Gender Schema and Social Judgments: A Developmental Study of Children from Hong Kong

Thalma E. Lobel

Thalma E. Lobel [1]

The purpose of the study was to examine inferences and judgments of gender schematic and aschematic children. Kindergarten, third grade, and sixth grade Hong Kong Chinese children (n = 196) were categorized into schematic and aschematic groups on the basis of their reaction times to gender-stereotypic and counterstereotypic items. The children were then asked to make several inferences and judgments about four male and female targets behaving gender-stereotypically or counterstereotypically. Both age and gender schematicity were related to the children's inferences. Older children and genderschematic children relied more on individuating information and younger children relied more on the gender label. The results are discussed within the framework of gender-schema theories.

INTRODUCTION

Gender schema theories (e.g., Martin & Halverson, 1981, 1987) maintain that gender is a major component around which children organize information (Martin, 1989; Martin & Halverson, 1981). These theories assert that gender-related information is organized in the form of a schema, an abstract knowledge structure that serves as implicit theory, and expectations that guide attention, retrieval, behavior, and social judgment (Martin & Halverson, 1981; Stangor & McMillan, 1992). The present study is concerned with the purported influence of gender schema on children's inferences and judgments.

Many studies have illustrated that children make gender-related inferences and judgments (Berndt & Heller, 1986; Biernat, 1991; Martin, 1989; Martin, Wood & Little, 1990). These studies focused on developmental issues and, although their findings somewhat differ, they all concluded that the relative extent to which children's inferences rely on gender label and on individuating information, i.e., targets' interests and activities, varies with age. Whereas kindergarten children rely mainly on gender label, older children rely more upon individuating information. Biernat (1991) offered two explanations as to why kindergarten children rely solely on gender label. First, as was evidenced in past research, the gender label is a very salient judgment cue upon which children base their predictions about the characteristics of a given individual (Martin, 1989). Second, it is likely that, given younger children's relatively limited cognitive capacity, they are not able to integrate multiple sources of information or reconcile contradictions, such as the ones inherent in counterstereotypic behavior (Harter & Buddin, 1987; Higgins, 1981; Martin, 1989).

Gender schema theories (e.g., Martin & Halverson, 1981) maintain that individual differences in gender schematicity should be considered in order fully to understand children's gender-related behavior and judgments (for a review see Signorella, Bigler, & Liben, 1993). Researchers in gender development agree that individual differences exist in various dimensions of gender typing including knowledge, flexibility, gender schematicity, and gender-typed behavior (e.g., Katz & Boswell, 1986; Katz & Ksansnak, 1994; Signorella, 1987; Turner & Gervai, 1995). However, although current research has established clear developmental differences in children's inferences and judgments, surprisingly little empirical research directly examines how and when these judgments might be influenced by individual differences in children's gender schematicity. This is especially surprising considering the fact that individual differences in the use of gender-related knowledge were acknowledged "It is intriguing that some children spo ntaneously guessed the sex of the target children, indicating that they may have used gender to mediate their judgments. Perhaps gender is very salient for these children" (Martin et al., 1990, p. 1897).

The present study focuses on the relationship between individual differences in children's gender schematicity, and inferences and judgments in kindergarten, third grade, and sixth grade children. In order to formulate our hypotheses we shall first try to define gender schematicity by integrating several theoretical frameworks, and then examine individual differences in gender schematicity from a developmental perspective.

Conceptualization of Individual Differences in Gender Schematicity

Investigators have elaborated upon definitions of individual differences in gender schematicity by distinguishing conceptually among different aspects of the gender schema. Although specific definitions vary, two general components typically define individual differences in schematicity: accessibility of the gender schema and its complexity and richness.

Individual differences in gender schematicity can be conceptualized as differences in the global and prominent tendency to use the gender schema construct, i.e., differences in its importance when attending and judging the social world (Ruble & Stangor, 1986). According to Bem (1981), gender-schematic individuals have greater readiness to process information on the basis of gender than do gender-aschematic individuals. Higgins (1996; Higgins & King, 1981) maintained that there are individual differences in the chronic accessibility of the stored cognitive constructs that influence the information processing of related stimuli (e.g., Bargh & Thein, 1985; Bargh, Bond, Lombardi, & Tota, 1986; Higgins & Bargh, 1987). Therefore, gender-schematic and aschematic individuals differ in the chronic accessibility of their gender schema.

Another distinction between gender-schematic and gender-aschematic individuals is the richness and the complexity of the gender schema (Martin & Halverson, 1981; Ruble & Stangor, 1986). According to Smith and Zarate (1992), similarity between a perceiver's mental representation and a target stimulus determines the judgments of that target. They assert that this similarity is not a fixed property of the stimulus, but is mediated by the perceiver's relative allocation of attention to various stimulus dimensions such as gender or age. The more attention a person pays to a certain dimension of a stimulus, the more familiar this aspect of the stimulus becomes. The relative allocation of attention is affected by individual differences such as past experience and self-schemas. Thus, gender-schematic individuals are exposed to more gender-related material, which hence becomes more familiar to them. Furthermore, not only schematicity, but also age affects amount of children's exposure to gender-related material. The older they become, the more children are exposed to gender-related stimuli, and thus they become more familiar with them. According to Linville, Fischer, and Salovey (1989), greater familiarity with a category leads to a greater perceived differentiation and variability of this category, which, in turn, results in a richer and more differentiated set of experiences. These experiences are reflected in the mental representation of the group. Thus, gender-schematic individuals are expected to have a more elaborated, richer, and more differentiated gender schema.

In this work we suggest that at different ages the individual differences in gender schematicity are differently manifested depending on the capabilities and the limitations that are typical of each age group. According to Martin (1993), gender stereotypes develop through a series of phases. In the first phase, children learn what kinds of things are directly associated with each sex as well as acquire the basic ability to identify the sexes and to label them (Martin, 1993). In our view, what differentiates genderschematic from gender-aschematic children at this early stage is how accessible the gender-related information is to them, i.e., individual differences in the prominent tendency to use the sex of the target and to rely on it when attending to and judging the social world. Thus, we predicted that for kindergarten children, the more schematic they are, the more they rely on gender label information. Evidence for such individual differences at very young ages was indicated in the work of Fagot and Lein bach (1989), who found that children differed in the age they acquired gender labels.

