This item is only available as the following downloads:
WHAT A GIRL WANTS : INCORPORATING BELIEFS IN TO THE STUDY OF GENDER SELF SOCIALIZATION BY JESSA BAKER MOSS A Thesis Submitted to the Division s of Psychology and Gender Studies New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Michelle Barton Sarasota, Florida April, 2013
Idiosyncratic Beliefs and Gender Self Socialization ii Acknowledgements I would like to thank my academic sponsor, Dr. Michelle Barton, as well as Dr. Steven Graham and Dr. Catherine Cottrell for spending countless hours guiding me through this project and preparing me for all of my future academic endeavors. Thanks are also due to D r. Amy Reid for sponsoring all three of my ISPs, each of which gave me the opportunity to explore the world and myself. Thank you to the staff and students of Girls Inc. of Sarasota for opening up your doors to me and providing me with the kindness, wi sd om and inspiration without which my research would not have been possible. I am constantly in awe of the work that you do for our community. Special thanks to the girls who took part in this study for being patient with me even though I know you were tired from school and my surveys were long and boring. You helped me because you believe that it s important to support those around you, and because of that belief you will make the whole world better. Thanks to Dr. Anne Fausto Sterling, Dr. Irene Frieze, and Dr David Perry for leading the field of Feminist Psychology, publishing the research that my study is grounded upon, and for responding to my e Mom and Dad & the rest of my family thank you for bei ng my role mod els, my best friends, my proof r eaders, and for financing my out of state tuition. You taught me unconditional love. Thank you Lacy for all the nights you slept with the lights on and all the mornings (literally all of them) that you made t he coffee, for reading/editing my thesis a million times and rubbing my back and letting me watch Xena and being my partner. To Emily Adams, Zoe Rayor, Arielle Scherr, Tyler Whitson, Jordan Campbell, Zach Eidelman, Zoe Posner, Gabe, Laura, and Annemarie, thank you for being my friends Finally, thank you to the guy in the library who loaned me a dollar for printing because it was a thesis support and empathy, which runs so deeply that it can be sensed even among strangers. I think rather than by competition between students for better grades and test scores, a place where we learn because of our social community rather than in spite of it. Thank you New College. I will miss you when I go.
Idiosyncratic Beliefs and Gender Self Socialization iii Table of Contents AC ... INTROD The Gende .. Self Percep 17 .. 27 Testing the three GSSM Free Response Scales: An Alternative?...............................................................47 THE DISCUSSI Tests of the th REFEREN TABLE S APPENDIX A APPENDI X B
List of Tables TABLE 1 TABLE 2 .. TABLE 3 ............................................. .122 TABLE 4.......... ................................. TABLE 5 ....... 26 TABLES 6 10
Idiosyncratic Beliefs and Gender Self Socialization v WHAT A GIRL WANTS : INCORPORATING NCRATIC BELIEFS INTO THE STUDY OF GENDER SELF SOCIALIZATION Jessa Baker Moss New College of Florida ABSTRACT The Gender Self Socialization Model (GSSM; Tobin et al., 2010) is an innovative theory of gender development which delineates gender stereotype beliefs self perceptions and gende r identity as distinct cognitive constructs, each of which is hypothesized to be dependent on the multiplicative properties of the other two. However, few researchers have tested this prediction, and fewer still have heeded call to co on two assessments of stereotyping; one which used a traditional format and one which used a free response format. Then, gender egalitarianism scores were derived from both assessments of self perceptions and gender identity. Participants were 45 girls aged 7 11 Results f ound that the free complex beliefs about gender, which were not reflected on the traditional scale. Overall, iefs, self perceptions, and gender identity would be mutually interactive. The importance of understanding discussed. __________________ ___ Professor Michelle Barton Division of Social Sciences
Idiosyncratic Beliefs and Gender Self Socialization 1 What a Girl Wants : Incorporating Ch i nto t he S tudy o f Gender Self Socialization The notion that the sexes are irreconcilably different from one another is pervasive in contemp orary Western culture. As feminist theorist Judith Lorber (2000) explains, the fact that every society will inevitably divide men and women into two categories of people remains a modern doxa com es without 1977, p. 167; original emphasis). The rigidity of this dichotomy is emphasized especially to children F or example, a visit to any toy store in the United States will be likely to reveal one side marked by shades of blue to denote tha with the fuchsia hues usually reserved for girls. In fact, these colors are used so male children prefer their respective shade because of neurological differences, when in reality this trend was reversed pink was for boys and blue for girls until the mid 20th century (Maglaty, 2011). Yet despite the arbitrary origins of many of the attributes that society associates with sex even the smallest deviation from traditional gender roles can incite social outrage.
Idiosyncratic Beliefs and Gender Self Socialization 2 Consider the controversy evoked by a 2011 spread in JCrew magazine which s toenails neon pink (shown above). The ensuing backlash against the photo was covered by conservative organizations such as Fox News, Media Research Network Center (MRNC), and One Million Moms, as well as mainstream news outlets including ABCNews The Wal l Street Journal CNN, and the Los Angeles Times (Wade, 2011). The article published by MRNC was ostentatiously titled News family photo : This is a dramatic example of the way that our culture is being encouraged to abandon all trappings of gender identity homogenizing males and females when seem like no big deal, it wil l be a very big deal if it turns out that neither gender is very comfortable anymore nurturing children above all else, and neither gender is motivated to rank creating a family above having great sex forever and neither gender is motivated to protect the nation by marching into combat against other men and risking their lives. This controversy raises important questions about the process by which children come to understand themselves as male or female: What attributes do children associate with each sex? Are these attributes interdependent? F or example, do people directly mentally associate preference for the color pink with a care giving personality, and thus, would a child who considers pink to be associated with boys also associate care giving with bei or female influence their own gender identity and how they feel about themselves in attitu
Idiosyncratic Beliefs and Gender Self Socialization 3 future behaviors? These issues are of primary concern in the current investigation. I n an attempt to answer questions such as these, psychologists have developed a plethora of theoretical frameworks to explain how, when, and why sex differences emerge. Some theorists focus on the role of biological and anatomical differences on psychological gender differentiation (Fausto Sterling, 1997 ), some study the role of social influences ( Bussey & Bandura 1999), and others assess the role of mental processes of perception, judgment, memory, and reaso ning (Bem, 1981; Spence, 1993; Martin, 2000). The approach a researcher chooses to take guides his or her choice of research questions, methodologies, and interpretations of data (Leaper & Bigler, 2011). Today, no single theoretical framework can, or claims to be able to, describe every aspec t of early gender role development (Levy, Barth, & Zimmerman, 1998; Liben & Bigler, 2002). Rather, theories of gender development can often be viewed as individual pieces of a much larger puzzle, often used in tandem with one another to create an integrati ve view of the gender development process. Despite the fact that an abundance of psychological literature has been dedicated to investigating the processes through which children come to understand and enact gender (for reviews, see Liben & Bigler, 2002; Zosuls, Miller, Ruble, Martin, & Fabes, 2011), a recently developed theoretical framework which integrates previous conceptualizations of gender differentiation, the Gender Self Socialization Model (GSSM; Tobin, Menon, Menon, Spatta, Hodges, & Perry, 2010) has been under tested by psychological researchers (Patterson, 2012 ; Perry, 2011 ). The Gender Self Socialization Model is innovative because it differentiates between three psychological constructs:
Idiosyncratic Beliefs and Gender Self Socialization 4 stereotypes (attitudes about appropriate roles for men a nd women) self perceptions and gender identity (2010) argue that these constructs have been wrongly conflated and confused by pr evious researchers, and suggest that psychologists distinguish between these three constructs and investigat e the interactive properties between them. Furthermore, the authors assert that previous psychological literatu re has inadequately explored the idiosyncratic content of ppropriate traits for each sex, and argue that previous researchers have not adequately examined how these attitudes influence evaluations of the ir own attributes or how children feel about being male or female. conceptualizations of the acce ptable attributes for each sex which, the authors note was originally suggested by Edlebrock and Sugawara (1978) has been mirrored by many other scholars in the field (Corby, Hodges, & Perry, 2007; Martin, 2000; Signorella, of gender identity for adju stment may hinge crucially on the particular messages about the cognitive aspects of gender development stressed the importance of understanding the dynamic nature of these belief syst ems. Furthermore, Signorella et al. (1993) concluded their meta gender related cognitions by stressing the need for attitudes towa
Idiosyncratic Beliefs and Gender Self Socialization 5 th ese calls for the assessment of idiosyncratic attitudes about gender this topic remains relatively unstudied in the psychological literature, particularly in relation to the constructs suggested by Tobin et al. (2010). The present study investigated these aspects of gender development by (1) assessing pre usefulness of a measure of stereotypes that allows children to express their own attitudes about gender to a traditional stereotyping measure and (3) examining the hypothesis of alization Model, that stereotype beliefs self perceptions, and gender identity are independent constructs with interactive properties Ultimately, the goal of this investigation was to e mpower young girls who are rarely included in the production of scientific knowledge t hat purports to speak for them (Checkoway, 2011 ) to express thei r own beliefs, and in the process to learn how their views coincide with the interplay of the constructs defined by The hypotheses tested in the study at hand are predicated on theoretical framework, and therefo re it is critical to understand the complexity of the Gender Self Socialization Model (Tobin et al., 2010) and the three cognitive constructs that it unites. The following section offers a brief description of the conceptual model under investigation. Then e ach of the three constructs delineated by the GSSM (Tobin et al., 2010) are described in detail followed by a n examination of the evidence which supports the three core hypotheses of the G SSM (Tobin et al., 2010) This framework offers exciting insights into the gender development process, and may be useful in answering the questions implicit in the news coverage of JCrew spread (2011) and other s like it in the popular media.
Idiosyncratic Beliefs and Gender Self Socialization 6 The Gender Self Socialization Model: An Overvi ew The Gender Self Socialization Model (GSSM), proposed by Tobin et al. (2010), integrates several pre existing theories about the process of gender development into a unified theoretical framework. It delineates between three distinct constructs which Tob in et al. (2010) argue have been confused in previous research : attribute self perceptions (the aspects that individuals consider to be important to their self concept ) gender identity ( the degree of ), and gender stereotypes (prescriptive beliefs about the attributes of men and women). Tobin et al. (2010) a ssert that distin guishing between these three factors theoretically and empirically is necessary in order to answer key questions about the process of gender development. These elements are theorized to be independent yet interdependent, and although researchers can choose to focus on an individual dimension or test multiple dimensions of them stereotypes, self perceptions, and gender identity cannot be combined into a single variable. Importantly, Tobin et al. (2010) clarify that each of these constructs must be measured explicit, and unique views into account, rather than conforming to the practice of using trait inventories (e.g. Bem, 1981) to aggregate responses on items that have been classified a priori By distinguishing between these three factors, Tobin et al. (2010) assert that three interactive hypotheses can be made, such that each construct can be predicted by the interaction between the other two constructs. Thus, in add ition to investigating the mechanisms which cause individuals to adopt gender typical traits the question which
Idiosyncratic Beliefs and Gender Self Socialization 7 has been of great historical importance to researchers Tobin et al. (2010) hypothesize that the way that children come to form gender identities and stereotype beliefs can be understood through this framework Hypothe ses of the Gender Self Socialization Model Following the research of Greenwald (2002) and Kohlberg (1966), Tobin et al. (2010) assert that individuals are motivated to maintain cognitive consistency such that each of the three psychological constructs associated with gender development (stereotypes, self perceptions, and gender identity) is of the same strength and valence. The researchers provide a useful mathematical m odel for predicting the directional relationship between these interactions: Each of the three central constructs (cognitive associations) may be conceptualized as varying in strength, from very positive to very negative. For ease of explication, gender i dentity may be considered positive when children identify more strongly with their own gender than with the other gender (+), neutral when children identify equally with both genders (0), and negative when children identify more strongly with the other gen der ( -). Gender stereotype is then construed as positive when children perceive an attribute to be more common or suitable for their own gender than for the other (+), as neutral when children do not view the attribute as varying with gender (0), and as n egative when children view the attribute as cross gender typed ( -) Self perception of an attribute is positive when children incorporate the attribute into their self concept (+), neutral when children neither claim nor deny the attribute (0), and negati ve when chi ldren disavow the attribute ( -). Each construct is expected to be a multiplicative function of the other two; that is, the sign and strength of a given association for a e other two associations. Thus, a particular cognitive association (e.g. self perception of an attribute) should be positive if both of the other associations are either positive or negative but should be negative if the two other associations are di fferen t (positive or negative) (p. 606). This mathematical model is useful because it demonstrates the process by which individuals can be expected to be motivated to adapt one of the three cognitive constructs in order to conform to the other two and achieve c ognitive consistency, and offers operational definitions about the directional properties of each construct. Based on this
Idiosyncratic Beliefs and Gender Self Socialization 8 model, three hypotheses are made about the multiplicative influence of two of the constructs on the third. First, the stereotype emulation hypothesis feelings about their membership in a gender category (gender identity) and their feelings about how members of this category should be (gender stereotypes) will influence the way that they view themselves (s elf perceptions) For example, if a boy believes that men should be good at football, and he strongly identifies as a boy, then the stereotype emulation hypothesis predicts that whether or not he likes football should be an important factor in his view of himself. Tobin et al. (2010) note that this hypothesis is of particular importance to the model because the brunt of gender developmental research is concerned with the way that children come to adopt gender typed attributes. Second, the stereotype constr uction hypothesis about how men and women should behave (stereotypes) are shaped by their self knowledge (self perceptions) and the way that they view their membership within their gender category (gender identity) For exa mple, research has found that when children are shown unfamiliar objects and asked whether they believe that boys and girls would like each object, they are more likely to predict that members of their gender will prefer the objects that they themselves pr efer (Martin et al., 1995). Yet, t his relationship has been under examined in the literature; in fact, Tobin et al. (2010) assert that no previous literature has tested this hypothesis. Third, the identity construction hypothesis suggests that the degree to which individuals identify with their gender category (gender identity) is based on the degree of
Idiosyncratic Beliefs and Gender Self Socialization 9 similarity between their perceived attributes (self perceptions) and the traits that they believe that members of their gender category should possess (ste reotypes) For example, a girl who strongly believes that girls should like the color pink would have a stronger sense of gender identity if she likes the color pink than if she d oes not. The researchers suggest that an omnibus test of the identity construction hypothesis could be designed by perceptions and calculating the similarity between the two, which would sense of gender typicality (the study at hand included this analysis). These predictions are central to the GSSM (Tobin et al., 2010), and therefore a comprehensive test of the model requires that self perceptions, gend er identity, and gender stereotypes be assessed in relation to one another; yet, very few research studies have assessed all three variables (Tobin et al., 2010). Tobin et al. (2010) point out that in research that has assessed the relationship between the portrayal of gender typed attributes and beliefs about gender stereotypes gender identity is not evaluated; it is simply assumed that individuals are motivated to internalize gender roles because they feel pressure to conform to gender stereotypes (feeli ng pressure to conform to gender stereotypes is a correlate of gender identity which will be described with more detail in the following sections) In research that has assessed the relationship between individuals feel ings about their gender role s in r elation to their own characteristics, stereotype beliefs have not been measured; it is simply assumed that all individuals have similar beliefs about the shared characteristics of men and women. In research that has compared beliefs about gend er stereotypes to how they feel about themselves in relation to their gender category, self perceptions have not been evaluated; it is simply
Idiosyncratic Beliefs and Gender Self Socialization 10 assumed that individuals feel typical of their gender because their attributes are similar to the stereotypes they hold about their gender group; or, conversely, that individuals feel pressure to conform to the stereotypes they hold about their gender group because their characteristics are dissimilar to other members of their gender (Tobin et al, 2010; 614). Furthe rmore, Tobin et al. (2010) theorize that very little has been done to consider the and instead psychologists have historically relied on the use of trait prescriptive stereotypes The dilemma presented by this scarcity of research is complicated by the fact that each of the key constructs included in the model can be conceptualized in a multiplicity of ways, making the possibilities for comparing them endless. Therefore, an in depth review of each of the three constructs must predicate further discussion of the psychological literature that has tested the interactions between them. The following three sections explore what we know about self perceptions, gender identity, and gender stereotypes in relation t o the GSSM (Tobin et al. 2010), followed by a description of the research which has conducted full or partial testing of the three G SSM hypotheses Self Perceptions Prior to related traits were studied based on the degree to which they corresponded with societal stereotypes, with the purpose of investigating the manner by which children come to enact these stereotypes. Not long ago, psychologists believed that
Idiosyncratic Beliefs and Gender Self Socialization 11 the set of traits associated with males and females were unidimensional ; that is, that masculinity and femininity were two ends of the sam e spectrum. This model assumed that all of the traits associated with each respective gender were inter correlated (so, for example, that people who enjoy a feminine typed activity such as baking would be more likely to enjoy other feminine typed activitie s, such as childcare), and that masculine traits and feminine traits were negatively related to one another. This conceptualization of gender was revised by Sandra Bem (1981), a cognitive theorist who proposed that children develop gender schemata or cogni tive framework s of associations which they internalize from social messages about gender and use as mental tools to encode novel information. Bem (1981) hypothesized that gender schemata were bidimensi onal arguing that female and male typed attributes we re not oppositional to one another, but instead were two in dependent, unrelated factors. She asserted that masculinity was associated with instrumental traits (sometimes called agentic traits, these attributes demonstrate independence, such as logical reas oning skills and athletic ability), while femininity was associated with communal traits (sometimes called expressive traits, these are attributes which emphasize interpersonal relations, such as care giving and kindness) (Bem, 1981). Bem (1981) theorized that individuals internalize these cultural sets of stereotypes about gender to varying degrees and modify their own behavior to match the schema they have formed for their own gender, a process which is known as gender typing In a tradi tional measure of gender typing, participants are given a list of attributes, each of which the researcher has predetermined ( unbeknownst to the participant) to be associated with either masc ulinity, femininity, or neither. The participant s rates degree to which
Idiosyncratic Beliefs and Gender Self Socialization 12 they associate themselves with each item and the researcher derives an overall score to designate each participant as masculine, high scores on low scores on both scales ) Gender schema theory ha s been advanced and promoted by many subsequent theorists (Martin, 2000; Martin & Halverson, 1984; Spence & Hall gender typing by using trait inventories remains among the most prominent assessments of gende r typing within the psychological literature. has faced widespread criticism (Perry & Pauletti, 2011; Tobin et al, 2010). Research has largely refuted ty ped and female typed attributes inter correlate into separate orthogonal factors (Perry & Pauletti, 2011). Inst ead, many including Spence and Hall (1 996 ) and Signorella (1999) have asserted that gender schemata are multifactorial; that is, that each indivi dual develops a unique, heterogeneous set of associations with each gender. For example, one person might think that men should be effective leaders and be good at chess but never be emotionally express ive; another may think that it is alright for men to b e emotionally expressive, but that they must also be heterosexual. Today, many researchers agree that the set of characteristics associated with masculinity and femininity differs for each individual (Perry & Pauletti, 2011). Additionally, Tobin et al. (20 tautological. Bem (1981) believed that people are motivated to behave like other members of their gender because of social pressure to conform to societal gender roles. measurement of gender typing assumes that if a
Idiosyncratic Beliefs and Gender Self Socialization 13 participant is found to be characterized by stereotypic traits, he or she believes that trait is an important attribute for his or her sex and has internalized that characteristic because of a strong feeling model does not distinguish between how individuals view themselves (self perceptions), how they view their gender category (gender stereotypes), and how they feel about themselves in relat ion to their gender category (gender identity). This lack of different iation makes it impossible for typing to prove a causal relationship between self perceived traits, beliefs about gender stereotypes, and f eelings of pressure to conform to those stereotypes; to do so would require that each of these factors be measured separately (Tobin et al., 2010). Thus, Tobin et al. (2010) assert that psychologists must not assess gender typing ch, but instead must first identify the traits which characterize an individual and subsequently study how these traits interact with prescriptive stereotypes and gender identity. And, because these traits cannot be assumed to uniformly correspond with either one gender or the other in a predictable pattern, Tobin et al. (2010) argue that it is necessary degree to which these traits correspond with th eir unique understandings of gender so that we may infer which gender stereotypes they have internalized. The authors also assert that individuals will judge their own traits based on the salience of that trait, or the importance that it has to the individ s identity (Tobin et al., 2010). For example, if a girl loves to play football and practices the sport every day, yet believes that only boys should enjoy sports, her sense of gender identity would be deflated (this would be an s 2010 identity construction hypothesis). Yet if she merely liked
