Journal of Experimental Child Psychology 145 (2016) 120–138

Contents lists available at ScienceDirect

Journal of Experimental Child

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

The effects of fantastical pretend-play on the

development of executive functions: An
intervention study
Rachel B. Thibodeau a,⇑, Ansley T. Gilpin a, Melissa M. Brown b,
Brooke A. Meyer a
a
Department of Psychology, University of Alabama, Tuscaloosa, AL 35487, USA
b
Department of Family Sciences, Texas Woman’s University, Denton, TX 76204, USA

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

Article history: Although recent correlational studies have found a relationship

Received 29 June 2015 between fantasy orientation (FO; i.e., a child’s propensity to play
Revised 6 November 2015 in a fantastical realm) and higher order cognitive skills called exec-
utive functions (EFs), no work has addressed the causality and
directionality of this relationship. The current study experimen-
Keywords:
tally examined the directionality of the observed relationship
Executive functions
Fantasy orientation
between FO and EF development in preschool-aged children
Pretend-play through an innovative play intervention employing a randomized
Intervention controlled design. A sample of 110 children between the ages of
Cognitive development 3 and 5 years were randomly assigned to one of three conditions:
Pretense fantastical pretend-play intervention, non-imaginative play inter-
vention, or business-as-usual control. Results revealed that chil-
dren who participated in a 5-week fantastical pretend-play
intervention showed improvements in EFs, whereas children in
the other two conditions did not. Within the fantastical pretend-
play condition, children who were highly engaged in the play
and those who were highly fantastical demonstrated the greatest
gains in EFs. These data provide evidence for the equifinal relation-
ship between fantasy-oriented play and EF development, such that
engaging in fantasy-oriented play may be one of many ways to
directly enhance EF development.
Ó 2016 Elsevier Inc. All rights reserved.

⇑ Corresponding author.
E-mail address: rbthibodeau@crimson.ua.edu (R.B. Thibodeau).

http://dx.doi.org/10.1016/j.jecp.2016.01.001
0022-0965/Ó 2016 Elsevier Inc. All rights reserved.
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 121

Introduction

Due to recent declines in academic performance throughout the United States, there has been a
surge of research during the past 10 years aimed at understanding the development of cognitive con-
trol. This accumulation of research demonstrates the importance of cognitive control not only to
everyday functioning but also to academic achievement, job performance, and overall well-being
(Dunn, 2010; Eakin et al., 2004; Moffitt et al., 2011; St Clair-Thompson & Gathercole, 2006). One of
the main foci that has emerged from the body of research on cognitive control is a category of cogni-
tive abilities referred to as executive functions (EFs).
Executive functions are defined as higher order thinking processes that allow individuals to over-
ride more automatic thoughts and behaviors for more adaptive and goal-directed responses (Carlson,
2005). Some of the most important cognitive processes included under the umbrella term of executive
function are working memory, inhibitory control, and cognitive flexibility (Miyake et al., 2000; St
Clair-Thompson & Gathercole, 2006). Working memory is defined as the temporary storage of infor-
mation, which allows individuals to manipulate information as they cognitively process it
(Baddeley, 1983, 1992). Inhibitory control refers to individuals’ ability to suppress an automatic pre-
potent response (Stroop, 1935; Wright, Waterman, Prescott, & Murdoch-Eaton, 2003). Finally, cogni-
tive flexibility, also referred to as attentional shift, refers to individuals’ ability to shift their attention
back and forth between two different domains (Monsell, 1996).
Unfortunately, research indicates that EF abilities do not automatically develop and mature over
the lifespan but rather benefit from rich environmental experiences (Center on the Developing
Child at Harvard University, 2011). In fact, children raised in adverse environments (e.g., environments
characterized by abuse and neglect) exhibit serious deficits in cognitive, attentional, and behavioral
control, suggesting that EFs are at risk for disruption at an early age (Center on the Developing
Child at Harvard University, 2011). Therefore, the literature calls for research investigating the devel-
opment of EFs during early childhood.
As Vygotsky (1978) theorized, complex pretend-play may provide a natural environmental experi-
ence in which cognitive skills can be developed. Specifically, Vygotsky reasoned that imaginative play
is instrumental to the development of children’s ability to think about objects and events that are not
immediately present (i.e., internal systems of representation; Vygotsky, 1967). As children enter the
preschool years, their play becomes more abstract and less dependent on actual objects or props
(Woolley & Tullos, 2008). For example, children at this age will pretend to bounce an imaginary ball
or pretend to cook and eat an imaginary meal. Because they are able to use abstract concepts in their
play, children at this age are able to enrich their play with fantastical themes.
The term fantasy orientation (FO) refers to children’s propensity to play in a fantastical realm and is
often operationalized in children as the extent to which they engage in imaginary play and whether or
not they have imaginary companions (Taylor, 1999). Although there are clear individual differences in
children’s level of FO (Taylor & Carlson, 1997), to date few studies have investigated the impact of chil-
dren’s level of FO on their cognitive control. One possibility is that having more advanced EFs provides
individuals with the capacities needed to engage in fantasy-oriented behaviors. However, this explana-
tion is less likely given that incidence rates of high fantastical thinking are consistent during childhood
and adulthood (Woolley, 1997) and that fantasy is measured as a part of the openness personality trait
that is stable across the lifespan (McCrae, 1987, 1993). By contrast, research indicates that EFs develop
as individuals progress through childhood (Diamond & Taylor, 1996). Another possibility is that the act
of engaging in fantasy-oriented behaviors facilitates the development of EFs. Similar to the cognitive
flexibility needed for bilingualism, the act of engaging in imaginary play and having imaginary com-
panions requires that children switch back and forth between fantasy and reality (Estes, Wellman, &
Woolley, 1989; Golumb & Kuersten, 1996) and, thus, use working memory to remember pretense rules
and scripts, inhibit using pretense scripts in real life, and shift attention back and forth between reality
and pretense. Therefore, it seems logical that FO would be related in some fashion to EF development.
Sound empirical evidence of a relationship between FO and EF has only recently emerged. Specif-
ically, Pierucci, O’Brien, McInnis, Gilpin, and Barber (2014) interviewed preschoolers between the ages
122 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