According to Martin, after children have acquired gender labels, they can start structuring and elaborating their gender-related information. As children grow older, more domains of knowledge become gender-typed. Researchers have demonstrated that young children associate gender labels with various components such as appearance (Thompson, 1975), activities and toys (Blakemore, LaRue, & Olejnik, 1978; Fagot & Leinbach, 1989; Thompson, 1975), and occupations (Kuhn, Nash, & Brucken, 1978; O'Keefe & Hyde, 1983; Papalia & Tennet, 1975). Martin maintains that at around the age of 4-6 years, children begin to develop indirect and complex associations in the processing of information relevant to their own sex. By the age of 8 years, children have also acquired the associations relevant to the other sex. By this time they have acquired an understanding of the gender concepts of masculinity and femininity such that links are evident both within and between the various content domains (Martin, 1993). We assume that ind ividual differences in the readiness to process information on the basis of gender leads to differences in the representation and organization of gender-related knowledge. The more attention a child allocates to gender-related information, the more elaborated his/her gender schema will be. Indeed, studies have shown that the ability to distinguish between the sexes at a young age was related to a broader gender-related knowledge at an older age (Fagot and Leinbach, 1989; Martin and Little, 1990). Thus, those who were more schematic at a younger age, manifested by their greater reliance on the gender label, were also more schematic at an older age, which was manifested in their having a more complex and a richer gender schema. Martin (1993) further suggested that developmental changes in gender knowledge relate to developmental changes in children's ability to recognize variability within gender groups. For instance, as children grow older, their understanding of variability within gender groups increases so t hat they are more likely to attend to individuating information.

Thus, we predicted that at an older age, the more schematic children are, the more they will rely on individuating information when making inferences and judgments.

The Present Study

The present study compared the inferences and judgments of gender-schematic and aschematic children of three age groups in Hong Kong, kindergartners, third graders, and sixth graders. These particular age groups were sampled since previous studies have shown that at around these ages children move from one phase of gender-related information processing to another (Martin, 1993). For example, Martin (1993) maintains that at around age 4--6 years, children move to the phase where they begin to develop more complicated and indirect associations regarding their own sex. At about 8 years of age, children move to the next phase, where they acquire associations regarding the other sex. Indeed, it has been shown (Berndt and Heller, 1986; Biernat, 1991) that the gender-related inferences and judgments of these particular age groups differ.

Most previous studies examining children's inferences (e.g., Berndt & Heller, 1986; Biernat, 1991) focused on what are called cognitive inferences (Lobel, 1994). Children were presented with a target that behaved either gender-stereotypically or counterstereotypically and were asked to make inferences and judgments about other dimensions of the gender stereotype such as traits, roles, and occupations. Some studies also examined the perceived popularity of the targets (e.g., Berndt & Heller, 1986). However, most studies did not examine children's emotional--motivational judgments, such as affinity, willingness to engage in activities with the target, and perceived similarity to the target. One exception was Martin (1989), who asked young children to indicate how much they liked the various targets. The only two studies that examined not only affinity, but also willingness to engage in activities and perceived similarity to the target were studies conducted with preadolescents (Lobel, Bempechat, Gewirtz, Shoke n-Topaz, & Bach, 1993; Lobel, 1994) investigating the relationship between preadolescents' self-perception and inferences and judgments. The present study examines cognitive and emotional--motivational inferences of gender-schematic and aschematic children. Regarding emotional motivational judgments, we expected that age and schematicity will interact and affect differently the relationship between how stereotypic the target is perceived and how he/she is liked and preferred. More specifically, as mentioned above, we assumed that for younger children the more schematic they are, the more they rely on gender labels. We therefore predicted that for kindergarten children, the more schematic they are, the more they prefer the target of their own sex. Hence, we expected that the more schematic they are, the more there will be a negative relationship between how stereotypic the target is perceived by them and how much they like and prefer that target. In contrast, since we assume that for older children the more sc hematic they are, the more they rely on individuating information, we expected that the more schematic they are, the more there will be a positive relationship between how stereotypic the target is perceived by them and how much they like and prefer that target. This will be true only for targets of the same sex.

The current study measured schematicity with an adapted form of a measure developed by Carter and Levy (1988, 1991; Levy, 1994; Levy & Carter, 1989) based upon children's reaction times to toy choices. Carter and Levy reasoned that since gender is more salient for gender-schematic individuals than for gender-aschematic individuals, the former would base their choices mainly on gender, while the latter would be more likely to base their choices on dimensions other than gender. Therefore, genderschematic individuals should have a relatively faster reaction time when asked to choose between masculine and feminine items (heterogeneous pairs). In contrast, when asked to choose between two homogenous gendertyped items (i.e., two masculine or two feminine toys), reaction times should be relatively longer for these individuals. Thus, the difference between reaction time to homogenous pairs and reaction time to heterogeneous pairs should be greater for schematic than for aschematic individuals.

Carter and Levy (1988, 1991; Levy, 1994; Levy & Cater, 1989) found that gender schematization was associated with children's learning, memory, and toy choice. These findings illustrate that, even at a young age, there are individual differences in children's gender schematicity that are related to various behaviors. The present study used an adapted computerized form of Carter and Levy's measure that was further developed in two major issues. First, it took into consideration the multidimensionality of the gender schema (Deaux & Lewis, 1984; Martin, 1993). Thus, children's schematicity was determined from their reaction times not only to pairs of toys, as in the original measure, but also to pairs of activities, items of clothing, and occupations. Second, computerization of the gender schematicity measure allowed for more finely tuned assessments of individual differences in the use of gender.

Cultures play an important role in setting social norms (Rubin, 1998). However, most studies done on children's inferences and judgments have been conducted in Western societies. Developmental Psychology has recently devoted a special issue to the cross-cultural perspective of social and emotional development (Rubin, 1998) and showed the importance of culture in these developmental processes. It has been suggested that cultures vary along the dimension of individualism and collectivism (Markus & Kitayama, 1991; Shweder, 1990; Triandis, 1989). Collectivistic cultures are defined as cultures in which the most valued aspects of the self concept are those related to the norms, role definitions, and values of a particular collective. These cultures focus on roles, status, and fitting in. In contrast, in individualistic cultures, personal goals are valued more than group norms.

Many Asian cultures, among them the Chinese (Hsu, 1985), can be clearly defined as collectivistic cultures (Ames, Dissanayake, & Kasulis, 1994; Kondo, 1990; Lebra, 1976). Hong Kong is a Chinese society that has been exposed to both Chinese and Western cultures, but on the whole, its values are similar in many respects to those of traditional Chinese culture (Chiu, 1988, 1989; Chiu & Hong, 1992). Thus, many aspects of life in Hong Kong as a collectivist culture are shaped according to the demand that individuals conform to the group norms and values.

