Monsters vs. superheroes: aggression in children's storytelling

Storytelling has become an increasingly popular addition to preschool curricula. It provides children with an opportunity to dictate an original story that is transcribed verbatim by an adult. These stories provide teachers and researchers with valuable insight into children's thinking. Though very little research has examined the nature and content of children's stories, aggression is a common theme. This study provides a descriptive analysis of the aggressive acts, aggressors, and victims that children include in their stories. In addition, this study examines how age, gender, creativity, and aggressive behavior are related the use of aggression in storytelling. Over 1100 stories were collected from 239 children from 2003-2008. A subset of 75 children, Accounting;; for 513 stories, also participated in a creativity study where data on creativity and behavioral aggression were collected. Boys included twice as many aggressive acts in their stories than girls. Boys also included fighting, weapons, killing more often than girls, whereas girls included more instances of capturing than boys. Such gender differences may be explained by gender-targeted media. Age, creativity, and behavioral aggression were not significantly associated with aggression in children's stories. The lack of an association between behavioral aggression and aggression in stories supports the idea that many children use storytelling as a means of exploring otherwise threatening topics, like aggression, even if they do not normally act aggressively.

STORYTELLING by The University of Utah of the in of Family August 2009 MONSTERS VS. SUPERHEROES: AGGRESSION IN CHILDREN'S STORYTELLING by Michael Bradley Burton A thesis submitted to the faculty of in partial fulfillment requirements for the degree of Master of Science In Human Development and Social Policy Department and Consumer Studies The University of Utah Copyright © Michael Bradley Burton 2009 Rights Reserved C> All T H E U N I E R S I T Y U T A H G R A D U A T E S C H O OL SUPERVISORY COMMITTEE APPROVAL of a thesis submitted by Michael Bradley Burton majority vote has been found to be satisfactory. Chair: Cheryl Wrfolit U THE UNI V ERSlTY OF UTAH GRADUATE SCHOOL This thesis has been read by each member of the following supervisory committee and by satisfactory. 6 /11/ 200'1 Marissa Diener = Beverly Brehl T H E U N I V E R S I T Y OF U T A H G R A D U A T E S C H O OL APPROVAL I have read the thesis of. Michael Bradley Burton m j t s fm a l fo rm and have found that (1) its format, citations, and bibliographic style are consistent and acceptable; (2) its illustrative materials including figures, tables, and charts are place; and (3) the final manuscript is satisfactory to the supervisory committee and is ready for submission to The Graduate School 6/11/100°! Date Cheryl Wright $ ' \ Chair; Approved for the Major Department Cheryl Wright (J Chair/Dean Approved for the Graduate Council <-*7 r Charles A. Wight Dean of The Graduate School THE UNI SITY UTAH GR ADUA TE SCHOOL FINAL READING APPROVAL To the Graduate Council of the University of Utah: of in its final form fonuat, in fin al School. 1'l/lO oCf Cheryl Wrigh; z/ \ y ~ '"a" (' Chair: Supervisory Committee Cheryl w rigit! - () ......". - Z~ insight into children's thinking. Though very little research has examined the nature and content of children's stories, aggression is a common theme. This study provides a descriptive analysis of the aggressive acts, aggressors, and victims that children include in their stories. In addition, this study examines how age, gender, creativity, and aggressive behavior are related the use of aggression in storytelling. Over 1100 stories of 75 stories. The lack of an association between behavioral aggression and aggression in stories supports the idea that many children use storytelling as a means of exploring otherwise threatening topics, like aggression, even if they do not normally act aggressively. ABSTRACT Storytelling has become an increasingly popular addition to preschool curricula. It provides children with an opportunity to dictate an original story that is transcribed verbatim by an adult. These stories provide teachers and researchers with valuable were collected from 239 children from 2003-2008. A subset of75 children, accounting for 513 stories, also participated in a creativity study where data on creativity and behavioral aggression were collected. Boys included twice as many aggressive acts in their stories than girls. Boys also included fighting, weapons, killing more often than girls, whereas girls included more instances of capturing than boys. Such gender differences may be explained by gender-targeted media. Age, creativity, and behavioral aggression were not significantly associated with aggression in children's TABLE OF CONTENTS LIST OF FIGURES viii 1 2 Development and Aggression 3 4 Creativity and Aggression Behavioral Aggression and Fantasy METHODS 10 10 Measures 12 RESULTS 17 Unit of Analysis 17 Descriptive Analyses 18 Regression Models 21 DISCUSSION 30 Content of Children's Stories 30 Child Characteristics 33 Strengths and Limitations 36 TABLE OF ABSTRACT ... ................................................................................................................. iv LIST OF TABLES ..... .................................................................................................... vii ...................................... ................................................................ INTRODUCTION ........................................................................................................... 1 Aggression ...... .... ... ............................. ................................................................. Development and Aggression .............................................................................. 3 Gender and Aggression ................... .................................................................... .4 Creativity and Aggression .................................................................................... 5 Behavioral Aggression and Fantasy ..................................................................... 7 HYPOTHESES ................ ................................................................................................ 9 .................................................................................................................... Participants ......... .............................. ................................................................. Story Collection ......................... .... ... ... ... .. .......... ................ .................... ....... .... 11 Measures ................................................................. ........................... .. .............. 12 RESULTS .......................... ............................................................................................ 17 Unit of Analys is ... ............................ ... ... ................ ...... ...................................... 17 Descript ive Analyses .......................................................................................... 18 Regression Models ............................................................................................. 21 DiSCUSSiON ................................................................................................................. 30 Content of Children's Stories ................................................................. .... ........ .30 Child Characterist ics .......................... .... ............................................................ 33 Strengths and Limitations ....... .... ........... ...................... ..................................... .36 CONCLUSION REFERENCES CONCLUSION ............... .... ............................ ................... ................... ... .............. ....... .38 REFERENCES ........................................... ....... .................... ......................................... .39 VI Prompts Gender Victim Deviations by Coefficients: Coefficients: LIST OF TABLES Table Page 1. Story Pr0l11pts .............................................................. ... .. .. .... ... .. ... .. .. ... ... .. .. .. .......... 