STRUCTURE AND UNDERLYING VALUE OF PREFERABLE GENDER CHARACTERISTICS AMONG THE STUDENTS OF BELGRADE UNIVERSITY2

The aim of the study was to analyse preferable gender characteristics based on underlying gender stereotypes among the students from the University of Belgrade. The specific objective was to explore the prescribed value and the structure of those stereotypes. The participants were 261 students (69.7% female) from the University of Belgrade, 55.2% from Departments of social sciences and humanities and 44.8% from Departments of technical sciences, from the first to the fourth year of undergraduate studies. In the first phase of the study, we have used the list of eight pairs of desirable male and female prototypical attributes, and in the second phase, we have used the semantic differential scale for evaluative assessment of those attributes. The results have indicated that typical gender stereotypes are still predominant even in the academic environment. Stereotypes are more pronounced in the male sample than in the female one, i.e. women perceive the sexes as more similar to each other than men do. Both men and women evaluated the desirable “male” characteristics more positively than “female” ones, but men valued “female” characteristics significantly lower than women. We identified four groups of students with different structures of stereotypes. The predominant groups of stereotyping expressed masculinisation of both genders, or clear polarisation based on patriarchal tradition. If we consider students as the important strength for development of gender equality, more efforts should be made for understanding influence and planning policies, and programs targeting gender equality on faculties.

In social psychology, stereotypes are defined as structuralized schemas of beliefs about a particular social group (Hamilton, 1979).One of the most important groups of stereotypes refers to the characteristics of the sexes.Gender stereotypes, as cognitive schemas that organize perceptual processes and behavioural patterns (Bem, 1981), have the function of descriptive and proscriptive norms for the sexes.These norms specify desirable and appreciated types of behaviour for either sex, and can be used as a guide towards those behavioural patterns that are expected to be effective in particular situations (Cialdini & Trost, 1998;Eagly, Wood, & Diekman, 2000).Their effect is especially pronounced in ambiguous or confusing situations, in which people can follow these norms in order to behave in a way typical of their sex.Such norms dictate types of behaviour that are likely to elicit approval of other people and provide a personal sense of pride or, if not followed, they cause shame.Notions for religious or cultural discourses are often quoted as "proof" that these traditional roles are "natural" (Lynch & Nowosenetzb, 2009).These norms are denoted through discourse as a system of statements that construct an object and represent a set of meanings, metaphors, representations, stories, or statements that produce a particular version of events, (Parker, 1992).Gender identity, as a part of child's identity, is formed gradually, beginning with the understanding of speech, and within a specific discourse.The contested concept of a "genderrole identity" is usually used to describe the extent to which an individual identifies with and conforms to the cultural, locally accepted standards of femininity or masculinity.In a given society, determination of gender roles specifies types of behaviours, beliefs and attitudes that males and females are expected to have and exhibit (Basow & Rubin, 1999;Kessler & McKenna, 1978;Wood, Christensen, Hebl, & Rothgerber, 1997;Wood & Eagly, 2002).
Traditional construction of "masculinity" and "femininity" usually define men as "providers", who should maintain existential and financial security of the family, while women are defined as "caregivers" who should perform domestic chores in the household and take care of the children.Over the years, this has changed in modern society, and in the public and professional sphere, women's role has shifted more in the "masculine" direction.At the same time, the construction of men's role has changed less.For example, the possibility of a role reversal, such as the possibility of a man being financially dependent on a woman is hardly acceptable for men (not even approved by women), and they usually represent the opinion of how a change in traditional roles would be "unnatural".In transitional societies, as is the case in Serbia, transition from a traditional society and values to the modern ones is more likely to be reflected in the differences between gender roles in the private sphere (a partner and family relations), and less so in the public sphere (Radović, 2014).
Nevertheless, the impact of gender stereotypes on all domains of a society continues to be significant, albeit less overt.Perception of gender differences is a matter of comparison and is context dependent.For example, men are usually used as the "norm" in the domain of leadership, due to the spontaneous posiprimenjena psihologija, str.395-414 tioning of higher-status groups as the standard for comparison.Furthermore, the group positioned as the norm is perceived as more powerful.(Bruckmüller, Hegarty, & Abele, 2012).
There are numerous theoretical paradigms for the study of gender differences in psychology (Mitrović & Trogrlić, 2014), from evolutionary and psychoanalytic perspective, theories of gender socialization (Bussey & Bandura, 2004) to the cognitive developmental theories like Theory of Gender Schemas (Bem, 1981) and Theory of Social Role (Cialdini et al., 1998;Eagly, Wood, & Diekman, 2000).For example, Theory of Social Role emphasizes that gender hierarchy based on physical differences influences a position in the labour market which empowers the status of men.That produces different expectations about the psychological characteristics of bothgenders, and their both expected desirable behaviours in many social situations.Depending on how much a person deviates from the gender roles learnt through the process of socialization, he/she will be looked on with moreor less approval.Also, in the process of self-evaluation, the person will compare himself/herself with the perceived norms in his/her environment (Radović, 2014).
Studies of gender stereotypes are usually based on scales that are designed to measure gender characteristics used for descriptions of both sexes.A most popular study is the Sex Role Inventory (Bem, 1974).These scales include both female and male characteristics that are generally defined as gender-appropriate in Western cultures (Absi-Semaan, Crombie, & Freeman, 1993;Rossi & Rossi, 1990).Items tapping into female characteristics usually include the following attributes: considerate, emotional, affectionate, compassionate, tender, sensitive to the needs of others and eager to help others, while masculine include: self-confident, competitive, like to be in charge of things, ambitious, self-reliant, dominant and mastery-oriented (Brutsaert, 2006;Williams & Best 1990).

