A conceptual analysis of sex-role identity and its interaction with gender and chronological age

MISSING IMAGE

Material Information

Title:
A conceptual analysis of sex-role identity and its interaction with gender and chronological age
Physical Description:
vii, 115 leaves : ; 28 cm.
Language:
English
Creator:
Suzuki, Tetsuko Fujita, 1929-
Publication Date:

Subjects

Subjects / Keywords:
Sex role   ( lcsh )
Self-perception   ( lcsh )
Age (Psychology)   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 110-114).
Statement of Responsibility:
by Tetsuko Fujita Suzuki.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000090496
notis - AAK5888
oclc - 05802019
System ID:
AA00003486:00001

Full Text












A CONCEPTUAL ANALYSIS OF SEX-ROLE IDENTITY AND ITS
INTERACTION WITH GENDER AND CHRONOLOGICAL AGE













By

TETSUKO FUJITA SUZUKI


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY













UNIVERSITY OF FLORIDA


1979















ACKNOWLEDGMENTS


I would like to express my appreciation to the members of my

doctoral committee, Dr. Jacquelin R. Goldman, Chairman;

Dr. Patricia T. Ashton; Dr. Cynthia D. Belar; Dr. Walter R. Cunningham;

and Dr. Marvin E. Shaw. I would like to thank Dr. Robert L. Isaacson

and Dr. Betty L. Siegel, who also served on my committee.

Dr. Goldman's persistent emphasis on clarity of thinking served as

a valuable objective and stimulus.

My sincere thanks are extended to Dr. Charles E. Holzer who

patiently taught me the intricacies of computer language. He was almost

unlimited in his willingness to discuss critically the results of the

analysis.

The seed of this study was planted by the intelligent, troubled

women who allowed me to participate in their struggle to search for

their identity. I was enlightened and educated by my association with

all of those who took the time to contribute their views as subjects for

the study. They must all be considered as collaborators in this endeavor.

The prompt, accurate, and efficient work of Linda Stallings in

typing the final draft is gratefully acknowledged.

This work is dedicated to my children, Georganne, Joan, Jimmy, and

Stanley, and to my husband, Howard, whose endurance was bottomless.

Their support was constant as I shared my joys and anxieties and often

sacrificed my attention at home. They steadfastly allowed me to do it

my way.
















TABLE OF CONTENTS


Page
ACKNOWLEDGMENTS........................................... ........ ii

ABSTRACT.............................................................. v

CHAPTER

1 INTRODUCTION............................... ................ i

The Developmental Process of Psychological
Sexual Differentiation................................ 2
The Concept of Sex-Role Identity........................ 15
Integration of Changes in Sex-Role Definition
With Changes in Developmental Tasks.................. 20
Hypotheses Concerning the Relationships Between
Self-Concept and Sex-Role Style and Their
Interaction With Gender and Age....................... 32

2 EXPERIMENTAL ANALYSIS OF THE INTERACTION BETWEEN
CHRONOLOGICAL AGE AND GENDER AND ITS EFFECT ON
SEX-ROLE IDENTITY........................................ 35

Method................. ... .. ...................... ..... 35
Subjects................................................. 36
Assessment Instruments.................................. 39
Theoretical Relationships Between Assessment
Instruments and Derivation of Specific Hypotheses..... 47

3 RESULTS........................ ...... .... ................. 59

Analyses of Data Related to the Bem Sex-Role Inventory.. 59
Analyses of Relationships Between Independent
Variables and Dependent Variables..................... 62

4 DISCUSSION...... ........................................... .80

Statistical Characteristics of the Self-Identity
Measures ......... ..... ..................... ......... 82
Sex-Role Identity and Its Relationship to Gender,
Chronological Age, and Other Measures of Self-
Identity...................................................... 85
Relationship of Sex-Role Identity to Demograhic Data.... 88










CHAPTER Page

5 SUMMARY AND CONCLUSIONS..................................... 90

APPENDICES

A UNIVERSITY OF FLORIDA/DEPARTMENT OF PSYCHOLOGY CONSENT
FORM...................................................... 93

B DEMOGRAPHIC DATA............................................. 95

C BEM SEX-ROLE INVENTORY..................................... 97

D TEXAS SOCIAL BEHAVIOR INVENTORY............................. 99

E LOCUS OF CONTROL (I-E SCALE)...............................104

F ROLE CONSISTENCY...........................................107

REFERENCES..........................................................110

BIOGRAPHICAL SKETCH............................................... ..115










Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy



A CONCEPTUAL ANALYSIS OF SEX-ROLE IDENTITY AND ITS
INTERACTION WITH GENDER AND CHRONOLOGICAL AGE

By

Tetsuko Fujita Suzuki

June 1979

Chairman: Jacquelin R. Goldman
Major Department: Psychology


Sex-role identity, unlike biological gender, is subject to

influence by the social environment. The data indicate that our

culture not only differentiates males and females along particular

psychological dimensions but that the effect differs contingent on

the individual's age-role.

The college-aged female is of particular interest in terms of

sex-role identity because she is confronted with the prospect that

achievement success may be seen as conflicting with her internalized

feminine values by which she measures her success as a woman. Data

from past research consistently demonstrate that feminine-typed

college students are at a psychological adjustment disadvantage. A

meaningful issue is whether these results are at least partially a

consequence of the particular phase of the life cycle.

This analysis extended over a broad spectrum of chronological ages

from the college-aged individuals to persons over 60 years of age coping

with retirement. Sex-role identity was viewed within the new conceptu-

alization of sex-roles that allows for the possibility that a person may









develop both masculine and feminine attributes. Sex-role identity was

considered one aspect of "self-identity" and was related to three

measures (self-esteem, locus of control, and role consistency) that

have a theoretical relationship to psychological health and a sense

of self.

Self-esteem scores were highly correlated with masculinity scores,

regardless of gender or age. The college-age feminine-typed females

were not only lower in self-esteem but also lower in role consistency

scores and the most external in locus of control. On the other hand,

within the other three age groups, there were no consistent or

significant differences in role consistency or locus of control scores

associated with the feminine-typed category. There is some indication

that the psychological adjustment disadvantage accruing to the

feminine-typed college female may be at least partially a consequence

of the developmental stage. This is a stage when sex-role identity

plays a major role in the establishment of self-identity.

Additional support for the hypothesis that the import of sex-role

identity changes with developmental stages was found with reference to

men in the oldest age group. A significant proportion of men over 60

years of age diminished their endorsement of "masculine" characteristics

and maintained the most internal locus of control and had relatively

high role consistency scores.









These findings support the hypothesis of a significant inter-

relationship between age role prescriptions sanctioned by society

and an individual's global self-concept of masculinity and femininity.

Sex-role identity not only impacts on self-identity with changing

ascendency, but the influence differs for males and females.






7 -Chairman















CHAPTER 1
INTRODUCTION


Sex is a highly salient status. With rare exceptions, human

beings are unambiguously assigned to the male or female gender at

birth, and it is the first status to be announced. Once the assign-

ment is made, the new born infant is reared according to this

assignment and the individual is not only subject to a host of

expectations but is differentially rewarded and punished by the social

environment contingent on this assignment. An individual's biological

gender identity is constant. However, his or her sex-role identity

(the psychological dimension of masculinity or femininity congruent

with the individual's self-concept) is constantly subject to influence

by the social environment.

This study undertakes the examination of the concept of sex-role

identity and its interaction with gender and chronological age. First,

the developmental process of psychological sexual differentiation is

explored. Second, the concept of sex-role identity is reviewed in

the context of questions being raised by psychologists (Bem, 1974;

Block, 1973; Mednick, Tangri, & Hoffman, 1975; Spence & Helmreich,

1978) about the assumptions underlying psychological research on

masculinity and femininity conducted over the past three or four

decades. Third, a theoretical framework is presented which con-

ceives of sex-role identity within the larger perimeter of ego

identity and attempts to integrate changes in sex-role definition









with changes in developmental tasks. Fourth, according to this

formulation, hypotheses are derived that attempt to examine the

relationships between personality variables associated with a sense

of self, an individual's sex-role definition, and the interaction

of these variables with gender in four age groups.


The Developmental Process of Psychological
Sexual Differentiation


Since learning to be a psychological male or female is one of

the earliest and most pervasive tasks imposed on an individual in

our culture, the developmental process of sex-typing has generated

much research. Generally, researchers have attempted to clarify two

major questions: Are there psychological differences between males

and females? And, if psychological differences do exist, how did

they come about?


Psychological Differences Between Males and Females

The Hampsons and Money (Hampson & Hampson, 1961; Money, 1965b;

Money, Hampson, & Hampson, 1957) believe that humans are psycho-sexually

neutral at birth and that gender identity is an end product of a sexually

dimorphic developmental sequence; i.e., the genetic code does not program

gender identity and role as a male or a female in the human species. They

posit a critical period phenomenon which asserts that the development of

normal adult sexual behavior is contingent on having been socially

assigned to a given sex before the age of three or four. Hermaphrodites

assigned at birth to one sex because of external genital characteristics

have later been reassigned to the other sex so that their social sex identity









will be more consistent with their internal sex characteristics.

If this is done before the age of three or four, the child's later

sexual adjustment seems to be normal. However, their sample is

small and their data largely retrospective and based on pediatric

reports, parent reports, etc.

Diamond (1965) after reviewing a variety of clinical and

experimental evidence reaches the following conclusion:

The evidence and arguments presented show that,
primarily owing to prenatal genic and hormonal
influences, human beings are definitely predis-
posed at birth to a male or female gender orien-
tation. Social behavior of an individual and thus
gender role, are not neutral and without initial
direction at birth. Nevertheless, sexual pre-
disposition is only a potentiality setting limits
to a pattern that is greatly modifiable by onto-
genetic experiences. Life experiences most likely
act to differentiate and direct a flexible sexual
disposition and to mold the prenatal organization
until an environmentally (socially and culturally)
acceptable gender role is formulated and established.
(p. 167)

Maccoby and Jacklin (1974) after reviewing a massive body of

research findings of a more psychological nature conclude that with

our present body of knowledge, it is not possible to classify the

differential behaviors as being either innate or learned. They

propose that a genetically controlled characteristic may take the form

of a greater readiness to learn a particular kind of behavior. Aside

from some fairly well established differences in intellectual skills,

there is unequivocal evidence for greater male aggressiveness. The

authors state:

The sex difference in aggression has been observed in
all cultures in which the relevant behavior has been
observed. Boys are more aggressive both physically and
verbally. They show attenuated forms of aggression
(mock-fighting, aggressive fantasies) as well as the









direct forms more frequently than girls. The sex
difference is found as early as social play begins--
at age 2 or 2. Although the aggressiveness of both
sexes declines with age, boys and men remain more
aggressive through the college years. Little infor-
mation is available for older adults. (p. 352)

However, they point out that if one sex is more biologically predis-

posed than the other to perform a behavior, this fact would be

reflected in popular beliefs about the sexes, so that innate tendencies

help to produce the cultural lore that the child learns.

It is beyond the scope of this study to review the empirical

evidence, but there seems to be general agreement that biological-

genetic factors and cultural-environmental factors interact in the

process of psychological sexual differentiation. A biological basis

for psychological differences between the sexes has not been unambigu-

ously established and even where the evidence most often favors a

biological explanation (as in aggression) all that has been clearly

established is that there is a sex-linked differential readiness to

respond in aggressive ways to the relevant experiences. An individual's

aggressive behavior can be strengthened, weakened, redirected, or

altered in form by his or her unique pattern of experiences.

The relationship of sex differences to age is more complicated

and vague. There are great variations with age not only in what is

measured but in how the measurement is taken. In the very young

children, naturalistic behavioral observations and parent ratings are

fairly frequent. From the time the child enters school through adult-

hood, the measurements are usually based on questionnaires, self-

reports or on experimental situations with a restricted set of

eliciting conditions and behavioral measures.






5


However, what clearly emerges are some underlying presumptions.

In the past, psychological investigations have largely been predicated

on the assumption that biological gender, masculine and feminine sex-

role behaviors and the psychological attributes of masculinity and

femininity are tightly intercorrelated and that the distribution of

the sexes on these variables has implicity been assumed to be bimodal.

A concomitant assumption has been that cross-sex behaviors and attri-

butes are in a sense pathological with negative implications.

Within the general culture, a frequent assumption has been that

sex-role behaviors are not only correlated with psychological charac-

teristics but may also have causal connections. Thus, parents often

insist that their children behave according to traditional sex-

appropriate standards (such as playing with "sex-appropriate" toys)

to forestall the possibility that their children might adopt cross-sex

behaviors and become sexually deviant. The goal of socialization has

been to inculcate culturally defined sex-appropriate characteristics

and behaviors in each sex.


Psychological Sexual Differentiation

Given that there are cultural expectations about how a male or a

female ought to behave, theories concerning the process of psycho-

logical sexual differentiation have been formulated. Freud's libido

theory posits that the basic patterning of sexual attitudes is

instinctual in its origins. In outline, psychoanalytic theory states

that both sexes typically form strong emotional attachments to the

mother in the early years due to her nurturing role. At the age of

three to five, the Oedipal complex emerges within the boy as he is









sexually attracted to his mother and wishes to displace his father.

However, his recognition of his father's greater power and the fear

of castration by his father generates an intolerable conflict which

he resolves by renouncing the mother and identifying with the father.

This identification leads the boy to adopt the masculine character-

istics he sees in his father and eventually internalizes through his

father the values and standards set by society for males. In the

girl, the sequence of events is more involved and not as clearly

outlined. Holding her mother responsible for her castrated state,

she turns from her mother and develops a sexual attachment to her

father. Due to the realistic barriers that prevent her from gratifying

her desire for her father and fearing the loss of her mother's love,

the girl resolves the Oedipal conflict by reidentifying with the

mother. Freud theorized that the girl's Oedipus complex tends to

persist, although in modified form, and her identification with the

mother is less intense that the boy's with his father.

Mischel and other social learning theorists stress the impor-

tance of modeling and conditioning. Mischel (1966) states:

According to social-learning theory the acquisition
and performance of sex-typed behaviors can be
described by the same learning principles used to
analyze any other aspect of an individual's behavior.
In addition to discrimination learning, these prin-
ciples include the patterning of reward, nonreward
and punishment under specific contingencies and the
principle of direct and vicarious conditioning.
(pp. 56-57)

Kohlberg, on the other hand, proposes a cognitive developmental

theory. He elaborates a theory which assumes that basic sexual atti-

tudes are not patterned directly by either biological instincts or

arbitrary cultural norms but by the child's active cognitive









organization of his social world along sex role dimensions. Kohlberg

(1966) states:

We shall point out that this learning is cognitive
in the sense that it is selective and internally
organized by relational schemata rather than
directly reflecting associations of events in the
outer world. In regard to sex-role these schemata
that bind events together include concepts of the
body, the physical and social world, and general
categories of relationship (causality, substan-
tially, quantity, time, space, logical identity,
and inclusion). (p. 83)

Major reviews on sex-typing (Kagan, 1964; Maccoby, 1966; Mussen,

1969; Sears, Rau, & Alpert, 1965) all agree that children acquire

behaviors through imitation of parents, but there is little evidence

that children systematically imitate same-sex models. In an investi-

gation in which children observed an adult model dispensing toys and

snacks to an adult of the opposite sex, Bandura, Ross, and Ross (1963a)

reported that girls 3-5 years of age preferred to imitate the powerful

controlling model more than the adult recipient, whether the model

was male or female. Boys on the other hand tended to imitate the

male even when faced with the controlling female adult model and a

male recipient. Hetherington (1965) failed to show any consistent

tendency for children ranging from 4-11 to imitate the same-sex

parent in the choice of aesthetic preferences after they had observed

the parents express their choices. There is some empirical evidence

that lead researchers to believe that it is not the sex of the

model but certain characteristics of the model; namely, nurturance

and powerfulness, which promote imitation. Bandura and Huston (1961)

report that when 3-5 year old children were exposed to either a

rewarding interaction or a cold nonnurturant relationship with a model









prior to performing a task, with the model present, the children

tended to imitate the explicit, but functionless behaviors of the

nurturant model significantly more than the nonnurturant model.

