Group Title: influences of family background variables versus subject variables on the sex roles, sex role stereotypes and sex role orientations of college students
Title: The influences of family background variables versus subject variables on the sex roles, sex role stereotypes and sex role orientations of college students
Full Citation
Permanent Link:
 Material Information
Title: The influences of family background variables versus subject variables on the sex roles, sex role stereotypes and sex role orientations of college students
Physical Description: Book
Language: English
Creator: Huchendorf, Malia Ann
Copyright Date: 1990
 Record Information
Bibliographic ID: UF00102742
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltuf - AHZ0697
oclc - 24917227

Full Text








The writer wishes to thank the members of her committee

for the guidance and assistance that was rendered during the

writing of this dissertation. Particular thanks go to Dr.

Patricia H. Miller, the chair of the dissertation committee,

whose insightful comments and insight into the area were

invaluable in finishing this project. Dr. W. Keith Berg,

Dr. Fonda Eyler, Dr. Monica Biernat, and Dr. William

Marsiglio truly enriched this project. They should remain

blameless for any deficiencies that may remain.

Acknowledgements go to the faculty in the Psychology

Department of Illinois State University who became

colleagues in the true sense of the word. Specific

acknowledgements go to Dr. Laura E. Berk and Dr. James J.

Johnson for their guidance, instruction and colleagueship.

Acknowledgements also go to Dr. Steven C. Huchendorf who

served as statistical consultant and typist.

Most importantly, the writer wishes to acknowledge the

patience and perseverance of her family during the writing

process. Thanks go to Lynnay and K. C. who did not

understand why Mommy always had to work. Finally, thanks to

Steve who did understand.



ACKNOWLEDGEMENTS.............. ............... ........... ii

LIST OF TABLES.......................................... vi

ABSTRACT................................................. vii


1 INTRODUCTION................................... .... 1

Introduction to Sex Role Research.................. 1
Development of Sex Roles............................ 4
Family Background Variables......................... 5
Sex Role Stereotypes and Sex Role Orientation
Ratings........................................... 6
Plan of Study........................................ 7

2 LITERATURE REVIEW .................................. 10

Sex Role Research with Adults...................... 10
Definitions ................................... 11
The Measurement of Sex Roles.................... 14
Research Findings on Sex Roles................... 19
Methodological Problems in the Sex Role
Literature.................................... 24
Summary......................................... 32
The Development of Sex Typing....................... 32
Development of Sex Role Stereotypes............. 33
Development of Sex Typed Behavior............... 36
Summary......................................... 38
Developmental Theories............................. 39
Psychoanalytic Theory............................ 40
Social Learning Theory........................... 42
Cognitive-Developmental Theory.................. 49
Gender Schema Theory............................ 52
Social-Role Theory............ .................. 56
Summary......................................... 62
Family Background.................................. 63
Education....................................... 67
Maternal Employment............................. 75
Divorce........................................... 79
Religion....................................... 85


Summary. ...................................... 92
Sex Role Stereotype and Sex Role Orientation
Ratings....................................... 93
Discrepancies in Sex Role Stereotype Ratings.... 94
Gender Differences in Sex Role Orientation
Ratings....................................... 97
Summary......................................... 100
Statement of Hypothees.e............................ 101
Personal Sex Roles.............................. 103
Sex Role Stereotypes and Sex Role Orientations.. 104
Education....................................... 104
Maternal Employment............................. 106
Divorce.......... ................................... 107
Religion............................................ 107
Discrepancies in Sex Role Stereotype Ratings.... 108
Gender Differences in Sex Role Orientation
Ratings....................................... 109

3 METHODOLOGY........................................ 110

Subjects........................................... 110
Education. ...................................... 110
Maternal Employment............................. 111
Divorce.......................................... 111 i
Religion......................................... 112
Materials............................................... 112
Procedure.......................................... 114

4 RESULTS............................................ 116

Personal Sex Roles.................................. 119
Sex Role Stereotypes and Sex Role Orientations..... 126
Education. ........................................ 127
Maternal Employment................................ 130
Divorce .......................................... 131
Religion................................................ 132
Discrepancies in Sex Role Stereotype Ratings....... 134
Gender Differences in Sex Role Orientation Ratings. 135
Multiple Regression Analyses....................... 136
Family Background Versus Subject Variables...... 136
Personal SROS Scores: Subject Variables........ 143
Personal SROS Scores: Family Background
Variables.................................... 145
Personal SROS Scores: Relative Importance
of Variables................................. 145
Other Dependent Variables....................... 146

5 DISCUSSION......................................... 148

Discussion of Hypotheeses........................... 148
Personal Sex Roles, Sex Role Stereotypes and
Orientations................................ 148
Family Background............................... 152

Sex Role Stereotype and Sex Role Orientation
Ratings ........ .............................. 157
Theoretical Integration.............................. 162
Problems and Future Directions...................... 166


REFERENCES.... ........... ............................... 175

BIOGRAPHICAL SKETCH.. .................................. 186


Table page

1. Order Effects Analysis............................. 118

2. Sex Role Orientation Scale: Item Analysis.......... 120

3. Sex Role Orientation Scores By Education Variables. 128

4. Sex Role Orientation Scores By Religion Variables.. 133

5. Personal SROS Regression Results:
Full Model.................................... 137

6. Personal SROS Regression Results:
Family Background Variables Only............... 141

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



Malia Ann Huchendorf

December, 1990

Chair: Dr. Patricia Miller
Major Department: Psychology

The purpose of this investigation is to explore three

areas of sex role research. The first goal is to document

whether personal sex roles, sex role stereotypes and sex

role orientations still exist. The second goal is to

explore the effects of family background variables on the

development of personal sex roles. The third goal is to

explore discrepancies in sex role stereotype ratings and

gender differences in sex role orientation ratings.

The subjects consisted of 191 undergraduate college

students. Each subject was first administered a background

questionnaire followed by five administrations of the Sex

Role Orientation Scale (SROS). Each subject first answered

the SROS based on his or her own beliefs. The next four

administrations were counterbalanced and requested responses


for male and female sex role stereotypes, and male and

female sex role orientations.

The results of the investigation demonstrated the

continued existence of personal sex roles, sex role

stereotypes, and sex role orientations. Each sex role area

demonstrated the expected sex difference, with male ratings

indicating more traditional beliefs than those of females.

The results also indicated that present social roles

outweigh the impact of previous family background variables.

Present educational level, anticipated final educational

level, current level of religiosity and gender were all

influential variables. Parental educational level, maternal

employment, divorce and religious affiliation all showed

insignificant relationships to sex roles.

Sex role stereotypes were not found to reflect the

self-ratings of either males or females. Male stereotypes

were rated more traditionally than male self-ratings

indicated, while females were also rated more traditionally,

but to a lesser degree. Sex role orientations were not

rated differently by males and females. Both sexes rated

male sex role orientations more traditionally than female

sex role orientations.

The results were interpreted within the social-role

theoretical framework. Family background variables showed

little relationship to current sex roles, while present

social situations were related to current sex roles. Thus,

support for social-role theory was deemed appropriate.



Introduction to Sex Role Research

Sex roles are the unwritten yet well-defined rules that

are deemed appropriate by society to guide male/female

behavior. How these appropriate standards of behavior are

acquired has been under systematic investigation by

developmental psychologists and other scientists since the

early 1960s. The pervasiveness of sex roles was described

by Mussen (1969), who stated that "no other social role

directs more of an individual's overt behavior, emotional

reaction, cognitive functioning, covert attitudes, and

general psychological and social adjustment" (p. 707). The

development of sex roles, then, plays an important part in

the overall development of the individual and is deemed

worthy of investigation.

The importance of sex roles and their accompanying

stereotypes should not be underestimated. A recent Supreme

Court ruling (Price Waterhouse v. Hopkins, 1989) stated that

the practice of sex role stereotyping may be used as

evidence in sex discrimination suits. The case involved a

female who claimed that she was denied a partnership in an

accounting firm because she did not fit her sex role


stereotype. The Supreme Court ruled in her favor. Sex role

stereotyping, then, may be used in court as evidence of

discrimination. Hence the documentation of the continuing

existence of sex roles and their accompanying stereotypes is


The volume of information on sex roles has grown

steadily since the onset of systematic empirical

investigation. According to D. N. Ruble and T. L. Ruble

(1982), the 1965 Psychological Abstracts reported 50 studies

dealing with sex role stereotypes. By the 1975 issue,

Psychological Abstracts reported over 500 articles.

Computer searches employing the "PsycLit" program listed

over 1,000 references in English under the sex role

attitudes search for the time period of January 1983 through

June 1989. Furthermore, the most recent Handbook of Child

Psychology chapter on sex typing by Aletha Huston (1983)

listed over 500 references. In 1975 the premier issue of

the journal Sex Roles appeared. Now the area of sex roles

has an entire publication devoted to the empirical

investigations of sex typing and its ramifications. Most

recently, the cover of the May 28th, 1990, issue of Newsweek

and the lead story were devoted to explaining sex

differences between boys and girls. Thus, there is a

proliferation of information available to the scientific

community as well as to the lay public on the topic of sex


A rather startling conclusion of this expanse of

scientific research is that sex role stereotypes are still

very dominant and that they have changed so very little over

the last 20 plus years of study. As examples of this lack

of change, Lewin and Tragos (1987) found sex role

stereotyping to be as significant in 1982 as it was in 1956

among adolescent males and females. The work of T. L. Ruble

(1983) also concluded that sex role stereotypes are

extensive. T. L. Ruble employed the 54-item Personal

Attributes Questionnaire (PAQ). The PAQ was developed by

Spence, Helmreich and Stapp (1974) and has been employed in

numerous empirical investigations. T. L. Ruble found

statistically significant differences on 53 of the 54

measures when subjects rated the "typical" male and female.

The only variable not distinguishing males from females was


In general, stereotypical males are rated as

instrumental, meaning they are competent, self-assured,

independent and rational. Stereotypical females are labeled

as expressive, which includes such traits as emotional,

caring, gentle and devoted to others above self. These are

the same traits that were ascribed to males and females when

the trait approach to stereotyping was begun by Rosenkrantz,

Vogel, Bee, Broverman, and Broverman in 1968. The

persistence of sex role stereotypes despite the radical

political and social activism of the 1970s and early 1980s

is proof of how deeply society has ingrained these


Development of Sex Roles

Due to the vast influence that sex roles impart, it is

not surprising that the indoctrination into the appropriate

sex role frequently begins only moments after birth when the

parents are informed first thing that they are the proud

parents of either a boy or a girl (Shaffer, 1985). The

child is often then fitted with an "appropriately" pink or

blue wrist bracelet with a name identification and then

wrapped in either a pink or blue blanket. Since

socialization into the appropriate sex role begins so early,

it is not surprising that stereotypical sex role behavior is

exhibited by children as young as two years of age (Lewis &

Brooks-Gunn, 1979).

Personal sex roles do not develop in a continuous,

monotonic fashion, but rather fluctuate across the life

span. In childhood they are viewed as inflexible, immutable

rules. By middle childhood increased flexibility as well as

increased pervasiveness of sex roles is evident (Marantz &

Mansfield, 1977). During adolescence it appears that the

reformulation of the self-concept once again puts the

individual's sex role in a state of flux (Streitmatter,

1985). Once adulthood has been attained, sex roles appear

to reflect the social roles that the individual is currently

occupying, for example, career woman versus housewife

(Eagly, 1987; England, 1988). The overall findings on sex

roles and aging, however, are not consistent. Some studies

report that older adults are more traditional (e.g.,

Dambrot, Papp, & Whitmore, 1984; Swatos & McCauley, 1984),

while other studies indicate they are less traditional

(e.g., Urberg, 1979). Thus personal sex roles develop early

in life and are a life long but inconsistent influence on

social behavior.