of 4 and 6 years using measures of EF (i.e., inhibitory control, attention shift, and working memory)
and FO. Pierucci and colleagues found that children exhibiting high FO (i.e., engaging in fantasy-
oriented cognitions, belief in fantasy-oriented entities) displayed better cognitive inhibition and atten-
tional shift than children exhibiting low or moderate levels of FO. In addition, Carlson, White, and
Davis-Unger (2014) found positive correlations between EFs (i.e., inhibitory control, attention shift,
and working memory) and the understanding of pretend versus reality as well as the ability to per-
form pretend actions among preschool children. Furthermore, when subjected to a regression analysis,
Carlson and colleagues found that the understanding of pretense versus reality significantly predicted
EF scores and that the ability to perform pretend actions marginally predicted EF scores.
Despite the observed relationship between FO and EF in preschoolers, the correlational nature of
these findings leaves open questions regarding the directionality of these relationships (Carlson
et al., 2014; Pierucci et al., 2014). A recent meta-analysis in the field called for experimental investi-
gations of the effects of pretend-play on cognitive control (Lillard et al., 2013) because many previous
studies that have explored EFs in the context of pretend-play have several methodological limitations
and, therefore, are inconclusive. In addition, all previous experimental studies on play and EFs have
specifically investigated pretense (Lillard et al., 2013). Pretense is a type of imaginative play in which
children pretend to be an animal or another person (Woolley & Tullos, 2008) but does not necessarily
involve the same fantastical elements as fantasy-oriented activities and cognitions. For example, pre-
tense could involve pretending to drive a car, whereas fantasy-oriented play would involve flying a car
in outer space. Interestingly, Pierucci and colleagues (2014) found that, unlike fantasy-oriented activ-
ities and cognitions, pretense did not correlate with any measures of EF. Thus, it appears that fantasy
may be a crucial element to the observed relationship between play and EFs.
Therefore, the purpose of the current study was to experimentally determine the causality and
directionality of the observed relationship between FO and EFs through the training of fantasy-
oriented play in preschool-aged children. To accomplish this, the first aim of the current study was
to investigate how pretend-play with a fantastical component affects development differentially from
non-imaginative play. The second aim was to examine how fantasy-oriented imaginative play differs
from maturation alone with regard to the development of EFs. A sample of 110 preschoolers between
the ages of 3 and 5 years were pre-tested on several measures of FO, pretense, and EF (i.e., working
memory, inhibitory control, and attention shift). Participants were then randomly assigned to one
of three conditions: fantastical pretend-play, non-imaginative play, or control. After 5 weeks of inter-
vention, participants were tested using similar pre-test measures of EF and FO. Pre-test to post-test
changes in EF and FO scores were compared among conditions.
Because previous research demonstrates a relationship between being fantasy oriented and perfor-
mance on various EF tasks (Pierucci et al., 2014), it was hypothesized that engaging in fantastical
pretend-play would facilitate the development of EFs. Specifically, it was expected that there would
be a significant difference among intervention conditions on post-test measures of EFs, such that chil-
dren in the fantastical pretend-play condition would show an increase in EF abilities beyond that of
children in the non-imaginative play and no-intervention conditions. It should be noted, however, that
FO was not hypothesized to change over the course of the intervention period. In fact, it is highly unli-
kely that FO would show any changes over time due to its perceived stability. Rather, it was expected
that engaging in fantastical pretend-play would create an enriched environment through which EFs
could be scaffolded.

Method

Participants

A total of 121 participants were recruited for the current study. Of these children, 2 were excluded
from the study because they scored below the 20th percentile on the Peabody Picture Vocabulary Test,
a measure of receptive vocabulary, and 9 were excluded because they did not complete post-test
assessments (5 from the non-imaginative play condition and 4 from the control condition). Therefore,
the final sample of the current study consisted of 110 preschoolers between the ages of 3 and 5 years
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 123

Table 1
Demographics data.

Fantastical pretend-play Non-imaginative play Control Total

Gender
Boys 49% (n = 19) 50% (n = 16) 41% (n = 16) 46% (n = 51)
Girls 51% (n = 20) 50% (n = 16) 59% (n = 23) 54% (n = 59)
Age (in 50.62 54.06 52.37 52.22
months)
Vocabulary 106.74 110.93 109.36 108.84
Ethnicity
Caucasian 95% (n = 37) 88% (n = 28) 85% (n = 33) 89% (n = 98)
African 2.5% (n = 1) 6% (n = 2) 10% (n = 4) 6% (n = 7)
American
Other 2.5% (n = 1) 6% (n = 2) 5% (n = 2) 5% (n = 5)
Parent M = 4-year college or M = 4-year college or M = 4-year college or M = 4-year college or
education bachelor’s degree bachelor’s degree bachelor’s degree bachelor’s degree
Income M = $90,000–$99,999 M = $80,000–$89,999 M = $90,000–$99,999 M = $80,000–$89,999

(Mage = 52 months, SD = 9.70; 59 girls and 51 boys). Most (89%) of the children were Caucasian, 6%
were African American, and 5% were ‘‘other” or not reported. All participants were recruited from local
preschools and day-care centers in the southern United States. The mean level of parental education
for this sample was 4 years of college with annual family income ranging from $10,000 to more than
$100,000 (M = $80,000–$89,999) (see Table 1).

Measures

Data were obtained pre- and post-intervention from direct child assessment, teacher question-
naires, and interventionist report about children’s pretense and fantasy orientation, their executive
functions, and their play style and engagement.

Pre-intervention measures
Teacher questionnaires. Teachers completed a Fantasy Orientation Questionnaire (Gilpin, 2009) in
order to assess each child’s FO level. Teachers were asked about each child’s belief in various fantas-
tical figures (e.g., Santa Claus, tooth fairy) and their favorite books, games, and activities. Answers to
the belief questions received a score of 1 if teachers indicated that the child believed the entity was
real, .5 if they thought the child was unsure of the entity’s existence, and 0 if they thought the child
believed the entity was pretend. The total score for this first item was calculated as a percentage of the
total points possible (determined by the number of questions answered) in order to accommodate
responses that were left blank by teacher respondents. Teacher reports of a child’s favorite books,
games, and so forth were coded from a standardized list of responses, with 0 indicating a reality-
based response, 1 indicating a response that was representational or anthropomorphic but not fantas-
tical, and 2 indicating a high fantasy-based response. These questions were independently coded by
two trained raters, and discrepancies were resolved by a third coder (jpre = .753, jpost = .802). In addi-
tion, teachers were asked to rate a child’s level of FO using a 5-point scale, with 1 indicating a strong
interest in reality (e.g., plays sports) and 5 indicating a strong interest in fantasy (e.g., often engages in
pretense, enjoys fantastical books). Finally, teachers were asked to indicate whether or not a child had
an imaginary friend. All of these responses were used separately as individual reports from teachers
about a child’s FO. Teachers were not specifically instructed to skip items they did not know, so items
left blank by teachers were considered missing data.

Child interview fantasy orientation and pretense measures. In addition to teacher reports of a child’s FO,
two child questionnaires were used to assess participants’ level of FO: Singer and Singer’s (1981)
Imaginative Play Predisposition interview and Taylor and Carlson’s (1997) Imaginary Companion
interview. These interviews asked children questions about the fantastical nature of their play, their
124 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

thoughts, and their play preferences as well as questions about their experiences with an imaginary
companion. Responses indicating interest in fantastical toys, stories, and/or games (e.g., fairies, super-
heroes) were given a score of 2; responses involving anthropomorphized/animated toys, stories, and/
or games (e.g., stuffed animals, toy cars) were given a score of 1; and responses involving realistic toys,
stories, and/or games (e.g., checkers, Bingo) were given a score of 0. These questions were indepen-
dently coded by two trained raters, and discrepancies were resolved by a third coder (jpre = .915,
jpost = .806). Children who reported having an imaginary companion and who could describe it
received a score of 1 on the Imaginary Companion interview.
Pretense, or pretending based in reality, was measured using Taylor and Carlson’s (1997) Imperson-
ation Interview and the Toy Phone task (Tahiroglu, Mannering, & Taylor, 2011; Taylor, Sachet, Maring,
& Mannering, 2013). During the Impersonation Interview, children were asked whether they pretend
to be an animal, a different person, or anything unrelated to their self (e.g., an airplane). Participants
received a score of 1 for each question to which they answered ‘‘yes”, with a total possible score of
3. During the Toy Phone task (Tahiroglu et al., 2011; Taylor et al., 2013), participants were presented
with a toy phone and asked whether they would like to pretend to call a friend with whom they like
to play. Children were then allowed to play with the phone as they wished. During this time, the exper-
imenter recorded children’s behaviors on five dimensions: dialing a number, holding the phone to their
ear, talking on the phone, listening to the other person, and generating a conversation that went
beyond standard greetings such as ‘‘hi” and ‘‘how are you?” For each action that children performed,
they received a score of 1. The total score for this measure ranged from 0 to 5. Higher scores indicate
a greater propensity to engaging in pretense. This measure was independently coded by two trained
raters, and discrepancies were resolved by a third coder (jpre = .972, jpost = .790).

