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 b

Department of Psychology, University of Alabama, Tuscaloosa, AL 35487, USA Department of Family Sciences, Texas Woman’s University, Denton, TX 76204, USA

a r t i c l e

i n f o

Article history: Received 29 June 2015 Revised 6 November 2015

Keywords: Executive functions Fantasy orientation Pretend-play Intervention Cognitive development Pretense

a b s t r a c t Although recent correlational studies have found a relationship between fantasy orientation (FO; i.e., a child’s propensity to play in a fantastical realm) and higher order cognitive skills called executive functions (EFs), no work has addressed the causality and directionality of this relationship. The current study experimentally examined the directionality of the observed relationship between FO and EF development in preschool-aged children through an innovative play intervention employing a randomized controlled design. A sample of 110 children between the ages of 3 and 5 years were randomly assigned to one of three conditions: fantastical pretend-play intervention, non-imaginative play intervention, or business-as-usual control. Results revealed that children 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 pretendplay 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 relationship 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: [email protected] (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 control. 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 cognitive abilities referred to as executive functions (EFs). Executive functions are defined as higher order thinking processes that allow individuals to override 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 information, which allows individuals to manipulate information as they cognitively process it (Baddeley, 1983, 1992). Inhibitory control refers to individuals’ ability to suppress an automatic prepotent response (Stroop, 1935; Wright, Waterman, Prescott, & Murdoch-Eaton, 2003). Finally, cognitive 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 development of EFs during early childhood. As Vygotsky (1978) theorized, complex pretend-play may provide a natural environmental experience 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 children’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 explanation 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 companions 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. Specifically, 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 fantasyoriented cognitions, belief in fantasy-oriented entities) displayed better cognitive inhibition and attentional 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 perform 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 investigations 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, pretense 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 activities 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 fantasyoriented 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 intervention, 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 performance 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 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. It should be noted, however, that FO was not hypothesized to change over the course of the intervention period. In fact, it is highly unlikely 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.

Gender Boys Girls Age (in months) Vocabulary Ethnicity Caucasian African American Other Parent education Income

Fantastical pretend-play

Non-imaginative play

Control

Total

49% (n = 19) 51% (n = 20) 50.62

50% (n = 16) 50% (n = 16) 54.06

41% (n = 16) 59% (n = 23) 52.37

46% (n = 51) 54% (n = 59) 52.22

106.74

110.93

109.36

108.84

95% (n = 37) 2.5% (n = 1)

88% (n = 28) 6% (n = 2)

85% (n = 33) 10% (n = 4)

89% (n = 98) 6% (n = 7)

2.5% (n = 1) M = 4-year college or bachelor’s degree M = $90,000–$99,999

6% (n = 2) M = 4-year college or bachelor’s degree M = $80,000–$89,999

5% (n = 2) M = 4-year college or bachelor’s degree M = $90,000–$99,999

5% (n = 5) M = 4-year college or bachelor’s degree 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 questionnaires, 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 fantastical 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 realitybased response, 1 indicating a response that was representational or anthropomorphic but not fantastical, 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 addition, 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, superheroes) 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 independently 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) Impersonation 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 experimenter 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 experimenter 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 experimenter spoke. The PPVT-4 consists of 19 sets with 12 items in each set for a total of 228 items. Participants 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 standardized 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 measures 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 participants 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, encouragement, questions that inspired imagination such as ‘‘What do you see on the moon?”). In the nonimaginative 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 fantastical pretend-play condition, research assistants offered encouragement throughout the play sessions 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 childinitiated 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 condition). Group size differed daily based on school attendance. After 5 weeks of intervention, all participants 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 control 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, gender, 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 measure 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 series 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 considered 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.

127

R.B. Thibodeau et al. / Journal of Experimental Child Psychology 145 (2016) 120–138 Table 2 Descriptive data. Measure

Condition

Possible range

npre

Fantastical Non-imagination Control


1 to +1

39 32 37

.132 .116
.251

.682 .706 .770

39 30 35

.123
.201 .037

.820 .702 .687

FO Toys & Games

Fantastical Non-imagination Control


1 to +1

39 32 37


.013 .027
.010

.553 .595 .554

39 30 35

.089
.076
.042

.518 .467 .485

FO Pretense

Fantastical Non-imagination Control


1 to +1

39 32 37

.041 .126
.151

.737 .813 .733

39 29 33

.096
.151 .036

.733 .755 .693

FO Entities

Fantastical Non-imagination Control


1 to +1

39 32 39


.074 .155
.101

.704 .690 .726

39 32 39

.117 .020
.105

.716 .636 .525

FO Pretense Behaviors

Fantastical Non-imagination Control


1 to +1

39 32 37


.034 .120
.072

.704 .667 .756

39 31 35

.255
.041
.233

.699 .751 .744

Fantastical Non-imagination Control

0 to 7

39 31 37

3.95 4.06 4.30

.887 1.03 1.05

39 31 35

4.69 3.94 4.38

1.76 1.26 1.52

Attention Shift (CS errors)