Family Background Variables

In addition to the social roles that an individual

occupies (especially those in adulthood), the social roles

and environmental influences surrounding the individual also

appear to impact sex role development. While peers,

teachers and school environments have an impact, the

majority of the work done on environmental influences has

examined the impact of family members on the developing

child due to the pervasiveness of family influence as

compared to other social institutions. Major areas of

research have included the influence of parental and

personal educational attainments, the effects of maternal

employment, the impact of divorce, and the differential

effects of religion. Numerous studies have been completed

in each of these areas. Most studies found effects on sex

role development such that traditional family styles, lower

levels of education, and increased religiosity were

correlated with more traditional sex roles. Thus, the

family environment has been found to influence the

development of sex typing.

Sex Role Stereotype and Sex Role Orientation Ratings

Sex roles are influential in our perceptions of others

in that we all engage in sex role stereotyping of others as

well as ourselves. In fact, Eagly (1987) has suggested that

stereotypes in general are the basis of person perception.

Fabes and Laner (1986) agree with Eagly and go one step

further by stating that sex role stereotypes guide our

individual interactions with others. Hence an understanding

of how members of the opposite sex are perceived would be of

value in interpersonal experiences. If men and women hold

discrepant sex role stereotypes of each other, then

interpersonal interactions will be ripe for misunderstanding

and misinterpretation. Therefore, it is important to

evaluate the commonly held sex role stereotypes of each


Beyond sex role stereotypes, which are perceptions of

"typical" individuals, is the area of sex role orientations.

Sex role orientations reflect opinions about "ideal"

individuals (Brogan & Kutner, 1976; T. L. Ruble, 1983). As

is the case with stereotypes, ideal opinions are also of

importance in dealing with members of the opposite sex as

they once again influence our actions and behaviors.

Individuals respond in certain situations based on what they

believe the appropriate response to be. For example, in

studies involving interactions between members of both

sexes, the level of aggression, compassion or submission to

leadership is frequently a result of the sex of the partner

or opponent (Spence & Helmreich, 1980). The problem with

sex role orientations is that an overview of studies in this

area indicates no consensus regarding what traits and

behaviors are considered ideal. Studies also suggest

(Deutsch & Gilbert, 1976; McPherson & Spetrino, 1983; Scher,

1984) that orientation ratings vary based upon the sex of

the rater. Hence, continued investigation into the area of

sex role orientations is warranted.

Plan of Study

The purpose of this investigation is threefold. The

first goal is to document the continued existence of

personal sex roles, sex role stereotypes and sex role

orientations. The second goal of this investigation is to

explore the effects of family background variables on the

development of personal sex roles. The last goal is to

explore the discrepancy in sex role stereotype ratings and

to explore gender differences in sex role orientation


Despite recent political and social changes, the

scientific literature indicates that sex roles, sex role

stereotypes and sex role orientations continue to exist, and

show little change over the past 20 years of investigation.

The vast majority of the studies that document this finding

has employed sex role measures that assess psychological and

personality traits (e.g., the PAQ). This study will employ

a sex role measure that assesses attitudes toward specific

behaviors, the Sex Role Orientation Scale (SROS, Brogan &

Kutner, 1976). Instead of simply continuing the

documentation of the existence of sex roles, sex role

stereotypes and orientations, this study will expand our

knowledge by moving closer to the behavioral domain.

As previously stated, the second goal of this

investigation is to explore the effects of differing family

background variables on the development of personal sex

roles. The effects of parental and personal education,

maternal employment, parental marital status and religious

affiliation and practice will all be investigated. Again,

the employment of a sex role measure that assesses attitudes

toward behavior will expand the current literature.

Additionally, it is possible that previously documented

familial effects will not be replicated due to changes that

have occurred over the last 20 years (e.g., increased

educational levels, increased maternal employment, and

increased divorce rates). Thus, continued investigation of

the effects of family background variables will be


The third area of investigation deals with sex role

stereotypes and sex role orientations. If there are

discrepancies between personal sex roles and either

stereotypes or orientations, then miscommunications in

interpersonal interactions are possible. Additionally,

ratings of sex role orientations have been found to vary

depending upon the sex of the rater. It is likely that

measurement of behavioral attitudes rather than traits will

be of more benefit when dealing with these areas due to the

specificity of behavioral attitudes versus the abstractness

of traits.


This investigation explores personal sex roles, sex

role stereotypes, and sex role orientations. The

development of sex roles has been well documented. Much of

this attention has focused on family background as a cause

of differential sex role development. Sex role stereotypes

have been the most thoroughly researched, while sex role

orientations have been only recently addressed. This

section will begin with a look at the methodology of sex

role research. The development of sex roles and sex role

theoretical formulations follows. The impact of family

background on the development of sex roles will then be

examined, followed by an examination of discrepancies in sex

role stereotype ratings and gender differences in sex role

orientation ratings. The section will close with a

statement of the hypotheses under investigation.

Sex Role Research with Adults

As noted, the research on sex roles is extensive. This

section will attempt to summarize many of the ideas,

empirical results and difficulties associated with research

in this field. The first area to be addressed will be the


definition of terms. The next area to be discussed will be

the differing measurement scales in use, followed by the

overall empirical findings of research in the area.

Finally, the methodological problems associated with the

area will be discussed.


Before investigating the area of sex roles, it is

necessary to define and clarify several of the terms now in

use. Due to the connotations associated with the use of the

word "sex," some investigators have determined that the use

of the word "gender" is more appropriate (e.g., Stevens,

Barton & Gardner, 1983). However, the preponderance of

literature in the area employs the term "sex." These terms

will be used synonymously in this work due to their

interchangeable usage in the literature.

There are several different terms utilized in the sex

role literature, each with a specific meaning. To begin

with, sex typing is the overall "process by which children

acquire the values, motives, and behaviors viewed as

appropriate to either males or females in a specific

culture" (Hetherington & Parke, 1986, p. 623). It is a

developmental process that begins shortly after birth and

continues throughout life as social roles change across the

life span.

The terms sex role identity and sex role are

interdependent. Sex role identity "has a private

connotation. It refers to the individual's perception of

the self as relatively masculine or feminine in

characteristics, capabilities and behaviors" (Berk, 1989, p.

551). Sex role identity is a personal perception of oneself

compared to societal standards of masculinity and

femininity. Thus an individual's sex role identity is a

hypothetical psychological construct that extends beyond the

"simple cognitive awareness of his or her biological sex to

psychologically identifying with it" (Pleck, 1981, p. 12).

Sex roles are the public face of one's sex role identity.

Sex roles are "everything that a person says and does, to

indicate to others or to the self the degree that one is

either male, or female" (Money & Ehrhardt, 1972, p. 4). So

then, sex role identity "is the private experience of sex-

role, and sex-role is the public expression of sex-role

identity" (Money & Ehrhardt, 1972, p. 4).

Sex role stereotype is the most commonly researched

term in the literature. Yet Ashmore and Del Boca (1979)

found that a conceptual definition for sex role stereotype

was lacking in most research studies. The definition put

forth by these authors includes four characteristics. First

of all, a sex role stereotype is regarded as a cognitive

construct. It is a perception, an expectation or a

judgment. Second, a sex role stereotype is a collection of

beliefs or judgments, a set of cognitions that are

interrelated, and so grouped together. Third, sex role

stereotypes are beliefs about the psychological make-up, the

personalities or the traits that men and women possess.

Finally, in order to be a stereotype, it must be a set of

beliefs that are shared by members of some group. Putting

these four characteristics together, Ashmore and Del Boca's

definition of sex role stereotype would be "the structured

set of beliefs about the personal attributes of women and of

men" (p. 222).

A newer term being utilized in the research is that of

sex role orientation. The major proponents of the usage of

this term are Brogan and Kutner (1976). They define sex

role orientation as "attitudes about what is 'right' for

males and females to do" (p. 32). Sex role orientation has

a moral aspect, how things "should be," versus the sex role

stereotype which is merely a reflection of how things are.

Thus, when requesting subjects to rate a male or female

based on how typical they are, the request is for a

stereotypical reply. If the request is to rate the subject

based on how males or females should behave, the request is

for a sex role orientation response. T. L. Ruble (1983)

also makes this distinction; however, instead of employing

the term orientation, he refers to "should behavior" or

ideal/desirable behavior as sex role attitudes. Pleck

(1981) also concurs with this usage; however, he too employs

a different term. Pleck refers to ideal behavior as sex

role norms. Thus, there are several different terms in

common usage in the field at present. This study will

employ the term sex role orientation to reflect

ideal/desirable sex role attitudes due to the usage of

Brogan and Kutner's Sex Role Orientation Scale as the

study's measurement tool.

Original work on sex typing assumed that individuals

over the course of development became either masculine or

feminine in their sex role adoption. Sex roles were viewed

as containing bipolar traits. Sandra Bem (1974) took

exception to this conceptual viewpoint and hypothesized

androgynous sex typing. The term androgyny is Greek, with

"andro" meaning male, and "gyne" meaning female. The

androgynous individual, then, has some masculine and some

feminine traits, making the individual situationally

flexible. Sex role traits then are no longer considered to

be bipolar; rather, they are considered independent traits

that can be clustered to form the individual's sex role

identity as each person desires. Androgynous individuals

can be both instrumental and expressive; the categories are

not mutually exclusive.

The Measurement of Sex Roles

Just as there are numerous terms and a plethora of

studies, there are also several measurement tools available


to evaluate the extent of sex role stereotypes and sex role

orientations. An overview of the literature, however,

indicates that only five of these measures are used with any

regularity. An examination of these five instruments will

be undertaken in order to familiarize the reader with the

design and purpose of each. Examination of the three

measures that test sex role traits will be undertaken first.

Examination of the two measures that test for sex role

behaviors will follow. Limitations and problems associated

with sex role measurement tools will be discussed under the

section on methodological problems.

The two sex role measures most commonly used are

actually the same instrument, one merely a revision of the

other. Systematic investigation of sex roles began with the

advent of the Sex Role Stereotype Questionnaire (SRSQ)

devised by Rosenkrantz et al. (1968). The SRSQ was devised

to assess the psychological traits and sex role behaviors of

males and females. The questionnaire contained 122 trait

descriptions listed in bipolar form. Raters were then asked

"to indicate the extent to which each item characterized an

adult man (masculinity response), and an adult woman

(femininity response)" (Broverman et al., 1972, p. 62).

Initial and subsequent usage of the SRSQ has shown that male

and females do differ substantially in their assigned sex

role stereotypes, and furthermore, agreement between males

and females concerning the assignment of these traits is


extremely good (correlations of r = .95 for feminine traits,

and r = .96 for masculine traits as reported by Rosenkrantz

et al.).

The SRSQ was revised by Spence et al. (1974). The

scale was narrowed from 122 items down to 55 and renamed the

Personal Attributes Questionnaire (PAQ). The 55-item

listing still required the subjects to rate male and female

traits on a bipolar scale. Spence and her colleagues then

broke down the PAQ into three subscales: "Male-valued (mean

ratings for the ideal man and woman both fell towards the

stereotypic masculine pole), Female-valued (both towards the

stereotypic feminine pole), and Sex-specific (ideal man and

woman fell toward different poles)" (p. 44). The authors

concluded that masculinity and femininity were separate

dimensions. The femininity scale reflected an expressive

character, high in warmth, caring and emotionality. The

masculinity scale reflected an instrumental character, high

in competence, independence and aggressiveness.

In opposition to the notion that masculinity and

femininity are bipolar opposites, Bem (1974) created a

measurement device based on the hypothesis that a person

need not be either masculine or feminine, but could display

traits of both, which is labeled androgynous. Bem's

conception of sex roles, then, reflected masculinity and

femininity as independent, not mutually exclusive. It is

possible to be both instrumental and expressive. In fact,


Bem suggested that the androgynous person would demonstrate

greater situational flexibility and so be more likely to

display superior mental health.

