. For each block, the number of disadvantageous choices was subtracted
from the number of advantageous choices, yielding difference scores ranging from 10 to +10. Positive
difference scores indicate relatively advantageous performance, whereas negative scores indicate rel-
atively disadvantageous performance.

Color Word Stroop

Participants were presented with two laminated cards, each containing 21 words, and were asked
to name the color in which the word was printed (Stroop, 1935). Two conditions were administered in
a fixed order: congruent color word and incongruent color word. In each condition, participants were
shown a laminated card with 21 words arranged in two columns and were instructed to name the col-
ors, starting with the first column, as quickly and accurately as possible without skipping any. Words
were presented in three colors: red, blue, and green. In the congruent condition, color words were pre-
sented in colors that were congruent with the word (e.g., the word RED in red ink), whereas in the
incongruent condition, color words were presented in incongruent colors (e.g., the word RED in blue
ink). As in other common versions of the Color Word Stroop (e.g., the Delis–Kaplan Executive Function
System [Delis, Kaplan, & Kramer, 2001]), performance on each card was timed using a stopwatch and
the number of errors was recorded. Following the procedure outlined by Stroop (1935), we created
adjusted reaction time scores by subtracting twice the average time per word for each error from
the total reaction time per condition. No children had difficulty reading the words. However, to control
for individual reading ability (and to obtain a measure of the effect of interference), reaction time on
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 627

congruent trials was subtracted from reaction time on incongruent items. Lower scores indicated bet-
ter performance.

Delay Discounting
A computerized adjusting amount procedure was used to assess the discounting of delayed rein-
forcers (Richards et al., 1999). Discounting describes the decline in the subjective value of a reward
when its availability is delayed. Participants were given a series of choices between a variable imme-
diate amount of money and a fixed amount of $10 that would be delayed by 1, 2, 30, 180, or 365 days.
Both the amount of immediate reward and the order of presentation varied with participant choices;
for each choice, an adjusting amount algorithm adjusted the magnitude of the immediate reinforcer
(immediate amounts started at 50 cents and increased in 50-cent intervals) until it was subjectively
equivalent in value to the delayed reward. This value of the immediate reinforcer was identified as the
indifference point (ID). Participants’ data were fitted to the hyperbolic function, ID = A/(1 + kD), de-
scribed in Mazur (1987), to calculate values for k, which provides a measure of the rate at which each
participant discounted reinforcers as a function of delay. In the hyperbolic function, A is the nominal
amount of the delayed reward ($10) and D is the length of the delay (1–365 days). The distribution of k
values was skewed, as is typical for this type of data (e.g., Johnson & Bickel, 2002), so k values were
log-transformed for the purpose of analysis (Alessi & Petri, 2003). Lower scores indicated better
performance.
To help ensure that participants were appropriately engaged by the task, they were told that fol-
lowing completion of the task, one of their choices would be granted at random. Regardless of perfor-
mance, however, participants received $10 at the end of the task.

Digit Span
The Digit Span subtest from the WISC-III (Wechsler, 1991) is composed of two types of assess-
ment, Forward Digit, which requires participants to repeat a sequence of digits verbatim, and Back-
ward Digit, which requires participants to repeat the digits in reverse order. Participants completed
both in a fixed order (Forward and Backward) with sequences ranging from two to nine digits in the
Forward Digit component and from two to eight digits in the Backward Digit component. Each se-
quence was presented by the experimenter at a rate of one digit per second, after which partici-
pants were required to recite the digits from memory. Two trials were administered for each
sequence length; if participants were correct on either trial, then they advanced to the next se-
quence length, with the number of digits increasing by one. Scores were computed by counting
the total number of sequences successfully remembered in each condition. Forward Digit Span
and Backward Digit Span were analyzed as separate dependent variables (Lezak, 1995). For relations
among tasks, only Backward Digit Span was considered because this is widely believed to be a bet-
ter index of EF insofar as it requires both the maintenance and manipulation of information in
working memory. Published norms for performance on this task cover a wide age range (6–16 years
[Wechsler, 1991]).
For each task, participants were considered as outliers if their mean scores were more than three
standard deviations above or below the mean for their age group. Four outliers (distributed across age
groups) were deleted from the Iowa Gambling Task analysis, two outliers were deleted from the Color
Word Stroop analysis, and one outlier was deleted from the Digit Span analysis. These outliers were
not deleted from the exploratory factor analysis (see below) because they were not found to influence
the results.
The analysis plan involved several steps. First, to address age-related changes in performance,
performance on each measure of EF was regressed onto age in months, and then separate analyses
of variance (ANOVAs) were conducted on scores for each task by age group (8- and 9-year-olds,
10- and 11-year-olds, 12- and 13-year-olds, and 14- and 15-year-olds) and sex. Post hoc comparisons
for all analyses were performed using a Bonferroni correction, with the critical value set at p < .05.
Second, to explore associations among performance on hot and cool tasks, partial correlations among
the tasks (controlling for age) were examined, followed by an exploratory factor analysis.
628 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

Results

Age-related trends in performance

Table 1 presents the standardized parameter estimates for performance on each measure regressed
onto age (in months). Age was related to performance on all of the tasks (p < .005).

