A Comparison of Turkish and South Korean Preschool Children’s Gender Self-Concepts

Abstract

personal standards for behavior.Children internalize lens of their larger culture for viewing gender and this process becomes the basis for a network of unchanged, rigid, and internalized knowledge, which Bem (1993) terms as gender schema.Piaget (1952) also stated that, like any other cognitive schema, gender schemas develop at early ages.They are derived from cultural values and prevalent way for children to develop gender schemes is through socialization.
The gender schema formed during early childhood can develop into gender stereotypes held by older men and women in relation to gender roles (Beal, 1994).Gender stereotypes are well developed by the age of 5 and are fairly rigid between the ages of 5 and 7 (Martin & Ruble, 2004).According to Trautner et al. (2005), gender stereotyping shows a developmental pattern that occurs in three ordered phases.First, children start to learn gender-related characteristics; this usually takes place in toddler and preschool years.Secondly, children reinforce their newly acquired gender knowledge and its rigidity reaches the peak between 5 to 7 years.Third and finally, after reaching the peak of rigidity, the phase of relative flexibility follows.In other words, as children get older, gender stereotypes become more flexible than their younger ages.However, even though these children have the cognitive ability to understand the flexible nature of gender roles and stereotypes, they still remain attached to these schemas -stereotypes-and this emerges as a spontaneous process (Banse et al., 2010).
There are two broad theories of gender that direct research on gender in the early years: Sex-role socialization theory and relational theory.While sexrole socialization theory defines gender as socially learnt, relational theory defines gender as socially constructed.To examine the gender self-concepts of Turkish and Korean preschool children in the present study, the conceptual framework used consisted of Bussey and Bandura's (1999) social cognitive theory of gender role development.Bussey and Bandura (1999) proposed that children develop concepts of gender from interactions with environmental influences.Children interact with people, institutions and objects that have effect on formation of their self-concept.These interactions have a crucial role on gender self-concept of the children.Sex-role socialization researchers including Bussey and Bandura identify four key agents in sex-role socialization: family, peer group, media and school (or early childhood setting).The messages they have received from these agents play a role in children's gender self-concept.They put forward that through imitation, observation and modeling, children learn how to behave in ways appropriate to their sex roles (Bussey & Bandura, 1999;McNaughton, 2000).
In this context, according to Bussey and Bandura (1999), gender is equated with a cultural rather than biological origin, supporting the role of environmental influence in its formation.On the other hand, Kohlberg (1966) suggested that gender identity did not stem solely from either biology or culture, but rather the product of the child's cognition regarding both.Kohlberg proposed three stages of children's gender identity process.Kohlberg's (1966) first stage, identity, was expected of children between the ages of 2 to 3 years and involved a child self-labeling as a girl or a boy.The second stage, stability, was expected of children between the ages of 4 to 5 years and involved anticipating identity in youth to predict that for adulthood, such as a girl expecting to grow up to be a woman.The final stage, constancy, was expected of children between the ages of 5 to 7 years and involved a child's ability to self-perceive oneself as female even if a strict adherence to gender roles was not demonstrated.According to Kohlberg's perspective, acquisition of gender consistency is crucial for children's development for gender identity (Green, Bigler & Catherwood, 2004).After they understand that gender is fixed regardless of the time, appearance and conditions, children focus their interests on behaviors, objects, activities, roles and characteristics that meet their cultural gender stereotypes with an increasing motivation (Arthur, Bigler & Ruble, 2009).
Both Turkish society and South Korean society continues to place limits upon women that have served to restrict access and choices.Preschool children develop self-concept about their identity, and they acquire gender perceptions in preschool period.It is vital to scrutinize how the children who grown in a world full of gender definitions evaluate themselves.The main purpose of current study was to better understand the self-concepts of preschool-age children in Turkish and South Korean cultures, between the ages of 3 to 6 years, regarding gender identity and gender roles.In this way, the study examined whether preschool children's self-concepts regarding gender is similar in Korea and Turkey since both countries are conservative in regard to gender norms.Age and cultural differences in children's gender expectations and gender role development were explored.

Research Model
This research was designed by a qualitative method, namely by a phenomenological research pattern, to collect in-depth information on thoughts and perceptions of preschool children about gender identities.One of the main contributions of the phenomenological approach in qualitative research is that it reveals how the individuals perceive and interpret the world and their own experiences (Yıldırım & Şimşek, 2011).