The Bem Sex Role Inventory (BSRI, Bem, 1974) is made up

of 60 items, 20 masculine, 20 feminine and 20 neutral.

Subjects are asked to rate themselves on each item using a

7-point Likert-type scale. Total scores are tallied for

three scales, the masculinity scale, the femininity scale,

and the androgyny scale.

The Masculinity and Femininity scores indicate the
extent to which a person endorses masculine and
feminine personality characteristics as self-
descriptive. The Androgyny score reflects the relative
amounts of masculinity and femininity that the person
includes in his or her self-descriptions. (Bem, 1974,
p. 158)

Therefore, an androgynous person endorses both masculine and

feminine traits to equal degrees in their self-descriptions.

This scale is widely used in the literature.

Another earlier creation by Spence and Helmreich (1972;

Spence, Helmreich & Stapp, 1973) was the Attitudes toward

Women Scale (AWS). The purpose of this scale was to assess

the "appropriate" rights and assigned roles of women today.

The original scale contained 55 items and was rated (from 0

to 3) in terms of amount of agreement. High scores reflect

a profeminist attitude. The AWS can be separated into six

subscales based on attitude content. The subscales reflect

attitudes toward "1) vocational, educational and

intellectual roles; 2) freedom and independence; 3) dating,

courtship, and etiquette; 4) drinking, swearing, and dirty

jokes; 5) sexual behavior; and, 6) marital relationships and

obligations" (Borges, Levine & Naylor, 1982, p. 408). This

scale has been revised into a shorter form (Spence et al.,

1973), and both scales continue to be utilized in sex role


The most recent scale in use is the Sex Role

Orientation Scale (SROS) by Brogan and Kutner (1976). This

scale was devised in order to investigate the "normative

prescriptions for the behavior of males and females" (p.

32). The SROS consists of 36 items, rated on a 6-point

Likert-type scale. The item statements describe specific

behaviors which are based on one of six content areas. The

six areas are 1) division of labor in marriage, 2) sex based

power structure, 3) employment of women, 4) political status

of women, 5) sex role socialization of children, and 6)

miscellaneous sex role behavioral areas such as morals or

standards of dress. Each of the 36 items on the scale is

scored from 1 to 6. A subject with a higher total score is

rated as having a more nontraditional sex role. This scale

has been used in the sex role literature and should continue

to be employed due to its emphasis on behavior rather than


Research Findings on Sex Roles

The research on sex roles can be divided into three

areas, research concerning 1) sex role stereotypes (beliefs

about "typical" males and females), 2) sex role

orientations (beliefs about "ideal" or "desirable" males and

females), and 3) personal sex roles (self-perceived beliefs

about oneself). The vast majority of the literature deals

with sex role stereotypes. However, there are studies in

each of the other areas as well. Stereotypical findings

have been surprising in their lack of change over the years,

whereas in general, sex role orientations and personal sex

roles have shown some movement away from traditional sex

roles and toward more egalitarian sex roles, although these

findings are inconsistent.

The systematic investigation of sex role stereotypes

began in earnest with the development of the SRSQ by

Rosenkrantz et al. (1968). Their initial findings concluded

that there were substantial differences between typical

males and typical females in terms of sex role stereotypes.

Subsequent investigations have employed not only the PAQ

(the updated SRSQ), but other instruments as well, and have

replicated these early findings, even though the studies

have been reported as recently as 1987. So, for nearly 20

years, the status of sex role stereotypes has remained

fairly constant. Examples of this work include Spence,

Helmreich and Stapp (1975) and T. L. Ruble (1983). Both

sets of researchers employed the PAQ and found substantial

evidence of sex role stereotypes. In fact, T. L. Ruble

found significant differences on 53 of the 54 traits the PAQ


Recent work on sex role stereotypes has expanded from

the typical psychological trait approach to the use of

behavioral and interest scales. Canter and Meyerowitz

(1984) asked college students to respond to a behavioral

questionnaire. The results indicated typical sex role

stereotypes in perceptions of male and female behaviors.

Subjects also indicated that male sex role stereotypes were

stronger than female sex role stereotypes. Interests and

personality factors were explored by Lewin and Tragos

(1987). Their subjects were 15 to 18 years of age and were

compared to adolescents tested in 1956. Results showed no

change in sex role stereotypes for females, while males

indicated increasing traditionalism.

The final study reviewed here concerning sex role

stereotypes compared college students in 1978 to a 1957

sample by employing a checklist of sex role adjectives

(Werner & LaRussa, 1985). Their results found that 62% of

the adjectives used to describe males in 1957 were still

employed in 1978, while 77% of the adjectives used to

describe females in 1957 were still in effect. None of the

adjectives switched over from one sex to the other, and the

female adjectives that did change showed an increase toward


more favorable evaluations, while the male adjective changes

showed an increase in more unfavorable evaluations. The

overall indication of changes in sex role stereotypes has

generally shown little movement. The majority of the

empirical evidence has found that sex role stereotypes still

remain strong after more than 20 years of investigation.

Sex role orientation research has received less

attention, and the findings are not as consistent. The lack

of consensus in this area is not surprising in that sex role

orientations have a moral aspect. Orientations are opinions

of how people should behave and thus are more likely to

deviate over time than perceptions of actual behavior (sex

role stereotypes). The upcoming section on sex role

stereotype and sex role orientation ratings deals with this

topic in detail. Hence at this point only a summary will be


Studies of sex role orientation ratings have shown

contradictory results. Some suggested that orientations

should be sex typed and traditional, while others have

suggested the opposite finding, that orientations should be

androgynous and egalitarian. The sex of the rater appears

to be an important factor in determining sex role

orientations. Females are more likely to rate orientations

less traditionally than males (Deutsch & Gilbert, 1976;

McPherson & Spetrino, 1983; Scher, 1984). Yet, several

studies indicated that orientations are sex typed (T. L.

Ruble, 1983; Silvern & Ryan, 1983; Urberg & LaBouvie-Vief,

1976). Finally it should be noted that the recent studies

(e.g., T. L. Ruble, 1983), are not finding different results

from the older studies (e.g., Urberg & LaBouvie-Vief, 1976).

Overall there appear to be no consistent conclusions

regarding sex role orientations.

The final area of investigation concerns personal sex

roles, or the ratings given oneself on sex role measures.

Again it appears that there is no consistent trend in how

individuals perceive themselves, and it has been suggested

that the reason for this inconsistency is because

subcategories of sex roles have changed, and much of the

research has not dealt with subsections of sex roles, only

the totality of sex role scores (Thornton, Alwin, & Camburn,

1983). There is one finding on personal sex roles that has

been consistent from study to study, and that is that males

rate themselves as more traditional than females (Brogan &

Kutner, 1976; Helmreich, Spence, & Gibson, 1982; Storms,


Beyond that one reliable finding, research requesting

subjects to evaluate themselves has found sex typed

individuals, movements toward egalitarianism, and movements

toward more traditional sex roles. Employing the PAQ, and

using college students as subjects, both Spence et al.,

(1975) and Storms (1979) found that subjects believed

themselves to be sex typed. The work of Canter and


Meyerowitz (1984) and Robinson and Follingstad (1985) echoed

the stereotypical sex role for behavioral aspects rather

than psychological trait aspects of the sex roles. Canter

and Meyerowitz found that college students endorsed

stereotypical behavior in terms of ability, enjoyment of

activity choice, opportunity to act and competence in these


On the other hand, personal sex roles have indicated

less traditionalism. McBroom (1984, 1987) found that both

males and females endorsed less traditional domestic sex

roles over the course of a five year longitudinal study.

Initial support for this finding was offered by Helmreich,

et al. (1982). These authors found increasing levels of

egalitarianism between the years of 1972 to 1976. However,

between 1976 and 1980, females showed a tendency

(statistically significant) toward becoming more

conservative, while the males showed no change in their

perceived sex roles. Thus the field is not clear concerning

changes in personal sex roles.

Research on personal sex roles and sex role orientation

has not generated a common consensus, whereas studies on sex

role stereotypes have. Males and females consistently rate

the stereotypical male as more traditional than the

stereotypical female (e.g., Brogan & Kutner, 1976; Larsen &

Long, 1988; Storms, 1979; Swatos & McCauley, 1984). Studies

testing sex role stereotypes have utilized different

measures testing psychological traits, behaviors,

personality factors, interests, and attitudes. The subject

populations tested have been variable and research has

spanned over 20 years. Thus the pervasive existence of sex

role stereotypes has been extensively documented.

In addition to these indications of the pervasiveness

of sex roles stereotypes, there are other indicators of

their generality. Swatos and McCauley (1984) documented sex

role stereotypes in the working class, extending the finding

beyond the typical college sample. Canter and Ageton (1984)

found that while there are differences between races in

terms of amount of sex role stereotyping, no race is free

from this phenomenon. Williams and Best (1982) extended the

findings cross culturally by exploring sex roles in 30

nations and found that women were regarded as expressive and

men were regarded as instrumental in every country studied.

Hence sex role stereotypes are deeply ingrained in culture

and continue to be of interest to developmental

psychologists as well as other scientists due to the

pervasiveness, generality, and apparent general level of


Methodological Problems in the Sex Role Literature

There are four methodological issues that will be

addressed. The first is the problem involving definition of

terms in the sex role literature. A second problem concerns


measurement difficulties. The third problem area has to do

with the conception of sex roles as a unitary versus a

multidimensional phenomenon. Lastly, the issue of the

magnitude of the differences will be investigated. These

issues are not the only methodological problems addressed in

the literature; however, they appear to be the most commonly


There are two definitional issues that must be

addressed in this section. The first controversy surrounds

the use of the terms sex role stereotype versus sex

stereotype (D. N. Ruble & T. L. Ruble, 1982). The use of

the word "role" implies "appropriateness" of behavior to

some, and thus suggests a moral element. However, the

generally accepted definition of sex roles states nothing

about appropriate behavior; rather, it states that sex roles

are the actions and verbalizations one displays to others to

indicate one's gender (Money & Ehrhardt, 1972). Thus the

use of the term sex role stereotype is controversial.

The second definitional problem is more complex. It

involves how psychologists have defined and operationalized

the concept of sex differences (Eagly, 1987). In order to

achieve the best experimental control, psychologists have

attempted to control all outside variables, with the

exception of gender in order to determine what sex

differences exist. Thus subjects are put into social

interactional situations with people they do not know, in a

place they are unfamiliar with, for a short period of time.

If males and females react in the same manner, sex

differences are not inferred. However, Eagly pointed out

that this situation is exceptionally contrived, and hence

the ecological validity may be totally lacking. As an

example, Eagly referred to Maccoby and Jacklin's (1974)

conclusion that males and females do not differ on level of

nurturance, based on experimental (assumed laboratory)

evidence. However, one look at reality would indicate that

this conclusion is a complete violation of common sense.

Women outnumber men in nurturing occupations by vast

amounts. In general women are the homemakers, teachers

(especially at the elementary level), and nurses. Hence, to

conclude that men and women do not differ on level of

nurturance appears to violate observation of real life

roles. Thus, Eagly suggested that psychologists need to

confront reality in the social world not just laboratory


There are several different measurement problems that

deal with sex role scales. D. N. Ruble and T. L. Ruble

(1982) compared different sex role scales and concluded

that, in part, the differing results between studies were

due to the different styles of measurement. They suggested

that there are four main types of measurement: open ended

styles such as interviews or questionnaires, adjective

checklists that assign traits to each sex, bipolar rating

scales, and Likert-type multiple point rating scales.

Research in the area has shown that the forced-choice

methods of ratings increase the number and magnitude of sex

role stereotypes (Marantz & Mansfield, 1977). Hence

different measurement scales may result in varying findings.