Gender Stereotypes in University Setting
Being tasked to provide higher education and conduct scientific research in each community, university setting has a privileged position in the complex process of forming future generations' attitudes, therefore having a special responsibility.Research within the European Union countries has shown that a the main risk factor in the process of achieving gender equality in the academic context is the existence of gender bias and stereotypes, which points to the need for gender sensitization of the teaching contents and methods, as well as to the lack of gender role models outside the dominant stereotypes, i.e. the presence of small numbers of women in the leading academic positions (Cacace, 2009).
Despite an appreciation for the need to increase gender sensitivity and awareness among the students, there is a lack of research in Serbia that ex-plores how they construct gender.A greater understanding of such constructions can assist in transforming gender relations and creating a more gendersensitive learning environment for students.A discourse analysis identifies various restrictive discourses that resist female participation in the academic setting and value traditional gender roles for men and women, having a tendency to marginalize voices that support gender equity.Research conducted internationally indicates that female students are still generally under-represented in higher education in the fields of science, engineering and technology (Bebbington, 2002;National Science Board, 2008).They account for 31.1% of science, engineering, technology and mathematics (STEM) graduates in the UK, and 33.5% of them in the USA (Langen & Dekkers, 2005).Disparities in female participation are particularly pronounced at graduate and postgraduate degree level.For example, in the US female students account for less than 29% of doctoral enrolments in SET, with only 20% of engineering doctoral degrees awarded to female students (National Science Board, 2008).Dominant public, as well as academic discourses, often construct the field of SET in masculine terms and often depict the field as being more appropriate for men than for women.Such constructions of SET can impact on how both male and female SET students construct their professional identities and may affect their understanding of how gender operates within the public, professional and domestic environments.
It is obvious that gender differences in access to higher education have decreased significantly in many European countries.Although women are currently over-represented in upper secondary education, and more frequently attain a general qualification for university entrance, research in Germany has showen that they still enrol in higher education less frequently.Even then, their choices of study field differ from those of men (Lörz, Schindler, & Walter, 2011).
These findings have also been confirmed at the University of Belgrade.Although women constitute the majority of students in the fields of social sciences and humanities, while men are predominant in technical sciences, when the distribution of teaching staff is observed, one can notice the predominance of men at all Departments.Differences between men and women in terms of motivation for studying have shown that girls place greater interest in certain professions based on the opportunity to help others, while the prospect of employment is more important for men (Džamonja, Žegarac, Popović, & Duhaček, 2009).These findings testify to the difference in the social power of men and women and reflect traditional stereotypical gender roles, as well as the selfstereotyping of "caring women", where the male role includes financial power.Regarding the attitudes toward participation of the sexes in different social roles, the results of this research suggest conflation of gender roles among students.This is reflected in the acceptance of equal contribution to a family budget, inclusion of both sexes in housework, childcare and caring for elderly or ill family members.Still, men express such convictions with less intensity, while the traditional gender roles in partnerships show greater resilience to change than the attitudes toward family roles.
In order to explore gender stereotypes in academic settings, the research of the student population has been conducted in Serbia (Džamonja, Duhaček, & Popović, 2010).The results of the research has showen that the prototype of the desirable female profile is characterized by success, self-confidence, rationality, focus on family, sociability and education from the standpoint of both sexes, although female students tend to emphasize these characteristics more than the male ones.While women see themselves as more rational, self-confident and capable, compared to emotional, tolerant and attractive, men have ambivalent opinion about which of those characteristics are more favourable for women.In male perception of an ideal woman, she should be tender rather than strong, while women perceive strength as a more desirable female attribute.Also men more frequently think that it is more desirable for women to be focused on the family rather than socially involved, while women have a split opinion in respect to those characteristics.Approximately half of them believe that it is more desirable to be family oriented, whereas the other half preferred that women should rather be socially involved.The prototype of the desirable male profile is composed of the attributes of success, strength, self-confidence, rationality, equal focus on work and family, sociability and education.Respondents of both sexes agree about this image, except that a higher percentage of female respondents choose emotionality as desirable, although even among them there is a dominant preference for rationality, albeit by a significantly lower percentage than men.
While such attributes and expected behaviour may be more or less preferable for different sexes, the mere existence of gender stereotypes does not say anything about the value of those stereotypes.In other words, people can have different perceptions of the characteristics of the same and the opposite sex, but that does not necessarily imply that these perceptions are evaluated unequally.In addition, importantly, the contribution of this research is precisely evaluative assessment of gender stereotyping.For example, people can prescribe that men should be competent and women should be tender, but although both attributes are positive, somebody can evaluate the competence ascribed to men as more valuable than tenderness.For that reason, the assessment of stereotypes should encompass both the conceptual and the evaluative component.
Based on the idea that the difference in the profiles of typical man and woman is by itself insufficient to conclude whether there is a difference in evaluating these profiles, we have developed a short instrument based on descriptive stereotypes in order not only to assess "content" of male and female attributes among students, but also to identify how they value different gender characteristics.Also, our intention has been to explore if there are specific groups of students with qualitatively different structures of gender stereotypes.

