Elaborated Role Play and Creativity in Preschool Age Children

Abstract

Elaborated role play (i.e., pretending in which children imagine and act out the part of another individual on a regular basis) is often considered an early indicator of creativity, but there is not strong research evidence of a relation between this type of pretend play and performance on creativity tasks during the preschool years. One possible reason is that the measures of creativity that are commonly used are not appropriate for young children. To address this, we developed 2 new measures of creativity based on a storytelling task, in which children were asked to complete a story, and a drawing task, in which children were asked to draw an imaginary person. Of the 75 4-and 5-year-old children who participated, those who engaged in elaborated role play had higher creativity scores on both measures (controlling for age and language ability). In contrast, children's performance on a measure of pretend play development that did not involve imaginary others (i.e., the action pantomime task) was not related to either measure of creativity. These results suggest that the storytelling and drawing measures were effective in assessing children's creativity, and that they were specifically associated with elaborated role play, rather than the developmental level of children's ability to pretend.

Full text

Elaborated Role Play and Creativity in Preschool Age Children
Candice M. Mottweiler and Marjorie Taylor
University of Oregon
Elaborated role play (i.e., pretending in which children imagine and act out the part of another individual
on a regular basis) is often considered an early indicator of creativity, but there is not strong research
evidence of a relation between this type of pretend play and performance on creativity tasks during the
preschool years. One possible reason is that the measures of creativity that are commonly used are not
appropriate for young children. To address this, we developed 2 new measures of creativity based on a
storytelling task, in which children were asked to complete a story, and a drawing task, in which children
were asked to draw an imaginary person. Of the 75 4- and 5-year-old children who participated, those
who engaged in elaborated role play had higher creativity scores on both measures (controlling for age
and language ability). In contrast, children’s performance on a measure of pretend play development that
did not involve imaginary others (i.e., the action pantomime task) was not related to either measure of
creativity. These results suggest that the storytelling and drawing measures were effective in assessing
children’s creativity, and that they were specifically associated with elaborated role play, rather than the
developmental level of children’s ability to pretend.
Keywords: creativity, role play, imaginary companions, preschool age
Many young children exhibit a striking ability and inclination to
invent imaginary characters and act out fantasy scenarios. For
some children, particular characters become regular features in
their everyday pretend activities, either as imaginary companions
(including invisible friends and personified objects) or as pretend
identities that children act out (Harris, 2000). The descriptions of
these characters are often vivid and detailed (Gleason, 2004;
Gleason, Sebanc, & Hartup, 2000;Taylor, 1999), and can be stable
across a period of months (Taylor, Cartwright, & Carlson, 1993)or
even years (Taylor, Carlson, Maring, Gerow, & Charley, 2004), as
demonstrated in both children’s verbal accounts and their draw-
ings. Taylor, Sachet, Mannering, and Maring (2013) refer to pre-
tend play involving such imaginary characters as elaborated role
play to distinguish this type of pretending from more fleeting acts
of role play that are common during the preschool years (e.g.,
occasionally pretending to be Batman or pretending to feed a doll).
Given the high level of generativity involved in elaborated role
play, one might expect that children who engage in this type of
pretending would score higher than other children on creativity
tasks, either because the practice of inventing imaginary characters
might have a training effect or, alternately, because creative chil-
dren might be the ones who find this type of pretend play partic-
ularly enjoyable. There is some evidence that is consistent with
this hypothesis. Elaborated role play is correlated with creativity in
10-year-olds (Hoff, 2005), early role play is related to later cre-
ativity in adolescence (Mullineaux & Dilalla, 2009;Schaefer,
1969), and retrospective reports of childhood role play are com-
mon in adults who pursue careers as actors (Goldstein & Winner,
2009) or fiction writers (Taylor, Hodges, & Kohanyi, 2002).
There is also research investigating the relation between pretend
play more generally and creativity (e.g., Dansky, 1980;Pepler &
Ross, 1981). For example, Russ and Schafer (2006) found that the
frequency of emotional expression in a free-play session with
puppets was related to higher creativity scores in first- and second-
grade children (also see Russ, Robins, & Christiano, 1999). How-
ever, in a recent review of the literature on the impact of pretend
play on development, Lillard et al. (2013) criticized many of the
creativity studies on methodological grounds and concluded that
“the evidence that pretend play enhances creativity is not convinc-
ing” (p. 8). In addition, research with young children that has
focused more narrowly on elaborated role play (specifically, imag-
inary companions) has not found a relation with creativity (Manos-
evitz, Fling, & Prentice, 1977;Pearson et al., 2001).
In our view, the inconsistent and null results of research with
preschool children could be at least partly due to the challenges of
measuring creativity in this age group (Busse, Blum, & Gutride,
1972;Manosevitz et al., 1977;W. C. Ward, 1968). In past re-
search, creativity has most commonly been assessed with diver-
gent thinking tasks in which children are asked to answer a
question and then answer it again and yet again, each time gener-
ating a new response. For example, in the Alternative Uses Task,
participants are asked to list as many uses as possible for an
everyday object, such as a brick or newspaper. Fluency (i.e., the
number of uses that are generated) and uniqueness (i.e., the num-
This article was published Online First March 31, 2014.
Candice M. Mottweiler and Marjorie Taylor, Department of Psychology,
University of Oregon.
This research was supported by a National Science Foundation Graduate
Research Fellowship awarded to Candice Mottweiler.
Correspondence concerning this article should be addressed to Candice
M. Mottweiler, Department of Psychology, 1227 University of Oregon,
Eugene, OR 97401. E-mail: mottweiler@gmail.com
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Psychology of Aesthetics, Creativity, and the Arts © 2014 American Psychological Association
2014, Vol. 8, No. 3, 277–286 1931-3896/14/$12.00 DOI: 10.1037/a0036083
277

ber of uses that are not mentioned by other participants in the
study) are then used to index creativity (Wallach & Kogan, 1965).
But generating multiple solutions for a problem is an unfamiliar
task that is cognitively challenging for preschool children (Beck,
Robinson, Carroll, & Apperly, 2006;Guajardo & Turley-Ames,
2004). In addition, divergent thinking tasks have been criticized as
measurements of creativity because inappropriate and bizarre re-
sponses tend to be identified as unique and thus inflate creativity
scores (Silvia et al., 2008).
The purpose of this study was to develop creativity measures
appropriate for preschool age children, as well as to investigate the
relation between individual differences on these measures and
children’s elaborated role play. Our tasks were adapted from
Amabile’s (1982) consensual assessment technique, in which par-
ticipants are asked to generate a product that is scored for overall
creativity by appropriate judges who are instructed to use their
own subjective criteria for creativity. According to Amabile, the
use of subjective ratings is a valid method for measuring creativity
because it is similar to how creativity is assessed in the real world
(compared with supplying coders with a specific definition of
creativity). This approach avoids asking children to generate mul-
tiple solutions for the same task and has the flexibility to allow for
comparisons of creativity across different types of activities.
We asked children to create stories and drawings—two different
types of products that involve familiar activities for preschool
children. Both storytelling and drawing have been used for assess-
ing children’s creativity in past research, but with mixed success.