11 2. Aggression Categories ........................... .. ..... ...... ........ .. .. .. .. .. ... .. .. .. .. ... .. ..................... 13 3. Character Categories ..... ... .... ... ... .. .. .... ... .. .. ... .. ...................................................... ...... 14 4. Aggressive Theme Means and Standard Deviations by Gender ... .. ..... .. ... .. .............. 22 5. Aggressor and Victim Category Means and Standard Deviations by Gender. ........... 23 6. Aggressor and Victinl Gender Means and Standard Deviat ions Child Gender .... 24 7. Correlations between All Independent and Dependent Variables ............................. 25 8. Regression Coeffi cients: Aggressive Acts per Story as Dependent Variable ........... 27 9. Combined Regression Coeffi cients: Aggressive Acts per Story as Dependent Variable ....... ... ... .. .. ... .. .. .. .. ... ..... ...... ... ............................................... 28 LIST OF FIGURES 2. Percent of Stories Including Common Aggressor Categories 3. Percent of Stories Including Common Victim Categories Figure Page 1. Percent of Stories Including Common Aggressive Themes ...................... ................ 19 ......... .......................... 20 ....... ................................. 20 The storytelling INTRODUCTION Once upon a time, it was snowing. There's a kid making a snowman, and then he feeled the ground move, then he looked, a big monster, then Batman come and killed it. Then Spiderman killed Batman, tons of aliens killed Spiderman, aliens killed Spiderman. The end. (5-year-old boy) Storytelling has become an increasingly popular addition to preschool curricula, largely due to the work of Vivian Paley. In her books Wally's Stories (1981) and The Boy Who Would Be a Helicopter (1990), Paley describes the benefits of ~torytelling and dramatization, including providing teachers with insight into children's thinking and providing children with a safe way to explore ideas and themes that might otherwise be too threatening. Similarly, Vygotsky (1978) argued that play provides a medium in which children can easily remember, imagine, and explore ideas from their previous experiences, even though these same children might not be able to think through or discuss these ideas without the help of play. The telling and dramatization of stories mimics play in this way. Engel (1995) suggested that when children tell stories they combine fantasy with real experiences and real concerns, allowing them to take possession of those ideas and concerns. As a child tells a story, he gains mastery over those concerns or experiences that are included. Stories also provide teachers, parents, and researchers with insight into how children experience the world and how children think or feel. arise in their children's unprompted stories. One such theme that seems to come up repeatedly in children's stories is aggression. Though some researchers (Wright, Bacigalupa, Black, & Burton, 2007) suggest that the use of aggression in preschool children's stories should not be worrisome, little research has been done on the children who include aggression in their stories. This study provides a descriptive analysis of are associated with behavioral aggression are related to the use of aggression in 2 Despite the increase in the use of storytelling in preschool curricula, little research has examined the nature of the stories children tell when they are unprompted. Most previous research on preschooler's stories has examined stories told with a story stem or other prompt. Furthermore, some parents become upset with the topics that the aggression children include in their stories and examines how characteristics that storytelling. These characteristics include age, gender, and creativity. This study also attempts to determine if behavioral aggression is related to the aggression children include in their stories. Aggression In a review of aggression research, Coie and Dodge (1998) cited several definitions of aggression; the most fitting for this study came from Brain (1994). He suggested that an aggressive act needed three parts. First, the act must have the potential to do harm. Second, the act must be intentional; accidents are not considered aggressive. Finally, the act must be potentially aversive to the victim. These criteria are intentionally broad in order to include acts that result in a variety of undesirable outcomes, including, but not limited to, making a person uncomfortable, damaging their 3 & 2004). Since Crick and Grotpeter (1995) introduced the concept of relational aggression, which involves exclusion and attempts to damage a person's social standing, many researchers have included it in their studies as well. Studies have shown that with preschoolers, verbal aggression is more common than relational or physical aggression (Ostrov & Keating, 2004), but preschool children commonly use all three types of aggression (Crick & Grotpeter, 1995; Crick, Casas, & Ku, 1999). retaliate aggressively when provoked. From 12 to 18 months nearly half of children's peer exchanges involve conflict, usually over the possession of objects. Through the preschool years physical aggression decreases and verbal aggression increases. In one study (Dionne, Tremblay, Boivin, Laplante, & Perusse, 2003) it was suggested that the decline in physical aggression and increase in verbal aggression is due to the development of better language skills. They found that as expressive vocabulary increased, children were more likely to use verbal instead of physical aggression during conflicts. social standing, depriving them of something they had or wanted, and actually causing someone or something physical harm. 3 Aggressive acts can often be divided into categories. Many studies have focused on differences between physical and verbal aggression (Tapper Boulton, Development and Aggression Some clear trends emerge when looking at the literature on development and aggression. In a review of research on aggression, Coie ·and Dodge (1998) reported that aggression develops early in life. By the time a child reaches 12 months he is likely to 4 aggression moves from real life into the realm of fantasy; children begin to explore physical aggression through play, rather than in reality. Because storytelling is similar to play (Engel 1995; Paley 1981, 1984, 1990), it is probable that older children will include more aggressive acts in their stories than younger children. & & relationally & & differences appear across socioeconomic groups and cultures (Archer, 2004; Coie & Dodge, 1998). These differences also appear to be consistent across development (Capara, etal., 2001). As preschool children get older, they also become more comfortable talking about and using more aggressive themes in their play (Tegano, Moran, & Sawyers, 1991). It is possible that as children develop more advanced symbolic ability, physical Gender and Aggression In addition to developmental changes, there appear to be gender differences in children's behavioral aggression. Many studies have shown that boys are typically more physically and verbally aggressive than girls (Archer, 2004; Caprara, Barbarnelli, Pastorelli, 2001; Tapper Boulton, 2004) and that girls tend to be more relationally aggressive than boys (Crick, Casas, Mosher, 1997; Crick Grotpeter 1995). These development et aI., Ostrov and Keating (2004) had similar findings when observing preschool children during both free play and structured activities. They confirmed that boys were generally more physically and verbally aggressive than girls. They also found that boys were more often victims of physical and verbal aggression than girls. In addition to naturalistic observations of aggression, several studies have examined gender differences in children's use of aggression in structured play. 5 8 lA). sixth-graders. a l , different Tallandini (2004) used dollhouse play to explore gender differences in aggression for children in preschool and beginning school-age (ages 4 - 8 Y2). By counting the number of aggressive acts the children used