Hetherington and Frankie (1967) examined the effects of parental

dominance, warmth, and conflict on imitation in children. They

found that, in homes in which conflict had been observed and both

parents were low in warmth, children 4-6 years old modeled the

behavior of the dominant parent on an imitation task regardless of

the sex of the parent or the sex of the child. In homes in which

the conflict was low or the nondominant parent was warm, girls

imitated the warm parent regardless of sex but boys imitated the

dominant parent regardless of sex. The modeling process is certainly

operative in the acquisition of a wide variety of potential behaviors;

however, knowledge about what behaviors are sex-appropriate is crucial

in the performance of an action out of this repertoire of potential

behaviors.

The psychoanalytic theorists assume that the sex-typing of

behaviors comes about as a result of identification with the same

sexed parent. The social learning theorists postulate that sex-

typing of behaviors is a product of differential reinforcement (of

the children and others) of sex-appropriate behaviors; while Kohlberg's

view is that the child's growing understanding of the context of the

culturally prescribed roles determines the child's sex-typed behavior

choices. The manner in which parents proceed to socialize a child in

terms of sex-appropriate behaviors must, of course, depend on the

parent's own definition and values if he or she is to differentially

reinforce behaviors emitted by the child.









Whatever the underlying mechanisms are that foster sex-appropriate

socialization, boys seem to receive more pressures against engaging in

sex-inappropriate behaviors, where as the sex-inappropriate activities

of girls are less clearly defined and less firmly enforced. This can

be seen not only in the obvious matter of dressing the two sexes but

also in direct monitoring of behaviors. For example, a parent is more

upset if a boy wears high heels and plays house than when a girl wears

boots and plays cowboy and Indians. Two studies have been directed at

this question (Lansky, 1967). Parents of preschool children were

asked to judge a same-sex parent's reaction to a child's preference for

1 of 2 sex-linked objects, names or activities. No differences were

found in fathers' and mothers' attitudes towards boys' and girls'

same-sex choices. The primary finding was a significantly more nega-

tive parental reaction to cross-sex choices of boys. Fling and

Manosevitz (1972) obtained similar results in their study of parents

of nursery-school children. Parents were asked to make activity

choices from a standardized list of sex-linked activities, and in a

subsequent interview they were asked how strongly they would object to

their children engaging in any of the activities they had omitted from

their choices. Scores were derived that represented the extent of

their discouragement of a child's sex-inappropriate activities. Both

parents chose more sex-appropriate activities for their sons than they

did for girls and much more strongly discouraged sex-inappropriate

behavior in sons than daughters.

Mussen and Rutherford (1963) designed a study to test not only

the developmental identification hypothesis (i.e., appropriate sex

typing of children is a consequence of identification with same sex









parents) but also the hypotheses that parental encouragement of sex-

appropriate behaviors of parents self-acceptance of his or her own

sex-role both facilitated appropriate sex-typing. The subjects (5!1-6

year olds) were divided into high and low masculine boys and high and

low feminine girls by means of a projective test of sex-role preferences

(the IT scale). Structured doll play with the children and structured

interviews with the parents were used to assess nurturant or punitive

relationships between parents and children. Parental personality was

assessed with two scales of the California Psychological Inventory.

Parental encouragement of appropriate sex-typed activities was esti-

mated with the use of a questionnaire which inventoried each parent's

reactions to a list of activities that have been known to differentiate

significantly between the preferences of the two sexes. According to

the data of this study, the most crucial determinant of the development

of masculinity in young boys is the nature of the father-son

relationship--boys who see their father as a powerful person in their

lives were likely to develop high sex-appropriate responses. None of

the other variables had any significant effect on the boy's masculini-

zation. The process of female-typing in the little girls, although

directly related to a positive mother-daughter relationship, is also

facilitated by a highly adequate mother as the feminine model, and a

father who possesses a high degree of masculinity of interests and

attitudes and actively encourages the daughter's participation in

sex-typed activities. The authors conclude:

From the point of view of learning theory, the role
of the parents in masculinizing their sons is
primarily that of providing initial motivation to
acquire masculine characteristics and behaviors.
The general social-cultural milieu further implements










the masculinization by presenting numerous, well
articulated and distinct cues for the male sex
role and rewards the boy for learning these .
For girls, the social-cultural milieu gives
less support in the assimilation of her sex role.
Due to the relatively less value of the feminine
role in middle class American culture and the
relative paucity and nondistinctiveness of cues
associated with the female sex role among young
children, her parents must assist her in several
ways if she is to achieve a high degree of
femininity. More specifically, parents are
forced to assume three feminizing functions with
their daughters, only one of which is like the
parents' role in masculinizing boys. They must
evoke motivation to acquire femininity, and,
in addition, they help the feminizing process
by presenting some cues for discriminating the
sex roles and by directly encouraging the girls
to adopt at least certain kinds of behavior
characteristics of the feminine role. (p. 604)


Another View of Parental Variables and the Development of
Masculinity and Femininity

Spence and Helmreich (1978) point out some of the problems within

the literature on the development of masculine and feminine personality

attributes. First they indicate that masculine and feminine "sex roles"

have been used as umbrella terms, referring to all the internal char-

acteristics and overt patterns of behavior that may be presumed to dis-

tinguish between men and women. They believe this fact has fostered

the search for single theoretical or empirical models to account for

all gender-related phenomena and the use of measures that tap only a

limited behavioral domain to define the multi-faceted concept of

masculinity-femininity. Most frequently these measures have been

heavily, if not exclusively, concerned with sex-typed interests and

activities and have an uncertain relationship with masculine and









feminine personality characteristics. Second, until recently even

masculinity-femininity measures that contain only personality trait

names or dimensions have been set up as bipolar scales.

Two studies provide suggestive data on the influence of parental

behaviors on the development of masculine and feminine characteristics.

Baumrind and Black (1967) identified three major patterns of parental

behaviors that were labeled authoritarian, authoritative, and per-

missive. The prototypic authoritarian parent imposes demands with

little explanation or allowance for the child's needs and opinions,

is generally nondemocratic in disciplinary procedures, but may or may

not be affectionate and emotionally responsive. The authoritative

parent exercises firm control but recognizes the child's needs and

shares with the child the reasons underlying disciplinary decisions.

The permissive parent attempts to be highly accepting toward the

child's impulses but makes little or no attempt to shape or alter the

child's ongoing or future behavior. The nursery school children of

these same parents had initially been separated into three groups

according to their social and emotional behaviors; the parent behavior

clusters associated with each group of children were then determined.

The group of children observed to be high in self-reliance, self-control,

and had friendly relationships with peers tended to have parents from

the authoritative cluster. Authoritarian parents were associated with

children who were relatively discontented, moody, and unsuccessful in

peer relationships. Children in the third group were relatively more

cheerful and less irritable than those in the second group, just

described, but had the least self-reliance and self-control. These








data suggest that the development of instrumental characteristics

related to self-esteem (also related to masculinity) and social

competence is facilitated when the parents combine warmth, reasoning,

and acceptance of the child with the imposition of standards of

behavior that they expect the child to meet.

Baumrind (1971) again studying nursery school children and their

parents identified the aforementioned cluster of parent behaviors

plus a fourth, the rejecting pattern, in which the parent is relatively

unrestrictive and neglectful. A clear association was found in girls

between dominant, purposive, and independent (often associated with

stereotypic masculinity) females and authoritative parental behaviors.

In boys, the relationships were similar but there was some indication

that the extremely firm control of the authoritative parents impaired

the development of independence in these otherwise competent boys.

Block (1973) summarizes results that are relevant to sex-role

socialization from a larger body of data accumulated for a longitu-

dinal study at the Institute of Human Development. The subjects were

between the ages of 30-40 years at the time they were administered

the California Psychological Inventory but they were part of a longi-

tudinal study and a great deal of antecedent data was available on

these subjects. The subjects were divided into four groups on the

basis of two of the CaliforniaPsychological Inventory scales, the

Femininity Scale (Above mean females = highly sex-appropriate, below

mean females = low sex-appropriate, below mean males = high sex-

appropriate, and above mean males = low sex-appropriate); and the

Socialization Scale (Individuals on the high end of the socialization

scale are conforming; while individuals on the low end are nonconforming

and asocial).









Males and females classified as being highly sex-appropriate and

highly socialized appear to have internalized parental characteristics

with respect to both sex-role and cultural proscriptions through the

process of identification with same-sex parents in a context of familial

harmony and traditional parental role definition. The men were

relaxed, competent, and comfortable with their masculinity and possessed

a rule-respecting, other-respecting value system. The women typified

the traditional concept of femininity but were somewhat dissatisfied,

indecisive, vulnerable, and lacking in spontaneity. The low-sex appro-

priate, high-socialized individuals had parents who were less traditional

in their sex-role definition and appears to offer a wider range of

behavioral and attitudinal options to their children but had established

an emotionally satisfying and value-inculcating home. Despite their

relatively low sex-appropriate scores, both men and women were comfort-

able within their sex-role and represented a blending of agentic

(instrumental) and communal (affiliative) concerns and the women lacked

the tendency to passivity that characterized the high sex-appropriate,

high socialized women. The high sex-appropriate, low-socialized indi-

viduals had like-sex parents who were neurotic, rejecting and provided

poor models for identification and cross-sex parents who were charac-

terized as somewhat seductive. Both males and females exhibited

exaggerated sex-role characteristics, were self-centered, and irre-

sponsible. Low sex-appropriate, low socialized individuals had like-

sex parents who were emotionally uninvolved and cross-sex parents who

were salient and conflict inducing. The men in this group are

described by Block as "caricatures of the 'weaker' sex" (p. 524) and

the women as assertive, rebellious, expressive, and demanding of

independence and autonomy.









The author concludes that the socialization process appears to

have differential effects on the personality development of males and

females.

For males, socialization tends to enhance experimental
options and to encourage more androgynous sex role
definitions since some traditionally feminine concerns
are emphasized along with the press to renounce nega-
tive aspects of the masculine role. For women, the
socialization process tends to reinforce the nur-
turant, docile, submissive, and conservative aspects
of the traditionally defined female role and dis-
courages personal qualities conventionally defined
as masculine. The sex role definitions and behavioral
options for women, then, are narrowed by the sociali-
zation process, whereas, for men, the sex role
definitions and behavioral options are broadened by
socialization. The achievement of higher levels of
ego functioning for women is more difficult because
individuation involves conflict with our prevailing
cultural norms. (Block, 1973, pp. 525-526)


The Concept of Sex-Role Identity


A review of the literature on the psychological differentiation

of masculinity and femininity appears to indicate that until recently

researchers have accepted the societal belief that there are clusters

of "appropriate" masculine and feminine characteristics and that they

are bipolar opposites. The implicit assumption has been that girls

ought to acquire the characteristics such as nurturance and inter-

personal sensitivity prescribed by the culture as feminine character-

istics and inhibit qualities such as independence and competitiveness

which belong in the masculine domain; and boys ought to acquire

masculine characteristics and suppress feminine characteristics.

Nearly all the published scales devised to measure masculinity and

femininity treat these two dimensions as two ends of a single continuum

and as though they were inversely correlated. That is, an individual









is defined as a sex-typed female or a sex-typed male or in some cases

sex-reversed. However, in the last few years, there has been a pro-

liferation of research designed to validate a new conceptualization

of the psychological dimensions of masculinity and femininity that

allows for the possibility that a person may develop both masculine

and feminine attributes. This theoretical model contends that while

masculine and feminine attributes differentiate the sexes to some

degree, they are not bipolar opposites; but in each sex are separate

and essentially orthogonal dimensions. Most current measures approach

masculinity and femininity in terms of socially desirable character-

istics for males and females (Bem, 1974; Berzins, Welling, & Wetter,

1978; Heilbrun, 1976; Spence, Helmreich, & Stapp, 1975). An individual

is considered sex-typed (masculine or feminine) to the extent that the

individual endorses either of the sex-stereotyped characteristics to

the relative exclusion of the other. The androgynous individual is

one who endorses relatively equal proportions of both masculine and

feminine characteristics. Within this latter category, a distinction

is made between an individual who scores high in both masculinity and

femininity (Androgynous) and one who scores low in both (Undifferen-

tiated).

Bem (1972, 1974, 1975) questions the traditional assumption that

it is the sex-typed individual who typifies mental health. She

theorizes that the androgynous individual is psychologically healthier

for he or she has at his or her disposal a larger and more diverse

number of behaviors which enables the individual to engage in situa-

tionally effective behaviors in a more adaptable fashion. By contrast,









the highly sex-typed male or female is motivated to maintain a

culturally sex-stereotyped self-image and for them cross-sex behavior

may be problematic even when it would be more appropriate for that

particular situation.


Behavioral Validation

Bem (1975) has found some behavioral support for her hypotheses

in regard to androgynous subjects and males. Feminine-typed females

did not perform well even on some tasks designed to evoke feminine

expressiveness and affection.

Bem also examined the behavior of college students in two situa-

tions: The first was a standard conformity deception paradigm

intended to evoke independence (stereotypically masculine attributed);

and the second involved an opportunity to play with a kitten intended

to evoke nurturance and expressiveness (stereotypically feminine

attributes). The students had been divided into the following four

groups by means of their scores on the Bem Sex-Role Inventory (BSRI):

Masculine (high masculine-low feminine); Feminine (high feminine-low

masculine); Androgynous (high masculine-high feminine) and Undifferen-

tiated (low masculine-low feminine). Androgynous subjects of both

sexes exhibited the hypothesized behaviors; i.e., independence when

under pressure to conform and nurturance when given the opportunity to

play with the kitten. The "masculine" males displayed independence

but not nurturance and the "feminine" males displayed nurturance but

not independence. However, the pattern for the females was less

clear. The "masculine" females did display independence as predicated

but also exhibited a moderate degree of playfulness and nurturance,









falling between the "androgynous" females and the "feminine" females

on this dimension. The "feminine" females failed to show independence

as predicated but also failed to exhibit nurturance toward the kitten.

In another study, Bem (1975) examined the relationship between

sex-roles and expressive behavior. Expressive behavior was assessed

by rating the interaction of subjects with a human infant. As

expected, androgynous and feminine-typed males were significantly

more responsive than musculine-typed males. However, the female sex-

role categories did not differentiate on the responsiveness score.

In a second study, responsiveness was estimated by observer ratings

of the subjects' verbal and nonverbal responsiveness to a same-sex

partner's description of his or her loneliness, isolation, and

homesickness. (The partner was always a confederate of the experi-

menter.) As hypothesized, androgynous and feminine-typed males were

again significantly more responsive than the masculine-typed males.

And, in this study the hypothesis was also confirmed for the females,

with masculine-typed females rated as less responsive than androgynous

and feminine-typed females.

In a third study, Bem and Lenney (1976) investigated the relation-

ship between sex-role type and willingness to engage in cross-sex

behaviors. Subjects were required to select and perform one activity

from each of 30 pairs of tasks. The pairs were: (1) a neutral and a

masculine activity, (2) a neutral and a feminine activity, and (3) a

masculine and a feminine activity. In addition, small cash payments

were given for performing the tasks and arranged such that sex-

inappropriate choices always paid more. As predicated, for each

gender, the sex-typed subjects were significantly more stereotyped in









their activity choices than the androgynous or sex-reversed subjects

despite the fact that the sex-appropriate tasks paid less money.

This series of behavioral validation studies by Bem and her

colleagues appears toconfirm her hypothesis about behavioral flexi-

bility in terms of males but the results are not as distinct for

females.


Personality Correlates to Sex-Role Mode

Several researchers have explored the relationship between sex-

role categorization and measures of self-esteem. Unfortunately,

different scales have been used to measure the same criterion variable,

but all three studies found similar results.

Spence et al. (1975) used the Personality Attributes Question-

naire (another instrument for sex-role assessment based on similar

theoretical assumptions as the BSRI) to assess the relationship of

sex-role categorization and self-esteem scores and report that

androgynous males and females report the highest levels of self-esteem,

and the undifferentiated subjects report the lowest levels. The sex-

typed subjects were intermediate between them with the feminine-typed

person of each gender being slightly lower than the masculine-typed

subjects. Wetter (1975) using the ANDRO scale (another measure for

sex-role dimensions) found that, for each gender, the masculine-typed

and androgynous subjects did not differ from one another but reported

significantly higher self-esteem than the feminine-typed and undifferen-

tiated subjects. The latter two groups did not differ from each other.