Idiosyncratic Beliefs and Gender Self Socialization 14 to watch football on occasion but did not consider it a core attribute of her identity, it is less likely that her sense of gender identity would be affected. Therefore, in order to accurat based on the degree to which the child perceives that the trait defines who he or she is; D omains of self perceptions The Gender Se lf Socialization Model (Tobin et al., 2010) specifies that self perceptions can exist across a variety of dimensions. This evaluations of their traits and attributes, but the authors explain that the model also incorpo rates self perceptions across numerous guides (e.g., T esting self perceptions is perhaps done by considering specific domains; h owever, Tobin et al. (2010) specify that tests of the GSSM can include omnibus, aggregated measures of self account ; this was done in the current study Clearly, there is a great deal of variance in the specif ic traits and attributes that each unique person might consider important to their predictable. Developmental antecedents of self perceptions. The depth and content of c Montemayor and Eisen (1977) asked 136
Idiosyncratic Beliefs and Gender Self Socialization 15 es. Responses were coded in a 30 category scoring system developed to capture the major varieties of self representations by two independent raters. The categories were designed to be exclusive and exhaustive, and included descriptions such as (but not lim ited to): age, name, religion, likes and dislikes, sense of moral worth, social status, possessions, sense of competence, occupational role, body image, artistic activities, membership in an No reliable differences were found between the sexes. Overall, adolescents were more likely to describe themselves in abstract ways, such as their interpersonal style, ideological beliefs, and sense of self determination. Ten year olds, on the other hand, were significantly more likely to use concrete self descriptions than their older counterparts, including self perceptions that were reflective of their citizenship (for example, living in a specific neighborhood) and were more likely to list possession s and resources. Conversely, younger children were much more likely to describe themselves in te rms of physical self/body image Nearly identical patterns were reported by Brinthaupt and Lipka (1985), Harter (198 8 1999; for review, see: Harter, 2003 ), and Gara and Rosenberg (1979), who found that self descriptions become increasingly abstract across the span of development. based on the order in which they appeared in However, unlike the former researchers, Brinthaupt and Lipka (1985) did report gender differences in the responses; fema les were more likely than males to describe kinship
Idiosyncratic Beliefs and Gender Self Socialization 16 role, group membership, interpersonal style, and age, whereas males were more likely than females to mention activities that they enjoyed. Linking self perceptions with a ctual b ehavior. The stereotype em ulation hypothesis of the GSSM (Tobin et al., 2010) provides a model for the motivations that particularly vexing to gender theorists. This prediction hinges on the ass umption that if an perceptions are not cognitively consistent with his or her gender identity and gender stereotype beliefs, he or she will be motivated to adopt new traits. But do self perceptions correspond with behaviors? In a longitud inal study, Davis Kean, Huesmann, Jager, and Collins (2009 perceptions of their own self efficacy were predictive of their future success or failure. Self efficacy related to academics was predictive of their grades in math cour s es, and in a separate sample, self efficacy for assess the influence of academic self efficacy on academic performance the researchers analyzed archival data from 28 pu blic schools in Michigan between the years of 1988 1997. Beginning in the 1988 school year and continuing for a three year period, children in grades 1, 2 and 4 responded to five questions related to their self concept of their ability to achieve in mathem school records) for each school year were recorded. In a follow up study in 1994, the same information was collected from the same participants for another three year period. The researchers f ound that the relationship between efficacy and academic achievement increased sharply between early (kindergarten grade 2) and late childhood (grade 3), at
Idiosyncratic Beliefs and Gender Self Socialization 17 which point self efficacy was a strong predictor of behavioral outcomes throughout adolescence. T hese results, which confirm that beliefs about the self correspond with hypothesis that the influence of stereotype attitudes and gender identity will alt er ch children to adopt stereotypic al attributes based on that change in self perception. Although the GSSM (Tobin et al., 2010) does not give a clear model for the transition of mere self knowledge to the adoption of trait s, this evidence suggests that self perceptions do indeed have the potential to predict future behaviors. Yet, Davis study does not address the important mitigating a construct which is ex plored further in the following section. Gender Identity conveyed a variety of meanings throughout the span of its academic study. The Gender Self Socialization Model (GSSM; Tobin et al., 2010) incorporates a broad span of Identity Theory, the most recent conceptual framework of gender identity, which was proposed by Egan and Perry in 2001 and has been used broadly throughout psychological literature in the short time since its inception. Gender Identity Theory Egan and Perry (2001) proposed a multi factorial model in order to assess the factors that compose gender identity, which separated the variable into four parts: gender contentedness ategory),
Idiosyncratic Beliefs and Gender Self Socialization 18 gender typicality felt pressure to conform to gender stereotypes (internalized social pressure to behave in accordance with gender stereotypes), and intergroup bias (gender group attitudes that of which contained between 6 10 self report items comprehen sible to children as young as 7 years old. Unlike the practice of inferring gender identity fr om gender typing (e.g. Bem, 1981) awareness of each gender identity factor. To avoid the possibility that this explicit style might lead children to respond in socially desirable w ays, each item pres ents two dichotomous hypotheses. These scale items frame either directional response as socially participants ch oose which response suits them and then indicate whether they are somewhat or very much like that group. T herefore, a total of four options for each item is presented, which can be reduced to a lik ert type scale in data analysis l test of the model, 235 children in grades 4 8 completed a preliminary 92 item self report questionnaire which measured global self worth (Harter, 198 8 ), self perceived peer social competence, gender compatibility (a combined variable of gender typicality and gender contentedness) felt pressure, intergroup bias, male typed activities, female typed activities, agentic traits (masculine typed), communal traits (feminine typed), and heterosexual identity. A 6 month follow up study also evaluated the partici acceptance from male peers, an d acceptance from female peers. Results demonstrated that gender compatibility (gender typicality combined with contentedness), felt pressure,
Idiosyncratic Beliefs and Gender Self Socialization 19 and intergroup bias were indepe nde nt and unrelated to one another supporting a multidimension al approach to gender identity and found that associations between the domain specific measures of gender typing were moderate to non significant. As children grew older, the y demonstrated more gender typical attributes and less compatibility and intergroup bias, counter to the hypothesis that children would feel more content and typical of their gender as they adopted attributes that were gender typical Felt gender contentedness was positively related to high self esteem and acceptance by peers, while felt pressure to conform to stereotypes and intergroup bias were indicative of negative psychosocial affect. The GSSM (Tobin et al., 2010) predicts that three of the factors proposed in Egan an felt typicality, contentedness, and felt press ure to conform to stereotypes described, these predictions are rooted in the belief that individuals desire cognitive consistency between each of the three constructs. Inconsistency is expected to result in distress and negative mental health outcomes (Tobin et al., 2010). Other gender theorists have also made the claim that gender identity and the negative mental health consequences associated with it are responsible for motivating children to adopt sex typed behaviors. Bem (1981) asserted that this motivation was derived from the stress of felt pressure to conform to stereotypes, while Martin (2000 ) hypothesized that a low feeling of typicality was responsible. Therefore, it is important to consider the ways that each of these gender identity variables are associated with adjustment and mental wellbeing in order to demonstrate the causal predictions of the GSSM (Tobin et al., 2010). In the years since Gender Identity Theory (Egan and Perry, 2001) was proposed,
Idiosyncratic Beliefs and Gender Self Socialization 20 its psychometric properties have been studied extensively, primarily by Egan and Perry (2001) and their colleagues (Caver, Egan, and Perry, 2004; Carver, Yunger, & Perry, 2003; Corby, Hodges, and Perry, 2007; Yunger, Carver, & Perry, 2004). Gender identity and adjustment. Carver, Yunger, and Perry (2003) were the first examined the relationship between each gender identity construct with various the importance of these vari wellbeing which support predictions made by previous gender resea rchers (Liben and Bigler, 2002), that children would adopt sex stereotypic traits as a result of the internalized pressure that they feel to adhere to those roles. Two hundred and six children in grades 3 through 8 ( M = 11.5 years), who were pr contentedness, felt pressure, and intergroup bias. Additionally, participants completed a peer nomination inventory, which assessed aspects of social behavior, including internalizing problems (self harming emotions such as depression), victimization (being the target of bullying), externalizing problems (such as aggressive behaviors), agentic traits (male typed), communal traits (female typed), and filler items. Responden ts checked off the names of same sex classmates who fit the description of each item, yielding a score for each child based on the percentage of nominations they received for each category from their classmates. The participants also completed additional m eas ures of global self worth, self esteem, and self perceived peer social competence. In order to test the hypothesis that children who feel more typical of their gender actually exhibit the qualities associated with the traits that society associates wit h
Idiosyncratic Beliefs and Gender Self Socialization 21 prediction, agentic traits in boys (but not girls) and communal traits in girls (b ut not boys) were associated with a sense of typicality. Descriptive analyses revealed that boys scored higher than girls on measures of gender typicality, gender contentedness, and felt pressure, but scored lower than girls on intergroup bias. Boys also s cored higher than girls on internalizing and externalizing problems and agentic traits, but lower than girls on communal traits. Older children reported higher typicality and contentedness but less felt pressure and intergroup bias than their younger count erparts. Multiple regression analyses evaluated the relation of each gender identity component with each adjustment variable (24 analyses in total), finding that intergroup bias was not significantly associated with adjustment. Gender typicality, on the o ther hand, was inversely correlated with wellbeing ; children who reported feeling less typical of their gender were rated by their peers as having greater internalizing problems, a relationship which increased in strength when sex and age were accounted fo r. Furthermore, higher gender contentedness was related to positive self esteem, and additional regression analyses revealed that contentedness and felt pressure had an interactive relationship that is, gender dysphoric children who experience strong press ure to adhere to gender roles (regardless of their sex) are harmed the most: they experience the most internalizing problems, the lowest global self worth, and the lowest peer perceived social competence. Girls who were not content with their gender were c lassified by their peers as having more externalizing problems, which was not the case for boys Female participants (but not males ) who reported higher pressure to conform to
Idiosyncratic Beliefs and Gender Self Socialization 22 gender stereotypes also reported lower self worth and lower self perceived peer social competence. Importantly, the interaction of felt pressure and sex significantly predicted the degree to which peers rated participants as adopting the characteristics acceptable for their sex (i.e. agentic or communal characteristics). This researc h on psychopathological correlates of gender identity reotype emulation hypothesis on an underlying desire for cognitive consistency by de monstrating the negative mental health correlates which can arise when participants have high felt pressure to conform to gender stereotypes The accuracy of these findings was tested in the following year by Yunger, Carver, and Perry (2004). Participants were 280 3rd 7th grade students ( M = 11.1 years once and repeated the process at the same time du ring the following school year, and were tested using the same design as C Overall, low gender typicality, low gender contentedness, and high felt pressure to conform were related to a deterioration of psychosocial adjustment. Consistent with the findings of Carver et al. (2003), high felt pressure pr eceded the development of internalizing problems and less peer acceptance; yet, it did not predict lower self esteem or externalizing behaviors. Low gender contentedness foreshadowed lower self esteem and was the only identity construct related to decrease d acceptance by peers. Low gender typicality in year 1 led participants to have reduced self esteem in year 2, but typicality did not directly affect any of the othe r adjustment variables. However, it did act in conjunction with high felt pressure to incre internalizing behaviors. The authors
Idiosyncratic Beliefs and Gender Self Socialization 23 reflected on this f ) model of self discrepancies, which delineates between ideal self discrepancies, in which the individual fails to meet the ir own intrinsic standard and is thought to respond with dejection related emotions (such as low self esteem), versus ought self discrepancies, which reflect failure at meeting a social standard and are thought to lead to agitated affect (such as shame), w hich might be more visible to others than dejection. Yunger et al. (2004) suggested that feeling both gender a typicality and high felt pressure to conform are ought self discrepancies ; participants affectively responded with agitated emotions that were vis ible to their peers. Generalizability of gender identity research It is important to note that the majority of the research conducted in this field has tested European American samples. The generally consistent patterns found in these samples have led some researchers, including Yu and Xie (2010) to wonder whether these aspects of gender identity are universal to people across all cultures or if they are instead influenced by societal messages about the appropriate roles for men and women. The researchers examined among 9 12 ye ar old children in Mainland China. Four measurements of psychological wellbeing were administered, including measures of self worth, scholastic competence, social competence, athletic competence, physical appearance, and behavioral conduct. Additionally, t he researchers measured the relationship between gender identity and the four generalizabi however, Chinese boys expressed more gender contentedness than did their female peers,
Idiosyncratic Beliefs and Gender Self Socialization 24 there were no sex differences in intergroup bias, and gender typicality was by far the stronge st indicator of psychological adjustment. The authors concluded that cultural appropriate behavior and therefore wellbeing However Corby, Hodges, and Perry (2007) recently suggested that this model of gender identity cannot be effectively used to predict adjustment unless the researcher researchers invest self regulatory mechanisms that hu mans have developed as a result of evolution or on the other hand, if these factors and their relationship to wellbeing are socially constructed, and th us vary according to the unique experiences of each individual. The Hispanic 5th grade students ( M = 11.1 years old), and compared the relationship between scores on meas ures of gender identity (comprised of felt gender typicality, felt gender contentedness, and felt pressure to conform to stereotypes) and social adjustment. Adjustment was tested through a 40 item peer nomination inventory, which was comprised of 5 subscal es of behavioral descriptions: agentic traits (such as bravery and assertiveness), communal traits (such as friendliness and empathy), internalizing problems (such as social withdrawal and anxiety), victimization (such as being a target of bullying), and e xternalizing problems (such as aggression and dishonesty). Children were asked to indicate which of their same sex classmates fit the description provided by sex ngs of him or her.
Idiosyncratic Beliefs and Gender Self Socialization 25 The results for White children were consistent with the previous literature; feelings of typicality, contentedness, and freedom from pressure to conform were associated with positive social adjustment. The results of Black and Hispanic children, however, varied from this trend significantly Both groups reported higher pressure to conform than did White children. For Black children, there was almost no association between gender identity (typicality or felt pressure) and adjustment, thou gh contentedness was highly related to self esteem. Unlike findings in White samples, ernalizing problems, and in Hispanic boys felt pressure was negatively related to this variable; both genders otherw ise demonstrated results that were predicted by the Gender Identity model. The authors struggled to make sense of these results and their implications, warning researchers to be cautious about attributing the responses to variability between the cultures, as each ethnic group tested was actually quite heterogeneous and the authors did not assess factors related to culture. Instead, they suggested that the results might be explained by the process of identifying as a racial minority, which could preclude the children to be more sensitive to in group norms. The authors resigned from further interpretation of these controversial results, concluding that the model of Gender Identity established by Egan and Perry (2001) does not apply well to Black and Hispanic children. They also called to future researchers to examine the individual messages that children receive about gender, which has had great influence on the present study: model, gender identity is cast as a direct influence on adjustment. However, the implications of gender identity for adjustment may hinge crucially on the particular messages about gender that children experience and internalize (depending on many variable s, some of which are culturally correlated, such as family practices and relationships, socioeconomic status, religion, neighborhood,
Idiosyncratic Beliefs and Gender Self Socialization 26 peer groups, and media exposure). The extent to which gender identity contributes to healthy adjustment may depend on the degree to which children internalize gender norms that are healthy and adaptive. For children with socially maladaptive or inherently self limiting gender beliefs (e.g., a boy who defines maleness in terms of violence, dangerous risk taking, and defiance o f adults, or a girl who defines femaleness in terms of subservience to men, helplessness, and self deprecation), stronger gender typicality, gender contentedness, and felt pressure may all promote unhealthy adaptation (including externalizing problems, at that Egan additional cognitive construct are important for persons of each sex to possess In this revised mod el, interacting components of a causal cognitive system, with the effects of gender identity on adjustment moderated by the specific content (and flexibility ) of conceptions of gender roles along with assessments of their gender identity in future research would allow determining not only the degree to which gender conceptions differ across ethnic/racia l subgroups but also the degree to which such conceptions actually account for between and within ethnic/racial group variability in the impact of gender identity on adjustm ent. (Corby et al., 2007, p. 265, emphasis added). Therefore, while researchers sho uld be wary of attributing these results to cultural variables on the hypoth eses made in the GSSM (Tobin et al., 2010). Acknowledging this need is part of the overall movement which has been described throughout this review for will be explo red further throughout the remainder of this review. One way in which it is possible to conduct such an analysis is to explicitly ask children about their stereotypes, and to compare this information to gender identity outcomes (Tobin et al., 2010); indeed this is the method utilized in the current research, and is explored further in the following section s
Idiosyncratic Beliefs and Gender Self Socialization 27 Gender Stereotypes suggestion that the content of gender stereotypes might influence gender identity as well as the mental health outcomes associated with gender identity variables emphasizes the importance of examining the variation and flexibility attitudes about stereotypes more closely. These findings, and the identity construction hypo thesis of the GSSM (Tobin et al., 2010) imply that examining the eelings of gender typicality and contentedness and to decrease their felt pressure to conform to these roles, subsequently improving their mental health outcomes. Developmental antecedents of stereotyping Most children demonstrate knowledge of difference s between the sexes across a range of dimensions by age 3, A variety of models have been proposed to explore how and why children come to adopt prejudicial sex role att itudes. Recen tly, Arthur, Bigler, Liben, Gelman, and Ruble (2008 ) proposed a cognitive approach called Developmental Intergroup Theory (DIT), which suggests that stereotyping is a byproduct of normal childhood development. Arth ur et al. (2008 ) asserted tha t children grow to understand and learn about the world through the use of cognitive categories, which help to organize novel information in the memory category in the chi Based on this social information, children come to group certain traits with social groups
Idiosyncratic Beliefs and Gender Self Socialization 28 DIT might be used to explain variability in the influence of gender identity on mental health outcomes found by Corby et al. (2007): a variety of environmental and experiential differences early in life lead children to form different concepts of masculinity and femin inity. Stereotype knowledge vs. attitudes: A critical distinction Though young children tend to view sex stereotypes as moral imperatives, as children develop cognitive constancy, their opinions about stereotypes become increasingly flexible (Signorella, Bigler, & Liben, 1993). However, it is of crucial importance to distinguish between development of the knowledge that cultural gender stereotypes exist and the development ana lysis of research related to gender stereotyping in childhood, the authors investigated the development of stereotyping by age and sex, as well as the influence of various methodological approaches of assessing stereotype beliefs on the outcomes of partici true beliefs about how men and women should be, rather than assessing only their understanding of the existence of societal stereotypes about gender. Of primary interest was the effect of the wording of explicit stereotype measures and whether or not the answer was forced ach stereotype item, versus non forced choice measures in whi
Idiosyncratic Beliefs and Gender Self Socialization 29 te levision preferences, maternal employment, stereotyping of toy preferences, and memory for gender related material. The results of the meta analysis found that forced choice measures tended to elicit stereotype knowledge, whereas non forced choice measure s tended to elicit attitudes age, regardless of the type of question that was asked, and matching stereotyped traits to the correct gender was associated with a higher IQ score and with more time spent watching non educational television. Girls were also found to have greater knowledge of societal stereotypes on forced choice measures than bo ys were. On non forced choice measures, the type of question that was asked was found to significantly influence stereotype kn responses followed the same patterns found in forced choice measures; correct mat ches increased with age. On the although girls were found to have significantly more e galitarian responses than boys. Furthermore, egalitarian responses on non forced choice measures were unrelated to IQ scores, but were associated with decreased gender related memory biases, maternal employment (egalitarian attitudes were related to having a mother who worked outside the home), and non traditional television preferences. These findings suggest that, aside from the correlates of attitudes about gender roles with age and sex, culture plays a ped or non stereotyped views about men and women.