during depictions of normal daily activities, she found that boys used more physical aggression than girls during dollhouse play. Studies have also examined gender differences in children's stories. Strough and Diriwachter (2000) studied gender differences in stories written by pairs of sixth­graders. They found that boy dyads included more physical aggression than girl dyads and mixed-gender dyads. Paley (1984) noted similar results in preschool and kindergarten aged children. She commented that boys tend to include more physical aggression than girls when telling individual stories. This study explores gender differences in preschool children's stories. Specifically, the differences in the use of physical, verbal and relational aggression, common categories of aggressive acts, and the gender of aggressors and victims are compared by gender. Creativity and Aggression Another child characteristic that may influence the use of aggression in stories is creativity. Creativity is a complex and oft debated topic among social scientists. It can be defined as the process of coming up with original, high-quality, and genuinely significant products (Tegano, et al., 1991). In young children, creativity is different because children are more concerned with the process of creation rather than the quality or significance of the final product (Tegano, et al., 1991). 6 dimensions of creativity and found ways to measure them. Two commonly examined dimensions in children are fluency and imagination (Tegano, et al., 1991; Torrance 1981) because they involve the process of creativity, rather than final products. Fluency is an aspect of divergent thinking. This is a type of thinking that produces multiple solutions to a problem, compared to convergent thinking where the goal is to converge on one "right" answer. A good example of a divergent thinking task would be to give a child a small drinking cup and ask them to come up with as many uses as possible for that cup. For example, a common use would be to use the cup to drink with, while a more unique response would be to use the cup as a telephone. Imagination is the ability of a child to freely create ideas and take on roles in pretend play. children are less constrained by social norms than less creative children (Dacey & Lennon, 1998). The theory suggests that a creative child is not afraid to push boundaries and do things that might not be entirely socially acceptable. Applying this to aggression, more creative children would be more likely to act out aggressively because they are less concerned with aggression being an unacceptable social norm. They might also be more willing to explore aggression in play, or storytelling, for the same reason. An opposing theory is that creative children are better able to come up with multiple responses to any given situation than are less creative children (Butcher & Niec, 2005). Creativity is difficult to measure, but Torrance (1981) identified several aI., converge on one "right" answer. A good example of a divergent thinking task would be to give a child a small drinking cup and ask them to come up with as many uses as possible for that cup. For example, a common use would be to use the cup to drink with, while a more unique response would be to use the cup as a telephone. Imagination is the ability of a child to freely create ideas and take on roles in pretend play. Few studies have been done on the link between creativity and aggression. However, there are two competing theories in the literature that can be applied to the link between aggression and creativity. One theory of creativity is that highly creative 7 child would be able to come up with more ideas on how to resolve a problem in a story, and be less likely to include aggression. Thus, these two theories of creativity suggest opposite predictions about the association between aggression and creativity, with one suggesting that creativity will be associated with greater aggression, and the other implying that creativity will be associated with less aggression. and the theories of creativity, it is likely that creative children will not be constrained by social norms in their stories. This trend is likely to be found in preschool children, in addition to second graders. The present study will examine how children's creativity is related to the aggression they include in their stories. fantasy When applied to aggression, this theory suggests that a creative child would have multiple responses available to choose from and be less likely to rely on aggression as a problem solving strategy. This might also appear in play and storytelling as a creative Little research has been done on the link between aggression and creativity in children. However, Tacher and Readdick (2006) studied the relation between creativity and aggression in second graders. They found a significant positive correlation between creativity and observed physical and verbal aggression. Based on this study Behavioral Aggression and Fantasy Despite clear relationships between age, gender, and creativity with behavioral aggression, it is less clear how behavioral aggression relates to the use of aggression in storytelling. It is not uncommon for parents to be concerned that children who use aggressive themes in their play will develop aggressive behaviors. However, few recent studies have examined the connection between this use of aggression in fantasy 8 in response to pictures, Lesser (1950) also found a positive correlation between aggression in children's stories and behavioral aggression as reported by peers. A more recent study showed that the use of aggression in the play narratives of 5-year-olds was correlated with behavioral problems (Klitzing, Kelsay, Emde, & Schmitz, 2000). The present study examines whether more behaviorally aggressive children use more and aggressive behaviors. Several older studies showed positive correlations between fantasy and behavioral aggression. For example, Kagan (1955) showed elementary school boys pictures and asked them to generate a story of what was happening. Boys rated as more aggressive by their teacher used more fighting themes in their stories than those who were rated as less aggressive. In a similar study that used stories generated aggression in their stories. 1) girls. will creative children. stories. HYPOTHESES I) Older children will have more aggression in their stories than younger children. 2) Boys will have more physical and verbal aggression in their stories than gi rl s. 3) Girls will have more relational aggression in their stories than boys. 4) More creative children have more aggression in their stories than less 5) Children with more behavioral aggression will have more aggression in their (M=53.7, SD=1.9). CFDC preschool, the sample was predominately White, middle class. The children from the Head Start classroom were more diverse, but again, no demographic data were available. In the 2003-2004 school year, a subset of 75 children (36 girls, 39 boys) from CFDC classrooms were rated on behavioral aggression by their teachers, and assessed for creativity. These children did not differ significantly from the total sample in age (/(228)=-1.56,/?= 12), story length (f(237)=-.81, p=.42), or aggression in stories(/(237)=1.61,;?