Bem (1977) using the BSRI obtained results identical to Wetter. All

three studies were conducted with college students as subjects.









Spence and Helmreich (1978) conducted a parallel analysis with a

sample of high school-aged students and found an exact duplication

of the college student results. That is, within each gender, the

order of mean self-esteem scores from highest to lowest was

Androgynous, Masculine-typed, Feminine-typed and Undifferentiated.


Integration of Changes in Sex-Role Definition
with Changes in Developmental Tasks


Feminine Sex-Typing--An Unclear Dimension


It is evident from a cursory review of the literature on the

developmental process of psychological sexual differentiation that

the social-cultural milieu affords the female in the American culture

less support in the assimilation of her sex-role. The fact that there

are cultural expectations about how a male or a female "ought" to

behave is clear but for the female it appears that the parameters of

the expectations are less distinct. Whether one espouses the psycho-

analytic identification theory of sexual differentiation or the social

learning theory or the cognitive developmental theory, there are some

inherent difficulties for the female. Modeling is certainly operative

within each of the three theories and the models themselves appear to

be less definite about what characteristics and behaviors should be

differentially instilled. Thus, researchers (Lansky, 1967; Fling &

Manosevitz, 1972) have demonstrated that parents are more definite

about what they consider to be sex-appropriate and inappropriate for

boys.









The previously cited study by Mussen and Rutherford (1963)

indicates that the feminization of girls involves a greater number

of, and more complex, determinants than the masculinization of boys.

The authors hypothesize that this is true because the male role is

more highly valued in the middle class American culture and because

behaviors considered sex-appropriate for males are more clear cut and

well-defined.

Hypothesizing that social reward for conformity should tend to

strengthen interpersonal habits, whereas social punishment for non-

conformity should have a weakening effect, Heilbrun (1964) explored

the relationship between a measure of masculinity-femininity to

perceived social role consistency. (Social role consistency (RC)

has been empirically related to level of adjustment in previous

research; i.e., high RC has been related to better psychological

adjustments.) Heilbrun predicted that the more masculine boys would

show higher RC than the less masculine boys since there was pervasive

expectancies in our culture that men should display masculine attri-

butes and these would be reinforced by his social environment and be

strengthen. On the other hand, the low-masculine boy's behavior

may be subjected to greater social censure and possibly pressure

toward the modification of his behavior. He found a significant

difference in the RC measures in the expected direction for this group

of adolescent boys. In his female population, it was found that both

high-feminine and low-feminine girls perceived themselves as more

stable in their interpersonal behaviors than those falling into an

intermediate group. The author speculates that the intermediate group









is lower in RC because girls who fulfill neither the traditional

feminine role nor a more masculine role might not provide social

behavior patterns of sufficient stability to be systematically

reinforced. He offers no speculation about why a cross-sexed

behavior pattern in girls should result in high role consistency

except to point out that: Not only is a female permitted consid-

erably more freedom as a child to engage in cross-sex behaviors but

that the feminine role itself is undergoing revision; therefore, a

female adolescent would be exposed to conflicting social rewards and

punishments. (Neither does he offer a rationale for dividing the

boys into a high and low group and the girls into three groups.)

Another piece of evidence can be found in the previously cited

longitudinal research by Block (1973). The data presented suggest

that the highly sex-appropriate, highly socialized parents produced

competent well adjusted males but somewhat dissatisfied, indecisive,

passive females despite the fact that these women typified the

traditional concept of femininity. On the other hand, the low sex-

appropriate, highly socialized parents appeared to offer a wider range

of behavioral and attitudinal options and produced both men and women

who were competent and comfortable with their sex-roles.

It appears that boys growing up in the American culture not only

have more distinct and easily discriminable cues for sex-appropriate

and inappropriate behavior but those "masculine" behaviors and attri-

butes that he ought to display (such as competence and independence)

are valued by society and he is rewarded for them. On the other hand,

a girl is allowed more latitude in sex-appropriate and inappropriate

behaviors and is often exposed to conflicting social rewards and









punishments for her behaviors and attitudes. Moreover, even when she

endorses the "feminine" behaviors and attitudes that she ought to dis-

play, this does not necessarily lead to a feeling of adequancy or are

they necessarily valued and rewarded by her social environment.


The Enigma of Femininity

Although females may not be consistently reinforced or punished

for exhibiting "feminine" behaviors and attitudes and despite the

fact that she has more latitude in engaging in cross-sex behaviors,

there is a very consistent consensus in our society about what char-

acteristics are most desirable for each or that would describe an

"ideal" or "average" man or woman.

Rosenkrantz, Vogel, Bee, Broverman, and Broverman (1968)

administered sex role questionnaires with 130 bipolar items to college

students who were asked to rate the typical adult male and the typical

adult female on each item. The student populations came from five

different settings which differed with respect to religion and social

class. There were no differences across the student subgroups; the

average adult man responses (masculinity) and the mean adult woman

responses given by male subjects and female subjects were nearly

perfectly correlated. Two different samples of students were asked to

indicate the pole of each item that they considered to be the more

socially desirable behavior for the population at large. Twenty-nine

male-valued items and 12 female-valued items were chosen by a majority

of each sample regardless of the sex of the subject. Additional samples

of men and women were given the questionnaires with instructions to

indicate that point on each item scale they considered most desirable









for an adult, sex unspecified. The same 29 stereotypic items were

closer to the masculine pole. It appears that not only are existing

stereotypic differences between men and women approved of by a large

segment of society, but there seems to be some indication that mascu-

line characteristics are more highly valued than are feminine

characteristics.

Broverman, Broverman, Clarkson, Rosenkrantz, and Vogel (1970)

administered the same questionnaire to a sample of practicing mental

health clinicians, of both sexes, with one of three sets of instruc-

tions: (1) to describe a mature, healthy, socially competent adult

man; (2) a mature, healthy, socially competent adult woman; and

(3) a healthy, mature, socially competent adult person. There was

high agreement, within each set of instructions, about which pole

reflected the more healthy behavior. The clinicians' ratings of a

healthy adult and a healthy man did not differ from each other.

However, a significant difference was found between the ratings of

the healthy adult and the healthy woman. The authors state:

Acceptance of an adjustment notion of health, then,
places women in the conflictual position of having
to decide whether to exhibit those positive charac-
teristics considered desirable formen and adults,
and thus have their "femininity" questioned, that
is, be deviant in terms of being a woman; or to
behave in the prescribed feminine manner, accept
second-class adult status, and possibly live a
lie to boot. (p. 6)

Horner (1972) focused our attention on this differential valuation

with her research on the incidence of "fear of success" found in college

women. Horner conceptualized the Motive to Avoid Success, within the

framework of an expectancy value theory of motivation and described it

as "a latent, stable personality disposition acquired early in life.. .









disposition to become anxious about achieving success" (p. 159). She

proposed that women, as a result of their sex-role training,have the

belief that successful achievement can only be accomplished at the

price of a loss of femininity and social rejection. Her procedure

involves giving subjects a verbal cue describing a woman competing

in a mix-sex situation. The cue was: "After first term finals,

Anne finds herself at the top of her medical school class." The

subjects were required to respond in a manner similar to responses

elicited with Thematic Apperception Tests and the protocols were

scored according to the presence or absence of negative imagery

related to the success. Horner reported negative imagery in 65-89%

of the women and 9% of the men. (The male cue substituted John for

Anne.)

Hoffman (1974) replicated Horner's research and introduced three

variations in the story cue: In one, the setting was changed to child

psychology, a less masculine academic area; in another, the achievement

was communicated privately rather than publicly; and finally, the

competitive aspect was minimized. None of the variations diminished

fear of success for females but 77% of the males also produced fear of

success protocols. The most common theme for females was fear of

affiliative loss as in Horner's study; for males, it was questioning

the value of the achievement. Tresemer (1976) reviewed 46 studies

involving fear of success and found no consistent tendencies for females

to report more fear of success than males. Weston and Mednick (1970)

compared black and white college women and two class levels. Social

class differences were not found but black women reported considerably


less fear of success themes.









Zuckerman and Wheeler (1975) reviewed 16 studies and found that nine

showed more fear of success imagery in women while seven reported

more such imagery in men. It is difficult to assess these data

because of differences in procedures and measuring devices but

"fear of success" does not appear to be a stable personality dispo-

sition but does seem to be related to sex-role standards and the

anticipated consequences of deviating from them, at least for women.

Wiggins and Holzmuller (1978) hypothesized that the flexibility

associated with Bem's definition of psychological androgyny is a more

general personality characteristic that subsumes sex role stereotypes;

i.e., the flexibility of androgynous persons may be part of a broader

pattern of flexibility that is expressed in all or most dimensions of

interpersonal behavior. By this reasoning, the androgynous person's

profile of interpersonal variables would be relatively flat, and the

stereotyped person's profile would be both positively and negatively

spiked on variables that are highly sex stereotyped. This hypothesis

of a flat profile for the androgynous person received considerable

support for male subjects but was strikingly disconfirmed for female

subjects. The index of profile variability was highest for androgynous

females and a mirror image of the pattern for stereotyped females.

The authors suggest that the greater profile variability of androgy-

nous females may reflect a more differentiated self-perception on the

part of androgynous females.

Jones, O'C. Chernovetz, and Hansson (1978) conducted a series of studies

to test hypotheses derived from Bem's theory of androgyny with respect

to conventionality, adaptability, social competence, and adjustment.

The subjects were eight separate samples of college psychology students.









In no case were androgynous males found to be significantly more

adaptive, flexible, or competent than masculine males, across a wide

variety of personality, adjustment, and intellectual variables. The

feminine-typed male subjects were less secure and flexible, had lower

self-esteem, were more sensitive to criticism and had more problems

with alcohol in comparison to the masculine-typed males. Androgynous

males scored significantly lower on creativity than the feminine-typed

males, whereas, there was no difference between the latter and the

masculine-typed males on creativity. Thus, with the exception of

intellectual functioning, masculine-typed males were described as more

competent and confident over a wide range of variables, whereas the

less traditionally sex-typed males were generally more limited and

restricted, less effective and more vulnerable and less secure. A

similar pattern emerged for the females. The more masculine in orien-

tation, the more adaptive, competent, and secure the female subject

was; with the masculine-typed females scoring in the most desirable

direction. The authors propose that what is being devalued in society

is perhaps not female gender but, rather, feminine behaviors. And

perhaps, they suggested,

that the important issue becomes not whether one has
internalized the traits and behaviors appropriate to
one's gender but the extent to which one has assimi-
lated the tendencies most highly valued by society .
In a society that prefers the agentic role, it becomes
reasonable to conclude that individuals high in agentic
tendencies will not only be more successful within the
context of such society's values, but such persons
will feel more confident due to a history of differen-
tial application of social rewards. (p. 311)









Sexual Identity A Facet of Self-Concept

The bulk of the research on the development of sexual identity

appears to assume that the ultimate goal is the achievement of mascu-

linity or femininity or androgyny. However, there is some evidence in

the developmental research conducted by Baumrind (1971) and Block (1973)

which indicates that the parents of the most competent, well-adjusted

children who were also comfortable with their sexuality directed their

efforts not at instilling a particular sex-role style but at providing

an environment that facilitated the development of self-identity.

Block (1973) conceives of sexual identity within the framework of

the larger developmental tasks of ego and cognitive development. She

states, "sexual identity means, or will mean, the earning of a sense

of self in which there is a recognition of gender secure enough to

permit the individual to manifest human qualities our society, until

now has labeled as unmanly or unwomanly" (p. 512).

Kelly (Maher (Ed.) 1969) points out that man's thought and behavior

is not the residue of biographical incidents, nor are they projected

facsimilies of reality. He theorizes that each man constructs guidelines

that given behavior its directionality--reference axes are devised by

each man for establishing a personal orientation toward the various

events he encounters. He states, "With such personal constructs a

man can make his entrance into the world of reality by acting with

initiative and ingenuity. Failing to erect them he can only repeat

concretely what has been 'reinforced,' in the circular manner that

psychological journals describe" (p. 37). If we can accept Kohlberg's

theory that sex-typing is dependent upon certain aspects of cognitive









growth and development, and that each individual develops his or her

own reference axis or construct of sex-role, it seems logical to assume

that the individual may revise, reverse, and add new connotations or

attributes to the concept of self as new reference axes are added upon

which to project the events that are encountered in the environment.

In addition, it seems reasonable to conjecture that, for the

bulk of the population, an individual's gender and all the nuances

associated with it, would be "figure" with changing degrees of

ascendency at different critical periods in life. Most of the recent

research generated by Bem's concept of "androgyny" has been conducted

with the college population as subjects. This is a period of life

when most young adults are not only in the process of establishing

their vocational identity but also exploring the possibilities of

forming relatively enduring sexual relationships. Perhaps it is a

period in life, more than any other, in which individuals tend to define

themselves in terms of "femininity" or "masculinity." Thus, the

fact that most of the subjects fall into a particular phase of the

life cycle may have some effect on the results derived from research

on sex-role concept.


Sex-Role and Aging

Although there are a number of useful references in the literature

that relate to personality changes in adulthood and old age, the find-

ings are usually suggestive or descriptive and often limited to obser-

vations drawn from an institutionalized older population or limited to

one gender and not easily generalizable. Kelly (1955) as part of his

investigation on the consistency of adult personality, retested 176









males and 192 females who were part of another study some 20 years

earlier. The two instruments used to assess personality variables

were the Bernreuter Personality Inventory and Strong's Vocational

Interest Inventory. Very few significant changes were noted but

there was a small but statistically significant shift toward greater

self-confidence in women and a small but significant shift in the

masculine direction for both men and women on the Masculinity-

Femininity scale and the Strong Vocational Interest Inventory

(nonvocational interest scales).

Neugarten and Gutmann (1958) explored the relationship between

role image and personality in men and women ages 40-70 years

through the use of the Thematic Apperception Test (TAT). Although

the data represented affective connotations of role behavior rather

than objective role descriptions, the authors present findings that

imply that there are personality differences that are fairly general

to the age groups (40-54; 55-70). For example, the younger group of

women produced protocols that centered around enforcing nurturant views

on events, guaranteeing benevolent outcomes to the young; in the older

group, although they still acknowledged the relevance of nurturance

and generativity, they also allow themselves to be more self-

assertive and domineering. The younger group of men produced themes

about struggles with problems relating to assertion, guilt, nurturance,

and affiliativeness-conflicts he attempts to solve in terms of complex

role patterns that integrate the various elements. The older men reached

a solution through relinquishing the assertive role elements and

abandoned any attempts at active manipulation of the environment. The









authors theorize, "For example, women, as they age, seem to become

more tolerant of their own aggressive, egocentric impulses; whereas

men, as they age, of their own nurturant and affiliative impulses"

(p. 89).

Reichard, Livson, and Petersen (1962) utilized interview data

to examine changes in personality, self-concept, and personal adjust-

ment during the period from the middle age to old age in a population

of eighty-seven men. They conclude: "The way that a man grows old

depends to a degree on his personality--on what his psychological

drives are and his ability to satisfy them in old age. Personality

has an important effect on whether a man grows old successfully and

how he goes about it" (p. 170). They were able to identify and describe

three personality types among men who adjusted well to aging and two

personality types who adjusted poorly. The three well-adjusted types

were: (1) Mature men who were relatively free of neurotic conflict,

were able to accept themselves realistically and found genuine satis-

faction in activities and personal relationships. (2) Rocking-chair

men characterized by their general passivity. They welcomed the oppor-

tunity to be free of responsibility and to indulge their passive needs

in old age. (3) Armored men who maintained a highly developed but

smoothly functioning system of defenses against anxiety about growing

old by keeping active. The two poorly adjusted types were: (1) Angry

men who were bitter over having failed to achieve their goals earlier

in life. They blamed others for their disappointments and were unable

to reconcile themselves to growing old. (2) The self-haters blamed

themselves for their misfortunes. Growing old underscored their feelings

of inadequacy and worthlessness and these men tended to be depressed.