Idiosyncratic Beliefs and Gender Self Socialization 30 The content of gender stereotypes. Unfortunately, very few studies have measured the content of trait inventories derived by the researchers who conduct the study in which the liste d attributes are classified a priori ased on pilot testing. Spence and Hall (1996) commented in their review of research on stereotyping that according to their estimation, this method has been utilized hundreds of t imes Lueptow, Garovich found empirical data to support this assertion in their meta analysis of studies on gender stereotyping, which included 30 papers from 1974 1997 and found that only list inventories were used to measure s tereotyping, never free res ponse measures. Of these measurements, only two differentiated between self stereotypes compared to stereotypes about others, and none distinguished between knowledge of versus attitudes towards stereotypes (or at least, this dif ference was not recognized by Lueptow et al., 2001). Lueptow et al. (2001) did not focus their review on the developmental differences of stereotype content and only briefly discussed the methodologies of each study reviewed thus it is difficult to draw c onclusions from their review about the methodological trends in the field or to assess the developmental patterns in stereotype content. The flexibility of gender s tereotypes. Few researchers have studied historical patterns of change in the content of ge nder stereotypes. Slightly more have investigated 1980s
Idiosyncratic Beliefs and Gender Self Socialization 31 has arguably been to dismantle rigid and marginalizing cultural gender stereotypes, with the goal of reducing sexism among future generations ( Lueptow et al., 2001). Lueptow et analysis of 30 studies of gender stereotyping in 1974 1997 found that stereotyping did not decrease and that, in fact, feminine gender typing in female participants increased. The authors conclude that attitudes about males and females are developed as a result of evolution, and c onjecture that males and females are theorized to be more capable of transmitting their genes to future generations when males are aggressive and dominant and females care for offspring. However, as previously described, this meta analysis was significantl y limited by the use of trait inventories which did not differentiate between the measurements of knowledge versus attitudes about stereotypes, restrained the quantity and types of stereotypes that participants could rate, and did not assess differences be tween stereotype domains. Katz and Ks ansnak (1994) explored the correlates of gender stereotypes across two domains: stereotypes for the self and stereotypes for others. Four hundred and seventy nine middle and high school students reported information re lated to their cognition, physiological development, and socialization Cognitive factors included age, self reported academic ability, and gender flexibility the latter was measured by four open were coded by independent raters into 6 chronological stages of gender role concept development. Biologically based factors included gender and pubescent stage and socialization factors included perceptions of parents, perceptions of pee rs, media preferences, perceptions of siblings and perceptions of teachers. Dependent variables were future vocational preferences, future domestic preferences, leisure time activity
Idiosyncratic Beliefs and Gender Self Socialization 32 preferences, academic subject preferences, and attitudes towards others. Overall, gender schema flexibility for the self and for others was predicted most by age and socialization factors, including higher ra tings on sex parent; perceived parental atraditionality; perceived parental respo gender atraditional activities, jobs, and domestic chores, degree of perceived atraditionality of a same sex friend, degree of identification with an opposite sex teacher, number of favorite cross sex media characters and degre e of identification with them; and perceived influence of opposite p. 277). While it is likely that biological, cognitive, and socio cultural factors all play a role related attit udes, these results suggest that despite the consistency of stereotype content over time found in Lueptow analysis, theorists of gender development should not forgo social theories of gender development in favor of an evolutionary mode l just yet Additional research has found roles are in fact quite malleable. Master, Markman, and Dweck (2012) investigated the possibility that pre school aged children could learn to avoid categorization a cognitive precursor to stereotyping (see Arthur et al., 2002) The experimenters recruited 32 4 year olds (16 male) to participate. All of the children were shown the same three sets of six horizontally aligned pictures, which depicted the figure gradually changing; in a practice condition, the pictures shown were of a tadpole growing into a frog, while in the test condition the children viewed a frowning face with eyebrows that expressed anger(a basic line drawing) changing into a smiling face. The researchers showe d the chi ldren the images of the tadpole and discussed the set with the participants, a conversation that
Idiosyncratic Beliefs and Gender Self Socialization 33 varied listened to the experimenters read a short paragraph, which e xplained that some people are nice while others are mean and that people are either one or the other. The continuum condition heard each image in the set be described as slightly different indicating the gradual growth of the frog, and then heard a paragra ph explaining the variability between people wh o are nice and mean. The depende nt measure was cted a dichotomous or continuous viewpoint. Results showed that children who heard images being described in more nuanced terms were more susceptible to doing so themselves, while those who were primed to think of the images as binary did not elaborate on the changing characteristics of the image. These results have important implications for the way that stereotypes are understood and confronted; they suggest that educating children to think outside of dichotomous categories and to see variation within gro ups may reduce the likelihood that they will stereotype others or think about social groups in na rrow and generalized terms. These findings also represent a fusion of social and cognitive models: by demonstrating that children can learn to avoid categoriza tion, a cognitive process associated with gender stereotyping (Liben & Bigler, 2002), th ese data demonstrates that cognitive processes that precede stereotyping are not psychologically intrinsic and unchangeable but instead can be unlearned. Summary. A plethora of research on stereotyping has been conducted the most recent of which has concluded that, counter to previous models, stereotyping occurs
Idiosyncratic Beliefs and Gender Self Socialization 34 across a variety of dimensions including knowledge of stereotypes, attitudes about stereotypes, and stere otypes for the self versus for others and across a variety of domains, including those which have been largely ignored by previous research, such as appearance. Recent studies also suggest that these beliefs can be t aught and by extension, untaught. Howev er, there is a severe lack of research relating to the variability of the variety of domains into account and allows participants to report the stereotype attributes that are important to them. The current study attempted to fill this gap. However, as previous research has shown (Liben & Bigler, 2002; Spence, 1999; Tobin et al, 2010) attitudes about stereotypes alone do not determine gender development; the attributes that individuals perceive themselves as having and the feelings that they have about themselves in relation to their gender group play an interactive role in the process of developing gender related attitudes and behaviors. Tests of the Gender Self Social In the Gender Self Socialization Model (Tobin et al., 2010), it is predicted that the degree to which children will emulate the attributes that they believe are characteristic of their same gender collective is motivated by the extent to which they identify with their gender group (stereotype emulation hypothesis), that the attributes that children associate with their gender collective will be determined by their own perceived characteristics and the extent to which they identify with their gender group (stereotype construction hypothesis) and that children identify with members of their gender more strongly if they perceive themselves as having the traits that they believe are characteristic of their gender (identity con struction hypothesis). Two criteria must be met in order for a study to be
Idiosyncratic Beliefs and Gender Self Socialization 35 considered a full test of these hypotheses: it must include and analyze the relationship between all three gender constructs (gender stereotypes, self perceptions, and gender identi ty) and it must use explicit measurements. However, the number of research studies which have conducted such analyses is limited (Tobin et al., 2010), and therefore it is sometimes necessary to study the relationships between only two of the variables. Lit tle research has related gender identity to self perceptions and stereotypes, but the previous sections of this writing have explained that incongruent findings on gender identity outcomes between different social groups have suggested that stereotype atti tudes do influence identity (Yunger, Carver, & Perry, 2007; Yu & Xie, 2009). Additionally, two other studies (Leaper & Van, 2008; Patterson, 2012) have found that one specific dimension of gender identity gender typicality plays the interactive role betwee n the other two constructs that the GSSM (Tobin et al., 2010) expects. However there is considerably more research focusing on the relationship between self perceptions and stereotypes; those findings while not a full test of the GSSM (Tobin et al., 2010) provide considerable support for both the stereotype emulation hypothesis and stereotype construction hypothesis. The R elationship between self perceptions and s tereotypes Tobin et al. (2010) partially base d their integrative model on the assertion that self perceptions and stereotypes have historically been conflated under models of gender typing (e.g. Bem, 1981), which assume that self reports of gender typed traits are indicative of belief in gender stereo types. However, more recently, researchers have separated the two variables and studied the relationship between them, finding a mutual relationship between the two constructs that gives partial support to the stereotype emulation and construction
Idiosyncratic Beliefs and Gender Self Socialization 36 hypothes es. The influential role of stereotype attitudes on self perceptions was originally investigated by Martin and Halverson (1981), but was revised by Liben and Bigler (2002) to highlight the differences in children and to specify the relevance of the domain (delineated by the researchers into traits, activities, and occupations) of One of the most prominent tests of the relationship between stereotypes and self perceptions has been conducted by Liben and Bigler (2002), who propose d a dual pathway model between the two constructs in which children not only incorporate gender stereotypes into their self concepts, but also project their self perceptions onto their beliefs about gender stereotypes. The researchers de monstrated the validity of their hypotheses with longitudinal data, which reflects similar findings elsewhere in the literature. To test the first half of their dual pathway model, the attitudinal pathway, the researchers administered surveys to 78 child ren (divided evenl y by sex) four times over a two th and 7 th grade s The scales administered to test gender stereotypes and self perceptions Occupations, Activities, and Traits Attitude Measure and Personality Measure (COAT AM and PM; Liben & Bigler, 2002) were developed specifically for this study, and have since been used frequently in research. Each of the scales in the COAT AM and PM incorporate a list of 25 traits, activities or occupations and either ask participants to rate the extent to which they are described by each item (for the personality measure, which measures self perceptions ) or
Idiosyncratic Beliefs and Gender Self Socialization 37 ask items each gender ( boys/girls/neither/b oth) should do (for the attitudinal measure, which evaluates stereotype beliefs ). For the personality measure, items are categorized as masculine or used to indicate their gende r orientation. However, for the attitudina l responses are simply classified in terms of the proportion of items that they indicate are gender, and these ratings are used to evaluate the ext stereotype attitudes are egalitarian or rigid. It is important to note that the COAT AM relies on prescriptive gender stereotypes appro priate roles for men and women rather than their mere understanding of gender each scale. On the personality measure, which gauges self perceptions the researchers found that children generally reported greater associations with traits than occupations or activities. This main effect was moderated by the domain of the measure and the gender of each item, such that children expressed stronger identif ication with masculine activities than feminine activities, but the reverse effect was true for occupations, in which children gave higher ratings to feminine than to masculine items. Overall, children expressed greater identification with masculine person ality items than with feminine feminine items dropped significantly from sixth to seventh grade. Children generally reported an increase in gender egalitarian attitudes over time, but their self endorsement of sex typed traits, activities, and behaviors remained consistent across time. The primary
Idiosyncratic Beliefs and Gender Self Socialization 38 egalitarian attitudes. Researchers foun d a strong relationship between the personality measures (self perceptions) and attitude measures (stereotypes), such that for girls (but not for boys), holding egalitarian attitudes towards masculine activities in the spring of seventh grade was associate d with higher ratings of their own interest in masculine occupations and activities. The researchers concluded that children who endorsed a rigid stereotype about a particular item on the attitude measure were more likely to associate that item with their own personal interests. Liben and Bigler (2002) also tested the second half of their dual pathway theory, the personal pathway, which predic ted that self perceptions influence their attit udes about sex role stereotypes. In the same longitudina l study that tested the attitudinal pathway model, t he researchers found that at time 1 (the beginning of sixth endorsement of feminine traits was unassociated with their stereotype flexibility. In contrast, self endors ement of feminine traits was associated with their increased advocacy of gender egalitarian ster eotypes at time 4 (in the spring of 7 th grade), leading the researchers to infer that gender atypical self Additionally, c hildren who expressed interest in pursuing cross sex careers were far more likely to e xpress gender egalitarian views about the appropriate careers for men and airplane pilot, and out of these respondents, none expressed a personal interest in becoming a pilot; conversely, of those who indicated that boy and girls should aspire to be pilots, 45 percent expressed that they would personally aspire to that career. Overall,
Idiosyncratic Beliefs and Gender Self Socialization 39 17 significant associations were reported between self perception and stereotype resp onses on cross sex typed items for girls. The researchers speculated that children probably would not be motivated to pursue a traditionally gender atypical career simply because they thought that it would be acceptable for their gender, but that they woul d be more likely to adopt an egalitarian attitude about that occupation if they themselves were interested in pursuing it. Based on these results, the authors conclude that self perception s influence stereotype beliefs, yet qualify that these effects are c ontextually limited on the basis of participant sex and domain of the items. Martin, Eisenbud, and Rose (1995) reported more robust findings to support the studies assessed childre were not explicitly labeled by typed toys (the masculine typed toy was a transformer, and the feminine typed toy w as a dollhouse) and eight additional objects and toys which were unusual products originally designed for adults (such as a prism, a magnetic ball sculpture, and colored interlocking ge ars). Preschool aged children ( N = 22; 11 female; M age = 59 months) vi ewed each seen the object before, and rated how much they liked it and how much they thought oth er boys or girls would like it on a four point scale. Results showed t predictions about what other children would like was contingent on the sex of the target child and their own preference for the toy, such that they estimated that children of their own sex would express similar attitudes to their own, and ch ildren of the other sex would
Idiosyncratic Beliefs and Gender Self Socialization 40 design by using Seventy one 4 and 5 year old children (38 males; M age = 58 months) viewed 12 objects which had been selected based on their unfamiliarity to the children and attractiveness as determined by two adult judges using positively correlated with their predictions of the attitudes of same sex peers and negatively correlated with their predictions for the preferences of other sex peers. Because children had no a priori knowledge of these items and their classifications of sponded with their own gender and their construction hypothesis. More recently, Martin and Dinella (2011) assessed the influence of self perceptions and stereoty pe beliefs in girls aged 7 perceptions of their activity preferences would be congruent with their gender stereotypes about girls. Additionally, unlike the previously described studies, the researchers considered a v ariable which was related to gender identity: whether or not the female participants considered themselves to be tomboys. While this factor is not included in the gender identity variable, it does move one step closer towards meeting the criteria for being a full test of that model. Participants were 112 girls at an after school program at public elem entary school, and the majority were European American. The researchers assessed self identification as a tomboy, stereotype variability (an 8 item questionnaire whether the participants believed that there were also some girls or boys who did not
Idiosyncratic Beliefs and Gender Self Socialization 41 enjoy that activity; higher scores indicated greater flexibility), activity preference for self (participants rated on their enjoyment of a variety of masculine and feminine type d items on a likert type scale), and personal stereotypes about g (participants rated the same list of activities, but indicated the degree to which they thought boys or girls would enjoy them). Tomboys and non tomboys exhibited the same degree of gender egalitarianism, but regardless of their status as a tomboy, girls who had greater interest in masculine activities were found to have more inclusive gender stereotypes for girls but not for boys. On the other hand, girls who preferred female typed activities did not differ in their gen der egalitarian beliefs for boys versus girls. Though their study did not prove a causal relationship between self perceptions and beliefs, Martin and Dinella (2011) conjectured that this pattern of correlations between the two constructs suggested that pa rticipants who perceived themselves to enjoy activities male typed activities subsequently developed increasingly flexible stereotype beliefs. These results, when considered together, indicate that individuals project their own self evaluations onto membe rs of their same gender group in a manner consistent (2002) dual pathway model construction hypothesis -which is tested in the current study also makes this claim, but adds further predictions about the interactive role of a third variable: gender identity. Liben and Bigler (2002) and Martin (2000) both theoretically account for the role of gender identity in this process, but differ in the gender identity construct that they emphasize Liben and Big ler (2002) suggest ed that the relationship between self perceptions and stereotypes is catalyzed by feelings of pressure to conform to gender stereotypes, whereas Martin (2000) hypothesized that the relationship between
Idiosyncratic Beliefs and Gender Self Socialization 42 stereotypes and self perceptions wou ld be mediated by the extent to which children feel gender a typ ical, yet none of these theorists have formally me asured explicit gender identity Perry and Pauletti (2011) described five reasons why neglecting to consider the independent role of gender ide ntity is problematic: (1) the tautological nature of the presumption that gender identity causes gender typing when both factors are measured using the same scale, (2) the degree to which individuals are gender typical varies across domains (for example, a man may be gender typical in that he is physically strong yet gender atypical in that he enjoys cooking), ( 3) different youth possess different gender stereotypes and thus have individualized bases of comparison for assessing themselves in relation to the ir gender (for e xample, many youth may not care about instrumental and expressive traits, but place great importance on physical app earance or sexual orientation), (4) this practice does not consider the different aspects of gender identity or the strength that each play s, and (5) self perceptions of instrumental and expressive traits do not predict other gender phenomena that should be predictable from gender identity The GSSM (Tobin et al., 2010) seeks to unite each of these viewpoints by viewing gender as multidimensional (e.g. Egan & Perry, 2001) Fortunately, a small but notable number of studies have compared self perceptions and stereotype beliefs to gender identity, as is Complete tests of the Gender Self Socialization Model. Leaper and Van (2008) demonstrated that gender identity does have a moderating role on the relationship towards gender roles, perc eived gender typicality, and self efficacy and interests (self perceptions) in academic fields that were traditionally male typed (e.g. engineering),
Idiosyncratic Beliefs and Gender Self Socialization 43 non traditional (e.g. literature), or neutral (e.g. hi story). Three hundred and forty two male undergrad uate students completed self report assessments of endorsement of masculine ideology (divided into five categories: achievement and status, avoidance of femininity, rejection of homosexuality, restrictive emotionality, and non relational attitudes towards sex), sexist attitudes (11 items addressing sexism in contemporary cale. The participants also reported their academic majors, nominated one of 11 academic areas that interested them the most, and rated their self efficacy in the following areas: computer science, physics, math, psychology, literature, writing, performin fields were significantly and positively correlated with covert sexism, gender typicality, and masculine ideology. Follow up analyses showed an interactive relationship between gender typicality and masculine ideology on this measure, such that men who perceived themselves as typical members of their gender and had rigid beliefs about the male sex role were more likely to express interests in traditional disciplines. The same pa ttern was Overall, the authors found significant evidence of the Gender Self Socialization sculine ideology, covert sexism, and gender fields, and negatively correlated with their self efficacy in non traditional fields. However, subsequent analyses whic h took all three predictors into account simultaneously found that none had significant main effects on any academic variable.
Idiosyncratic Beliefs and Gender Self Socialization 44 Instead, the researchers found that beliefs about gender roles and gender identity were interactive, suggesting that men may beco me more interested in traditionally masculine self efficacy in non traditional fields was highest for those who reported low gender typicality and low endorsement of sex ist beliefs. The authors reflect that perhaps men adopt gender egalitarian views in order to justify their own gender non conformity. It is also possible that egalitarian gender attitudes lead men to select non traditional fields, and subsequently recogniz e that they are not like other men. Although it is difficult to make assumptions about the causal relationship of these variables, these findings support In a similar study, Leaper, Farkas, and Br own (2011) determined that girl s motivation in STEM curricula was positively correlated with greater gender egalitarian beliefs and exposure to feminism, while their motivation in English was associated with their felt pressu re from parents and negatively related to felt pressure from peers. Five hundred and seventy nine girls aged 8 13 years from a wide variety of ethnic achieved level of educa tion, academic grades, perceived academic support, gender parents, gender typicality, and gender role contentedness), gender egalitarian beliefs (on a 10 tem scale of at titudes towards women), and exposure to feminism (assessed using a 10 item scale which asked participants about the degree to which they had learned about typicality and conte ntedness would be negatively related to motivation in STEM courses
Idiosyncratic Beliefs and Gender Self Socialization 45 and positively related to success in English courses, neither gender identity variable was related to motivation in either academic field The researchers suggested could have been explained by the fact that although STEM courses are traditionally male typed, the participants may not have viewed being successful in math and science as being incompatible with a typical girl, indicating an increasing degree of acceptance for wome n in STEM fields. The researchers also found evidence for the GSSM (Tobin et al., 2010) stereotype emulation hypothesis: felt pressure to conform to gender stereotypes moderated the re from their parents to conform to stereotypes did less well in STEM fields and performed better in English (which is traditionally female typed ), and greater gender egalitarian attitudes significantly increased motivation in math and science, though thes e findings were not interactive with the gender identity variables under analysis More recently, Patterson (2012) studied the relationship between gender stereotype endorsement, gender typed attributes, and self perceived gender typicality in children, e xploring the connections between the three variables in a participant group that is younger than is typically studied. One hundred elementary aged children (6 12 years old) were interviewed individually by a female experimenter. Participants completed a me asure of gender typicality (10 items) which was developed for the study and asked the degree to which they felt they were similar to their perceptions of other children of the same gender in various domains. Gender typed interests and attributes were asses sed with Attitude Measure (COAT AM; Liben and Bigler, 2002), and their stereotype
Idiosyncratic Beliefs and Gender Self Socialization 46 endorsement was assessed with the activity and occupation scales of the COAT AM ( Lib en and Bigler, 2002). Gender typicality was found to be related to gender stereotype endorsement such that lower feelings of gender typicality were associated with more egalitarian gender attitudes. Gender a typical children were also more interested in o ther typed activities than were children who perceived th emselves as more gender typical, and t hese relationships did not differ by age or sex. However the researcher noted that although results found a main effect of gender typicality and gender stereoty pe endorsement, the causal pathway between these two constructs remains unclear. This correlation may be due to the possibility that i eotype construction hypothesis children develop their stereotype attitudes ab out others based on their own traits and interests. Conversely, it may also be possible that harboring more flexible stereotypes decreases the anxiety that is associated with being gender at ypical leading children to feel comfortable adopting atypical tra its The author also noted that perceptions of gender appropriateness and the impact of gender typicality on adjustment are both likely to be influenced by cultural context and therefore that the relationship between gender typicality and stereotype beli efs may be subject to change in a different environment. Patterson (2012) closed by reflecting that while the COAT AM is based on cultural norms of gender typing, researchers should use a measure of gender stereotyping idiosyncratic prescriptive gender stereotypes into account This conclusion replicates the assertion of Tobin et al. (2010) that the Gender Self Socialization Model must only be tested using measurements, which allow children to assert their own beliefs about gender rol es. Unfortunately, very few studies hav e given
Idiosyncratic Beliefs and Gender Self Socialization 47 children the option to express their unique viewpoints. Two studies that have successfully done so will be discussed in the following section. Free Response Scales: An Alternative to Traditional Stereotyping Measures? Miller, L urye, Zosuls and Ruble (2009) asserted that the traditional list inventory approach to measuring stereotyping is highly problematic because it does not take into account the importance that children ascribe to stereotypes or different iate between stereotyping domains For example, though boys have been generally considered by psychological researchers to have more stereotypical behavioral preferences than girls, hat include primarily items about toys, objects, and activities. However, Miller et al. (2009) assert that these results may be skewed by the possibility that girls may also endorse stereotypes strongly but care more about stereotypes in other domains, su ch as appearance. Miller et content of 256 3 response inter views about their Miller et al. (2009) had several objectives in their evaluative re search of of stereotype knowledge (the degree to which knowledge of stereotypes exists in the memory) and accessibility (the degree to which stereotypes are salient and can be readily recalled). Though the researchers did not empirically differentiate between the two, they
Idiosyncratic Beliefs and Gender Self Socialization 48 asserted that list inventories of stereotypes did not measure the accessibility of stereotypes (due to the fact that, by presenting stereotype ite ms, list inventories make the stereotypes accessible); in fact, the researchers point ed out that no other studies had done so, and thus wished to assess which domains of stereotypes were the most accessible to children. The researchers also investigated t he effects of the differences between the gender of participants, age of participants, and gender of the stereotype target on the quantity of stereotypes listed, the order of stereotypes listed, and the domain of stereotypes listed by participants. It was predicted that girls would report more gender gender stereotypes than descriptions for the other gender, and that older children would report more gender stereotypes overal l than younger children. The research used archival data from a study conducted in 1988 1991, which had ended responses to a female interviewer were coded by two independent raters into fo ur general gender related domains (it was unclear how the gender typicality of domains were determined) as well as 7 subcategories within those domains: activities and to ys, appearance, found that of those categories, four (interpersonal, oc cupational, social role, and biological) were reported by participants less than 2% of the time and were excluded from analysis. Statistical analyses of the first responses mentioned by participants found that the most frequent domain of stereotypes descri bed for female targets was appearance,
Idiosyncratic Beliefs and Gender Self Socialization 49 while for male targets, the most frequent domain described was traits. Participant gender did not have a main effect on this trend; however, there was a significant interaction of participant age and gender on the typ es of stereotype domains reported overall such that as they progressed in age, boys tended to report more trait related stereotypes while girls tended to report more appearance related stereotypes for both target genders. In addition to measuring the part the quantity of gender stereotypes that each child listed in proportion to the overall nder of the target, domain of the stereotype and gender of the target were found to be interactive such that, in consistency with the analyses of the first responses, overall female targets were primarily described in terms of appearance and male targets w ere primarily described in terms of activity preferences (such as sports) and traits. Female participants also tended to report more appearance related stereotypes for girl targets than boys did, and older participants (4 th and 5 th graders; regardless of g ender) tended to describe female targets by appearance more than younger children. Younger children also described fewer traits than older children, and kindergarteners reported the most activity related stereotypes of all of the age groups. For both ster eotypes about male and female targets, participants of all ages and genders reported the same average quantity of gender stereotypes. However, statements provided by girls and older children contained a significantly higher proportion of gender stereotypes overall, and at every grade level girls provided more gender stereotypes than boys did. Unfortunately, limited by several factors, as
Idiosyncratic Beliefs and Gender Self Socialization 50 and therefore, that the relevant information as not reported For example, the researchers did not address which appearance related qualities female participants felt were the most important for their gender to emulate. This study was also limited by the fact that the researchers measured knowledge rather than attitudes about stereotypes, and that the data analyzed had been collected over 20 year s ago at the time of the author s writing. However, a vital contribution to the literature on gender development because of consideration of the differences in the content and y match stereotyped items with the correct gender. attitudes about them other studies have tested the effectiveness of measuring prescriptive stereotypes using open ended assessments Eagly, M ladinic and Otto (1992) expressed dissatisfaction wi th the rating scales that have traditionally been u sed to measure social attitudes and instead examined the potential validity of using a free response t echnique to undergraduate psychology students to respond to a questionnaire pertaining to one of four target groups women, men, Democrats, and Republicans. In order t o assess attitudes about stereotypes, the participants first rated the target group in comparison to ten other groups (e.g. teenagers, alcoholics), and subsequently indicated their like or dislike for the group on five 7 point items. Then, the participants completed the measurements at study: they created two 10 item lists one of characteristics that they believed to be typical of the group, and one of feelings that they typically had towards members of the group.