Although there are several sex role measures in the

literature, Orlofsky (1981) stated that there are no good

measures of sex role interests and sex role behaviors. The

mainstream of the psychological literature has focused on

traits or personality dimensions. Canter and Meyerowitz

(1984) also believe that the focus on traits is

inappropriate. Traits are hypothetical constructs that must

be inferred from behavior. Due to the need for inference on

the part of the subject, the use of psychological traits may

be vague or misleading. Each subject rating traits must

interpret what that trait means. What is independent? What

is caring? These traits mean different things to different

subjects and so may be inaccurate. Brogan and Kutner (1976)

echoed this sentiment and believed the logical conclusion

was to measure attitudes toward specific behaviors, rather

than measuring traits and then inferring behavior.

A final measurement problem deals with the inability

of the measurement instruments to keep pace with changes in

the field. Orlofsky (1981) had two complaints in this area.

The first had to do with the change in the

conceptualizations of masculinity and femininity. These

concepts were originally operationalized as bipolar traits

(Rosenkrantz et al., 1968). However, since then masculinity

and femininity have been restructured as independent traits.

Yet the old measures (e.g., SRSQ and the PAQ) are still in

use. The other problem that Orlofsky described was the

failure of the area to update the measurement instruments.

The existing scales are viewed as dated by Orlofsky, and due

to the cultural changes of the last 20 years many of the

items no longer reflect sex role stereotypes. It is also

possible that the scales reflect ceiling effects, with the

item statements being so outdated that everyone either

completely agrees or completely disagrees with the idea.

Hence measurement instruments are in need of updating in

order to stay current with the field and with societal


It has been suggested that sex roles are

multidimensional (D. N. Ruble & T. L. Ruble, 1982). Sex

roles and their stereotypes are found when studying

attitudes, interests, occupational preferences, personality

dimensions, and behavior. Yet several authors have found

that the correlations between these various aspects of sex

roles are quite low (Deaux, 1984; Komarovsky & Mayer, 1984;

Mussen, 1969; Orlofsky, 1981). Steven et al. (1983) found

that attitudes as measured by a sex role scale did not

correlate with interview answers that tapped behavioral

dimensions of sex roles. Komarovsky and Mayer found weak

relationships between sex role stereotypes, orientations,

and personal sex roles, and urged researchers to be very

careful in their use of terminology in order to avoid

confusion between studies.

If, as suggested, sex roles are multidimensional, then

the inconsistency among studies may be due to the different

dimensions being tested with each measure. Brogan and

Kutner (1976) constructed their SROS based on six different

aspects of sex role orientation such as division of labor in

marriage, employment of women, and political status of

women. Thornton et al. (1983) clustered their questions

into eight categories covering such things as working

women's relationships with their children, happiness of

working versus at home mothers, and the wife's

responsibility toward helping the husband's career. With

differing scales tapping different dimensions of sex roles,

and with perhaps differing weights to each dimension, it is

not surprising that a comparison of total scores between

studies often shows conflicting results. Hence, the

dimensionality of sex roles is an area that deserves

methodological attention (Deaux, 1984).

The final area of methodological problems revolves

around the magnitude of the differences that are noted in

sex role studies. Maccoby and Jacklin's (1974) overall

conclusions after reviewing the sex differences literature

were that there were very few actual differences between the

sexes, that existing differences were very small in degree,

and that, for the most part, males and females were more

alike than different. Deaux (1984) agrees with these

findings. Furthermore, Deaux states that even when main

effects for sex are found, they are relatively weak. The

percentage of variance accounted for by sex is in most cases

under 5%. Hence researchers should be searching for more

important influences on social behavior.

The issue of the magnitude of sex differences has

received in-depth attention from Eagly (1987). Eagly

suggested that

judgments about size should not be delivered upon the
mere inspection of a sex difference expressed in a
particular metric, because meaningful judgments should
take into account other known research findings,
aspects of research methods, and, if possible, the
evaluation of the behavior in the society. (p. 114)

Even the meta-analysts who have attempted to ascertain what

is a "small" difference and what is a "big" difference do

not agree. Therefore, Eagly suggested several guidelines

for the interpretation of effect sizes.

One suggestion was to compare mean sex differences to

mean differences between males and females in other areas of

social behavior, such as aggression or conformity. Another

idea was to employ binomial effect sizes. "To apply this

technique to the interpretation of sex differences, an

investigator determines the percentage of each sex above the

average response in the combined group of women and men"

(Eagly, 1987, p. 118). Eagly's last suggestion relied on

the common sense of the investigator. She recommended an

examination of the utility of the behavior. Differences

that appear to be small in magnitude may be of major

importance if the behavior is one that is highly valued or

closely monitored by others.

The importance of the magnitude of size effects should

not be slighted. It appears that many researchers have

forgotten the basic caveat of an introductory statistics

class, which is that statistical significance does not mean

practical significance. Studies in the area frequently

report statistically significant findings, and they report

the appropriate analysis of variance or multiple regression

statistics, but they often fail to give the basic

descriptive data necessary to evaluate the research findings

on a practical level. It is common to find reported means

for the groups of interest missing. Hence, judgments are

not possible on behalf of the reader.

The other side of the situation occurs when means are

given, but apparently ignored by the researcher. As an

example, two studies by McBroom (1984, 1987) reported that

"longitudinal comparisons over the five year period of 1975-

1980 show that both men and women have significantly

lessened in sex-role traditionalism" (1987, p. 439). Yet it

should be noted that the mean changes in sex role scores

over a five year period for both men and women were less

than one full point. Thus, McBroom's conclusion that men

and women are "significantly less traditional" (1987, p.

448) is highly circumspect. Again, a critical reading of

the literature must be undertaken in order to properly

evaluate the results as presented by each researcher.


There is an abundance of literature, terms, measurement

devices, and not surprisingly, methodological problems in

the study of sex roles. The conflicting empirical results

can be untangled if careful analysis of each study is

undertaken. Problems may result from misuse of terminology,

measuring devices that test separate dimensions of sex

roles, or the mistaken interpretation of the magnitude of

the effects that are found. Such a vast area requires

careful attention to detail in order to lessen confusion and

strengthen findings. Any of the areas just discussed could

lead to interpretive errors and so extreme caution should be

employed when researchers draw conclusions regarding sex


The Development of Sex Typing

Developmental research in the area of sex typing has

taken many forms, employed several sex role measures, and

studied a wide variety of age groups. Not surprisingly,

these studies have arrived at a variety of conclusions.

Shaffer (1985) suggested dividing developmental sex role

research into three areas. These three areas are gender

identity, the development of sex role stereotypes and the

development of sex typed behavior. The area of gender

identity focuses on cognitive-developmental sex role theory

as espoused by Kohlberg (1966) which will be discussed later

in the developmental theory section. The developmental

trends and changes in sex role stereotypes across the life

span will be summarized first. Then the development of sex

typed behavior which deals with the activities and interests

in which males and females engage in to varying degrees will

be discussed.

Development of Sex Role Stereotypes

"In general the available research suggests that, on

most dimensions, children's knowledge of sex stereotypes

develops at an early age. .children make few errors in

assigning sex stereotypic labels to activities, occupations,

and playthings" (D. N. Ruble & T. L. Ruble, 1982, p. 215).

The overall developmental trends are not clear but appear to

present a curvilinear trend with sex roles beginning early

and strong, then lessening and then again becoming quite

robust (Streitmatter, 1985). The next section will attempt

to document these chronological changes in sex role


As soon as children begin to acquire gender identities,

they also begin to show signs of sex role stereotypes. And


children who display the most stereotyping are the ones with

the most mature gender identities (Kuhn, Nash, & Brucken,

1978). The trend to stereotype increases over the next

several years with children showing an increase in the

rigidity of their sex role stereotypes. Sex roles are

viewed as immutable rules that must be followed, and no

deviation is allowed. However, this rigidity relaxes as the

child reaches the approximate age of ten (Marantz &

Mansfield, 1977; Ullian, 1976). At this point in

development, sex role stereotypes become more pervasive,

spreading to more abstract areas such as personality traits

and achievement. Children also become more flexible in that

it is now realized that personality traits, attitudes,

behaviors and the like, while associated with one gender

more than the other, are not required based on biological

sex (Ullian).

During adolescence the trend fluctuates. The

individual starts off as stereotyped, relaxes, then becomes

increasingly stereotyped once again (Sigelman, Carr, &

Begley, 1986; Streitmatter, 1985). Streitmatter suggested

that sex role measures employing total scores blur the

issue. It is individual components of the total scores that

are changing over time with some components increasing, some

decreasing, and some remaining constant. Streitmatter

argued that this supported Erikson's stage of identity

formation. At this stage, the adolescent is reevaluating

and incorporating new roles into the newly forming self-


A study that did not support the curvilinear trend was

conducted by Canter and Ageton (1984). They found a steady

decline in traditional sex role stereotypes for females ages

13 to 19. While their results did show statistical

significance, it should be noted that the mean scores

collected as part of the National Youth Study dropped over

the seven year period by a total of 1.34 points on a scale

with a range of 40 points and this trend did not apply to

males. There are changes in sex role stereotyping during

the adolescent years, however, the direction of these

changes is not consistent.

In adulthood, there were also some inconsistent

findings, but the overall trend was for younger adults to be

less traditional and older adults more traditional in their

sex role stereotypes. These results were all based on

cross-sectional data and hence should not be viewed as

developmental stages or changes. A more likely explanation

would be the finding of generational effects with older

subjects displaying a continuing attitude rather than

movement toward conservatism. Brogan and Kutner (1976)

found that males over the ages of 23 were more traditional,

while Swatos and McCauley (1984) found that males and

females over the ages of 35 were more traditional. Dambrot

et al. (1984) tested three generations of women and found

that only the eldest (mean age of 69) were significantly

more traditional in their sex role beliefs. Mothers (mean

age of 43) were more like their daughters (mean age of 19)

than the eldest group (the grandmothers). A contrary

finding by Urberg (1979) indicated that adults over 25

(through age 65) actually stereotyped less than 12- and 17-

year old subjects. Hence, the results cited in the

literature are not consistent throughout all adulthood

cohort groups.

Development of Sex Typed Behavior

Children show preferences for sex typed toys, games and

even same sex playmates before the age of three (Strayer,

1977). Investigations of children's bedrooms are often very

telling (Rheingold & Cook, 1975). Girls rooms are often

pink, frilly and filled with dolls and stuffed animals.

Boys rooms tend to be bright colors, plain and loaded with

balls and cars. The strong sex typing of these early years

gives way temporarily for girls. Between the ages of four

to ten, there is an increasing awareness of sex role

stereotypical behavior. However, girls become increasingly

interested in engaging in cross sex activities, games and

toys (Hall & Halberstadt, 1980). The male role is more

strongly stereotyped, and pressures to conform to it are

strong. Boys are taught at early ages to be tough, not to

cry and not to be a "sissy." Girls however are allowed to

deviate from their sex roles. Playing soccer, climbing

trees and being a "tomboy" are tolerated, although perhaps

not totally accepted. Thus girls are allowed to experiment

with male roles, while males are not allowed to experiment

with female roles.

There are also sex differences in aggressive, compliant

and dependent behavior. Even in the preschool years, boys

instigate and retaliate with more physical and verbal

aggression than girls (Maccoby & Jacklin, 1980). The higher

level of male aggressiveness is a stable trait. In

adolescence and adulthood, males are five times more likely

than females to be involved in antisocial behavior and

criminal activity (Johnson, 1979). Also beginning in the

preschool years, girls display more compliant and dependent

behavior than boys (Block, 1976). Girls are more compliant

to peers as well as adults. Furthermore, girls are more

likely to seek adult assistance than boys. Hence, sex

differences in emotionally based behaviors also exist.