For research using storytelling, there is a trade-off between pro-
cedures in which children are simply asked to tell a story and more
constrained storytelling tasks. Spontaneously generated stories
with minimal or no prompts can provide rich qualitative informa-
tion about the creative content of children’s narratives (Ahn &
Filipenko, 2007), but the wide variability in the content and length
is problematic for a systematic investigation of individual differ-
ences in creativity. The use of consistent stimuli and prompts can
provide more structure and thus make it easier to compare and rate
children’s narrative responses. For example, Hennessey and
Amabile (1988) showed 5- to 10-year-old children a picture book
with a readily understood plot and asked them to say something
about each page in the book. However, the picture book provides
all of the story’s content, leaving little scope for creativity. In that
study, creativity scores were strongly correlated with the length of
the children’s stories, suggesting that the procedure might not have
clearly distinguished creative innovation from talkativeness. Alex-
ander et al. (1994), in an effort to address these challenges, told 4-
to 7-year-old children two stories involving a protagonist in trou-
ble and asked them to finish the story by thinking of multiple ways
to help him. This task provided some structure, while allowing for
a wide range of responses. However, as in other divergent thinking
tasks, children were required to generate multiple variations for a
single story. The preschoolers had difficulty doing the task, and it
was also a challenge for most of the children to develop solutions
that were original or effective.
Similar to Alexander et al. (1994), we provided the beginning of
a story and then asked children to finish it. However, our task did
not involve generating alternatives—children provided a single
completion for a story stem that was then coded for creativity. Our
method was based on the MacArthur Story Stem Battery, in which
an experimenter tells the beginning of a story with the use of dolls
and props, and then asks children to finish the story (“Show me
and tell me what happens now”; Emde, Wolf, & Oppenheim,
2003). We presented children with a story stem that was designed
to elicit a range of narratives that could be assessed for creativity
using Amabile’s (1982) consensual assessment technique.
In addition, we included a measure of drawing creativity that
was less verbally demanding than the storytelling task. Our draw-
ing task was adapted from Karmiloff-Smith’s (1990) procedure for
assessing the development of drawing, in which children are asked
to draw real and pretend versions of the same object (e.g., a real
person and a pretend person). Matuga (2004) used the Karmiloff-
Smith task to assess children’s creativity by having judges rate the
“real” drawings for drawing ability and the “pretend” drawings for
creativity. In our view, information about creativity is lost with this
procedure because no attention is paid to the approaches children
take in solving the problem of depicting real and pretend people.
For example, some children draw a “pretend” person that is vir-
tually identical to the “real” person, whereas other children deviate
from the “real” person dramatically in their depiction of the
“pretend” person. The level of deviation from the baseline “real”
object is useful information when considering creativity. In our
study, children were asked to draw a picture of a person and a
picture of a pretend person. The children’s solutions to the task of
depicting real and pretend people were assessed for creativity
using the consensual assessment technique.
Children’s engagement in elaborated role play (i.e., creating
invisible friends, personified objects, and/or pretend identities)
was assessed using a procedure that involves interviewing both
children and their parents (Taylor et al., 2004). In addition, we
were interested in the range of creativity in children’s descriptions
of these characters. Past work has demonstrated that role play
characters vary dramatically in their originality; some are based on
real people (e.g., a pretend version of the child’s best friend) or
media characters (e.g., the Little Mermaid), while others are idio-
syncratic and unique to the child (e.g., a tiny tie-dyed veterinarian
named Elfie Welfie; Hoff, 2005;Taylor, 1999). In this study, the
elaborated role play characters were coded for creativity in order to
assess the extent that individual differences in children’s descrip-
tions were related to creativity on the other tasks.
We also wanted to examine the possibility that creativity is
related to the general ability to engage in pretend play, rather than
more specifically to elaborated role play. For this purpose, we
included the action pantomime task, a measure of children’s de-
velopmental ability to represent pretend objects (Overton & Jack-
son, 1973). Unlike elaborated role play, this type of pretending
task does not involve the creation of a character or other
social/emotional content. Instead, the focus is on the child’s
ability to represent a physical object that is used as a tool in a
pretend action.
We predicted that engaging in elaborated role play would be
related to creativity, whereas performance on the action panto-
mime task would be related to verbal ability and age. In addition
to investigating children’s pretend play and creativity, we included
a structured narrative task to assess and control for the possibility
that superior language ability and/or the ability to structure narra-
tives was related to creativity ratings (Reilly, Bates, & Marchman,
1998).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
278 MOTTWEILER AND TAYLOR

Method
Participants
Eighty-one children and their parents were recruited from a
database primarily based on birth announcements published in the
local newspaper. Six children were excluded from analyses; five
because they were unable to complete the procedure and one
because of outlier scores possibly related to his diagnosis of
Asperger’s syndrome. The final group of 75 children (38 boys, 37
girls) ranged in age from 4 years 0 months to 5 years 11 months
(mean age ⫽5 years 0 months; SD ⫽7 months) and were all
native English speakers with no known developmental disorders.
The participants were primarily from European American, middle-
class backgrounds, reflective of the local demographics where the
study was conducted. Parents provided written consent for their
children’s participation and the children provided verbal assent.
The study was approved by the institutional review board at the
university at which it was conducted.
Procedure
Children and their parents came to the laboratory to participate
in an hour-long session. Before beginning the tasks, two experi-
menters developed rapport with the children by interacting and
playing with them. Then an experimenter escorted children to a
separate room, where they completed tasks to assess pretend play,
creativity, language ability, and the ability to structure a narrative.
These measures were part of a larger study examining the corre-
lates of elaborated role play. The tasks were presented in a fixed
order for all of the children. According to Carlson and Moses
(2001), fixed orders are standard practice for individual differences
research because “it is critical that the individuals be exposed to
identical stimulus contexts” (p. 1035), including the stimuli and
the order in which the stimuli are presented (see Carlson & Moses
for an explanation of the interpretive problems that arise if coun-
terbalanced orders are used for individual differences research).
Data sheets, transcriptions, and the parent questionnaire were
identified with participant numbers, and names were always ex-
cluded from transcriptions (e.g., if a child said his or her own
name, it was omitted in the transcription).
Pretend Play Assessment
Child role play interview. Using the procedure developed by
Taylor et al. (2004), the experimenter asked children about imag-
inary companions in the following way:
I’m going to ask you some questions about pretending. Some friends
are real, like the kids who live on your street, the ones you play with.
And some friends are pretend friends. Pretend friends are ones that are
make-believe, that you pretend are real. Do you have a pretend friend?
If children said “no,” they were asked if they had ever had a
pretend friend. If children reported having a pretend friend, either
current or past, they were asked a series of questions about it (e.g.,
name, age, gender, appearance, whether it was an invisible friend
or a personified object, the activities that the child engaged in with
the pretend friend, and what the child liked and disliked about the
pretend friend).
Next, children were asked about pretend identities:
Now I’m going to ask you about another type of pretending. Some-
times children like to pretend they are someone else. They like to talk
and act like another person or an animal. Do you pretend to be
someone else—like another person or an animal?