Collectivistic cultures emphasize behavior according to social norms. Gender roles constitute a subdomain of normative behavior that is particularly stressed. In collectivistic cultures the goals of harmony and fitting in make counterstereotypic behavior even more salient than in Western cultures and it is therefore judged more harshly. It has been shown that in Hong Kong society gender stereotypes are particularly distinct and the expectations for males and females clearly differ from one another (Keyes, 1983, 1984). It would therefore be especially interesting to examine the relationship between individual differences in gender schematicity and inferences and judgments in such a collectivist culture, where individual differences are relatively less encouraged.

METHOD

Participants

One hundred and ninety-six Chinese children (103 boys and 93 girls) from Hong Kong participated in the study. After receiving consent from the Bureau of Education, principals, teachers, and parents, the children were recruited from a Chinese primary school and a Chinese kindergarten located in the Kowloon Peninsula. Participants were sampled from three age groups, 60 kindergarten children, aged 4-6 years (mean age = 62 months), 70 third graders, aged 8-9 years (mean age = 105 months), and 66 sixth graders, aged 11-12 years (mean age = 141 months). Response rates in the classrooms ranged from 92% to 98%.

Materials

The Computerized Gender Schematicity for Children (CGSC)

In order to construct a measure of gender schematicity suitable for the Chinese society, a pretest was conducted whereby pictures of various items representing several of Deaux and Lewis' (1984) components of the gender stereotype--toys, activities (role component), parts of clothing (physical appearance component)--and occupations (occupational component)--were presented to 40 Chinese children from the three age groups of the study. The children were asked to indicate on a 5-point scale the masculinity and femininity of each item and its attractiveness.

Based on the pretest, the Computerized Gender Schematicity for Chinese Children (CGSC) was designed. Only those pictures that were clearly deemed as either masculine or feminine (i.e., were given a score of 5 on the scale of masculinity or femininity) were included. In addition, each pair included items that were perceived as equally attractive, as well as equally masculine or equally feminine. Altogether, 28 pairs of colored pictures were presented, representing toys and objects (e.g., cars, hammer and nails, toy gun, saw, tractor, dolls, sewing kit, cosmetics, jewelry, toy oven), activities (e.g., ironing, dishwashing, sewing, boxing, sawing) occupations (e.g., fireman, soldier, teacher, nurse), and physical appearance (e.g., traditionally feminine or masculine bags, shirts, shoes, glasses, pants, watches). In a pretest, the correlations among the different schematicity scores were high and ranged from .59 to .81. Each picture was 9X10 cm in size and the distance from its pair was 8 cm. Each pair of pictur es was either homogenous (i.e., both items were masculine or both were feminine) or heterogeneous (i.e., one masculine and one feminine item).

The pairs of items were displayed in a randomized order on a PC computer of 28 X 23 cm in size. Each child was tested individually and presented with the various pairs of pictures. For each pair, children were asked to choose the picture they preferred most, as fast as possible. Two computer keys were highlighted such that the one on the left corresponded to the answer choice of the picture on the left and the one on the right corresponded to the picture on the right. Children were asked to indicate their choice by pressing a computer key of the chosen item. Reaction time in milliseconds was automatically recorded by the computer.

In accordance with Carter and Levy (1991), participants' reaction times to the relevant pairing were standardized and a z score for each child was computed using the scores on the relevant items, the child's mean reaction time across all items, and the standard deviation of the individual child's reaction times. Each child received two scores, an inhibited and a facilitated score. Schematicity score was computed by computing the difference between the two scores. The larger the difference between the two scores, the more schematic the child was.

The Inference and Judgment Questionnaire

Four stories, two of a male target and two of a female target, with either traditionally masculine or traditionally feminine interests, were composed. The masculinity and femininity of the various interests were determined in a pretest. Some of the items chosen as gender-typed by Chinese children were similar to those chosen by Western children, such as dolls (feminine) and cars (masculine). In contrast, other items were perceived as gender-typed only by Chinese children, such as singing (feminine). The following stories were composed:

Masculine boy version: Chi Wai is a boy your age who lives around here. He loves to play with cars and sometimes in the afternoon he likes to build model airplanes.

Feminine boy version: Chung Wai is a boy your age who lives around here. He loves to sing and, sometimes, in the afternoon, he likes to play with dolls.

Feminine girl version: Yan Yen is a girl your age who lives around here. She loves to sing and, sometimes, in the afternoon she likes to play with dolls.

Masculine girl version: Yan Lim is a girl your age who lives around here. She loves to play with cars and sometimes in the afternoon she likes to build model airplanes.

The stories told to the sixth grade children differed slightly from those told to the kindergarten and third grade children. The feminine activity of playing with dolls was changed to needlework, and the masculine activity of playing with cars was changed to football. We chose to replace these activities since no common activities to all age groups were found and a pretest showed that older children do not usually play with dolls and cars, and younger children do not usually play with needlework and football. It was therefore important to choose different activities for the sixth graders to prevent the perception of older children engaging in younger activities as childish. At the same time it was important to maintain the extent of the masculinity/femininity of the activity as well as its popularity. We conducted a pretest and found that the sixth grade activities were deemed equally masculine or feminine to those of the third grade and kindergarten. In addition it was found that playing with dolls for girls and cars for boys were very popular activities among younger children, whereas needlework and football were very popular activities among sixth grade girls and boys, respectively. Thus, although the stories differed in one activity, the masculinity/femininity of the activities as well as their popularity were very similar.

Following each story, participants were asked to make several cognitive and emotional--motivational inferences and judgments.

Cognitive Inferences

Task 1. Predicted play and after-school activities: Participants were asked to predict on a 5-point scale the target's choice of traditionally feminine and masculine play and after-school activities (e.g., Does this child take a course of Tibondo--a masculine Chinese activity? Does this child take a course of ballet? Does this child play mainly with boys? Does this child play mainly with girls?).

Task 2. Birthday present: The children were asked to choose on a 5-point scale traditionally masculine and feminine gifts for the target (Would you buy this child a toy gun? Would you buy this child origami papers?).

Task 3. Predicted occupations: Children were asked to rate on a 5-point scale the probability that the target child would choose one of the following future occupations: policeman or fireman (traditionally masculine occupations), teacher or nurse (traditionally feminine occupations).

Task 4. Predicted popularity among peers: Participants were asked to rate on a 5-point scale (1) how popular the target was among other boys and (2) how popular the child was among other girls, ranging from 1 (not at all popular) to 5 (very popular).

Emotional--Motivational Judgments

Task 5. Affinity for target: Participants were asked to rate on a 5-point scale how much they liked the target, ranging from 1 (not at all) to 5 (very much).

Task 6. Willingness to befriend the target: Children were asked to indicate on a 5-point scale the extent they would like to befriend the target.

