FindArticles > Sex Roles: A Journal of Research > June, 2003 > Article > Print friendly
Positive and negative androgyny and their relationship with psychological health and well-being
Brenda Mae WoodhillGender role development is one of the central tasks of the developing person according to Erikson (1950). It is a central feature of our identity. Most people are masculine or feminine; some people are masculine and feminine (Pearson & Cooks, 1995). The social construction perspective on gender role identity asserts that social conditions prescribe major amounts of observable behavior (Eagly & Wood, 1999; Mead, 1935) Are sex, gender, and gender role identity different? Many authors use sex and gender interchangeably, and despite their enormous popularity, these basic constructs remain unclear to some and especially confusing in the nonprofessional world. In a study of how nonpsychologists define gender (in the New England area of North America) Pryzgoda and Chrisler (2000) showed that the majority of their participants used sex to describe what is biological and gender to describe what is social and psychological. Gender indicates something about socialized patterns of behaving, and sex refers to the biologic al differences between men and women. Therefore we do gender in a gendered world (Kimmel, 2000).
As children and as adults, people adopt characteristics and behaviors that display to the outside world which gender they are. Doing gender involves gender roles and gender displays that focus on the behavioral aspects of being a man or a woman. People express their gender through their social world in gendered display (West & Zimmerman, 2000), that is they adopt a gender role identity. Boswell (1997) suggested that no behavioral trait is intrinsically masculine or feminine, rather each society assigns different groupings of traits to each sex and leaves some as gender neutral. So people understand that if a person is female, she should behave in a feminine way; if a person is male, he should behave in a masculine way. We accept these distinctions and maintain that sex is not gender and that gender is not the whole sum of personality, but that masculinity and femininity can be measured by comparing the behavior of people.
If the sex of a person is determined biologically, then gender role identity is culturally and socially constructed. Gender does not equal behavior, although we suggest that people behave in a gendered way and that gendered behavior follows from gender role identity. Traditionally, instrumental or agentic behavior has been associated with being male, and expressive or communal behavior has been associated with being female. This has led to the building of the borders to separate gender, behavior, and personality, but it hardly seems correct to suggest that a women who acts instrumentally is not feminine and a man who acts expressively is not masculine.
The traditional psychological gender role identities of feminine and masculine are based largely on the child-bearing role of women, the provider role of men, and the assumed psychological differences between them. These traditional stereotypes have allowed for the acceptance of some nonadaptive behavior in men and women. Such notions as "women are not capable of being resourceful and assertive," and; men are neither vulnerable nor capable of compassion" are typical. Assumed psychological differences between the sexes have contributed to the hierarchy of gender and have been made the basis of many cultural prohibitions (Lewis, 1978). Gender inequality is blamed on assumed differences between the sexes, and these assumed differences are then used to justify gender inequality. Bem's ideology (Bem, 1974, 1975, 1998) aimed to promote an androgynous world where gender distinctions were reduced to a minimum and women and men shared equally the roles and characteristics usually associated with only one sex. Even tho ugh androgyny was once called the developmental ideal (Mead & Ignico, 1992), empirical support for its outcomes remains ambiguous.
Much of past research has concentrated on the relationships of a single category of androgyny and various measures of well-being. Positive outcomes for androgynous individuals have often been found (Antill, 1983; Campbell, Steffen, & Langmeyer, 1981; Cheng, 1999; Green & Kendrick, 1994; Kirchmeyer, 1996; Rose & Montemayor, 1994; Sawrie, Watson, & Biderman, 1991; Shaver, Papalia, Clark, & Koski, 1996; Shimonaka, Nakazato, Kawaai, & Sato, 1997; Stake, 1997; Wubbenhorst, 1994). Other studies have shown mixed results (Burchardt & Serbin, 1982; Lubinski, Tellegen, & Butcher, 1981, 1983; May & Spangenberg, 1997; O'Heron & Orlofsky, 1990; Skoe, 1995; Wulff & Steitz, 1999).
Researchers have also focused on the individual constructs of positive and negative masculinity and positive and negative femininity and again found confusing results (Desrochers, 1995; Marsh & Myers, 1986; Ricciardelli & Williams, 1995; Russell & Antill, 1984; Spence, Helmreich, & Holahan, 1979). As well, researchers have often found support for masculinity as the healthier gender identity (Antill & Cunningham, 1979,1980; Cook, 1985; Jones, Chernovetz, & Hansson, 1978; Kopper & Epperson, 1996; Markstrom-Adams, 1989; Powell & Butterfield, 1979; Whitley, 1983) especially when adopted by women (Anderson, 1986; Baril, Elbert, MaharPotter, & Reavy, 1989; Heilbrun, 1981a).
Traditional androgyny is seen as a gender role identity that consists of a balance of positive feminine and positive masculine traits. Androgyny is thought to be a balanced identity that combines the virtues of both genders. However, gender stereotypes do not include only desirable aspects of femininity and masculinity (Kelly & Worrell, 1977; Ricciardelli & Williams, 1995). Socially undesirable feminine and masculine traits are also important to gender stereotypes and may even be dominant (Spence et al., 1979). Logically then, an androgynous gender role identity may also consist of a balance of negative feminine and negative masculine traits, that is, an identity that combines the failings/defects of both genders and thus creates the possibility of an undesirable or negatively androgynous gender role identity. The presence of significant levels of negative feminine and negative masculine traits in androgynous individuals may have a detrimental effect on the androgynous gender role identity to the extent that such negative behaviors may override any of the positive benefits proposed for the androgynous person (Woodhill & Samuels, in press).