There have been few studies in which a systematic set of variables

has been used to study the organization of personality among aged

persons. However, there seems to be suggestive evidence that disen-

gagement from central life roles during senescence may be basically

different for men and women. Kastenbaum (1964) states:

In general we might say that a woman's life long
training to a role that is primarily socio-emotional
but nevertheless includes adaptive skills leaves her
more diffusely adaptable than a man's working career
leaves him, because he does not automatically need
integrative skills. The disposition toward the
instrumental role can remain after retirement, but
the specific skills lose relevance. (p. 68)

Perhaps in senescence, the disjunction for women is far less acute

than it is for men. A man has no clearcut roles upon retirement, no

special place to go to perform it and even if he plays an instrumental

role relative to his wife, it loses its public label and utility.


Hypotheses Concerning the Relationships Between Self-Concept
and Sex-Role Style and Their Interaction With Gender and Age


According to Bem's theory of androgyny, an androgynous person has

at his or her disposal a larger and more diverse number of socially

approved behaviors which enables the individual to engage the environ-

ment in a more adaptable fashion. A logical outgrowth is an increase

in the probability of reinforcement, either self-generated or externally

delivered, which should result in a high evaluation of self and lead to

higher self-esteem. However, it has also been shown that the American

culture consistently values agentic or instrumental behaviors associated

with "masculinity" (Broverman et al., 1970; Rosenkrantz et al., 1968)

and at least one study (Jones et al., 1978) indicates that competence

and adaptability are largely accounted for by "masculinity."









It is hypothesized that a Feminine-typed female consistently

scores lower in self-esteem relative to Androgynous and Masculine-typed

individuals, not only because Feminine-typed behaviors have a lesser

potential for leading to social reinforcements in our society, but

also because she is not secure in her self-identity. That is, there

is a consensus in our culture about the attributes a female ought to

possess but her social environment does not necessarily reinforce

these behaviors when they are displayed. In addition, she may be

punished for displaying those behaviors that are valued by society and

cast doubt on the very identity she is trying to establish. This con-

flictual situation is not likely to lead to high self-esteem. Granted

this, it would be reasonable to hypothesize that older women, who

should presumably have resolved the issue of sexual identity, may

describe themselves in a feminine fashion and maintain a fairly high

self-esteem.

Erikson's (1950, 1959) theory of development represents a systematic

formulation of the effects of maturation, experience, and social inter-

action on personality organization. He proposes that the establishment

of identity involves a feeling of continuity with past and future, an

assimilation of the rapid physical changes, an adoption of sexual roles,

and a formulation of a dominant social role. These same factors that

Erikson considers crucial to identity formation in the adolescent seem,

theoretically, to confront the aged population and may pose the task of

preserving a stable self-image despite the disruption in life patterns.

For aging poses the problem of adjusting to physical and mental changes

in roles that have been central to one's view of "self" throughout









life, little social instigation to maintain sexual roles and a

relinquishing of one's vocational identity, particularly for the

males in our society.

This coupled with the fact that there is some indication that

men tend to decrease their agentic qualities and take on more com-

munal ones would lead to the hypothesis that men after retirement

would endorse more equal numbers of masculine and feminine charac-

teristics and most probably score low in both.

Kalish (1975) introduces the concept of "age role" in terms of

an individual's age group "calling forth responses and expectations

different from those they have encountered" (p. 50). Most people

are effectively socialized to the particular age roles prescribed for

their age group by society. Hence, it might be reasonable to assume

that individuals of any age group share certain qualities, feelings,

experiences, roles and changes in roles with their age cohorts. It

is also evident that many age role expectations of our culture are

imposed along sexual lines and consequently, behaviors are differen-

tially reinforced. The general hypothesis to be explored is: The

nature of the developmental task engaged in (age role) at critical

periods in life, particularly during those periods of large disruptive

changes, such as adolescence and senescence, interacts with the

individual's gender. This interaction will be reflected in the

individual's sex-role identity and its relationship to self-esteem

and other personality variables.
















CHAPTER II
EXPERIMENTAL ANALYSIS OF THE INTERACTION BETWEEN CHRONOLOGICAL AGE
AND GENDER AND ITS EFFECT ON SEX-ROLE IDENTITY


This study explores the hypothesis that an individual's sex-role

identity, unlike biological gender, is subject to the influence of

cultural norms; and the effect may not be equivalent for men and women

due to the presence of social forces that differentiate males and

females along particular psychological dimensions. It is further

hypothesized that the nature of this effect will differ contingent

on the character of the developmental task engaged in, particularly

during certain critical periods of life.


Method


Hales and females within four age groups were administered paper

and pencil questionnaires designed to measure a variety of personality

variables. In addition to the usual college population, an attempt

was made to survey individuals in the following age groups: 25-49

years of age; 50-60 years of age; and persons older than 60 years of

age. All subjects were living in their own households and functioning

within the community. The clinical and institutionalized populations

were deliberately excluded. Anonymity was guaranteed by a strict

coding scheme. The following questionnaires were administered: Bem

Sex Role Inventory, Texas Social Behavior Inventory, I-E Scale, a

measure of Role Consistency, along with a questionnaire designed to

elicit demographic information.










Subjects


The College Population

The subjects in the 18-24 year old group were male and female

college students enrolled in the Introductory Psychology course at

the University of Florida in the Spring and Fall quarters of 1977.

This population was available to participate in various psychological

experiments as part of their course requirement. Fifty-eight subjects

(40 females and 18 males) took part in the experiment in the Spring

quarter. In the Fall quarter, 22 male subjects were recruited to

yield an equal number of males and females (N=80; 40 females, 40

males). The subjects were administered the packet of questionnaires

in two evening group sessions. Instructions were delivered to the

group as a whole at the beginning of the sessions and the experimenter

remained available for any questions that might arise. This college

sample is analogous to the criterion samples utilized by Bem (1974)

for the Bem Sex Role Inventory; by Helmreich et al. (1974) for the

Texas Social Behavior Inventory; and by Rotter (1966) for the I-E

Scale.


The Older Age Groups

The college students participate in psychological experiments for

course credit and are relatively homogeneous not only in motivation

but on many other relevant characteristics. Once an experimenter leaves

the classroom in search of an analogous adult population, it is never

certain, in what manner and to what extent the sample will be selective

by virture of its being composed of individuals who volunteer to par-

ticipate in a psychological experiment. The intent was to study a










sample of adults resembling the college population and fairly represent-

ative of the general population. Data were collected over a span of

nine months, from June, 1977,to February, 1978. Organized groups

were reluctant to divulge their membership and when permission was

obtained, the members were not willing to contribute their time during

the meeting. Therefore, instructions were given, consent forms signed,

and the volunteers were permitted to complete the questionnaires at

home. Thirty-seven individuals provided data in this manner. The

"snow-balling" technique proved to be a more efficient means for

collecting data. Subjects themselves provided other subjects who met

the criteria for inclusion in this study and the questionnaires were

individually administered. The nonstudent samples consisted of 77

subjects 25-49 years of age (39 females and 38 males); 55 subjects,

50-60 years of age (30 females and 25 males); and 62 subjects, over

60 years of age (32 females and 30 males). Demographic data are

presented in Tables 1, 2, and 3.


Table 1
Marital Status


25-49 Years 50-60 Years Over 60 Years
Marital
Status Males Females Males Females Males Females
Married 26 21 23 23 21 18
Single 3 4 1 4
Divorced 5 7 2 3
Widowed 1 3 5
Separated
Remarried 4 5 2 4 5 2
Other 1*
TOTALS 38 38** 25 30 30 32


* Open Marriage


**One individual failed to complete t re


-










Table 2
Educational Level


25-49 Years


50-60


Years


Over 60 Years


Education Males Females Males Females Males Females
College Plus 25 18 13 11 20 18
College 4 9 3 7 5 8
Partial
College 8 8 3 7 1 2
High School 1 2 6 5 2 4
Partial High
School 1 2
TOTALS 38 38 25 30 30 32



Table 3
Work Status of Women



25-49 Years 50-60 Years Over 60 Years

Full-Time
Housewife 11 16 14
Working
Full-Time 21 9 1
Working
Part-Time 5 5 2
TOTALS 37* 30 17**


* 2 women failed to complete this portion of the questionnaire
** 15 women over 60 years of age had been employed full-time until
retirement and were now unwilling to categorize themselves as
Full-Time Housewives

Twenty-one males over 60 years of age were retired, 8 men were

working full time and 1 male was retired but working part time.

This is evidently a very well educated sample, with 13% of the non-

student population receiving formal training beyond a standard college

education. An attempt to deliberately sample a population of less

educated individuals was abandoned because professionals involved with


---










Older Americans in various programs advised the experimenter that these

questionnaires designed for the college students might intimidate their

members. In general, individuals with some college experience were

more likely to serve as volunteers in this psychological experiment.

Five women over 60 years of age volunteered under fortuitous circum-

stances but were not able to complete the task.


Assessment Instruments


Demographic Data Sheet

A questionnaire designed to elicit relevant demographic informa-

tion. The data have been summarized in the previous section.


Bem Sex-Role Inventory (BSRI) (Bem, 1974)

The BSRI treats masculinity and femininity as two independent

dimensions. This makes it possible to characterize individuals as

Masculine, Feminine or Androgynous as a function of the difference

between his or her endorsement of masculine or feminine personality

characteristics. Bem's scale contains 60 personality characteristics:

20 masculine-typed items, 20 feminine-typed items, and 20 items that

are neutral with regard to sex. In the development of the scale, the

masculine and feminine characteristics were so designated on the basis

of sex-typed social desirability. That is, a characteristic was

selected for the masculine category if it was judged to be more desir-

able for a man than a woman in our society; and feminine if it was

judged to be more desirable for a woman than a man.










Bem conceptualizes a sex-typed individual, whether masculine or

feminine, as one who has internalized society's sex-typed standards of

desirable behaviors for men or women and endorses one group of sex-

stereotyped characteristics to the relative exclusion of the other on

a self-descriptive inventory such as the BSRI. The androgynous

individual, on the other hand, endorses relatively equal numbers of

masculine and feminine items. Bem hypothesizes that an androgynous

self-concept might permit an individual to engage in both "masculine"

and "feminine" behaviors depending on the situational appropriateness

of the behavior. Thus, her concept of androgyny denotes a flexibility

which enables an individual to respond to shifting situational demands

and in turn leads to greater social competence.

Spence et al. (1975) and Strahan (1975) extended Bem's concept of

androgyny by proposing that the absolute number of response options co-

varies with sex-role diversity to enhance behavioral flexibility. That

is, an individual whose self-definition restricts both masculinity and

femininity (low-masculine--low-feminine) would have fewer behavioral

alternatives and thus be at an adjustment disadvantaged, particularly

in comparison to an androgynous individual (high-masculine--high-feminine).

In response to this proposal, Bem (1977) amended her concept to endorse

Spence et al.'s designation of "Undifferentiated" individuals, who due

to a low degree of both masculine and feminine characteristics is seen

as less adaptive than androgynous individuals.

Normative data for the BSRI were obtained from 444 male and 279

female Introductory Psychology students at Stanford University and 117

male and 77 female paid volunteers at Foothill Junior College. The










internal consistency scores (coefficient alpha) were found to be highly

reliable, both in the Stanford sample (Masculinity a = .86; Femininity a =

.80; Androgyny a = .85) and in the Foothill sample (Masculinity a = .86;

Femininity a = .82; Androgyny a = .86). Masculinity (Stanford sample

r = .42 for males and .19 for females; Foothill sample r = .23 for males

and .19 for females) and Femininity (Stanford sample r = .28 for males

and .26 for females; Foothill sample r = .15 for males and .15 for

females) correlated with Social Desirability. In contrast, the much

lower correlations between Androgyny and Social Desirability (Stanford

sample r = .12 for males and .03 for females; Foothill sample r = .07

for males and .06 for females) suggest that the Androgyny score is not

a measure of a general tendency to respond in a socially desirable

direction. The BSRI was administered for a second time, four weeks

later, to 28 males and 28 females from the Stanford normative sample.

Test-retest product correlations were: r = .90 for Masculinity;

r = .90 for Femininity; and r = .93 for Androgyny.


Texas Social Behavior Inventory (TSRI) (Helmreich et al., 1974)

The TSBI is a 32-item multiple choice scale designed to assess

individual perceptions of social competence and self-esteem. Subjects

are given five alternative choices (Not at all, Not very, Slightly,

Fairly and Very characteristic of me) in response to declarative

statements describing behaviors associated with either high or low

social competence. Each item is scored from 0 associated with low

social competence to 4 the response characteristic of high self-esteem.

The total score for each subject is the some of all items with a pos-

sible range of 0 to 128.










The TSBI was standardized on male and female students in Intro-

ductory Psychology at the University of Texas at Austin. The test

developers report no significant differences between the sexes on

total scale scores. Test-retest reliability is reported as r = .94

for males and r = .93 for females.

An oblique four factor rotation was computed for males and

females. Helmreich et al. (1974) reported the following factors:

72.3% of the variance for males and 72.8% for females was found to be

associated with the first factor, composed primarily of items dealing

with confidence. The second factor for both sexes was weighted most

heavily with items concerned with dominance and the third with social

competence. A fourth factor was related to social withdrawal for males

and relations with authority figures for females.

The authors report a strong correlation between the TSBI and

the California Personality Inventory (another measure of self-esteem)

r = .50 ( p < .001) for males and r = .52 (p < .001) for females. The

TSBI was not significantly related to intelligence as measured by the

Scholastic Aptitutde Test. For males, the TSBI was orthogonal to the

Marlow-Crowne Social Desirability Scale (r = .09, not significant),

however, a significant correlation was found for females (r = .32,

p < .01). The authors suggest that a definite though modest relation-

ship exists between the expression of socially desirable characteristics

and self-reports of social competence in females.

Further, the authors report a strong relationship of the TSBI to

the Personal Attributes Questionnaire (Spence et al., 1974), a measure

of sex-role identity. For males, the correlation (r = .54, p < .001)









indicates that the more stereotypically masculine a male rates himself,

the higher his self-esteem. The correlation for females (r = .59,

p < .001) is in the same direction and stronger, suggesting that the

more a woman attributes masculine characteristics to herself, the

higher her self-esteem.


Locus of Control (I-E Scale) (Rotter, 1966)

Locus of control is a construct generated within Rotter's social

learning theory which refers to the extent to which an individual per-

ceives events in his life as being a consequence of his own actions

and therefore under his control. The I-E Scale is a forced-choice

test and contains 23 items plus 6 filler items intended to make some-

what more ambiguous the purpose of the test. All the critical items

are concerned with the subject's expectations about how reinforcement

is controlled and the test is intended to measure generalized expectancy.

It is an additive scale and the subject's score is the total number of

external choices.

The I-E Scale was standardized on a large number of Elementary

Psychology students at Ohio State University. Test-retest reliability

over one and two month periods range from r = .49 to r = .83. The

internal consistency estimates are relatively stable, with correlations

ranging from r = .65 to r = .79. Correlations with the Marlow-Crowne

Social Desirability Scale range from r = .12 to r = -.41.

An estimated 600-650 articles have been published dealing with

some aspect of the internal-external control dimension. In general

the studies lend support to the construct of internal-external control

by forming clusters of personality characteristics: The externally










controlled individual is relatively more anxious, aggressive, dogmatic,

less trustful, and more suspicious of others because he perceives him-

self as having very little control over his environment. The internally

controlled individual is more sociable, achievement oriented, and more

trustful of others because he thinks of himself as being able to con-

trol his environment. One interesting finding (Hersch and Scheibe,

1967) was that internals were a more homogeneous group and consistently

linked to indices of social adjustment and personal achievement; while

the externals were more diversified psychologically.


Role Consistency (Block, 1961)

The core meaning of Erickson's concept of "ego identity" is

expressed in his statement that "the sense of ego identity is the

(individual's) accrued confidence that (his) inner sameness and con-

tinuity are matched by the sameness and continuity of (his) meaning

for others. ." (p. 228). Three elements are presented in this

definition. First, an individual must perceive himself as having

"inner sameness and continuity" (i.e., he must, over time, presume

himself to be essentially the same person he has been). Second,

people in his social environment must perceive a "sameness and con-

tinuity" in the individual. Third, the individual must have "accrued

confidence" in a correspondence between the two lines of continuity;

i.e., the person he sees himself as being must be validated by feedback

from his interpersonal experiences.