Study Group
Data sources in phenomenological researches are the individuals or groups who have experienced the phenomenon that the research is focused on and so can express or reflect it (Yıldırım & Şimşek, 2011).Typical case sampling method, one of the purposeful sampling methods, was used for study group selection.The characteristics of the sampling is that it is formed with typical one of many situations in the universe related to the research problem (Büyüköztürk, Çakmak Kılıç, Akgün, Karadeniz, & Demirel, 2011).This research was conducted with 80 preschool children who are between 36-72 months old, including 40 Turkish children and 40 South Korean children.Participants were evenly divided among two age groups (36-53 months old and 54-72 months old) and between male and female children within both countries.For each age group, 20 Turkish and South Korean children (10 female and 10 male) were included.
All children in the study were typically developing children from families of a middle class, urban community in Istanbul (Turkey) and Seoul (South Korea).While determining the participants, permission was obtained firstly from children's families, then from themselves.In both countries, children were recruited by sending home information about the study through their schools.Children whose parents signed and returned the consent forms were interviewed.

Data Collection
Semi-structured interview as a qualitative data collection method was used in this research.The participants answered a total of 11 questions.Four questions of the interview are open-ended while remaining 7 are close ended.Open-ended ones are related to children's thoughts about themselves, their likes and dislikes as girls or boys and things that they think a boy, or a girl cannot do.Questions generated to prepare and invite children for talking about themselves create a proper atmosphere to examine their gender-related self-perceptions.
Close-ended questions were adopted from the questions of Slaby & Frey (1975)'s Gender Constancy Interview that intend to measure gender constancy levels of the children.This inventory which contains highly structured binary and semi-structured reasoning questions was created to measure children's levels of Kohlberg's (1966) proposed stage of gender identity formation.The original inventory was designed to assess young children's notions of gender labeling, stability, and constancy (Kohlberg, 1966) regarding themselves and others.In the present study, only the second portion of the inventory comprised of 7 questions was used.

Data Analysis
Descriptive analysis technique was used for analyzing the data obtained from interviews.With this technique, obtained data were summarized and interpreted according to previously determined themes (Yıldırım & Şimşek, 2011).Interviews were recorded and transcribed.Findings were organized and defined in line with the previously determined themes, including "gender constancy" and of "perceptions related to one's own and opposite gender".The data were examined carefully considering the similarities and differences of the participants by the themes.
First, for "gender constancy" theme, children's answers to 7 close-ended questions were coded according to three levels of gender constancy, including gender identity, gender stability and gender constancy.Children who labeled their own gender correctly were accepted as they acquired the "gender identity level".For "gender stability level", children were asked whether they were same gender when they were babies and whether they would be same gender when they would become adults.If their gender labels were stable through time, they were accepted as they acquired the "gender stability level".For "gender constancy level", children were asked if they wear another friend's dress who is of opposite sex, his or her gender would be changed.If their gender labels were stable through situations, they were accepted as they acquired the "gender constancy level".
Second, for "perceptions related to one's own and opposite gender" theme, children's answers to 4 open-ended questions were evaluated.When children were asked the things that they do not like since they are boys or girls, the things that they think their gender cannot do and the things that they think opposite gender cannot do, their answers fell into four categories.First, some children considered themselves as boys or girls at first before being a child and they compared themselves with the opposite gender.Second, some children considered themselves as children at first and they compared themselves with adults regardless of their gender.Third, some children did not make any comparison and they only talked about themselves in general and gave examples from daily life.Fourth, some children thought there is nothing children of the same or opposite gender cannot do.

Validity and Reliability
Aiming to increase validity and reliability of the research, researchers spent two hours a-day in the classroom of participant children for children to feel confident with the researchers.Obtained data were analyzed by two researchers.Miles and Huberman (1994) formula [Reliability = Consensus / (Consensus + Dissidence) x100] was applied to the two researcher's encodings.The concordance between the researchers was calculated as 92%.70% or above scores are accepted as sufficient for reliability claims.

Procedure
Firstly, a "family consent form" was sent to the families of the children who will participate to the study through their teachers.In the family consent form, the purpose of the research and the interview process were expressed.In the week before the interviews, the researchers went to the schools and played a game with the children in their classrooms.Before the interviews, the researcher took verbal consent from children to get a conversation with them.
In both countries, children were interviewed individually at their schools in a private room.Interviews ranged from 15 to 20 minutes for each child.All interviews were audio-recorded for transcription.Children in Turkey were interviewed by the researchers.Children in Korea were interviewed by a native researcher, who is fluent in both Turkish and Korean.Interviews were coded by both authors.The data from the Korean sample was translated into Turkish by same South Korean researcher.Descriptive analysis was conducted on the data.Turkish children were named as T1, T2 and so on, while South Korean children were named as K1, K2 and so on.