Child interview EF measures. To assess working memory, the Forward Digit Span task was used (Davis
& Pratt, 1996). During this task, the experimenter spoke a series of digits (0–9), and children were
instructed to repeat the digits back to the experimenter. The trials started with two digits and became
progressively longer as children correctly recalled the previous number of digits. A child’s total score
on this task was equal to the highest number of digits that he or she successfully repeated.
The Day/Night task was used to measure inhibitory control (Gerstadt, Hong, & Diamond, 1994).
This Stroop-like task presents participants with two pictures: a picture of a moon and stars (i.e.,
‘‘nighttime”) and a picture of a blue sky and grassy field (i.e., ‘‘daytime”). Participants were instructed
to say ‘‘day” when the experimenter pointed to the picture of the moon and stars and to say ‘‘night”
when the experimenter pointed to the picture of the blue sky and grassy field. The numbers of correct,
incorrect, and self-corrected (i.e., initially incorrect but self-corrected to appropriate answer)
responses were recorded. Correct answers were given a score of 1, self-correct answers were given
a score of .5, and incorrect answers were given a score of 0, resulting in a maximum possible score
of 16. Higher scores on this measure indicate better cognitive inhibition.
Finally, participants’ ability to shift their attention was assessed using the standard Dimensional
Change Card Sort (Card Sort) task (Zelazo, 2006; Zelazo, Muller, Frye, & Marcovitch, 2003), following
the procedures of Pierucci and colleagues’ (2014) study. Children were presented with two small
boxes labeled with two different target images (e.g., a blue square and a red star). The experimenter
began by explaining the shape game. Children were told to place all of the star cards in the container
labeled with the red star and to place all of the square cards in the container labeled with the blue
square. After children correctly sorted five cards consecutively (maximum of 25 trials), the experi-
menter told them to stop playing the shape game and start playing the color game. The experimenter
explained to children that they should now place all of the blue cards in the container labeled with a
blue square and place all of the red cards in the container labeled with a red star. The task ended when
children correctly sorted five cards consecutively (maximum of 25 trials). Children were not reminded
of the rules after the task started, and the order of the color/shape game was randomized. This task
was scored as the number of inaccurate card sorts after the rule switch, with higher scores indicating
poorer attention shift performance.

Vocabulary measure. Because previous research demonstrates that vocabulary is strongly related to EF
performance in preschool children (Carlson & Moses, 2001), vocabulary level was assessed during
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 125

pre-testing in order to ensure that the experimental conditions did not differ on vocabulary level prior
to the intervention. The Peabody Picture Vocabulary Test–Fourth Edition, Form B (PPVT-4; Dunn &
Dunn, 2007), was administered to participants. Children were shown a colored book with four pictures
on each page. Children were instructed to point to the picture that best matched the word the exper-
imenter spoke. The PPVT-4 consists of 19 sets with 12 items in each set for a total of 228 items. Par-
ticipants were tested until they incorrectly identified eight vocabulary words within one set of 12
items. The PPVT-4 was coded using standard procedures outlined in the Form B manual, and the stan-
dardized score was used to assess children’s receptive vocabulary.

Post-intervention measures
During post-test assessment, teachers and children completed the same EF and FO/pretense mea-
sures used during the pre-test period. The Grass/Snow task was used as a measure of inhibitory control
during post-testing in place of the Day/Night task used during pre-test in order to rule out practice
effects (Carlson & Moses, 2001). Similar to the Day/Night task, this Stroop-like task presented partici-
pants with two pictures: a picture of a grassy field (i.e., ‘‘grass”) and a picture of a snowy field (i.e.,
‘‘snow”). The Grass/Snow task was administered and scored in the same way as the Day/Night task with
the same number of trials (i.e., 16). Higher scores indicate better cognitive inhibition performance.

Procedure

Research assistants interviewed children individually on the pre-intervention measures described

above in order to assess baseline levels of FO and EF. The order of administration of the measures was
randomized across participants. Children received stickers for participating. Teachers also completed
pretense and FO questionnaires at this baseline period.
Participants were then randomly assigned to one of three intervention conditions: fantastical
pretend-play, non-imaginative play, and control. Even though children were recruited from schools,
the conditions were not nested in classrooms. Teachers and research assistants who completed the
assessments were blind to the condition of the participants and the purpose of the study. Prior to
the intervention, we ensured that the conditions did not differ on pre-test EF, FO, or vocabulary levels.
In the fantastical pretend-play condition, research assistants worked with small groups of children
(n = 3–6) to come up with a fantastical script and then encouraged them to act it out. For example,
children may have pretended that they were birds flying around a forest or pretended to go on an
adventure to the moon and interact with space creatures. Children were encouraged to generate
the play scripts, but research assistants provided scaffolding when needed (e.g., suggestions, encour-
agement, questions that inspired imagination such as ‘‘What do you see on the moon?”). In the non-
imaginative play condition, small groups of participants (n = 3–6) engaged in action-oriented group
activities that did not require any imagination and were not designed to specifically tap EF skills.
Research assistants were given a list of commonly played activities including, but not limited to, four
corners, action songs (e.g., ‘‘Wheels on the Bus”), coloring, and ball games. It is possible that some of
the activities required some amount of EF, but the games chosen did not heavily tax EFs. For example,
some commonly played games such as Simon Says heavily rely on EF skills such as inhibitory control.
Such games were not included in the non-imaginative play intervention condition. Similar to the fan-
tastical pretend-play condition, research assistants offered encouragement throughout the play ses-
sions and often asked questions to inspire engagement (e.g., ‘‘How should we pass the ball around
the circle this time?”). Games in the fantastical pretend-play and non-imaginative play conditions
were designed to be similar in terms of child interest, activity level, and opportunities for child-
initiated choices. Finally, in the control condition, participants did not participate in any intervention
games, proceeding with their normal classroom activities as usual.
Throughout the intervention period, research assistants who led the intervention groups (who
were also blind to the purpose of the study) were asked to reflect on children’s behavior and come
to a consensus recording what occurred during each training session using a daily log provided by
the researchers (i.e., how engaged children were and how realistic or fantastical their play was). Each
day of the intervention, children received a 2 if they were highly engaged, a 1 if they were somewhat
engaged, and a 0 if they were not engaged in the play. Scores were averaged across all 25 days of the
126 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

intervention. Similarly, each day children’s style of play was rated on a 3-point scale. Children received
a 2 if their play was highly fantastical (e.g., involving fairies, dragons, and/or princesses), a 1 if their
play was pretense based (i.e., imaginative but not necessarily fantastical such as pretending to cook
or pretending to be an animal), and a 0 if their play was reality based (i.e., non-imaginative). Scores
were averaged across all 25 days of the intervention.
Based on previous EF training studies that have successfully increased EF (Bergman Nutley et al.,
2011; Thorell, Lindqvist, Bergman Nutley, Bohlin, & Klingberg, 2009), the fantastical pretend-play
and non-imaginative play conditions underwent 5 weeks of intervention. Training occurred every
school day for a total of 25 training sessions that lasted 15 min each. During each session, small groups
of children (3–6 children per group) mixed between classrooms played one game such as pretending
to fly to the moon (fantastical pretend-play condition) or four corners (non-imaginative play condi-
tion). Group size differed daily based on school attendance. After 5 weeks of intervention, all partic-
ipants underwent post-testing using the measures described above. Interviewers and teachers were
blind to the hypotheses of the study and the conditions of the children.