Goals and objectives
The main goal of this study was to examine closely preferable gender characteristic as specific gender stereotypes among students of University of Belgrade.We were also interested in the relation between the academic setting and gender stereotypes, and its potential role in changing these stereotypes.Therefore, we proposed specific objectives of the study: 1.To examine how students value preferable gender characteristics, and if there is any difference between male and female students in evaluating those stereotypes.For example, if self-confidence is ascribed more to one sex, and tolerance to another, it is important to know which attribute in the respondents' evaluative space is valued more highly, regardless of which sex it is ascribed to.If both attributes are valued equally, then there is no evaluative difference regardless of the conceptual one.2. To identify different structures of preferable gender characteristics.We have hypothesized that there are specific, qualitative differences among the structure of students' gender stereotypes, differences which underlie a preference of gender attributes.Therefore, if we focus just on the average profile of men and women, differences among the structure of their attitudes would be blurred.

Method Sample
Participants were 261 students (69.7% female) from 4 faculties of University of Belgrade: Faculty of Political Sciences, , Faculty of Philosophy, Faculty of Mechanical Engineering, and Faculty of Architecture.They were divided in two groups of 55.2% from social sciences and humanities (Faculty of Political Sciences and Faculty of Philosophy) and 44.8% from technical sciences (Faculty of Mechanical Engineering and Architecture).The respondents' age was between 18 and 31 (M = 20.48,SD = 1.72).There were 24.5% of the first year students, 49.0% of the second year students, 11.1% of the third year students, and 14.9% of the fourth year students.