Idiosyncratic Beliefs and Gender Self Socialization 51 Then, they indicated the strength of that a ssociation by estimating the percentage of members of the target group for whom the y thought this would be true To elicit comparison information for the free response lists, participants also completed tw o additional traditional scales in which they rated the target group on a list of 32 traits that might characterize its members and a list of 32 emotions that they believed interacting with a member of the target group might elicit for them. Coefficient alphas for the means of the evaluative ratings that t he participants provided for the free response items were calculated for beliefs and attitudes separately: for beliefs, the researchers reported alphas of .62 for women, .77 or men, .88 for Democrats, and .86 for republicans; for affects, .61 for women, .8 1 for men, .91 for Democrats, and .94 for Republicans. These findings demonstrated a satisfactory internal consistency for the free response measures that nearly matched the coefficients obtained for the attitude scales (.80 for women, .82 for men, .91 for Democrats, and .94 for Republicans). Overall, regressions analyses of the response measures were an accurate and The current study utilized a similar open ended approach to the measurement of stereotyping as was implemented by Miller et al. (2009 ) and Eagly et al. ( 1992) with the has been employe d only rarely in psychological literature, it was expected that by allowing children to describe the attributes of each gender that they felt were the most important, Tobin et al. (2010) asserted that the Gender Self Socialization Model could only be
Idiosyncratic Beliefs and Gender Self Socialization 52 therefore a second goal of the current investigation to compare the useful properties of an open ende d stereotyping measure with a traditional trait inventory of the style, which the GSSM eschews. Finally, the under studied hypotheses predicted by the GSSM were examined. The Current Study Although the study of when, why, and how children come to underst and and enact societal gender roles has grown exponentially throughout the past half century, many questions raised by modern researchers remain unanswered, two of which are central to the study at hand. First, because the vast majority of literature has e mployed a pre conceived framework of cultural gender norms in order to measure gender typing, little is known about the unique value that individual children ascribe to traits that are traditionally associated with either gender. Traditional measures of about gender, such as the Bem Sex Role Inventory (BSRI; Bem, 1981) and its (COAT ratings as eit her instrumental or expressive based on research conducted with adults which has been found to confirm their gender typed social meaning. However, many modern theorists b elieve that gender development i s a unique process for each individual. It is possible that children, especially those who are raised in settings that are gender progressive or those in minority populations might ascribe different meanings to traditionally gender typed traits. For both of these aforementioned reasons, some of the most recent literature in the field has called for the termination of the use of traditional gender schema inventories in favor of a mor e qualitative approach (Perry & Pauletti, 20 11; Tobin et al, 2010). Second
Idiosyncratic Beliefs and Gender Self Socialization 53 the role of these unique beliefs in the development of gender identity is unclear. Although researchers have investigated the interaction between gender contentedness, typicality, and felt pressure to conform to gender stereo types, recent findings (Yunger et al., 2007) suggest that the relationship between these three factors is moderated by the content of tion which appropriate roles for men and women. The present study examined the meaning that girls in mid late childhood ascribe d to their gender category and, in light of these findings assessed the relationship perceptions. Participants completed two pen and paper open ended self report measures of self perceptions and gender stereotypes which have b een created for the study at hand. One (2001) measures of gender typicality, gender contentedness, and felt pressure to conform to gender stereotypes, and two subscales from Traits Attitude Measure (COAT scale were compared to assess the following research questions: 1. How do girls today conceptualize gender roles? What traits and attributes do they consider to be important for boys and girls to possess? How do these qualitatively de Measure (COAT AM; Liben & Bigler, 2002)? H1: It was predicted that, while the aggregated responses will reflect the findings of the COAT AM and PM with some accuracy, participants would also generate
Idiosyncratic Beliefs and Gender Self Socialization 54 a plethora of other responses, including non tradit ional stereotype items in part as a result 2. Is the use of open ended stereotype assessments better at predicting measures of gender identity and self perception s than the COAT AM (Liben & B igler, 2002)? H2: Each stereotype assessment was used to generate a score indicating the degree to which the participants hold non traditional stereotype beliefs, which were based on the frequency of items on each list that the participant indicates is ac ceptable for both sexes. It was predicted that gender egalitarian scores which had been generated from open ended assessments would be stronger predictors of scores on assessments of self perceptions and gender identity than the COAT AM because the free re (Liben & Bigler, 2002). 3. How do stereotypes, self perceptions, and gender identity relate to one another? H3: It was hypothesized that if the open ended self perception assessment created for this stud y assessed the intended construct low scores would predict low gender typicality ratings. Previous research has agreed that children judge their gender typicality based on their gender typed attributes (Patterson, 2012). H4: Egalitarian attitudes were e xpected to decrease the strength of the relationship between gender typicality and gender contentedness; girls would be expected to continue to feel content with their gender, even when they fe lt that they were atypical of their gender if they had flexibl e beliefs about gender egalitarian attitudes were positively related to one another, and offers an alternative explanation for that results. H5: Rigid stereotypic views we re expected to predict self perceptions that were self and higher felt pressure to conform was expe cted to explain this variance. These hypotheses reflect only a few possible relationships between a wide range of variables under investigation. It is important to keep in mind that in the current analysis, there are two variables related to stereotypes (free response gender egalitarianism score and COAT gender role flexibility score), two scores related to self perceptions (the proportion of free response items describing the self that matched each
Idiosyncratic Beliefs and Gender Self Socialization 55 n that matched her stereotypes about boys) and three gender identity variables (gender typicality, gender contentedness, and felt pressure to conform to gender stereotypes) and so a total of 36 multiple regression analyses were conducted. Thus, it was dif ficult to produce further directional hypotheses about all possible combinations of the variables at study Rather, the most perceptions of gender, which have not be en studied in relation to gender identity or self perceptions. These components may influence key gender related cognitions. Method Participants Participants were 4 5 girls aged 7 11 years old who were recruited based on their enrollment at an after school center for girls in Southwestern Florida where the research was conducted The facility serves over 100 girls aged 5 14, and with a motto that girls ach children to oppose gender stereotypes. Demographic information about the individual participants in the study was not obtained in order to ensure their anonymity; however, the daycare center as a whole is quite diverse, serving girls from 20 schools fr om across the county. Sixty percent of the girls who attend the daycare receive d some level of financial assistance, 45% wer e White, 28% were Hispanic, 15% were African American, and 11% were Multi racial. The center conducts its regular programs on Monday s Fridays from 3:30 6pm, and also offers summer programming. The regular curriculum includes homework help, art, science, economic literacy, media literacy, and a leadership program.
Idiosyncratic Beliefs and Gender Self Socialization 56 Materials Gender stereotypes and self perceptions. The questionnaires us ed to assess viewed in Appendix A. These measurements were developed for the research at hand. They were distributed on pen and paper surveys, which consisted of a brie f set of instructions followed by a series of 4 5 spaces ( 15 spaces each for self perceptions, stereotypes about girls, and stereotypes about boys, respectively) for respondents to fill in answer s The self perceptions assessment, which was presented first, asked participants to The stereotype assessments were given after the self perceptions assessment, and had a very similar format to the self perceptions assessment but asked participants to make lists of attributes that described boys and girls (presented in a counter balanced order). The stereotype assessments also included a three point scale next to each item for the participants to indicate the importance that th ey felt that each item was to being a member of each respective gender (1= not important, 3= very important), followed by the boys/girls; the opposite sex of the stereotype being described and a place for the respondent to circle 3 item ranking was determined to be too confusing for the participants and was excluded from further analysis. The instructions on stereotype assessments was formatted based on the instructions given on the COAT (Liben & Bigler, 2002; in Appendix B), and like the COAT, asked participants what they thought each gender should be like. The use of the analytic findings that ttitudes about rather than knowledge
Idiosyncratic Beliefs and Gender Self Socialization 57 of gender stereotypes. The instructions on the free response prompt for stereotypes about girls read: Make a list of at least 10 things that YOU think a girl should be! THEN, circle important for girls to be like that; circle number 2 if you think it is kind of important for girls to be like that; circle number 3 if you Then circle whether you think boys should also be like that. There are no right or wrong answers! We just want to know what YOU think. These instructions, as well as the designs of these measures, closely mirror those Occupations Activities, and Traits A ttitudinal and Personality Measures (COAT AM & COAT PM; Liben & Bigler 2002). Liben and Bigler (2002) advise that the COAT AM and PM can be administered to groups of children by formatting the scale in large type and reading each item aloud while the children follo w along. The researchers assert that the personality measures ( which assess self perceptions) be a dministered before the attitudinal measures ( which assess stereotypes) to avoid making gender related beliefs highly salient prior to self ratings. Only one subscale, a combination, or all three for each test can be implemented because all three subscales are highly correlated and load together when facto r analyz ed; therefore, the open ended questions in this study allowed children to list occupations, activities, or traits for each scale. This omnibus design, in which a wide variety of attributes across many domains are tested, was chosen for the current study in stead of focusing on a single focal beliefs about gender to be assessed. Tobin et al (2010) assert ed researchers wish to make more general state ments about how GSSM processes work, p. 613).
Idiosyncratic Beliefs and Gender Self Socialization 58 Scoring. were scored in two ways. First, i n order to quantitatively compare the particip and gender identity, a gender egalitarianism score was derived by summing the total acceptable for the opposite gender. Thus, before taking into account the proportion of egalitarian responses to total responses, the highest possible g ender egalitarianism score was 3 0, and the lowest was 0 (the lower scores would indicate that the participant has very r igid views about gend er roles). The total number of egalitarian response on both stereotype lists was then divided by the total number of stereotypes that each participant had listed. ended stereotype responses were also aggregated and ranked based on the importance that each child assigned to their stereotype in order to derive a prototype of the meaning of gender identity. This qualitative information was of general interest, and was used to inform the interpretation of the results of the correlational d ata between the three target variables. All of t he responses of the participant s stereotypes for girls were combined into a single list, and then two research ers reviewed the list and compiled a list of attribute categories under which multiple stereotype s could be grouped Items were placed into the same stereotype category if they were judged different versions of the same word if they contained the same word within a phrase, or were very similar in meaning. Gender identity. Th assessed gender typicality, felt gender compatibility, and felt pressure to conform to gender stereotypes. The first two measures are 6 items long and the third is 10 items. All
Idiosyncratic Beliefs and Gender Self Socialization 59 three scales are in t he same format; each item gives two oppositional statements about and the participant chooses which statement they agree with and whether that identity can be viewed in Appendix B. Scoring. Responses on each item translate to a score of 1 4, such that high scores reflect greater gender typicality, greater gender contentedness, and greater felt pressure to conform. Reverse scores are given to items 3 and 5 on the gender typicality scale, 1 and 3 on the contentedness scale, and 1, 2,4,8,9, and 10 on the felt pressure scale. The items have been developed to be understandable to participants in middle childhood and early assessment, it was excluded fr om this research because it is not included as part of Tobin Two (COAT AM; Bigler & Li performance on a traditional style measure of gender typing. Each subscale contained a list of 25 items (either traits or ac tivities, respectively) and asked children to designate whether they b AM can be administered to groups of children by formatting the scale in large type and rea ding each item aloud while the children follow along. Only one subscale, a
Idiosyncratic Beliefs and Gender Self Socialization 60 combination, or all three of the original measure can be given because all three subscales response s to these items were not the primary focus of the study, the COAT AM was administered in order to create a basis of comparison for the novel stereotypes that the children generated. The full questionnaires can be viewed in Appendix B. Procedure Participants were selected to be involved in the current study if they were enrolled at the daycare center where the research was conducted, met the age criteria, and had completed participant assent and parental consent forms. The study was conducted over a two week period during the end of January, 2013. The experimenter received permission to conduct the research after developing a relationship with the staff. Participants were tested in groups of 10 15 children at a time, and two groups were tested per day on two days each week ( Tue sd ays and Thur sd ays ), and the order in which each group was tested was consistent at week #1 and week #2. The staff and the experimenter worked together closely to schedule testing during a time period in which no unusual even ts were occurring at the daycare center in order to ensure that the conditions of the survey administration were as standardized as possible for each of the groups. Each group met in the same classroom, which was located at the daycare facility, and the sa me regularly scheduled activities were occurring at the center during both sessions. Each participant completed the survey over the course of two sessions, which were spaced apart by approximately one week. Dividing the administration of the survey into t wo sessions spaced apart by at least one week was necessary in order to minimize the possibility that
Idiosyncratic Beliefs and Gender Self Socialization 61 measures and to reduce the possibility of fatiguing the children. Partic ipants completed the self perception and open ended stereotype assessments during the first testing period, and the gender identity measures and COAT stereotype scales during the second assessment. On the first assessment, the order of the stereotype lists about boys and girls were counterbalanced randomly in order to ensure that completing one list of stereotypes would not influence the responses on the second list. During both sessions, the participants were seated so that they could not view one s responses and were instructed not to communicate with one another. The experimenter explained to the children that they would be asked to answer questions about themselves, that they should try to answer as honestly as possible, and that their participat write their names, but instead wrote down personal identification numbers that the daycare center had assigned to them in order to identify their survey while ensuring anonymity. The researcher read the instructions aloud on each page to the participants, who wrote down their answers. Prior to their completion of the gender identity scale durin g week #2, the experimenter demonstrated how to respond to the questions using an example question which modified an item on the gender typicality scale to pertain to the degree to which the participants liked broccoli; this example item was necessary beca use of the potentially confusing format of the gender identity subscales. Completion of the survey took approximately 30 45 minutes per session. Because the center frequently conducts internal evaluations using the same pen and paper self
Idiosyncratic Beliefs and Gender Self Socialization 62 report technique this activity was expected to be familiar and comprehensible to the participants. After the research, the participants were given the opportunity to ask questions, and the researcher debriefed them on the intention of the study. The children were told tha t if they had any further questions or concerns, they could speak privately with an administrator at the daycare center. Finally, the participants were given a small Results Descriptive statistics Stereotype lists. Forty four p articipants generated 901 stereotypes, with an average of 20.47 ( SD = 6.37 ) total stereotypes per participant. A total of 467 stereotypes about girls were listed, with an average of 10.61 ( SD = 3.12) per participant. A total of 434 stereotypes for boys were listed, with an average of 9.86 ( SD = 3.58) items per participant. Two independent coders grouped the attributes into 57 ca tegories for the stereotypes of girls and 52 categories for the stereotypes of boys, with 94% agreement for were considered to be in the same category if they were different or modified versions of the same word or were considered by both raters to be identical in meaning. Due to the limited sample size, items which occurred more than once and could not be grouped with any other items were considered to be independen t categories. Idiosyncratic items that did not share a meaning with an existing category were classified as miscellaneous. However, because the primary interest of this analysis is in the idiosyncratic items themselves, they were not excluded from analysis all together.