Fantastical Non-imagination Control

0 to 25

38 30 36

2.89 1.57 3.14

4.69 3.11 3.70

38 28 35

1.58 3.14 2.83

3.21 4.78 4.00

Inhibitory Control (Day/Night & Grass/Snow)

Fantastical Non-imagination Control

0 to 16

27 26 26

4.31 2.49 5.04

34 26 31

FO/pretense measures FO Cognitions

EF measures Working Memory (FDS)

Mpre

12.3 13.5 10.8

SDpre

npost

Mpost

11.5 11.7 12.0

SDpost

4.72 4.40 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 preand 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 pretendplay 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 intervention 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, children 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.

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8 * **

Pre-test variable

1

2

3

4

5

6

7

FO Cognitions FO Toys & Games FO Pretense FO Entities FO Pretense Behaviors Working Memory Attention Shift Inhibitory Control

.155 .279** .241* .120
.013 .084 .148

.161 .065 .031 .276** .152
.001

.085 .259**
.046 .112 .113


.043
.032 .011 .023

.223*
.110
.106


.297** .135


.234*

Post-test variable

1

2

3

4

5

6

7

FO Cognitions FO Toys & Games FO Pretense FO Entities FO Pretense Behaviors Working Memory Attention Shift Inhibitory Control

.173 .189 .166 .070 .197*
.158 .049

.189 .242*
.052 .052
.031

.060 .077
.076 .019

.077
.163 .034


.325** .420**


.284*

.060 .017 .118 .271** .008 .198

p < .05. p < .01.

1 Fantastical Pretend-Play Non-Imaginative Play Control

0.8 0.6

** **

Post-test z-scores

0.4 0.2

0 -0.2 -0.4 -0.6 -0.8

FO

-1 Cognitions

FO Toys & Games

FO Pretense

FO Entities

FO Pretense 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 fantastical 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

1

FO pretense behaviors z-scores

0.8 0.6

129

Fantastical Pretend-Play Non-Imaginative Play Control

0.4 *

0.2 0 -0.2

Pre-Test

Post-Test

-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 working 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 fantastical 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 fantastical 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 difference 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 difference 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 repeatedmeasures 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 Non-Imaginative Play Control

16

Post-test scores

14 12 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.

Highest number of digits recalled

7 6

Fantastical Pretend-Play Non-Imaginative Play Control

5

**

4 3 2 1 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

25

Errors after rule switch

20

131

Fantastical Pretend-Play Non-Imaginative Play Control

15

10

5

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 interested in determining what aspects of the fantastical pretend-play intervention were driving these benefits. 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. Interestingly, at post-test there was a significant difference between children in the fantastical pretendplay 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 imaginary 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 Post-test z-scores

0.4 0.2

Highly Engaged

0 -0.2

Somewhat Engaged

-0.4 -0.6 -0.8 -1

FO Cognitions

FO Toys & Games

FO Pretense

FO Entities

FO Pretense Behaviors

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 marginally 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 average 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/pretense 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 regardless 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 consistently 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

Post-test scores

12 10 8

†

Highly Engaged

6 Somewhat Engaged

4 2

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

Post-test Scores

12 10 8

* Fantastical

6

Pretense-Based

4 2 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 children in the non-imaginative play condition after controlling for pre-test scores. Interestingly, no statistically 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 Post-test z-scores

0.4 0.2 0 -0.2 -0.4

Fantastical

-0.6

Pretense-Based

-0.8 -1

FO Cognitions

FO Toys & Games

FO Entities

FO Pretense

FO Pretense Behaviors

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 (Working Memory) Card Sort errors after rule Switch (Attention Shift) Grass/Snow task (Inhibitory Control)

Pretense based Non-imaginative Pretense based Non-imaginative Pretense based Non-imaginative

17 30 15 26 9 22

4.01 3.94 2.35 3.45 12.42 11.33

1.46 1.46 4.57 4.56 4.28 4.21

0.022

.88

.000

0.556

.46

.014

0.409

.53

.014

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 fantasy 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 supported, 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 relations 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 fantasyoriented 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 typically 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 children in the fantastical pretend-play condition who typically engaged in more pretense-based or sociodramatic 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 differ from their everyday behaviors. Although all of these children were consistently engaging in imagination, 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. Interestingly, 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 entities 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 imagination, 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 several potential limitations. Of first note is the nature of the sample used in the current study. Specifically, 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, attentional, 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 postintervention. These ceiling effects may have prevented us from observing any changes in these measures 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 important 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 necessary 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 equifinal, 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).