Task 7. Willingness to engage in activities with the target: Children were asked to rate on a 5-point scale ranging from 1 (not at all) to 5 (very much) the extent to which they would like to engage in various activities with each candidate: engage in social activities with him or her, reveal a secret to him or her, and invite the target to their home.

Task 8. Perceived similarity to the target: Children were asked to rate on a 5-point scale how similar they perceived themselves to be to the target, ranging from 1 (not at all similar) to 5 (very similar).

Martin's (1989) method of rating task was employed to help the kindergarten children in their ratings. The children were shown a large poster board with five boxes of varying size. The smallest box was described as meaning not at all, the next largest as a little bit, the middle box as a medium amount, the next largest as pretty much, and the largest box as very much. The children were then asked to point to a box in order to state their rating. The internal reliability of the questionnaire was quite high, and Cronbach alpha reliability score was .71.

Procedure

The study was conducted in two stages within an interval of 3 days. Each session lasted between 30 and 40 mm.

Stage 1

The CGSC was administered to each child individually. Before the test itself began, two practice pairs of pictures were shown in order to verify the comprehension of the instructions. The children were then presented with the pairs of pictures on the computer and asked to choose the item that they liked most, as fast as they can, by pressing the marked key on the side of the picture on the computer keyboard. Preceding each set of pictures, specific instructions were given. For example, in the activities set, the children were told, "I'm going to show you pictures of all sorts of activities. Each time you will see two pictures. Choose the one you would like to do more. Answer as fast as you can." The participants' choices were recorded by the computer and reaction times were measured in milliseconds from the appearance of the pair of pictures on the computer until the key representing the child's choice was pressed.

Stage 2

Each child was presented with the four stories. Following each story the child was asked to answer the various questions. The stories and the questions were presented in a randomized order. The experimenter repeated each story in the middle of the questionnaire to refresh the children's memory. Similarly to Biernat (1991), each kindergarten child was interviewed and tested individually by an experimenter, whereas the third and sixth graders were tested in a group. Although it is preferable to use the same procedure with all ages, most studies that compare preschoolers to older children use such a procedure. Since kindergarten children cannot read, the only alternative would have been to read the questions to all age groups. We believe that in the present study, reading the questions to all children would have caused other problems. For example, older children are more sensitive to social desirability and hence the presence of the experimenter might influence their answers more than in the case of kindergarte n children. In addition, it might seem odd to older children that the questions are being read to them. It should be noted that we tried to make conditions as similar as possible by reading the kindergarten children exactly the same questions as presented in the questionnaires. We made sure that the kindergarten children were read the instructions and the questions in exactly the same words with no additional explanations. This procedure was similar to the one used in other studies dealing with gender-related inferences (e.g., Biernat, 1991).

RESULTS

Cognitive Inference

A set of regression analyses was performed to examine the effect of children's age, sex, and schematicity on their cognitive judgements. More specifically, we examined the effect of these variables on children's femininity and masculinity judgments of the counterstereotypic targets (i.e., the feminine male target and the masculine female target). In order to avoid confusion, since the sex of both the target and the participants is discussed, we differentiate between the two by referring to the target's sex as male and female and to the participant's sex as boy and girl. Table I presents the correlation matrix among the three independent variables (age, sex, and schematicity) and the four dependent variables (perceived masculinity and perceived femininity of the feminine male target, and perceived masculinity and perceived femininity of the masculine female target).

Table I shows that there are high positive correlation coefficients between judgments of femininity of the feminine male target and masculinity of the masculine female target, and between femininity of the masculine female target and the masculinity of the feminine male target. In addition, while age was positively correlated with the femininity attributed to the feminine male target and the masculinity attributed to the masculine female target, it was negatively correlated with the masculinity attributed to the feminine male target and with the femininity attributed to the masculine female target.

To examine the effects of the three independent variables on each of the four dependent variables, four three-step hierarchical regression analyses were performed. In the first step, the effect of the three main effects was assessed. In the second step, the three first-order interactions were introduced. Finally, in the third step, the second-order interaction was assessed. Following Dunlap and Kemery's (1987) suggestion concerning the reduction of multicollinearity, schematicity was standardized before the respective cross-products were computed. All four analyses as a whole achieved significance. The results show two different patterns. On one hand, the analyses on the masculinity attributed to the masculine female target and the femininity attributed to the feminine male target yielded only a main effect of age (b = .41, t 4.82, and p[less than].01, and b = .56, t = 6.73, and p[less than].01, respectively). On the other hand, the analyses on the femininity attributed to the masculine female target and the masculinity attributed to the feminine male target analyses yielded a main effect of age (b = - .52, t = 6.13, and p[less than].01, and b = - .49, t = 5.78, and p[less than].01, respectively) and an interaction effect of age by schematicity (b = - .24, t = 3.03, and p[less than].01, and b = - .21, t 2.58, and p = .01, respectively). The second-step equation of the analysis of masculinity of feminine male is

Y' =.06 + A * (-.50) + B * (-.09) + C * (-.00) + AB * (.03) + AC * (-.21) + BC * (.11)

The equation for the analysis of femininity of the masculine female target is

Y' = .06 + A * (-.53) + B * (-.16) + C * (-.07) + AB * (.05) + AC * (-.24) + BC * (.04)

Here A stands for participant's age, B for participant's sex, and C for participant's schematicity; the rest are the three first-order interaction terms among the three variables. For interpreting the source of the first-order interaction terms, two sets of two regression coefficients were calculated separately for each age group: masculinity attributed to the feminine male target and femininity attributed to the masculine female target on participants' schematicity. The regression coefficients were calculated using the equation obtained in the second step. Specifically, the regression coefficient of participants' schematicity was added to that of the interaction term after the latter was multiplied by -1, 0, or +1 (Cohen & Cohen, 1983). The results of the analyses show that in the case of masculinity attributed to the feminine male target, for the kindergarten children, the regression coefficient was positive (b = .21), whereas for sixth graders it was negative (b = -.21), and for third graders there was no relationship (b = .00). That is, the more schematic the younger participants were, the more masculinity they attributed to the feminine male target. Conversely, the above relationship was inverted in the case of the sixth grade rs: the more schematic they were, the less masculinity they attributed to the feminine male target. A similar pattern emerged for the attribution of femininity to the masculine girl. While the regression coefficient was positive for the youngest age group (b = .31), it was negative for the oldest age group (b -.17) and in between for the middle age group (b = .07). That is, the more schematic the youngest children were, the more femininity they attributed to the masculine girl, whereas the more schematic the oldest children were, the less femininity they attributed to the masculine girl.