Idiosyncratic Beliefs and Gender Self Socialization 63 Tables of categories for each gender stereotype, listed by participant frequency and including frequencies of each item included in each category can be viewed in Tables 1 and 3 respectively. F requencies for each category were derived by ta llying the number of times that each category appeared in the data and dividing that sum by the total number of participants ( N =44) thus percentages in each table should not add up to 100% and in cases in which participants listed the same categories mult iple times, frequencies exceed 100%. The limitations of this approach to calculating the frequencies of free response data will be enumerated in the discussion of these analyses. Free Response Gender Egalitarianism Score. The degree of g ender egalitarianism as measured by the Free Response scales was calculated by summing the number of t imes that respondents answered on the Free Response stereotyping measure, indicating that the stereotype she listed would be acceptable for the opp osite sex. This score was then divided by the total number of stereotypes that the respondent of responses. The mean Gender Egalitarian ism score was .807 ( SD = .18), wi th an observed range of .33 to 1.0 and a possible range of 0 to 1.0, with higher scores indicating more flexibility in stereotype beliefs. This indicates that participants felt a great deal of gender egalitarianism in relation to the items that they had li sted as characterizing males and females. Despite the wide range of responses, however, the Free Response Gender Egalitarianism scores were quite skewed; 17 participants earned the maximum score. The o verall mean of flexibility of gender stereotypes as measured by the activities and traits subscales of the
Idiosyncratic Beliefs and Gender Self Socialization 64 SD = 22.65) with a possible and observed range of 0 to 100 (with higher scores indicat ing greater flexibility of gender stereotypes). This figure indicates that participants answered that the items listed on the of the time and thus had a high level of g ender egalitarianism Like the Free Response stereotyping scores, the COAT scores were quite skewed; 16 participants received the maximum score. Self Perceptions. Self perceptions were scored by calculating the number of elf Perceptions scale which matched the items listed on generated list of prescriptive stereotypes for girls and for boys, respectively. Thus, two separate Self Perception scores were calculated based on the degree to which the items scale matched her lists of stereotypes for each gender. Items were considered to be matching if they were the same word, or if they were different forms of the same word with a shared meaning. Two independent coders judged 20% o f the data (selected randomly) and agreed on 100% of the Self Perception scores. Because each participant listed between 3 and 15 items for the Self Perceptions scale, each Self Perception score was averaged by the total number of self perceptions listed. Therefore, if a participant had listed 10 self descriptors, and 5 of those items matched the items on her list of stereotypes for girls while only 2 matched the items she had listed on her stereotypes for boys, her Self Perceptions score for girls was .5 a nd her Self Perceptions score for boys was .2. For both scales, the minimum possible score with either of her stereotype lists, and the highest possible score was 1, suggesting that
Idiosyncratic Beliefs and Gender Self Socialization 65 every item on her self perception list corresponded with her list of stereotypes. The mean Self Perception score in comparison to stereotypes for girls was .34 ( SD = .21), and the mean Self Perception score in comparison to stereotypes fo r boys was .30 ( SD = .24). Both scores had an ob tained range of 0 .93. Consistent with the finding that both stereotyping scales revealed a higher degree of gender egalitarian stereotype attitudes a phenomenon that has been observed in other populations (Liben & Bigler, 2002; Signorella et al., 1993), the two subscales of Self to her lists of stereotypes about girls and stereotypes about boys, respectively) were found to be strongly correlated to one another ( r (42) =.76, p .001). This suggests that the participants generally envisioned themselves to be about as similar to their prescriptive stereotypes for boys as they did for their prescriptive stereotypes for girls. Gender Identity. The overall mean of gender typicality as scored by the Gender Typicality subscale of the Gender Identity Scale was 2.29 (highest possible score 4; SD = .65) with an o btained range of 1 to 3.6. Higher scores on the typicality scale indicated gender contentedness as scored by the Gender Contentedness subscale of the Gender Identity Scale was 2.32 ( SD = .72) out of 4 with an observed range of 1.17 4.0, with overall mean of felt pressure to conform to gender stereotypes as measured by the Felt Pressure subscale of the Gender Identity scale was 2.08 ( SD = .77) out of 4 with an observed range of .9 3.8, with higher scores indicating increasing feelings of pressure
Idiosyncratic Beliefs and Gender Self Socialization 66 Pearson correlation coefficients were calculated between each of the three Gender Identity variables. Consistent with previous findings in the literature for Caucasian samples (Corby, Hodges, & Perry, 2007; Egan & Perry, 2002; Yunger, Carver, & Perry, 2004), Typicality was correlated with Contentedness, though this observe d correlation was much stronger than has been found in previous research ( r (43 ) = .74, p .001). The previous literature has found mixed results regarding the relationship between Felt Pressure with the other two Gender Identity variables; in this case, it was found that Felt Pressure was negatively correlated with both Typicality ( r (43)= .45, p = .002) and Contentedness ( r (43)= .45, p =.00 2). Importantly p revious research has found that the relationships between the three constructs differ greatly depending on the ethnicity of the sample (Corby, Hodges, & Perry, 2007; Yu & Xie, 2009). The meaning of these results will be discussed later. A list of correlations bet ween all of the variables under study can be found in Table 5 These scores were then used to assess the efficacy of each stereotyping measure and were correlated with the Stereotyping and Self Perception scores in order to test the three interactive hypo theses of the Gender Self Socialization Model (GSSM; Tobin et al., 2010). Relationship b etween the stereotyping measures. It was hypothesized that the gender egalitarianism scores derived from the free response stereotype scales would be a better measure ctivities, and Traits Attitudinal M easure (COAT AM; Liben & Bigler, 2002) because the free response scales would reflect the attributes that each individual participant felt were most important for each ge nder to emulate. Therefore, it was expected that both quantitative
Idiosyncratic Beliefs and Gender Self Socialization 67 stereotyping measures would be mildly correlated with one another, but that the free perceptions and gende r identity. On the contrary, the gender egalitarianism scores derived from the free response stereotyping lists were uncorrelated (though approaching significance) with the gender flexibility scores yielded by the COAT measure ( r ( 42)= .22, p = .15). Test ing the Gender Self Socialization Model Hypothese s In order to test the hypotheses of the Gender Self Socialization Model (Tobin et al., 2010) that significant interactions between two of the three constructs (stereotype beliefs, self perceptions, and gender identity) would influence the outcome of the third, Pearson correlation coefficients were first calculated between each pair of variables (Free Response Stereotype score, COAT AM Gender Flexibility score, Self Perceptions related to stereotypes abo ut boys, Self Perceptions related to stereotypes about girls, Gender Typicality, Gender Contentedness, and Felt Pressure to Conform to Gender Stereotypes). Then, multiple regression analyses were conducted. Correlations A full list of correlations betwe en all of the v ariables can be found in T able 5 The only significant correlations between the primary constructs in the current analysis were between t he Free Response Gender Egalitarianism score s with the Self Perceptions Boys scores ( r (42)= .30, p = .04 5), as well as a significant, negative correlation between the COAT with Felt Pressure ( r (43)= .34, p = .02). The Free Response Gender Egalitarian scores were unrelated to Typicality, Contentedness, Felt Pressure, and Self Perceptions Girls, and the COAT was unrelated to Typicality, Gender
Idiosyncratic Beliefs and Gender Self Socialization 68 Contentedness, or either Self Perceptions measure. These results were surprising; it was expected that egalitarian stereotype beliefs would be related to greater contentedness, greater typicality, and lower felt pressur e, as has been found in previous literature (Egan & Perry, 2001; Patterson, 2012). Self Perceptions Girls and Boys were unrelated to any of the Gender Identity variables, counter to the hypothesis that higher similarity between d their stereotypes about girls would predict higher Typicality would predict lower Typicality. However, the two Self Perceptions variables were highly correlated with one another ( r (42) = .76, p .0001), indicating that there was a great degree of similarity between how participants described themselves in relation to how they thought that boys should be and how they thought that girls should be. Additionally, there were significant relationships between all three gender identity variables with one another: typicality was highly correlated with contentedness ( r (43)= .74, p .001), typicality was negatively related to felt pressure ( r (43)= .48, p =.002), and contentedness was negatively related to felt pressure ( r (43)= .45, p =.002). This is generally consistent with previous findings on the relationship between gender identity variables (Egan & Perry, 2001; Yunger et al., 2004). Multiple regression analyses. As a further test of the Gender Self Socialization Model Hypotheses, multiple regression analyses were performed with two of each of the three constructs serving as predictor variables and the third construct serving as a criterion variable; a total of 36 analyses were conducted. Unfortunately, conducting so many analyses likely inflated the type 1 error rate; this limitation will merit discussion in the next section. Results for all significant regression mo dels can be found in Tables 6 10
Idiosyncratic Beliefs and Gender Self Socialization 69 Due to the extr emely high volume of regression analyses conducted, only the values of significant models are reported. Importantly, because both stereotyping variables (and, to a lesser extent, both self perceptions variables) were quite skewed, logarithm transformation s were conducted on the criterion variable in each analysis. Because some of the values in the Self Perceptions and COAT measures were equal to 0, it was necessary to convert these values to .0001 in order to conduct the logarithm transformations. Stereot ype emulation hypothesis. The stereotype emulation hypothesis of the Gender Self Socialization Model suggests that Stereotypes and Gender Identity will interact to predict Self Perceptions. For example, it was expected that children who expressed egalitari an stereotypes and felt typical of their gender would be more likely to have high Self Perceptions Boys scores because they would feel more comfortable transgressing traditional gender boundaries (Patterson et al., 2012). Therefore, 12 regression procedure s were conducted with either Self Perceptions Girls or Self Perceptions Boys as the criterion variable and two of the following as predictor variables (one stereotyping and one Gender Identity variable in each analysis): Free Response Gender Egalitarianism score, COAT score, Gender Typicality score, Gender Contentedness score, and Felt Pressure score. Of these, four models were significantly predictive; all of which predicted the Self Perceptions Boys score s These findings can be viewed in Tables 6 9. The COAT and Typicality scores predicted Self Perceptions Boys, explaining only 15% of the data but achieving statistical significance ( R 2 =.15, F(2, 42)=3.57, p= .038). Only the influence of
Idiosyncratic Beliefs and Gender Self Socialization 70 the COAT was significant ( = 2.49 t (42 )=2.45, p =.01), while the effect of Typicality was not. Likewis e, a second multiple regression analysis found that the COAT scores and Contentedness scores were significantly predictive of Self Perceptions Boys scores, and explained 16.2% of the variance ( R 2 = .162 F (2,4 3 )=3.95, p =. 027). It was found that the COAT scores significantly predicted Self Perceptions Boys scores ( = 2.45, t (41 ) = 2.64, p =.018) but Contentedness did not. A third multiple regression analysis found that the COAT and Felt Pressure were also predictive of Self Perceptions Boys, and explained 15% of the variance ( R 2 =.15, F (2,43)=3.62, p =.036). Again, only the effect of the COAT was significant ( = 2.63, t ( 4 1)= 2.67, p =.01), while the effect of Felt Pressure was not. The Free Response Gender Egalitarianism score had a similar predictive capacity on the Self Perceptions Boys scores as the COAT did, which was surprising given that the COAT and Free Response measures were uncorrelated and did appear to share any similar properties prior to this analysis. The Free Res ponse score and the Contented ness score were predictive of 15 % of the variance associated with Self Perceptions Boys ( R 2 =.15, F (2 42 )=3.61, p =.036), and only the Free Response score was predictive of Self Perceptions Boys ( = 2.92, t (40 )=. 2.51, p =.02). While the gender identity variables did not play their expected role as predictors of Self Perceptions, these results demonstrate that when Self Perceptions Boys scores Self Perceptions Boys scores increased by about 2.5 points though importantly that value reflects the scores post logarithm transformation, and are not a fully accurate representation of the raw scores This finding is notable considering that the mean Self Perceptions Boys score was .29; so, it is likely that the few participants who did attain
Idiosyncratic Beliefs and Gender Self Socialization 71 high Self Perceptions Boys scores also had very high egalitarianism scores (as most participants did) and this could explain why these findings only account for such a small amount of the variance on Self Perceptions Boys. However, both Self Perceptions scores were transformed using logarithms, which should have helped to limit the effect of this skewed data. Identity construction hypothesis The identity constru ction hypothesis of the Gender Self Socialization Model suggests that Stereotypes and Self Perceptions will interact to predict Gender Identity. To test this theory, 12 regression analyses were conducted with either Gender Typicality, Gender Contentedness, or Felt Pressure as the criterion variable and two of the following as predictor variables (one stereotyping and one Self Perception variable in each analysis): Free Response Gender Egalitarianism score, COAT score, Self Perceptions Girls, and Self Percep tions Boys. None of these models were significant, and thus the identity construction hypothesis could not be confirmed. However, one model was approaching significance: COAT scores and Self Perception Girls scores were close to being predictive of Felt Pr essure and explained 12.5% of the variance ( R 2 = .125 F (2,43)= 2.94, p =.063). However, only the effect of the COAT was significant ( = 1.12, t (41 )= 2.22, p =.032) while the effect of Self Perceptions Girls was not. Thus, when Felt Pressure was held constan t, for every 1 point increase in COAT scores of stereotype flexibility, Felt Pressure could be expected to decrease by 1.12 points (on a 4 point scale ; these values reflect the scores post logarithm transformation ). All t hough no previous research has tested the influence of high gender egalitarianism on Felt Pressure, theorists of gender development (e.g. Bem, 1981; Liben & Bigler, 2002) have a rich history of making the reverse prediction; that children form
Idiosyncratic Beliefs and Gender Self Socialization 72 rigid stereotypes because they feel strong societal pressure to do so. This finding suggests more pressure they feel to conform to them. However, no causal relationship can be established based on these data. S tereotype construction hypothesis Finally, some support was found for the stereotype construction hypothesis of the Gender Self Socialization Model, which suggests that Self Perceptions and Gender Identity will predict Stereotype beliefs. Twelve regressio n analyses were conducted with either Free Response Gender Egalitarianism scores or COAT scores as the criterion variables and two of the following as predictor variables: Self Perception Girls, Self Perception Boys, Gender Typicality score, Gender Content edness score, and Felt Pressure score. Of these models, one was significant: Typicality and Sel f Perceptions Boys predicted 15% of the variance of the Free Response Gender Egalitarianism scores ( R 2 = .15, F (2, 42)= 3.57, p = .04). Only Self Perc eptions Boy s had a significant e ffect ( = .06. t (40 )=2.22, p =.03), while the effect of T ypicality was not significant. Thus, when Free Response scores were held constant, for every 1 point increase in Self Perceptions Boys scores, the Free Response scores could be ex pected t o increase by .06 points (with .0001 being the lowest possible score and 1 being the highest ; these values reflect the scores post logarithm transformation ). Two other models were quite close to approaching significance: Self Perceptions Boys and Felt Pressure were almost predictive of the Free response score ( R 2 =.13, F (2, 42)= 2.96, p =.063). In this model, only Self Perceptions Boys was significant ( = .06.
Idiosyncratic Beliefs and Gender Self Socialization 73 t (40 )=2.04, p =.05) while Felt Pressure was not, suggesting that when the criterion variable was held constant, for every point that Self Perceptions Boys increased, Free Response Gender Egalitarianism scores increased by .06 points (with .0001 being the lowest possible score and 1 being the highest ; these values reflect the scores post logarithm transformation ). Self Perceptions Boys and Felt Pressure were also almost very close to significantly predicting 14% of the variance of the COAT scores ( R 2 = .14, F (2, 43)= 3.21, p =.0508). However, unlike in the other models, neither variable was signific antly effective when predicting COAT scores, although the influence of Felt Pressure was approaching significance ( = .02. t (41 )= 1.92, p =.06). Though the effect of Felt Pressure was insignificant, it is interesting that Felt Pressure generally had a neg ative (if non significant) effect on Gender Egalitarianism scores. Discussion This investigation was guided by two primary goals: (1) to discover the idiosyncratic attributes that young girls associate with femininity and masculinity and to establish differences between these findings and previous assumptions about the research, and (2) to investigate the hypothesis of the Gender Self Socialization Model which claims that gender stereotypes, self perceptions, and gender identity would each interact with the other two in order to achieve cognitive constancy as well as to which included wn idiosyncratic stereotype beliefs would influence the outcome of the interaction between these three constructs. The first analysis was fruitful and painted a clear vision of the unique views that the participants held about gender roles, which often
Idiosyncratic Beliefs and Gender Self Socialization 74 de viated from the characteristics that have been historically associated with men and women in the psychological literature. In the second analysis, correlations were calculated between each of the three constructs A significant negative correlation was fou nd between high scores on the Measure (COAT AM; Liben & Bigler, 2002) and feelings of pressure to conform to gender stereotypes and a significant positive correlation was found between the Free Respo nse Gender Egalitarianism scores and Self Perceptions Boys scores All of the gender identity variables were correlated with each other in a manner that was consistent with the previous literature ( Egan & Perry, 2001; typicality and contentedness were posi tively correlated with one another and negatively correlated with felt pressure), and the two self perception variables were highly inter correlated. However, no other meaningful relationships between the three gender related constructs were identified, de viating from findings in the previous literature and leading to a rejection of the hypothesis that the free response measure was a more useful stereotype assessment than the COAT AM and a rejection of the hypothesis that low self perception scores in compa rison to stereotypes about girls would be related to lower feelings of gender typicality. Multiple regression analyses were also conducted on each of the variables, with each variable serving as a criterion variable and a pair of two other variables servin g as predictors (one from each of the two constructs not represented by the criterion stereotype construction, identity construction, and stereotype emulation hypotheses. I n the following sections, the implications of these results are discussed.
Idiosyncratic Beliefs and Gender Self Socialization 75 Gender Stereotype A nalyses The first assessment conducted was of the unique views that the children had ith one another, AM; Liben & Bigler, 2002). Although the COAT AM is not intended to be viewed as an exhaustive list of highly stereotyped items for each gender, viewing the responses of the current study in comparison to the items on the COAT AM illuminates the similarities and discrepancies between the way that the children who participated in thi s study viewed gender in contrast with the researchers who established the formal psychological measure. Due to the fact that the free response stereotype scales included a quantitative assessment which replicated the coding schema used to assess gender fl exibility in the COAT AM, comparisons of these numeric quantitative scores will be discussed first, followed by a descriptive analysis of the items listed on each free response scale. Egalitarian scores on the free response and COAT AM measures P articip ants rated the items on the COAT 75% of the time and rated the items on the free response scales as acceptable for both boys and girls 85% percent of the time, suggesting that the attitudes of gender flexibility on th e traditional trait inventory were mirrored by attitudes of gender egalitarianism on the free response stereotype measure. In contrast, in the research used to establish the COAT AM, the researchers found that fewer than half of the children who rated the items on the scale considered them to be appropriate for both men and women (Liben & Bigler, 2002). The first hypothesis predicted that gender egalitarianism scores generated from these
Idiosyncratic Beliefs and Gender Self Socialization 76 assessments would be more successful at predicting scores on the self perception and gender identity measures than would a traditional trait inventory (the COAT AM) because considered to be most important characteristics of males and females rather than based on their attitudes about limited number of items on the COAT AM This hypothesis was rejected; the Free Response Gender Egalitarianism Score was significantly correlated with Self Perceptions Boys, but was not associated with Self Percep tions Girls or with any of the gender identity variables Higher ratings of gender flexibility on the COAT AM had a significant, negative relationship with felt pressure to conform to gender stereotypes, but did not predict gender contentedness, gender ty p icality, or self perceptions. The scores yielded by the two stereotype assessments were also unrelated to one another, despite the fact that both scores indicated high levels of non traditional beliefs about gender roles among most participants. The lack of relationship between the two stereotype measures is unsurprising; no prediction was made about the correlation between the two because it was expected that if one was more accurate than the other, or if neither w ere particularly effective, the two scores would not be related. However, from these findings it is not clear which is the case. Both scores yielded far more egalitarian responses than have been observed in other populations (Liben & Bigler, 2002; Signorella et al., 2002), which could ha ve been explained by several factors. Participant fatigue may have played a role in the abnormal distribution of gender egalitarian scores as the COAT AM measures were administered after the Gender Identity scale, which was particularly long and difficult to c omprehend for the participants. It is also possible that the COAT AM was an effective measure and
Idiosyncratic Beliefs and Gender Self Socialization 77 the free response stereotyping measure was not. Findings from previous research suggest that the format of the free response questionnaire should have b een adequate at assessing the same construct as the COAT AM scale; following the findings of Signorella et al. (1993) that non forced choice measures are more adequate at assessing attitudes about stereotypes (as opposed to knowledge that stereotypes exist ) The questions on both the free given the option to select that each item listed could be applicable to both genders. The format of the free response stereotype scale nearly r eplicates that of the COAT AM (both scales and accompanying instructions can be viewed in Appendices A and B) and was administered using the same guidelines for that scale. Furthermore, previous researchers have had success using free response scales ; Eag ly et al. (1992) and Miller et al. (2009) found that using free response lists to assess stereotypes was comparable to using traditional scales (though unlike that study, in this case valance ratings were not given), and similar open ended options have suc cessfully been used to gauge self perceptions in a variety of research (Brinthraupt & Lipka, 1985; Montemayor & Eisen, 1977). Additionally, it is possible given the high number of maximum scores on both stereotyping measures that social desirability bias played a role in leading children to select egalitarian answers The role of such bias might make sense given that the atmosphere of the daycare center where the research was conducted was infused with regular anti gender stereotype education. On the oth er hand, this atmosphere may have compelled the participants to genuinely internalize the anti stereotyping message egalitarian is supported by the finding that as i s indicated later on in the discussion of the
Idiosyncratic Beliefs and Gender Self Socialization 78 descriptive analysis, most of the items on the stereotype lists for boys and for girls were very similar, suggesting that it is possible that the participants did hold highly gender egalitarian beliefs. Limitat ions of the quantitative stereotyping scores. Several limitations may have impact ed the comparability of the two stereotyping scores. First, the theoretical approa ches to each stereotyping scale differ slightly. The COAT AM assumes that gender schemata are interrelated, and gender typed items are included on the activity and trait measures because they correlate highly with other items of the same domain and gender type on the measure. In contrast the GSSM follows the assumption based on the findings of Spence (1985) and other similar researchers that gender schemata are not necessarily related, but instead that each individual child learns to associate their gender with unique characteristics bas domains that were measured on the free response scales were simply too broad; while the Gender Self Socialization Model asserts that global, generalized assessments can be used to measure stereoty pe attitudes so long as they are used only to come to broad, authors do assert that the construct is better tested in one or two specific domains (for example, academic s uccess or body image). Therefore, it is possible that the two stereotyping scales measured different constructs. If either of the two measures had been successful, though, one might expect that their scores would be related to the gender identity and self perception variables in the manner that is predicted by the Gender Self Socialization Model and supported by previous research. For example, Patterson (2012)
Idiosyncratic Beliefs and Gender Self Socialization 79 found that non traditional gender stereotype attitudes were associated with lower feelings of gen der typicality. However, this relationship was not observed in the current analysis. The two stereotyping scores might also measure different constructs As Miller et al. (2009) point out, there is an important distinction between the accessibility of ste reotypes and the availability of stereotypes the free response measure judges which stereotypes are most accessible, while the COAT AM tests only availability (due to the fact that stereotypes are made accessible to participants by providing them in the me asure). Unfortunately, Miller et al. (2009) did not compare the free response measures used in that study to a traditional scale, and no other study (to the best knowledge of the researcher) has done so, and thus it is impossible to determine whether parti stereotypes would be more or less rigid based on the style of scale used. It is also intuitive to assume that because the free response scale asked children to list the stereotypes that were the most important to them, they may have chosen stereot ypes that were more rigid, rather than less, though again there is no way to determine this. Because both stereotyping measures elicited the same general pattern of data, it is more likely that the r (b) due to fatigue (though this is unlikely given that the measures were conducted on separate days) or (c) social desirability. The first is possible after all, Signorella et al. (1993) as well as Katz and Ksansnak (1994) both found that children who w ere surrounded by a more gender egalitarian social environment had more flexible views, and indeed the daycare center where the research was conducted participated in frequent anti stereotype teachings. It is also possible for this reasons that participant s desired to give socially desirable responses and thus rated the items more egalitarian.
Idiosyncratic Beliefs and Gender Self Socialization 80 Limitations of qualitative free response data. The descriptive findings of the free response stereotyping measure, which will be discussed in the following section, may also have been affected by these and other factors In addition to the limitations mentioned in the discussion about gender egali tarianism scores, the qualitative data may have been limited by the fact that, although the format of the instructions mi rrored those (1993) meta order to gauge stereotype attitudes, some of the responses seemed to gauge pa des ires rather than their beliefs about the general characteristics of each gender. Ultimately, because as Tobin et al. (2010) note, from a theoretical perspective, the interactive properties of the stereotype construct with gender identity and self perceptions should be based on f a girl thought that females in society are generally expected to be fashionable, she would not be expected to adopt that trait unless she personally believed that girls should care about their appearance. However, it could also be argued that if a parti consider herself to be more similar to boys than the Self Perceptions Boys scale would measure. Since Signorella et al. (1993) found that important for measuring attitudes, it is not clear how this dilemma could have been avoided. Additionally, any discussion of the findings yielded by the free response stereotyping measure should consider the caveat that it i s possible that the most common
Idiosyncratic Beliefs and Gender Self Socialization 81 stereotypes listed by the participants were not actually the attributes that were the most important to their perceptions of each gender. Although several studies have had success open ended scale s (Eagly et al., 1992; Katz & Ksans nak, 1994; Miller et al., 2009), Lambert, Graham, and Fincham (2009) studied the concepts that adults associated with gratitude using a prototype approach (Rosch, 1975 ). The researchers surveyed 94 under graduate students over the course of two studies. In the first, participants were instructed to write down characteristics that came to mind when they thought of the word gratitude and to rate the valance of each feature on a 6 point scale; these features were then coded by two independent raters into attribute categories in an approach that was similar to that which was used in the current study. In the second study, 91 participants indicated the degree to which the features derived from the first study we re central to their concept of gratitude on an 8 point scale. The most commonly mentioned features in the first study were found to be only modestly correlated with these centrality ratings, suggesting that the importance that participants assigned to each feature was not strongly related to t he frequency with which that feature was mentioned. Additionally, Kearns and Fincham (2004) used a similar methodology with their prototype analysis of forgiveness, and found that the frequency of features participants associated with forgiveness was not significantly correlated with the overall importance that participants in a follow up study attributed to each feature. Although there is no way to measure whether the features that participants listed in the current st udy were truly central to their gender stereotypes, future researchers using open ended scales to assess stereotypes should conduct a follow up study to assess the centrality of these features for participants.