Results

Preliminary analyses

There were a total of 39 children in the fantastical pretend-play condition (19 male and 20 female),
32 children in the non-imaginative play condition (16 male and 16 female), and 39 children in the con-
trol condition (16 male and 23 female). The average age of children was 51 months (SD = 10.09) in the
fantastical pretend-play condition, 54 months (SD = 9.09) in the non-imaginative play condition, and
52 months (SD = 9.74) in the control condition. The conditions did not differ significantly on age, gen-
der, or ethnicity and were evenly distributed from the 20 classrooms and four preschools included in
the study (see Table 1). In addition, there were no differences in PPVT vocabulary scores among the
play conditions prior to the intervention, F(2, 100) = 1.66, p = .20, g2 = .032.
Before any pre-test analyses were conducted, FO items were categorized into distinct components
using a principal component analysis (PCA) with varimax rotation, as has been done in previous
research to form FO constructs (Pierucci et al., 2014). Previous literature has found four FO constructs:
FO cognitions, FO toys and games, FO entities, and FO pretense (Pierucci et al., 2014). Our data replicated
these four constructs and added a fifth with Toy Phone scores (not included in Pierucci et al.’s [2014]
battery) loading onto a separate FO pretense behaviors construct (eigenvalues = 1.489, 1.862, 1.673,
2.652, and 3.772, respectively). The five constructs were created by averaging the z-scores of each mea-
sure in the component. All future analyses use these five FO constructs rather than individual FO items.
To examine any pre-existing differences among the play conditions prior to the intervention, a ser-
ies of one-way analyses of variance (ANOVAs) comparing all three conditions was conducted on all
pre-intervention FO, EF, and vocabulary measures. See Table 2 for pre-intervention descriptive data,
and see Table 3 for correlations among all measures. Groups did not differ on any measures prior to
the intervention with the exception of the FO cognitions construct, F(2, 105) = 3.32, p = .04,
g2 = .059. Specifically, post hoc analyses indicated that, prior to the intervention, the control condition
reported engaging in marginally fewer FO cognitions (M =
.251, SD = .770) than the fantastical
pretend-play condition (M = .132, SD = .682), p = .058, d =
0.527. No differences were found with
the non-imaginative play condition (M = .116, SD = .706). On all future post-test analyses comparing
the play conditions, pre-test scores are included as a covariate.
Finally, throughout the intervention, attendance records (e.g., intervention dosage) were recorded.
The mean number of days missed out of the 25-day intervention was 5.7 days, with a range from 1 to
17 days absent. There was no effect of dosage on any of the variables of interest; thus, it is not con-
sidered in future analyses.

Differences among conditions at post-test

At post-test, FO/pretense and EF scores were compared among all three intervention conditions using
a series of analyses of covariance (ANCOVAs) with pre-intervention scores included as covariates.
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 127

Table 2
Descriptive data.

Measure Condition Possible npre Mpre SDpre npost Mpost SDpost

range
FO/pretense measures
FO Cognitions Fantastical
1 to +1 39 .132 .682 39 .123 .820
Non-imagination 32 .116 .706 30
.201 .702
Control 37
.251 .770 35 .037 .687
FO Toys & Games Fantastical
1 to +1 39
.013 .553 39 .089 .518
Non-imagination 32 .027 .595 30
.076 .467
Control 37
.010 .554 35
.042 .485
FO Pretense Fantastical
1 to +1 39 .041 .737 39 .096 .733
Non-imagination 32 .126 .813 29
.151 .755
Control 37
.151 .733 33 .036 .693
FO Entities Fantastical
1 to +1 39
.074 .704 39 .117 .716
Non-imagination 32 .155 .690 32 .020 .636
Control 39
.101 .726 39
.105 .525
FO Pretense Fantastical
1 to +1 39
.034 .704 39 .255 .699
Behaviors Non-imagination 32 .120 .667 31
.041 .751
Control 37
.072 .756 35
.233 .744

EF measures
Working Memory (FDS) Fantastical 0 to 7 39 3.95 .887 39 4.69 1.76
Non-imagination 31 4.06 1.03 31 3.94 1.26
Control 37 4.30 1.05 35 4.38 1.52
Attention Shift (CS errors) Fantastical 0 to 25 38 2.89 4.69 38 1.58 3.21
Non-imagination 30 1.57 3.11 28 3.14 4.78
Control 36 3.14 3.70 35 2.83 4.00
Inhibitory Control (Day/Night & Fantastical 0 to 16 27 12.3 4.31 34 11.5 4.72
Grass/Snow) Non-imagination 26 13.5 2.49 26 11.7 4.40
Control 26 10.8 5.04 31 12.0 5.52

Note. The means in the table reflect the full-sample raw means. Means reported in the text may be slightly different because
ANCOVA analyses corrected for pre-test scores and the repeated-measures analyses considered only children who had both pre-
and post-intervention scores. FDS, Forward Digit Span task number of digits recalled; CS Errors, Card Sort task number of errors
after rule switch.

Planned contrasts were performed on all findings, comparing performance in the fantastical pretend-
play condition with performance in the other two conditions. In addition, to explore the nature of
changes observed, post hoc repeated-measures analyses comparing pre- and post-test scores within
a condition were performed on significant findings. See Table 2 for all post-intervention descriptive
data.

FO/pretense post-test differences

To assess the hypothesis that FO would not change throughout the course of the intervention
among any conditions, a series of ANCOVAs controlling for pre-test FO/pretense scores was conducted
among the three play conditions with the post-test FO constructs as the outcome variables. Among all
FO/pretense constructs, only one significant difference among conditions was found at post-test. An
ANCOVA controlling for pre-test scores as a covariate revealed a significant difference between inter-
vention conditions on post-test FO pretense behaviors scores, F(2, 100) = 4.63, p = .012, g2 = .085.
Planned contrasts among the three conditions revealed a significant difference between the fantastical
pretend-play and non-imaginative play conditions (p = .022, d = 0.560) and between the fantastical
pretend-play and control conditions (p = .006, d = 0.654). In other words, after the intervention, chil-
dren who had participated in the fantastical pretend-play condition engaged in more pretending
behaviors (M = .261, SD = .656) than children in the non-imaginative play condition (M =
.108,
SD = .663) and the control condition (M =
.169, SD = .659) (see Fig. 1).
128 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

Table 3
Correlations among all measures.

Pre-test variable 1 2 3 4 5 6 7
1 FO Cognitions
2 FO Toys & Games .155
3 FO Pretense .279** .161
4 FO Entities .241* .065 .085
5 FO Pretense Behaviors .120 .031 .259**
.043
6 Working Memory
.013 .276**
.046
.032 .223*
7 Attention Shift .084 .152 .112 .011
.110
.297**
8 Inhibitory Control .148
.001 .113 .023
.106 .135
.234*
Post-test variable 1 2 3 4 5 6 7
1 FO Cognitions
2 FO Toys & Games .173
3 FO Pretense .189 .060
4 FO Entities .166 .017 .189
5 FO Pretense Behaviors .070 .118 .242* .060
6 Working Memory .197* .271**
.052 .077 .077
7 Attention Shift
.158 .008 .052
.076
.163
.325**
8 Inhibitory Control .049 .198
.031 .019 .034 .420**
.284*
*
p < .05.
**
p < .01.

1
0.8 Fantastical Pretend-Play
Non-Imaginative Play **
0.6 Control **

0.4
Post-test z-scores

0.2
0
-0.2
-0.4
-0.6
-0.8
FO FO Toys FO FO FO Pretense
-1 Cognitions & Games Pretense Entities Behaviors

Fig. 1. Difference among play conditions on FO measures at post-test after controlling for pre-test scores. Errors bars reflect 95%
confidence interval. After the intervention, children in the fantastical pretend-play condition were engaging in significantly
more pretending behaviors than children in the non-imaginative play and control conditions. **p < .01.