=l 1). METHODS Data for this study came from two existing sources: a collection of stories told by preschool children and a collection of data on creativity and aggression from a subset of the same preschoolers. Hypotheses 1-4 will be tested using the entire collection of stories. Hypotheses 5 and 6 will be tested using stories from only those children on which creativity and aggressive data have been collected. Participants Stories were collected from 239 children (121 girls, 118 boys) from preschool classes in the Child and Family Development Center (CFDC) at the University of Utah from 2003-2008, and from one Head Start classroom in Salt Lake City, Utah, in 2006- 2007. Children's ages ranged from 33 months to 79 months SD=7.9). Demographic data were unavailable. However, based on average enrollment in the of75 from (t(228)=-1.56, p=.12), (t(23 7)=-.p=.42), stories(t(23 7)= 1.61, p= .11). Children voluntarily participated in the storytelling and were free to tell stories with very little prompting from the story transcriber (see Wright, et al., 2007). Prompts that were used included opening prompts, continuation prompts, and finishing prompts (Table 1). Children were free to tell stories involving any characters and plots they down. I! a...' ii1 Then happened? action)? n right' 11 Story Collection Storytelling is a part of the regular curriculum in the two preschools where stories were collected. Children told stories which were transcribed by teachers, parents, and volunteers during a time when children were free to choose from a number of offered activities. Typically, storytelling was available to the children once a week. aI., could think of. No character or situation prompts were given. Opening Prompts Continuation Prompts Closing Prompts Table 1 Story Prompts "Would you like to tell me a story?" "You tell me the words in your story and I'll write them down." If child still cannot begin, say "Some stories begin with 'Once upon a time', or 'Once there was a ... ' but you can start your story any way you want." "Then what happened?" Read last sentence and say "What happened next?" If child continues to repeat the same action with the same characters say "What happened after (character) did (action)?" "Let me read the story to you to make sure I got it right" "Thank you for telling me the story" following coded. After an initial training period, reliability for all categories (average Cohen's K=.812) was determined to be very good and coding continued on an individual basis. Questions arising during the coding process were brought to a meeting where a consensus of four coders and one noncoder was reached. 12 As the children told the stories, they were transcribed verbatim. When the child finished the transcriber read the story back to the child to ask if it was correct. Later in the day, the children acted out the stories in front of the class, using classmates as characters in the story. The opportunity for dramatization of the stories seemed to be the major motivation for the children to tell the stories. The stories were then digitized using a word processor and printed for coding. Measures Aggression in Stories Storytelling data were coded by four independent coders for number of aggressive acts. Each aggressive act was also categorized into one of the following common categories: ate, captured, fought, cast spell, killed, used weapon, stole, verbal aggression, and other (see Table 2 for categories and examples). During a pilot coding period, very few stories were found to include relational aggression, so relational aggression was dropped from the coding scheme. The category and gender of the aggressors and victims of each act (see Table 3 for categories and examples) were also Cohen's Jc=.13 Table 2 fitting Aggression Categories Category Includes words like: Ate Ate, Swallowed, Chomped, Bit, etc. Captured Captured, Locked up, Chased, Got, etc. Fought Got in a fight, Fought, Beat up, Crashed, Squashed, Stepped on, Ran over, Broke, Ripped, Destroyed, Blew up, Hit, Punched, Slapped, Kicked, Spat, Hurt, Pushed, Shoved, Threw, etc. Cast Spell Cast spell, Used Magic, etc. Killed Killed, or any aggressive action resulting in death Used Used a lightsaber, Shot, Stabbed, Poisoned, Threw into magma, etc. Weapon Stole Stole, Took away Verbal Threatening, Aggressive animal sounds directed at other characters Aggression (e.g. Bear growled at the man), Was mean to, Yelled, Used a bad word Other Aggressive actions not fitting into any of the above categories unicorns, fluency and imagination. Subtest A was used because Torrance (1981) reported that it had the highest test-retest reliability coefficient. Subtest B was chosen because it is the only test that gives an imagination score. Researchers spent a minimum of 2 weeks in each class getting to know the children before administering the tests. child. Then, the researcher asked the child if he or she could cross the room in a Category Real Human Fantasy Human Animal Other Creature Table 3 Character Categories Includes characters like: Daddy, a little girl, Kim, guys, cowboys, Indians superheroes, royalty, robots, ninja turtles, TV characters, Disney-type animals (e.g. those who speak); giants, bad guys, knights, wizards, witches bear, crocodile, bunny, sharks dinosaurs, monsters, dragons, unIcorns, aliens, ghosts, vampire, troll 14 Inanimate Car, truck, ball, the number 4, separate body parts (eye, nose, pinky, etc.) Object Creativity Creativity was measured using Torrance's (1981) "Thinking Creatively in Action and Movement (TCAM)." Subtest A and B were used to give a score for fluency In subtest A, the child was taken to a room with two tape lines on the floor. The researcher warmed the child up by walking back and forth between the lines with the 15 until the child indicated he was finished, usually by saying "that's it" or "that's all I can think of." fun SD=\IA) from S7_)=5.8) 018, 5!D=.a=.different way. After each time, the researcher would ask questions like "what else?" "what else can you do?" and "how else can you get across the room?" This continued until 1 of." The researcher then told the child that they were going to do another fun activity. The child was again warmed up by pretending to be a bird, an elephant and pretending to throw and catch a ball with the researcher. The child was then given instructions to act like six different characters in different situations. These included being a tree with a strong wind blowing, trying to push an elephant, and driving a car. Children's responses to both subtests were coded following the codebook provided by Torrance (1981). A fluency score (M=10.2, SD=II.4) was generated from subtest A, and was comprised of the total number of ways the child demonstrated for crossing between the two lines. Each part of subtest B was scored by the researcher on a scale of one to five, with one being no movement, three being average, and a score of five being given if the child used sound effects or showed a unique elaboration. For example, children received scores of one when they said "I can't" or shook their head. Children who received higher scores included sounds like grunting when pushing the elephant or elaborated the tree blowing in the wind by pretending the tree got blown over and crashed to the ground. Scores for all six parts were added together to form an imagination score (M=21.7, SD=5.8) for each child. A scale was created for creativity (M=.OI8, SD=.913) by standardizing and averaging each child's fluency and imagination score (Chronbach's 0.=.75). 16 TASB) l .74, SD=M). Behavioral Aggression Teachers rated children's behavioral aggression using the Cassidy and Asher (1992) Teacher Assessment of Social Behavior (T ASB) questionnaire. This questionnaire included six statements about aggression rated on a 5-point scale from 1 (very uncharacteristic of the child) to 5 (very characteristic of the child). The items were 1) This child starts fights; 2) This child interrupts other children; 3) This child is mean to other children; 4) This child bothers other children when they are working on things; 5) This child hurts other children; and 6) This child disrupts other children's activities. The responses on the six items about aggression were then coded so that higher scores indicated more aggression and the scores were averaged (M=1.74, SD=.84). Word Count It is plausible that the number of aggressive acts in a story could be affected by the length of the story (Strough & Diriwachter, 2000). Longer stories give children more opportunities to include aggression. Consequently, the number of words in each story was included in the analyses as a control variable. of the who told those stories. Thus, the analyses on gender and age differences in aggression in stories, and on the associations among creativity and aggression in stories, were conducted with child as the unit of analysis. When child was used as the unit of analysis, in order to account for the fact that some children told multiple stories, the characteristics of each child's stories were averaged, taking an aggregate story for each child. For example, if a child