Instruments
Masculinity-femininity stereotype list.The technique of prototype profile was used in order to identify male and female profiles (Appendix A).Gender stereotypes were measured with Masculinity-femininity stereotype list of 8 pairs of socially desirable, positive attributes (e.g."successful" as opposed to "stylish") that were created based on prototypical male and female characteristics identified from the previous research (Džamonja et al., 2010).Each pair had to contrast one "male" vs. one "female" attribute.The task was force choice type, so participants had to choose which attribute from each pair was more desirable as a male and a female characteristic.If a male characteristic was chosen, a score for that pair was 1, and if a female characterstic was chosen, the score for that pair was -1.The participants separately assessed preferable characteristics for men and for women.The total score on this scale had a range from -8 to +8, where -8 denoteed the profile of maximum femininity, +8 denoteed the profile of maximum masculinity, and 0 suggested equally represented masculine and feminine characteristics.
Masculinity-femininity value scale.In order to obtain an evaluative dimension of male and female prototype profiles, we have created Masculinityfemininity value scale.The scale has the same pairs of attributes as the Masculinity-femininity stereotypes list, but it is constructed as a 5-point semantic differential scale (from -2 to +2, in which the midpoint (0) signifies equal desirability of the attributes).Respondents have evaluated which attribute in a pair is more desirable in general (regardless of which sex it is attributed to).The total evaluative score for each profile has a range from -16 to +16, depending on the assessed desirability attributes chosen in profile, where -16 represents maximum devaluation of the profile, and +16 is maximum valuation of the profile.

Procedure
The Masculinity-femininity stereotype list and the Masculinity-femininity value scale were given to students in two phases by researchers.The Masculinity-femininity stereotype list was given first, and one month later, the Masculinity-femininity value scale was given to the same group of participants.They were asked to complete the instruments anonymously, under a specific code though, so that we could match participants from two phases of testing.The scale was administrated in groups, during the regular exercise by teaching assistants of both sexes, to minimise the influence of administrator's sex.Since the students were examined during regular classes, the sample was convenient, but there was no reason to suppose that gender structure was affected.

Data analysis
First, the difference between male and female profiles was examined.Two-way ANOVA was used in order to compare differences between men and women in their assessment of masculinity-femininity stereotypes, and the value they ascribed to the male and female profile.Second, Latent Class Analysis -LCA was applied in order to examine subtypes of gender perception.LCA is a statistical method commonly used to test the existence of the discrete groups of cases with similar characteristics based on multivariate data.LCA has no requirements for the distribution of data, offering robust statistical tests for determining the optimal number of classes and classification accuracy (Nylund, Asparouhov, & Muthén, 2007).LCA is a widely used latent structure model for categorical and interval data, but it has been seldom applied to gender studies, although it may be especially useful for examining gender stereotypes.
Model selection was based on Bayesian Information Criterion (BIC) and Lo-Mendell-Rubin's adjusted likelihood ratio test, while measure of entropy was used to assess the accuracy of classification.Afterwards, MANOVA was applied in order to identify if differences existed between latent classes in terms of perception and value of profiles, and masculinity-femininity score.Statistical analysis was conducted by SPSS 19 and Mplus software.

Relation of gender and choice of studies
A significant relation was found between sex of participants and choice of studies, χ 2 (1, N = 261) = 40.83,p < .001,Cramer's V = .40.While women were the majority in humanities and social sciences (86.1%), they were minority at the Faculty of Mechanical Engineering (10.5%).It is interesting that there is prevalence of women at the Faculty of Architecture (68.4%) which is a part of technical and engineering disciplines.However, if architecture could be considered as an art, that is not such a surprise.Gender structure of the faculties confirms the assumption that students generally prefer the studies that are more "expected" for their sex.

Preferable gender attributes
In order to check whether there was a difference in perception of male and female students regarding masculinity-femininity as preferable characteristics for genders, we employed 2 (sex of students) x 2 (gender preferable profile) mixed analysis of variance (two-way ANOVA), with sex as the between-subjects variable.As expected, the analysis confirmed the existence of the main effect of the profile, F(1, 259) = 123.78,p < .001,η 2 = .32.The interaction effect was also significant, F(1, 259) = 80.65, p < .001,η 2 = .24,with male students showing a greater preference for masculinity as shown in Figure 1.It was interesting that women perceived the genders as more similar (in more masculine terms), while men perceived greater differences between genders.They tended to see women as much more feminine then the women saw themselves.
primenjena psihologija, str.395-414 In order to check whether there was a difference between men and women in students' valuation of male and female characteristics, we employed 2 (sex of students) x 2 (gender profile) mixed analysis of variance (two-way ANOVA), with sex as the between-subjects variable.This analysis showed almost equal effects to the previous analysis.The analysis confirmed the existence of the main effect of the profile, F(1, 233) = 151.33,p < .001,η 2 = .39,with the male profile being more valued than female.The interaction effect was significant, F(1, 233) = 80.65, p < .001,η 2 = .23.Men perceived a greater difference between the value of male and female characteristics as shown in Figure 2. The analysis of the evaluative dimension showed that the female profile wassignificantly less valued than the male profile.This tendency was present both among men and women, i.e. both men and women valued the male profile higher.Still, the difference in valuing different profiles was much greater among men than among women, and this difference contributeed to the intensity of the main effect of the profile, i.e. to the difference in evaluating preferable profiles.