To observe the nature of the observed changes, post hoc repeated-measures analyses comparing
pre- and post-test FO pretense behaviors scores in one condition at a time were performed. In the fan-
tastical pretend-play condition, there was a significant increase in pretending behaviors from pre-test
(M =
.034, SD = .704) to post-test (M = .255, SD = .699), F(1, 38) = 4.68, p = .037, g2 = .110. By contrast,
there were no changes in pretending behaviors from pre-test to post-test in the non-imaginative play
condition (Mpre = .123, SDpre = .678; Mpost = –.041, SDpost = .751), F(1, 30) = 1.31, p = .262, g2 = .042, or
the control condition (Mpre = –.139, SDpre = .747; Mpost = –.224, SDpost = .754), F(1, 33) = .674, p = .418,
g2 = .020 (see Fig. 2). This serves as a manipulation check demonstrating that the fantastical
pretend-play condition, but not the other conditions, encouraged pretending behaviors.
As expected, there were no differences among play conditions on FO cognitions, F(2, 99) = 2.23,
p = .112, g2 = .043, FO toys and games, F(2, 99) = 1.36, p = .262, g2 = .027, FO pretense, F(2, 96) =
2.35, p = .100, g2 = .047, or FO entities, F(2, 106) = 1.48, p = .232, g2 = .027 (see Fig. 1).
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 129

1
Fantastical Pretend-Play
0.8 Non-Imaginative Play

FO pretense behaviors z-scores

0.6 Control
0.4
0.2 *

0
Pre-Test Post-Test
-0.2
-0.4
-0.6
-0.8
-1

Fig. 2. Follow-up repeated-measures analyses to examine the nature of the changes on the FO Pretense Behaviors construct
from pre-intervention to post-intervention in each play condition. There was a significant increase in pretending behaviors from
pre-intervention to post-intervention in the fantastical pretend-play condition. *p < .05.

EF post-test differences
To assess the hypothesis that there would be significant differences among intervention conditions
on post-test measures of EFs, such that children in the fantastical pretend-play condition would show
an increase in EF abilities beyond that of children in the non-imaginative play and no-intervention
conditions, a series of ANCOVAs controlling for pre-test EF scores was conducted among the three play
conditions on EF post-test variables. As expected, an ANCOVA controlling for pre-test Forward Digit
Span scores revealed a significant post-test difference among play conditions on this measure of work-
ing memory, F(2, 99) = 3.16, p = .047, g2 = .060. Planned contrasts among the three conditions revealed
a significant difference between the fantastical pretend-play and non-imaginative play conditions
(p = .017, d = 0.590). In other words, after the intervention, children who had participated in the fan-
tastical pretend-play condition recalled significantly more digits (M = 4.81, SD = 1.34) than children in
the non-imaginative play condition (M = 4.02, SD = 1.34). There was no significant difference between
the fantastical pretend-play and control conditions (M = 4.28, SD = 1.35, p = .099, d = 0.394) (see Fig. 3).
To observe the nature of the observed changes, post hoc repeated-measures analyses comparing
pre- and post-test Forward Digit Span scores were performed in one condition at a time. In the fantas-
tical pretend-play condition, there was a significant increase in working memory performance from
pre-test (M = 3.95, SD = .887) to post-test (M = 4.69, SD = 1.76), F(1, 38) = 10.41, p = .003, g2 = .215.
By contrast, there were no changes in working memory performance from pre-test to post-test in
the non-imaginative play condition (Mpre = 4.03, SDpre = 1.03; Mpost = 3.97, SDpost = 1.27), F(1, 29)
= 0.065, p = .801, g2 = .002, or the control condition (Mpre = 4.32, SDpre = 1.09; Mpost = 4.45,
SDpost = 1.49), F(1, 33) = 0.420, p = .522, g2 = .013 (see Fig. 4).
With regard to attention shift, an ANCOVA controlling for pre-test Card Sort task scores did not
reveal an overall significant post-test difference among the three play conditions, F(2, 92) = 2.59,
p = .080, g2 = .053. However, planned contrasts among the three conditions did reveal a significant dif-
ference between the fantastical pretend-play and non-imaginative play conditions (p = .031,
d = –0.561). In other words, after the intervention, children who had participated in the fantastical
pretend-play condition made significantly fewer errors after a rule switch (M = 1.44, SD = 3.89) than
children in the non-imaginative play condition (M = 3.63, SD = 3.92). There was no significant differ-
ence between the fantastical pretend-play and control conditions (M = 2.86, SD = 3.90, p = .131,
d = –0.365) (see Fig. 3). To perceive the nature of the observed changes, post hoc repeated-
measures analyses comparing pre- and post-test Card Sort task scores were performed within each
condition. There were no significant changes in attention shift scores from pre-test to post-test in
the fantastical pretend-play condition (Mpre = 2.84, SDpre = 4.75; Mpost = 1.51, SDpost = 3.23), F(1, 36)
= 2.54, p = .119, g2 = .066, the non-imaginative play condition (Mpre = 1.65, SDpre = 3.33; Mpost = 3.38,
130 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

18 Fantastical Pretend-Play
16 Non-Imaginative Play
Control
14
12

Post-test scores 10
8 * *
6
4
2
0
Working Memory Attention Shift Errors Inhibitory Control

Fig. 3. Difference among play conditions on EF measures at post-test after controlling for pre-test scores. Errors bars reflect 95%
confidence interval. Children in the fantastical pretend-play condition performed significantly better on the measure of working
memory and made fewer attention shift errors than children in the non-imaginative play condition after the intervention.
*
p < .05.

7 Fantastical Pretend-Play
Non-Imaginative Play
6
Highest number of digits recalled

Control
5
**
4

0
Pre-Test Post-Test

Fig. 4. Follow-up repeated-measures analyses to examine the nature of the changes on the Forward Digit Span task of working
memory from pre-intervention to post-intervention in each play condition. Throughout the course of the intervention, there
was a significant increase in working memory performance in the fantastical pretend-play condition but no changes in working
memory performance in the other two conditions. **p < .01.

SDpost = 4.88), F(1, 25) = 2.34, p = .139, g2 = .085, or the control condition (Mpre = 2.97, SDpre = 3.67;
Mpost = 2.97, SDpost = 4.08), F(1, 32) = 0.000, p = 1.00, g2 = .000 (see Fig. 5).
Finally, an ANCOVA controlling for pre-test inhibitory control scores (measures Day/Night and
Grass/Snow tasks) did not reveal an overall significant post-test difference among the three play
conditions on inhibitory control performance, F(2, 67) = 1.76, p = .180, g2 = .050. Similarly, planned
contrasts among the three conditions did not reveal any significant differences (see Fig. 3).

Individual differences in related variables’ influence on primary analyses

In addition to collecting data before and after the intervention relating to FO/pretense and EFs,
observations were made throughout the intervention on several aspects of children’s play, including
the quality of children’s play (i.e., how engaged children were) and how fantastical/pretense based
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 131

25
Fantastical Pretend-Play
Non-Imaginative Play
Control
20

Errors after rule switch

15

10

0
Pre-Test Post-Test

Fig. 5. Follow-up repeated-measures analyses to examine the nature of the changes on the Card Sort task of attention shift from
pre-intervention to post-intervention in each play condition. There were no significant changes in attention shift performance
throughout the course of the intervention.

(i.e., style of play) children’s play was. After each play session, research assistants, blind to the
hypotheses of the study, rated each child on these two dimensions. Because the data above indicate
advantages to engaging in fantasy-oriented play over non-imaginative play, we were specifically inter-
ested in determining what aspects of the fantastical pretend-play intervention were driving these ben-
efits. Analyses were conducted on how individual differences in engagement in play and style of play
related to the outcome variables in the fantastical pretend-play condition.