told two stories, and included two aggressive acts in one, RESULTS Three major analyses are presented in this section. First, a descriptive analysis ofthe types of aggression included by the children, including the prevalence of aggressive themes, aggressive character types, and victim character types, is reported. Second, gender differences in aggressive themes, character types, and character genders are explored. Third, a series of regression models are examined in order to test the research hypotheses. Unit of Analysis For this study, multiple units of analysis were used. First, in describing the content of the stories, it was most appropriate to describe the findings with story as the unit of analysis. Thus, for the descriptive analyses statistics will be referenced per story (e.g., number of aggressive acts per story). However, the descriptive analysis of gender differences and the research hypotheses are not about the stories, but about the children suggesting that the aggregated data set is representative of the children who participated in the study. various units of analysis. Noteworthy differences are noted. (SD=2.2) 7 5% themes/and four aggressive acts in another, his final aggregate number of aggressive acts would have been three. 18 To verify that these aggregate stories did not bias the results, analyses were also conducted several other ways. First, analyses were completed on all the stories, with the story as the unit of analysis. The analyses were also completed on a sample consisting of one random story from each child. Although sample size was smaller, the data were not aggregated in these analyses. There were no major differences in the results, Descriptive Analyses Aggression To describe the content of children's stories, the data in this section was analyzed with story as the unit of analysis. This is a more understandable way to describe what the stories are about than using the aggregate stories or random selection of stories. It is important to note that very few differences were found in using the Aggression was a common theme in children's stories. Fifty-one percent of stories included an aggressive act. Stories contained an average of 1.4 aggressive acts 2.2) but ranged from 0 to 19 aggressive acts. Additionally, with child as the unit of analysis, 75% of children told at least one story with aggression themes/references. Analyses also revealed common aggressive themes (see Figure 1). The most common act was fighting (included in 22.4% of stories), followed by captured( 18.0%), 19 20 o oo 15 o c 8 10 M PH 0 • 111 1 1 I l l I I I 1 1 . Aggression Category using a weapon (10.4%), killing (10.3%), and eating (9.0%). The most common category of aggressor (see Figure 2) was fantasy human (included in 25.8% of stories), followed other creature (13.5%), and animal (11.3%). The most common category of victim (see Figure 3) was fantasy human (included in 23.8% of stories), followed by other creatures (11.7%), real humans (10.8%). It is important to note that very few stories (2.0%) included real human aggressors acting aggressively against real human victims. Word Count Looking at all stories, children's stories averaged 66 words (SD=41), but stories ranged from 1 to 283 words. There was no gender difference (/(l 187)=.97,/?=.33) in 25 ~ v ·c .o-.0.n. 15 0 t: ev 10 v p.. 5 o Figure 1. Percent of Stories Including Common Aggressive Themes . by I (1(1187)=.97, p= .33) Human Human Object • • • 30 -I 25 GO CD •C 20 s 2 1 5 l u * 5 n • _ | • • • U i i i i Fantasy Other Animal Real Human Inanimate Human Creature Object V i c t im Figure 3. Percent of Stories Including Common Victim Categories 30 25 ~ ·-0c~ 20 VJ '0- 15 -c ~ 0 10 ~ ~ "- 5 0 Fantasy Other Creature Animal Aggressor Category Real Inanimate Object Figure 2. Percent of Stories Including Common Aggressor Categories ~ ~ ·c 20 .9 VJ '- 15 0 "~ 10 e ~ "- 5 0 Victim Category 20 46 66, 31). l 197)=.24,;?<.001) 00\). Differences kinds of aggressive acts in stories (see Table 4). Boys included more total aggressive acts in their stories than girls. Boys included more fighting, killing, and weapons than girls. The difference in including capturing approached significance, with girls including more capturing than boys. Gender differences also appeared in the aggressors and victims that boys and girls included in their stories (Table 5). Boys included more fantasy humans, animals, and objects as aggressors, than girls. Boys also included more animals and creatures as victims than girls. and victims than boys. length of story (Males: M=68, SD=46 ; Females: M=66, SD=37). Word count was positively correlated with both age (r(1197)=.24,p<.OOI) and creativity (r(513)=.22, p<.OOI). Gender Differences 21 Examining gender differences required looking at the data with the child as the unit of analysis. Using the aggregate story for each child, boys' and girls' stories were compared using t-tests. Several gender differences were found in the total number and Gender differences were also present in the gender of aggressors and victims children included in their stories (Table 6). Boys included more male characters as aggressors and victims than girls. Girls included more female characters as aggressors Regression Models It is again important to note that the hypotheses were about children. Thus, the aggregate stories, with child as the unit of analysis, were used in the following section. Table 4 by Boys /(1.88 1.96 .90 1.36 4.48*** Fight .84 1.42 .13 .28 .26 .62 .06 .14 3.51*** Killed .25 .60 .12 .33 2.07* .17 .29 .25 .40 -1.852' Spell .03 .16 .06 .19 Stole .03 .13 .02 .10 1.28 Ate .22 .49 .21 1.00 .12 Verbal .07 .18 .07 .33 p<.\ ****p<.00\ 22 Aggressive Theme Means and Standard Deviations by Gender Girls M SD M SD t(237) Total Aggressive Acts 4.48"'** .1 3 5.45*** Used Weapon 3.51 *** 2.07' Captured -.06 -.06 p<. ! ·p<.05 ... ·p<.OO I Table 5 by /(237) .16 .55 .08 .24 .80 1.30 .36 .27 .53 .40 2.20* Creature .31 .72 .19 .87 1.12 .12 .36 .03 .16 2.53* .23 .54 .16 .44 1.16 .53 .95 .39 .57 1.41 .22 .44 .09 .23 .39 .83 .09 .22 3.76*** .21 .44 .11 .63 1.36 lp<A 05 0l ***p<.00\ 23 Aggressor and Victim Category Means and Standard Deviations by Gender Boys Girls M SD M SD 1(237) Aggressors Real Human 1.49 Fantasy Human .59 3.36*** Animal .13 2.20' Other Crealure Inanimate Object 2.53' Victims Real Human Fantasy Human Animal 2.97** Other Creature 3.76*" Inanimate Object p <. l *p<.05 **p<.OI ***p<.OOl 24 /(4 42*** 01 ***/7<.001 Words indicating that multicollinearity was unlikely to be a problem. Correlations (see Table 6) between the independent variables and the dependent variable (Number of Aggressive Acts per Story) ranged from .04 (Behavioral aggression) to .44 (Words) indicating that the data were likely suitably correlated with the dependent variable for examination through multiple linear regression. 001), T, 29,/?<.word count and aggressive acts (r(239)=.55,/K.001), word count and age (r(239)=27, Table 6 Aggressor and Victim Gender Means and Standard Deviations by Child Gender Boys Girls M SD M SD t(237) Aggressors Male .87 1.28 .25 .87 4.42*** Female .10 .28 .29 .57 -3.24** Victims Male .71 1.25 .19 .43 4.35*** Female .09 .23 .33 .56 -4.35*** **p<.OI ***p<.OOI Before the regression analysis was performed, the correlations (Table 7) between all independent variables were examined. Pearson's correlations ranged from .01 (Words per Story and Behavioral Aggression) to .45 (Words per Story and Creativity) Words) Significant correlations were between aggressive acts and gender (boys=l, r(239)=0.28, p<.OOl), behavioral aggression and gender (boys=l, r(239)=.29, p<.05), 55,p<.r(239)=.25 Aggressive Creativity* Behavioral Aggression"1" 0.13£ Creativity* 0.204 0.211 Aggression"1" 027*** +Subset of 75 children p<A *p<.05 **p<M ***/?<001 00\), 29,/?<more creative, and including more aggressive acts. four previously mentioned hypotheses were first tested individually, with only the indicated independent variable and word count as a control variable for length of story (see Table 8). The independent variables were then combined (see Table 9) in order to determine the unique contribution of each to aggression in stories. 