According to the traditional definition of androgyny (Bem, 1975), an androgynous person is more competent on a wide variety of tasks independent of with what sex the task may usually be associated. Nowhere in the literature has it been suggested that humans never behave in an undesirable fashion, therefore the androgynous personality may also include competence in undesirable behaviors or vices. For example a positively androgynous person may demonstrate high levels of independence (m+), compassion (f+), ambition (m+), and tolerance (f+), whereas a negatively androgynous person may demonstrate high levels of submissiveness (f-) and selfishness (m-), or be temperamental (f-) and aggressive (m-).
Androgynous people are sensitive to both masculine and feminine cues and as such may respond to a wider range of positive or negative stimuli than traditional people. A negatively androgynous person would have a bigger repertoire of undesirable behaviors from which to choose a response. A negatively androgynous person may, for example, react in an undesirable feminine way in one situation (e.g., submissively) and in an undesirable masculine way in another situation (e.g., aggressively). It becomes clearer that a single category of androgyny cannot capture these important differences because such a category by definition usually ignores gender associated "vices" (i.e., failings, deficits, or negative traits; Woodhill & Samuels, in press).
Some theorists (Helgesen 1994; Hoffman 2001; Spence & Buckner 2000) have objected to the use of the terms feminine and masculine in the definition of androgyny. They have argued that continued use of these terms reinforces socially sanctioned restrictions on behavior and may obfuscate results. Shaffer (2002) and Woodhill and Samuels (in press), on the other hand, argued that androgyny itself is associated with equality between the sexes and promotes equality of opportunity. In defence of the terms feminine and masculine we argue that the media and people generally are so preoccupied with them and that they are so entrenched in the professional and nonprofessional world, that it would be futile to attempt to remove them completely from discourse. Feminine and masculine are dominant and compulsory terms in our social fabric. Bem (1998) suggested it is better to work with them than to invent something new. However, there may be a case for their reassignment away from research based on gender role identity measur es. Spence (1984) argued that the BSRI and the FAQ are actually measures of behavioral traits and not of contemporary femininity and masculinity per se. She also argued for a separation of gender identity and behavior as studied in gender role research, and maintained that gender identity is an ineffable quality that is difficult to put into words (Spence, 1999). Perhaps the theory of intelligent design (Looy, 2001) can clarify the notion of gender. Where social construction theorists assert that psychological differences between the sexes are almost entirely the result of differential socialization, evolutionary psychologists assert that they are the result of the different reproductive roles evolved by each sex. Looy (2001) asserted that there may be a third possibility where a spiritual dimension is responsible to some extent; the purpose of being human is to form loving relationships, and there is more to being gendered than sexuality. The fundamental claim of this theory is that intelligent supernatural causes are necessary to explain the complex structures of biology. Looy (2001) suggested that an intelligent design approach might free us from endless attempts to define an ineffable feminine or masculine nature, presumably by having us accept as a given that there are two genders, that they preexist in people, that they are a part of the supernatural world, and that they are well matched and necessary for people to develop loving relationships.
Other criticisms of androgyny have been summarized by Matlin (2000), with the pragmatic difficulties of being multiskilled as prominent. However, androgyny is more correctly related to psychological traits and not physical skills per se (traits such as ambition and confidence, not behaviors such as wood chopping and cooking). Bem's early example (Bem, 1977) where an androgynous person may be capable of firing an employee if necessary but with a sensitivity for the human emotion that inevitably follows, helps us to understand this point. Androgyny is a call for the degendering of traits, not a call for the degendering of people. As others have called for a moving on beyond gender (Spence, 1999), we believe that research would be best served by a separation of gender from the study of androgyny and a definition of androgyny as a personality trait-based identity. For the facilitation of this study however, we will maintain the term "gender role identity."
Other critics have suggested that if we all adopted an androgynous identity, society would become unisexed and lacklustre (Hefner, Rebeca, & Oleshansky, 1975; Kimmel, 2000). These theorists seem not to make allowances for the fact that everyday social behavior and romantic behavior are necessarily different in order for people to maintain healthy relationships. Androgynous behavior is still valuable in romantic relationships in that the androgynous person is capable of giving way to the other or adopting whatever behavior the other is not.
Other pleas for a movement away from androgyny specify what is known as gender role transcendence. Gender role transcendence is a call for a dynamic and flexible orientation to life in which assigned gender is irrelevant (Hefner et at., 1975). Being positively or negatively androgynous can accommodate the call to transcend gender, to allow people to choose behaviors that are personally meaningful. We suggest that a differentiated androgyny is gender role transcendence.
The majority of gender role identity research to date has used either the Bem Sex-Role Inventory (BSRI; Bem, 1974) or the Personal Attributes Questionnaire (FAQ; Spence, Helmreich, & Stapp, 1974). The logic that humans have both desirable and undesirable traits has been extended into the constructs of feminine and masculine and into gender role identity scales or measures such as the Australian Sex Role Scale (ASRS; Antill, Cunningham, Russell, & Thompson, 1981) and the
Extended Personal Attributes Questionnaire (EPAQ; Spence et al., 1979) but not into the construct of androgyny. There are no gender role identity inventory methods that differentiate androgyny into positive and negative androgyny (Woodhill & Samuels, in press).