Iowa Gambling Task

Means (and standard deviations) for the number of net advantageous choices across all five
blocks for each of the age groups were 7.92 (14.99), 6.93 (21.58), 10.11 (16.98), and 2.00
(19.42) for 8- and 9-, 10- and 11-, 12- and 13-, and 14- and 15-year-olds, respectively. Net scores
for each trial block were analyzed using a 4 (Age Group)  2 (Sex)  5 (Block: 1, 2, 3, 4, or 5)
mixed-model ANOVA.
A main effect of age group, F(3, 90) = 3.71, p < .05, g2 = .11, was qualified by a significant Age
Group  Block interaction, F(12, 360) = 2.62, p < .01, g2 = .08. Post hoc tests revealed that perfor-
mance differences between age groups were evident only in the last block, F(3, 94) = 4.53, p < .05,
g2 = .13, where the oldest group performed better compared with all other age groups (see Fig. 1);
the only improvement in performance over trial blocks occurred among participants in the oldest
age group, who chose more advantageously in each of the last four trial blocks compared with the
first block.
The analysis also revealed a marginally significant Age Group  Sex interaction, F(3, 90) = 2.36,
p = .077, g2 = .07. Given that past research has found sex differences on this task (e.g., Overman,
2004), we chose to investigate this trend further. Post hoc tests revealed that the oldest boys made
more advantageous choices compared with the 8- and 9-year-olds and 12- and 13-year-olds, but
not when compared with the 10- and 11-year-olds. However, gender comparisons within each age
group were not significant.
The t distribution was used to compare the means for each age and block with what would be ex-
pected if participants responded randomly (i.e., a score of 0). Only participants in the oldest age group
had net scores in the last block that were significantly higher than chance, t(25) = 2.23, p < .05.
Finally, we investigated whether children’s performance on this task reflected a bias to avoid decks
that delivered more frequent loss outcomes, as reported in previous work (e.g., Carlson et al., 2009;
Crone et al., 2005; Huizinga, Crone, & Jansen, 2007). A 4 (Age Group)  2 (Sex)  2 (Gain: advanta-
geous or disadvantageous)  (Frequency: high or low)  2 (Block: Trials 1–50 or Trials 51–100)
mixed-model ANOVA was conducted on the number of selections per deck. This analysis revealed a
main effect of frequency, F(1, 90) = 22.32, p < .001, g2 = .20, reflecting a bias for low-frequency loss
decks. The main effect of frequency was qualified by the following interactions: Frequency  Age
Group, F(3, 90) = 4.09, p < .01, g2 = .12, Frequency  Gain, F(1, 90) = 73.18, p < .001, g2 = .49, Fre-
quency  Block, F(1, 90) = 12.13, p < .001, g2 = .12, Frequency  Age Group  Block, F(3, 90) = 2.75,
p < .05, g2 = .08, and Frequency  Gain  Block, F(1, 90) = 3.94, p < .05, g2 = .04, all of which were qual-
ified by an overall Frequency  Age Group  Gain  Block interaction, F(3, 90) = 2.85, p < .05, g2 = .09.
Post hoc tests (p < .05, Bonferroni corrected) revealed that by the second trial block, participants in the
oldest age group made more selections from the low-frequency loss deck not only for the disadvan-
tageous decks but also for the advantageous decks.

Table 1
Standardized parameter estimates (and confidence intervals) of performance on
key tasks regressed on age (in months).

Backward Digit Span +.406 (+.568, +.244)

Color Word Stroop .515 ( .658, .373)
Delay Discounting .257 ( .438, .076)
Iowa Gambling Task +.246 (+.428, +.064)
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 629

Fig. 1. Iowa Gambling Task: Net advantageous selections (±1 standard error) per trial block by age group.

Color Word Stroop

To examine age-related changes in performance on this task, we conducted a 2 (Condition: congru-
ent or incongruent)  4 (Age Group)  2 (Sex) mixed-model ANOVA to examine reaction time. This
analysis revealed a main effect of condition, F(1, 92) = 330.8, p < .001, g2 = .78, as well as a Condi-
tion  Age Group interaction, F(3, 92) = 13.47, p < .001, g2 = .30. A comparison of congruent and incon-
gruent trials revealed that all participants were faster to respond on the congruent trials relative to the
incongruent trials. To explore the Condition  Age Group interaction, a difference score was computed
by subtracting reaction times on the congruent condition from reaction times on the incongruent con-
dition. This measure of the effect of interference was then used as the dependent variable in a 4 (Age
Group)  2 (Sex) ANOVA. As seen in Fig. 2, there was a highly significant effect of age group on the
interference measure F(3, 99) = 13.47, p < .001, g2 = .31. Post hoc comparisons found that the youngest
group (8- and 9-year-olds) showed a significantly greater effect of interference relative to the three
older age groups (p < .002, p < .001, and p < .001 for 10- and 11-, 12- and 13-, and 14- and 15-year-
old groups, respectively). The 10- and 11-year-old group also showed a significantly greater effect
of interference relative to the oldest group (p < .05).