Children who reported having a pretend identity were asked
additional questions similar to those asked about imaginary com-
panions. The interviews with the children were video recorded and
then transcribed for coding.
Parent role play questionnaire. While the children were
being interviewed, in a separate room, parents completed a ques-
tionnaire about their children’s involvement in pretend play. The
questions were similar to the questions in the child role play
interview, focusing on imaginary companions and pretend identi-
ties.
Follow-up interviews. A second experimenter supervised the
child while the experimenter who had interviewed the child ini-
tially reviewed the parent questionnaire to identify any discrepan-
cies with what the child had reported and asked the parent
follow-up questions, as necessary. For example, if a child said that
he or she had a pretend friend, but the parent reported that the child
did not, the experimenter asked the parent about the imaginary
companion described by the child (e.g., if the child had a real
friend by that name). Then, the experimenter returned to where the
child was playing and asked the child additional follow-up ques-
tions, as necessary. For example, if a child said that he or she did
not have a pretend friend but the parent reported that the child did,
the experimenter asked the child about the imaginary companion
named by the parent.
Coding of elaborated role play. After the data were collected
from all of the families, the child participants were categorized as
having invisible friends, personified objects, pretend identities, or
engaging in no form of elaborated role play, based on the tran-
scriptions of the child role play interviews, the parent role play
questionnaires, and the follow-up interviews.
Children were categorized as having an invisible friend if (a)
they said that they had an invisible friend and provided a good
description of it, or (b) they said that they had an invisible friend
and their parents confirmed the invisible friend and provided a
good description. The criteria for coding children as having a
personified object were similar, with one additional condition to
differentiate personified objects from transitional objects (e.g., a
blanket or teddy bear that a child holds or carries for comfort; see
Winnicott, 2005). To be categorized as having a personified object,
the description had to go beyond the physical appearance of the
object to include psychological details (e.g., “She is nice and
listens to me”). The criteria for coding children as having a pretend
identity also included an additional condition: Descriptions of
pretend identities had to go beyond a description of a costume that
a child might wear to include psychological details of the pretend
identity.
The coding was done by the first author (who conducted 50% of
the interviews), the second author (who did not conduct any of the
interviews and was unaware of the children’s information prior to
coding), and a graduate student whose role in this study was
restricted to coding. Although data sheets and transcriptions were
deidentified, it is possible that the first author’s interactions with
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
279
ELABORATED ROLE PLAY AND CREATIVITY

the children introduced bias. Therefore, we analyzed the scores
from the latter two coders in addition to all three coders. Reliability
was acceptable for all coding (role play category, role play cre-
ativity, narrative creativity, and drawing creativity), and the pattern
of results was the same for two and three coders. Given that the
results are the same regardless, the scores from all three coders are
reported.
The overlap in agreement for the three coders was 91% for
invisible friends, 84% for personified objects, and 73% for pretend
identities. Disagreements were resolved by assigning the scores of
the majority (two out of three coders). Eight children met the
criteria for more than one type of elaborated role play. For five of
these children, the most predominant type was identified by dis-
cussion. For three of the children, it was not clear which type of
role play was predominant (e.g., child frequently pretended to be
“Jena” while playing with an invisible “Balto”). These children
were included as “role players” in analyses comparing role play
with nonrole play. However, they were excluded from analyses
comparing different types of elaborated role play because their
idiosyncratic experiences could not readily be categorized.
In addition to categorizing invisible friends, personified objects,
and pretend identities, the creativity of the role play characters was
assessed by the same three coders. The coders were instructed to
independently read through all of the character descriptions to see
the full range of responses (presented in different random orders),
and then read them a second time to rate the characters for
creativity on a scale from 1 (not creative)to5(highly creative)
(Amabile, 1982). Coders were instructed to use the full range of
the scale and to avoid using the amount of content included as an
indicator of creativity, but were otherwise allowed to use their own
definitions of creativity. This procedure yielded good reliability
among the three coders (Cronbach’s alpha ⫽.88). The mean
average of the three scores was then calculated and used as an
indicator of role play creativity.
Action pantomime task. The action pantomime task was
used as a behavioral measure of children’s developmental level of
pretend play (Overton & Jackson, 1973). For this task, children are
asked to engage in pretend actions with an imagined object (e.g.,
“Pretend to brush your teeth with a toothbrush”). During the
preschool years, children tend to use body parts (e.g., when pre-
tending to brush their teeth, they use a finger as the imaginary
toothbrush), but by 8 years of age, most children are able to
integrate imagined objects with the pretend actions (e.g., they
pretend to hold an invisible toothbrush). Previous research has
shown that holding a pretend object is easier for young children
than performing other actions with the pretend object (Dick, Over-
ton, & Kovacs, 2005). Therefore, in order to elicit a range of
scores, we included two trials that only required pretending to hold
the imaginary object, as well as seven trials that involved more
complex actions. Children were asked to perform the nine pretend
actions in the following order: (a) brush teeth with a toothbrush,
(b) hold a pencil, (c) put on a pair of sunglasses, (d) hammer this
[a wooden block] with a hammer, (e) hold a knife, (f) cut this [a
piece of paper] with scissors, (g) pour water from a pitcher, (h) fan
yourself with a fan, and (i) flip a pancake with a spatula. Chil-
dren’s responses were coded for the use of a body part to represent
the object required for the action (e.g., a finger to represent a
toothbrush) or an invisible object (e.g., the child pretends to hold
an invisible toothbrush). Children received a score for the number
of times they used an invisible object (0 to 9). One child declined
to participate in this task.
Narrative Creativity
For this task, children were presented with the beginning of a
story and then asked to complete it (adapted from the MacArthur
Story Stem Battery; Emde et al., 2003). To acquaint children with
the demands of the task, they were first asked to provide a
completion for a warm-up story stem (“The Birthday Party”) that
could be easily completed using a script familiar to most children.
Two small female dolls (“Susan” and “Jane”) were used to act out
the story stem for female participants, and two small male dolls
(“George” and “Bob”) were used for male participants. Then
children were asked to complete “The Magic Key” story stem that
was designed to elicit responses that could potentially vary in
creativity. The experimenter placed a felt path and a small key on
the table and walked the dolls down the path. “Susan/George and
Jane/Bob are going for a walk outside when they see a key.
Susan/George says, ‘What’s this on the ground?’ Jane/Bob says,
‘It’s a key. I wonder if it’s magic.’” Then the experimenter asked,
“Can you show me and tell me what happens now?”
Three coders (the two authors and a graduate student who was
not otherwise involved with this study) independently rated the
creativity of children’s completions for “The Magic Key.” The
coders were instructed to read through all of the transcribed story
completions to see the full range of responses (presented in dif-
ferent random orders) and to read the transcriptions a second time
to rate each for creativity on a scale from 1 (not creative)to5
(highly creative;Amabile, 1982). Coders were instructed to use the
full range of the scale (1 to 5) and to avoid using the children’s
verbal ability or talkativeness as an indicator of creativity, but were
otherwise allowed to use their own definitions of creativity. This
procedure yielded high reliability among the three coders (Cron-
bach’s alpha of .95). The mean average of the three scores was
calculated and used as an indicator of narrative creativity. Table 1
provides examples of children’s completions.