Classes of gender stereotypes
We expected that there were specific structural differences among the students' gender stereotypes.To identify if there were different types of gender stereotypes, we applied latent class analyses -LCA.LCA was conducted on 16 (from 8 pairs) of socially desirable characteristics of men and women.Based on Lo-Mendell-Rubin's adjusted likelihood ratio test -LRM (Table 1) the 4-class solution was identified as superior to the 3-class solution and 5-class solution.Also the 3-class and 5-class solution did not contribute to accuracy of classification (entropy) in regard to the 4-class solution.Based on this, the 4-class solution was identified as best fitting the data.Results of a MANOVA suggested that there was a significant difference between members of the latent classes in terms of masculinity-femininity stereotypes, Wilks' Lambda = .32,F(6, 512) = 65.83,p < .001,and in terms of valuing profiles, Wilks' Lambda = .50,F(6, 460) = 31.37,p < .001,as shown in Table 2. Post hoc comparisons using the Bonferroni test indicateed that, regarding the differences in masculinity-femininity score for the male profile in preferable attributes, there were significant differences between all classes, except between Class 1 and Class 2, as well as Class 3 and Class 4. Regarding the value assigned to male profile, post hoc comparisons showed significant differences between all classes, except between Class 1 and Class 2. Regarding the differences in masculinity-femininity score for female profile in preferable attributes, there were significant differences between all classes.Regarding the value assigned to female profile in valuation, post hoc comparisons showed significant difference between all classes, except between Class 1 and Class 4.
Based on the results of MANOVAs and descriptive statistics of each class profile we can describe identified classes as following: • Class 1 -"Masculinization" of both genders.Preferable characteristics of men and women are seen as highly masculine, and women are less valued than men.Both genders are more frequently described as successful, strong and capable vs. good looking, tender and well dressed, as well as oriented toward social and professional spheres, as opposed to being focused on the family.• Class 2 -Traditional polarisation of gender stereotypes.According to the attitudes of this class, men should be more masculine, whilewomen should be more feminine.Men are masculine in each aspect, slightly more than in Class 1.However, students from this class have almost equalised the importance to be family and socially involved.Women are perceived as family oriented, tender and more extroverted than men, but also as edu-cated and successful.So, it could be said that this traditional polarization is softened or less sharpened.• Class 3 -"Feminization" of both genders.This group perceive that women and men should be predominantly focused on the family and tender, but also rational, capable and extroverted.However, even in this group, men are more valued and perceived as slightly more successful.• Class 4 -"Mixture" of gender characteristics.In this group, women are seen as more successful and self-confident, but also as focused on the family.Men are perceived as focused on the family as well, but also as emotional and tolerant.Both sexes share characteristics such as being successful, extroverted and educated, so we can speak about a mixture of gender characteristics that could be less essential or prescribed for both genders.
Results of a chi-square test suggested a strong relationship between membership in classes and gender, χ 2 (3, N = 261) = 36.66,p < .01,Cramer's V = .38.If we look at the distribution of male students by classes, it can be seen that men predominantly belong to Class 2, which represents the traditional polarisation of gender stereotypes, and to some extent to Class 1, which represents masculinisation of both genders (Table 3).There is no such clear polarisation between women.In relation to men, a greater number of women belong to Class 1 and Class 4, and less to Class 2.