Fig. 2. Color Word Stroop: Mean difference (±1 standard error) in reaction time between incongruent and congruent trials by
age group.
630 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

Fig. 3. Delay Discounting: Mean untransformed k values (±1 standard error) by age group.

Delay Discounting
A 4 (Age Group)  2 (Sex) ANOVA on log-transformed k values indicated a main effect of age group,
F(3, 94) = 3.00, p < .05, g2 = .09. Post hoc tests comparing the age groups revealed higher k values for
the youngest group compared with the 14- and 15-year-olds (p < .05) (see Fig. 3 for the untransformed
k values). No other comparisons were significant.

Digit Span
A 2 (Direction: forward or backward)  4 (Age Group)  2 (Sex) mixed-model ANOVA was con-
ducted on forward and backward scores (the number of sequences completed correctly). This analysis
revealed a main effect of direction (forward vs. backward), F(1, 93) = 264.93, p < .001, g2 = .74, with
better performance on Forward Digit Span than on Backward Digit Span (see Fig. 4). The analysis also
revealed a main effect of age group, F(3, 93) = 9.07, p < .001, g2 = .226. Post hoc analyses revealed that
children in the youngest age group remembered significantly fewer sequences than the 10- and 11-
year-olds and 14- and 15-year-olds (p < .05 and p < .001, respectively). There was also a trend suggest-
ing that the youngest age group had poorer memory performance relative to the 12- and 13-year-old
group (p = .06). Thus, results suggest that improvements in memory span scores occurred between the
youngest age group and the three older groups. There was no significant interaction between digit
span direction and age group.

Fig. 4. Digit Span: Mean number of sequences completed (±1 standard error) for Forward Digit Span and Backward Digit Span
by age group.
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 631

Table 2
Partial correlations (controlling for age) between scores on cool and hot tasks.

Measure 1 2 3 4
1. Backward Digit Span – –.25** –.15 .21
2. Color Word Stroop – – .09 –.26**
3. Delay Discounting – – – –.19
4. Iowa Gambling Task (net advantageous choices) – – – –

p < .10 (two-tailed).

**
p < .01 (two-tailed).

Table 3
Factor loadings for each task overall and for younger versus older children.

Overall Younger Older

(8–15 years) (8–11 years) (12–15 years)
1. Backward Digit Span .62 .63 .48
2. Color Word Stroop .65 .95 .36
3. Delay Discounting .36 .17 .49
4. Iowa Gambling Task .50 .32 .63

Note. >.70 = excellent; .63–.69 = very good; .55–.62 = good; .45–.54 = fair; .32–.44 = poor (Comrey &
Lee, 1992).

Associations among hot and cool tasks

Partial correlations, controlling for age, were computed to investigate whether performance on the
two hot tasks was related to performance on the two cool tasks (see Table 2). These correlations pro-
vide some support for overlap in performance variance associated with hot and cool tasks. For exam-
ple, net advantageous choices on the Iowa Gambling Task were significantly correlated with
performance on the Color Word Stroop (p < .01). In addition, both the Iowa Gambling Task and Delay
Discounting were correlated with Digit Span scores at a trend level (p < .10). However, the correlation
between Delay Discounting and Stroop scores did not approach significance. The two cool tasks, Digit
Span and Stroop, were significantly correlated (p < .01), whereas the two hot tasks, Delay Discounting
and the Iowa Gambling Task, were not.
An exploratory factor analysis was used to assess whether different factors would distinguish per-
formance on hot versus cool tasks. To obtain appropriate sample size restrictions, age groups were
combined to create a younger age group (8–11 years) and an older age group (12–15 years). The re-
sults of the maximum likelihood factor analysis for the overall sample, and separately by age group
(younger vs. older), are presented in Table 3. A single-factor model was sufficient to account for the
data from the overall sample and from each age group, with 100% of the variance accounted for by this
factor. However, the pattern of item (task performance) loadings within the factor structure varied by
age group. For example, as shown in Table 3, Color Word Stroop performance loaded most strongly for
the younger group, whereas it fell last in the item-loading pattern for the older group. An examination
of standardized estimates of these factor loadings revealed that all item loadings were statistically sig-
nificant except for Delay Discounting for the younger group.

Discussion

The current study examined the development of EF from middle childhood through mid-adoles-
cence using a battery of tasks assessing both relatively cool and relatively hot EF. The Iowa Gambling
Task and Delay Discounting both require the flexible appraisal of motivationally significant stimuli
and, therefore, were chosen to measure relatively hot EF. In contrast, the Color Word Stroop and Digit
Span task were chosen to investigate cool EF. Given a theoretical framework suggesting that children
632 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

and adolescents may have special difficulty in engaging in reflective processing in the context of
motivationally salient information (Zelazo & Cunningham, 2007), we hypothesized that developmen-
tal changes on the relatively hot EF tasks would occur later than changes on relatively cool EF tasks.
This prediction was supported by the current data. Performance on all tasks improved with age, but
improvements on the cool EF tasks were more robust and occurred earlier in the age range tested.
The three older groups outperformed the youngest group on the Color Word Stroop and Digit Span
task. Improvements on the relatively hot tasks, in contrast, were evident only in the oldest group.
The oldest group performed better than the three younger groups on the Iowa Gambling Task and per-
formed better than the youngest group on Delay Discounting. This pattern of results is consistent with
the suggestion that EF may be expressed earlier in relatively cool contexts, with relatively hot contexts
posing additional regulatory challenges. Despite these different developmental trajectories, however,
performance on hot and cool tasks does not appear to reflect different underlying factors based on the
results of the exploratory factor analysis.