Table 1
Examples of Children’s Story Stem Completions
“Then they tried to open a door with the key and it opened and it had
fake scary stuff, like fake scary xxxxx and costumes or Frankenstein
costumes or mummies or pumpkin ones. Then they went out and
locked the door. Then they went and then they went into jungle and
with the key and they rode in the front until a board that came by and
it took the key out of Bob’s hand. And then it, then it put it down the
chimney. And inside the chimney was on fire. It dropped the key and
it fell in the fire.”
Average creativity rating: 4.5
“She picks the key up and then they find something to unlock, like
maybe, like that lock over there....Itdoesn’t work. And then they
pick up the key and then they found a lock and they put the key in it
and they unlock it and the door opens. And then they say, ‘Mom,
Mom, a door opens!’ And then they said, ‘Come on’ and then they
went inside. And then they went and they set the key back where it
was and she put it back there and they went home. Because they were
suppose to be right here.”
Overall creativity: 2.5
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
280 MOTTWEILER AND TAYLOR

Drawing Creativity
For this task, children were first asked to draw a picture of a
person and then to draw a picture of a “pretend person, a person
that couldn’t exist, a person that is made up” (adapted from
Karmiloff-Smith, 1990). After children completed the drawings,
they were asked to describe what they had drawn. Three children
were excluded due to experimenter error and seven children did
not want to participate in the task (possibly because this was one
of the last tasks in the session). Of the 64 children who participated
(87%), eight did not complete a drawing of a “pretend person.”
For the 56 children who completed two drawings, the three
coders independently rated them for creativity. The coders were
instructed to look through the children’s drawings and descriptions
to see the full range of responses (presented in different random
orders), and then go through them a second time rating the cre-
ativity of the children’s solutions to the task of drawing a pretend
person from 1 (not creative)to5(highly creative;Amabile, 1982).
Coders were instructed to use the full range of the scale (1 to 5)
and to avoid using drawing ability as an indicator of creativity, but
were otherwise allowed to use their own definitions of creativity.
This procedure yielded good reliability among the three coders
(Cronbach’s alpha of .88). The mean average of the three scores
was calculated and used as an indicator of drawing creativity.
Language Ability
Frog, Where Are You? (Mayer, 1969) is a picture book that
communicates a readily understood plot about a boy, a dog, and a
lost frog without the use of any text. This book is often used to
elicit narratives from young children for studying language devel-
opment (Slobin, 2004). Children were asked to look at each page
of the book and describe what was happening (“This book has a
story about a boy, a dog, and a frog. We’re going to look at all the
pictures in the book and you’re going to tell the story of what’s
happening”). If a child did not give a response for a page, the
experimenter would prompt him or her by asking, “What do you
think is happening on this page?” Two children did not complete
this task. Children’s narratives were later transcribed for coding.
To obtain an index of children’s vocabulary, the number of
nonrepeating nouns, verbs, and adjectives included in the chil-
dren’s transcribed narratives were totaled, as described by A.
Nicolopoulou (personal communication, April 8, 2009). To mea-
sure children’s syntax, children’s narratives were coded for the
number of different types of complex sentences (coordinate sen-
tences, adverbial clauses, verb complements, relative clauses, and
passive sentences) for a possible score of 0 to 5 (see Reilly et al.,
1998). Half of the narratives were double-coded, with 92% overlap
reliability.
Narrative Structure
In past research, Trionfi and Reese (2009) found that children
with imaginary companions (mostly invisible friends) did not
differ from other children in vocabulary or story comprehension,
but included more dialogue when asked to retell a story told to
them in the lab, and used more temporal-locative-causal terms
when asked to describe an event they had experienced in the past
year. Trionfi and Reese interpreted these results as indicating that
children with imaginary companions have more advanced narra-
tive skills than other children. In order to control for children’s
ability to organize and structure a narrative in this study, we used
a coding scheme that has previously been used for children’s
narrations of Frog, Where Are You? (Reilly et al., 1998). Children
received up to 2 points for the initiating event (i.e., the frog escapes
and the boy looks for him in his bedroom), up to 2 points for search
episodes (i.e., the boy looking for the frog in the woods), and up
to 2 points for the resolution (i.e., the boy finds his frog and goes
home), for a total possible score of 0 to 6. Half of the narratives
were double-coded, with 85% overlap reliability.
Data Analysis
Due to concerns about missing data (most notably for the
drawing task), independent samples ttests and chi-square tests
were conducted to compare children who completed all tasks with
children who had missing data. As there were no significant
differences as a function of missing data, all available data were
included in the analyses: 75 observations for story stem creativity,
whether or not the child engaged in elaborated role play, role play
creativity, and demographic information; 74 for pretend play de-
velopment; 73 for vocabulary, syntax, and narrative structure; 72
for type of elaborated role play; and 56 for drawing creativity.
Results
In this study, we measured two types of pretend play: elaborated
role play and general pretend play development (i.e., object sub-
stitution). The primary difference between the two is that elabo-
rated role play involves representing social roles and characters,
whereas object substitution involves representing physical trans-
formations of objects without social content. Although there are
likely many ways that these two types of pretend play are similar,
we hypothesized that elaborated role play would be related to
creativity, while object substitution would be related to age and
language abilities.
Elaborated Role Play
The predominant form of elaborated role play was identifiable
for 72 of the 75 children. Of the 72 children, 22 were identified as
primarily having invisible friends (30.56%), 10 were coded as
primarily having personified objects (13.89%), 12 were coded
as primarily having pretend identities (16.67%), and 28 were
coded as not engaging in any form of elaborated role play
(38.89%). There was a trend for role play category to be related to
sex, ␹
2
(3, N⫽72) ⫽6.60, p⫽.09, ␸⫽.30, due to the greater
number of girls than boys who had personified objects, and the
greater number of boys than girls who had pretend identities.
Consistent with previous research in this area, children de-
scribed a range of different types of characters (see Table 2 for
examples of invisible friends, personified objects, and pretend
identities). The characters also varied in creativity ratings (M⫽
2.66, SD ⫽1.15, range ⫽1to5).Table 3 provides means,
standard deviations, and the range of scores as a function of type
of role play (invisible friend, personified object, pretend identity,
or no role play).
One-way ANOVAs examining differences for age, vocabulary,
syntax, and narrative structure (dependent measures) as a function
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
281
ELABORATED ROLE PLAY AND CREATIVITY

of role play (between-group factor) yielded no significant effects.
In addition, the creativity of role play characters was not correlated
with age, vocabulary, syntax, or narrative structure (see Table 4).