Discussion
The results indicate that there are gender stereotypes in the academic environment endorsed by both women and men.Stereotypes are more pronounced among the men, which means that women perceive genders as more similar to each other than the men do.Beliefs held by men and women about the desirable gender characteristics indicate that female attributes approach the image of preferred male attributes in terms of rationality, sociability, success and education.Yet, these attributes do not lessen the importance of the "traditional" ones, namely that a woman should be tolerant rather than self-confident, and focused on the family rather than being socially involved.The traditional image of woman is more pronounced among men, who, unlike women, prefer appeal and tenderness, rather than capability and a woman's strength of personality.The previous study (Džamonja et al., 2010), has shown that the attributes associated with women's passive position (depression, reserve, conformism, passivity, dependence, insecurity) are dismissed as undesirable for the respondents of both sexes, while women dismiss these attributes more strongly.
Male "macho" gender stereotype still persists as desirable, regardless of their more active participation in the family roles and a decreased preference for domination.Women evaluate certain more "active" attributes as desirable, as well.From the perspective of male students, it seems that these attributes are moderately desirable as female characteristics only if they do not endanger traditional gender attributes of women as tolerant, non-aggressive, non-dominant etc.
In this study we tried to develop the innovative approach to measure gender stereotypes.Besides a direct description of desirable characteristics, we also measured an evaluative dimension of them, regardless of which sex they were attributed to.Therefore, we wanted to go beyond the surface of the description that might be a reflection of just socially acceptable attitude, and tap into the semantic dimension of stereotypes.
In general, both men and women evaluate the desirable male profile highly positively.However, even though women evaluate the characteristics of their sex positively, they tend to perceive the characteristics of the male sex as more positive.On the other hand, men evaluate characteristics ascribed to women as significantly lower.Those differences in the preference of profiles are based on the perception of male characteristics as dominant and superior.
In Serbia, where masculine characteristics pose as the "ideal type", i.e. socially dominant and desirable forms of behaviour, they have the function of producing wide support for men and their dominant position.This is especially significant in the academic world, which still gives legitimacy to most social and political options.Both men and women strive for obtaining and maintaining their respective traditional social roles or masculinized roles.Greater presence of stereotypes among men is a consequence of the perpetuation of male dominance in the society, which may be a broad social explanation of why there is no direct motivation for changing this position, and by extension no motivation for substantial change in the social and political framework of our society.
Greater consensus among the respondents of both genders on the prototypical image of man, and relatively higher gender difference in description and evaluation of female attributes, testifies to the greater confusion regarding the female identity.It is obvious that women are in fact those who strive to change their position and roles by approaching the male prototype, while men prefer traditional differences.
The results have also confirmed different structures of the stereotypes.The stereotypes do not present a simple picture of two typical, opposite profiles of male and female characteristics, or an equal portrait of both genders.Four different classes of stereotype structure have been marked off, what could be seen as consistent with changes in a feminist perspective.The first one represents similarity of both genders according to the masculine type (Class 1 -the attitude more typical for women students) that is consistent with the first phase of feminism in which equality is seen in terms of women being "the same" as men.The next one (Class 2), represents a typical patriarchal view of gender differences that values masculine characteristics, that is predominant in our student sample (36.4%).The third class could be seen as a "feminization" of both genders, which is a trend in modern societies where it is desirable for men to be tender and focused on the family.This is the smallest group in our sample (8.8%).The last group (Class 4) represents "less prescribed" gender characteristics, that is in accordance with contemporary "anti-essentialist" approach (Butler, 1990;Riley, 1988;Spellman, 1988).
Generally speaking, the student population is still far away from antiessentialism in terms of gender identity, in that they are basically restricted in terms of individual choice and idiosyncratic forming of personal identity, remaining under the dominant traditional patriarchal culture.Although changes are evident, the situation still reflects polarization and different valuation of gender characteristics in favour of masculinity.When dominant traditional gender stereotypes have begun to fade, the trend of "masculinization" as preferable gender position begins to take place.If we suppose that students are usually the most progressive part of the community, we may speculate that Serbia is still a traditional patriarchal society.The empirical evidence thus support the theoretical debates in the gender studies field.
Although there are a lot of factors, such as a family, friends, media, commercials, etc., which all influence gender stereotyping and reinforce traditional gender stereotypes, still more attention should be paid to education policy in order to improve gender equality.Considering the results of our previous study that involved the content analyses of a sample of university handbooks which show a lack of sensitivity to gender issues (Baćević et al., 2010), universities should take a more active role in education and promotion of gender equality through development of gender sensitive policies and programs.
We want to emphasize the importance of linking gender equality policy to the education policy, particularly at higher levels, with the purpose of promoting positive social values and creating incentives for women to achieve professional fulfilment.,In turn, this will allow the society as a whole to utilize all of its human resources.

Figure 1 .
Figure 1.Interaction between sex of students and gender profile in perception of masculinity-femininity.

Figure 2 .
Figure 2. Interaction between sex of students and gender profile in valuation of male and female characteristics.

Table 2
Descriptive statistics for analysis of variance with masculinity-femininity score and valuation score as dependent variables

Table 3
Relationship between class membership and sex