Iowa Gambling Task

The results from the Iowa Gambling Task are consistent with research suggesting that children and
adolescents are less likely to reflect on long-term outcomes relative to adults (Crone & van der Molen,
2004; Crone et al., 2005; Garon & Moore, 2004; Kerr & Zelazo, 2004). Only the oldest adolescents in the
current study learned to choose advantageously by the last 20 trials (i.e., trials 80–100). This is
generally consistent with the findings of Hooper and colleagues (2004), who reported that 14- to
17-year-olds performed better than younger participants by the second half of the task (by Trials
60–80); it is possible that the current study would have also reported an earlier shift to advantageous
decks if the upper limit of the oldest age group were extended to 17 years.
The finding that all participants were biased to avoid frequent punishment confirms previous work
(Carlson et al., 2009; Huizinga et al., 2007), although it is somewhat discrepant with the results of
Hooper and colleagues (2004), who reported that only older participants in their study were more
biased toward the low-frequency loss decks overall. By the second half of the task, the oldest partic-
ipants in the current study began to favor low-frequency losses even for the advantageous decks, con-
sistent with the suggestion that the ability to respond to more ambiguous or marginal levels of
punishment improves over adolescence (e.g., Bjork et al., 2007).

Color Word Stroop

Substantial improvements in performance on the Color Word Stroop emerged relatively early in
the age range tested, with the largest improvements occurring between the two youngest age groups
(i.e., by 10–11 years). These results are in accord with previous findings of age-related improvement in
an adolescent sample (Adleman et al., 2002) and suggest that EF, as measured by this task, develops
relatively early. In fact, the results for the oldest age group are typical of Stroop effects reported in
adults (MacLeod, 1991; Uttl & Graf, 1997).

Delay Discounting

The findings from the Delay Discounting task are consistent with previous research suggesting that
impulsive choices decrease gradually over an extended age span. The measure of impulsivity used in
this task, k values, was of a greater magnitude and exhibited greater variance than is typically reported
for normative adult populations (Dixon, Marley, & Jacobs, 2003; Green, Myerson, & Ostaszewski, 1999;
Steinberg et al., 2009), further supporting the hypothesis that the ability to make adaptive, future-
oriented decisions in the context of reward stimuli is not yet fully developed even during adolescence.

Digit Span

Performance on Backward Digit Span was strongly related to age. As with the Color Word Stroop,
the largest improvements on Backward Digit Span occurred relatively early in adolescence, between
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 633

the two youngest age groups. This finding is generally consistent with the findings of Hooper and
colleagues (2004).
More generally, studies of working memory during adolescence have yielded variable results, with
some studies reporting adult levels of performance by early adolescence (e.g., Asato, Sweeney, & Luna,
2006; Crone, Wendelken, Donohue, van Leijenhorst, & Bunge, 2006) and others showing continued
improvement through the late adolescent years (e.g., Hooper et al., 2004; Huizinga et al., 2006; Lamm
et al., 2006). Possible reasons for discrepancies include the modalities of stimulus presentation
(auditory or visual) and recall (visual–spatial or verbal). Other relevant factors include the nature
and complexity of the stimuli that must be remembered and the rate of stimulus presentation. Even
the 10- and 11-year-olds in the current study performed as well as typical adults on the measure used
here (Janssen, Krabbendam, Jolles, & van Os, 2003; Kinsella et al., 1996).