Idiosyncratic Beliefs and Gender Self Socialization 82 Finally, it is notable that the frequencies of the free response items (which appear in T ables 1 and 3 ) were calculated by tallying every appearance of each category in the data whether or not the same participant listed the same category multiple times These tallies were then divided by the numbe r of participants ( N =44) in order to depict the number of times, on average, that each participant listed each category. I n cases where the same participant listed multiple items which fell into the same category (such as that category w as tallied multiple times, and thus some of the percentages that are reported in the descriptive analyses exceed 100% In previous studies which have compiled prototypical features, such as those of Lambert et al. (2009) and Kearns and Fincham (2004), the frequencies of items are calculated such that each category can only be counted once for each participant. The latter approach is especially advantageous in situations in which a single participant listed variations of the same category many times, in whic h case a high frequency for that category would have been misleading. Indeed, the approach utilized in the current study of reporting the frequencies of the free response items may have been suscept ible to this issue. However, when the independent coders in the current study reviewed the descriptive data, they did not find that items were skewed such that a few participants listed the same category many times. Therefore, it was decided that counting the same category multiple t imes per participant, would have the benefit of yielding data which represented the overall prevalence of the category in the aggregated stereotype lists. With these considerations in mind, the content of the free response data may now be enumerated. Desc riptive Results of the Free Response Stereotyping Measure
Idiosyncratic Beliefs and Gender Self Socialization 83 Interestingly, none of the items included on the short scale of the COAT AM activity measure were included in any of the total 901 stereotypes generated by go horseback Bigler, 2002). While that scale is certainly not intended to be an exhaustive list of stereotypes, the fact that none of those items frequently occurred to the participants as response scales suggests that even if the participant ha COAT AM questionnaire, their feelings about the importance of that item in relation to gender roles was probably not very strong. The items which did appear on the free response scales g enerally indicated notions of gender egalitarianism. Overall, in contrast to accessibility of stereotype domains, an overwhelming proportion of the responses for both male and female stereotype targets were traits, and many others were activities. Though descriptions of appearance were prominent stereotypes about other girls, far more trait related categories emerged than were derived
Idiosyncratic Beliefs and Gender Self Socialization 84 by Miller et al. (2009 ), who categori groups into domains based on whether the stereotypes were masculine or feminine typed and whether they were related to traits, activity/toy preference, and appearance. In that study, interpersonal, occupati onal, and social role domains had to be excluded because less than 2% of the population had mentioned traits which could fall under those categories; however, participants in the study at hand frequently mentioned interpersonal traits (such as manners and bullying behaviors) and occupations appeared on the lists with some frequency as well, though social roles were rarely mentioned in either Miller In the current study, the most frequently stereotyped cate gory for both boys and occurring 134% of the time (when the frequency of categories was divided by the number of participants, N =44) on the lists of stereotypes about girls (meaning that most the lists of stereotypes about boys, more than double the seco nd most frequent and 52.2% on the boys. AM, similar items do appear on that scale, in lists provides partial support for the supposition that the high levels of gender egalitarian attitudes generated in
Idiosyncratic Beliefs and Gender Self Socialization 85 provide socially acceptable responses (given that the daycare environment stressed the importance of res isting stereotypes), but was indeed due to an internalized notion that boys and girls should share the same general qualities regardless of their sex. These findings were replicated in similar categories; for example, the most frequently occurring items o each occurring 16 times for stereotypes on the trait subscale of the COAT scale of the COAT participants in this study mentioned that particular sport). Items included in that category degree of gender typing that did occur, other items grouped into the athletic category for once), demonstrating a variety of ways that the participants perceived that their gender could be physically active. Many of the other cate gories which occurred frequently for none of these or similar traits are listed on the COAT AM. Interestingly, for stereotypes about boys, the c frequent (occurring 17 times), though that trait is classified on th e COAT
Idiosyncratic Beliefs and Gender Self Socialization 86 has been considered to be more community oriented and prone to care giving (Bem, 1981). This finding is consistent with previous findings that children hold more flexible beliefs about masculine it ems than feminine items (Liben & Bigler, 2002; Signorella et al., 1993). A surprising number category was more frequ ent for the list of stereotypes about boys (10 items) than it was applicable to the girls lists AM, several items relate to anti frequency of the anti bullying categories on the lists of stereotypes about boys could be interpreted to suggest that the participants stereotype boys as being rough in congruence with the masculine typed items on the COAT AM or as being gentle in congruence with the feminine typed items on the COAT AM. Ultimately, because the stereotypes of greatest inte rest to the research at hand are the prescriptive attitudes about how the participants believe each gender should be (Liben & Bigler, 2002; Signorella et al., 1993; Tobin et al., 2010), these findings can be interpreted to be congruent with the feminine ty ped items listed on the COAT AM. This interpretation is consistent with the prevalence (traditionally a feminine 9
Idiosyncratic Beliefs and Gender Self Socialization 87 AM as Not all of the free response stereotypes maintained this spirit of gender egalitarianism, however. For example, items related to appearance were l isted much more versus 18 times for boys). By comp arison, one item on the COAT ems in the occurred 7 times, and 5 items related to having nice hair and makeup. Mi scellaneous same spirit as the rest of the items in that category and connote fem ininity. By cont rast, for boys, the most common items related to appearance included twice). Idiosyncratic items on that list included (but were not limi a greater degree of flexibility for which the participants perceived that boys could appear acceptable in comparison to girls. Further analyses sugg ested that the girls had, at least to some extent, internalized the Feminist message of the after school care program that gir Smart, and B response stereotype lists.
Idiosyncratic Beliefs and Gender Self Socialization 88 (occurring 25 times for stereotypes about girls and 20 times for stereotypes about boys). fact that participants emphasized this item so strongly on their lists of stereotypes for girls leads to the assumption that they had either provided that item because it was a sal ient or socially acceptable response, or because they had truly internalized the culture of their after school care environment. While it is not possible to determine which motivation predicated the responses, items related to general academic success (gro uped in a boys than for girls (12 times for boys versus 6 times for girls), suggesting that to some degree participants had retained the cultural stereotype that boys should be more academically successful than girls, though the limited sample size and infrequency of these items makes it impossible to conclude so definitively. However, although a limited number of participants listed occupations as stereotypes (for bot h boys and girls), the 7 scientist, and veterinarian; each occurred 1 time) these responses illuminate the wide variety of career options that the participants viewed a s available to them. the possible confounding role of social desirability bias However, the sentiment of the term could be interpreted to mean either physically or emotionally robust, and both interpretations are present elsewhere on the stereotype lists physical strength is trength is reflected
Idiosyncratic Beliefs and Gender Self Socialization 89 again leading to the conclusion that the participants had taken the message of the after school program to heart at least to some de the stereotypes of girls, appearing as a category 12 times (compared to 6 times in the also reflected attitudes about stereotypes, some of which endorsed traditional stereotypes the participants were reflecting on the status of women in relation to men in society. One of the most striking findings of the free response stereotype questionnaires was the quantity of idiosync ratic items that children prescribed for each gender The could not be grouped with other items of the same meaning, occurred in 125% of lists for stereot ypes about girls and in 100% of the lists for stereotypes about boys, suggesting that every child on average considered at least one attribute important for each gender that none of the other participants had considered. This raises important questions abo ut the mechanisms used by psychologists to assess gender stereotyping, such as whether trait inventories provide sufficient information about the attributes that children believe are the most important for each gender to possess Although the gender
Idiosyncratic Beliefs and Gender Self Socialization 90 egalit arianism scores yielded by the free response measures were roughly equivalent to those yielded by the COAT AM, the former measures contained far more information about the specific domains which children associated with males and females, and illuminated t he similarities and differences that they identified between the two. Possible influence of the demographics of the sample. The participants in the study were indeed more egalitarian in their attitudes about gender than the participants in the study condu c ted by Liben and Bigler (2002) both quantitatively and qualitatively, meriting an evaluation of the differences between the sample in the latter experiment versus the sample in the study at hand. The sample in the current study was consid erably less homog enous wide age range (7 11 years old) participated, and all respondents were female and came from a wide variety of racial and socioeconomic backgrounds. Regarding the influence of age differences in the two samples, is notable that the authors of the COAT AM do specify that the scale has been tested successfully with children as young as 7, and the administration of the scale in this study followed the specific guidelines that the a uthors suggested for using the device to measure the attitudes of younger children, including displaying the scale with large lettering on alternating white and shaded lines to make the items easier for younger children to read and differentiate, reading e ach item aloud, and questions. Furthermore, if the participation of much younger children in the current ndings, it would be expected that the responses to the stereotype measures would have been less ega litarian in congruence with Sign analytic finding that elementary school
Idiosyncratic Beliefs and Gender Self Socialization 91 y flexible with age. Their research also found that there was a effect of sex on stereotype response measures such that girls gave significantly more egalitarian responses with age than did boys (though the inclusion of male participants in study influenced the greater proportion of egalitarian responses yielded in the study at hand. However, Signorella et al. (1993 ) also concluded stereotypes were influenced by their cultural environment which was supported by their finding traditional television watching and having a m other who worked outside of the home The participants in this study came from a wide variety of backgrounds; some came from families with extreme wealth, and many others were being raised by a single parent who worked full time to support their household. Signorella et al. (1993) did not assess the influence of ethnic heritage or socio economic status on stereotype s and few others have done so making it difficult to make inferences about the influence of the heterogeneity o f this sample on the findings. Ul timately, it was outside of the realm of analysis to take into account race or socioeconomic cl ass when analyzing the results; in fact, doing so would have been problematic because it would have required the researcher to make hasty assumptions about the e ffect of those variables Such an analysis would have nec essitated a reliance on stereotypes about gender relations in various racial and socioeconomic groups.
Idiosyncratic Beliefs and Gender Self Socialization 92 In spite of this limitation, the use of the qualitative data described in this section provides an excellent opportunity to assess the influence of a cultural background which all of the participants did share the daycare center where the research was conducted. With an explicit Feminist pedagogy, the center provides anti stereotype messaging to gi rls of all ages, including classes on media literacy which teach the children to identify and reject gender based stereotypes, as well as courses in traditionally male fields such as math, science, economic literacy, leadership, and p hysical education. Fur thermore, it repeated frequently to the students by instructors at the center. Therefore, the occurrence of those three words (or words with similar meaning) as well as the occurrence of stereotype based items which show an awareness and rejection of gender based stereotypes can serve as an indicator of the influence of the cultural environment on Self P erceptions Self percept ions were also measured using descriptive, free response data; however, it was outside of the scope of the current study to use the same categorization classification system of analysis used to assess the descriptive stereotype data. Instead, self percepti ons were analyzed quantitatively in relation to the degree to which they matched the stereotypes for girls and boys, respectively. This analysis was chosen assessments was in determining how similar ly they viewed themselves in comparison to their stereotypes of gender groups. However, lists was .30, suggesting that this approach may not have been ideal.
Idiosyncratic Beliefs and Gender Self Socialization 93 If the measure of self perc eptions utilized in this research was an accurate assessment of the construct it intended to measure, it was predicted that lowe r sco res on the self perception girl s scale and higher scores on the self perceptions boys scale would be related t o stronger fe elings of gender a typicality. Patterson (2012) found that gender non normative self perceptions of activity preferences were linked with lower of appropriate traits and behaviors for their gender could be predicted to be negatively related to feelings of gender contentedness based on a pattern of similar findings in the literature on gender identity (Carver et al., 2003, 2004; Menon, 2011). In opposition to these assu mptions, neither self perceptions relating to stereotypes about girls nor to those about boys were found to be related to typicality or contentedness. Therefore, t he most likely explanation for the lack of correlations between gender identity and self perc eptions found in this study was that the self perceptions measure was simply invalid. Limitations of Self Perceptions Scores This analysis is perhaps limited by the fact that self perceptions scores are empirically and conceptually inseparable from Free Response Gender Egalitarianism Scores. This approach was taken in the current research because the Gender Self Socialization Model eschews the practice of the researcher assigning participant responses to gender categories on the basis that it is empirica lly unsound to assume that gender schemata are universal and inter correlated ( Tobin et al., 2010; Spence, 1984 ). Thus, it was necessary to have a Self Perceptions measure which ulinity and femininity. The current method of analysis has the advantage of yielding numeric al quantitative data that are statistically comparable to the gender stereotyping and gender
Idiosyncratic Beliefs and Gender Self Socialization 94 erms of their own personal views about gender roles. On a similar note, the current study was limit ed by the fact that the categories ( or domain s) of the free response stereotype items were not taken into account in analyzing the quantitative results; th is analysis was not possible given the constrained resources of this investigation, but future researche rs should be sure to conduct such an investigation. Future researchers should also their ow n stereotypic views while still ensuring that the measure is methodologically sound. It is unfortunate that analyzing the domain content of the self perceptions in this study was outside of the bounds of the current analysis, as it would have been interes ting those in the stereotypes scales. For example, one might expect that girls who have more appearance related stereotypes for their gender would also hold themselv es to rigid beauty standards and that this relationship would be heighte ne d by feelings of pressure to conform to gender stereotypes and mitigated by feelings of contentedness. It would certainly be fruitful to compare how all of the different domains of s tereotypes for the same and opposite gender interact with self perceptions and gender identity in this way. Finally, it is possible that the Self Perceptions differences in the instructions on the Self Perceptions and Stereotyping scales influenced the res ults. The wordings on the tter were more idealistic than were the
Idiosyncratic Beliefs and Gender Self Socialization 95 of self discrepancies to Gender Identity in order to argue that both gender typicality and felt pressure to conform are ought se lf discrepancies, which are associated with the failure to meet a social standard, as opposed to ideal self discrepancies, which related to how the individual intrinsically thought she should be. It would seem that the Self Perceptions scores, which measur self discrepancy. Therefore, it is possible that the Self Perceptions scores did tap into the correct construct. Gender Identity Surprisingly, felt pressure to conform to gender stereotypes had a much higher mean (2.02 on a scale of 1 4) than has been found in previous research, in which the mean of felt pressure has generally been around 1.6 (e.g. Egan & Perry, 20 01; Yunger et al., 2004). It would have been expected that, given the anti stereotyping messaging of the day care facility felt pressure would have been greatly reduced. Perhaps children who do feel pressure to conform to gender stereotypes from their fam ily and peers are more stereotype education. It is also possible that the inclusion of younger children increased the mean of self perceptions scores; although very little research on gender iden tity has been done on children as ( 2004 ) 2 year longitudinal study found that girls (but not boys) in 3 rd 7 th grades had lower self perceptions scores as they aged. This explanation is likely, as the means of typicalit y and contentedness were lower than are generally found in research, and Yunger et al. (2004) also found that for both sexes, scores on those measures increased with age. Yunger et al. (2004) also found that gender
Idiosyncratic Beliefs and Gender Self Socialization 96 identity research has tended to find that boys have higher results on all three measures and older children score higher on typicality but lower on felt pressure, so these scores might have been higher in a sample that included male participants. However, Patterson (2012) investigated predictive effect of age and gender (in participants 6 12 years old) on the correlation between typicality and gender typed interests and stereotype beliefs, and found that neither participant variable significantly changed those relationships. So, if age did influen could have changed the correlations between typicality or, by logical (if not empirically supported) extension, contentedness and felt pressure and self perceptions or stereotyp e beliefs. Limitations of the Gender Identity S cores. One minor limitation for the gender (2001) scales is quite confusing, and although the scores these measures yie lded were comparable to other findings in the literature, in practice the participants did have a very difficult time interpreting the questions and required help from the adult instructors to answer them, which may have influenced their responses. Future researchers should use a more simplified version of the scale when assessing these constructs. Tests of the Gender Self Socialization Model Hypotheses The goal of the current study was to assess the interaction between stereotype beliefs, self perception s, and gender identity. Thus, Pearson correlations were used to compare eac h pair of variables (see Table 5 for a full list) and multiple regression analyses were performed with one variable from each construct serving as a criterion
Idiosyncratic Beliefs and Gender Self Socialization 97 variable and two other variables (one from each construct) as the predictor variables. Overall, the findings provide limited support for the Gender Self (Tobin et al., 2010) three hypotheses. Limitations of statistical analyses. Prior to reviewing the fi ndings of these analyses, it should be noted that a total number of 36 regression analyses were conducted to test the hypotheses of the Gender Self Socialization Model (Tobin et al., 2010). Therefore, the type 1 error rate for these analyses was likely inf lated; if a .05 alpha level was assumed, it could be expected that 2 in 40 analyses would occur by chance alone. It was necessary to perform so many analyses because this study was interested in the interactive properties of all three constructs of the GSS M, yet one must consider the risk of alpha inflation when interpreting these results. Future researchers can prevent this dilemma by combining gender typicality and gender contentedness, which were found to be highly correlated and had similar interactive properties with stereotypes and self hypotheses. Due to constraints in resources, it was also outside of the realm of the current research to assess the interactions betw een each pair of variables in each regression analysis. However, future researchers should extend the literature by considering the interactions as well as the general multiple regression models in order to eses. Stereotype emulation hypothesis. is perhaps the most focal and detailed prediction made by the model, and the authors explain that this is due to the fact that the hypothesis attempts to provide a concrete answer to one of the most central and studied questions in the literature on gender
Idiosyncratic Beliefs and Gender Self Socialization 98 development: why do children come to behave in sex typed ways? Thus, in the GSSM, self perceptions on r should be serve as the depende nt variable in this hypothesis. Correlations. I t was expected that the similarity between self perceptions with stereotypes for girls would be positively correlated with gender typicality and contentedness, while t he similarity between self perceptions and stereotypes for boys would be negatively correlated with these gender identity variables. This prediction was that for girls, gender typicality and contentedness were positively correlated with communal (stereotypically feminine) and female typed activities, and negatively correlated with agentic (stereotypically male) t raits and male typed activities Yet, this was not the case no correlation was found between self pe rceptions and gender identity variables. If this finding was due to the fact study, it would be expected that the stereotyping scores would explain the relationship betw een self perceptions and gender identity; thus, it is most likely the case that the self perceptions measure used in this study was not accurate at assessing the intended construct. Multiple regression analyses. The results of the multiple regression anal yses in which Self Perceptions Girls and Self Perceptions Boys served as separate criterion variables found that generally, higher gender egalitarianism scores were predictive of higher Self Perceptions Boys scores. It is surprising that the COAT (Liben & Bigler, 2002) was found to be more predictive of the Self Perceptions Boys variable than was the Free Response Gender Egalitarianism measure in these regression models, given that the
Idiosyncratic Beliefs and Gender Self Socialization 99 Free Response measure was correlated with Self Perceptions Boys and the COAT was not. These findings may be due to the fact that most participants achieved very high COAT scores and very l ow Self Perceptions Boys scores. However, the variability of both stereotyping measures was controlled by conducting logarithm transformatio ns on the criterion variables, and Self Perceptions Girls scores which had a very similar mean, obtained range, and correlation with the Self Perceptions Boys scores were not predicted by any of the regression models, suggesting that the two Self Perceptions scores may have indeed successfully measured different constructs. Additionally, when the Free Response Gender Egalitarianism Score and Contentedness were used as pred ictors of Self Perceptions Boys, it was found that higher egalitarianism scores yielded similar results to the COAT, suggesting that both measures of stereotyping assessed the same construct (though this finding would be more persuasive if all three of the regression models which included the Free Response variable had been significantly predictive.) The gender identity variable s did not have any significant e ffects on self perceptions in the regression analyses. However, because the interactions between g ender identity and stereotyping measures were not tested (this will be discussed further in the limitations section) it is difficult to know the role that the gender identity variables played in the regression models. The fact that COAT scores and self per ceptions scores were uncorrelated, but that COAT scores were predictive of Self Perceptions Boys scores on models which included the interaction variables does suggest support for the stereotype emulation hypothesis. It makes sense that higher egalitarian ism would predict the degree to which girls would perceive themselves as being similar to how they think that boys should be. This
Idiosyncratic Beliefs and Gender Self Socialization 100 pathway model (2002), which predicts that children with rig id gender stereotypes will feel less able to express or adopt non traditional behaviors. In support of their hypothesis Liben and Bigler (2002) found that girls (but not boys) who had more egalitarian stereotypes also had higher ratings in interest in masc uli ne typed activities. Similar results have been found in previous research. Patterson (2012) found that children who had higher stereotype flexibility on the COAT tended to view themselves as a typical of their gender. Leaper and Van (2008) found that men who had low typicality and high egalitarianism were more interested in feminine typed academic fields. Martin and Dinella (2011) found that girls who had more inclusive stereotypes tended to have more interest in masculine activities Finally, Leaper et a l. (2011) found that girls with more egalitarian stereotypes were more motivated to achieve in masculine typed academic fields. Al t hough the influence of gender identity was not found to be significant in the multiple regression analyses performed in the current study, evidence in the previous literature suggest s that higher Typicality scores should have been predictive of higher Self Perceptions Girls scores and lower Self Perceptions Boys scores, while low Typicality scores should have had the opposite e ffect. While the influence of gender identity on the adoption of gender normative traits has only been tested to a limited degree in the literature, Egan and Perry (2001) found that for girls, interest in gender typical activities was, in general, associat ed with higher typicality and contentedness (but not with felt pre ssure), while the adoption of a typical traits was associated with lower felt pressure. The results at hand are also incongruent with the findings of Leaper and V an (2008) and Patterson (2012 ) that typica lity and contentedness were significantly
Idiosyncratic Beliefs and Gender Self Socialization 101 as sociated with the adoption of a typical traits. Thus, contentedness could not be expected to influence t he adoption of self perceived a typical traits, lower typicality could be expected to play a low to moderate role in predicting the adoption of a typical traits, and lower felt pressure could be expected to predict a h igher association with gender a typical traits. In terms of the potential interactions between egalitarianism and gender identity on self perceptions, then, it could be expected that girls who have rigid stereotype beliefs and feel typical of their gender and content with their gender role, as well as a high level of pressure to conform to gender stereotypes should not feel partic ular moti vation to adopt gender a typical traits because they would not feel the need to deviate from their gender role. On the other hand, girls who have rigid gender stereotypes but lower typicality and contentedness might feel conflicted about whether or not to deviate from their gender role (2001) findings and the GSSM hypotheses, they would experience negative mental health outcomes and perhaps might be more motivated to do so if they experienced less societal pre ssure to conform. Girls with egalitarian stereotypes could be expected to f eel more at ease adopting gender a typical traits, particularly if they ha d low typicality and contentedness (in which they would feel motivated to deviate from their gender role) and low felt pressure to conform (in which case they would feel more capable of deviating). However, the current findings cannot support these conjectures since the influence o f gender identity variables on self p ercepti ons was not significant. Identity Construction Hypothesis. identity construction hypothesis suggests that self perceptions and stereotypes should, together, influence gender identity. This hypothesis predicts that t he more individuals
Idiosyncratic Beliefs and Gender Self Socialization 102 believe they possess an attribute that is an important characteristic of their gender group, the more typical and content they will feel with their gender (the GSSM does not make a directional hypothesis about felt pressure) (Tobin et al. 2010). Correlations Due to the findings of Egan and Perry (2001), Patterson (2012), and Leaper and Van (2008) that gender identity variables were associated with egalitarian stereotyping, it was expected that higher typicality and contentedness would be related to higher Self Perceptions Girls scores and lower Self Perceptions Boys scores. One potential explanation for this hypothesized relationship could be that girls who had more gender typed traits recognized the congruence between their behavior an d their gender role. However, these relationships were not found in the current study. Girls with more egalitarian stereotypes were expected to have lower gender typicality, as was found by Patterson (2012), yet this relationship was not found. The relati onship between contentedness and stereotyping has not been tested in the previous research, so no hypothesis was made about that relationship and no significance between the two was found. Interestingly, felt p ressure was found to be negatively related t o egalitarian COAT scores, suggesting that perhaps girls who have more egalitarian stereotypes are less aware of or less influenced by pressure from their peers and family to conform to stereotypes. No known research has studied both variables, and the Gen der Self Socialization Model does not make predictions about this relationship. However, it makes sense that girls who have more egalitarian beliefs might hold those views because their social environment does not place value on traditional gender roles, a nd thus their feelings of pressure from peers and family would also be decreased.