In regards to achievement and career choice, again

there are sex differences. Career choice follows the same

path that behavior does. As boys and girls mature

traditionally male dominated careers are considered more

acceptable for both males and females, while males do not

accept the idea of engaging in traditionally female

dominated occupations (Hensley & Borges, 1981). Achievement

offers one of the most dramatic examples of sex differences.

Not only do different areas show sex differences (spatial

ability, verbal ability, math ability), but more telling are

the explanations behind achievements and failures as offered

by males and females (Nemerowicz, 1979). When females fail,

it is because of a lack of ability. When they succeed it is

because of effort. For males, failure is due to lack of

effort or failure to encounter a learning opportunity, while

success is due to ability. Additionally, female achievement

is devalued by members of both sexes, especially if the

success is in a traditionally masculine area. Thus

differences in behavior are evident from early ages and

explanations for these differences are dependent on the

gender of the individual.


The inconsistencies in the literature regarding the

development of sex typing may have several explanations.

The measures employed often require a forced choice on the

part of the subject. This methodology has been implicated

in increasing the amount and the magnitude of differences

between the sexes (Marantz & Mansfield, 1977; D. N. Ruble &

T. L. Ruble, 1982). Additionally, the use of total scores

may be deceiving. Sex role measures are often made up of

several very different areas such as employment of women,

political status of women, or division of labor in marriage.

It has been suggested that these factors may be

differentially salient across the life span. Hence, total

scores obscure the changes that may be taking place at a

more specific level or at a specific point in time.

Furthermore, differences between studies may be due to

differential loadings of these factors between measures. A

final reason for variability between studies may be due to

the idea that sex roles fluctuate across the life span

(Minnigerode & Lee, 1978). The actual social roles that an

individual becomes involved in over the course of a lifetime

change dramatically, and with these changing roles comes a

different set of behavioral and attitudinal expectations.

For example, parenthood tends to increase traditional sex

roles, while grandparenthood tends to relax sex roles (Cowan

& Cowan, 1983). Thus there are many reasons why the

literature appears to have inconsistent findings and more

work in the area of the development of sex roles is


Developmental Theories

There are several theoretical formulations that have

been advanced to explain the development of sex typed

behavior. The oldest would be psychoanalytic theory as

developed by Freud. The main emphasis for Freud is on

identification of the child with the same sex parent. The

two theories that have amassed the most research are social

learning theory and cognitive-developmental theory. Social

learning theory espouses basic principles of learning to

explain sex typing. Cognitive-developmental theory centers

on the development of gender identity and the cognitive

processes needed to acquire a mature sense of gender. Newer

theories are gender schema theory and social-role theory.

Gender schema theory blends social learning with cognitive-

developmental and adds information processing aspects as

well. Social-role theory explains sex roles as based on

adaption to other social roles rather than sex roles as

being primary. These five theories will be elaborated with

supporting and refuting evidence offered where available.

Psychoanalytic Theory

The basis of Freud's theory of sex typing is the

process of identification. According to Freud, sex typing

occurs during the third psychosexual stage of development,

the phallic stage. Children are generally between the ages

of three to six at this point, and the beginning of sex

typing occurs once the child has noticed the biological

differences in genitals between the sexes. Having noticed

the difference between the sexes, the child must now set out

to resolve the Oedipus Complex if the child is a boy, and

the Electra Complex if the child is a girl. Both

resolutions revolve around identification with the same sex

parent. Boys are hypothesized to have a stronger and

earlier resolution to their complex due to the increased

severity and tension the complex causes. In the end, the

boys imitate and internalize the attitudes and behaviors of

their fathers, while the girls model their mothers. Thus,

sex typing of the next generation has successfully begun.

In general, most of the current work in sex roles does

not employ a psychoanalytical framework. However, there is

evidence that does support some of Freud's ideas concerning

sex typing. For example, Shaffer (1985) supplies three

pieces of supportive evidence. First, the ages that Freud

hypothesized (ages three to six) for the phallic stage fit

in very well with studies on gender identity. Second, Freud

suggested that the developmental trends between males and

females will differ, with males showing an earlier and

stronger sex typing. This indeed seems to be the case.

Studies have shown that boys are more strongly sex typed at

earlier ages than girls. Finally, Freud indicated that

fathers are especially important in the sex typing of not

only boys, but girls as well. Evidence cited by Shaffer has

shown that at these early ages boys need a male role model

to emulate, and girls need a male model for long term

interpersonal social adjustment.

Yet, there is damaging evidence as well. Most

importantly, several studies (Katcher, 1955; McConaghy,

1979) have shown that children between the ages of three and

six cannot differentiate male/female genitalia. Since this

is the basis for the beginning of the sex typing process, it

seems impossible to proceed any further according to

psychoanalytic theory without this most valuable

information. Further, Freud stated that boys identify best

with a hostile father figure. Conversely, evidence has

shown that boys identify best with a warm and caring father,

rather than a hostile father (Hetherington & Frankie, 1967).

While there is support for some of Freud's assertions, the

damaging empirical evidence is fairly strong. Thus the

psychoanalytic theory of sex typing is no longer frequently


Social Learning Theory

The interpretation of social learning theory in the sex

typing literature is extremely varied. Some authors

consider it to include operant conditioning, observational

learning, identification, imitation, all of the general

principles of learning (such as reinforcement, punishment,

generalization, extinction, and discrimination), and the

most recent writings in the area contain a strong cognitive

component as well. Arguments have ensued regarding these

areas. However, for the most part there is agreement on the

basics of the social learning theory interpretation of sex


The basic premise of social learning theory in regard

to sex typing is that sex roles are a learned phenomenon.

Thus, they are not inevitable nor are they impervious to


change (Bem, 1983). Children can be taught to be sex typed

as masculine or feminine, or they can be taught to be

androgynous. Given the appropriate learning environment,

and correct models from which to glean information, a child

can become sex typed in the fashion that is appropriate or

desired by each society or culture. Bem gives extensive

instructions on how to raise "gender-aschematic" children,

by providing them with the "right" environment, both

physically and psychologically.

The manner in which sex typing is learned according to

social learning theory is based on the general principles of

learning. Shaffer (1985) breaks down the learning of sex

typing into two categories: direct tuition or teaching and

observational learning. Direct teaching of sex roles

involves the processes of reinforcement and punishment.

When children display sex-appropriate behavior, they are

reinforced and hence the behavior is more likely to recur.

When children engage in sex-inappropriate behavior, they are

punished, and hence the behavior is less likely to recur.

The most salient teachers in this regard are the parents.

And the sex role expectations regarding children's behavior

are quite clearly communicated to children, even at young

ages. As an example, consider the color and style of dress

of a newborn on the way home from the hospital. Also

consider the toys already received. For girls, the flowers

and balloons are pink, along with her frilly dress and pink


baby blanket. The toys are dolls and stuffed animals. The

newborn boy received blue flowers and balloons, a blue

sports outfit of some type, and probably even sporting

equipment such as balls or basketball hoops (with suction

cups to be attached to the crib). Even the stuffed toys

little boys receive are likely to be balls rather than dolls

of some type.

As the child grows, participation in and with sex-

appropriate items will be encouraged while playing with sex-

inappropriate items will be discouraged, especially for boys

(D. N. Ruble, 1984). D. N. Ruble has pointed out that

adults have differing expectations for girls and boys, and

so they treat them differently. Furthermore, children

themselves report that their parents do indeed treat them

differentially, and that their parents display different

attitudes toward them based on their gender (Mischel, 1970).

Not surprisingly, then, children form differing sex role

stereotypes about mothers and fathers, with mothers viewed

as more nurturant and fathers viewed as more powerful and

more punitive. D. N. Ruble also pointed out that boys

receive more pressure to conform to sex roles than do girls,

and that the pressure is exerted more often by the father

than the mother.

In addition to the differential effects of

reinforcement and punishment as methods of direct teaching

of sex roles, the principles of generalization and

discrimination are also employed. The principle of

generalization states that once a response has been learned,

it will be demonstrated in other situations that are similar

to the original circumstance. Since girls are often

rewarded for obedience and quiet activity such as coloring,

it is likely that this behavior will be generalized to other

similar situations such as nursery school. Boys, on the

other hand, are often rewarded for rough and tumble play,

such as sports activities, and so it is likely that when

playing with other boys at school these sex-appropriate

behaviors will again be utilized and reinforced. Mussen

(1969) suggested that as cognitive development continues,

children learn to label activities that are sex-appropriate,

and so generalization will be verbally mediated. This

mediated generalization serves to strengthen and further

facilitate sex typing in the growing child.

The flip side of generalization is discrimination.

Discrimination involves the enactment of sex role behavior.

At this point the behavior has been learned, but the actor

must decide whether or not this is the appropriate time and

place in which to engage in this specific behavior. Even

sex role behavior that is extensively culturally regulated

is not appropriate in every situation. Therefore, sex role

behavior is also situationally specific (Eagly, 1987).

Mischel (1970) rather generally stated that enactment of sex

role behavior is based on motivational considerations. A

more specific set of factors was outlined by Orlofsky,

Cohen, and Ramsden (1985). Orlofsky et al. suggested that

sex role enactments are influenced by the fact that

conforming to societal norms is highly sanctioned, that

individuals behave in order to escape negative sanctions,

and that people often behave in order to use the situation

to their best advantage. The best example of discrimination

of sex role behavior would be displayed by an androgynous

person as defined by Bem (1974). According to Bem, the

purpose of androgyny is flexibility. Androgyny allows the

individual to behave according to the situation first, and

not according to the sex role. So enactment of sex role

behavior is situationally constrained, and individuals that

are defined as androgynous are the most situationally

flexible or discriminating in their behavior.

The other major tenet of social learning theory is

observational learning. Mischel (1970) suggested that sex

roles are initially acquired by observing the actions of

others and then the subject models or imitates these

actions. Reinforcement or punishment need not be personally

experienced to gain knowledge about appropriate and

inappropriate sex role actions.

The main concern for social learning theorists is the

nature of the model. At what point do children pay

attention to models, and are some models more salient than

others? The original hypothesis was that children pay

attention to same sex models at very early ages and so

observe differences in treatment of the sexes early on.

However, it has been found that children do not pay

attention to same sex models until they are six or seven

years of age, long past the original stages of sex role

development (D. N. Ruble, Balaban, & Cooper, 1981). The

main characteristics of a model that increase the chances of

a child paying attention and hence learning are nurturance

and power (Mischel, 1970). This appears to be especially

true for boys, but it also holds for girls. It is also

important to note that children do not model only adults.

Other children play a substantial role in sex typing as the

sanctions placed by children against other children can be

very harsh. Playgrounds abound with taunts and name calling

by children. In order to avoid these negative sanctions by

ones peers it is best to behave in accord with gender

expectancies. Otherwise one runs the risk of being labeled a

"sissy" or a "tomboy" and becoming an outcast. Therefore,

children model other children that they believe are powerful

and are nurturant in hopes of being rewarded and accepted.

The latest theoretical position that falls into the

social learning category (very broadly defined) would be the

social-cognitive appraisal of sex role development as

defined by Bussey (1983). Bussey attempted to update the

social learning theory position on same sex modeling, while

also incorporating a stronger cognitive component into the


model. Bussey contended that past research was not totally

supportive of children engaging in same sex modeling. Her

work has shown that if children were exposed to multiple

models of both sexes, they would differentiate between the

two sexes and then determine what proportion of behaviors

are engaged in by each sex. This proportion gives the child

the answer as to whether or not the behavior is sex

appropriate for him or her, and thus the child can respond


Bussey (1983) employed a four step model of sex role

development. Step one involved the realization that there

were two groups of people, males and females. Step two was

to recognize which group they belong to. Step three

required the cognitive capacity to label and then encode

behaviors as either male or female appropriate. That is,

one must have the ability to categorize sex role behavior.