Relations among tasks

Previous research with both adults (Miyake et al., 2000) and children (Brocki & Bohlin, 2004; Lehto,
Juujärvi, Kooistra, & Pulkkinen, 2003) suggests that EF encompasses a number of related but distinct
factors, but these studies did not include both hot and cool measures of EF. Neuropsychological and
neuroimaging data strongly suggest that performance on hot and cool EF reflects dissociable functions,
at least in adults (e.g., Bechara et al., 1998). The results of the exploratory factor analysis in the current
study, however, did not find evidence for the differentiation of hot versus cool EF. Instead, a one-factor
model was found to be sufficient to account for the data. Age-partialed correlations also revealed a
robust association between the Iowa Gambling Task and the Color Word Stroop and between the Color
Word Stroop and Backward Digit Span. As in earlier work by Hooper and colleagues (2004), however,
only a marginally significant relation (p < .10) was found between the Iowa Gambling Task and Back-
ward Digit Span. Nonetheless, taken together, these findings are consistent with the suggestion that
hot and cool EF work in concert, rely on the same underlying mechanisms (e.g., the iterative reprocess-
ing of information in lateral prefrontal cortex), and may be best viewed as falling along a continuum
(Zelazo & Cunningham, 2007).
Age variation in the pattern of item (task performance) loadings within the factor structure may
reflect the relative contribution of different subfunctions to EF performance at different ages. For
example, the relatively strong factor loading for Color Word Stroop for younger participants in this
study suggests that the monitoring of response conflicts, associated with anterior cingulate cortex
and strongly implicated in this task (e.g., Pardo, Pardo, Janer, & Raichle, 1990), may make a greater
contribution to EF at this age. The shift in relative factor loadings for Stroop and Iowa Gambling Task
performance between the younger and older participants may reflect a greater reliance, with age, on
more dorsal and lateral regions of prefrontal cortex involved in more complex rule use (e.g., Bunge &
Zelazo, 2006).
Delay Discounting did not load significantly on the factor structure for the younger group. One pos-
sibility is that this task is less reliable for younger children, perhaps because these children fail to
appreciate fully the nature of the delays involved. Overall, however, despite age fluctuations in the fac-
tor loadings associated with performance on each task, the current set of results fails to provide evi-
dence that performance on these tasks can be differentiated into separate factors corresponding to hot
and cool aspects of EF. Future work using more tasks may reveal separate factors, as found in other
research on the development of EF (see Best, Miller, & Jones, 2009, for a review).
A more direct test of the hypothesis that hot and cool EF can be differentiated by separate factor
structures was not possible due to limitations in our dataset; a confirmatory factor analysis was not
advisable in the current study because it is recommended that each latent variable be assessed with
at least three indicator variables and preferably four or five (see Hatcher, 1994, p. 260). Future work
should include a greater variety of hot and cool tasks and larger sample sizes to assess the factor struc-
ture using confirmatory factor analysis and to determine whether hot and cool tasks load onto differ-
ent factors, although any factor-analytic results should still be interpreted with caution (see Miyake
et al., 2000; Zelazo & Müller, 2002). For now, the single-factor structure in the current study suggests
that the set of relatively hot and relatively cool tasks presented here may rely on similar underlying
abilities. If supported by the results of future research, this finding, together with the finding of
634 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

relatively late developmental changes on hot versus cool measures of EF, suggests that although both
hot and cool EF rely on similar underlying processes, hot EF tasks may be relatively more challenging
during childhood and adolescence.

Limitations

The current study has several limitations that should be considered. First, to ensure that children
remained engaged in the tasks throughout the session, only a limited number of tasks were selected to
assess relatively hot and relatively cool EF. Thus, the current findings should be considered prelimin-
ary and limited in their generalizability. Future research should include a broader range of hot and
cool tasks and should examine interrelations using a confirmatory factor analysis to test the hypoth-
esis that hot and cool EF represent separable factors at the behavioral level.
Furthermore, to make tasks more directly comparable across age, future work should vary motiva-
tional and affective salience within tasks (e.g., Carlson et al., 2005; Crone et al., 2008; Kohls, Peltzer,
Herpetz-Dahlmann, & Konrad, 2009; Prencipe & Zelazo, 2005). This would allow more direct compar-
isons of the development of relatively hot and cool EF.

General discussion

In the Color Word Stroop and Digit Span, both relatively cool tasks that lack any obvious motiva-
tional significance, substantial improvements in performance were observed early in development.
This is in contrast to performance on the Iowa Gambling task and Delay Discounting, which require
reflective responding in the context of reward information. Although the objective of each of these rel-
atively hot tasks was understood by all of the children, only the oldest children made responses in an
effective, goal-directed way.
An exploratory factor analysis yielded a single factor, consistent with the suggestion that hot and
cool EF work in concert, rely on the same underlying mechanisms (e.g., the iterative reprocessing of
information in lateral prefrontal cortex), and may be best viewed as falling along a continuum (Zelazo
& Cunningham, 2007). Nonetheless, hot EF may develop later than cool EF, and the ability to imple-
ment cognitive control in relatively hot contexts may be hampered by motivationally salient informa-
tion (e.g., Dahl, 2004; Steinberg, 2005).
In the current study, even relatively inconsequential reward information interfered with decision
making during middle childhood and early adolescence. This suggests that although children and ado-
lescents may implement considerable cognitive control in some contexts, this ability is still easily
undermined by information that is motivationally salient. Given that young adolescents are often in
situations that require rational reflective responses in the context of strong motivational factors
(e.g., peer pressure), this difficulty may have important implications for the kind of risky decisions of-
ten observed during adolescence. Adolescents may have the ability to apply EF in relatively cool sit-
uations but may have difficulty making reflective advantageous decisions in situations that more
closely resemble real-life decision making.

Acknowledgments

We thank Selahadin Ibrahim for his expert assistance with the data analyses and Andrea Reynolds
for her help with data collection.