Differences in play engagement

Prior to comparing post-intervention differences between children who were highly engaged and
children who were somewhat engaged in the fantastical pretend-play condition, a series of one-way
ANOVAs was calculated to determine whether there were any pre-existing differences between these
two groups within the fantastical pretend-play condition. No child received an average score of 0, so
the following analyses examine only differences between children who received an average score of 2
(highly engaged) and children who received an average score of 1 (somewhat engaged). There were no
differences on any measures of FO/pretense or EF prior to the intervention between children in the
fantastical pretend-play condition who were highly engaged and children in the fantastical
pretend-play condition who were somewhat engaged throughout the intervention.
A series of ANCOVAs controlling for pre-test FO/pretense and EF scores were conducted between
the two levels of engagement in the fantastical pretend-play condition on all post-test variables. Inter-
estingly, at post-test there was a significant difference between children in the fantastical pretend-
play condition who were highly engaged in the fantastical pretend-play intervention and children
who were only somewhat engaged in the FO pretense construct, F(1, 36) = 6.68, p = .014, g2 = .156.
Specifically, children who were highly engaged reported engaging in more pretense (e.g., pretending
to be an animal) (M = .305, SD = .632) than children who were somewhat engaged (M =
.237,
SD = .635) after the intervention (see Fig. 6).
There was also a marginal difference at post-test on the FO entities construct, such that children in
the fantastical pretend-play condition who were highly engaged throughout the intervention were
more likely to report believing in fantastical entities (e.g., Santa Claus, tooth fairy) and having an imag-
inary companion (M = .266, SD = .607) than children who were somewhat engaged (M =
.120,
SD = .612), F(1, 36) = 3.59, p = .066, g2 = .091 (see Fig. 6).
Similarly, there was a marginal difference on post-test Forward Digit Span scores between levels of
engagement in the fantastical pretend-play condition, F(1, 36) = 3.53, p = .068, g2 = .089. Children who
were highly engaged throughout the intervention tended to remember more digits (M = 5.03,
SD = 1.41) than children who were somewhat engaged (M = 4.16, SD = 1.41). Level of engagement
132 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

1
0.8 †
*
0.6
0.4
Post-test z-scores

0.2
Highly
0 Engaged

-0.2 Somewhat
Engaged
-0.4
-0.6
-0.8 FO FO Toys FO FO FO Pretense
Cognitions & Games Pretense Entities Behaviors
-1

Fig. 6. Difference between children in the fantastical pretend-play condition who were highly engaged and those who were
somewhat engaged on FO measures at post-test after controlling for pre-test scores. Errors bars reflect 95% confidence interval.
Children in the fantastical pretend-play condition who were highly engaged reported higher beliefs in fantastical entities and
engaged in more pretense than children in the fantastical pretend-play condition who were somewhat engaged. *p < .05;
y
p < .07.

did not influence FO pretense behaviors, FO cognitions, FO toys and games, Card Sort task, or Grass/
Snow task scores (see Fig. 7).
To rule out the possibility that the observed benefits in EF observed in the fantastical pretend-play
condition were driven by differences among conditions in engagement level, post-test EF scores were
compared between highly engaged children in the fantastical pretend-play condition and highly
engaged children in the non-imaginative play condition. Even after equating levels of engagement
between these two conditions, we still see that children in the fantastical pretend-play condition
are performing better on the measure of working memory, F(1, 49) = 8.71, p = .005, g2 = .151, and mar-
ginally better on the measure of attention shift, F(1, 45) = 3.83, p = .056, g2 = .078. These results, which
mirror the results above, indicate that children who were highly engaged do not show equivalent
gains across the two intervention conditions, thereby further highlighting the observed benefits of
engaging in fantastical pretend-play.

Differences in style of play

Prior to comparing post-intervention differences between children in the fantastical pretend-play
condition who consistently engaged in fantastical play and children in the fantastical pretend-play
condition who consistently engaged in pretense-based play (i.e., realistic sociodramatic play), a series
of one-way ANOVAs was calculated to determine whether there were any pre-existing differences
between these two groups within the fantastical pretend-play condition. No child received an average
score of 0, so the following analyses examine only differences between children who received an aver-
age score of 2 (fantastical) and children who received an average score of 1 (pretense based). Prior to
the intervention, there were no differences on any FO, pretense, or EF measures between these two
subsets of children. It is also important to note that within the fantastical pretend-play condition,
there was no relationship between children who scored high on pre-intervention measures of FO/pre-
tense and their average style rating (all v2 values P .225, all ps P .130). In other words, all children in
the fantastical pretend-play condition were equally likely to engage in highly fantastical play regard-
less of their initial trait FO.
An ANCOVA controlling for pre-test Forward Digit Span scores revealed a significant difference in
working memory performance between children in the fantastical pretend-play condition who consis-
tently engaged in highly fantastical play and children in the fantastical pretend-play condition who
typically engaged in more pretense-based play, F(1, 35) = 6.48, p = .015, g2 = .156. Specifically, children
in the fantastical pretend-play condition who were more fantastical throughout the intervention
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 133

16

14

12
Post-test scores
10

8 † Highly
Engaged
6
Somewhat
4 Engaged

0
Working Memory Attention Shift Errors Inhibitory Control

Fig. 7. Difference between children in the fantastical pretend-play condition who were highly engaged and those who were
somewhat engaged on EF measures at post-test after controlling for pre-test scores. Errors bars reflect 95% confidence interval.
Children who were highly engaged tended to perform better on the measure of working memory after the intervention than
children who were somewhat engaged. yp < .07.

16

14

12
Post-test Scores

10

8 *
Fantastical
6
Pretense-Based
4

0
Working Memory Attention Shift Errors Inhibitory Control

Fig. 8. Differences in post-test EF scores between children in the fantastical pretend-play condition who consistently engaged
in fantastical play and children in the fantastical pretend-play condition who consistently engaged in pretense-based play.
Errors bars reflect 95% confidence interval. Children in the fantastical pretend-play condition who consistently engaged in
fantastical pretend-play performed significantly better on the measure of working memory than children in the fantastical
pretend-play condition who engaged in pretense-based pretend-play. *p < .05.

period remembered significantly more digits on the Forward Digit Span task (i.e., better working
memory) (M = 5.16, SD = 1.36) than children in the fantastical pretend-play condition who engaged
in more pretense-based play (M = 4.03, SD = 1.36) (see Fig. 8).
There were no other significant differences between these different styles of play in the fantastical
pretend-play condition on any FO, pretense, or EF variables (see Figs. 8 and 9).
In addition, EF scores of children in the fantastical pretend-play condition who consistently
engaged in pretense-based imaginative play (as opposed to fantastical play) were compared with chil-
dren in the non-imaginative play condition after controlling for pre-test scores. Interestingly, no sta-
tistically significant differences were observed between these subsets of children (all Fs 6 .556, all
ps > .46) (see Table 4). In other words, children in the fantastical pretend-play condition who engaged
134 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

1
0.8
0.6
0.4
Post-test z-scores

0.2
0
-0.2
-0.4 Fantastical

-0.6 Pretense-Based

-0.8 FO FO Toys FO FO FO Pretense

Cognitions & Games Pretense Entities Behaviors
-1

Fig. 9. Difference between children in the fantastical pretend-play condition who consistently engaged in fantastical play and
those who engaged in pretense-based play on FO measures at post-test after controlling for pre-test scores. Errors bars reflect
95% confidence interval. There were no differences in levels of FO pretense after the intervention among these subsets of
children.

Table 4
Pretense-based pretend-play versus non-imaginative play condition.

Measure Play style/Condition n M SD F Significance g2

Forward Digit Span Pretense based 17 4.01 1.46 0.022 .88 .000
(Working Memory) Non-imaginative 30 3.94 1.46
Card Sort errors after rule Pretense based 15 2.35 4.57 0.556 .46 .014
Switch (Attention Shift) Non-imaginative 26 3.45 4.56
Grass/Snow task Pretense based 9 12.42 4.28 0.409 .53 .014
(Inhibitory Control) Non-imaginative 22 11.33 4.21

in more pretense-based imaginative play throughout the intervention performed at similar levels to
children in the non-imaginative play condition on measures of EF after the intervention. Thus, the fan-
tasy component of the fantastical pretend-play condition seems to be relevant to the benefits in EFs
observed in this condition.