In contrast to the findings for role play, performance on the
measure of pretend play development (the action pantomime task)
was significantly correlated with age, vocabulary, syntax, and
marginally with narrative structure (see Table 4). There was no
significant difference in action pantomime scores as a function of
gender. Controlling for age, vocabulary, syntax, and narrative
structure, a one-way ANOVA found that children who engaged in
elaborated role play scored significantly higher on the action
pantomime task than nonrole players, F(1, 66) ⫽5.80, p⫽.019,
␩
2
⫽.08. For the comparisons of each type of role play with
nonrole play (controlling for age, vocabulary, syntax, and narrative
structure), we used a Bonferroni correction to set the significance
level at .016. These analyses show a trend for children with
invisible friends to score higher than children who engaged in no
form of role play, F(1, 42) ⫽6.11, p⫽.018, ␩
2
⫽.12. The scores
for children with personified objects and pretend identities did not
differ significantly from the scores of children who engaged in no
form of role play (p⫽.07, p⫽.63, respectively).
Creativity
The creativity ratings for children’s story stem completions
were positively correlated with vocabulary and syntax, suggesting
that children with superior language skills produced stories that
were rated as more creative. Narrative creativity was not related to
age, sex, or narrative structure. The creativity ratings for the
drawing task were positively correlated with age, but were not
related to any of the other variables (sex, vocabulary, syntax, or
narrative structure; see Table 3 for means and SDs and Table 4 for
correlations).
These results suggest that aspects of general development or
ability might be associated with enhanced creativity or influence
the way adults code children’s responses for creativity. Therefore,
partial correlations were conducted to evaluate the relations among
the narrative, drawing, and role play creativity, controlling for age,
vocabulary, and syntax. (For this analysis, children who did not
engage in any form of role play were assigned role play creativity
scores of 0.) These analyses indicated that all three measures of
creativity were positively correlated: narrative creativity and draw-
ing creativity, r(49) ⫽.42, p⫽.002, narrative creativity and role
play creativity, r(68) ⫽.36, p⫽.002, and drawing creativity and
role play creativity, r(49) ⫽.32, p⫽.02. These findings demon-
strate associations between the measures of creativity that are not
better accounted for by developmental level or verbal ability.
Table 2
Examples of Children’s Elaborated Role Play
Invisible friends:
Ava (4-year-old girl): An older girl who has black hair. She likes to
play with the child’s dolls, colors pictures for the child, and
decorates the child’s room. Sometimes Ava doesn’t want to play
with the child and prefers to play with her brother.
Big Kittens and a Little Kitten (4-year-old boy): Pink and brown
kittens that can do “all the tricks” that the child likes, play with a
ball of string, and chase a pretend mouse that flies a pretend
airplane and “goes out the window.”
Personified objects:
Nuni (5-year-old girl): A large, stuffed, purple unicorn as big as the
child that can “almost fly” and likes to play checkers. There’s
nothing about Nuni that the child doesn’t like.
Froggy (4-year-old boy): A stuffed green and white frog that likes to
play the guitar, watch television, and likes to be fed fly soup. The
child doesn’t like it when “he’s a bad boy.”
Pretend identities:
Superman (4-year-old boy): A superhero who is 10 years old and
likes to rescue the town. He has special powers that allow him to
“save the world.” The child doesn’t like snoring when he pretends
to be Superman.
Fairy (5-year-old girl): The child pretends to be a fairy with wings
that can fly away from people and can fly to the moon. The child
likes to pretend that she can touch the clouds and stand on the
clouds.
Table 3
Summary of Means, Standard Deviations, and Ranges for Measures as a Function of Role Play
Total
(n⫽75) No role play
(n⫽28) Invisible friend
(n⫽22) Personified object
(n⫽10) Pretend identity
(n⫽12)
Gender (n⫽75) 38 boys 14 boys 11 boys 2 boys 9 boys
37 girls 14 girls 11 girls 8 girls 3 girls
Age, in months (n⫽75) 59.84 (7.04) 59.00 (6.97) 61.14 (6.88) 59.30 (9.31) 59.75 (6.31)
49 to 71 49 to 71 50 to 71 49 to 71 50 to 68
Vocabulary (n⫽73) 44.79 (12.50) 42.22 (9.88) 46.23 (16.84) 44.44 (10.63) 46.92 (17.33)
12 to 84 26 to 71 12 to 84 18 to 56 22 to 83
Syntax (n⫽73) 2.22 (.93) 2.07 (.92) 2.32 (.95) 2.00 (1.00) 2.25 (.87)
0to4 1to4 0to4 1to4 1to4
Narrative structure (n⫽73) 3.77 (1.44) 3.70 (1.26) 3.82 (1.33) 3.78 (1.48) 3.75 (1.87)
0to6 0to6 1to6 1to5 0to6
Pretend play development (n⫽74) 5.61 (1.83) 4.78 (1.89) 6.32 (1.32) 6.10 (1.10) 5.42 (2.15)
0to9 0to8 4to9 4to8 3to9
Drawing creativity (n⫽56) 2.76 (1.20) 2.33 (1.22) 3.08 (1.22) 3.08 (1.00) 2.70 (1.06)
1 to 5 1 to 5 1 to 5 1.33 to 4.33 1.33 to 4.33
Narrative creativity (n⫽72) 2.22 (1.25) 1.76 (.83) 2.56 (1.47) 1.83 (.98) 2.82 (1.42)
1 to 5 1 to 3.67 1 to 5 1 to 4 1 to 5
Role play creativity (n⫽45) 2.70 (1.17) — 3.29 (1.09) 2.03 (.81) 2.03 (.90)
1 to 5 1 to 5 1 to 3.67 1 to 3.67
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
282 MOTTWEILER AND TAYLOR

Are Children Who Engage in Elaborated Role Play
More Creative Than Other Children?
In order to examine this question, two ANCOVAs were con-
ducted with role play status (any form of elaborated role play vs.
no role play) as the between group factor and creativity scores
(drawing and narrative) as dependent measures, controlling for
age, vocabulary, and syntax. Separate ANCOVAs were conducted
for drawing creativity and narrative creativity because of the
difference in sample size for these comparisons (73 vs. 54 chil-
dren). The analysis for the narrative creativity measure yielded a
significant effect for role play status, F(1, 68) ⫽6.31, p⫽.039,
␩
2
⫽.06, with role players scoring higher than nonrole players.
The analysis for the drawing creativity measure yielded a trend for
the effect of role play status, F(1, 49) ⫽3.08, p⫽.09, ␩
2
⫽.06,
with role-players scoring higher than nonrole players.
We also compared children in each of the role play groups (i.e.,
invisible friends, personified objects, and pretend identities) with
children who did not engage in any form of elaborated role play.
In order to evaluate these differences, we used a Bonferroni
correction to set the significance level at .017. For narrative
creativity, children with pretend identities were rated as more
creative than nonrole players, t(38) ⫽3.01, p⫽.005, and there
was a trend for children with invisible friends to be more crea-
tive than nonrole players, t(48) ⫽2.44, p⫽.019. The comparison
with nonrole players was not significant for children with person-
ified objects (p⫽.83). For drawing creativity, there was a trend
for children with invisible friends to be rated as more creative than
nonrole players, t(33) ⫽1.81, p⫽.08). The comparison with
nonrole players was not significant for children with personified
objects (p⫽.14) or for children with pretend identities (p⫽.42).