References

Adleman, N. E., Menon, V., Blasey, C. M., White, C. D., Warsofsky, I. S., Glover, G. H., et al (2002). A developmental fMRI study of
the Stroop Color-Word task. NeuroImage, 16, 61–75.
Alessi, S. M., & Petri, N. M. (2003). Pathological gambling severity is associated with impulsivity in a delay discounting
procedure. Behavioural Processes, 64, 345–354.
Asato, M. R., Sweeney, J. A., & Luna, B. (2006). Cognitive processes in the development of TOL performance. Neuropsychologia, 44,
2259–2269.
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 635

Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity to future consequences following damage to
human prefrontal cortex. Cognition, 50, 7–15.
Bechara, A., Damasio, H., Tranel, D., & Anderson, S. W. (1998). Dissociation of working memory from decision making within the
human prefrontal cortex. Journal of Neuroscience, 18, 428–437.
Best, J. R., Miller, P. H., & Jones, L. L. (2009). Executive functions after age 5: Changes and correlates. Developmental Review, 29,
180–200.
Bjork, J. M., Smith, A. R., Danube, C. L., & Hommer, D. W. (2007). Developmental differences in posterior mesofrontal cortex
recruitment by risky rewards. Journal of Neuroscience, 27, 4839–4849.
Blair, C., & Razza, R. A. (2007). Relating effortful control, executive function, and false belief understanding to emerging math
and literacy ability in kindergarten. Child Development, 78, 647–663.
Brocki, K. C., & Bohlin, G. (2004). Executive functions in children aged 6 to 13: A dimensional and developmental study.
Developmental Neuropsychology, 26, 571–593.
Bunge, S. A., & Crone, E. A. (2009). Neural correlates of the development of cognitive control. In J. Rumsey & M. Ernst (Eds.),
Neuroimaging in developmental clinical neuroscience (pp. 22–37). Cambridge, UK: Cambridge University Press.
Bunge, S., & Zelazo, P. D. (2006). A brain-based account of the development of rule use in childhood. Current Directions in
Psychological Science, 15, 118–121.
Carlson, S. M., Davis, A. C., & Leach, J. G. (2005). Less is more: Executive function and symbolic representation in preschool
children. Psychological Science, 16, 609–616.
Carlson, S. M., Zayas, V., & Guthormsen, A. (2009). Neural correlates of decision making on a gambling task. Child Development,
80, 1076–1096.
Casey, B. J., Trainor, R. J., Orendi, J. L., Schubert, A. B., Nystrom, L. E., Giedd, J. N., et al (1997). A developmental functional MRI
study of prefrontal activation during performance of a Go–No–Go task. Journal of Cognitive Neuroscience, 9, 835–847.
Comrey, A. L., & Lee, H. B. (1992). A first course in factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
Crone, E. A. (2009). Executive functions in adolescence: Inferences from brain and behavior. Developmental Science, 12,
825–830.
Crone, E. A., Bullens, L., van der Plas, E. A. A., Kijkuit, E. J., & Zelazo, P. D. (2008). Developmental changes and individual
differences in risk and perspective taking in adolescence. Development and Psychopathology, 20, 1213–1229.
Crone, E. A., Bunge, S., Latenstein, H., & van der Molen, M. W. (2005). Characterization of children’s decision-making: Sensitivity
to punishment frequency, not task complexity. Child Neuropsychology, 11, 245–263.
Crone, E. A., & van der Molen, M. W. (2004). Developmental changes in real life decision making: Performance on the gambling
task previously shown to depend on the ventromedial prefrontal cortex. Developmental Neuropsychology, 25, 251–279.
Crone, E. A., Wendelken, C., Donohue, S., van Leijenhorst, L., & Bunge, S. A. (2006). Neurocognitive development of the ability to
manipulate information in working memory. Proceedings of the National Academy of Sciences, 103, 9315–9320.
Cunningham, W. A., & Zelazo, P. D. (2007). Attitudes and evaluations: A social cognitive neuroscience perspective. Trends in
Cognitive Sciences, 11, 97–104.
Dahl, R. E. (2004). Adolescent brain development: A period of vulnerabilities and opportunities. Annals of the New York Academy
of Sciences, 1021, 1–22.
Damasio, A. R. (1999). The feeling of what happens. San Diego: Harcourt Brace.
Delis, D. C., Kaplan, E., & Kramer, J. H. (2001). The Delis–Kaplan executive function system. San Antonio, TX: Psychological
Corporation.
Diamond, A. (2002). Normal development of prefrontal cortex from birth to young adulthood: Cognitive functions, anatomy,
and biochemistry. In D. T. Stuss & R. T. Knight (Eds.), Principles of frontal lobe function (pp. 466–503). London: Oxford
University Press.
Diamond, A., Barnett, W. S., Thomas, J., & Munro, S. (2007). Preschool program improves cognitive control. Science, 318,
1387–1388.
Dixon, M., Marley, J., & Jacobs, E. (2003). Delay discounting by pathological gamblers. Journal of Applied Behavioral Analysis, 36,
449–458.
Duckworth, A. L., & Seligman, M. E. P. (2005). Self-discipline outdoes IQ in predicting academic performance of adolescents.
Psychological Science, 16, 939–944.
Durston, S., Davidson, M. C., Tottenham, N., Galvan, A., Spicer, J., Fossella, J. A., et al (2006). A shift from diffuse to focal cortical
activity with development. Developmental Science, 9, 1–20.
Eigsti, I. M., Zayas, V., Mischel, W., Shoda, Y., Ayduk, O., Dadlani, M., et al (2006). Predicting cognitive control from preschool to
late adolescence and young adulthood. Psychological Science, 17, 478–484.
Ernst, M., & Hardin, M. (2010). Neurodevelopment underlying adolescent behavior: A neurobiological model. In P. D. Zelazo, M.
Chandler, & E. Crone (Eds.), Developmental social cognitive neuroscience (pp. 165–190). New York: Psychology Press.
Eshel, N., Nelson, E. E., Blair, R. J., Pine, D. S., & Ernst, M. (2007). Neural substrates of choice selection in adults and adolescents:
Development of the ventrolateral prefrontal and anterior cingulate cortices. Neuropsychologia, 45, 1270–1279.
Galvan, A., Hare, T. A., Parra, C. E., Penn, J., Voss, H., Glover, G., et al (2006). Earlier development of the accumbens relative to
orbitofrontal cortex might underlie risk-taking behavior in adolescents. Journal of Neuroscience, 26, 6885–6892.
Garon, N., & Moore, C. (2004). Complex decision-making in early childhood. Brain and Cognition, 55, 158–170.
Giedd, J. N., Blumenthal, J., Jeffries, N. O., Castellanos, F. X., Liu, H., Zijdenbos, A., et al (1999). Brain development during
childhood and adolescence: A longitudinal MRI study. Nature Neuroscience, 2, 861–863.
Gogtay, N., Giedd, J. N., Lusk, L., Hayashi, K. M., Greenstein, D., Vaituzis, A. C., et al (2004). Dynamic mapping of human cortical
development during childhood through early adulthood. Proceedings of the National Academy of Sciences of the United States
of America, 101, 8174–8179.
Green, L., Fry, A. F., & Myerson, J. (1994). Discounting of delayed rewards: A life-span comparison. Psychological Science, 5, 33–36.
Green, L., Myerson, J., & Ostaszewski, P. (1999). Discounting of delayed rewards across the lifespan: Age differences in individual
discounting functions. Behavioral Processes, 46, 89–96.
Hatcher, L. (1994). A step-by-step approach to using SAS for factor analysis and structural equation modeling. Cary, NC: SAS
Institute.
636 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637