25 Table 7 Correlations between All Independent and Dependent Variables I Acts per Story Age I Gender I I Behavioral Creativity + Aggression + Age O.13 t Gender (boys=l) 0.28*** 0.04 Creativity + 0.20t 0.21t -0.16 Behavioral Aggression + 0.05 -0.01 0.29** 0.05 Words 0.55*** 0.27*** 0.012 0.29* -0.03 p<.l p<.OI ***p<,OOI p<.OOI), and word count and creativity (r(239)=.29,p<.05). These correlations indicated that boys included more aggression in their stories and were rated as more aggressive by teachers than girls. More words in a story were related to being older, With the child as the unit of analysis, multiple linear regression was employed with the number of aggressive acts per story as the dependent variable. Each of the p=.61) significant. if boys average, boys included one more aggressive act per story than girls (B = .95,/?<.001). Hypothesis 1 To test whether older children included more aggression in their stories, a regression was performed with age as the independent variable and number of aggressive acts per story as the dependent variable with word count as a control variable. The regression coefficient (B=-.005, 67) was not significant. 26 It is important to note that the averaging of the data created the average age for the child for all stories they told. Thus, for children who told stories at different ages or during different school years, the average age was used. This aggregated data may have affected the results. However, there were no differences in the results when using story as the unit of analysis (which used the age of each child each time they told a story) or one random story for each child (which required no averaging). Hypothesis 2 To determine ifboys included more aggression than girls, a regression was performed with child gender (Male = 1, Female = 0) as the independent variable and aggressive acts as the dependent variable and word count as a control variable. On .95,p<.Hypothesis 3 Because so few stories included relational aggression, the hypothesis that girls would include more relational aggression than boys was not tested. Model: 1 3.01*** 1.53*** 1.62*** - - - Gender(= 1) - - - + Creativity + - - 0.198 Subset children p<A */K.**p<.0\ ***/7<.001 27 Table 8 Regression Coefficients: Aggressive Acts per Story as Dependent Variable 2 3 4 Words/story (in hundreds) 3.01 *** 2.96*** 1.53*" 1.62*** Age -.01 Gender (boys ~ I) .95*** Creativity+ .17 Behavioral Aggression .09 R-squared 0.300 0.379 0. 198 0.198 Adj. R-squared 0.294 0.374 0.176 0.176 +Subsct of 75 chi ldren lp<.! *p<.05 "p<.Ol ·"p<.OO I 28 1.75*** Creativity* Aggression4 Adj. +Subset of 75 children 'p<A **/?<.01 ***/?<.001 a regression was performed the creativity scale as the independent variable and creativity, controlled, children their stories (B=.17,p=50). 28 Table 9 Combined Regression Coefficients: Aggressive Acts per Story as Dependent Variable Model: 5 Words/story (in hundreds) 1.75*"" Age -0.01 Gender .72*** Creativity+ .30 Behavioral Aggression + -0.04 • R-squared 0.303 Adj . R-squared 0.253 of75 ~<.I *p<.05 **p<.OI ·"p<.OOJ Hypothesis 4 To test whether more creative children include more aggression in their stories, perfonned with aggressive acts as the dependent variable with word count as a control variable. The regression coefficient was not significant for creativity. indicating that once word count was ~ontro ll ed. more creative chi ldren did not include significantly more aggression in (B~. 1 7, p='.the independent variable and aggressive acts as the independent variable with word count as a control variable. The regression coefficient (B=.09,/>=.51) was not significant, indicating that behavioral aggression was not predictive of aggression in stories. 53,/?<.significant. 2 8% story. Hypothesis 5 To test whether more behaviorally aggressive children included more aggression in their stories, a regression was performed with behavioral aggression as 09,p=.29 However, due to the lack of variability in behavioral aggression, this result was uncertain. To further test this hypothesis, participants were divided into half using a median split on behavioral aggression. Using this new variable and controlling for word count, the coefficient (B=.53, p<.05) was significant, such that children who scored in the top 50 percentile in behavioral aggression included more aggression in their stories than children scoring in the lower 50 percentile. Predicting Aggression in Stories from Child Characteristics By combining the above independent variables into one model it was possible to determine the regression coefficients while controlling for all other independent variables. When including all variables, age, gender, and word count were significant. Together, these factors accounted for about 28% of the variance in aggressive acts per revealing. Even children who do not normally behave aggressively may like to tell stories about aggression. In this study, the average teacher ratings of behavioral aggression were very low, yet three-quarters of the children in this study told at least one story with aggressive themes and half of all stories contained some type of aggression. Thus, children who rarely act aggressively towards peers may enjoy telling and acting out stories involving things they would never intentionally do. like fairytales than diary pages. Very few of the stories involve a real-life person acting aggressively towards a real-life victim. Instead, children often tell stories that are similar to other stories they have heard, read, or seen. Though most stories that children tell are not exact retellings of other stories, many children borrow characters from the media and use them in new, but similar plots. For example, one boy said, DISCUSSION Content of Children's Stories Children's stories provide an opportunity to see what children are thinking about. When children are free to talk about anything, what they choose can be very At first look, the fact that children include so much aggression in their stories may be concerning to some. However, examining the context of aggression in the stories can help alleviate some of that concern. It becomes clear that children's aggressive stories are fantasy-based. The characters and events they describe read more ofthe the" Once upon a time there was a Spiderman and he needed to get some bad dead... Vi-year-guys. And then he bumped into some bad guys and he fighted some bad guys. And then the bad guys fighted him back and then he punched them and they all got dead ... (Four-year-old boy) 31 This story is probably not an exact retelling of something this child saw on TV, yet the main character, Spiderman, fights bad guys like he would probably do on TV or in a book. Similar examples can be found in girls' stories. One 4 Yz-year-old-girl said, "There was a princess named Belle but her dad was really really mad at her stepsisters. Then the stepsisters were really really upset and then the Beast came and he destroyed the stepsisters." Again, the child borrowed characters from other stories, but changed the plot around. Even if children do not use the names of characters from other stories, the plots are often similar. Consider this girl's story, One time there was a princess and there was a queen and a king and there was a prince. And they loved the princess and then the witch came and cast a spell on the princess. And then the prince came and saved the princess and the prince got cast a spell on too. And the king and the queen uncast the spell. The end. (four-year-old girl) Note the generic princess that is the main character of her story. This princess finds herself in a plot very similar to other common princess stories. The media may also account for gender differences in the types of aggression included in stories. Boys may choose to include fighting so often because the characters they use come from the media to which they are exposed. Many of the television shows, movies, and books that target little boys portray fighting and weapons. For example, boys often tell stories about Transformers fighting each other, or various superheroes fighting "bad guys." Girls' stories included a third more instances of being captured than boys' stories. The influence of the media may also 32 are very common from the girls who participated in this study. the same way that this study did. However, the authors did not allow the children to re­tell stories. They only accepted what they deemed to be original stories, those without characters and plots from other stories. Even when excluding stories about characters from the media, a large number of the stories in their study also included aggression. This suggests that children use storytelling to explore aggression, even without the prompts that arise when using characters that perform aggressive acts in media. the story. For example, one 4-year-old boy included in his story, " . . . a bad guy just killed one girl, a princess. And then a prince come and kissed her and she woke u p . . ." Young children often do not understand the irreversibility of death (e.g., Hunter & Smith, 2008). They often portray those who die as just temporarily asleep or away. So, though the killing and death present in the stories may seem rather violent or severe to an adult, it may not be so disturbing to a child. Other studies can be done to understand how children portray death in their stories. 