Discussion

The purpose of the current study was to experimentally determine whether or not the repeated act
of engaging in fantasy-oriented behaviors facilitates the development of EFs beyond what would be
observed due to simple maturation or engaging in non-fantasy play. The main hypothesis was sup-
ported, such that children in the fantastical pretend-play condition showed improvements in working
memory (i.e., Forward Digit Span task) and attention shift (i.e., Card Sort task) performance beyond that
of children in the non-imaginative play condition. In addition, the data indicate that our fantastical
pretend-play manipulation was successful in that children in the fantastical pretend-play condition
were engaging in higher levels of pretense (i.e., FO pretense behaviors) after the intervention, whereas
children in the non-imaginative play and control conditions did not show any changes in pretending
behavior (i.e., FO pretense behaviors). As hypothesized, we did not observe any changes in children’s
FO, which is consistent with previous literature suggesting that FO may be a stable trait. The second
aim of the current study compared changes in EF performance between children in the fantastical
pretend-play condition and children in the business-as-usual control condition. The differences
between these two conditions at post-test did not reach significance. However, as expected, data show
improvements in EFs in the fantastical pretend-play condition and no changes in the control condition.
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 135

The following results add to and extend existing literature that suggests a relationship between FO
and EF. Specifically, both Carlson and colleagues (2014) and Pierucci and colleagues (2014) found rela-
tions between FO and better cognitive skills using a correlational design. Our experimental data help
to elucidate the directionality of this relationship. Specifically, we found that the act of engaging in
fantasy-oriented pretend-play facilitates development of EFs. In addition, the data of the current study
also indicate that a high level of engagement in the fantastical pretend-play condition is related to the
observed benefits. Interestingly, our fantastical pretend-play condition was made up of individuals
with varying levels of EF skills prior to the intervention, with 22% and 28% scoring at chance or below
on the Day/Night task and Forward Digit Span, respectively. The fact that we observed improvements
in many domains of EF in the fantastical pretend-play condition after engaging in 5 weeks of fantasy-
oriented play not only indicates that fantasy-oriented play facilitates EF development but also may
challenge the idea that a certain level of EF skill is necessary in order to engage in fantasy-oriented
behaviors.
In addition to highlighting the importance of engaging in fantasy-oriented behaviors, the results of
the current study also emphasize that certain types of imaginative behaviors may be more beneficial
than others. Specifically, we found that children in the fantastical pretend-play condition who typi-
cally engaged in more fantastical play throughout the intervention (e.g., pretending to be fairies or
to play with dragons) scored better at post-test on measures of working memory compared with chil-
dren in the fantastical pretend-play condition who typically engaged in more pretense-based or socio-
dramatic play (e.g., pretending to be a mom or to go to a restaurant). This finding is interesting
because both fantastical pretend-play and pretense-based play involve children engaging in behavior
that is guided by their imaginative thoughts rather than external stimuli and following norms that dif-
fer from their everyday behaviors. Although all of these children were consistently engaging in imag-
ination, there were clear differences in their post-intervention EF scores when their imaginative
behaviors involved non-realistic or fantastical elements than when their imaginative behaviors
involved more realistic themes that they might experience in everyday life.
Perhaps this is the case because fantastical cognitions vary widely from reality as opposed to
pretense-based imaginations, which may vary only slightly from children’s everyday behaviors.
Engaging in fantastical cognitions that differ greatly from reality likely involves greater EF demands
such as working memory to remember the rules of the realm in which children are playing. Interest-
ingly, these results are consistent with previous literature showing that pretense (i.e., pretending to be
an animal or another person) is not correlated with measures of EF, as are believing in fantastical enti-
ties and engaging in fantastical cognitions (Pierucci et al., 2014). Therefore, it appears that the fantasy
element to the fantastical pretend-play intervention is what was driving the observed improvements
in EF abilities. Anecdotally, parents and teachers often worry about the development of children who
are constantly daydreaming. To be clear, we posit that the act of switching between reality and imag-
ination, specifically highly fantastical imagination, is what is driving the observed benefits in EFs,
much like the cognitive benefits observed when switching between languages in a bilingual household
(Bialystok, 1999, 2011; Bialystok & Martin, 2004).
Although the current study adds to and extends existing literature, it is important to address sev-
eral potential limitations. Of first note is the nature of the sample used in the current study. Specifi-
cally, children who participated came from predominantly middle-class households and varied little
in ethnicity. Therefore, our results are not generalizable outside of this narrow sample. Future studies
should incorporate children living in at-risk populations who may exhibit deficits in cognitive, atten-
tional, and behavioral control. Because our data indicate that engaging in fantasy-oriented play creates
an enriched environment through which EFs can be scaffolded, we would hypothesize that children in
at-risk environments would show even greater gains in EF after participating in a fantasy-oriented
play intervention. It should also be noted that some of the EF measures used in the current study
(e.g., Card Sort task) were not sufficiently difficult for children in our sample given that the majority
of participants made few errors on these measures, resulting in ceiling effects at pre- and post-
intervention. These ceiling effects may have prevented us from observing any changes in these mea-
sures from pre-intervention to post-intervention at times. Thus, employing more robust measures in
future studies may reveal even more benefits of fantasy-oriented play on EF development. For
example, the rigor of the Card Sort task could be improved by adding additional sorting dimensions
136 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

(e.g., color, shape, number) and by switching more often. Finally, the two control conditions used in
the current study (i.e., non-imaginative play and control) were not designed to encourage pretense
separately from fantastical play. Future experimental studies should include a fourth condition that
specifically encourages imaginative play that is sociodramatic and pretense based, but does not have
a fantasy component, in order to experimentally confirm the effect of the fantasy component of the
fantastical pretend-play condition on development that our observational engagement and play style
data suggest.
Despite these limitations, the current study is the first to use a randomized controlled trial design
to innovatively test the effects of play on cognitive development, and the results have several impor-
tant theoretical implications. Previous literature has laid out three possibilities for the influence of
pretend play on cognitive development (Lillard et al., 2013). First, Vygotsky (1978) would have argued
that pretend play is a crucial element to cognitive development. In other words, pretend play is a nec-
essary component through which cognitive abilities develop. The second possibility is that pretend
play is equifinal (cf. Smith, 2010). According to this view, pretend play helps to promote cognitive
development, but it is not the only route through which cognitive abilities can develop. Finally,
Piaget (1962) suggested that pretend play may be epiphenomenal, such that it is simply a byproduct
of some other factor that is actually linked to cognitive development. According to epiphenomenalism,
pretend play does not make any direct contributions to cognitive development. Rather, pretend play
simply coincides with some other factor that is directly related to cognitive development. In such
cases, pretend play may be mistakenly thought to cause the development of cognitive abilities. Up
to this point, methodological limitations have prevented researchers from conclusively determining
which theoretical perspective best describes the relationship between pretend play and cognitive
development (Lillard et al., 2013). By using a randomized controlled experimental design in which
all experimenters and teachers were blind to the hypotheses and conditions of each child, our data
suggest that the relationship between fantasy-oriented play and EF development is most likely equi-
final, such that engaging in fantasy-oriented play is one of many ways to directly enhance EF
development.
In conclusion, the current research has identified fantastical pretend play, a ubiquitous experience
during childhood, as a facilitator of EFs and cognitive control. The results of this study indicate that
encouraging fantasy play in children’s everyday lives will facilitate development in the cognitive
domain. These findings have the potential to improve school-readiness in preschool children and to
help close achievement gaps throughout the nation. Because fantasy play is easily implemented into
existing classroom curricula at little or no cost, the findings from the current study may have even
broader impacts on curriculum development, especially in at-risk populations such as Head Start.
Children who are well prepared for school are much more likely to be occupationally successful.
Indeed, a conservative estimate for the return on investment for preparing children for school is 7:1
based on several longitudinal school-readiness interventions such as the Chicago and High/Scope
Perry Preschool Projects (Bruner, 2004). Thus, legislators and education specialists should be very
interested in learning about the effects of a low-cost intervention involving fantastical play to prepare
children for school success.

Acknowledgment

The authors are grateful for funding for this research from the University of Alabama Research
Grants Committee and to the schools, families, teachers, and research assistants who made this
research possible.