Findings
In this section, the responses obtained from Turkish and South Korean preschool children were presented under four groups, of girls and boys 36-53 months old (as young children) and 54-72 months old (as older children).Findings were categorized under two themes of "gender constancy" and of "perceptions related to one's own and opposite gender".

Findings on "Gender Constancy" Theme
Table 1 shows the frequencies, by age and sex, of the responses of Turkish and South Korean preschool children to the questions asked in order to determine their gender constancy levels.
There is nothing he or she thinks that the children of opposite gender cannot do -- as children before being a boy or a girl and 2 of Turkish children stated that there is nothing, he or she thinks that the children of opposite gender cannot do.

Discussion and Conclusion
When Turkish and South Korean preschool children were asked whether they are girl or boy, they all express their gender correctly and state that they are pleased to define themselves as a boy or a girl.Their labeling their gender correctly is an expected result.Because, Kohlberg (1966) stated that labeling one's own gender correctly is a developmental phenomenon that takes place in early ages, it can be concluded that these children have reliable gender labeling.In other words, all the children have acquired the gender identity level.
On the other hand, children have shown a stability varying due to their ages and cultures on perception of being a girl or a boy.This constitutes the second level, namely the gender stability level of gender identity development which takes place about the ages of 4-5 according to Kohlberg (1966).While 8 of Turkish boys and girls from each age group are in gender stability level, 3 young South Korean girls and 2 young and 2 older South Korean boys are in gender stability.Gender labels of these children are stable through time but not stable through situations.Because these children thought that if they wear another friend's dress who is of opposite sex, his or her gender would be changed.There is only one young Turkish boy and 2 boys and 2 girls from older age group who have thought that their gender will not change, that is, who are in gender constancy level.On the other hand, 6 South Korean girls and boys from young age group and all girls and 8 of the boys from older age group are in gender constancy level.Therefore, the gender consistency of Turkish and South Korean children is at different levels.The last stage, according to Kohlberg (1966) takes place at the ages of 5-7, while recent studies show that children acquire the gender constancy notion developmentally at the ages of 4-5 (Bee & Boyd, 2007).Most of South Korean children participated to this research have fully acquired the gender constancy notion, while Turkish children have not yet.It seems that South Korean preschool children in this study comprehend better that gender is independent from physical appearance.
When participants were asked questions related to the things they do not like since they are boys or girls, and the things they think boys and girls cannot do, it was found that more South Korean children than Turkish children considered themselves as boys or girls before being a child and compared boys and girls.
In addition, it was also found that in both Turkish and South Korean cultures, more boys than girls and older children than younger children considered themselves as boys or girls before being a child.The results indicate that although in both cultures children have gender stereotypes regarding gender expectations and gender roles, more South Korean children in this study perceive themselves according to not their age but to the gender stereotypes.
In explaining perceptions of what girls and boys could not do, the most of the children within this study saw themselves foremost as their gender, described in comparison to opposite gender, rather than as children, described in comparison to adults.This result points to self-concepts not defined by a variety of personal experiences of childhood, but rather, defined by gender roles.The participants' selfdescriptions suggest that the children within this study may be controlled by gender stereotypes.It can be suggested that in pursuing their interests, the children did perceive blockades associated with gender roles.
It could be argued that the responses of children to the questions may lie in the patriarchal nature of Turkish and South Korean society.Both Turkey and South Korea have been found to have strict social norms (Oyserman, Coon & Kemmelmeier, 2002), especially about women's roles (United Nations, 2002).Regarding gender roles in the family, the traditional tendencies are still continuing in both South Korea and Turkey (Chin & Chung, 2010;Eun & Lee, 2005;Gedikli, 2014;Dedeoglu, 2010).Turkey can be argued to be still under the influence of patriarchal ideologies and traditional values are operating to the detriment of women and, eventually, of the whole society (Gedikli, 2014).While the policies aiming at the emancipation of women generated a small group of highly educated and economically active women in the urban areas of the country, they did not impact on rural women`s lives to the same extent (Gündüz-Hoşgör & Smits, 2008).Although some rapid social changes have also occurred in South Korea, its traditional gender roles structure has been maintained (Baek, 2009).
Self-concept of children is related to their experience, and gender is a socially constructed notion (Bandura, 1969;Bandura & Bussey, 1999).Gender identity develops as a result of learning process through observing the persons accepted as role models in genderrelated areas.Learning through observation conveys information on gender.This information guides children to develop self-efficacy beliefs, which will help to orient their gender-related behaviors (Bandura & Bussey, 2004).Therefore, the children in this study are influenced to a great extent by their environment in internalizing gender.If an environment transmits positive or negative messages about being a girl or a boy, children receive and interpret these messages.A research conducted by Ambady, Shih, Kim & Pittinsky (2001) shows that even 5-year-old girls can internalize the negative stereotypes about being a girl.Another research conducted by Epstein & Ward (2011) shows that children's gender stereotypes have great similarities with their parents' ones.Recent research has focused on parents' perceptions and expectations of gender roles in children, and indicates that children's gender typing is related to that which is expressed by their parents (Turner & Gervai, 1995;Heyman & Legare, 2004;Adams, Coltrane & Parke, 2007).In addition, classroom environments and teachers' expectations and feedback may further reinforce gender stereotypic beliefs and perceptions (Meece, Glienke & Burg, 2006).Children participated to this research are from families with middle socioeconomic status.It was also observed in the study of Dilek (1997) that despite their high economic status and education level, most of the parents convey their children to acquire a gender identity.
Results also conclude that the gender stereotypes increase by age in both Turkish and South Korean cultures.Similar to this finding, many studies find that children's gender stereotypes increase between the ages of three and five (Halim, Ruble, Tamis-LeMonda & Shrout, 2013;Ruble, Lurye & Zosuls, 2007, Trautner et al., 2005).In other studies (Çelebi Öncü & Ünlüer, 2012;Yağan Güder & Alabay, 2016), it was also found that gender stereotypes increase by age.Awareness or knowledge of gender norms increases with age through the preschool years (Blakemore, 2003).At the same time, the researches examining the differentiation of activity choices of children by time conclude that children's preference of gendered activities increase or remain constant by the age at early childhood (Golombok et al., 2008;Martin & Fabes, 2001;Martin & Ruble, 2009).
It is important to attend to the concepts preschool children incorporate into their self-perceptions for gender identity.The purpose of this research is not to generalize the findings.Being qualitative with relatively small number of participants who are homogeneous prevent making any generalizations.It can be expected that further research with the children from the families of different cultural, economic and education levels will give different results.Hence, more knowledge can be produced on how their environment influence children's gender identity acquisition.While a certain knowledge accumulation exists abroad, the research on this issue in Turkey is limited.It is needed additional qualitative and quantitative researches.Further studies should aim greater number of participants and heterogeneous groups.They can also examine how the parents perceive gender-related expectations of the society.They can create a wider perspective through comparing gender role perceptions of children with gender identity expectations of parents.Moreover, in today's world where the social pressure of technology increases, the effect of messages transmitted by media on children's gender roles can be examined.