Idiosyncratic Beliefs and Gender Self Socialization 103 their stereotype attitudes and feelings of pressure to conform to those stereotypes, perhaps s pressure ratings and stereotype beliefs. Or, future researchers could employ a longitudinal young age, and then sending participants in the experimental condition to an after school program (such as the one in the current study) that teaches children to resist gender stereotyping and measuring their change in felt pressure with age in comparison to that of a control group that attends a more traditional after school care program. Multiple regression analyses. Only one model approached significance; COAT scores and Self Perceptions Girls scores were found to be close to predicting Felt Pre ssure. In this model, only the e ffect of COAT scores was significant, a relationship which makes sense given the correlation between the two, and lends very limited support to the identity construction hypothesis. Theorists have a long history of asserting that felt pressure is a cause of rigid stereotype beliefs (i.e. Bem, 1981), but it is also reasonable to assume that children who have more rigid standards for each gender to begin with will place more pressure on themselves to conform to those standards. Howev er, no directional or causal inference can be made from th ese data. Stereotype construction hypothesis The Gender Self Socialization Model their gender identity and self perceptions. Thus, in 12 multiple regression analyses, the COAT and Free Response Gender Egalitarianism scores served separately as criterion variables.
Idiosyncratic Beliefs and Gender Self Socialization 104 Correlations. The significant negative relationship between higher gender egalitarian scores on the COAT (Liben & Bigler, 2002) and felt pressure to conform to gender stereotypes was also interesting in consideration of the stereotype construction stereotype beliefs with felt pressure to conform. Hypothetically, this relationship could have been expected to go in the opposite direction; it could be possible that children who feel pressure to conform to rigid stereotypes develop more egalitarian ste reotypes in order to dec rease their stress. However, the results of the current study also make sense, and could mean that children who feel less pressure from their parents and peers to conform simply learn to place less value on rigid gender roles, or th at children behaviors in order to fit in with their gender. To test either hypothesis would require knowledge of a directional relationship between these variable s with self perceptions; thus, future research should test this model in comparison to a more effective self perceptions measure. Multiple regression analyses. One significant finding of this study which can be considered to offer tentative support to the stereotype construction hypothesis of the Gender Self Socialization Model is that feelings of Gender Typicality and Self Perceptions Boys were, together, predictive of the Free Response Gender Egalitarianism score. Only the effect of Self Perceptions Boys had a significant, positive influence on the increases in egalitarianism, consistent with the theoretical models suggested by Tobin et al. (2010) that children would change their stereotype beliefs to match their self perceptions in order to attain cogniti ve constancy. Although this analysis does not include
Idiosyncratic Beliefs and Gender Self Socialization 105 information about the causal or interactive rel ationship about these variables and the influence of typicality was insignificant, Patterson (2012) found that children with lower typicality also had more egalitarian stereotypes, and suggested as a possible interpretation that it was possible that children who perceiv ed themselves as being gender a typical widened their stereotype views in order to maintain their self esteem. When Felt Pressure was include d with Self Perceptions Boys in a separate regression analysis with Free Response Gender Egalitarianism scores as the criterion, the results were found to be approaching significance, but again only th e influence of Self Perceptions Boys was significant. F inally, when Self Perceptions Boys and Felt Pressure served as predictors and COAT scores served as the criterion variable, neither predictor variable was independently influential. However, these results cannot be compared to previous data because none ex ists on the relationship between Felt Pressure and COAT scores. Theoretically, if causation had been found, these results could be interpreted to mean that children with interests that they associated with the opposite sex might form more egalitarian belie fs to justify these behaviors and that either (a) this effect would be increased by felt pressure because they would feel a stronger amount of dissonance or (b) felt pressure would decrease this effect because children who feel more pressure from significa nt social groups to adhere to rigid stereotypes would place greater value on those stereotypes (in which poorer mental health outcomes would be expected). These predictions should be tested by future researchers stereotype beliefs are malleable based on a change in felt pressure or interest in cross gender activities If indeed this is the case, it is possible that by educating parents to have more egalitarian worldviews or by encouraging children to engage in non traditional
Idiosyncratic Beliefs and Gender Self Socialization 106 activities, stereo type rigidity could be reduced. This is a powerful possibility indeed, as research has found that egalitarian gender related beliefs are associated with a number of positive outcomes, including futur e higher academic motivation (Leaper and Van, 2008; Leaper et al., 2011), and adjustment (DiDonato & Berenbaum, 2011). Conclusion Overall, limited but notable support was found for the Gender Self Socialization e theory that stereotypes, self perceptions, and gender identity are independent but interactive constructs. Only 5 of 36 regression analyses were found to be significant, yet those which were significant generally multiplicative properties of two of the primary constructs delineated by the GSSM would be predictive of the third. Though proposed only recently, this model has rece ived too little attention in psychological literature. Because research has consistently Perry, 2001; Yunger et al., 2004) and future aspirations (Leaper & Van, 2008; Leaper et al., 2011) are linked to these cognitions of the potential by which related stress can be reduced it is critical (Tobin et al., 2010) hypotheses. Counter to the hypothesis that a free response stereotyping measure would be more highly corre analyses found that neither a traditional stereotyping measure n or a free response stereotyping measure were more or less predictive of ge nder related outcomes; and,
Idiosyncratic Beliefs and Gender Self Socialization 107 generally, the t wo measures yielded comparable gender egalitarianism scores. However the free gender related beliefs, much of which deviated from classical understandings of gender roles. Participant s responses demonstrated the complexity of their values, beliefs, and experiences. Some responses reflected traditional societal ideologies about gender, such ed from hegemonic gender roles. no n reflected the items that are included on traditional trait inventories such as the COAT AM (Liben and Bigler, 2002). While this free response instrument was by no means perfect, the findings at hand should compel future researchers to continue th e practice of In recent years, many have made a similar call unique attitudes and beliefs about gender roles into gender development research (Egan & Perry, 2001; Corby et al., 2007; Miller et al., 2009; Perry & Pauletti, 2011; Tobin et al., 2010). Th e se researchers argue, as th e findings of the current study suggest, that traditional measures of stereotyping limit the ways in which gender related cognitions can be understood. However, it is also important to note that the paradigmatic practice of using trait inventories to infer related beliefs and attitudes is one that not only limits our knowledge of gender development, it is an approach to research that denies children the opportunity to have their own views represented in
Idiosyncratic Beliefs and Gender Self Socialization 108 research that purports to speak for them This methodology perpetuates the problematic exclusion of young people from the production of knowledge that has pervaded scientific research throughout history (Prillenstensky, 2010). Psychologists in particular must recognize the pos ition of power they hold and ensure that they do not impose their own beliefs in research that purports to measure the attitudes of marginalized groups. As Checkoway (2011) argued the fundamental right of children to express their own values in scientifi c research that affects them Declaration of the Rights of the Child, which children are full fledged persons who have the right to express their views in all matters affecting them and requires thos It is imperative that scholars of developmental psychology treat children not as passive subjects, but instead empower youth to have their voices heard.
Idiosyncratic Beliefs and Gender Self Socialization 109 References Ablow D. K. (2011, April 11). J. Crew Plants the Seeds for Gender Identity. Fox News Retrieved April 16, 2013, from http://www.foxnews.com/health/2011/04/11 /j crew plants seeds gender iden tity/ Arthur, A. E., Bigler, R. S., Liben, L. S., Gelman, S. A., & Ruble, D. N. (2008). Gender stereotyping and prejudice in young children: A developmental intergroup perspective. I n S. R. Levy, M. Killen (Eds.) Intergroup Attitudes and Relations in Chil dhood through A dulthoo d, 66 86. New York, NY US: Oxford University Press. Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development, 55 (4), 1456 1464. doi:10.2307/1130015 Bem, S. (1981). Gender schema theory: A cognitive a ccount of sex typing. Psychological Review, 88 354 364.1981 25685 00110.1037/0033 295X.88.4.354. 10.1037/0033 295X.88.4.354 Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting & Clinical Psychology, 42 155 162.1974 27631 00110.1037/h0036215 Bigler, R. S. (1995). The role of classification skill in moderating environmental influences on children's gender stereotyping: A study of the functional use of gender in the classroom. Child Development, 66 1072 1087.1996 09661 00110 .2307/1131799. 10.2307/1131799
Idiosyncratic Beliefs and Gender Self Socialization 110 Bourdieu, P. (1977) Outline of a Theory of Practice trans. R. Nice. Cambridge: Cambridge University Press. Brinthaupt, T. M., & Lipka, R. P. (1985). Developmental differences in self concept and self esteem among kindergarte n through twelfth grade students. Child Study Journal, 15 (3), 207 221. Brown, E. (2011, April 8). J.CREW Pushes Transgendered Child Propaganda | Media Research Center. Media Research Center. Retrieved from http://www.mrc.org/articles/jcrew pushes transgend ered child propaganda Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and differentiation. Psychological Review, 106 676 713.1999 11924 00210.1037/0033 295X.106.4.676. 10.1037/0033 295X.106.4.676 Carver, P. R., Egan, S. K ., & Perry, D. G. (2004). Children who question their heterosexuality. Developmental Psychology, 40 (1), 43 53. doi:10.1037/0012 1622.214.171.124 Carver, P. R., Yunger, J. L., & Perry, D. G. (2003). Gender identity and adjustment in middle childhood. Sex Roles, 49 (3 4), 95 109. doi:10.1023/A:1024423012063 Checkoway, B. (2011). What is youth participation?. Children And Youth Services Review 33 (2), 340 345. doi:10.1016/j.childyouth.2010.09.017 Corby, B. C., Hodges, E. E., & Perry, D. G. (2007). Gender ide ntity a nd adjustment in black, h ispanic, and white preadolescents. Developmental Psychology, 43 (1), 261 266. doi:10.1037/0012 16126.96.36.1991
Idiosyncratic Beliefs and Gender Self Socialization 111 Davis Kean, P. E., Jager, J., & Collins, W. (2009). The self in action: An emerging link between self beliefs and behaviors in middle childhood. Child Development Perspectives, 3 (3), 184 188. doi:10.1111/j.1750 8606.2009.00104.x DiDonato, M. D., & Berenbaum, S. A. (2011). The benefits and drawbacks of gender typing: How different dimensions are related to psychological adjustm ent. Archives Of Sexual Behavior, 40 (2), 457 463. doi:10.1007/s10508 010 9620 5 Edelbrock, C. S., & Sugawara, A. I. (1978). Acquisition of sex typed preferences in preschool aged children. Developmental Psychology, 14 (6), 614 623. doi:10.1037/0012 1649.14. 6.614 Egan, S. K., & Perry, D. G. (2001). Gender identity: A multidimensional analysis with implications for psychosocial adjustment Developmental Psychology, 37 (4), 451 463. doi:10.1037/0012 16188.8.131.521 Eagly, A. H., Mladinic, A. and Otto, S. (1991), A re women evaluated more favorably than men ? Psychology of Women Quarterly, 15 203 216. doi: 10.1111/j.1471 6402.1991.tb00792.x Fausto Sterling, A. (1997). Beyond difference: A biologist's perspective. Journal o f Social Issues, 53 (2), 233 258. doi:10.1111/ 0022 4537.00016 Gara, M. A., & Rosenberg, S. (1979). The identification of persons as supersets and subsets in free response personality descriptions. Journal Of Personality And Social Psychology 37 (12), 2161 2170. doi:10.1037/0022 35184.108.40.2061
Idiosyncratic Beliefs and Gender Self Socialization 112 Greenwald, A. G., Bahaji, M. R., Rudman, L. A., Farnham, S. D., Nosek, B. A., & Mellott, D. S. (2002). A unified theory of implicit attitudes, stereotypes, self esteem, and self concept. Psychological Review, 109 3 25. doi: 10.1037//0033 295X.220.127.116.112 00351 00110.1037/0033 295X.109.1.3. 10.1037/0033 295X.109.1.3 Harter, S. (1988). Developmental processes in the construction of the self. In T. D. Yawkey, J. E. Johnson (Eds.), Integrative Processes and Socialization: Early to Middle Childhood 45 78 H ill sd ale, NJ England: Lawrence Erlbaum Associates, Inc. Harter, S. (1999). The Construction of the Self: A Developmental P erspective. New York, NY US: Guilford Press. Harter, S. (2003). The development of self representations during childhood and adolescence. In M. R. Leary, J. Tangney (Eds.) Handbook of Self and I dentity (pp. 610 642). New York, NY US: Guilford Press. Higgins, E. T., Klein, R. L., & Strauman, T. J. (1987). Self discrepancies: Distinguishing among self states, self state conflicts, and emotional vulnerabilities. In K Yardley, & T. Honess (Eds.), 173 186 Oxford England: John Wiley Sons. Kohlberg, L. (1966). A cognitive role concepts and attitudes. In E. E. Maccoby (Ed.), The Development of S ex Differences (pp. 82 173). Stanford, CA: Stanford University Press.
Idiosyncratic Beliefs and Gender Self Socialization 113 Katz, P. A., & Ksansnak, K. R. (1994). Developmental aspects of gender role flexibility and traditionality in middle childhood and adolescence. Developmental Psychology, 30 272 282.1994 25045 00110.1037/0012 1618.104.22.1682. 10.1037/0012 1622.214.171.1242 Kearns, J. N. & Fincham, F.D. (2004). A prototype analysis of forgiveness. Personality and Social Psychology Bulletin, 30 838 855. Lambert, N.M., Graham, S.M., & Fincham, F.D. ( 2009). A prototype analysis of gratitude: Varieties of gratitude experiences Personality and Social Psychology B ulletin, 35 1193 1207. Leaper, C., & Bigler, R. S. (2011). Gender. In M. K.Underwood & L. H.Rosen (Eds.), Social D evelopment (pp. 289 315). N ew York: Guilford.2011 26924 012. Leaper, C ., Farkas, T., & Brown, C. (2011 related beliefs in relation to their motivation in mat h/science and English. Journal o f Youth a nd Adolescence, 41 (3), 268 282. doi:10.100 7/s10964 011 9693 z Leaper, C., & Van, S. R. (2008). Masculine ideology, covert sexism, and perceived gender typicality in relation to young men's academic motivation and choices in college. Psychology of Men & Masculinity, 9 139 153. doi:10.1037/1524 922 0.9.3.139.2008 09203 00310.1037/1524 9126.96.36.199. 10.1037/1524 9188.8.131.52
Idiosyncratic Beliefs and Gender Self Socialization 114 Levy, G. D., Barth J.M., Zimmerman, B.J. (1998). Associations among cognitive and The Journal of Genetic Psychology 15 9 (1), 121 126 Liben, L. S., & Bigler, R. S. (2002). The developmental course of gender differentiation: Conceptualizing, measuring, and evaluating constructs and pathways. Monographs of the Society for Research in Child Development, 67 (2, Serial No. 269). doi: 10.1111/1540 5834.t01 1 00187.2002 08295 00110.1111/1540 5834.t01 1 00187. 10.1111/1540 5834.t01 1 00187. Lorber, J. (2000). Using gender to undo gender: A feminist degendering movement. Feminist Theory, 1 (1), 79 95. doi:10.1177/14647000022229074 Lu eptow, L. B., Garovich Szabo, L., & Lueptow, M. B. (2001). Social change and the persistence of sex typing: 1974 1997. Social Forces, 80 (1), 1 36. doi:10.1353/sof.2001.0077 Lyons, Jenna. (2011, April 8). Saturday with Jenna. JCrew Style Guide 2011 (4). Retr ieved from http://thenewcivilrightsmovement.com/wp content/uploads/2011/04/J Crew Ad1.jpg Martin, C. (2000). Cognitive theories of gender development. In T. Eckes, H. M. Trautner (Eds.) The developmental social psychology of gender (pp. 91 121). Mahwah, NJ US: Lawrence Erlbaum Associates Publishers. Martin, C., Eisenbud, L., & Rose, H. (1995). Children's gender based reasoning about toys. Child Development, 66 (5), 1453 1471. doi:10.2307/1131657
Idiosyncratic Beliefs and Gender Self Socialization 115 Martin, C. L., & Dinella, L. M. (2011). Congruence between ge nder stereotypes and activity preference in self identified tomboys and non tomboys. Archives of Sexual Behavior, 41, 599 610. doi:10.1007/s10508 011 9786 5.2012 11500 01310.1007/s10508 011 9786 5. 10.1007/s10508 011 9786 5. J ournal Citation Maglaty, J. (2011, April 8). When Did Girls Start Wearing Pink?. Smithsonian Magazine. Retrieved April 16, 2013, from http://www.smithsonianmag.com/arts culture/When Did G irls Start Wearing Pink.html Martin, C. L., & Halverson, C. F. (1981). A schematic processing model of sex typing and stereotyping in children. Child Development, 52 (4), 1119 1134. doi:10.2307/1129498 Martin, C. L., & Halverson, C. F. (1983). Gender consta ncy: A methodological and theoretical analysis. Sex Roles, 9 (7), 775 790. doi:10.1007/BF00290030 Master, A., Markman, E. M., & Dweck, C. S. (2012). Thinking in categories or along a Child Development 83 (4), 1145 1163. doi:10.1111/j.1467 8624.2012.01774.x Menon, M. (2011). Does felt gender compatibility mediate influences of self perceived gender nonconformity on early adolescents' psychosocial adjustment? Child Development, 82 1152 1162. doi:10.1111 /j.1467 8624.2011.01601.x.2011 14715 01010.1111/j.1467 8624.2011.01601.x. 10.1111/j.1467 8624.2011.01601.x
Idiosyncratic Beliefs and Gender Self Socialization 116 Miller, C., Lurye, L. E., Zosuls, K. M., & Ruble, D. N. (2009). Accessibility of gender stereotype domains: Developmental and gender differences in children. Sex Roles, 60 (11 12), 870 881. doi:10.1007/s11199 009 9584 x Montemayor, R., & Eisen, M. (1977). The development of self conceptions from childhood to adolescence. Developmental Psychology, 13 (4), 314 319. doi:10.1037/0012 16184.108.40.2064 Patterso n, M. M. (2012). Self perceived gender typicality, gender typed attributes, and gender stereotype endorsement in elementary school aged children. Sex Roles, 67 (7 8), 422 434. doi:10.1007/s11199 012 0184 9 Perry, D. G., & Pauletti, R. E. (2011). Gender and adolescent development. Journal Of Research On Adolescence, 21 (1), 61 74. doi:10.1111/j.1532 7795.2010.00715.x Prilleltensky, I. (2010). Child wellness and social inclusion: Values for action. American Journal Of Community Psychology 46 (1 2), 238 249. doi:10.1007/s10464 010 9318 9 Signorella, M. L. (1999). Multidimensionality of gender schemas: Implications for the development of gender related characteristics. In W. B.Swann, J. H.Langlois & L. A.Gilbert (Eds.), Sexism and stereotypes in modern society: The gender science of Janet T. Spence (pp. 107 126). Washington, DC: American Psychological Association.1998 06496 00510.1037/10277 005. 10.1037/10277 005 Signorella, M. L., Bigler, R. S., & Liben, L. S. (1993). Developmental differences in children's gen der schemata about others: A meta analytic review. Developmental
Idiosyncratic Beliefs and Gender Self Socialization 117 Review, 13 147 183.1993 37161 00110.1006/drev.1993.1007. 10.1006/drev.1993.1007 Spence, J. T. (1984). Masculinity, femininity and gender related traits: A conceptual analysis and critique o f current research. In B. A.Maher & W. B.Maher (Eds.), Progress in experimental personality research (Vol. 13, pp. 1 97). New York: Academic. Spence, J. T. (1993). Gender related traits and gender ideology: Evidence for a multifactorial theory. Journal of Personality & Social Psychology, 64 624 635.1993 25426 00110.1037/0022 35220.127.116.114. 10.1037/0022 3518.104.22.1684 Spence, J. T., & Hall, S. K. (1996). Children's gender related self perceptions, activity preferences, and occupational stereotypes: A test of three models of gender constructs. Sex Roles, 35 659 692.1997 07897 00110.1007/BF01544086. 10.1007/BF01544086 Tobin, D. D., Menon, M., Menon, M., Spatta, B. C., Hodges, E. E., & Perry, D. G. (2010). The intrapsychics of gender: A model of self socializat ion Psychological Review, 117 (2), 601 622. doi:10.1037/a0018936 Images. The Society Pages. Retrieved April 16, 2013, from http:/ /thesocietypages.org/ socimages/2011 Whitley, B. E., Jr. & Kite, M. E. (2009). The psychology of prejudice and discrimination (2nd ed.). Belmont, CA: Thomson Wadsworth.