References

Baddeley, A. D. (1983). Working memory. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 302,
311–324.
Baddeley, A. (1992). Working memory. Science, 255, 556–559.
Bergman Nutley, S., Söderqvist, S., Bryde, S., Thorell, L. B., Humphreys, K., & Klingberg, T. (2011). Gains in fluid intelligence after
training non-verbal reasoning in 4-year-old children: A controlled, randomized study. Developmental Science, 14, 591–601.
Bialystok, E. (1999). Cognitive complexity and attentional control in the bilingual mind. Child Development, 70, 636–644.
 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 137

Bialystok, E. (2011). Coordination of executive functions in monolingual and bilingual children. Journal of Experimental Child
Psychology, 110, 461–468.
Bialystok, E., & Martin, M. M. (2004). Attention and inhibition in bilingual children: Evidence from the Dimensional Change Card
Sort task. Developmental Science, 7, 325–339.
Bruner, C. (2004). Many happy returns: Three economic models that make the case for school readiness. Des Moines, IA: State Early
Childhood Policy Technical Assistance Network. Retrieved from http://www.finebynine.org/uploaded/file/SECPTAN_
MHR_final.pdf.
Carlson, S. M. (2005). Developmentally sensitive measures of executive function in preschool children. Developmental
Neuropsychology, 28, 595–616.
Carlson, S. M., & Moses, L. J. (2001). Individual differences in inhibitory control and children’s theory of mind. Child Development,
72, 1032–1053.
Carlson, S. M., White, R. E., & Davis-Unger, A. C. (2014). Evidence for a relation between executive function and pretense
representation in preschool children. Cognitive Development, 29. http://dx.doi.org/10.1016/j.cogdev.2013.09.001.
Center on the Developing Child at Harvard University. (2011). Building the brain’s ‘‘air traffic control” system: How early
experiences shape the development of executive function (Working Paper No. 11). Cambridge, MA: Author. Retrieved from
http://www.developingchild.harvard.edu.
Davis, H. L., & Pratt, C. (1996). The development of children’s theory of mind: The working memory explanation. Australian
Journal of Psychology, 47, 25–31.
Diamond, A., & Taylor, C. (1996). Development of an aspect of executive control: Development of the abilities to remember what
I said and to ‘‘do as I say not as I do”. Developmental Psychology, 29, 315–334.
Dunn, J. R. (2010). Health behavior vs. the stress of low socioeconomic status and health outcomes. Journal of the American
Medical Association, 303, 1199–1200.
Dunn, L. M., & Dunn, D. M. (2007). Peabody Picture Vocabulary Test (fourth ed.). Minneapolis, MN: NCS Pearson.
Eakin, L., Minde, K., Hechtman, L., Ochs, E., Krane, E., Bouffard, R., ... Looper, K. (2004). The marital and family functioning of
adults with ADHD and their spouses. Journal of Attention Disorders, 8, 1–10.
Estes, D., Wellman, H. M., & Woolley, J. (1989). Children’s understanding of mental phenomena. In H. Reese (Ed.), Advances in
child development and behavior (pp. 41–86). New York: Academic Press.
Gerstadt, C. L., Hong, Y. J., & Diamond, A. (1994). The relationship between cognition and action: Performance of children 3½–
7 years old on a Stroop-like Day-Night test. Cognition, 53, 129–153.
Gilpin, A. T. (2009). Mechanisms for overcoming reality status biases. Unpublished doctoral dissertation, University of Texas,
Austin.
Golumb, C., & Kuersten, R. (1996). On the transition from pretense play to reality: What are the rules of the game? British Journal
of Developmental Psychology, 14, 203–217.
Lillard, A. S., Lerner, M. D., Hopkins, E. J., Dore, R. A., Smith, E. D., & Palmquist, C. M. (2013). The impact of pretend play on
children’s development: A review of the evidence. Psychological Bulletin, 139, 1–34.
McCrae, R. R. (1987). Creativity, divergent thinking, and openness to experience. Journal of Personality and Social Psychology, 52,
1258–1265.
McCrae, R. R. (1993). Openness to experience as a basic dimension of personality. Imagination, Cognition, and Personality, 13,
39–55.
Miyake, A., Friedman, N., Emerson, M., Witzki, A., Howerter, A., & Wager, T. D. (2000). The unity and diversity of executive
functions and their contributions to complex ‘‘frontal lobe” tasks: A latent variable analysis. Cognitive Psychology, 41,
49–100.
Moffitt, T. E., Arseneault, L., Belsky, D., Dickson, N., Hancox, R. J., Harrington, H., et al (2011). A gradient of childhood self-control
predicts health, wealth, and public safety. Proceedings of the National Academy of Sciences of the United States of America, 108,
2693–2698.
Monsell, S. (1996). Control of mental processes. In V. Bruce (Ed.), Unsolved mysteries of the mind: Tutorial essays in cognition
(pp. 93–148). Hove, UK: Lawrence Erlbaum.
Piaget, J. (1962). Play, dreams, and imitation in childhood (G. Gattegno & F. M. Hodgson, Trans.). New York: Norton.
Pierucci, J. M., O’Brien, C. T., McInnis, M. A., Gilpin, A. T., & Barber, A. B. (2014). Fantasy-oriented constructs and related executive
function development in preschool: Developmental benefits to executive functions by being a fantasy-oriented child.
International Journal of Behavioral Development, 38, 62–69.
Singer, D. G., & Singer, J. L. (1981). Television, imagination, and aggression: A study of preschoolers. Hillsdale, NJ: Lawrence
Erlbaum.
Smith, P. K. (2010). Children and play. West Sussex, UK: Wiley-Blackwell.
St Clair-Thompson, H. L., & Gathercole, S. E. (2006). Executive functions and achievements in school: Shifting, updating,
inhibition, and working memory. Quarterly Journal of Experimental Psychology, 59, 745–759.
Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643–662.
Tahiroglu, D., Mannering, A. M., & Taylor, M. (2011). Visual and auditory imagery associated with children’s imaginary
companions. Imagination, Cognition, and Personality, 31(1), 99–112.
Taylor, M. (1999). Imaginary companions and the children who create them. New York: Oxford University Press.
Taylor, M., & Carlson, S. M. (1997). The relation between individual differences in fantasy and theory of mind. Child Development,
68, 436–455.
Taylor, M., Sachet, A. B., Maring, B. L., & Mannering, A. M. (2013). The assessment of elaborated role-play in young children:
Invisible friends, personified objects, and pretend identities. Social Development, 22, 75–93.
Thorell, L. B., Lindqvist, S., Bergman Nutley, S., Bohlin, G., & Klingberg, T. (2009). Training and transfer effects of executive
functions in preschool children. Developmental Science, 12, 106–113.
Vygotsky, L. S. (1967). Play and its role in the mental development of the child. Soviet Psychology, 5(3), 6–18.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University
Press.
138 R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138

Woolley, J. D. (1997). Thinking about fantasy: Are children fundamentally different thinkers and believers from adults? Child
Development, 68, 991–1011.
Woolley, J. D., & Tullos, A. (2008). Imagination and fantasy. In M. Haith & J. Benson (Eds.), Encyclopedia of infant and early
childhood development (pp. 117–127). Oxford, UK: Elsevier.
Wright, I., Waterman, M., Prescott, H., & Murdoch-Eaton, D. (2003). A new Stroop-like measure of inhibitory function
development: Typical developmental trends. Journal of Child Psychology and Psychiatry, 44, 561–575.
Zelazo, P. D. (2006). The Dimensional Change Card Sort (DCCS): A method of assessing executive function in children. Nature
Protocols, 1, 297–301.
Zelazo, P. D., Muller, U., Frye, D., & Marcovitch, S. (2003). The development of executive function in early childhood. Monographs
of the Society for Research in Child Development, 68(3, Serial No. 274).