32 explain this difference. Many television shows, movies, and books that target girls include examples of princesses getting locked up to be saved by a prince. These stories There is some evidence that the media may not entirely explain the aggression in children's stories. Pitcher and Prelinger (1963) asked children to tell stories in much Another possible concern is that killing is so common in the stories. Killing was the fourth most common aggressive act, and was present in over 10% of the stories. This becomes less concerning when taking a closer look at the stories than include killing. In many instances, the person or creature that is killed comes back to life during ... up ... " 33 while they were telling them. It was very common to find girls' stories starting out with "once there was a princess named" and then having the girl's own name. They indicated that they were going to be a princess when the story was acted out. Another example of this is from a 5-year-old boy's story. He said, "Well there were four little ghosts. And it's going to be Sam, and Callie, Crista, and me. And they were going to a haunted house and a witch was going to get them, and it was Rachel..." Notice how he indicates in the story who will be acting out each part. This is evidence that some younger children. However, the results of the regression showed that age had very little if any association with aggression. Recall that age was correlated with story length, and story length was the strongest predictor of aggression in stories. Older children told longer stories, and thus had more aggression. However, controlling for length of story, older children were no more likely to include aggression than younger children. One possibility for this is that word count and age are measuring something very similar: The results from the analysis of gender differences in the stories also show that boys tell stories about male characters and girls tell stories about female characters. This may be explained by the fact that the children anticipated acting out their stories Rachel ... " children told the stories with the intent of acting them out later. Child Characteristics Now that we have explained the type of aggression that children use in their stories, we can look at which children included more aggression in their stories. It was suggested earlier that though older children generally engage in less physically aggressive behavior, their stories might have more aggression in their stories than of ways, including better language ability and better ability to engage in pretend play. Age is just a rough way to measure development, and word count may measure verbal ability, or symbolic ability, to a certain extent. When both word count and age were included in the same regression model, word count proved to be a better measure of symbolic ability, thus making the contribution of age insignificant. The fact that word count and age were highly correlated supports the idea that older children include more aggression in their stories because they have a higher symbolic ability. al., & many aggressive acts in their stories than girls. Perhaps more interesting is that girls also included so much aggression. On average, girls included one act of aggression in each story. Some studies have found that boys and girls often show the same amount of overall aggression, but that the types of aggression are typically different. Boys usually demonstrate more physical and verbal aggression and girls use more relational aggression. If the girls would have included more relational aggression in their stories, perhaps the results of aggressive acts would have reflected this finding. It appears that 34 symbolic ability. According to Piaget, as children develop, their symbolic ability increases (Crain, 2005). This development of symbolic ability reveals itself in a variety symbolic ability, thus making the contribution of age insignificant. The fact that word count and age were highly correlated supports the idea that older children include more aggression in their stories because they have a higher symbolic ability. The hypothesis that boys would have more aggression in their stories than girls was confirmed through the regression analysis. Many studies have shown that boys are typically more physically and verbally aggressive than girls (Archer, 2004; Caprara, et aI., 2001; Tapper Boulton, 2004). Other studies have also shown that boys have more aggression in their play (Tallandini, 2004). Since storytelling and dramatization is like play (Engel, 1995; Paley, 1981, 1984, 1990) it is not surprising that boys had twice as 35 turn-this study were not very behaviorally aggressive. A broader range of behavioral aggression might have produced more significant effects. even though these young children experience relational aggression (Crick, et al., 1997), that they are not developmentally able to talk about relational aggression occurring. The hypothesis that more creative children would have more aggression in their stories was not confirmed in this study. This may support the second theory of creativity presented in this study. That is, more creative children can come up with multiple solutions to a given problem and thus be less likely to resort to aggression. Though this is not supported through the analyses, there is some anecdotal evidence. In the following story, notice how the author (who has the second highest creativity score) resolves the conflict without aggression. Once there was a sister and a brother. And the sister and the brother always played together. And the sister and the brother did art together. One day their mother said to the sister and the brother, that they had to get online and they couldn't get online together. So the brother had to wait a minute because there was one computer. Then when the sister got online then the brother got online for a minute. Then their mom said they could play together. The end. (5-year-old girl) The conflict between the brother and sister in this story could very easily have resulted in some sort of aggressive conflict. Instead, the child chooses a tum-taking solution to the problem. The hypothesis that children rated as more behaviorally aggressive would have more aggression in their stories showed mixed results. However, when controlling for age, gender, and creativity, behavioral aggression was not a significant predictor of aggression in children's stories. This result could be due to the fact that the children in effects. play acting may be unfounded. Many children used storytelling as a way to explore aggression, yet not many were rated as aggressive by their teachers. Storytelling allows children to explore aggression in a safe and controlled setting, where the children know the teacher will prevent any real injury. This allows children to explore and develop their understanding of aggression, a topic that many children are clearly thinking about, without actually having to become aggressive. analyses. additional way to understand children's thoughts. this study have probably