Emotional--Motivational Judgments

In order to test our hypotheses, we performed regression analyses that examined how children's sex, age, and schematicity affected the relationship between how stereotypical the target was perceived and the degree to which that target was liked and preferred. The dependent measures in these analyses were the liking and preferences of the male and female targets perceived as being stereotypical. The liking of a target perceived as being stereotypical was calculated by subtracting the liking of the counterstereotypic target (male or female) from that of the stereotypic one. That is, the total liking and preference of a stereotypic female (LSF), for example, was created by subtracting the [zeta] score of the liking and preference of the masculine girl from that of the feminine girl. The creation of the predictors of total femininity and of the female target perceived as being stereotypical, and of the total masculinity of the male target perceived as being stereotypical, was calculated according the following f ormula:

Net masculinity of the stereotypic male target = (mmb - fmb) - (mfb - ffb) where mmb is the masculinity of the masculine male target, fmb is the femininity of the masculine male target, mfb is the masculinity of the feminine male target, and ffb is the femininity of the feminine male target. Similarly,

Net femininity of the stereotypic female target = (ffg - mfg) - (fmg - mmg)

where ffg is the femininity of the feminine female target, mfg is the masculinity of the feminine female target, fmg is the femininity of the masculine female target, and mmg is the masculinity of the masculine female target.

Table II presents the correlation matrix among the variables of the second set of the regression analyses. It can be observed that there are relatively low correlation coefficients between the judgments of targets' net gender stereotypicality and the liking and preference of the target from the same sex. Similarly, participants' sex and schematicity had low correlation coefficients with the liking of the stereotypical target of both sexes. On the other hand, participants' age group had a stronger relationship with the liking and preference of the stereotypic target person from both sexes.

To examine the moderating effect of age, schematicity, and participants' gender on the relationship between the perceived stereotypicality of the target and to what degree he or she was liked and preferred, two four-step hierarchical regression analyses were performed. In the first step, the effect of the four main effects was assessed. In the second step, the six first-order interactions were introduced. In the third step, the four second-order interactions were examined. Finally, the effect of the third-order interaction was evaluated in the fourth step. Based on Dunlap and Kemery's (1987) suggestion regarding the reduction of multicollinearity, schematicity and the perception of the target's net gender stereotype were standardized before the respective cross-products were computed. The results of each of the regressions will be described in the following.

Female Target

The results showed that as a whole, the analysis achieved significance, F(15,185) = 3 .8l, p [less than].01. The analysis revealed a main effect of participants' gender (b = .34, t = 5.80, p[less than].01) and a significant four-way interaction (b = -1.10, t = 2.55, p = .01). The fourth-step equation is

Y' = -.07 + A * (-.05) + B * (.42) + C * (-.11) + D * (.08) + AB * (.09) + AC * (-.04) + AD * (.05) + BC * (.13) + BD * (.02) + CD * (.17) + ABC * (.06) + ABD * (-.07) + BCD * (-.05) + ACD * (.07) + ABCD * (-.10)

Here A stands for the perception of the target's net gender stereotype, B for the participant's gender, C for the participant's schematicity, and D for the participant's age group; the rest are the 11 interaction terms among the four variables. For interpreting the source of the four-way interaction, three sets of four regression lines of liking of the female target perceived as being stereotypical on the perception of the target's net gender stereotype were calculated separately for males and females and for each age group, each high and low in their levels of schematicity. Low and high levels of schematicity were calculated according to one standard deviation below and above the mean. Because the analysis was based on the z scores of the independent variables, the values were -1 and 1, respectively. The regression coefficients were calculated using the equation obtained in the fourth step. Specifically, the regression coefficient of the perception of the target's net gender stereotype was added to that of t he interaction term after the latter was multiplied by either - 1 or 1 in the case of interactions involving schematicity and gender and by -1, 0, or +1 in the case of the age group.

The results show that for the youngest participants, kindergartners, schematicity had a different effect on boys and girls. For boys, high levels of schematicity (b = -.53) had a much more negative effect on the relationship between the perceived stereotypicality of the target and the liking and preference of the stereotypical female target than low schematicity (b = 0). For girls, high levels of schematicity had a more positive effect (b = .11) than low schematicity (b = .01). That is, for boys, the higher their level of schematicity, the less they liked and preferred the female target perceived as being more stereotypical. In contrast, for girls, the higher their level of schematicity, the more they liked and preferred the female target perceived as being more stereotypical.

A similar, although weaker, pattern to that of kindergarten children appeared for the third grade children. For third grade boys, high levels of schematicity had a much more negative effect (b = -.23) on the relationship between the perception of the target's net gender stereotype and the liking of the female perceived as being stereotypical than did low levels of schematicity (b = -.04). For girls, high levels of schematicity (b = .06) had a more positive effect than did low levels of schematicity (b = -.04).

Finally, the results show that for the oldest age group, sixth graders, the pattern changes so that boy's high levels of schematicity had a slightly more positive effect (b = .05) on the relationship between the perception of the target's net gender stereotype and the liking of stereotypical female than did low levels of schematicity (b = -.09). For girls, high levels of schematicity seemed to have a similar effect (b = .01) to that of low levels of schematicity (b = .03).

Male Target

The results show that, as a whole, the regression analyses achieved significance, F(15,181) = 3.17, p[less than].Ol. The analyses revealed a main effect of participants' sex (b = -.33, t = 4.83, p[less than].Ol) and the perception of the target's net gender stereotype (b = .10, t = 2.60, p[less than].0l). In addition, a three-way interaction of age group by schematicity by the perception of the target's net gender stereotype reached significance (b = .09, t = 1.97, p = .05). The third step equation is

Y' = -.04 + A * (.07) + B * (-.32) + C * (-.09) + D * (-.01) + AB * (-.05) + AC * (-.02) + AD * (.00) + BC * (-.01) + BD * (-.16) + CD * (.07) + ABC * (.00) + ABD * (-.01) + BCD * (.02) + ACD * (.09) + ABCD * (-.10)

The interpretation of the interaction was performed by calculating the regression coefficients of liking of the stereotypical male target on the perception of the target's net gender stereotype separately for each age group, each low and high in schematicity, in a similar way to that described above.

The results show that for the young age group, kindergartners, the regression coefficient of children low on schematicity (b = .18) is more positive than that of children high on schematicity (b = -.04). In contrast, for the older age group, sixth graders, the regression coefficient for children high on schematicity (b = .10) is both more positive and higher than that for children low on schematicity (b = .00). For the middle age group, third graders, the regression coefficient for children low on schematicity (b = .09) is quite similar to that for children high on schematicity (b = .05).