Block (1961) focuses on one aspect of ego identity, the dimension

he labels role variability. The meaning of role variability is clari-

fied by describing its extremes. At one end of the dimension is "role









diffusion," where an individual has no internal reference which can

affirm his continuity and self-integrity. This kind of person is highly

variable in his behaviors and is plauged by self-doubts. At the other

extreme, Block speaks of "role rigidity," where an individual behaves

uniformly in all situations, disregarding the different responsibilities

imposed by different circumstances. Block states: "Somewhere in between,

presumably, a proper balance can be struck in the struggle both for

identity and the capacity for intimacy" (p. 392).

In order to explore the problem of the amount of personal consis-

tency necessary for psychological adjustment, Block (1952) studied a

single subject intensively. She was asked to describe systematically

her interactions with a set of "relevant others." When these descrip-

tions were factor analyzed, it was observed that the factor dimensions

appeared to order and to summarize the several kinds of roles this

subject manifested. He states: "Although seeing herself as changing

from relationship to relationship, a general factor of some consequence

proved to underlie all her interactions" (p. 285). From this frame of

reference, he hypothesized that "the amount of interpersonal consistency

is curvilinearly related to the degree of maladjustment, as defined

independently" (p. 285).

The Role Consistency measure is a direct translation into psychometric

form of Erikson's first criterion of ego identity. Subjects are asked to

rank order a set of 20 self-descriptive adjectives from most to least

characteristic of themselves for each of eight interpersonal situations

in which they imagine themselves involved. The Role Consistency measure

(Coefficient of concordance) is a type of multiple rank-order correlation










and may range from .00 to 1.00, with higher values indicating greater

perceived consistency of behavior over interpersonal situations

Forty-one college students in a class on factor analysis collected

data on their Role Consistency measure plus a Psychoneuroticism scale

developed by the author (Block, 1961). The product-moment correlation

between the index of interpersonal consistency (role consistency) and

scores on the Psychoneuroticism scale was r = .52, p < .001. Indi-

viduals who tended to see themselves as varying from interaction to

interaction were also more maladjusted as measured by the Psycho-

neuroticism scale. The expectation that individuals with too little

role variability would also prove to have weaknesses in their person-

ality makeup was not confirmed.

To date, experimental results indicate that high Role Consistency

is positively related to psychological health. In a study of 50 Vassar

alumnae some 20 years after graduation, Block (1961) found Role Con-

sistency correlated (r = .29, p < .05) with a consensus rating of

degree of adjustment. Women with high Role Consistency scores were

relatively "indulgent and forgiving, protective of those close to her,

sympathetic, efficient, adequate in her sexual role, turned to for

advice and reassurance, facially and gesturally expressive, and con-

siderate"; while women with low Role Consistency scores were described

as "irritable and overreactive, talkative, ostentatious, and sarcastic"

(p. 395). Meltzer (1957) found that a large self-ideal self-discrepancy--

a reasonable measure of self-recognized maladjustment--was significantly

related to extreme role variability (cited in Block, 1961). Cartwright

(1957, 1961) found that Role Consistency increased with psychotherapy









independently rated as successful. Heilbrun and Lair (1964) found

that psychiatric patients with high Role Consistency scores were

rated as significantly better in both their ability to communicate

and to socialize in ward interactions than patients with low Role

Consistency scores.


Theoretical Relationships Between Assessment Instruments
and Derivation of Specific Hypotheses

Traditional formulations of sex-typing were predicated on the

assumption that the adoption of sex-roles sanctioned by society for

one's gender is not only desirable but that deviation from the cultural

norm is maladaptive. Recent approaches to the assessment of masculinity

and femininity consider these dimensions to be independent and ortho-

gonal. Within this model, Bem (1974) suggests that an androgynous

individual, one who endorses a relatively balanced and high number of

sex-typed characteristics, may be more flexible, adaptive and psycho-

logically healthier. Many investigators have designed research to

relate current measures of sex-role orientation to indices of psycho-

logical well-being.

This study attempts to integrate sex-role identity within the

larger framework of self-identity. Under this theoretical framework,

sex-role identity is viewed as one aspect of "self-identity," and

will be related to three other personality variables that have been

associated with "self-identity." These particular psychological func-

tions appear to have a logical theoretical relationship to each other

and to good psychological adjustment: (1) Unity--a sense of inner

sameness and continuity (Role Consistency measure); (2) A sense of










inner competence, which is here translated into a belief that events

are contingent on one's behavior (measure of Locus of Control); and

(3) A sense of personal importance or self-esteem (Texas Social

Behavior Inventory).

According to Erikson (1959), identity formation involves a

feeling of continuity with past and future, an assimilation of the

rapid physical changes an adoption of sexual roles, and a formulation

of a dominant social role. In this study, the measure of Role Con-

sistency, the individual's perception of himself as behaviorally

consistent across interpersonal situations, serves as a crude estimate

of "ego identity." In terms of normal psychological development, the

task of establishing a sense of one's own identity becomes acute in

late adolescence and early adulthood (the college population). The

bulk of research to date has indicated that students seeking therapy,

psychiatrically disturbed geriatric patients, and other samples of

psychologically maladjusted individuals tend to experience little

role consistency. Therefore, a relatively high Role Consistency score

will be regarded as an indication of good self-identity.

Although the Texas Social Behavior Inventory was designed by

the authors to assess individual perceptions of social competence and

self-esteem, their computed factor analysis indicates that a large

portion of the variance (72.3% for males and 72.8% for females) is

found to be associated with the first factor, primarily concerned with

"confidence." While self-esteem can be defined in terms of the value

with which one is held by others and by him or herself, confidence

reflects the expectancy held regarding one's ability to control his or










her environment to obtain the reinforcements that would lead to positive

self-esteem. Rotter's measure of Locus of Control (I-E Scale) is a

rough estimate of what can be considered a person's expectancies about

control. It would seem likely that an internal locus of control (or

confidence that reinforcements are contingent on one's own behavior)

would enhance the occurrence of positive self-esteem. Thus, an internal

locus of control would appear to increase the probability of a relatively

high self-esteem, which in turn is a necessary component of a good

self-identity.


Hypotheses Related to Sex-Role Identity and Age Group

The research data to date indicate that sex-role orientation may

have differing implications for the psychological well-being of males

and females, in the American culture. Group differences consistently

favor the presence of relatively high masculinity scores (either alone

or as a component of the androgyny score), regardless of sex, for good

psychological adjustment. However, the scales currently in use were

all constructed and standardized on middle-class American college

students; and most of the research on psychological correlates of

sex-role orientation has also been conducted on college populations.

It can be expected that individuals belonging to the same culture

or subculture (the college population) will be relatively similar in

identifying those personality attributes defining masculinity and

femininity. Central to individual sex-role identity is the degree to

which the individual believes he or she measures up to or believes it

is important to measure up to his or her abstract conception of what

it is to be masculine or feminine. Although a majority of individuals










may consider themselves as acceptable members of their sex, it is

reasonable to expect considerable variability amongmen and women in

the constellation of sex-typed characteristics they possess and in

the nature of their self definition of masculinity and femininity.

Further, one's sex-role identity, like other aspects of self, may be

expected to change in response to the experiences one encounters over

the life span. These changes may be most easily observed during

critical developmental phases, when the environmental demands for

adjustment are most severe. In late adolescence (college age sample),

when individuals are struggling to establish a comfortable sexual

identity, those traits generally associated with masculinity and

femininity by our culture are apt to be cogent forces in the indi-

vidual's self-concept.

On the other hand, Neugarten and Gutmann (1958), in their

Thematic Apperception Test study of age-sex role and personality,

present suggestive evidence which implies that for the aged male and

female, those attributes that the researchers have labeled "masculine"

and "feminine" may not be central to his or her self-definition. Their

most striking finding was the fact that with increasing age of respon-

dent, there seemed to be a role reversal in regard to authority in the

family. The authors state, "these older men seem to be relatively

acceptant of 'womanly' qualities in themselves and to feel little need

to deny or limit these qualities. .the majority of women see the

older woman in terms of assertive, intrusive qualities" (pp. 28-29).

Rosen and Neugarten (1960) used abbreviated Thematic Apperception

Test protocols to study ego functions in samples of fully functioning

individuals, aged 40-77 years. They hypothesized that with increasing









age there will be less energy available to the ego for maintaining

involvements in the outer world. The dimensions studied were:

(1) The ability to integrate wide ranges of stimuli; (2) The readiness

to perceive or to deal with complicated, challenging, or conflictual

situations; (3) The tendency to perceive vigorous and assertive

activitity; and (4) The tendency to perceive or to be concerned with

feelings and affects as they play a part in life situations (p. 63).

Rosen and Neugarten report that the aged population tends to respond

to inner rather than to outer stimuli, to withdraw emotional involve-

ments, to give up self-assertiveness, and to avoid rather than to

embrace challenge. These changes that occur in the aged population,

as reported by Rosen and Neugarten, would seem to have a greater

effect on the agentic or instrumental characteristics.

For the aged male, the "withdrawal of emotional investment, the

giving up of self-assertiveness, and the avoidance of challenge" is

in accord with the reported acceptance of "womanly qualities" reported

by Neugarten and Gutmann (1958).

Hypothesis 1: It is hypothesized that relative to all the
other age groups, a proportionally greater
number of the sample of men over 60 years of
age will endorse more "feminine" character-
istics and a lesser degree of "masculine"
traits.

There is a conflict in trends reported by the two studies with regard

to the females over 60 years of age. One study reports a greater

acceptance of assertiveness, while the other reports a lessening of ego

energy invested in self-assertiveness. Therefore, no comparable changes

can be hypothesized for women over 60 years of age.

Hypothesis 2: For females, it is hypothesized that there will
be no significant differences across age groups
in the endorsement of masculine and feminine
characteristics.










Hypotheses Relating Sex-Role Identity to Other Dimensions
of Self-Identity, and the Interaction With Age

Block (1973) presents a conceputalization of sex-role development

that establishes a relationship between sex-role identity and personal

maturity; specifically, Loevinger's (1966a) stages of ego development.

A brief sketch of her stages is presented: (1) In the earliest stage,

the infant's task is to distinguish self from nonself and Block con-

siders gender to be non-relevant at this stage. (2) At the impulse-

ridden level, gender identity is essentially sexless, although the

characteristic behaviors of the child include those that have been

defined as masculine (concern with self-assertion, self-expression,

and self-interest). (3) At the self-protective stage, the young child

is still concerned primarily with the extension and enhancement of

self, largely against the imposition of rules by socializing agents.

(4) At the conformity stage, Block first recognizes the bifurcation

in sex role development of boys and girls. Socializing patterns

impinge differentially on the two sexes. (5) At the conscientious

stage, the individual undertakes the examination and evaluation of

self with respect to certain abstract values and ideals. Block states,

"Notions about the 'kind of person I would like to be' are developed

and behaviors are moderated in accordance with internalized values"

(p. 514). (6) At the autonomous stage, the individual engages in a

series of attempts at conflict resolution and the differentiation of

self. At this stage, awareness develops of values, predispositions,

and behaviors that depart from traditional sex-role definitions.

(7) At the integrated stage (Loevinger's highest level of ego func-

tioning) the individual evolves for himself or herself an identity









consonant with history and aspiration. With regard to sex-role

identity, the definition derived by the individual represents an

integration of traits and values, both masculine and feminine. Block

refers to this individual as "androgynous."

Block (1973) reviewing her data from the longitudinal study

(cited earlier in the Introduction) concludes that socialization for

women, regardless of level of femininity as indexed by the California

Psychological Inventory, "becomes associated with control of impulse

expression and the renunciation of achievement and autonomy" (p. 523).

She reports an inverse relationship between upward occupational mobility

and femininity and adds,

Socialization tends to mitigate against career interests
in women, but among those women who elect to enter the
occupational arena, advancement in status is more likely
to be achieved by women who diverge from the traditional
feminine sex role stereotype. However, this advancement
is achieved at some personal cost since communal, inter-
dependent connectedness with others is suppressed and
agency is exaggerated." (p. 525)

This concept is supported by two other independent sources,

Jones et al. (1978) found that the more masculine in orientation the

female, the more adaptive, competent, and secure the subject. They

state, "What was unanticipated was that females who completely violated

societal sex-role expectations appear to be happier, more competent,

and more adaptive than either androgynous or sex-typed females" (p. 311).

Wiggins and Holzmuller (1978) found that the index of profile vari-

ability was highest for androgynous females. They suggest that, "The

greater profile variability of androgynous females may reflect a more

differentiated self-perception of their part" (p. 51). An implicit

corollary to the views and conclusions expressed by these researchers










is the notion that, because women need to counter social resistance

and/or personal ambivalence to arrive at a perception of themselves

as "masculine-typed" or "androgynous," they are more likely to have

developed a greater sense of individuality based on their own inner

characteristics.

Conversely, it is possible that the college-aged, "feminine-

typed" female selects for emphasis that constellation of characteris-

tics consistent with the demands of the role that she values most

highly, at that period in life, in order to validate her femaleness.

That is, her sense of self-identity is not secure or well-developed

and she relies on their correspondence to traditional standards of

behavior for her definition of "what is it to be feminine."

Bem hypothesizes that psychological androgyny affords greater

flexibility and consequently leads to better adjustment and high self-

esteem. Under the larger umbrella of self-identity, one can conceive

of androgyny resulting from a more integrated personality structure

or stronger self-identity which permits an individual to develop his

or her own values and to adopt or reject the sex-role behaviors pre-

scribed by our society, as they suit his or her own idiosyncratic

needs and capacities.

This study explores the general hypothesis that sex-role identity

is a central organizing force in the establishment of self-identity

during early adulthood (the college sample) but not necessarily a

dominant force in later life, particularly in old age. Chronological

age is introduced as a variable that may interact with sex-role identity.

Throughout this study, chronological age is used only as a convenient

index for representing events that occur with the passage of time and

has no connotation of developmental sequence.











Hypothesis 3:
















Hypothesis 4:











Hypothesis Relating


It is hypothesized that, within the college
age group, knowledge of sex-role category
membership will increase the predictability
of significant relationships between the
other psychological variables. That is,
individuals who are categorized as Undifferen-
tiated or Feminine-typed will have relatively
low self-esteem scores, low role consistency
scores and external locus of control. By con-
trast, individuals who are categorized as
Androgynous or Maculine-typed are more likely
to have higher self-esteem scores, higher role
consistency scores and a more internal locus
of control.

It is hypothesized that, within the other three
age groups, sex-role categorization will not
serve as an organizing influence and will not be
predictive of the relationships to the other
measure of self-identity. That is, there will
be no significant differences in self-esteem
scores, role consistency scores or locus of
control scores in relation to sex-role categori-
zation.


Sex-Role Identity to Demographic Data


In order to allow comparisons with the much researched college

population, the data for this study were obtained from a similar popu-

lation, which essentially stratified the sample with regard to education,

and by extension, socio-economic level. Aside from this practical

decision, however, it was also considered justifiable to assume that a

greater proportion of men and women in this intellectually elite popu-

lation would be facing or have faced the necessity of coping with

contradictory roles. (It is presumed that an individual who endorses

cross-sexed or androgynous characteristics has acknowledged to some

extent conflicting internal and external needs.) This is, of course,

an unsettled empirical question; but Spence and Helmreich (1978) present


data that lend some support to this speculation.


__ I









During the academic years 1974-76, 1,769 junior and senior high

school students (756 males and 1,013 females) were administered a

battery of personality questionnaires. These students represented a

broad range of socio-economic levels and religious backgrounds. The

authors report significant class differences in their measure of edu-

cational aspiration (an indication of the least amount of education

that would satisfy the individual). This in turn separated between

their sex-role categories.

For males, 64% of the Upper middle class have high
aspirations, in contrast to 54% of the Lower middle
class. (High aspiration here denotes a college
education.) For females, the corresponding figures
are 55% for the Upper middle and 47% for the Lower
middle. In both the Upper and Lower middle
classes a significant separation between PAQ categories
(Personal Attitudes Questionnaire Instrument for the
assessment of masculinity and femininity analogous to
the BSRI) and aspirations is found, with a higher per-
centage of Androgynous and Masculine individuals having
high aspirations. Females provide a more interesting
pattern of results than males, dividing significantly
into three distinct groups. Masculine females are
most likely to have high aspirations, followed at some
distance by Androgynous and finally by Feminine and
Undifferentiated females. (p. 96)

Hoffman and Fidell (1977) found two demographic characteristics

to differentiate middle class women along the dimension of sex-role

categories. In their sample of 369 middle class, 20-59 year old

women, they report significant differences in employment status.