Although the overall results demonstrate a connection between
engaging in elaborated role play and creativity, it appears that it
was primarily children with invisible friends and, for the narrative
creativity, children with pretend identities driving these findings.
This is mostly consistent with the creativity ratings of role play
characters: The invisible friends were judged to be more creative
than the personified objects, t(30) ⫽⫺3.26, p⫽.003, and the
pretend identities, t(32) ⫽3.42, p⫽.002. However, there was no
difference between the creativity of pretend identities and person-
ified objects, t(20) ⫽.02, p⫽.99.
Unlike the findings of relations between elaborated role play
and creativity, partial correlations (controlling for age and lan-
guage ability) indicated that the developmental level of pretend
play (action pantomime task) was not related to narrative creativ-
ity, r(67) ⫽.10, p⫽.42, drawing creativity, r(49) ⫽.20, p⫽.16,
or role play creativity, r(41) ⫽.19, p⫽.23.
Discussion
This study sheds light on how individual differences in creativ-
ity during the preschool years might be related to pretend play
activities. We found that elaborated role play was related to mea-
sures of narrative creativity and drawing creativity. In contrast, the
developmental level of children’s pretend play ability on the action
pantomime task was not related to narrative creativity, drawing
creativity, or role play creativity. The results comparing the spe-
cific types of elaborated role play with nonrole play suggest there
could be relative benefits associated with different types of role
play. Children with pretend identities and children with invisible
friends told the most creative stories, suggesting that these children
might be particularly inclined to act out various plots and include
inventive narratives in their play. Children with invisible friends
provided the most creative solutions to drawing a pretend person,
perhaps because inventing an invisible friend has some similarity
to the task of drawing a person who could not exist, and both tasks
involve visual imagery. In a study related to the latter point,
Tahiroglu, Mannering, and Taylor (2011) found that individual
differences related to visual imagery with invisible friends was
related to imagery use on laboratory tasks.
In addition to role play categories, we examined the creativity of
the children’s role play characters and found this to be correlated
with creativity on the two other creativity tasks. Invisible friends
were rated as more creative than personified objects or pretend
identities, probably because descriptions of invisible friends often
have idiosyncratic appearances, behaviors, and personalities. In
contrast, the descriptions of personified objects tend to be based
on the characteristics of the toys. And although pretend identities
can sometimes be quite inventive (e.g., an “old guy” who walks so
slow he can travel through time), they often include media figures
(e.g., Princess Jasmine or Spiderman) that are typically not con-
sidered particularly creative unless the child provides additional
elaboration.
Although our findings establish a relation between role play
activities and creativity, we cannot draw conclusions about cau-
sality. It could be that engaging in elaborated role play provides
Table 4
Summary of Correlations for Age, Vocabulary, Syntax, Narrative Structure, Pretend Play
Development, Drawing Creativity, Narrative Creativity, and Role Play Creativity
Measure 1 2 3 4 5 6 7 8
1. Age —
2. Vocabulary .20 —
3. Syntax .03 .58
ⴱⴱⴱ
—
4. Narrative structure .24
ⴱ
.06 .26
ⴱ
—
5. Pretend play
development .26
ⴱ
.29
ⴱ
.32
ⴱⴱ
.23 —
6. Drawing creativity .32
ⴱ
.21 .18 ⫺.17 .31
ⴱ
—
7. Narrative creativity .12 .29
ⴱ
.27
ⴱ
.08 .21 .46
ⴱⴱⴱ
—
8. Role play creativity .12 .16 .18 .08 .38
ⴱⴱ
.36
ⴱⴱ
.40
ⴱⴱⴱ
—
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
ⴱⴱⴱ
p⬍.001.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
283
ELABORATED ROLE PLAY AND CREATIVITY

children practice and opportunities to engage in the type of inven-
tion or imagery that promotes enhanced creativity on other types of
tasks. However, children who are more creative might also be
predisposed to engage in open-ended generative activities like role
play (as opposed to more concrete types of play, such as puzzles).
It is also possible that a third variable, such as flexible thinking,
leads children to be more creative and more likely to engage in
elaborated role play.
Our results contrast with those of previous studies that did not
find a relation between early role play activities and creativity
(Manosevitz et al., 1977;Pearson et al., 2001). However, past
research has relied on the Alternative Uses Task, which requires
children to generate multiple answers to the same question. For
this study, instead of relying on the quantity of products generated,
we developed measures of creativity that assessed the overall
creativity of products generated. In addition, the Alternative Uses
Task focuses on physical objects rather than social entities. The
story stem task and the drawing task both involved social content
(e.g., thinking about the behaviors of two children who found a
key, thinking about how to portray a person who is pretend), and
thus were possibly a closer match to the content of elaborated role
play in which children invent imaginary characters.
According to Mouchiroud and Lubart (2002), creativity in the
social domain is an understudied, but particularly important, area
of creativity to explore and understand. It could be that elaborated
role play is specifically related to creativity that includes social
content and not related to creativity that does not include social
content. This possibility is consistent with our finding that elabo-
rated role play was related to our creativity measures, whereas
children’s pretend play ability on the action pantomime task (a
measure of pretend play that is nonsocial in content) was not
related to our creativity measures. However, future research should
investigate this distinction more systematically by comparing chil-
dren’s performance on tasks that are different as a function of
social and nonsocial content (e.g., drawings of pretend vs. real
people and of pretend vs. real object.).
The limitations of this study should be kept in mind when
interpreting the results. The sample, although reflective of the local
demographics, was relatively homogenous and modest in size.
Another limitation was the substantial amount of missing data. We
chose to include all the available data rather than drop participants
with missing data, but this strategy limits the ability to include all
variables in a single analysis and opens up the possibility of bias
if the missing data were due to systematic factors. However, the
overall pattern of results found when using all the available data
was replicated when we conducted an analysis using only the 54
participants who had no missing data.
The issue of missing data was particularly problematic for
drawing creativity. Twenty five percent of the data were missing,
mostly due to children declining to participate or not completing
the drawing of the second picture. One possible explanation is that
some children became fatigued (the drawing task was one of the
last tasks in the procedure). Another possibility is that the children
who did not complete drawings of a pretend person might not have
been able to generate ideas for these drawings. However, this latter
explanation seems unlikely because these children did not score
significantly lower on the other measures of creativity. In future
research, the drawing task should be presented earlier in the
procedure, the procedure should be less lengthy, and/or children
who do not complete tasks should be queried about the reasons
they choose not to finish.
Another consideration for the interpretation of this study was the
selection of judges of creativity. According to Amabile (1982),
creativity ratings should be done by a group of “appropriate”
judges. For this study, we used three developmental psychologists
(one PhD and two graduate students), all of whom have expertise
in the development of children’s creativity and imagination. How-
ever, one of the judges had also served as an interviewer for half
of the participants. Although the data were deidentified, it is
possible there could be bias introduced by the codes from this
judge. However, the pattern of results was the same when we
conducted analyses in which only the means of the two other
judges were used.
Despite these limitations, this study provides new information
about young children’s creativity, how it is related to pretend play
and other characteristics, and considerations for measuring it in
this age group. An important direction for future research would be
to more fully explore the distinction between social and nonsocial
forms of creativity. This issue is related to a broader debate about
the extent that creativity is domain specific or domain general.