Step four was the realization that the sex role typing

complete with cultural sanctions, reinforcement and

punishment applied to oneself. In addition, Bussey noted

that step four must be qualified by stating that children of

different ages give different reasons for their sex-

appropriate behavior. By involving a stronger cognitive

component, Bussey attempted to create a more active

participant in sex role development than the traditional

social learning theory which contained a more powerful

environmental element, and hence a more passive subject.

Bussey's work shows promise as an updated social learning

theory of sex role development.

Cognitive-Developmental Theory

Kohlberg's (1966) theory of sex typing "assumes that

basic sexual attitudes are not patterned directly by either

biological instincts or arbitrary cultural norms, but by the

child's cognitive organization of his social world along sex

role dimensions" (p. 82). It is the child that is the

primary sex role socialization agent. The child actively

attempts to make sense of the social world. The child

organizes thoughts and perceptions into sex role knowledge

and understanding based on personal interpretations of the

social world. Kohlberg states that "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" (p. 83).

Observational learning, imitation and reinforcement are all

utilized. However, the concept of gender identity precedes

the use of these learning tools. Kohlberg hypothesized that

the internalized sex role identity develops first; then sex

appropriate behaviors follow.

There are three developmental stages of sex typing.

The first occurs around the age of two to three and is

called gender labeling. During this stage children learn

that there are two sexes and that they themselves belong to

one of these two groups. Kohlberg maintained, unlike

psychoanalytic theory, that the basis for understanding the

difference between sexes is not based on knowledge of

genitalia. Instead Kohlberg focused on the more outward

expressions of masculinity and femininity such as physical

size, hair style, or clothing.

The second stage of development is labeled gender

stability which occurs around the ages of three to five.

The cognitive achievement that occurs during this stage is

one of temporal stability. During gender stability,

children come to understand that gender is fixed across time

such that little girls grow up to be mommies and little boys

grow up to be daddies. Hence once a boy, always a boy.

The third stage is labeled gender consistency and

occurs between the ages of four and seven. At this point of

development the child becomes aware of consistency in gender

across situations. For example, superficial changes in hair

length or style of dress do not cause changes in gender.

Appearance, while perhaps deceiving, does not alter ones

biological sex. Thus the formation of a mature gender

concept has now been completed.

Huston's (1983) review of sex typing pointed out that

the development of sex roles and sex role stereotypes is not

a linear trend, but rather a curvilinear function. This

trend is based upon the child's cognitive developmental

level. Sex role stereotyping starts off weakly and


inconsistently until the child learns gender constancy. As

this learning is occurring, the rigidity of sex role

stereotyping is increasing. Once the child has moved into

Piaget's stage of concrete operations, the child can begin

to relax his views and become more flexible. But as

adolescence nears an end, sex role stereotypes are once

again pervasive.

According to Shaffer's (1985) review, cognitive-

developmental theory has received support for several of its

components. First, there is a relationship between

cognitive development and gender development as

hypothesized. Second, the three stage progression of gender

identification has been well documented. Finally, studies

have shown that gender understanding does affect sex role

stereotyping across the life span. Therefore, several of

Kohlberg's ideas have received empirical support.

There are also several problems associated with

cognitive-developmental theory. The most damaging empirical

evidence is that sex typing is well underway before gender

identity is completely established. For example, even

before the age of two, children are already showing a

preference for sex typed toys. Bem (1983) saw Kohlberg's

failure to explain why sex roles have a primacy over other

categories such as religion or race as a major problem.

According to Bem, Kohlberg viewed sex as more perceptually

salient than other categories, but failed to explain how the


categories are hierarchically structured. Another criticism

was levied by Mussen (1969). Mussen argued that cognitive-

developmental theory was descriptive rather than

explanatory. Hence, while it may provide interesting

reading, it fails to explain how development actually

occurs. Thus, Kohlberg's theory is not without problems.

Some of these problems appear to be especially damaging.

Gender Schema Theory

Gender schema theory is a true hybrid among theories.

It combines aspects of both cognitive-developmental theory

and social learning theory. However, gender schema theory

is considered primarily an information processing theory of

sex typing. Two sets of researchers have simultaneously

developed gender schema theories and for the most part the

theories are compatible (Bem, 1981, 1983; Martin &

Halverson, 1981). The basic construct in gender schema

theory is the schema which is defined as, "a cognitive

structure consisting of a set of expectations or a network

of associations that guide and organize an individual's

perceptions" (Huston, 1983, p. 399).

Schemas serve three main functions according to Martin

and Halverson (1981). Schemas regulate behavior in that

they set up a guide for anticipating future behavior. Other

areas have labeled this function a script. Scripts or

schemas define a temporal sequence of actions for familiar

situations. The second function of schemas is to organize

incoming information and direct attention selectively.

Schemas dictate what information should be paid attention to

and what information should be ignored. Information that is

consistent with the present gender schema will be

incorporated into the schema and further elaborated upon.

Information that is inconsistent with the gender schema may

be ignored or it may be misinterpreted and hence made to

fit. It is this function of schematic processing that

explains why sex role stereotypes are so hard to change and

why they are self-perpetuating. They become self-fulfilling

prophecies due to their ability to screen out incompatible

information and accept only confirming information. The

third function of schemas is to provide structure for

information that may be incomplete or ambiguous. Again in

this regard schemas mirror scripts. The ability to fill in

missing information is labeled a default function which is

automatically undertaken in order to clarify and complete

the incoming material. Thus gender schemas increase the

efficiency of information processing, but at the same time

they may also distort the information in order to make it


The development of sex typing in children requires one

basic assumption according to Martin and Halverson (1981).

This assumption is congruent with the cognitive-

developmental viewpoint, as it states that children have a


tendency to group or categorize information. By classifying

information children begin the process of self-definition.

Classification by sex is considered a pervasive phenomenon.

In fact, even Bem (1983) states that "no other dichotomy in

human experience appears to have as many entities linked to

it as does the distinction between females and males" (p.


The first developmental step (Martin & Halverson, 1981)

would be the ability to accurately identify people based on

gender distinctions. Once simple gender identity is

established, the "in-group" with which one identifies will

come to be evaluated positively while the "out-group" will

come to be evaluated negatively. Hence, the second step is

to associate values with the traits, behaviors and attitudes

that fit into the person's own gender schema. Continued

information processing leads to elaboration and extension of

the personal gender schema, while at the same time, gender

inconsistent information is ignored. This leads to a lack

of development of the opposite sex gender schema. As

development proceeds, the self-concept itself becomes

assimilated into the gender schema (Bem 1983). If the

individual's self-concept is organized on the basis of the

gender schema, the person will be viewed as extremely sex

typed. Thus an androgynous individual refrains from making

judgments based upon gender as a main variable, while a

highly sex typed individual will employ the gender based

explanations more freely. Development, then, is a

continuous process involving additions to the gender schema

and sex stereotyping is considered a normal cognitive


There is evidence to support the influence of gender

schemas in sex typed individuals compared to nonsex typed

individuals. Bem (1981) found that sex typed persons

recalled significantly more gender related words in a free

recall experiment. Highly sex typed persons were also found

to display faster information processing when making gender

schema-consistent judgments. Thus the availability of

gender as an extremely salient variable for sex typed

individuals has been supported.

There are also some problems with the gender schema

theory of sex typing. According to Martin and Halverson

(1981) androgynous persons are hard to explain. How is it

that they resisted sex typing? In fact, gender schema

theory would predict that androgynous individuals would show

a lack of competence, since inappropriate information would

be filtered out of their schemas. Additionally, Martin and

Halverson see many potentials for distortion in schematic

processing which could potentially lead to very distorted

sex typing. Individuals may use inappropriate schemas or

employ an illusory data base or make illusory correlations.

Any of these errors could serve to make sex stereotypes much

more extreme than reality would suggest. Therefore,


continued work in gender schema theory is needed to clarify

these issues.

Social-Role Theory

Social-role theory revolves around the concept of roles

within social systems and the influences of current social

situations on adult sex role behavior (Bee & Mitchell,

1984). The social systems within which everyone operates on

a daily basis are made up of a series of positions. Roles

are the contents or behavioral implications of these social

positions. Roles are more formally defined as "a set of

expected behaviors or qualities" (p. 23) that characterize

social positions. The differing social roles share some

specific characteristics. Roles are viewed as at least

partially culturally specific. Different cultures may share

similar views of a particular role, but the specifics of the

behavioral expectancies will be defined very explicitly by

the culture in which one resides. As a corollary to this

characteristic, Bee and Mitchell note that as a culture

changes, so do the culturally prescribed roles; hence roles

are not stagnant. The second characteristic of roles is

that they usually occur in complementary pairs. Hence in

the case of sex roles, there are males and females. The

area of research and debate would not exist if humans as a

race were unisexual. If all humans were identical in

biological make-up, gender as a category would cease to

exist. The third characteristic of roles is that people

occupy several at one time. In addition to gender roles,

there are also family roles, work roles and age roles.

Because of this plethora of roles, role conflict or

role strain may occur. Role conflict occurs when the

differing roles in which an individual is engaged do not fit

together well. There may be time conflicts because of

engaging in too many roles or there may be psychological

conflicts because different roles require different

responses. In these situations the individual must choose

whether to use, for example, the gender response or the

worker response in a given situation. Role strain occurs

when one individual role becomes too demanding and the

individual feels incompetent to live up to its demands. New

parents often feel overwhelmed by their responsibilities

shortly after the birth of a child, especially the first

child. Hence the number of roles and the type of roles

occupied may lead to problems for the individual. Overall

the social roles that each individual engages in define the

parameters of that person's behavior.

The major theoretical position utilizing social-role

theory to explain sex role behavior has been espoused by

Eagly (1987). Eagly stated that "the theoretical

orientation that is proposed considers sex differences to be

a product of the social roles that regulate behavior in

adult life" (p. 7). Furthermore, Eagly believed that the


importance of adult social roles and contemporaneous social

influences override the "more distal factors such as

childhood socialization pressures and biological

predispositions" (p. 9). Overall, Eagly concluded that

the social-role theory of sex differences promotes a
view of social life as fundamentally gendered...women
and men are subjected to somewhat different
expectations, to which they conform to some degree, and
they develop somewhat different skills as well as
attitudes and beliefs. (p. 31)

So the oft found conclusion that as a group women are more

communal or expressive than men is explained by the common

domestic role women occupy. In the same vein, men are more

agentic or instrumental because of their general role of

worker/economic provider.

Eagly (1987) defined sex roles in the same fashion as

defined by Bee and Mitchell (1984). Sex roles are "those

shared expectations (about appropriate qualities and

behaviors) that apply to individuals on the basis of their

socially identified gender" (Eagly, p. 12). However, there

are two aspects of sex roles that separate them from other

social roles. Sex roles have a larger scope or a greater

extensiveness than other social roles. Sex roles find

application in almost every aspect of daily life from

division of labor, to political status, to socialization of

children. Sex roles then appear to permeate almost every

aspect of life. Additionally, sex roles have a greater

generality than other social roles. Expectations about sex-

appropriate behavior can include personality dispositions,


moral and social norms, and even style of dress. Thus, sex

roles are pervasive in our daily lives.

A fundamental error in research methodology according

to Eagly (1987) has been made by psychologists attempting to

understand sex differences. The mistake was labeled by

Eagly as one of failure to establish ecological validity.

Tests of sex differences are usually conducted in a

laboratory setting for a short period of time with

strangers. The reason for this was to control extraneous

factors and increase the power and internal validity of the

experiment. Unfortunately, the situation may be so removed

from actual life situations that the conclusions cannot be

generalized to settings outside the laboratory. As an

example, Eagly referred to Maccoby and Jacklin's (1974)

conclusion that there was no clear cut evidence to show that

females were more nurturant than males. Maccoby and

Jacklin's conclusions were based on a compilation of studies

(laboratory variety) that "could not find evidence that

female and male research subjects, when faced with the same

eliciting stimuli, behaved with differing degrees of

nurturance" (p. 11). However, completely ignored by Maccoby

and Jacklin, (along with most other sex difference

researchers) is the fact that in natural settings women do

occupy nurturant roles (domestic, teaching, nursing) in much

higher proportions than males. Therefore, behavior in

natural settings is found to differ from behavior in

laboratory settings. Eagly concluded that the implication

of this conflict is that "women and men often behave

differently because they are carrying out dissimilar

organizational and familial roles" (p. 11).