In this study desirable or positive androgyny was defined as both F+ and M+ scores above a mean, and undesirable or negative androgyny was defined as both F- and M- scores above a mean. These were included in a reassessment of relationships between androgyny and mental health and well-being. Psychological health is defined as a combination of mental health with social effectiveness (Marsh & Myers, 1986) and well-being.
We propose that a differentiated androgyny would yield separate results on measures of mental health and well-being: that positively androgynous people would have the highest levels of mental health and well-being, followed by positively masculine, positively feminine, negatively masculine, negatively feminine, and negatively androgynous people. The seventh category of undifferentiated androgyny (Au) was used to account for the cases that were higher than the mean on all scores for F+, M+, F-, and M-. The term "undifferentiated" is borrowed from Spence who used it to account for people with no clear preferences, that is, those whose M and F scores were both below the median (J. Spence, personal communication, February 21, 2000). Here it is used also to account for those who have no distinct preferences but have all four scores above the mean. No a priori prediction was made for undifferentiated androgyny.
METHOD
Participants
A power analysis with effect size medium (.25) and power conventional (.80) yielded a suggested sample size of 240 participants. Adult participants (N = 196), 154 women and 38 men were recruited from various sources around Australia. Following a brief description of the research, 134 psychology students from a regional Australian university, 31 members of a city choir, 13 clients of an adult relaxation centre, and 14 friends and associates of the first author from varied walks of life volunteered to complete a series of seven questionnaires. Ten respondents (5.1%) were under 20 years old, 40 (20.4%) were 21-30 years, 57 (29.1%) were 31-40 years, 43 (21.4%) were 41-50 years, and 42 (21.4%) were over 50 years.
The sample included 5 (2.6%) bisexual, 12 (6.1%) lesbian, and 4 (2%) gay participants. Excluded from the study were four participants: one transsexual because s/he recorded (albeit correctly) sex as male and gender as female; one man who had many extreme scores; and two others with incomplete questionnaires. Seventy-two participants (36.7%) stated that they were married, 44 (22.4%) single, 14 (7.1%) separated, 27 (13.8%) divorced, and 33 (16.8%) in unmarried partnerships. Two were widowed. There were a total of 17 different religious categories identified.
Measures
A set of self-report questionnaires was used in this research.
Gender Role Identity
The EPAQ (Spence et al., 1979) is one of the most common scales used to measure gender role identity. It is a 40-item questionnaire; each bipolar item is scored on a 5-point scale from 0 to 4. Each item is keyed as a positive or negative trait, a balance of items are keyed in the feminine and masculine dimensions. Traditionally, raw scores on each of the four categories, F+, F-, M+, and M-, are summed and compared to the medians for the whole group, which yields four categories of gender role identity: feminine, masculine, androgynous, and undifferentiated. In this study we used the EPAQ as modified by the interpersonal perception method (Schullo & Alperson, 1984) where the items are reduced to 32 by removal of the eight bipolar M/F items.
In order to make valid, statistical comparisons between the positive and negative raw scores, which were expected to be widely divergent, all scores were transformed into z-scores. Participants were then classified into one of seven gender role identity categories: positively androgynous (A+), negatively androgynous (A-), positively feminine (F+), negatively feminine (F-), positively masculine (M+), negatively masculine (M-), and undifferentiated androgynous (Au). Androgynous participants were determined by a relative balance of positive feminine and positive masculine qualities, or a balance of negative feminine and masculine qualities. Those with high scores (z-scores above zero) on both F+ and M+ were positively androgynous (A+), and those with high F- and M-scores were negatively androgynous (A-). (A high score was defined as a positive or negative pair of z-scores above the z-score of zero, that is, both of either F+ and M+ or F- and M- are above a standardised sample mean). The third type of androgynous group, with all four z-scores above zero, was classified as undifferentiated androgynous (Au). As the more negative a z-score gets the lower is its associated raw score, the remaining nonandrogynous participants were classified into either of F+, F-, M+, or M- categories according to which of their z-scores on these scales had the highest positive value.
The categorical method of classification according to scores relative to a mean, is arguably a less than perfect way of identifying gender role types. It is insensitive in that members within a category are treated as though they have the same scores, and it does not account for the variation expected in the strength of people's gender role identities. There is a need to find ways to identify the relative strengths of gender role identities and examine within-group differences, such as by the use of multiple regression, however there are also problems with regression. Spence (1984) stated that there is no theory that proposes that androgyny is an interactive concept, that is, where androgyny is some type of M x F effect. Lubinski (1981, 1983), Taylor and Hall (1982), and Spence (1984) found little evidence of any interaction effects and Spence (1984) stated that the evidence did not justify the conclusion that M and F are intrinsically interactive. Further, Jaccard, Wan, and Turrisi (1990) suggested that regr ession assesses any interaction best when relationships are linear, however relationships within levels of gender role identity are most likely curvilinear. Jaccard et al. (1990) suggested the formulation and evaluation of interaction effects be best guided by theory. The categorical method of classification is by far the most common to date (Lenney, 1991), and its use allows us to link our results to the larger body of literature; in addition, the categories provided by our method provided convenient groupings for examination.