Masculine and Androgynous women tended to work while Feminine and

Undifferentiated women tended not to work (X= 19.49, p < .01).

Masculine and Androgynous women tended to work full time, while

Feminine and Undifferentiated women were employed part time or not at

all (X= 25.49, p < .001). The four sex-role categories in their

sample did not differ significantly in marital status, race, religion,


or birth order.









In this study, it is hypothesized that there will be a significant

difference in work status and in marital status among the sex-role

categories.

Hypothesis 5: It is hypothesized that a significantly greater
proportion of masculine-typed and androgynous
women will fall into the full time or part time
employment category.

Spence and Helmreich (1978) studied several unique populations

in which the distribution across the sex-role categories might be

expected to differ from those found in college students: 161 male

and female academic scientists; 110 male and female homosexuals; and

41 female varsity athletes. These special groups were selected for

conceptual validation of their measure of sex-role identity (Personal

Attributes Questionnaire) by demonstrating its ability to discriminate

between the college population and selected subgroups. In their college

population, they found the following percentage of each sex falling

into each category: Males 32% Androgynous, 34% Masculine, 8%

Feminine, and 25% Undifferentiated; Females 27% Androgynous, 14%

Masculine, 32% Feminine, and 28% Undifferentiated.

In the homosexual group, a dramatic reduction in the number of

conventionally classified individuals was found. The authors state,

Only 9% of the males are Masculine and only 13% of the
females are Feminine, but the redistribution into the
other categories differs greatly between the sexes. Of
the homosexual males, 50% were Undifferentiated, while
23% were classified as Feminine and 18% were designated
Androgynous. Among lesbians, the largest groups was
Androgynous (33%), but the percentages classified as
Undifferentiated (32%) and Masculine (22%) were also
elevated. (Spence & Helmreich, 1978, p. 67)

Within the group of female varsity athletes, they found the Feminine

category to be the smallest in number. The Masculine category was

greatly inflated while the Undifferentiated category was somewhat









reduced. The largest group was composed of those women classified as

Androgynous. The distribution for male scientists was similar to that

of college males, with a slight reduction in the proportions of

Feminine and Undifferentiated individuals. The results for female

scientists, however, showed a very different distribution from that

of college women. "The largest group is Androgynous, with 46% of the

women falling in this classification. The proportions of female

scientists in the Masculine category (23%) is also markedly higher than

in the college sample. The greatest reduction is in the Undifferentiated

classification, with only 8% falling into this group" (Spence & Helmreich,

1978, pp. 69-70).

The pattern of results for these groups is consistent with the

theoretical expectation that females who demonstrate a considerable

level of instrumental orientation to attain their sexual or social roles

will endorse a greater number of "masculine" characteristics. In our

culture, marriage is considered the traditional and expected role for

females. It can then be considered a break in convention for a female

to remain single, get divorced, or remarry; in contrast to those females

who are married or widowed.

Hypothesis 6: It is hypothesized that a greater proportion of
the women in the "un-traditional" marital status
categories will be Masculine-typed or Androgynous.
















CHAPTER 3
RESULTS



Analyses of Data Related to the Bem Sex-Role Inventory


The sex-role categories are established by first determining the

median masculinity and femininity scores for the sample. Those subjects

who score above the masculinity median and below the femininity median

are then classified as "masculine"; those who score above the femininity

median and below the masculinity median are classified as "feminine";

those who score above both medians are classified as "androgynous"; and

those who score below both medians are classified as "undifferentiated."

The median masculinity and femininity scores for Bem's normative sample

of college students were 4.89 for masculinity and 4.76 for femininity.

Table 4 presents the medians, means,and standard deviations for the

individuals in this sample.



Table 4
Bem Sex-Role Inventory Scores


Age Group
Entire Population (274)
18-24 years (80)
25-49 years (77)
50-60 years (55)
Over 60 years (62)
S.D. = Standard Deviation


Femininity Scores

Median Mean S.D.
4.95 4.88 .545
4.95 4.88 .564
4.83 4.83 .526
5.02 4.93 .583
5.05 4.90 .517


Masculinity Scores

Median Mean S.D.
5.10 5.06 .790
5.08 5.03 .774
5.20 5.16 .860
5.15 5.09 .691
4.98 4.93 .799









Table 5 presents comparable data separately for males and females.


Table 5
Bem Sex-Role Inventory Scores

Males Only


Age Group
18-24 years (40)
25-49 years (38)
50-60 years (25)
Over 60 years (30)


Femininity Scores

Median Mean S.D.
4.72 4.62 .440
4.65 4.60 .527
4.65 4.63 .508
4.70 4.72 .609


Masculinity Scores

Median Mean S.D.
5.28 5.34 .651
5.72 5.60 .644
5.40 5.44 .516
5.08 5.15 .714


Females Only


18-24 years (40) 5.10 5.14 .557 4.72 4.72 .772
25-49 years (39) 5.02 5.04 .431 4.80 4.74 .841
50-60 years (30) 5.32 5.17 .528 4.82 4.80 .692
Over 60 years (32) 5.12 5.06 .345 4.92 4.72 .828




There was a tendency for men between the ages of 25-49 years of age

to endorse more masculine characteristics and for women over 50 years of

age to endorse more feminine characteristics. Hypothesis 1 predicted

that, relative to all the other age groups, a proportionally greater

number of men over 60 years of age would endorse more "feminine" charac-

teristics and less "masculine" characteristics. Hypothesis 2 predicted

that for females there would be no significant differences across age

groups in the endorsement of masculine and feminine characteristics.

Hypotheses 1 and 2 were tested. A two-way analysis of variance was

performed, separately for masculinity scores and femininity scores, with

gender and age group as the independent factors. There was a significant


--








interaction for both masculinity scores (F = 6.778, 4 d.f., p< .001)

and for femininity scores (F = 11.840, 4 d.f., p< .001). The main effect

for gender was also significant for masculinity scores (F = 25.847,

1 d.f., p< .001) and for femininity scores (F = 44.608, 1 d.f., p< .001),

indicating that the masculinity and femininity scores are related to

gender. Across all age groups, males had higher masculinity score means

and lower femininity score means; while females had higher femininity

score means and lower masculinity score means. A one-way analysis of

variance was performed, separately for males and females, across the

four age groups, for masculinity scores and femininity scores. For

males, there was a significant (F = 47.25, 3 d.f., p< .001) difference

among the mean masculinity scores across the age groups, but no signifi-

cant differences in femininity scores. T-tests were performed on mean

masculinity scores, comparing the four age groups. The masculinity

score mean of men over 60 years of age differed significantly (t = 2.704,

66 d.f., p< .01) from that of males 25-49 years of age. Men 25-49 years

of age attributed higher levels of masculine traits to themselves than

men over 60 years of age. There were no significant differences in

masculinity or femininity score means across the four age groups for

the females.

The medians for the college age group (masculinity score means = 5.08,

femininity score median = 4.95) were almost identical to the medians for

the entire sample (masculinity score median = 5.10, femininity score

median = 4.95). Therefore, yielding to the convention of other researchers,

the college medians were used as the basis for deriving the four sex-role

categories.








Chi-square analyses were performed with respect to this categori-

zation, comparing the four age groups. There were no significant

differences in the distribution of females into the four sex-role

categories, across the four age groups. There was a significant

difference in distribution between males over 60 years of age and males
2
25-49 years of age (X = 8.48, 3 d.f., p< .05) and males 50-60 years of
2
age (X = 12.46, 3 d.f., p< .01). In both instances, a significantly

greater number of males over 60 years of age fell into the Undifferen-

tiated category and a fewer number in the Masculine category in compar-

ison to the other two age groups.


Analyses of Relationships Between Independent
Variables and Dependent Variables

In this study, the independent variables are attribute variables

and thus cannot be manipulated. Subjects were assigned to subgroups on

the basis of their status on the attribute variables resulting in unequal

numbers of subjects in the cells of the design. This signifies a cor-

relation between the independent variables. Therefore, the more general

multiple regression analysis technique was selected to analyze the data.

The following analyses were designed to test Hypotheses 3 and 4. Hypoth-

esis 3 predicted that within the college age group, the Undifferentiated

and Feminine-typed subjects would have relatively lower self-esteem and

role consistency scores and have a more external locus of control; while

the Androgynous and Masculine-typed subjects would have higher self-

esteem and role consistency scores and would be more internal in locus

of control. By contrast, Hypothesis 4 predicted that within the other

three age groups, there would be no significant differences in self-

esteem, role consistency scores, or locus of control, in relation to

sex-role categorization.









Multiple regression analysis is suited to studying the influence

of several independent variables on a dependent variable by helping to

"explain" the variance of the dependent variable. The predictor or

independent variables are gender, sex-role category, and chronological

age. The dependent or criterion variables are measure of self-esteem

(Texas Social Behavior Inventory), measure of locus of control (I-E

Scale), and a measure of role consistency. All analyses were based on

N = 254 due to a listwise deletion of missing data.

The first analysis was conducted to determine whether sex-role

category (knowledge of a subject's sex role category) accounts for a

portion of the variance of the dependent variables. Since past research

indicates that gender is likely to have a confounding effect, gender was

entered into the formula with Female = 1 and Male 0. Dummy variables

were created to represent sex-role category membership. Each category

was treated as a dichotomous variable and a 1 or 0 assigned depending

on membership in the category. Since there are four sex-role categories,

the fourth dummy variable is determined by the first three dummy vari-

ables (k 1) entered into the regression equation. The Undifferentiated

category was designated the reference category and all other categories

were interpreted in reference to the Undifferentiated category. Table 6

presents the output of the multiple regression analyses of the dependent

variables with sex-role category and gender.

With regard to the measure of self-esteem, sex-role categorization

accounts for 24% of the variance. Masculine-typed and Androgynous sub-

jects tended to have the highest self-esteem scores, followed by the

Feminine-typed and Undifferentiated subjects. However, there was a sex

difference in the pattern of relationships, such that within each cate-

gory, individuals of the female gender were slightly higher in self-esteem.









Table 6
Multiple Regression Analyses
With Sex-Role Category and Gender


Texas Social Behavior Inventory


Variables B Beta F(1,249) R2

Gender 4.891 0.140 4.928* .001
Androgynous 20.032 0.489 49.215** .070
Maculine 20.310 0.524 51.950** .163
Feminine 4.417 0.112 2.331 .007
Constant 74.855
(Undifferentiated)
The regression model is significant (F = 19.801, 4,249 d.f., p< .01)
with R2 = .241.


Role Consistency


Gender 0.257 0.088 1.571 .004
Androgynous 0.107 0.310 16.205** .061
Maculine 0.383 0.116 2.137 .006
Feminine 0.212 0.640 0.618 .002
Constant 0.447
(Undifferentiated)
The regression model is significant (F = 4.903, 4,249 d.f., p< .01)
with R2 = .073.


Locus of Control


Gender .939 0.114 2.613 .029
Androgynous -.225 -0.023 0.089 .007
Masculine .311 0.034 0.176 .001
Feminine 1.497 0.162 3.849 .015
Constant 7.322
(Undifferentiated)
The regression model is significant (F = 3.44, 4,249 d.f., p< .01)
with R2 = .053.

*p .05
**p .01









Sex-role categorization accounted for only 7% of the variance in

role consistency scores and 5% of the variance in locus of control

scores. Gender did not account for any significant portion of the

variance of the role consistency scores or the locus of control scores.

Androgynous subjects were significantly higher in role consistency

scores than the Undifferentiated subjects. None of the metric betas (B)

attained significance with respect to locus of control; however,

Feminine-typed females tended to have the most external locus of

control and Androgynous males the most internal locus of control.

Due to the nature of its computation, sex-role categorization

results in a loss of information about the effects of the actual magni-

tude of masculinity and femininity scores. Therefore, multiple regres-

sion analyses were run for all the dependent variables with actual

masculinity and femininity scores. In view of the fact that the

dependent variables are differentially affected by gender, interactions

of gender and masculinity and femininity scores were entered into the

formula. Table 7 presents the output for the multiple regression

analyses of the dependent variables with masculinity score, femininity

score, gender, interaction of gender and masculinity and femininity

score. The hierarchical decomposition method was employed with mascu-

linity and femininity scores and gender entered before the interaction

terms.

In no case were the interaction terms significant; therefore, the

simplified regression model is presented.

With regard to self-esteem, 47.65% of the variance was accounted

for by the analysis based on this regression model. High endorsement

of masculine characteristics accounted for 40.8% of the variance and








Table 7
Multiple Regression Analyses
With Bem Sex-Role Inventory Scores

Texas Social Behavior Inventory


Variable B Beta F(1,250) R2

Masculine 16.838 0.754 220.897** .408
Gender 9.683 0.278 25.481** .068
Feminine 0.566 0.017 0.115 .000
Constant -4.115
(Undifferentiated)
The regression model is significant (F = 75.862, 3,250 d.f., p < .001)
with R2 = .4765. Recalling that the regression model utilizing sex-
role categories accounted for only 24% of the variance, there is
indication that much information is lost in the categorization
procedure.


Role Consistency


Feminine 0.384 0.137 4.150* .027
Masculine 0.383 0.204 9.002** .028
Gender 0.267 0.091 1.519 .006
Constant 0.106
(Undifferentiated)
The re session model is significant (F = 5.420, 3,250 d.f., p < .01)
with R = .0611.


Locus of Control


Masculine -1.028 -0.195 8.246** .051
Feminine 0.492 0.063 0.869 .008
Gender 0.551 0.067 0.826 .003
Constant 10.758
(Undifferentiated)
The regression model is significant (F = 5.450, 3,250 d.f., p < .01)
with R = .0614.









gender accounted for 6.7% of the variance. This analysis confirms the

results presented in Table 6 and provides the additional information

that masculinity accounts for high self-esteem regardless of how the

individual stands on femininity.

The analysis based on this regression model was also significant

for role consistency, but explained only 6.1% of the variance. Indi-

viduals with high masculinity scores and high femininity scores

(Androgynous individuals) tended to have the highest role consistency

scores.

The analysis based on this regression model was significant for

the locus of control measure and accounted for 6.1% of the variance.

Individuals with high masculinity scores tended to have the most

internal locus of control.

The next analysis was conducted to determine whether age group

(membership in a particular age group) accounts for any of the variance

of the dependent variables. Table 8 presents the output for the

multiple regression analyses of dependent variables with gender and

age group. Dummy variables were created for age groups with the

25-49 year olds as the reference group. The following notations will

apply for age groups: Group 1 = 18-24 year olds; Group 2 = 25-49 year

olds; Group 3 = 50-60 year olds; and Group 4 over 60 years of age.

The analysis based on the regression model was not significant for

self-esteem or role consistency measures. With regard to the locus of

control measure, a greater percentage of the variance was accounted for

by age group than by sex-role category. A significantly greater number

of female students tended to endorse an external locus of control.

Although it did not attain significance, individuals over 60 years of

age tended to express the most internal locus of control.









Table 8
Multiple Regression Analyses
With Age Groups

Texas Social Behavior Inventory


Variable B Beta F(1,249) R2
Gender -1.103 -0.317 0.253 .001
Group 1 -3.586 -0.095 1.599 .011
Group 3 2.602 0.059 0.656 .004
Group 4 -0.859 -0.020 0.076 .000
Constant 90.377
(Group 2)
The regression model is not significant (F = 1.031, 4,249 d.f.) with
R2 = .016 indicating that knowledge of age and or sex of an individual
alone does not significantly improve predictability of self-esteem
scores.


Role Consistency


Gender .183 .062 .981 .004
Group 1 .127 .040 .284 .000
Group 3 -.264 -.007 .010 .003
Group 4 .439 .123 2.805 .011
Constant (Group 2) .479
The regression model is not significant (F = 1.125, 4,249 d.f.) with
R2 = .018.