Social learning theorists and gender schema theorists

have agreed with the notion of sex roles reflecting

situational behavior. Eagly and her colleagues (Eagly &

Steffen, 1984) as well as others (Abrahams, Feldman & Nash,

1978; Buckley & Hundleby, 1983; England, 1988) have found

substantial support for the idea that role demands, such as

differing occupations or varying situations, override the

effects or expectations of sex role behaviors and attitudes.

This becomes the main hypothesis of Eagly's social-role

theory. It is conformity to the social role and not gender

roles that determines behaviors and attitudes. However, as

stated, sex roles are pervasive and general in scope. Due

to the continuing division of labor between females and

males, sex roles continue to dominate behavior and attitude


Social-role theory, while relatively new, has received

empirical support. Yet, several caveats for its use need to

be enumerated. A potential difficulty with social role

theory concerns the all encompassing definition of gender

roles developed by Eagly (1987). Eagly stated that gender

roles are more extensive and more general than any other

social role. This statement gives the impression that

gender roles would be the most dominant social role in an

individual's life. Hence, when an individual experiences

role conflict, the gender role should dominate the

situational response. Yet Eagly's own research (Eagly &

Steffen, 1984) indicates this is not the case. Work roles

have been found to be more dominant in determining

situational reactions than gender roles. Thus there appears

to be a conflict between the definition of gender role and

the resolution of role conflicts as defined by Eagly.

Other potential problems concern the explanation of why

an individual becomes involved in a particular role and the

level of activity displayed by the individual. Social-role

theory has not yet addressed the mechanics of how a person

becomes involved in a particular role. The explanations to

date have revolved around the impact the role has on the

individual without addressing the individual's entrance into

that specific role. In addition, social-role theory views

the individual as a passive respondent to the situational

role. Interindividual differences need to be assessed in

order to examine the interaction of personal characteristics

with the situational characteristics. This would move the

social-role theory towards a more active conception of human

nature. Thus, social-role theory is not complete. The

theory needs additional clarification of its concepts.


The chronological development of theoretical frameworks

has progressed from an emphasis on biological foundations,

to basic principles of learning, to cognitive functions, to

social roles. In addition, the theories have progressively

become more eclectic in their orientations. Most

theoretical developments have also encouraged changes in the

conception of human nature which has progressed from the

notion of passive, mechanistic reactors to the environment

to active, organismic actors in the environment. While the

newer theories are not without their problems, substantial

progress towards a more complete understanding of sex role

development has occurred.

Differing types of empirical evidence would support the

varying theoretical positions presented. Research evidence

documenting the importance of early life experiences and the

family environment would support the psychoanalytic and

cognitive-developmental theoretical frameworks. Social

learning theory would best be supported by experimental

evidence. Laboratory experiments of reinforcement or

imitation would demonstrate the processes of social

learning. Gender schema theory would also be supported

through continued laboratory work. Experimental research in

the gender schematic framework demonstrates the functioning

of schemas. Little work has been done concerning the

developmental influences of gender schemas, although the

assumption made by both sets of authors in this area

suggests the importance of early childhood experiences in

forming and delineating gender schemas. Social-role theory

is best supported by either a comparison of the influences

of social roles, or by an examination of the relative

influence of current social roles versus previous family

life experiences. Social role theory assumes that current

social roles outweigh the influence of past family

experiences. Thus the differing theoretical approaches

require differing styles of empirical support.

Family Background

Socialization of sex roles as previously noted begins

very early in life. The family, with its own structure, and

value system and the personal experiences one gains therein,

in all likelihood plays the most important role in shaping a

child's sex role. Many variables have been studied

regarding family make-up and sex role development. For

example, Canter and Ageton (1984) found that whites were

less traditional than other races while Swatos and McCauley

(1984) found that the "working class" was more traditional

than those of other socioeconomic status groups. More

specific family background variables were studied by Bell et

al. (1985). Results of their work indicated that family

size, ordinal position, sibling gender, and sibling spacing

did not impact sex role development. The only variable of

importance was gender of the subject, with females being

less traditional than males. There are several areas of

family influence on sex role development that have been

empirically evaluated and found to be important. The areas

to be specifically investigated in this section are 1)

education, 2) maternal employment, 3) divorce, and 4)

religion. Each of these areas will be summarized below.

The area of education has been heavily researched and

can be subdivided into three areas 1) parental education

level, 2) personal education level, and 3) personal

educational goals. Increased education promotes greater

opportunities in terms of employment, advancement, and in

general lifestyle options. Therefore, education is thought

to have a liberalizing effect, and the greater the

educational level the greater the liberalization.

Maternal employment is expected to influence sex role

development due to the family role restructuring that takes

place when the mother works outside the home. In this case

both parents are viewed in an instrumental light since both

are out in the "real world" working and bringing home an

income. At the same time, it is likely that the father will

also take over some more of the domestic responsibilities

and hence he will be viewed in a more feminine role thus

reducing the sex role stereotype of division of labor in the

home (Baruch & Barnett, 1986).


Closely related to the situation of maternal employment

is the family restructuring involved after a divorce.

Divorce forces the custodial parent into role restructuring

since they are now "required" to do two roles, that of both

instrumental father and expressive mother. It should be

noted that while mothers most often are granted custody of

the dependent children, in approximately 11% of the cases

fathers are awarded custody (U.S. Bureau of Census, 1987).

Furthermore, the custodial parent role restructuring is

often temporary since 66% of divorced parents remarry (Glick

& Lin, 1987). In the majority of cases (89%), the effect of

divorce is best explained in terms of loss of the father

role. Sons do not have a father with which to identify and

daughters do not have a father with which they can learn

interpersonal skills that will benefit them socially in

adolescence. Again however, loss of the father figure is

frequently temporary. Furthermore the time frame for

remarriage is actually quite short, with about 50% of

divorced parents remarrying within three years (London &

Wilson, 1988). Whether the mother is awarded custody or the

father is awarded custody, the structure of a divorced

family differs from that of an intact family; hence, an

impact on sex role development may be expected.

The last area of investigation is that of religion.

The family unit is the main value dispensing system for

children and many of these values are taught within the


framework of religious instruction. In fact, some (Ruether,

1974; Wilson, 1978) have suggested that religion is probably

the most important influence in shaping and maintaining sex

roles, especially female sex roles which are typically

subordinate to that of the males. Since an individual's

main religious training comes from the family, the type of

religion (denomination) as well as the amount of religious

activity engaged in is an important area of study.

Thus, studying familial antecedents in sex role

development appears to be important for a complete

understanding of the developmental process. Huston (1983)

suggested that there are four potential sources of parental

influence. The first was that of role modeling, which is of

particular importance when investigating divorced families

and families in which both parents are employed. Huston

also suggested that parental attitudes are important. These

attitudes would be shaped themselves based in part on the

parents' educational level and also on the parents'

religiosity. Therefore these are prominent areas of study.

Parental personality differences could also be a source of

differential socialization. For example, the employed

professional mother is likely to differ in personality

structure from the voluntarily homebound housewife.

Huston's last idea was that childrearing practices will

differ between families and these differences may be a

reflection of the family structure. Again, an employed or


divorced mother will likely have different expectations and

place different demands on her children in terms of

independence or household responsibilities. The effects of

family structure and parental roles on socialization of sex

roles are, therefore, of prime importance, especially in the

areas of education, maternal employment, divorce, and



The impact of education on sex role development has

several aspects. Education can be analyzed as potential

effects due to the level of education achieved by the

subject's parents. Education can be investigated from the

subject's own achievement level or from the subject's

projection of the final level of education they wish to

achieve. All three of these aspects have been investigated

regarding their particular impact on sex role development.

Parental educational level is projected to affect a

child's sex role according to Brogan and Kutner's (1976)

"educational attainment model." Brogan and Kutner

hypothesized that highly educated mothers (they do not

project a paternal effect) will transmit a more

nontraditional sex role to their children. Thus, education

must first affect the mother's sex role and then after

incorporating this nontraditional sex role into her

attitudes and actions, the mother transmits these beliefs

and behaviors to her children.

The results of investigations on the impact of parental

education on children's sex roles have been equivocal. This

is in part due to the differing measures of sex roles that

have been employed, along with varying classifications of

parental educational level. Also, differing types of

statistical analyses have been employed and thus again the

results have varied from study to study.

The general conclusion of studies employing subjects

under 18 years of age supported the influence of higher

parental education level on the formation of egalitarian sex

roles. Richmond-Abbott (1984) employed subjects between the

ages of eight and fourteen and found that both paternal and

maternal educational levels were related to sex roles as

defined by the Traditional Family Ideology Scale (TFI).

This scale reflects attitudes towards male and female roles

and responsibilities. A longitudinal study by Thornton et

al. (1983) tracked children from birth through age 18 and

also concluded that both paternal and maternal educational

levels were related to the sex roles of their children.

Kiecolt and Acock (1988) studied 16-year-old subjects and

found that maternal educational level was related to sex

roles, for female children. Hence the link between parental

educational level and children's sex roles, is inconsistent.

Studies employing college aged subjects are also

inconclusive. Employing only female subjects, both Brogan

and Kutner (1976) and Tallichet and Willits (1986) found

that maternal educational level was related to the

daughter's sex role such that the higher the level of

education, the more egalitarian the sex role. Meier (1972)

used both male and female college students as subjects and

found the effect of maternal education held. Meier also

tested for the effects of paternal education, but was unable

to find a relationship for either sons or daughters.

Studies done by Tomeh (1979) and Zuckerman (1981) found

little relationship between parental educational level and

the sex role of college aged students. At best, Zuckerman

reported that paternal educational level in combination with

a Jewish upbringing accounted for 4% of the variance in the

daughter's sex role score (AWS). Tomeh found no statistical

relationship between parental education and any of the four

sex role measures tested. Thus, it is possible to once

again find studies that support a relationship between

education of parents and sex roles of their children, and at

the same time find studies that show no evidence of a link

between these two variables.

A closer comparison of studies utilizing college

students as subjects is needed in order to try and explain

this lack of consensus. It is interesting to note that

while all five of these studies used college students as

their sample, they each employed different dependent

variables. Three of the studies employed standard sex role

measures (AWS, FSE, and the SROS) while the other two

studies used listings of questions that had not been

standardized. Most interesting, however, is that the two

studies concluding no relationship (Tomeh, 1979; Zuckerman,

1981) employed the same statistical technique, a multiple

classification analysis. The research studies that found a

relationship (Brogan & Kutner, 1976; Meier, 1972; Tallichet

& Willits, 1986) all utilized variations of analysis of

variance techniques. So, while there appear to be

differences between groups in terms of parental educational

effects, these effects are not strong enough to predict

variation in sex role scores.

Another aspect of education and its potential impact on

sex role development comes from the arena of personal

experience. College attendance in particular has been

identified as affecting a person's attitudes and values. In

general, the impact of college is viewed as a "liberalizing

factor" in a person's general outlook, especially where

social issues are concerned (Funk & Willits, 1987).

Individuals who attend college, often even for only a
year, have been found to differ significantly from
their noncollege counterparts in areas such as
attitudes toward religion and traditional gender roles.
Overall, college attendees are less inclined to accept
the traditional roles of women in society and more
likely to approve of women working and participating in
public affairs. (p. 224)

The research in the area of personal experience then,

focuses on the level of education achieved, or more simply,

the year in school of the subject and compares sex role

beliefs to subjects at other educational levels.