Psychological Health and Well-Being
The Acceptance of Others Scale was developed by Fey (1955) as a measure of self-acceptance, acceptance of others, and feelings of acceptability to others. The scale consists of 20 attitude statements (e.g., "people these days have pretty low moral standards") with responses from almost always (scored as 1) to very rarely (scored as 5). Scores can range from low acceptance (20) to high acceptance (100). The mean score was 75.5 (SD = 8.4). Internal consistency is reported as .90 (Robinson, Shaver, & Wrightsman, 1991). There are no validity data available.
People's general level of happiness can also be said to reflect their level of mental health. The Affectometer 2, devised by Kammann and Flett (1983), is a 40-item measure of subjective well-being or general happiness and sense of well-being, based on the balance of positive and negative feelings in recent experience. It asks participants how often a feeling is present on a 5-point, graded response, from not at all to all of the time. Items are divided evenly between good and bad feelings, and participants get a score on both dimensions, which is mathematically adjusted to give an overall, net score. Reliability has been shown with an alpha level of .95. Convergent validity with the Beck Depression Inventory was .84, and test-retest reliability was shown to be between .53 and .80 (Robinson et al., 1991).
The notion of self-efficacy means believing that you can arrive at a personal goal through your own actions. Research has shown that it is generally beneficial for people to believe in their own self-efficacy (Cloninger, 1996; Nowicki & Duke, 1981). One comparable measure of this dimension of personality is "locus of control," where people believe that they are personally responsible for their life outcomes and that they are not totally at the mercy of uncontrollable, external events such as luck or fate. The Adult Nowicki-Strickland Internal-External Control Scale (Nowicki & Duke, 1981) assesses locus of control via forced-choice items that require participants to answer "yes" or "no" to indicate whether they have an internal or external locus of control. Internal consistency has been reported as between .74 and .86, and test-retest reliability as .56 to .83; convergent validity ranges from .44 to .68.
In other gender research (e.g., Spence et al., 1979) the Texas Social Behavior Inventory (Helmreich & Stapp, 1974) has been used as a measure of well-being, social competence, or social esteem. Participants are asked to respond to 16 items that address their degree of self-confidence in groups and their security in social situations. Items are keyed on a 5-point, Likert-type format from not at all like me to very much like me; higher scores indicate higher levels of social esteem. It has been suggested that the Texas Social Behavior Inventory measures both social and self-esteem (Robinson et al., 1991). Reliability is reported as between .89, and internal consistency as .92. Convergent validity has been found to be between .26 and .83.
The Rosenberg Self-Esteem Scale (Rosenberg, 1979) has enjoyed widespread use, is considered a reliable and valid measure of "self-esteem," and is the standard to which other measures are compared. The scale consists of 10 items; participants indicate their agreement on a 4-point scale that ranges from strongly agree to strongly disagree.
Procedure
Participants were asked to take the set of self-report questionnaires to complete in private and to return them via a collection box or by reply-paid mail. All participants were informed of the need to answer the questions in terms of how they actually saw themselves and not in terms of how they wished to be.
RESULTS
Participants were assigned to one of seven gender role identity identities, M+, M-, F+, F-, A+, A-, or Au using the methods previously described. Figure 1 indicates the percentage of participants in each category.
Observation of dependent variable scatter plots revealed eight extreme scores that were replaced with mean scores. Prior to analyses, all dependent measures were examined for distribution and the assumptions of multivariate analysis. Table I shows descriptive statistics for the dependent measures.
Distribution of scores for "acceptance of others," "locus of control," and "social esteem" appeared to be normal. Scores for "affect" (Affectometer 2) and "self-esteem" were both negatively skewed; self-esteem appeared substantially so. Linearity of variables was assessed via bivariate scatter plots between "acceptance of others" and the other variables. "Affect" and "self-esteem" showed the widest deviations on probability plots, and "self-esteem" appeared to be nonlinear on the bivariate scatter plot.
Homogeneity of variance was assessed as acceptable for all variables except "affect." The correlation matrix between the dependent variables (see Table II) showed a fair degree of intercorrelation, and thus there is a degree of redundancy in each. Therefore those high on "acceptance of others" were likely also to be high on "affect," internal "locus of control," and "social and self esteem."
A oneway MANOVA procedure revealed a main effect for the EPAQ identities, Wilks [lambda] = .649, p < .001, which indicates that the psychological gender role identities differ reliably on some combination of the measures used to assess psychological health and well-being. Plots of means for the MANOVA can be seen in Fig. 2.
To assess which and how the seven psychological gender role identity identities were associated with differences on the dependent measures, MANOVA results were followed up with a discriminant function analysis (DFA). DFA allows for a clearer and sometimes statistically more powerful understanding of group differences than do multiple post hoc tests on significant MANOVA results. No special problems are posed by unequal sample sizes of the classified groups in discriminant analysis as long as the smallest group is notably larger than the number of predictors (Tabachnick & Fidell, 1996). In this study the number of predictors is 5, and the smallest group is the Au group with 12 members.
A direct discriminant function analysis was performed on the whole sample using "affect," "social esteem," "self-esteem," "acceptance of others," and "locus of control" as predictors and psychological gender role identity as the grouping variable. One discriminant function was calculated as significant, Wilks Lambda = .636, p < .001, and it accounted for 67.7% of variance.
The structure matrix (see Table III), with a cut off point of 0.3 or less between predictors and discriminant function 1, suggests that all predictors were useful for distinguishing between gender role identity groups on the first and only significant function. People high in this function tend to have high levels of psychological health and well-being.