Idiosyncratic Beliefs and Gender Self Socialization 118 Yu, L., & Xie, D. (2010). Multidimensional gender identity and psychological adjustment in middle childhood: A study in China. Sex Roles, 62 100 113. doi:10.1007/s11199 009 9709 2.2010 01035 00910.1007/s11199 009 9709 2. 10.1007/s11199 009 9709 2 Yunger, J. L., Carver, P. R., & Perry, D. G. (2004). Does Gender Identity Influence Childr en's Psychological Well Being? Developmental Psychology, 40 (4), 572 582. doi:10.1037/0012 1622.214.171.1242 Zosuls, K. M., Miller, C. F., Ruble, D. N., Martin, C. L., & Fabes, R. A. (2011). Gender development research in Sex Roles: Historical trends and future directions. Sex Roles, 64 826 842. doi:10.1007/s11199 010 9902 3.2011 11483 00410.1007/s11199 010 9902 3. 10.1007/s11199 010 9902 3
Idiosyncratic Beliefs and Gender Self Socialization 119 Table 1 Prescriptive Stereotypes Abo ut Girls by Category (In Order o f % Participants ) List of categories for the prescriptive stereotypes about girls generated by participants in response with an average of 10.61 ( SD =3.17) per participant. Cat egories are listed by the percentage of participants who listed items in each category (number of items in each category divided by N =44), and the frequency of the items appearing in each category is indicated in bold NOTE: Many participants listed multiple items which fell into the same category; for example, a Therefore, percentages may exceed 100%. For a full list of the i tems in each category, see Table 3. Nice (59) 134% MISC (55) 125% Fun (27) 61.36% Smart (25) 56.8% Friendly (17) 38.63% Attractive (14) 31.82% Happy (14) 31.82% Bold (13) 29.54% Helpful (13) 29.54% Respectful (12) 27.27% Strong (12) 27.27% Caring (11) 25% Cool (11) 25% Good (11) 25% Playful (11) 25% Athletic gen (10) 22.72% Creative (8) 18.18% Generous (7) 15.9% Fashionable (7) 15.9% Independent (7) 15.9% Occupation (7) 15.9% Responsible (7) 15.9% Unique (7) 15.9% Anti bullying (6) 13.63% Healthy (6) 13.63% Good manners (6) 13.63% Honest (6) 13.63% Thankful (6) 13.63% Work hard (5) 11.36% Non judgmental (4) 9.09% Brave (4) 9.09% Drawing (4) 9.09% Outgoing (4) 9.09% Academic STEM (3) 6.8% Clean (3) 6.8% Gentle (3) 6.8% Loyal (3) 6.8% Makeup (3) 6.8% OMIT (3) 6.8% Sassy (3) 6.8% Successful (3) .6.8% Trusting (3) 6.8% Weird (3) 6.8% Academic general (2) 4.5% Careful (2) 4.5% Confident (2) 4.5% Dance (2) 4.5% Do the right thing (2) 4.5% Exciting (2) 4.5% Feminine hair (2) 4.5% Flexible (2) 4.5% Hygiene (2) 4.5% Ladylike (2) 4.5% Loves pets (2) 4.5% Listen (2) 4.5% Sing (2) 4.5% Talkative (2) 4.5%
Idiosyncratic Beliefs and Gender Self Socialization 120 Table 2 Prescriptive Stereotypes About Girls By Frequency And Category This list contains all of the items 10.61 ( SD =3.17) per participant. Items are grouped by categories generated by two independent rate rs and then listed in order of the frequency of the category. Bold indicates a category. Parentheses indicate frequency, and the frequencies of each category are the sum of the frequencies of each item in that category. Categories listed in blue are not ho mogenous in meaning, contain subcategories, and are not included in the descriptive analysis; they are listed here to be illustrative of meaning. Items listed in red are too similar to other features in the category to be miscellaneous, but too different f rom other items in the category to c onsider the category homogenous. _____________________________________________________ _________________________ Nice (59) Nice (32) Kind (17) Sweet (4) Be a nice girl (1) Agreeable (1) Be nice (1) Be nice to the things around them (1) Always be nice (1) Kindful (sic) (1) Appearance (32) Attractive (14) Pretty (10) Cute (3) Beautiful (1) Fashionable (7) Fashionable (4) Dress however you want (1) Dress nice (1) Dress up (1) Makeup (3) Wear makeup (2) Lipgloss (1) Feminine hair (2) Pretty hair (1) Long hair (1) Hygiene (2) Not sweaty (1) Good smell (1) Different hair (1) Big feet (1) Cool shoes (1) Nails (1) Jewelry (1) Taller (1) Wear earrings (1) Wear high heels (1) Fun (27) Funny (17) Fun (6) Funnier (1) Have fun (1) Humorous (1) Love to have fun (1) Smart (25) Smart (19) Smarter (3) Clever (2) Be smart (1) Friendly (17) A true friend (1) Friendly (7) Get along (2) Nice to one another (1) Nice to others (1) Nice to boys (1) Good sport (1) Good to others (1) Happy to other people (1) Join people (1) Athletic (16) General (10) Athletic (1) Active (1) Sports (1) Sporty (4) Fast (1) Play sports (1) Like to exercise (1) Dance (2) Dance (1) Learn to dance (1) Boxing (1) Golf (1) Gymnastics (1) Skateboarding (1) Happy (14) Cheerful (1) Glad (1) Happy (8) Joyful (3) Bold (13) Helpful (13) Helpful (12) Help others (1) Respectful (12) Respectful (10) Respect (2) Strong (12) Strong (11) Stronger (1) Caring (11) Caring (9) Loving (2) Cool (11) Cool (8) Be popular (1)
121 Popular (2) Good (11) Awesome (4) Be awesome (2) Better (1) Good (3) Lovely (1) Playful (11) Silly (3) Always playful (1) Playful (7) Creative (8) Generous (7) Generous (3) Sharing (3) Not selfish (1) Independent (7) Theirself (sic) (1) Themselves (1) Think for yourself (1) Freedom (1) Have their own opinions (1) However they want (1) Independent (1) Occupation (7) Actor (1) Army (1) Electrician (1) Have a job (1) Doctor (1) Scientist (1) Veterinarian (1) Responsible (7) Unique (7) Unique (5) Different (1) (1) Academic success (6) STEM (3) Good at math (2) Good at science (1) General (2) Study (1) Like school (1) Reading (1) Anti Bullying (6) Stop gossiping (1) Not mean (1) Healthy (6) Healthy (4) Be healthy (1) Eat healthy (1) Good manners (6) Polite (2) Be polite (1) Be proper (1) Have manners (1) Honest (6) Honest (3) Tell the truth (1) Truthful (1) Trustworthy (1) Thankful (6) Thankful (5) Grateful (1) Work hard (5) Non judgmental (4) Think good about people (1) Not judgmental (1) Judge based on character (1) Mind their business (1) Brave (4) Brave (2) Courageous (1) Be a hero (1) Drawing (4) (All 1 item) Color pretty Coloring Like drawing Love to draw Outgoing (4) Clean (3) Gentle (3) Loyal (3) Sassy (3) Successful (3) Trusting (3) Be trusting (1) Trusting (2) Weird (3) Careful (2) Confident (2) Confident (1) Happy with themselves (1) Do the right thing (2) Exciting (2) Flexible (2) Ladylike (2) Loves pets (2) Listen (2) Sing (2) Talkative (2) Chatter (1) Be loud (1) MISC (35) (All 1 item) Be spontaneous Better food Colorful Crazy Fabulous Faithful Family Feel bad Game boards Girly/Tomboy Nimble Not crazy Patient Pretty (on the inside) Less annoying Less confusing Less dramatic Like me Like pink and purple Love candy Loves nature Mad Magical Modest Resourceful Relax more often Sad Sell more things Serious Shop Try new things Thoughtful OMIT (4) (All 1 Item) Striping T.U. people Not load Happy/ni ce
122 Table 3 Prescriptive Stereotypes About Boys By Category (In Order Of % Participants ) List of categories for the prescriptive stereotypes about girls generated by participants in response boys icipants generated a total of 434 stereotypes, with an average of 9.86 ( SD =3.58 ) per participant. Categ ories are listed by the percentage of participants who listed items in each category (number of items in each category divided by N =44), and the frequency of the items appearing in each category is indicated in bold NOTE: Many participants listed multipl e items which fell into the same category; for example, a Therefore, percentages may exceed 100%. For a full list of the ite ms in each category, see Table 4 Nice (57) -129.54% MISC (44) -100% Fun (23) 52.27% Smart (20) 45.45% Helpful (17) 38.63% Good (13) -29.54% Respectful (13) -29.54% Athletic general -27.27% Cool (11) 25% Friendly (11) 25% Happy (11) 25% Anti bullying (10) 22.72% Nice to girls (9) 20.45% Caring (9) 20.45% Creative (9) 20.45% Good manners (9) 20.45% Strong (9) 20.45% Generous (8) 18.18% Honesty (7) 15. 9% Playful (7) 15. 9% Bold (6) 13.63% Brave (6) 13. 63% STEM (6) 13.63% Unique (6) 13.63% Healthy (5) 11.36% Responsible (5) 11.36% Work hard (5) 11.36% Clean (4) 9.09% Independent (4) 9.09% Short hair (4) 9.09% Thankful (4) 9.09% Careful (3) 6.8% Academic Gen(3) 6.8% Fast (3) 6.8% Gentle (3) 6.8% Weird (3) 6.8% Attractive fe m (3) 6.8% Reading (3) .6.8% Colorful (2) 4.5 % Doctor (2) 4.5% Flexible (2) 4.5% Handsome (2) 4.5% Love animals (2) 4.5% Loyal (2) 4.5% Non judgmental (2) 4.5% Outgoing (2) 4.5% Patient (2) 4.5% Quiet (2) 4.5% Sing (2) 4.5% Successful (2) 4.5% Talkative (2) 4.5%
123 Table 4 Prescriptive Stereotypes About Boys By Frequency And Category 34 stereotypes, with an average of 9.86 ( SD =3.58 ) per participant. Items are grouped by categories generated by two independent raters and then listed in order of th e frequency of the category. Bold indicates a category. Parentheses indicate frequency, and the frequencies of each category are the sum of the frequencies of each item in that category. Categories listed in blue are not homogenous in meaning, contain subc ategories, and are not included in the descriptive analysis; they are listed here to be illustrative of meaning. Items listed in red are too similar to other features in the category to be miscellaneous, but too different from other items in the category t o co nsider the category homogenous. ______________________________________________________________________________ Nice (57) Nice (32) Kind (16) Kind to each other (2) Nicer (2) Agreeable (1) Be sweet (1) Kindful (sic) (1) Sometimes nice ( 1) Some boys are kind (1) Fun (23) Fun (6) Funny (14) Have fun (3) Smart (20) Smart (16) Clever (3) Understand life (1) Athletic (19) Athletic general (12) Sporty (4) Active (2) Like sports (1) All sports (1) Enjoy sports (1) Work out (1) Love to sweat (1) Play outside (1) Fast (3) Fast (2) Run fast (1) Likes to play football (1) Run laps (1) Play kickball (1) Soccer (1) Appearance (18) Short hair (4) Attractive feminine (3) Pretty (2) Cuter (1) Handsome (2) Different hair (1) Glasses (1) Good smell (1) Has a Mohawk (1) Handsome/pretty (1) Not sweaty (1) Six pack abs (1) Wear chapstick (1) Helpful (17) Helpful (13) Help (2) Help out people in need (1) Help around (1) Good (13) Good (4) Not bad things (1) Awesome (6) Fantastic (1) Fabulous (1) Respectful (13) Academic success (12) General (3) School (1) Study (1) Go to school (1) STEM (6) Science (1) Math (4) Love math (1) Computers (1) Reading (3) Read (1) Do reading (1) Good at reading (1) Cool (11) Cool (10) Cool to hang out with(1) Friendly (11) Friendly (6) Play together (1) Good to others (1) Join people (1) Good to each other (1) Good sport (1) Continued on next page
124 Happy (11) Happy (6) Joyful (3) Be happy (2) Anti bullying (10) Not so mean (1) so mean (1) Not mean (2) Not be mean (1) (1) Not a bully (1) Not so rough (1) (1) Be nice to girls (9) Gentleman (2) Nice to girls (2) strong (1) Stop judging girls (1) Great to girls (1) Care for girls too (1) Caring (9) Caring (7) Have an open heart (1) Loving (1) Creative (9) Creative (6) Crafty (1) Be more creative (1) Get ideas (1) Good manners (9) More polite (1) Polite(1) weird stuff (1) Have manners (1) Be polite (1) Considerate (1) Proper (1) Be proper (1) Generous (8) Generous (3) sharing (3) share (2) Strong (9) Occupation (8) Doctor (2) Be a teacher (1) Scientist (1) Make buildings (1) Nice job (1) Artist (1) Actor (1) Honesty (7) Honest (4) Truthful (3) Trustworthy (1) Playful (7) Playful (3) Play games (1) Play hard (1) Silly (2) Bold (6) Brave (6) Have courage (1) Brave (4) Courageous (2) Unique (6 ) Unique (4) Different (1) Own personalities (1) Healthy (5) Healthy (4) Eat healthy (1) Responsible (5) Responsible (4) Responsibility (1) Work hard (5) o Work (1) o Work hard (4) Clean (4) Independent (4) Any way they want (1) Choose what they like not someone else (1) Have own opinion (1) Themselves (1) Thankful (4) Artistic (3) Paint (1) Draw (1) Make sculptures (1) Careful (3) More careful (1) Careful (2) Gentle (3) Weird (3) Colorful (2) Flexible (2) Love animals (2) Loves pets (1) Love dogs (1) Loyal (2) Non judgmental (2) Non judgmental Not care about looks Outgoing (2) Patient (2) Quiet (2) Sing (2) Successful (2) Successful (1) Achiever (1) Talkative (2) Loud (1) Chatter (1) Continued on next page
125 MISC (All have 1) Scary Aliens Less annoying Sleepy Mean Safe Not crazy Crazy Lovable More focused Gross Eaters Good sport Magical Less confusing Forgetful Recycle Family Optimistic Act Do test driving Resourceful Nimble Cook Listener (1) Make toys Sassy (1) Play on the trampoline Game boards Work around the house? Trust Donate food Thoughtful (1) Entertaining Non stereotype Be able to like pink Lonely (1) Sad (1) Mad (1) considerate OMIT Wear Boystins (1) Redspe (1)
126 Table 5 Correlations Among Stereotype, Self Perception, and Gender Identity Measures. Variable Free Response COAT Felt Pressure Typicality Content SP Girls SP Boys Free Response --.22 .21 .22 .15 .04 .30* COAT --.33* .03 .018 .06 .22 Felt Pressure --.45** .45* .14 .09 Typicality --.74** .20 .02 Contentedness --.17 .01 Self Perceptions Girls --.76** Self Perceptions Boys --Note. *p <.05, ** p < .01
127 Table 6 Test of Stereotype Emulation Hypothesis COAT and Typicality as Predictors and Self Perceptions (Boys) as Criterion (N=44) Variable t p Intercept 2.76 2.63 0.01 COAT 2.49 2.4 5 0.01 Typicality 0.14 0.43 0.67 Note. R 2 = 0.15 F = 3.57 P = 0.0 38 Table 7 Test of Stereotype Emulation Hypothesis COAT and Contentedness as Predictors and Self Perceptions (Boys) as Criterion (N=44) Variable t p Intercept 2.44 2.37 0.02 COAT 2.45 2.64 0.018 Contentedness 0.26 0.92 0.36 Note. R 2 = 0.16 2 F = 3.95 P = 0.0 27 Table 8 Test of Stereotype Emulation Hypothesis COAT and Felt Pressure as Predictors and Self Perceptions (Boys) as Criterion (N=44) Variable t p Intercept 3.51 3.04 0.004 COAT 2.63 2.67 0.01 Felt Pressure 0.15 0.52 0.61 Note. R 2 = 0.15 F = 3.62 P = 0.0 36 Table 9 Stereotype Emulation Hypothesis Free Response Stereotyping and Contentedness as Predictors and Self Perceptions (Boys) as Criterion (N=43) Variable t p Intercept 2.67 2.35 0.02 Free Response 2.92 2.51 0.02 Contentedness 0.39 1.34 0.19 Note. R 2 = 0.15 F = 3.61 P = 0.0 36
128 Table 1 0 Stereotype Construction Hypothesis Self Perceptions(Boys) and Typicality as Predictors and Free Response Stereotyping as Criterion (N=43) Variable t p Intercept 0.21 7.94 <0.0001 Self Perceptions Boys 0.06 2.22 0.03 Typicality 0.02 1.54 0.13 Note. R 2 = 0.15 F = 3.57 P = 0.04
129 ALL ABOUT YOU! YOUR PIN #: __________ Appendix A BEFORE WE GET STARTED: Thank you for agreeing to complete this survey! The goal of this activity is for me to learn what you think and how you feel. There are no wrong answers! This is a PRIVATE about your answers while you are taking it! Your answers will be kept secret, unless you write something that indicates that you are in which survey is yours! Please be honest and really try to think hard about your answers. Part I Jessa Baker Moss New College of Florida January 2013
130 WHAT I AM LIKE Make a list of at least 10 words that describe you There are no right or wrong answers. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
1 31 WHAT SHOULD GIRLS BE LIKE? Make a list of at least 10 things that YOU think a girl should be! THEN, circle number 1 if you think for girls to be like that; circle number 2 if you think it is kind of important for girls to be like that; circle number 3 very important for girls to be like that. Only circle one number for each! Then circle whether you think boys should also be like that. There are no right or wrong answers! We just want to know what YOU think. Girls should be: Not Sor t of Very Should Important Important Important Boys Too? 1. 1 2 3 Yes /No 2. 1 2 3 Yes /No 3. 1 2 3 Yes /No 4. 1 2 3 Yes /No 5. 1 2 3 Yes /No 6. 1 2 3 Yes /No 7. 1 2 3 Yes /No 8. 1 2 3 Yes /No 9. 1 2 3 Yes /No 10. 1 2 3 Yes /No 11. 1 2 3 Yes /No 12. 1 2 3 Yes /No 13. 1 2 3 Yes /No 14. 1 2 3 Yes /No 15. 1 2 3 Yes /No
132 WHAT SHOULD BOYS BE LIKE? Make a list of at least 10 things that YOU think a boy should be! THEN, circle number 1 if you think very important for girls to be like that; circle number 2 if you think it is kind of important for girls to be like that; circle number 3 very important for girls to be like that. Only circle one number for each! Then circle whether you think boys should also be like that. There are no right or wrong answers! We just want to know what YOU think. Boys should be: Not Sor t of Very Should Important Important Important Girls Too? 1. 1 2 3 Yes /No 2. 1 2 3 Yes /No 3. 1 2 3 Yes /No 4. 1 2 3 Yes /No 5. 1 2 3 Yes /No 6. 1 2 3 Yes /No 7. 1 2 3 Yes /No 8. 1 2 3 Yes /No 9. 1 2 3 Yes /No 10. 1 2 3 Yes /No 11. 1 2 3 Yes /No 12. 1 2 3 Yes /No 13. 1 2 3 Yes /No 14. 1 2 3 Yes /No 15. 1 2 3 Yes /No
133 Appendix B