Table 1 .
According to Table 1, two of young Turkish girls are in gender identity level while 8 of them are in gender stability level.In addition, eight of older girls are in gender stability level while 2 are in gender constancy level.Moreover, one of young boys is in gender identity level, while eight are in gender stability level and one in gender constancy level.Furthermore, eight of older boys are in gender stability level while two are in gender constancy level.Frequency Distributions of Participant Boys and Girls by Gender Constancy Levels and South Korean cultures, older children (9 Turkish and 13 South Korean children from older age group) than younger children (4 Turkish and 11 South Korean children from young age group) gave examples of things that they think the children of opposite gender do when asked the things that they do not like since they are boys or girls.As shown in Table4, when Turkish and South Korean children were asked the things that they think opposite gender cannot do, all South Korean children (n= 40) considered themselves as boys or girls before being a child and compared the things boys and girls can or cannot do, while 22 Turkish children in total did that.In addition, 16 of Turkish children considered themselves

Table 2 .
Frequency Distribution of Things Participant Boys and Girls Don't Like Since They Are Boys or Girls

Table 3 .
Frequency Distributions of Things that Participant Boys and Girls Have Though that Children of Same Gender Cannot Do

Table 4 .
Frequency Distributions of Things that Participant Boys and Girls Have Though that Children of Opposite Gender Cannot Do