As the mean is a summary measure of central tendency in any group, differences between the gender role identity groups on this function were examined by assessment of group centroids (see Table IV). Mean plots for this one function of psychological health and well-being can be seen in Fig. 3. Mean values for each group are as follows: A+ .756, M+ = .340, F+ = .034, Au = -.111, M- = -.123, F- = -.593, and A- = -.765). A follow-up Tukey's B test was performed to assess significant differences (see Table V). A+ was significantly different from A- and also from Au, M-, and F-, and not different from M+ and F+. The trend of results was as predicted.
DISCUSSION
These data support the theory of a differentiated androgyny as proposed by Woodhill and Samuels (in press). As predicted, results for positively androgynous and negatively androgynous people were significantly different. Further, positively androgynous, positively masculine, and positively feminine people had higher levels of a function involving all the measures of psychological health and well-being than negatively masculine, negatively feminine, and undifferentiated androgynous gender role identities.
"Self-esteem" scores were substantially skewed and appeared nonlinear. The authors of the Affectometer 2 (Kammann & Flett, 1983) noted that their scale gave consistently skewed results as do all measures of happiness. Scores for "affect" appeared less skewed than those for "self-esteem," and they appeared linear, however they were not homogenously distributed and may have also had a confounding influence in the final analysis. The sample size was smaller than indicated as desirable in the power analysis and may have been of concern in demonstrating significant results. As well, the small number of men in the sample may limit the generalizabiity of results.
Gender research has yielded ambiguous results for the androgynous gender role identity. Although there appears to be consensus among supporters of androgyny that femininity and masculinity consist of distinguishable traits, many have warned that their underlying structure may be complex and that incorrect assumptions inherent in gender identity inventories may seriously affect empirical findings (Cook, 1985; Marsh & Myers, 1986; Spence, 1984, 1999). It is here proposed that differentiating androgyny into positive and negative subcategories may help resolve this confusion. Our data support the notion that positive androgyny is the developmental ideal.
As predicted, results for A+ were higher than A- on all measures of psychological health and wellbeing. It was also predicted that after A+, significant results would first be found for the positive identities, then the negative identities, with masculine always first. No prediction was proposed for Au.
As a new construct in the measurement of gender role identity, differentiated androgyny has shown that the trait theory of gender role identity can be extended to classify androgynous people into two distinct subgroups that can then be used in reassessing the benefits of particular gender role identities. This is in spite of the possibility that people may not necessarily vary in such a tidy way. These seven categories of gender role identity take account of the fact that people do not always act in or combine their traits in virtuous ways, and it also accounts for the previously unaccounted category of undifferentiated androgyny. Obtained positive qualities associated with positive androgyny and a lack of association of these positive qualities with negative androgyny can go some way to answering past confusion in the literature as to the benefits of what could now be called singular or traditional androgyny.
The results of this study indicate that positively androgynous people obtained higher scores on indicators of mental health and well-being than negatively androgynous, negatively masculine, negatively feminine, and undifferentiated androgynous people. Mental health and general well-being are likely dependent on factors other than gender role identity, but these data support the theory that positive androgyny is associated with their attainment. Traditional androgyny has been treated as a global construct but it has been shown here that it may be as multidimensional as femininity and masculinity. The inability of prior research to differentiate between the positive and negative types of androgyny could have been responsible for masking the benefits of positive androgyny.
The results of this study have given empirical support to the constructs of positive and negative androgyny. Further, the results suggest that it is not just globally androgynous individuals but those who are positively androgynous who have the potential to live the more fulfilled and complete life that Bern (1975) once envisaged. On a global level, many societies and individuals still seem to encourage undesirable masculine traits, yet Bakan (1966) suggested that an unmitigated or overdeveloped masculinity can be destructive. Most incivilities and violent behavior originate from men's behavior and negative masculinity has a far reaching effect as it usually impacts on the lives of others (e.g., such traits as aggression, being hard-hearted, indifference, selfishness, showing off, and vindictiveness). Negative femininity severely affects only the individual who displays it (exemplified in such behaviors as being passive, submissive, temperamental, timid, modest, fragile, and mild). When compared to negative m asculinity, negative femininity does not in any real sense impact on those who are exposed to such individuals. With a balanced emphasis on the social world, on relationships, and on communal values, positive androgyny could temper crime and violence, which have been almost exclusively the domain of men (Heilbrun, 1981b) as well as reduce the competitive drive which causes greed, individualism, arrogance, and elitism (Biddulph, 1995). We propose a differentiated androgyny in which positive androgyny holds within it the hope of reducing the effects of undesirable masculinity to the innocuous effects of undesirable femininity. It may provide a key to unlocking the hidden desire of many men for more acceptable and adaptive levels of behavior.
The results of this study could revitalize the notion of androgyny as an identity whose time has come again. People tend to admire those among us who can express the virtues of both sexes (Woodhill & Samuels, in press); they express what most of us suppress. Support for the validity and utility of differentiating androgyny into positive and negative was here shown. Differences were also revealed for each of the other positive and negative identities. Positive androgyny, positive masculinity, and positive femininity were superior to all other gender role identities and positive androgyny was different from and superior to negative androgyny on measures of psychological health and well-being.
[FIGURE 3 OMITTED]
Table I
Descriptive Statistics of Measures of Mental Health and Well-Being
Mental health and well-being Mean SD
Acceptance of others 62.21 8.43
Affectometer 1.86 1.00
Locus of control 10.10 3.85
Self-esteem 32.99 5.64
Social esteem 42.11 9.11
Note. N = 192.