Some researchers suggest that creativity is a general cognitive
process or core ability that cuts across domains; if an individual
demonstrates creativity in one domain, he or she is likely to be
creative in other domains as well (T. B. Ward, Smith, & Finke,
1999). In contrast, other researchers argue that creativity tends to
be specific to a particular domain or task (Baer, 1998). The
empirical evidence does not provide consistently strong support
for either position. Studies that are interpreted as supporting a
domain-specific perspective, while demonstrating null correlations
between some domains, also typically find moderate correlations
between other domains (Baer, 1991;Han, 2003;Han & Marvin,
2002). In addition, studies that are interpreted as supporting a more
domain-general perspective, while demonstrating connections
across domains, typically find that the strongest connections are
within domain (Chen, Himsel, Kasof, Greenberger, & Dmitrieva,
2006;Conti, Coon, & Amabile, 1996). Some researchers opt for a
less polarized view, in which individuals have a general capacity
to be creative that is then developed in specific domains according
to individual interests and training (Kaufman, Cole, & Baer, 2009;
Plucker & Beghetto, 2004;Simonton, 2007). Overall, we suspect
that some aspects of creativity are domain specific, whereas other
aspects cut across domains, with generality and specificity shifting
somewhat depending on how creativity is defined and measured.
Our measures assessed creativity in the ways children com-
pleted a story, drew a pretend person, and described the characters
that served as their imaginary companions or pretend identities.
These tasks involve different types of activities and skills that,
arguably, could be related to different domains (narrative creativ-
ity, drawing creativity, and pretend play creativity). The positive
correlations among these measures are consistent with a domain-
general view of creative development, but this interpretation de-
pends upon how narrowly domains are defined and how one draws
the boundaries between domains. In recent research with adults,
there has been progress in identifying a five-factor structure of
creativity domains that is consistent with past adult research
(Kaufman, 2012). Some of the items used in this approach map
onto our tasks and are related to particular factors. For example, in
Kaufman’s model, the item “drawing a picture of something I’ve
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
284 MOTTWEILER AND TAYLOR

never actually seen (like an alien)” is strongly related to the
Artistic Domain. In contrast, the item “making up an original
story” loaded on three different factors with the strongest being
Scholarly Creativity. The Kaufman Domains of Creativity Scale
was developed for adults, but could be useful in guiding research
with children. Future studies that include batteries of creativity
tasks, clear accounts of what constitutes a domain, and descrip-
tions of how children’s performance on creativity measures change
over time will be important for understanding the role of domains
in the development of children’s creativity.
References
Ahn, J., & Filipenko, M. (2007). Narrative, imaginary play, art, and self:
Intersecting worlds. Early Childhood Education Journal, 34, 279–289.
doi:10.1007/s10643-006-0137-4
Alexander, P. A., Jetton, T. L., White, S. H., Parsons, J. L., Cotropia, K. K.,
Liu, H.-C., & Ackerman, C. M. (1994). Young children’s creative
solutions to realistic and fanciful story problems. The Journal of Cre-
ative Behavior, 28, 89–106. doi:10.1002/j.2162-6057.1994.tb00723.x
Amabile, T. M. (1982). Social psychology of creativity: A consensual
assessment technique. Journal of Personality and Social Psychology, 43,
997–1013. doi:10.1037/0022-3514.43.5.997
Baer, J. (1991). Generality of creativity across performance domains. Creativity
Research Journal, 4, 23–39. doi:10.1080/10400419109534371
Baer, J. (1998). The case for domain specificity of creativity. Creativity
Research Journal, 11, 173–177. doi:10.1207/s15326934crj1102_7
Beck, S. R., Robinson, E. J., Carroll, D. J., & Apperly, I. A. (2006).
Children’s thinking about counterfactuals and future hypotheticals as
possibilities. Child Development, 77, 413–426. doi:10.1111/j.1467-8624
.2006.00879.x
Busse, T. V., Blum, P., & Gutride, M. (1972). Testing conditions and the
measurement of creative abilities in lower-class preschool children.
Multivariate Behavioral Research, 7, 287–298. doi:10.1207/
s15327906mbr0703_2
Carlson, S. M., & Moses, L. J. (2001). Individual differences in inhibitory
control and children’s theory of mind. Child Development, 72, 1032–
1053. doi:10.1111/1467-8624.00333
Chen, C., Himsel, A., Kasof, J., Greenberger, E., & Dmitrieva, J. (2006).
Boundless creativity: Evidence for the domain generality of individual
differences in creativity. The Journal of Creative Behavior, 40, 179–
199. doi:10.1002/j.2162-6057.2006.tb01272.x
Conti, R., Coon, H., & Amabile, T. M. (1996). Evidence to support the
componential model of creativity: Secondary analyses of three
studies. Creativity Research Journal, 9, 385–389. doi:10.1207/
s15326934crj0904_9
Dansky, J. L. (1980). Make-believe: A mediator of the relationship be-
tween play and associative fluency. Child Development, 51, 576–579.
doi:10.2307/1129296
Dick, A. S., Overton, W. F., & Kovacs, S. L. (2005). The development of
symbolic coordination: Representation of imagined objects, executive
function, and theory of mind. Journal of Cognition and Development, 6,
133–161. doi:10.1207/s15327647jcd0601_8
Emde, R. N., Wolf, D., & Oppenheim, D. (2003). Revealing the inner
worlds of young children: The MacArthur Story Stem Battery and
parent-child narratives. New York, NY: Oxford University Press.
Gleason, T. R. (2004). Imaginary companions and peer acceptance. Inter-
national Journal of Behavioral Development, 28, 204–209. doi:10.1080/
01650250344000415
Gleason, T. R., Sebanc, A. M., & Hartup, W. W. (2000). Imaginary
companions of preschool children. Developmental Psychology, 36, 419–
428. doi:10.1037/0012-1649.36.4.419
Goldstein, T. R., & Winner, E. (2009) Living in alternative and inner
worlds: Early signs of acting talent. Creativity Research Journal, 21,
117–124. doi:10.1080/10400410802633749
Guajardo, N. R., & Turley-Ames, K. J. (2004). Preschoolers’ generation of
different types of counterfactual statements and theory of mind under-
standing. Cognitive Development, 19, 53–80. doi:10.1016/j.cogdev
.2003.09.002
Han, K.-S. (2003). Domain-specificity of creativity in young children:
How quantitative and qualitative data support it. The Journal of Creative
Behavior, 37, 117–142. doi:10.1002/j.2162-6057.2003.tb00829.x
Han, K.-S., & Marvin, C. (2002). Multiple creativeness? Investigating
domain-specificity of creativity in young children. Gifted Child Quar-
terly, 46, 98–109. doi:10.1177/001698620204600203
Harris, P. (2000). The work of the imagination. Oxford, UK: Blackwell.