Locus of Control



Gender 1.510 0.184 9.859** .029
Group 1 2.978 0.334 22.946** .114
Group 3 0.551 0.053 0.612 .005
Group 4 -0.609 -0.061 0.794 .003
Constant (Group 2) 6.576
The regression model is highly significant (F = 11.084, 4,249 d.f.,
p < .01) with R2 = .151.









Treating a continuous variable as a categorical variable leads to

loss of information as was evident with the results of sex-role categori-

zation. Therefore, multiple regression analyses were run on the dependent

variables with gender and age plus the interaction of age and gender.

The analysis based on this model was not significant for self-esteem nor

for role consistency, confirming the earlier analyses. Table 9 presents

the results for the locus of control measure.


Table 9
Multiple Regression Analysis
With Age and Gender

Locus of Control


Variables B Beta F(1,250) R2

Gender 1.458 0.177 1.605 .029
Age -0.629 -0.299 12.310** .088
Age/Sex 0.815 0.005 0.001 .000
Constant 10.086
The regression model is highly significant (F = 11.085, 3,250 d.f.,
p < .001) with R2 = .1174.

Finally, multiple regression analyses were run on each of the

dependent variables with sex-role category, gender, and chronological

age to determine the effect of each of the independent variables on the

dependent variables with adjustments made for all other independent

variables. The actual chronological age of the subject was entered

into the formula because the entry of age group category plus sex-role

category seriously violates the assumption of a normal distribution.

Table 10 presents the output from the analyses.

Although the relationship of sex-role category to self-esteem did

not change (i.e., Androgynous and Masculine-typed individuals with

higher self-esteem scores; Feminine-typed and Undifferentiated individuals










Table 10
Multiple Regression Analyses With All
Independent Variables

Texas Social Behavior Inventory


Variables B Beta F(1,248) R2
Feminine 4.529 0.115 2.472 .036
Gender 4.893 0.140 4.977** .004
Age 0.889 0.010 3.253 .004
Androgynous 20.116 0.491 50.064** .087
Maculine 20.782 0.536 54.412** .251
Constant 70.966
(Undifferentiated)
The regression model is significant (F = 16.635, 5,248 d.f., p < .01)
with RZ = .2512.

Role Consistency


Feminine 0.217 0.065 0.646 .002
Gender 0.257 0.088 1.570 .008
Age 0.387 0.052 0.704 .003
Androgynous 0.107 0.311 16.292** .054
Masculine 0.404 0.124 2.349 .009
Constant 0.430
(Undifferentiated)
The re ression model is significant (F = 4.059, 5,248 d.f., p < .01)
with R = .0756.

Locus of Control


Feminine 1.417 0.153 3.787 .042
Gender 0.938 0.114 2.859 .008
Age -0.626 -0.298 25.242** .089
Androgynous -0.284 -0.029 0.156 .001
Masculine -0.211 -0.002 0.001 .000
Constant 10.061
(Undifferentiated)
The regression model is significant (F = 8.098, 5,248 d.f., p < .01)
with R = .1404.








with lower self-esteem scores); the older females tended to have higher

self-esteem scores. With regard to locus of control, this analysis

confirmed the earlier analysis. The 18-24 year old Feminine-typed

females tended to have the most external locus of control and the males

over 60 years of age the most internal locus of control. Age was not

a very significant factor in the analysis of the role consistency

variable but the tendency was for older Androgynous females to have

higher role consistency scores.

Analyses of variance were also performed on each of the dependent

variables as another method for examining the data. In general, the

pattern of differences closely resembled that obtained from the multiple

regression analyses. Table 11 presents the means and standard deviations

for the Texas Social Behavior Inventory scores.


Table 11
Texas Social Behavior Inventory



Males Males
Age
Group N Mean S.D. Sex-Role N Mean S.D.

1 40 89.20 13.33 Androgynous 30 95.23 11.23
2 37 91.84 16.61 Masculine 62 95.69 14.29
3 25 91.32 20.10 Feminine 10 77.50 13.80
4 30 87.60 16.41 Undifferentiated 30 77.07 16.08

Females Females
1 39 83.33 19.61 Androgynous 31 98.55 15.56
2 39 87.23 17.39 Masculine 18 102.50 12.81
3 30 93.73 17.01 Feminine 63 83.14 18.34
4 32 89.47 20.96 Undifferentiated 28 78.18 16.90



There was no significant difference in Texas Social Behavior Inven-

tory means between males (89.98) and females (88.05). There were no








significant differences in means between age groups. There was a

highly significant difference in means among sex-role categories

(F = 50.15, 7,264 d.f., p< .01). The differences were also highly

significant for males only (F = 143.24, 3,128 d.f., p< .01) and for

females only (F = 58.74, 3,136 d.f., p< .01). Regardless of gender,

masculine and androgynous subjects attained significantly higher self-

esteem scores than feminine-typed and undifferentiated subjects with

no significant differences between masculine-typed and androgynous

subjects nor between feminine-typed and undifferentiated subjects.

Table 12 presents the means and standard deviations for the role

consistency scores.


Table 12
Role Consistency


Males


Males


Age
Group N Mean S.D. Sex-role N Mean S.D.

1 40 .472 0.14 Androgynous 30 .542 0.15
2 37 .486 0.13 Masculine 57 .492 0.12
3 20 .491 0.17 Feminine 8 .430 0.16
4 26 .527 0.13 Undifferentiated 28 .450 0.15



Females Females
1 40 .526 0.15 Androgynous 31 .588 0.15
2 38 .490 0.15 Masculine 16 .474 0.14
3 29 .485 0.13 Feminine 61 .497 0.17
4 29 .537 0.16 Undifferentiated 28 .470 0.15









There was no significant difference in mean role consistency scores

between males (.491) and females (.509). There were no significant

differences in means between age groups. The difference in means was

highly significant for sex-role categories (F = 40.07, 7,252 d.f.,

p < .01). Androgynous females were significantly higher in role con-

sistency scores than all the other sex-role category groups of both

genders except for Androgynous males. The Androgynous males were not

significantly different from individuals in the other sex-role categories.

Table 13 presents the means and standard deviations for the locus

of control measure.


Table 13
Locus of Control


Males


Males


Age
Group N Mean S.D. Sex-role N Mean S.D.

1 39 9.4 4.0 Androgynous 30 7.2 3.9
2 38 6.5 3.6 Masculine 62 7.9 3.9
3 25 8.0 3.6 Feminine 10 8.2 4.9
4 29 5.8 3.9 Undifferentiated 29 6.7 4.0


Females Females
1 39 11.1 4.2 Androgynous 30 8.0 4.2
2 39 8.3 3.4 Masculine 18 7.7 3.6
3 30 8.2 3.6 Feminine 64 9.8 3.9
4 32 7.8 4.0 Undifferentiated 28 8.8 4.0


There was a significant difference in mean locus of control scores

between males (7.5) and females (8.9). There were significant differences

in mean locus of control scores across age groups (F = 6.619, 7,263 d.f.,

p< .01). The female students were significantly more external in locus








of control than all the other age groups except the male students. The

males over 60 years of age were significantly more internal in locus of

control than all other groups except the males 25-49 years of age.

There were also significant differences in locus of control means

within the sex-role categories (F= 40.319, 7,264 d.f., p< .01). The

Undifferentiated males (largely composed of males over 60 years of age)

were significantly more internal in locus of control than the Undifferen-

tiated females and both males and females who fell into the Feminine-

typed category.

Table 14 presents the means and standard deviations for the self-

esteem (Texas Social Behavior Inventory) measure after breakdown by age

group and sex-role category.

Modified t tests, which takes into account the difference in variance

and Ns, were done to discern any significant differences in means whenever

the analysis of variance resulted in a significant F test. There were

no significant differences in the student group. The Androgynous males

25-49 years of age were higher in self-esteem than the Undifferentiated

males in this age group. There were no significant differences among the

females of this group. The Androgynous females 50-60 years of age were

higher in self-esteem than the Undifferentiated females of this age group.

There were no significant differences among the males of this age group.

Within the oldest age group, Masculine-typed males were higher in self-

esteem than the Undifferentiated males; and the Masculine-typed females

were higher in self-esteem than the Feminine-typed and Undifferentiated

females.

Table 15 presents the means and standard deviations for the role

consistency measure after breakdown by age group and sex-role category.

The only difference in means to attain significance, after adjusting for










Table 14
Texas Social Behavior Inventory


Students
Males Males
Sex-role N Mean S.D. Sex-role N Mean S.D.
Androgynous 10 90.30 7.42 Androgynous 9 93.11 12.53
Masculine 17 95.24 12.88 Masculine 5 98.80 15.69
Feminine 4 76.50 18.59 Feminine 19 78.05 20.57
Undifferentiated 9 82.22 12.18 Undifferentiated 6 72.50 17.12


25-49 year olds
Androgynous 7 102.43 2.94 Androgynous 8 96.38 19.80
Masculine 23 96.00 8.51 Masculine 6 97.17 7.88
Feminine 0 Feminine 15 84.93 15.02
Undifferentiated 7 67.57 22.31 Undifferentiated 10 77.40 18.16


50-60 year olds
Androgynous 5 92.00 18.76 Androgynous 7 106.14 14.11
Masculine 15 93.40 22.91 Masculine 1 128.00
Feminine 2 77.50 2.12 Feminine 15 88.60 16.55
Undifferentiated 3 89.00 15.39 Undifferentiated 7 87.43 8.50


Over 60 years old
Androgynous 8 97.12 12.04 Androgynous 7 100.43 14.92
Masculine 7 100.71 11.25 Masculine 6 106.67 10.01
Feminine 4 78.50 14.89 Feminine 14 82.28 20.09
Undifferentiated 11 75.64 12.39 Undifferentiated 5 73.60 22.01











Table 15
Role Consistency


Males
Sex-role N Mean S.I
Androgynous 10 .518 .1l
Masculine 17 .456 .11
Feminine 4 .430 .1l
Undifferentiated 9 .467 .14


Students Fe
Fema

Sex-role
'3 Androgynous
.6 Masculine
10 Feminine
8 Undifferentiated


les
N Mean
9 .641
5 .520
20 .480
6 .510


25-49 year olds
Androgynous 7 .582 .118 Androgynous 8 .535 .158
Masculine 23 .496 .106 Masculine 6 .483 .112
Feminine 0 Feminine 14 .491 .160
Undifferentiated 7 .357 .105 Undifferentiated 10 .455 .171


50-60 year olds
Androgynous 5 .482 .205 Androgynous 7 .538 .074
Masculine 11 .517 .151 Masculine 1 .269
Feminine 1 .134 Feminine 14 .479 .151
Undifferentiated 3 .362 .221 Undifferentiated 7 .476 .104


Over 60 years old
Androgynous 8 .575 .133 Androgynous 7 .632 .151
Masculine 6 .533 .127 Masculine 4 .456 .188
Feminine 3 .361 .109 Feminine 13 .547 .136
Undifferentiated 9 .535 .118 Undifferentiated 5 .442 .179


S.D.
.189
.125
.122
.142








the differences in variances and Ns, was within the student female

group. The Androgynous females were significantly higher in role

consistency scores than the Feminine-typed females of this group.

Table 16 presents the means and standard deviations for the locus

of control measure after breakdown by age group and sex-role category.

The only differences in means to attain significance were in the student

female group. The Androgynous females were significantly more internal

in locus of control than the Feminine-typed and Undifferentiated females

of this age group.


The Relationship of Demographic Data to Sex-Role Categorization


The distribution of subjects into educational categories (68% of

the females and 75% of the males were college graduates); and retirement

status (93% of the men over 60 years of age were retired and not working)

did not lend itself to statistical analysis.

Hypothesis 5 predicted that a significantly greater proportion of

Masculine-typed and Androgynous women would fall into the full-time or

part-time employment category. There was a significant difference in

2
sex-role category distribution (X = 8.24, 3 d.f., p< .05) between women

who had never worked (full-time housewife category) and women who have

worked or are now working. A significantly greater number of working

women fell into the Androgynous and Masculine-typed categories.

Hypothesis 6 stated that a greater proportion of the women in the

"un-traditional" marital status categories would be Masculine-typed or

Androgynous. Sixty-seven women were either married or widowed, while

thirty-three women were either single, divorced, or remarried. The dis-

tribution of women in the first group (traditional marital status) was

compared to the distribution in the second group ("un-traditional"











Table 16
Locus of Control


Students
Males Females
Sex-role N Mean S.D. Sex-role N Mean S.D.
Androgynous 10 8.2 4.0 Androgynous 8 8.1 4.6
Masculine 16 10.5 2.9 Masculine 5 8.4 4.7
Feminine 4 11.2 5.2 Feminine 20 12.4 3.7
Undifferentiated 9 8.0 4.8 Undifferentiated 6 13.3 2.1


25-49 year olds
Androgynous 7 5.6 3.6 Androgynous 8 8.2 3.4
Masculine 24 6.5 3.5 Masculine 6 6.3 1.5
Feminine 0 Feminine 15 8.8 3.8
Undifferentiated 7 7.4 3.9 Undifferentiated 10 8.7 2.8


50-60 year olds
Androgynous 5 6.4 2.6 Androgynous 7 8.6 4.3
Masculine 15 8.9 3.8 Masculine 1 11.0
Feminine 2 7.5 3.5 Feminine 15 9.0 3.2
Undifferentiated 3 6.3 4.0 Undifferentiated 7 5.7 3.1


Over 60 years old
Androgynous 8 7.8 4.7 Androgynous 7 6.8 5.3
Masculine 7 4.8 3.9 Masculine 6 8.0 4.2
Feminine 4 5.5 4.2 Feminine 14 8.2 3.4
Undifferentiated 10 5.1 3.1 Undifferentiated 5 8.0 4.6






79

marital status) across the four sex-role categories by means of Chi-

2
square analysis. The result was not significant (X = 6.32). However,

the distribution was skewed in the hypothesized direction. Fifty-one

percent of the women in the "un-traditional" marital status group was

either Androgynous (30%) or Masculine-typed (21%); whereas, only 27%

of the women in the traditional marital status group fell into the

Androgynous (18%) or the Masculine-typed (9%) categories.















CHAPTER 4
DISCUSSION


This study was designed to explore the nature of the influence of

societal age role prescriptions on sex-role identity and self-identity

for males and females. First, with regard to sex-role identity per se,

it was hypothesized that a proportionally greater number of men over

60 years of age would endorse more feminine characteristics and less

masculine characteristics. Not only did men over 60 years of age

attribute significantly lower levels of masculine traits to themselves

than men 25-49 years of age; but the hierarchy of age groups in terms

of masculinity score medians was 25-49 years olds =5.72, 50-60 years

olds= 5.40, 18-24 year olds = 5.28, and over 60 years old = 5.08. This

is not surprising since the items that comprise the masculinity sub-

scale have the underlying commonality of being instrumental in nature.

The years from 25-60 are probably the most productive years for males

in our society; thus, it is reasonable that more individuals in this

age range would endorse a high degree of those agentic or instrumental

characteristics that are most likely to lead to rewards within the

context of our society's values.

There was also a significant difference in sex-role categorization

for males. A significantly greater number of males over 60 years of age

fell into the Undifferentiated category and a fewer number into the

2
Masculine category in comparison to the 25-49 year old group (2 = 8.48)

2
and the 50-60 year old group (X = 12.46). The distribution of the

college age group fell somewhere in between the 25-60 year olds and the









oldest group of men. The difference in distribution did not attain

significance but the tendency was in the hypothesized direction. In

the college sample, the following percentages of individuals fell into

each category: 25% Androgynous, 42.5% Masculine, 10% Feminine, and

22.5% Undifferentiated. The following percentages were found for men

over 60 years of age: 26.7% Androgynous, 23.3% Masculine, 13.3%

Feminine, and 36.7% Undifferentiated. Only part of Hypothesis 1 was

confirmed. A greater proportion of males over 60 years of age endorsed

less "masculine" characteristics but did not endorse more "feminine"

characteristics. These results appear to be congruent with society's

age role expectations for men after retirement; i.e., achievement

demands imposed by society diminishes after retirement and this may

be related to the endorsement of less masculine (achievement oriented)

characteristics by the men over 60 years of age.