Table II
Intercorrelation Matrix of Measures of Psychological ZHealth and
well-Being
Acceptance of others Affectometer
Acceptance of others 1.000 0.363 (a)
Affectometer 0.363 (a) 1.000
Locus of control -0.231 (a) -0.404 (a)
Self-esteem 0.321 (a) 0.702 (a)
Social esteem 0.355 (a) 0.568 (a)
Locus of control Self-esteem Social esteem
Acceptance of others -0.231 (a) 0.322 0.355 (a)
Affectometer -0.404 (a) 0.702 (a) 0.568 (a)
Locus of control 1.000 -0.308 (a) -0.349 (a)
Self-esteem -0.308 (a) 1.000 0.620 (a)
Social esteem 0.349 (a) 0.620 (a) 1.000
(a)Correlation is significant at the <0.01 level (2-tailed). N = 192.
Table III
Discriminant Function Analysis--Structure Matrix
Function
1 2 3 4 5
Affectometer .906 (a) .127 -.113 -.352 .159
Social esteem .806 (a) -.252 .238 .461 -.129
Self-esteem .689 (a) -.024 -.159 .255 .659
Locus of control -.379 .776 (a) .443 .234 .051
Acceptance of others .411 .254 .7l7 (a) .366 -.345
(a)Largest absolute correlation between each variable and any
discriminant function. Variables ordered by absolute size of correlation
with function.
Table IV
Functions at Group Centroids (Mean Scores)
Function
EPAQ identity 1 2 3 4 5
F+ -.034 .308 -.298 .013 -.023
F- -.593 .628 .232 -.215 .004
M+ .340 .030 -.234 -.102 .035
M- -.123 -.532 .117 -.497 .004
A+ .746 -.048 .150 .026 -.011
A- -.765 -.265 .064 .155 -.008
Au -.111 .125 .082 .074 .062
Note. Unstandardized canonical discriminant functions evaluated at group
means.
Table V
Significant Differences Between Gender Identity Mean Scores On Function
1 From Turkey's B Test
A+ M+ F+ Au M- F- A-
.756a 340ab -.034abc -.111bc -.123bc -.593c -.765c
Note. Gender identities that share the same letter are not significantly
different from each other.
Fig. 1
Percentages of people categorized in each of the seven psychological
gender role identity identities according to the new scoring method for
the EPAQ.
EPAQ gender identity
F+ 15%
F- 9%
M+ 11%
M- 8%
A+ 27%
A- 23%
Au 6%
Note: Table made from bar graph
Fig. 2
MANOVA Means--gender role identity identities on each measure of
psychological health and well-being.
acceptance of others
epaqid
M+ 65
A+ 64
F+ 64
Au 62
A- 60
F- 59
M- 59
Note: Table made from bar graph
Fig. 2
MANOVA Means--gender role identity identities on each measure of
psychological health and well-being.
affectometer 2
epaqid
A+ 2.43
M+ 2.18
F+ 1.87
M- 1.84
Au 1.75
F- 1.43
A- 1.20
Note: Table made from bar graph
Fig. 2
MANOVA Means--gender role identity identities on each measure of
psychological health and well-being.
locus of control
epaqid
F- 13
Au 11
A- 11
F+ 10
A+ 9
M+ 9
M- 8
Note: Table made from bar graph
Fig. 2
MANOVA Means--gender role identity identities on each measure of
psychological health and well-being.
self esteem
epaqid
A+ 36
M+ 35
Au 33
F+ 33
M- 32
A- 31
F- 30
Note: Table made from bar graph
Fig. 2
MANOVA Means--gender role identity identities on each measure of
psychological health and well-being.
social esteem
epaqid
A+ 48
M+ 44
Au 41
M- 41
F+ 40
A- 39
F- 37
Note: Table made from bar graph
REFERENCES
Anderson, K. L. (1986). Androgyny, flexibility and individualism. Journal of Personality Assessment, 50, 265-278.
Antill, J. (1983). Sex role complementarity versus similarity in married couples. Journal of Personality and Social Psychology, 45, 45-155.
Antill, J., & Cunningham, J. (1979). Self esteem as a function of masculinity in both sexes. Journal of Consulting and Clinical Psychology, 47, 783-785.
Antill, J., & Cunningham, 3. (1980). The relationship of masculinity, femininity androgyny to self-esteem. Australian Journal of Psychology, 32, 195-207.
Antill, J., Cunningham, J., Russell, G., & Thompson, N. (1981). An Australian Sex Role Scale. Australian Journal of Psychology, 33, 169-183.
Bakan, D. (1966). The duality of human existence. Chicago, IL: Rand McNally.
Baril, G., Elbert, N., Maher-Potter, S., & Reavy, G. (1989). Are androgynous managers really more effective? Group and Organisation Studies, 14, 234-249.
Bem, S. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155-162.
Bem, S. (1975). Sex role adaptability: One consequence of psychological androgyny. Journal of Personality and Social Psychology, 31, 634-643.
Bem, S. (1977). On the utility of alternative procedures for the assessing of psychological androgyny, Journal of Consulting and Clinical Psychology, 45, 196-203.
Bem, S. (1998). An unconventional family. New Haven, CT: Yale University Press.
Biddulph, S. (1995). Manhood. Sydney: Finch.