Hooper, C., Luciana, M., Conklin, H. M., & Yarger, R. (2004). Adolescents’ performance on the Iowa Gambling Task: Implications
for the development of decision-making and ventromedial prefrontal cortex. Developmental Psychology, 40, 1148–1158.
Huizinga, H. M., Crone, E. A., & Jansen, B. J. (2007). Decision-making in healthy children, adolescents, and adults explained by the
use of increasingly complex proportional reasoning rules. Developmental Science, 10, 814–825.
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, 2017–2036.
Janssen, I., Krabbendam, L., Jolles, J., & van Os, J. (2003). Alterations in theory of mind in patients with schizophrenia and non-
psychotic relatives. Acta Psychiatrica Scandinavica, 108, 110–117.
Johnson, M. H. (in press). Interactive specialization: A domain-general framework for human functional brain development?
Developmental Cognitive Neuroscience, doi:10.1016/j.dcn.2010.07.003.
Johnson, M. W., & Bickel, W. K. (2002). Within-subject comparison of real and hypothetical money rewards in delay discounting.
Journal of the Experimental Analysis of Behavior, 77, 129–146.
Kerr, A., & Zelazo, P. D. (2004). Development of ‘‘hot’’ executive function: The Children’s Gambling Task. Brain and Cognition, 55,
148–157.
Kinsella, G., Murtagh, D., Landry, A., Homfray, K., Hammond, M., O’Beirne, L., et al (1996). Everyday memory following traumatic
brain injury. Brain Injury, 10, 499–507.
Kohls, G., Peltzer, J., Herpetz-Dahlmann, B., & Konrad, K. (2009). Differential effects of social and non-social reward on response
inhibition in children and adolescents. Developmental Science, 12, 614–625.
Lamm, C., Zelazo, P. D., & Lewis, M. D. (2006). Neural correlates of cognitive control in childhood and adolescence: Disentangling
the contributions of age and executive function. Neuropsychologia, 44, 2139–2148.
Lehto, J. E., Juujärvi, P., Kooistra, L., & Pulkkinen, L. (2003). Dimensions of executive functioning: Evidence from children. British
Journal of Developmental Psychology, 21, 59–80.
Lezak, M. D. (1995). Neuropsychological assessment. New York: Oxford University Press.
Liston, C., Watts, R., Tottenham, N., Davidson, M. C., Niogi, S., Ulug, A. M., et al (2006). Frontostriatal microstructure modulates
efficient recruitment of cognitive control. Cerebral Cortex, 16, 553–560.
Luciana, M., & Nelson, C. A. (1998). The functional emergence of prefrontally-guided working memory systems in four- to eight-
year-old children. Neuropsychologia, 36, 273–293.
Luciana, M., & Nelson, C. A. (2002). Assessment of neuropsychological function through use of the Cambridge
Neuropsychological Testing Automated Battery: Performance in 4- to 12-year-old children. Developmental
Neuropsychology, 22, 595–624.
Luna, B., Garver, K. E., Urban, T. A., Lazar, N. A., & Sweeney, J. A. (2004). Maturation of cognitive processes from late childhood to
adulthood. Child Development, 75, 1357–1372.
MacLeod, C. M. (1991). Half a century of research on the Stroop effect: An integrative approach. Psychological Bulletin, 2,
163–203.
Mazur, J. E. (1987). An adjusting procedure for studying delayed reinforcement. In M. L. Commons, J. E. Mazur, J. A. Nevin, & H.
Rachlin (Eds.), Quantitative analyses of behavior, Vol. 5: The effect of delay and of intervening events on reinforcement value
(pp. 55–73). Hillsdale, NJ: Lawrence Erlbaum.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. (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.
Olson, E. A., & Luciana, M. (2008). The development of prefrontal cortex functions in adolescence: Theoretical models and a
possible dissociation of dorsal versus ventral subregions. In C. A. Nelson & M. Luciana (Eds.), Handbook of developmental