Although the significant interaction did not involve participants' sex, since we had different hypotheses for the preferences of boys and girls, we decided further to explore to what extent the pattern reported above is typical of both genders. To examine this question, two three-step hierarchical regression analyses were performed separately for male and female participants. In the first step, the effect of the three main effects was assessed. In the second step, the three first-order interactions were introduced. Finally, in the third step, the second-order interaction was examined. The results show that, as a whole, the regression for girls was not significant, F(7,85) = .70, ns. For boys, however, the regression analysis as a whole was significant, F(7,96) = 3.09, p[less than].01. The analysis yielded a significant main effect of perception of target-boy's net stereotype (b = .14, t = 3.13, p [less than] .01) and a marginally significant three-way interaction (b = .11, t = 1.77, p = .08). The interpretat ion of the source of the interaction shows that for the youngest participants, for those with low levels of schematicity there was a higher and more positive regression coefficient (b = .25) than for those with high levels of schematicity (b = -.03). In contrast, in the oldest group, sixth graders, for those with low levels of schematicity the regression coefficient was lower (b = .05) than for those with high levels of schematicity (b = .21). In the middle age group, third graders, the results show that level of schematicity had relatively little effect on the relationship between the perception of how stereotypical the target was perceived as being and the emotional--motivational judgments toward that target, as can be seen from the similar values of the regression coefficients (b = .15 and .09, respectively).

DISCUSSION

Our findings show that several different factors, children's age and level schematicity as well as the target's and participant's sex, affect the types of inferences and judgments children make. These factors interact differently depending on the type of inferences and judgments, be they cognitive or emotional--motivational.

As hypothesized, both age and schematicity were related to children's inferences and judgements. The age-related differences were similar to those found in previous studies (e.g., Biernat, 1991; Martin, 1989; Martin et al., 1990), showing once again that older children relied more on individuating information, whereas younger children relied mainly on the gender label. Thus far, research on gender schematicity has been conducted primarily with European American children. The fact that similar developmental processes occur not only in Western cultures, but also in Eastern societies allows a further generalization of the development of children's inferences and judgments.

However, the main novel contribution made by our study is that we found age was not the only factor related to children's cognitive inferences, but it interacted with gender schematicity. As hypothesized, for kindergarten children, the more schematic they were, the more they relied on the gender label, whereas for sixth grade children, the more schematic they were, the more they relied on individuating information. This was true when they attributed stereotypic characteristics to the counterstereotypic targets (i.e., attributed femininity to the masculine female target and femininity to the feminine male target). In the case of third graders, no difference emerged between schematics and aschematics. Previous studies showed that kindergarten children relied only on the gender label when making inferences and judgments (Berndt & Heller, 1986; Biernat, 1991; Martin, 1989), whereas older children relied on the individuating information. Our results show that this is true especially for the schematic children. At a young age, when the gender schema is less complex and less elaborated, the most salient feature of the schema is the gender label. Therefore, the more schematic a child is, the more he or she relies on the gender label. As suggested by Martin (1989), young children seem to have "blanket rules" such as "all boys like trucks," making it difficult for them to distinguish between feminine and masculine boys. Our results suggest that schematic young children use these "blanket rules" more often, probably since the most salient feature for them is the gender label. In contrast, sixth graders have a more complex and developed gender schema and the associations between the various components of the gender schema are more complex. Therefore, the most salient feature of their gender schema is the individuating information. As such, the more schematic one is, the more he or she is likely to rely on individuating information.

Multiple classification skill, i.e., the cognitive ability to sort the same objects along alternative dimensions, might also be related to the individual differences that were found regarding the use of gender-related individuating information. It has been previously demonstrated that multiple classification skill is correlated with gender stereotyping (Liben & Signorella, 1992). Thus, it could be that in our study participant's difficulty in understanding that the same person can fit simultaneously into two categories influenced their ability to process counterstereotypic gender-related information (Liben & Signorella, 1992).

In their emotional--motivational judgments (the degree of affinity toward the target and willingness to befriend and engage in activities with the target), as predicted, not only did the children's age and level of schematicity interact, but the targets' and participants' sex did as well. Regarding the emotional--motivational judgments of the male target, different results emerged for the younger and older age groups. For kindergarten schematic boys, no relationship was found between how stereotypic the male target was perceived to be and how much he was liked and preferred. Conversely, for kindergarten aschematic boys, a positive relationship was evidenced between how stereotypic the target was perceived to be and how much he was liked. For sixth grade children this trend was reversed. The more stereotypic a male target was perceived to be, the more the schematic boys liked him, whereas no such relationship existed for the aschematic boys. These findings are in line with the results of the cognitive inferen ces suggesting that for young schematic children, the gender label is the most salient feature, and therefore individuating information does not play a part in their preferences and judgments of the target. Consequently, young schematic boys both liked and preferred the same-sex target, i.e., the male target, regardless of whether his interests were traditionally masculine or feminine. Since for aschematic kindergarten boys the gender label is less salient, they, more than schematic kindergarten boys, use information about the target's interests in their liking judgments. On the other hand, our results show that older schematic children (sixth graders) do rely on information regarding the target's interests when making liking judgments. These results are in accordance with the results of the cognitive inferences that show that for older schematic children, individuating information is the most salient feature. Therefore, schematic sixth graders, but not aschematics, base their preference of the target on his interests and not only on his sex.

Different results emerged for the female target. For the younger age groups (kindergarten and third grade children), the relationship between the degree to which the target was liked and whether the target was perceived to be stereotypic existed only for the schematic children. Schematic boys preferred the stereotypic girl the least and schematic girls preferred her the most. In contrast, this trend was reversed for sixth grade boys, where schematic boys preferred the stereotypic girl the most. It seems that for the younger children, the more schematic they are, the more salient the target's sex is to them and it is the target's sex that is the primary determinant as to whether he or she is perceived as belonging to the subject's in-group or out-group. The results suggest that while kindergarten and third grade schematic children clearly prefer their in-group, they still take individuating information into consideration and therefore boys prefer the feminine girl the least. The more schematic they are, the m ore boys dislike a feminine girl. In contrast, schematic sixth grade boys are entering the developmental stage of adolescence, a stage characterized by an increasing interest in the other sex, thus making its members more appealing. Therefore, the more schematic they are, the more they like the stereotypic feminine girl. It is also possible that at this age, the target's sex is not the most important feature, but, rather, how stereotypically the target behaves, whether the target is a boy or a girl. Therefore, the more stereotypically a female target behaves, the more she is preferred, even though she is a member of the out-group. Interestingly, researchers found mixed results regarding whether boys show an increasing preference for stereotyped behavior as they grow older (Ruble & Martin, 1998). Ruble and Martin (1998) suggest that these conflicting findings might be due to individual differences. Our results support this view and suggest that this increasing preference for stereotyped behavior is more pronou nced for schematic boys.