Boswell, H. (1997). The transgender paradigm shift toward free expression. In B. Bullough, V. Bullough, & J. Elias (Eds.), Gender blending (pp. 53-57). Amherst, NY: Prometheus.
Burchardt, C., & Serbin, L. (1982). Psychological androgyny and personality adjustment in college and psychiatric populations. Sex Roles, 8, 835-851.
Campbell, M., Steffan, J., & Langmeyer, D. (1981). Psychological androgyny and social competence. Psychological Reports, 48, 611-614.
Cheng, C. (1999). Gender role differences in susceptibility to the influence of support availability on depression. Journal of Personality, 67, 439-467.
Cloninger, S. (1996). Theories of personality: Understanding persons. Englewood Cliffs, NJ: Prentice Hall.
Cook, E. (1985). Psychological androgyny. New York: Pergamon. Desrochers, S. (1995). What types of men are most attractive and most repulsive to women? Sex Roles, 32, 375-392.
Eagly, A., & Wood, W. (1999). The origin of sex differences in human behavior. American Psychologist, 54, 408-423.
Erikson, E. (1950). Childhood and society. New York: Norton.
Fey, W (1955). Acceptance by others and its relation to acceptance of self and others: A revaluation. Journal of Abnormal and Social Psychology, SO, 274-276.
Green, B., & Kendrick, D. (1994). The attractiveness of gender type traits at different relationship levels: Androgynous characteristics may be desirable after all. Journal of Personality and Social Psychology, 20, 244-253.
Hefner, R., Rebecca, M., & Oleshansky, B. (1975). Development of sex-role transcendence. Human Development, 18, 143-158.
Heilbrun, A. (1981 a). Gender differences in the functional linkage between androgyny, social cognition, and competence. Journal of Personality and Social Psychology, 41, 1106-1118.
Heilbrun, A. (1981b). Human sex role behavior. New York: Pergamon.
Helgesen, V. (1994). Relation of agency and communion to wellbeing: Evidence and potential explanations. Psychological Bulletin, 116, 412-428.
Helmreich, R., & Stapp, J. (1974). Short form of the Texas Social Behavior Inventory, an objective measure of self-esteem. Bulletin of the Psychometric Society, 4, 473-475.
Hoffman, R. (2001). The measurement of masculinity and femininity: Historical perspective and implications for counselling. Journal of Counselling and Development, 79, 472-485.
Jaccard, J., Wan, C., & Turrisi, R. (1990). The detection and interpretation of interaction effects between continuous variables in multiple regression. Multi variate Behavioural Research, 25, 467-478.
Jones, W., Chernovetz, M., & Hansson, R. (1978). The enigma of androgyny: Differential implications for males and females. Journal of Consulting and Clinical Psychology, 46, 298-313.
Kammann, R., & Flett, R. (1983). Sourcebook for measuring wellbeing with the Affectometer 2. Dunedin, New Zealand: University of Otago.
Kelly, J., & Worrell, J. (1977). New formulations for sex roles and androgyny: A critical review. Journal of Clinical and Consulting Psychology, 45, 1101-1115.
Kimmel, M. (2000). The gendered society. New York: Oxford University Press.
Kirchemeyer, C. (1996). Gender roles and decision making in demographically diverse groups: A case for reviving androgyny. Sex Roles, 34, 49-663.
Kopper, B., & Epperson, D. (1996). The experience and expression of anger. Relationships with gender role socialization, depression and mental health functioning. Journal of Consulting Psychology, 43, 58-165.
Lenney, E. (1991). Sex roles: The measurement of masculinity, femininity and androgyny. In J. Robinson, D. Shaver, & L. Wrightsman (Eds.), Measures of personality and social psychological attitudes (pp. 573-660). San Diego, CA: Harcourt Brace Jovanovich.
Lewis, R. (1978). Emotional intimacy among men. Journal of Social Issues, 34, 108-121.
Looy, H. (2001). Sex differences: Evolved, constructed, and designed. Journal of Psychology and Theology, 29, 301-314.
Lubinski, D., Tellegen, A., & Butcher, J. (1981). The relationship between androgyny and subjective indicators of emotional wellbeing. Journal of Personality and Social Psychology, 40, 722-730.
Lubinski, D., Tellegen, A., & Butcher, J. (1983). Masculinity, femininity and androgyny viewed and assessed as distinct concepts. Journal of Personality and Social Psychology, 44, 428-439.
Markstrom-Adams, C. (1989). Androgyny and its relation to adolescent psychological well-being: A review of the literature. Sex Roles, 21, 25-341.
Marsh, H., & Myers, M. (1986). Masculinity, femininity and androgyny: A methodological and theoretical critique. Sex Roles, 14, 397-430.
Matlin, M. (2000). The psychology of women (4th ed.). Fort Worth, TX: Harcourt.
May, A., & Spangenberg, J. (1997). Androgyny and coping in men with a managerial orientation. South African Journal of Psychology, 27, 244-249.
Mead, M. (1935). Sex and temperament in three primitive societies. New York: William Morow.
Mead, B., & Ignico, A. (1992). Children's gender typed perceptions of physical activity: Consequences and implications. Perceptual and Motor Skills, 75, 1035-1013.
Nowicki, S., & Duke., M. (1981). The Nowicki-Strickland Life Span Locus of Control Scales: Construct validation. In H. M. Lefcourt (Ed.), Research with the locus of control construct (Vol. 2, pp. 16-43). New York: Academic Press.