Hennessey, B. A., & Amabile, T. M. (1988). Story-telling: A method for
assessing children’s creativity. The Journal of Creative Behavior, 22,
235–246. doi:10.1002/j.2162-6057.1988.tb00502.x
Hoff, E. (2005). Imaginary companions, creativity, and self-image in
middle childhood. Creativity Research Journal, 17, 167–180. doi:
10.1207/s15326934crj1702&3_4
Karmiloff-Smith, A. (1990). Constraints on representational change: Evi-
dence from children’s drawing. Cognition, 34, 57–83. doi:10.1016/
0010-0277(90)90031-E
Kaufman, J. C. (2012). Counting the muses: Development of the Kaufman
Domains of Creativity Scale (K-DOCS). Psychology of Aesthetics, Cre-
ativity, and the Arts, 6, 298–308. doi:10.1037/a0029751
Kaufman, J. C., Cole, J. C., & Baer, J. (2009). The construct of creativity:
A structural model for self-reported creativity ratings. The Journal of
Creative Behavior, 43, 119–134. doi:10.1002/j.2162-6057.2009
.tb01310.x
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. doi:10.1037/a0029321
Manosevitz, M., Fling, S., & Prentice, N. M. (1977). Imaginary compan-
ions in young children: Relationships with intelligence, creativity and
waiting ability. Journal of Child Psychology and Psychiatry, 18, 73–78.
doi:10.1111/j.1469-7610.1977.tb00418.x
Matuga, J. M. (2004). Situated creative activity: The drawings and private
speech of young children. Creativity Research Journal, 16, 267–281.
doi:10.1207/s15326934crj1602&3_10
Mayer, M. (1969). Frog, where are you? New York, NY: Dial Books for
Young Readers.
Mouchiroud, C., & Lubart, T. (2002). Social creativity: A cross-sectional
study of 6- to 11-year-old children. International Journal of Behavioral
Development, 26, 60–69. doi:10.1080/01650250042000591
Mullineaux, P. Y., & Dilalla, L. F. (2009). Preschool pretend play behav-
iors and early adolescent creativity. The Journal of Creative Behavior,
43, 41–57. doi:10.1002/j.2162-6057.2009.tb01305.x
Overton, W. F., & Jackson, J. P. (1973). The representation of imagined
objects in action sequences: A developmental study. Child Development,
44, 309–314. doi:10.2307/1128052
Pearson, D., Rouse, H., Doswell, S., Ainsworth, C., Dawson, O., Simms,
K., . . . Faulconbridge, J. (2001). Prevalence of imaginary companions in
a normal child population. Child: Care, Health and Development, 27,
13–22. doi:10.1046/j.1365-2214.2001.00167.x
Pepler, D. J., & Ross, H. S. (1981). The effects of play on convergent and
divergent problem solving. Child Development, 52, 1202–1210. doi:
10.2307/1129507
Plucker, J. A., & Beghetto, R. A. (2004). Why creativity is domain general,
why it looks domain specific, and why the distinction does not matter. In
R. J. Sternberg, E. L. Grigorenko, & J. L. Singer (Eds.), Creativity:
From potential to realization (pp. 153–167). Washington, DC: Ameri-
can Psychological Association. doi:10.1037/10692-009
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
285
ELABORATED ROLE PLAY AND CREATIVITY

Reilly, J. S., Bates, E. A., & Marchman, V. A. (1998). Narrative discourse
in children with early focal brain injury. Brain and Language, 61,
335–375. doi:10.1006/brln.1997.1882
Russ, S. W., Robins, A. L., & Christiano, B. A. (1999). Pretend play:
Longitudinal prediction of creativity and affect in fantasy in children.
Creativity Research Journal, 12, 129–139. doi:10.1207/
s15326934crj1202_5
Russ, S. W., & Schafer, E. D. (2006). Affect in fantasy play, emotion in
memories, and divergent thinking. Creativity Research Journal, 18,
347–354. doi:10.1207/s15326934crj1803_9
Schaefer, C. E. (1969). Imaginary companions and creative adolescents.
Developmental Psychology, 1, 747–749. doi:10.1037/h0028270
Silvia, P. J., Winterstein, B. P., Willse, J. T., Barona, C. M., Cram, J. T.,
Hess, K. I.,...Richard, C. A. (2008). Assessing creativity with
divergent thinking tasks: Exploring the reliability and validity of new
subjective scoring methods. Psychology of Aesthetics, Creativity, and
the Arts, 2, 68–85. doi:10.1037/1931-3896.2.2.68
Simonton, D. K. (2007). Creativity: Specialized expertise or general cog-
nitive processes? In M. J. Roberts (Ed.), Integrating the mind: Domain
general vs. domain specific processes in higher cognition (pp. 351–367).
New York, NY: Psychology Press.
Slobin, D. I. (2004). Relating events in narrative, Vol. 2:Typological and
contextual perspectives. Mahwah, NJ: Erlbaum.
Tahiroglu, D., Mannering, A. M., & Taylor, M. (2011). Visual and auditory
imagery associated with children’s imaginary companions. Imagination,
Cognition and Personality, 31, 99–112. doi:10.2190/IC.31.1-2.i
Taylor, M. (1999). Imaginary companions and the children who create
them. New York, NY: Oxford University Press.
Taylor, M., Carlson, S. M., Maring, B. L., Gerow, L., & Charley, C. M.
(2004). The characteristics and correlates of fantasy in school-age chil-
dren: Imaginary companions, impersonation, and social understanding.
Developmental Psychology, 40, 1173–1187. doi:10.1037/0012-1649.40
.6.1173
Taylor, M., Cartwright, B. S., & Carlson, S. M. (1993). A developmental
investigation of children’s imaginary companions. Developmental Psy-
chology, 29, 276–285. doi:10.1037/0012-1649.29.2.276
Taylor, M., Hodges, S. D., & Kohányi, A. (2002). The illusion of inde-
pendent agency: Do adult fiction writers experience their characters as
having minds of their own? Imagination, Cognition and Personality, 22,
361–380. doi:10.2190/FTG3-Q9T0-7U26-5Q5X
Taylor, M., Sachet, A., Mannering, A. M., & Maring, B. L. (2013). The
assessment of elaborated role-play in young children: Invisible friends,
personified objects and pretend identities. Social Development, 22, 75–
93. doi:10.1111/sode.12011
Trionfi, G., & Reese, E. (2009). A good story: Children with imaginary
companions create richer narratives. Child Development, 80, 1301–
1313. doi:10.1111/j.1467-8624.2009.01333.x
Wallach, M. A., & Kogan, N. (1965). Modes of thinking in young children.
New York, NY: Holt, Rinehart, & Winston.
Ward, T. B., Smith, S. M., & Finke, R. A. (1999). Creative cognition. In
R. J. Sternberg (Ed.), Handbook of creativity (pp. 189–212). Cambridge,
UK: Cambridge University Press.
Ward, W. C. (1968). Creativity in young children. Child Development, 39,
737–754. doi:10.2307/1126980
Winnicott, D. W. (2005). Playing and reality. New York, NY: Routledge.
Received November 13, 2011
Revision received December 11, 2013
Accepted December 20, 2013 䡲
E-Mail Notification of Your Latest Issue Online!
Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
286 MOTTWEILER AND TAYLOR