In conclusion, although the importance of individual differences in gender schema elaboration has been recognized (Ruble & Martin, 1998), most studies have focused on age differences. The current study indicates the greater complexity of the issue and suggests that individual differences in gender schematicity should also be considered. Our results suggest that there is a difference between cognitive and emotional--motivational judgments. They further suggest that at a young age when the schema is less elaborated, the more schematic a child is, the more he or she will rely on the gender label. At an older age, when the schema is more elaborated and complex, the more schematic the child is, the more he or she will rely on the individuating information. It should be noted that these results were reached in Hong Kong, a collectivistic culture. In such a culture, individuals are expected to behave according to the social norms, including the gender role norms. Therefore, any transgression from the norm is less a ccepted. In contrast, in the Western cultures of the 1990s, there is a movement toward androgyny and individuals are encouraged to express not only characteristics of their own sex, but those of the other sex as well. One can speculate that children in individualistic societies such as the United States will be more flexible in their inferences and judgments. Consequently, the relationship between gender schematicity and inferences and judgments might differ in individualistic cultures more than in collectivistic ones, especially for older children since they are more likely to have internalized their culture's expectations. The fact that there were such clear differences between the inferences and judgments made by schematic and aschematic children even in a collectivistic society such as Hong Kong lends additional support to our conclusion that individual differences in gender schematicity play an important part in information processing. Future studies should examine the questions raised in this study with in Western individualistic cultures.

(1.) To whom correspondence should be addressed at Department of Psychology, Tel Aviv University, Ramat Aviv, Israel 69697, e-mail: talma@freud.tau.ac.il.

REFERENCES

Ames, P. T., Dissanayake, W., & Kasulis, T. P. (1994). Self as person in Asian theory and practice. Albany, NY: State University of New York Press.

Archer, J. (1991). A methodological commentary on gender schema research. British Journal of Social Psychology, 30, 185-188.

Bargh, J. A., Bond, R. N., Lombardi, W. J., & Tota, M. E. (1986). The additive nature of chronic and temporary sources of construct accessibility. Journal of Personalty and Social Psychology, 50, 869-878.

Bargh, J. A., & Them, R. (1985). Individual constructs accessibility, person memory, and recall-judgement link: the case of information overload. Journal of Personality and Social Psychology.

Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354-364.

Berndt, T. J., & Helter, K. A. (1986). Gender stereotypes and social inferences: A developmental study. Journal of Personality and Social Psychology, 50, 889-898.

Biernat, M. (1991). Gender stereotypes and the relationship between masculinity and femininity: A developmental analysis. Journal of Personality and Social Psychology, 61, 351-365.

Blakemore, J. E., LaRue, A. A., & Olejnik, A. B. (1978). Sex appropriate toy preference and the ability to conceptualize toys as sex-role related. Developmental Psychology, 15, 339-340.

Bond, M. H., & King, A. Y. C. (1986). Coping with the threat of Westernization in Hong Kong. International Journal of Intercultural Relations, 9, 351-364.

Carter, D. B., & Levy, G. D. (1988). Cognitive aspects of early sex-role development: The influence of gender schemas on preschooler's memories and preferences for sex-typed toys and activities. Child Development, 59, 782-792.

Carter, D. B., & Levy, G. D. (1991). Gender schemas and the salience of gender: Individual differences in non-reversal discrimination learning. Sex Roles, 25, 555-567.

Chiu, C. (1988). The effects of personality and performance on reward allocation. Journal of Social Psychology, 128, 279-280.

Chiu, C. (1989). Reward allocation among Chinese high school students in Hong Kong. Journal of Social Psychology, 129, 751-757.

Chiu, C., & Hong, Y. (1992). The effects of intentionality and validation on individual and collective responsibility attribution among Hong Kong Chinese. Journal of Psychology, 126, 291-300.

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for behavioral sciences (2nd ed.). Hillsdale, NJ: Eribaum.

Deaux, K., Kite, M. E., & Lewis, L. L. (1985). Clustering and gender schemata: An uncertain look. Personality and Social Psychology Bulletin, 11, 387-397.

Deaux, K., & Lewis, L. L. (1984). Structure of gender stereotypes: Interrelationships among components and gender label. Journal of Personality and Social Psychology, 46, 991-1004.

Dunlap, W. P., & Kemery, E. R. (1987). Failure to detect moderating effects: Is multicollinearity the problem? Psychological Bulletin, 102, 418-420.

Fagot, B. I., & Leinhach, M. D. (1989). The young child's gender schema: Environmental input, internal organization. Child Development, 60, 663-672.

Farrar, M. J., & Goodman, G. S. (1990). Developmental differences in the relations between scripts and episodic memory: Do they exist? In R. Fivush & J. Hudson (Eds.), Knowing and remembering in young children (pp. 30-64). Cambridge: Cambridge University Press.

Graesser, A. C., Gordon, S. E., & Sawyer, J. D. (1979). Recognition memory for typical and atypical actions in scripted activities: Tests of a script pointer + tag hypothesis. Journal of Verbal Learning and Verbal Behavior, 18, 319-332.

Harter, S., & Buddin, B. J. (1987). Children's understanding of the simultaneity of two emotions: A five-stage developmental acquisition sequence. Developmental Psychology, 23, 388-399.

Hartup, W. W. (1989). Social relationships and their developmental significance. American Psychologist, 44, 120-126.

Higgins, E. T. (1981). Rote taking and social judgment: Alternative developmental perspectives and processes. In J. H. Flavell & L. Ross (Eds.), Social cognitive development: Frontiers and possible futures. Cambridge: Cambridge University Press.

Higgins, T. E. (1996). Knowledge activation: Accessibility, applicability, and salience. In T. E. Higgins & A. W. Kruglanski (Eds.), Social psychology; Handbook of basic principles. New York: Guilford Press.

Higgins, E. T., & Bargh, J. A. (1987). Social cognition and social perceptions. Annual Review of Psychology, 38, 39-425.

Higgins, E. T., & King, G. (1981). Accessibility of social constucts: Information-processing consequences of individual and contexual variability. In N. Cantor & J. F. Khilstrom (Eds.), Personality, cognition and social interaction. Hillsdale, NJ: Erlbaum.

Hsu, F. L. K. (1985). The self in cross-cultural perspective. In A. J. Marsella, G. DeVos, & F. L. K. Hsu (Eds.), Cultural and self (pp. 24-55). New York: Tavistock.

Katz, P. S., & Boswell, S. (1986). Flexibility and traditionality in children's gender roles. Genetic Social and General Psychology Monographs, 112, 105-147.