O'Heron, C., & Orlofsky, J. (1990). Stereotypic and non-stereotypic sex role trait and behavior orientations, gender identity and psychological adjustment. Journal of Personality and Social Psychology, 58, 134-143.
Pearson, J., & Cooks, L. (1995). Gender and power. In P. M. Kalbfieisch & M. J. Cody (Eds.), Gender, power, and communication in human relationships (pp. 331-349). Hillsdale, NJ: Erlbaum.
Powell, G., & Butterfield, D. (1979). The good manager: Masculine or androgynous. Academy of Management Journal, 22, 395-403.
Pryzgoda, J., & Chrisler, J. C. (2000). Definitions of gender and sex: The subtleties of meaning. Sex Roles, 43, 553-569.
Ricciardelli, R., & Williams, R. (1995). Desirable and undesirable gender traits in three behavioral domains. Sex Roles, 33, 637-656.
Robinson, J., Shaver, D., & Wrightsman, L. (1991). Measures of personality and social psychological attitudes. San Diego, CA: Harcourt Brace Jovanovich.
Rose, A., & Montemayor, R. (1994). The relationship between gender role orientation and perceived self-competency in male and female adolescents. Sex Roles, 31, 579-595.
Rosenberg, M. (1979). Conceiving the self. New York: Basic Books.
Russell, G., & Antill, J. (1984). An Australian sex-role scale: Additional psychometric data and correlations with self-esteem. Australian Psychologist, 19, 13-18.
Sawrie, S., Watson P., & Biderman, M. (1991). Aggression, sex role measures and Kohut's psychology of the self. Sex Roles, 25, 41-161.
Schullo, S., & Alperson, B. (1984). Interpersonal phenomenology as a function of sexual orientation, sex, sentiment, and trait categories in long term dyadic relationships. Journal of Personality and Social Psychology, 47, 983-1002.
Shaffer, D. (2002). Developmental psychology: Childhood and adolescence (6th ed.). Stamford, CT: Wadsworth.
Shaver, P., Papilia, D., Clark, C., Koski, L., Tidwell, M., & Nalbare, D. (1996). Androgyny and attachment security: Two related models of optimal personality. Personality and Social Psychology Bulletin, 22, 582-597.
Shimonaka, Y., Nakazato, K., Kawaai, C., & Shinichi, 5. (1997). Androgyny and successful adaptation across the life span among Japanese adults. Journal of Genetic Psychology, 158, 389-401.
Skoe, E. (1995). Sex role orientation and its relationship to the development of identity and moral thought. Scandinavian Journal of Psychology, 36, 235-245.
Spence, J. (1984). Masculinity, femininity, and gender-related traits: A conceptual analysis and critique of current research. Progress in Experimental Personality Research, 13, 1-97.
Spence, J. (1999). Thirty years of gender research: A personal chronicle. In W. Swann Jr., J. Langlois, & L. Gilbert (Eds.), Sexism and stereotypes in modern society: The gender science of Janet Taylor Spence (pp. 255-281). Washington, DC: American Psychological Association.
Spence, J., & Buckner, C. (2000). Instrumental and expressive traits, trait stereotypes, and sexist attitudes: What do they signify? Psychology of Women Quarterly, 24, 44-62.
Spence, J., Helmreich, R., & Holahan, C. (1979). Negative and positive components of psychological masculinity and femininity and their relationship to self-reports of neurotic and acting out behaviors. Journal of Personality and Social Psychology, 37, 1673-1682.
Spence, J., Helmreich, R., & Stapp, J. (1974). The Personal Attributes Questionnaire: A measure of sex role stereotypes and masculinity--femininity. JSAS Catalogue of Selected Documents in Psychology, 4, 43-44.
Stake, J. (1997). Integrating expressiveness and instrumentality in real life settings: A new perspective on the benefits of androgyny. Sex Roles, 37, 541-564.
Tabachnick, B., & Fidel, L. (1996). Using multivariate statistics (3rd ed.). Northridge, CA: Harper Collins.
Taylor, M., & Hall, J. (1982). Psychological androgyny: A review and reformulation of the theories, methods, and conclusions. Psychological Bulletin, 92, 347-366.
West, C., & Zimmerman, D. (2000). Doing gender. In M. Kimmel (Ed.), The gendered society reader (pp. 131-149). New York: Oxford University Press.
Whitley, B. (1983). Sex role orientation and self-esteem: A critical meta-analytic review, Journal of Personality and Social Psychology, 44, 765-778.
Woodhill, B., & Samuels, C. (in press). Desirable and undesirable androgyny: A prescription for the twenty-first century. Journal of Gender Studies.
Wubbenhorst, T. (1994). Personality characteristics of music educators and performers. Psychology of Music, 22, 63-74.
Wulff, M., & Steitz, J. (1999). A path model of the relationship between career indecision, androgyny, self-efficacy and self-esteem. Perceptual and Motor Skills, 88, 935-940.
Brenda Mae Woodhill (1,2)
(1.) School of Psychology, University of New England, Armidale, New South Wales, Australia.
Curtis A. Samuels (1,2)
(2.) To whom correspondence should be addressed at School of Psychology, University of New England, Armidale, New South Wales 2351, Australia; e-mail: b_m_w@mail.optusnet.com.au or csamuels@metz.une.edu.au.
COPYRIGHT 2003 Plenum Publishing Corporation
COPYRIGHT 2003 Gale Group