cognitive neuroscience (2nd ed., pp. 575–590). Cambridge, MA: MIT Press.
Overman, W. H. (2004). Sex differences in early childhood, adolescence, and adulthood, on tasks that rely on orbitofrontal
cortex. Brain and Cognition, 55, 134–147.
Overman, W. H., Frassrand, K., Ansel, S., Trawalter, S., Bies, B., & Redmond, A. (2004). Performance on the Iowa Card Task by
adolescents and adults. Neuropsychologia, 42, 1838–1851.
Pardo, J. V., Pardo, P. J., Janer, K. W., & Raichle, M. E. (1990). The anterior cingulate cortex mediates processing selection in the
Stroop attentional conflict paradigm. Proceedings of the National Academy of Sciences United States of America, 87, 256–259.
Prencipe, A., & Zelazo, P. D. (2005). Development of affective decision-making for self and other: Evidence for the integration of
first- and third-person perspectives. Psychological Science, 16, 501–505.
Richards, J. B., Zhang, L., Mitchell, S. H., & de Wit, H. (1999). Delay or probability discounting in a model of impulsive behavior:
Effect of alcohol. Journal of the Experimental Analysis of Behavior, 71, 121–143.
Séguin, J. R., & Zelazo, P. D. (2005). Executive function in early physical aggression. In R. E. Tremblay, W. W. Hartup, & J. Archer
(Eds.), Developmental origins of aggression (pp. 307–329). New York: Guilford.
Sowell, E. R., Thompson, P. M., Holmes, C. J., Jernigan, T. L., & Toga, A. W. (1999). In vivo evidence for post-adolescent brain
maturation in frontal and striatal regions. Nature Neuroscience, 2, 859–861.
Sowell, E. R., Trauner, D. A., Gamst, A., & Jernigan, T. L. (2002). Development of cortical and subcortical brain structures in
childhood and adolescence: A structural MRI study. Developmental Medicine & Child Neurology, 44, 4–16.
Steinberg, L. (2005). Cognitive and affective development in adolescence. Trends in Cognitive Sciences, 9, 69–74.
Steinberg, L., Graham, S., O’Brien, L., Woolard, J., Cauffman, E., & Banich, M. (2009). Age differences in future orientation and
delay discounting. Child Development, 80, 28–44.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 28, 643–662.
Uttl, B., & Graf, P. (1997). Color-Word Stroop test performance across the adult life span. Journal of Clinical and Experimental
Neuropsychology, 19, 405–420.
van Leijenhorst, L., & Crone, E. A. M. (2009). Paradoxes in adolescent risk taking. In P. D. Zelazo, M. Chandler, & E. Crone (Eds.),
Developmental social cognitive neuroscience (pp. 209–226). New York: Psychology Press.
Wechsler, D. (1991). Wechsler Intelligence Scale for Children (3rd ed.). San Antonio, TX: Psychological Corporation.
Welsh, M. C., Pennington, B. F., & Groisser, D. B. (1991). A normative-developmental study of executive function: A window on
prefrontal function in children. Developmental Neuropsychology, 7, 131–149.
 A. Prencipe et al. / Journal of Experimental Child Psychology 108 (2011) 621–637 637

Willcutt, E. G., Doyle, A. E., Nigg, J. T., Faraone, S. V., & Pennington, B. F. (2005). A meta-analytic review of the executive function
theory of ADHD. Biological Psychiatry, 57, 1336–1346.
Zelazo, P. D., Carlson, S. M., & Kesek, A. (2008). Development of executive function in childhood. In C. A. Nelson & M. Luciana
(Eds.), Handbook of developmental cognitive neuroscience (2nd ed., pp. 553–574). Cambridge, MA: MIT Press.
Zelazo, P. D., & Cunningham, W. (2007). Executive function: Mechanisms underlying emotion regulation. In J. Gross (Ed.),
Handbook of emotion regulation (pp. 135–158). New York: Guilford.
Zelazo, P. D., & Müller, U. (2002). Executive function in typical and atypical development. In U. Goswami (Ed.), Handbook of
childhood cognitive development (pp. 445–469). Oxford, UK: Blackwell.