Hypothesis 2 was confirmed. There were no significant differences

in the distribution of females into the sex-role categories across

the four age groups, nor were there any differences in the degree of

endorsement of masculine or feminine characteristics. In each age

group, the largest percentage of subjects fell into the feminine

category: 50% of 18-24 year olds, 38% of 25-49 year olds, 50% of 50-60

year olds, and 44% of females over 60 years of age.

With regard to sex-role identity, it is possible that the American

culture not only provides more distinct and easily discriminable cues

for sex-appropriate behaviors for males, but also provides more explicit

age role sanctions for males. During those years when the masculine

traits are important components of their age role prescription, a

greater proportion of males tend to endorse these masculine characteristics.








When their age role prescription diminishes the importance of the

masculine traits, a greater proportion of males may adjust to this

change by endorsing the lesser degree of masculine characteristics.

These results are also in accord with the findings of Rosen and

Neugarten (1960), who reported that the aged population tended to

give up self-assertiveness, and to avoid rather than to embrace

challenge. However, this sample of men over 60 years of age did not

seem to endorse more "womanly" qualities as reported by Neugarten

and Gutmann (1958).

In a sense, the results for the females may also reflect societal

prescriptions. Just as there was a conflict in trends reported by

Neugarten and Gutmann (1958) and Rosen and Neugarten (1960), the

changes, if any, in the age role prescriptions may be less clearcut

for females. This vagueness may be reflected in the lack of differences

among the four age groups.


Statistical Characteristics of the Self-Identity Measures

The Texas Social Behavior Inventory is strongly related to the

measure of masculinity. The authors (Helmreich et al., 1974) state:

For males, the correlation between the Personal
Attributes Questionnaire (PAQ a measure of sex-
role analogous to Bem's Sex-Role Inventory) and
the Texas Social Behavior Inventory is .54 (p< .001)
indicating that the most stereotypically masculine
a male rates himself, the higher his self-esteem.
The correlation is in the same direction but
stronger for females (r = .59, p< .001) suggesting
that the more a woman attributes masculine chara-
teristics to herself, the higher her self-esteem. (p. 4)

The data from this study confirm the correlation between the

masculine subscale and the measure of self-esteem. An individual's

masculinity score accounts for 46.4% out of 47.7% of the variance









explained by the masculinity and femininity scores. Femininity scores

had no input for self-esteem. Further, neither age nor gender contri-

buted greatly to the variance of the Texas Social Behavior Inventory

scores. The authors of the Texas Social Behavior Inventory have

defined what they consider to be the means for attaining high self-

esteem; i.e., social competence. They pose the question, "How

confident do you feel that you can elicit self-esteem from others by

the competent performance of these behaviors?" They define social

competence in terms of masculine characteristics; therefore, the

endorsement of a large number of masculine characteristics necessarily

leads to a high score on the Texas Social Behavior Inventory. The

results are intrinsic to the nature of the measure. It is plausible

that in our culture those behaviors characterized as masculine do

lead to more positive reinforcement and higher self-esteem. However,

the Texas Social Behavior Inventory does not directly measure self-

esteem defined as pride in oneself. In this study, masculinity was

the determining factor for high self-esteem but was not consistently

correlated with high role consistency or a more internal locus of

control. By the same token, Undifferentiated and Feminine-typed sub-

jects with low masculinity scores were consistently lower in self-

esteem but not necessarily lower in role consistency or more external

in locus of control.

Both masculinity and femininity scores contributed to the variance

in Role Consistency scores. The Androgynous individuals, regardless

of age or gender, tended to have the highest Role Consistency scores.

Neither age nor gender was consistently related to Role Consistency

scores; however, the subjects over 60 years of age tended to have higher

Role Consistency scores.








None of the college age students, but at least half of the sample

of individuals over 25 years of age, complained in writing or verbally

about the required forced choices in the Locus of Control measure.

These individuals expressed their frustration at having to decide

between "black and white" alternatives, when they felt the situations

presented often did not warrant such unqualified decisions. Gurin,

Gurin, Lao, and Beattie (1969) factor-analyzed Rotter's scale

(Appendix E) and found separate factors for personal as opposed to

general causality. The first factor "control ideology" includes items

16, 11, 6, 23, 7, 10, 26, 20, and 18; and seems to measure the

respondent's general beliefs about the role of internal and external

forces in determining success or failure in the culture at large.

The second factor "personal control" contains items 13, 9, 28, 25, and

15; and seems to measure the respondent's belief in his or her own

competence to control what happens in his or her own life. The

complaints were without exception aimed at questions regarding "control

ideology." The Locus of Control measure was introduced into this study

in an attempt to obtain some measure of the sense of confidence, or

lack of it, individuals possess that he or she can influence the envi-

ronment by competent performance to elicit desired results. The

utility of this measure is probably obfuscated by the multidimensional

character of the present Locus of Control measure. For the purpose of

this study, more precision in assessment can probably be gained by a

measure that focuses on and enlarges on the personal control dimension

of the scale.









This is the one measure in which chronological age accounted for

most of the variance. The linear relationship was such that the older

individuals were more internal in locus of control. Other variables

contributing to the variance in Locus of Control scores were gender

and femininity scores. Feminine-typed females tended to have the

most external locus of control.


Sex-Role Identity and Its Relationship to Gender,
Chronological Age, and Other Measures of Self-Identity


There seems to be some indication, as discussed earlier, that

cultural definitions of age role (developmental stage) influence an

individual's sex-role identity. The relationship between cultural pre-

scriptions and sex-role patterns appears to be clearer for males in our

society. Conversely, Hypotheses 3 and 4 predicted that the importance

of sex-role identity, itself, within the larger framework of self-

identity would also change with age. Hypothesis 3 predicted that

during late adolescence, when an individual is engaged in establishing

his or her self-identity, sex-role identity would have an organizing and

major impact on how one feels as a person. Hypothesis 4 predicted that

beyond the college age, sex-role identity would not play a major role

in determining an individual's self-concept.

Based on past research (Bem, 1977; Jones et al., 1978; Spence et al.,

1975; Wetter, 1975), Hypothesis 3 predicted that within the college sample,

individuals who are categorized as Undifferentiated or Feminine-typed

would have relatively low self-esteem scores, low role consistency

scores,and a more external locus of control. By contrast, individuals

who were categorized as Androgynous or Masculine-typed were predicted to

have higher self-esteem scores, higher role consistency scores, and a









more internal locus of control. The Feminine-typed and Undifferentiated

subjects of both genders were significantly lower in self-esteem. The

Feminine-typed females were significantly lower in role consistency

scores than the Androgynous females. The Feminine-typed and Undif-

ferentiated females were also significantly more external in locus of

control than the Androgynous females.

Hypothesis 4 proposed no significant differences in self-esteem

scores, role consistency scores, or locus of control scores in relation

to sex-role categorization, among the other three age groups. As was

discussed earlier, self-esteem scores are related to sex-role categori-

zation. The Masculine-typed and Androgynous subjects, regardless of

age group, had higher self-esteem scores; while the Feminine-typed and

Undifferentiated subjects had lower self-esteem scores. However, there

were no consistent differences in role consistency scores or locus of

control scores across the three older age groups associated with sex-

role categories.

Hypotheses 3 and 4 were partially confirmed. Self-esteem was almost

entirely determined by the masculinity subscale score regardless of

gender or age. The college age Feminine-typed subjects were signifi-

cantly lower in role consistency and the most external in locus of

control. These results may be a reflection of the conflict inherent

in attempting to maintain a feminine self-identity in a society that

rewards masculinity during a developmental phase when sex-role identity

is a major issue. On the other hand, within the other three age groups,

there were no consistent or significant differences in Role Consistency

scores or Locus of Control scores associated with maintaining a femi-

nine identity.








Additional support for the hypothesis that the significance of

sex-role identity in relation to self-identity changes with develop-

mental stages was found with reference to the oldest age group. A

large number of men over 60 years of age fell into the Undifferentiated

category. Past research has often supported the notion that the

Undifferentiated individual is at a psychological disadvantage because

he or she has a smaller repertoire of behaviors to adjust to environ-

mental demands. Kelly and Worrell (1977) reviewing the research on

the new formulation of sex roles state: "Androgynous persons were

highest in social poise and intellectuality, whereas undifferentiated

subjects were the most socially awkward and nonintellectual. the

two sex-typed categories fell in between but did not differ from one

another" (p. 1109). Jones et al. (1978) found that, "A pattern emerged

in which masculine males can be described as more competent and confi-

dent on numerous dimensions, whereas less traditionally sex-typed

males are generally more limited and restricted, less effective and

more vulnerable to influence, less sure of themselves, and perhaps

even less well adjusted" (p. 310). The Undifferentiated men over 60

years of age in this study not only had the most internal locus of

control but their role consistency scores did not differ from the

Androgynous and Masculine-typed men in this age group. It is speculated

that for these Undifferentiated men over 60 years of age, cultural age

role expectations tend to de-emphasize sex-role and allow these men

to diminish those qualities associated with masculinity without impinging

on their sense of identity. These men did not seem to face the diffi-

culties of the college age Undifferentiated subjects nor did they seem

to be at a psychological adjustment disadvantage.








This study conceives of sex-role identity as a part of the

larger framework of self-identity. Under this rubric, the Androgynous

individual is seen as one who has a sense of self strong enough to

counter or ignore cultural prescriptions associated with sex-

appropriate behaviors. The data lend support to this concept. There

appears to be a psychological advantage associated with Androgyny,

across all age groups. The Androgynous individuals were not only

high in self-esteem but also had high role consistency scores and

were internal in locus of control. Perhaps a more cogent and intriguing

piece of information in support of the notion that sex-role identity

interacts with age role is that the Feminine-typed and Undifferentiated

individuals were not at a psychological disadvantage, as is usually

true for the college age individual, when the developmental stage is

taken into account. Feminine-typed females were not necessarily at a

psychological disadvantage within the three older age categories; and

the Undifferentiated males over 60 years of age had the most internal

locus of control and relatively high role consistency scores.


Relationship of Sex-Role Identity to Demographic Data

Hypotheses 5 and 6 were concerned with demographic characteristics

that have been found to differentiate women along the dimensions of

sex-role categories. In general, it was hypothesized that those

women who have a self-identity secure enough to violate sex-role

expectations would be more likely to behave in a manner counter to

other cultural expectations. Hypothesis 5 predicted that a significantly

greater proportion of Masculine-typed and Androgynous women would be

employed full-time or part-time. Hypothesis 5 was confirmed. A








2
significantly greater number (X = 8.24) of working women fell into

the Androgynous and Masculine-typed categories in comparison to women

who fell into the non-working housewife category.

Hypothesis 6 predicted that a greater proportion of the women

in the "un-traditional" marital status category would be Masculine-

typed or Androgynous, in comparison to women in the more traditional

marital status category. The difference in distribution was not

significant but was skewed in the hypothesized direction with 51% of

the women in the "un-traditional" marital status category falling into

the Masculine-typed and Androgynous categories and only 30% of the

traditional marital status group falling into these same sex-role

categories. These results provide evidence that women who endorse a

high degree of "masculine" characteristics also demonstrate these

qualities in real-life behavior.

It is fairly clear that internalized images of what it is to be

a proper woman or man acquires the capacity to exert considerable

psychological pressure on our behaviors and our self-concept. There

is also an indication of some interrelationship between age role

prescriptions sanctioned by society and an individual's global self-

perception of masculinity and femininity. While a longitudinal study

is the method of choice for such research, the present findings are

provocative and suggest that future research might explore more fully

the developmental pattern of sex-role identity in adulthood and into

senescence.















CHAPTER 5
SUMMARY AND CONCLUSIONS


This study was undertaken to explore the concept of sex-role

identity and its interrelationship with gender and chronological age.

Sex-role identity was viewed within the new conceptualization of the

psychological dimensions of masculinity and femininity that allows for

the possibility that a person may develop both masculine and feminine

attributes. Thus, the measure of sex-role adopted for the study con-

ceives of masculine and feminine attributes as separate and essentially

orthogonal. The selection of older subjects was based on the desire to

approximate the much studied college population. Thus, the older sub-

jects are intellectually elite and more representative of the general

population.

The college female population is of particular interest in terms

of sex-role identity because data from past research consistently indi-

cate that Feminine-typed college females are at a psychological adjust-

ment disadvantage. In college, males and females alike are preoccupied

with questions about their identity, about their relations with others,

their worth, and about the direction their life is taking. A male

approaches late adolescence fairly secure in his masculinity because,

in our culture, the masculine role is more highly valued and more clearly

defined and there is not only a continuity of role expectations over his

lifetime but also a consistency of role models in the entire culture.










On the other hand, adolescence is probably the first critical period

for sex-role identity for females. Prior to adolescence, sex-identity

for females is not as clearly defined or societal sanctions as rigidly

enforced. Until late adolescence, appropriateness of behavior for

her is more often determined by age than by sex, and she is permitted

to participate in what will later be perceived as masculine activities.

During adolescence, the female is confronted with the prospect that

achievement success (the culturally masculine and more valued qualities)

may be seen as threatening affiliative success (internalized feminine

values by which she measures her success as a woman).

This study conceives the essence of this conflict not at the

behavioral level but as internal and psychological and as an important

part of the process of establishing a self-identity. To this end,

Block's conceptual framework was presented which integrates changes in

sex-role definition with changes in developmental tasks. The psychological

adjustment disadvantage accruing to the Feminine-typed college age

individual was regarded as a reflection of the conflict inherent in

maintaining a feminine-identity in a society that prefers and rewards

masculine characteristics at a time when her self-identity is in question.

It was hypothesized that once a female comes to grips with her sex-role

identity and establishes a comfortable identity as a person, whatever

the outcome, sex-role identity per se will cease to be such a salient

force within self-identity. As hypothesized, the college age Feminine-

typed individuals produced a pattern of scores on the measure of self-

identity that indicated poor psychological adjustment, relative to their

age mates. On the other hand, within the other three age groups there










were no consistent or significant correlations with the measures of

self-identity associated with maintaining a feminine identity.

Additional support for the hypothesis that the significance of

sex-role identity in relation to self-identity changes with develop-

mental stages, was found with reference to the oldest age group. A

large number of men over 60 years of age fell into the Undifferentiated

category. Past research has often supported the notion that the

Undifferentiated college age individual is at a psychological disad-

vantage because he or she has a smaller repertoire of behaviors to

adjust to environmental demands. The self-identity scores of the

Undifferentiated men over 60 years of age in this study were indicative

of psychologically secure individuals. It is speculated that because

our culture tends to de-emphasize sex-role identity for the aged indivi-

dual; men over 60 years of age can diminish their masculine qualities

without feeling "unmanly" or experiencing psychological disadvantages.

This study presents data which indicate that internalized images

of what it is to be a man or woman changes in response to changes in

developmental tasks. The cultural expectations and demands differ at

each developmental stage, with differential implications for males and

females. The present findings suggest a longitudinal study would be

the method of choice to clarify the interaction of sex-role identity

with gender and age role beyond late adolescence.















APPENDIX A


UNIVERSITY OF FLORIDA

DEPARTMENT OF PSYCHOLOGY CONSENT FORM


Subject's Name

Subject's Address

Project Number 907 Project Title Interaction of Age, Gender,

and Personality Variables

Principal Investigator Tetsuko Suzuki Date

I agree to participate in the research as explained to me below:

The development of personality in the early years of life
has been been extensively examined; but there has been
little developmental study of adulthood. There has been
widespread acceptance of the view that personality is
stabilized by the time early adulthood is reached. However,
there are some indications that personality may continue
to change over time. We are assuming that people continually
revise,reverse and add new connotations to their self-concept.
This study is designed to explore the changes, if any, that
take place in how a healthy, active, functioning individual
views his or her self between the ages of 18-70 years. You
will be asked to answer four questionnaires, which are self-
descriptive. You will be assigned a number and no individual
will be identifiable in the final analysis. If you are
interested, and so indicate, we will be glad to send you
our results when the study is completed.

The above stated nature and purpose of this research, including dis-

comforts and risks involved (if any) have been explained to me

verbally by Furthermore, it is agreed

that the information gained from this investigation may be used for

educational purposes which may include publication. I understand that

I may withdraw my consent at any time without prejudice.


Signed