not experienced serious aggression or violence or seen its 36 The lack of an association between behavioral aggression and aggression in fantasy has immediate implications. The worries of some parents and teachers that aggressive storytelling and playacting are indicative of aggressive behavior in children Strengths and Limitations One of the strengths of this study was the total number of stories that were collected. Over a thousand stories were analyzed for the descriptive analysis giving an informative look into how children portray aggression in their stories. The overall number of children who participated also added to the generalizability of these Another strength of this study is that storytelling provides a different way to look into children's thinking. Few studies have used storytelling as a way to understand what children are thinking about. Future research could include storytelling as an One potential limitation of this study comes from the sample. Most children who participated were white, middle to upper class families. As such, most children in stories that are very different from children in this sample. Such children may need more support in coping with aggression that storytelling is able to provide. from those who choose not to participate. For example, it is possible that more aggressive children are less likely to tell stories. If such were the case, the sample of stories included in this study may not reflect the thoughts of the children most likely to be aggressive information determine effects first hand. It is possible that children who experience serious violence may tell 37 Additionally, children in this study were always free to choose to participate in storytelling. It is possible that children who choose to tell stories are different from aggressIve Another limitation of this study was that each child did not tell the same number of stories. This fostered the possibility of bias in the choice of the unit of analysis. However, as the results were consistent across the different units of analysis, this probably did not affect the integrity of the study. Much of the discussion is based on the influence of media in the stories. Unfortunately, the available data did not include a way to measure the amount and kinds of media the children were exposed to. Future research should include this infonnation in order to detennine the effect that these media have on the content of children's stories. Children are exposed to aggression in many forms of media. Fairytales, television, movies, and books often contain aggressive themes. It should not be very surprising that when we ask children to tell a story that the characters in their stories do exactly the same kind of things. Gender differences in the types of aggression and characters used may also be a product of media targeting particular genders. who are not behaviorally aggressive may include just as much aggression in their stories as more behaviorally aggressive children. CONCLUSION Storytelling provides children with an opportunity to explore aggression in a safe and controlled environment. Aggression is a common theme in children's stories. Age and creativity were not significantly associated with the amount of aggression present in children's stories. Gender was a significant predictor of aggression in stories with boys including more aggression than girls. Surprisingly, aggressive behavior was not predictive of aggression in stories, suggesting that children J. Review of General Psychology, Understanding and preventing violence, Vol. Biobehavioral influences Washington, DC US: National Academy Press. J. C , & Prosocial J. J. R. 63, 350-365. J., Handbook of child psychology, 5th ed.: Vol Social, emotional, and personality development NJ John & Theories of development: Concepts and applications NJ J.social-psychological Child Development, R., J.R., J.J.Jossey- Bass REFERENCES Archer, 1. (2004). Sex differences in aggression in real-world settings: a meta-analytic review. o/General 8, 291-322. Brain, P. (1994). Hormonal aspects of aggression and violence. and 2: (pp. 173-244). Butcher, 1. L., & Niec, L. N. (2005). Disruptive behaviors and creativity in childhood: The importance of affect regulation. Creativity Research Journal, 17, 181-193. Caprara, G.V., Barbaranelli, c., Pastorelli, C. (2001). Pro social behavior and aggression in childhood and pre-adolescence. In A.C. Bohart & D.l. Stipek (Eds.), Constructive and destructive behavior (pp. 187-204). Washington, DC: American Psychological Association. Cassidy, 1. and Asher, S.R (1992). Loneliness and peer relations in young children. Child Development, 63,350-365. Coie, 1., & Dodge, K. (1998). Aggression and antisocial behavior. In W. Damon (Ed.), 3. and (pp. 779-862). Hoboken, Nl US: 10hn Wiley Sons Inc. Crain, W. (2005). (5th ed.). Englewood Cliffs, Nl US: Prentice-Hall, Inc. Crick, N.R., Grotpeter, 1.K. (1995). Relational aggression, gender, and social­psychological adjustment. 66, 710-722. Crick, N.R, Casas, 1.F., Ku, H. (1999). Relational and physical forms of peer victimization in preschool. Developmental Psychology, 35, 376-385. Crick, N.R, Casas, 1.F., Mosher, M. (1997). Relational and overt aggression in preschool. Developmental Psychology, 33, 579-588. Dacey, 1.S., & Lennon, K.H. (1998). Understanding creativity. San Francisco: 10ssey­Bass Publishers. & childhood. & communicative & Of & Gender-Adolescent Psychology, J.Social 13, 255-277. Dionne, G., Tremblay, R., Boivin, M., Laplante, D., Perusse, D. (2003). Physical aggression in 19-month-old twins. Developmental Psychology, 39, 261-273. 40 Engel, S. (1995). The stories children tell: Making sense of the narratives of childhood. New York: W.H. Freeman and Company. Hunter, S., Smith, D. (2008). Predictors of children's understandings of death: age, cognitive ability, death experience and maternal communicative competence. Omega: Journal of Death Dying, 57, 143-162. Kagan, J. (1995). The Measurement of Overt Aggression from Fantasy. Journal Of Abnormal Psychology, 52, 390-393. Klitzing, K, Kelsay, K., Emde, R., Schmitz, S. (2000). Gender-specific characteristics of 5-year-olds' play narratives and associations with behavior ratings. Journal of The American Academy Of Child And Adolescent Psychiatry, 39, 1017-1023. Lesser, G.S., (1950). The relationship between overt and fantasy aggression as a function of maternal response to aggression. Journal Of Abnormal Psychology, 55,218-221. Ostrov, 1.M., & Keating, C.F. (2004). Gender differences in preschool aggression during free play and structured interactions: an observational study. Social Development, J 3,255-277. Paley, V.G. (1981). Wally's stories. Cambridge, Massachusetts: Harvard University Press. Paley, V.G. (1984). Boys and girls: superheroes in the doll corner. Chicago: The University of Chicago Press. Paley, V.G. (1990). The boy who would be a helicopter. Cambridge, Massachusetts: Harvard University Press. Pitcher, E. G., & Prelinger, E. (1963). Children tell stories. New York: International Universities Press, Inc. Strough, J., & Diriwachter, R. (2000). Dyad gender differences in preadolescents' creative stories. Sex Roles, 43, 43-60. Tacher, E., & Readdick, C. (2006). The relation between aggression and creativity among second graders. Creativity Research Journal, 18, 261-267. Tallandini, M. (2004). Aggressive behavior in children's dolls' house play. Aggressive Behavior, 30, 504-519. 41 childhood movement. 1L: C , C , Education 35, 363-369. Tapper, K., & Boulton, M. (2004). Sex differences in levels of physical, verbal, and indirect aggression amongst primary school children and their associations with beliefs about aggression. Aggressive Behavior, 30, 123-145. Tegano, D. W., Moran, J.D., Sawyers, J.K. (1991). Creativity in early childhood classrooms. United States: National Education Association. Torrance, E. (1981). Thinking creatively in action and movement. Bensonville, IL: Scholastic Testing Services. Vygotsky, L.S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. Wright, c., Bacigalupa, c., Black, T., Burton, M. (2008). Windows into children's thinking: a guide to storytelling and dramatization. Early Childhood Education Journal, 35,363-369.