Katz, P. S., & Ksansnak, K. R. (1994). Developmental aspects of gender role flexibility and traditionality in middle childhood and adolescence. Developmental Psychology, 30, 272-282.

Keyes, S. (1983). Sex differences in cognitive abilities and sex role stereotypes in Hong Kong Chinese adolescents. Sex Roles, 9(8), 853-870.

Keyes, S. (1984). Measuring sex-role stereotypes: Attitudes among Hong Kong Chinese adolescents and the development of the Chinese sex-role inventory. Sex Roles, 10, 129-140.

Kohlberg, L. (1966). A cognitive-developmental analysis of children's sex role concepts and attitudes. In E. E. Maccoby (Ed.), The development of sex differences. Stanford, CA: Stanford University Press.

Kondo, D. (1990). Crafting selves: Power, gender, and discourses of identity in a Japanese work place. Chicago: University of Chicago Press.

Kuhn, D., Nash, S. C., & Brucken, L. (1978). Sex role concepts of two- and three-year-old children. Child Development, 49, 445-451.

Lau, S. (1988). The value orientation of Chinese university students in Hong Kong. International Journal of Psychology, 23, 583-596.

Lau, S., & Wong, A. K. (1992). Value and sex-role orientation of Chinese adolescents. International Journal of Psychology, 27, 3-17.

Lebra, T. S. (1976). Japanese patterns of behavior. Honolulu, HI: East-West Center.

Levy, G. D. (1994). High and low gender schematic children's release from proactive interference. Sex Roles, 30, 93-108.

Levy, G. D., & Carter, D. (1989). Gender schema, gender constancy, and gender-role knowledge: The role of cognitive factors in preschooler's gender-role stereotype attribution. Developmental Psychology, 25, 444-449.

Liben, L. S., & Signorella, M. L. (1993). Gender schematic processing in children: The role of initial interpretations of stimuli. Developmental Psychology, 29(1), 141-149.

Linville, P. W., Fischer, G. W., & Salovey, P. (1989). Perceived distributions of the characteristics of in-group and out-group members: Empirical evidence and computer simulation. Journal of Personality and Social Psychology, 57, 165-188.

Lobel, T. E. (1994). Sex typing and the social perception of gender stereotypic and nonstereotypic behavior: The uniqueness of feminine males. Journal of Personality and Social Psychology, 66, 379-385.

Lobel, T. E., Bempechat, J., Gewirtz, J., Shoken-Topaz, T., & Bach, E. (1993). The role of gender-related information and self-endorsement of traits in preadolescents inferences and judgments. Child Development, 64, 1285-1294.

Maccoby, E. E. (1988). Gender as a social category. Developmental Psychology, 24, 755-765.

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Bulletin, 98, 224-253.

Martin, C. L. (1989). Children's use of gender-related information in making social judgments. Developmental Psychology, 25, 80--88.

Martin, C. L. (1993). New directions for investigating children's gender knowledge. Developmental Review, 13, 184--204.

Martin, C. L., & Halverson, C. F. (1981). A schematic processing model of sex typing and stereotyping in children. Child Development, 52, 1119--1132.

Martin, C. L., & Halverson, C. F., Jr. (1987). The role of cognition in sex roles and sex typing. In D. B. Carter (Ed.), Current conceptions of sex roles and sex typing: Theory and research (pp. 123--137). New York: Praeger.

Martin, C. L., & Little, J. K. (1990). The relation of gender understanding to children's sex-typed preferences and gender stereotypes. Child Development, 61, 1427--1439.

Martin, C. L., Wood, C. H., & Little, J. K. (1990). The development of gender stereotype components. Child Development, 61, 1891--1904.

O'Keefe, E. S., & Hyde, J. S. (1983). The development of occupational sex-role stereotypes: The effects of gender stability and age. Sex Roles, 9, 481--492.

Papalia, D. E., & Tennet, S. S. (1975). Vocational aspirations in preschoolers. Sex Roles, 1, 197--199.

Rubin, K. H. (1998). Social and emotional development from a cultural perspective, Developmental Psychology, 34, 611--615.

Ruble. D. N., & Martin. C. L. (1998). Gender development. In W. Damon and N. Eisenberg (Eds.), Handbook of child psychology, Vol. 3: Social, emotional and personality development (5th ed., pp. 933--1016). New York: Wiley.

Ruble, D. N., & Stangor, C. (1986). Stalking the elusive schema: Insights from developmental and social-psychological analyses of gender schemas. Social Cognition, 4, 227--261.

Schank, R. C., & Abelson, R. (1977). Scripts, plans, goals and understanding. Hillsdale, NJ: Erlbaum.

Shweder, R. E. (1990). Cultural psychology--what is it? In J. W. Stigler, R. E. Shweder, & C. Herdt (Eds.), Cultural psychology: Essays on comparative human development. Cambridge: Cambridge University Press.

Signorella, M. L. (1987). Gender schemata: Individual differences and context effects. In L. S. Liben & M. L. Signorella (Eds.), Children's gender schemata: New directions in child development (pp. 23--37). San Francisco: Jossey-Bass.

Signorella, M. L., Bigler, R. S., & Liben, L. S. (1993). Developmental differences in children's gender schemata about others: A meta-analytic review. Developmental Review, 13(2), 147--183.

Smith, E. R., & Zarate, M. A. (1992). Exemplar-based model of social judgement. Psychological Review, 99, 3--21.

Spence, J. T. (1984). Masculinity, femininity and gender related traits: A conceptual analysis and critique of current research. In B. A. Maher & B. Maher (Eds.), Progress in Experimental Personality Research, 13, 1--97.

Spence, J. T. (1993). Gender related traits and gender ideology: Evidence for a multifactorial theory. Journal of Personality and Social Psychology, 64(4), 624--635.

Stangor, C., & McMillan, D. (1992). Memory for expectancy-congruent and expectancy-incongruent information: A review of the social and social developmental literature. Psychological Bulletin, 111, 42--61.

Sun, L. (1983). The deep structure of the Chances culture. Hong Kong: Yi Shan.

Thompson, S. K. (1975). Gender labels and early sex-role development. Child Development, 46, 339--347.

Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96, 506--520.

Turner, P. J., & Gervai, J. (1995). A multidimensional study of gender typing in preschool children and parents: Personality, attitudes, preferences, and cultural differences. Developmental Psychology, 31(5), 759--779.

Yau, J., & Smetana, J. G. (1996). Adolescent--parent conflict among Chinese adolescents in Hong Kong. Child Development, 67, 1262--1265.

COPYRIGHT 2000 Plenum Publishing Corporation
COPYRIGHT 2001 Gale Group