The conclusions of studies dealing with personal

educational achievement have in general reached a consensus

that college attendance is associated with an increase in

sex role egalitarianism. The strongest support for this

statement comes from studies by Dreyer, Woods and James

(1981), Mason and Bumpass (1975), and Morgan and Walker

(1983). All three of these studies employed regression

analysis. The results showed that educational level was the

number one factor for Dreyer et al., and Mason and Bumpass,

while it was the number two factor (behind age) for Morgan

et al. All three studies employed different measures of sex

roles and their subject populations were adult females.

However, the results of the Dreyer et al. study were

exceptionally strong, with education alone accounting for

61% of the variance.

More support came from the work of Funk and Willits

(1987) who followed subjects longitudinally for over ten

years. Their male and female subjects reported less sex

role traditionalism the longer they stayed in college. Also

of note from Funk and Willits is that the subject's sex role

attitudes in high school appeared to predict college

attendance. Those in high school with the lowest (most

traditional) scores, were the subjects that did not attend

college, while those with the highest scores did attend

college. According to the authors however, it was college

attendance per se that was related to sex role attitudes

based on analyses that controlled for initial attitude

level. Renzetti (1987) also found that among college women,

those listed as juniors or seniors differed on attitudes

towards feminism, traditional sex role attitudes and level

of support for the feminist movement compared to freshman

and sophomore subjects.

The most equivocal results concerning the effects of

educational level on personal sex roles comes from studies

employing the BSRI as the dependent measure. Martin and

Light (1984) found that among college males and females the

percentage of students in the masculine category increases

while the percentage of students in the feminine category

decreases over the college years. Lyons and Green (1988)

were only able to support the decline in the number of

students in the feminine category. However, Lyons and Green

only used female subjects. Thus, interpretation of these

two studies must be done carefully. Overall it appears that

there is fairly consistent and strong support for the idea

that additional college experience is related to more

egalitarian sex roles.

A final area of education is the anticipated level of

educational accomplishments for the subject. As suggested

by Morgan and Walker (1983),

women who are the most supportive of sex role
egalitarianism should be those who have alternative
rewards available to them. .and are likely to be
young, well educated, employed, and have feelings of
personal competence. (p. 148)

They proposed that this hypothesis was based in exchange

theory. According to exchange theory, people make decisions

or change attitudes based on cost-benefit considerations.

Situation are judged based upon the amount of rewards

offered or the number of costs encountered. A decision is

then based on the bottom line of the mental tally sheet.

Hence, additional college achievement that is looked upon as

increasing a person's potential rewards will be viewed

positively, while the costs of the alternative roles

available (e.g., wife/mother) will also be taken into

account, valued or devalued, and a decision reached. If the

value of the career wife/mother is found to hold more

rewards, and the role of wife/mother is found to hold more

costs, it is anticipated that the person would opt for

higher rewards, and thus higher levels of education (which

increase the changes of higher occupational skills as well).

Therefore, anticipated educational level is hypothesized to

be related to personal sex roles.

The research available indicates support for the

exchange theory prediction of egalitarian attitudes being

associated with higher levels of anticipated educational

achievement. Alper (1973) found that women who were labeled

as more feminine on the Wellesley Role-Orientation Scale

(WROS) were less likely to be planning a graduate school

education or a career. Martin and Light (1984) found that

for both male and female college students, scores on the

BSRI indicated that students classified as more masculine

were more likely to desire higher levels of education after

completing college. This finding is also supported by the

work of Zuckerman (1981) who found that for both males and

females, those that reported anticipating higher levels of

education (also higher career goals) scored higher on the

AWS, which reflects attitudes towards feminism. Hence, all

research to date supports the notion that educational goals

are related to personal sex roles.

All three aspects of education, parental educational

levels, personal educational levels, and anticipated

educational goals are apparently related to personal sex

roles. Overall findings for the area of education and sex

roles are generally supportive, but several problems were

also noted and discussed. Some of the findings are

inconsistent, especially in the area of parental educational

effects. Additional work in all three areas is warranted in

order to clarify the conclusions.

Maternal Employment

The number of married women with children both

preschool and school age that have entered the work force

over the last several decades has been dramatic. Recent

estimates (U.S. Department of Labor, 1987) indicate that at

least half of today's mothers are in the labor force. The

employment of married women with children results in effects

similar to divorced households in that both households may

require family role restructuring. Because of this

potential role restructuring, children's perceptions of sex

roles could be affected. Broverman et al. (1972) suggested

that the perception of sex roles is

influenced by the degree of actual role differentiation
that one experiences in one's own family. Maternal
employment status appears to be central to the role
differentiation that occurs between parents. If the
father is employed outside the home while the mother
remains a full-time homemaker, their roles tend to be
clearly polarized for the child. But if both parents
are employed outside the home, their roles are more
likely to be perceived as similar. (p. 73)

Employment of the mother necessarily implies less time for

domestic jobs. Hence the father will in all likelihood

increase his share of the domestic responsibilities and thus

alter the traditional masculine sex role (Marantz &

Mansfield, 1977). However, empirical research does not

substantially support this position. Time estimates of

husbands' domestic work when their wives are employed full-

time ranges from four to six hours per week, which only

represents 14 percent of the total time spent on household


chores (Bernardo, Shehan & Leslie, 1987). Furthermore, the

increase in husbands' household work was estimated to have

increased only 1.5 hours when the wife went to work (Barnett

and Baruch, 1983). Hence while a mothers entry into the

work force is likely to affect perceptions of women's sex

roles, due to her role restructuring, the perception of male

sex roles may not be substantially altered.

Due to the apparent importance and potential impact of

the effects of maternal employment on sex role formation in

childhood, it is not surprising that there is a fair amount

of literature in this area. And, while the majority of

studies agree that maternal employment is associated with

less traditional sex roles (Hoffman, 1974), there are

nonetheless studies that report no effects.

Kiecolt and Acock (1988) suggest three reasons for

these discrepant findings. The first reason is that studies

vary in their choice of dependent variables and hence may

not be comparable. The second reason is due to differing

subject populations. While most studies utilized college

freshman, this area studied children as young as two due to

the early formation of sex roles. Some research had

suggested that sex roles are not static, but rather the

amount of traditionalism exhibited varies with the age of

the subject (Urberg, 1979). Therefore, comparisons between

studies without accounting for subject age may produce

apparently contradictory findings.

Finally, the work in this area has not paid enough

attention to the timing of maternal employment. Studies

frequently simply dichotomize maternal employment into "Yes,

mom works" or "No, mom doesn't work." How long the mother

has been in the labor force is not accounted for nor is the

age of the subject when the mother joined/returned to the

labor force. Researchers have suggested that there are two

important time frames for sex role formation. The first is

before the age of six, when sex role identity is forming

(Weinraub et al. 1984). The second is during adolescence,

when the self-concept is being restructured (Streitmatter,

1985). Hence, the timing of maternal employment could be

crucial in interpreting the findings of each study.

Research supporting the impact of maternal employment

on the formation of less traditional sex roles is abundant.

Beginning with subjects as young as two and three year of

age, Weinraub, et al. (1984) found that their subjects were

less aware of sex differences in toy preferences the more

hours the mother worked outside the home. Female subjects

between the ages of five and eleven were found to assign

personality traits in a more egalitarian manner if their

mother worked (Marantz & Mansfield, 1977). Kiecolt and

Acock (1988) found that female subjects aged six and sixteen

devalued the traditional female sex role if their mothers

worked. Additionally, the teenaged females believed in

greater equality for women. Ten-year-old boys and girls

with employed mothers were found to differentiate less

between the sex roles of males and females in terms of

traits, behavior, jobs and authority relations (Gold &

Andres, 1978). Adolescents between 14 and 18 that came from

dual-career families had more egalitarian views than

adolescents from either single-earner or dual-worker

families. This finding was echoed by Gardner and LaBrecque

(1986) whose subjects were 17 to 18 years old. Hence during

childhood and adolescence there is evidence that maternal

employment is related to sex role development.

Employing college students as subjects confirms the

finding of less traditional sex roles when mothers are

employed. Vogel, Broverman, Broverman, Clarkson, and

Rosenkrantz (1970), and Meier (1972), found that males and

females held more traditional sex roles if their mothers

were not employed. Tomeh (1979), repeated these

conclusions, but only for male subjects while Broverman et

al. (1972) concluded that female subjects were more affected

than males. Thus there appears to be some discrepancy in

the findings of older subjects regarding the impact of

maternal employment. Again the resolution of this

discrepancy could be determined based on the dependent

variables employed and investigation of the timing of

maternal employment as suggested by Kiecolt and Acock


There is also research which does not support the

existence of a relationship between personal sex role and

maternal employment. Using college students as subjects,

Keith (1988) reported that maternal employment did not

directly impact the sex roles of students. Keith found that

maternal employment was not a factor of any great

consequence in predicting personal sex role scores based

upon multiple regression analysis. Weeks, Wise, and Duncan

(1984) also concluded that the sex roles of daughters are

not related to their mothers' employment. Analyses of

variance conducted on the daughter's sex role attitude

scores indicated no significant difference between daughters

of homemakers, homemaker plus part-time employment,

homemaker plus full-time employment, and career woman plus

some homemaking. Therefore, there are studies that show no

differences based on maternal employment, as well as studies

that show effects for both males and females. Due to the

complications involved in operationalizing the variable of

maternal employment, continued work in this area is



The traditional family structure consisting of an

employed husband, a housewife, and children appears to be a

thing of the past. According to the U.S. Bureau of Census

(1987), the American divorce rate tripled between 1960 and

1980. During the 1980s the divorce rate then stabilized

resulting in over one million new children experiencing

parental divorce or separation each year. Currently 13

million or 23% of American children live in single-parent

households. Of this group, 89% live with their mothers.

Yet these statistics do not tell the whole story.

Approximately 66% of divorced parents remarry (Glick & Lin,

1987). Remarriage of divorced parents has resulted in an

upsurge in the number of step-families. Current estimates

(Glick, 1989) suggest that over 17% of American households

are comprised of step-families or blended families. The new

wave of family structuring be it single-parent, step-parent,

or blended family, will certainly impact the development of

personal sex roles. Yet there is to date very little

empirical work in the area of the effects of divorce on sex

role development.

Two major theories have attempted to explain the

potential impact of divorce on sex role development. The

hypothesis behind the research in this area is that divorce

will lead to more egalitarian sex roles for all of those

involved, parents and children. The first theoretical

explanation comes from social learning theory and has been

labeled the "father-absence hypothesis" (McLanahan, 1985).

Social learning theory postulated that sex roles are learned

through identification or modeling of the same sex parent.

In most cases (89%) when a divorce occurs the mother

receives custody of the children (U.S. Bureau of Census,

1987). If the children are still quite young (under six

years of age), it has been suggested and empirically

supported (Hetherington, 1966) that boys will become more

dependent, less aggressive and less masculine in their sex

roles due to the loss of a male role model. Girls, on the

other hand, tend to be influenced very little due to the

continued presence of their mothers. However, there may be

a delayed effect in terms of the way she relates to males

later in life (Hetherington, 1972). In addition, "fathers

are particularly important for inculcating gender-role

orientations because, compared to mothers, they vary their

behavior more by sex of child and urge greater conformity to

traditional gender roles" (Kiecolt & Acock, 1988, p. 710).

The father-absence hypothesis would suggest that by

default researchers should investigate the sex role

attitudes of the mothers since for the most part they will

be in charge of the socialization of the children after a

divorce (MacKinnon, Stoneman & Brody, 1984). Again,

however, the notion of father-absence needs to be tempered.

As stated earlier, most divorcees remarry (Glick & Lin,

1987). Further, remarriage occurs rather quickly.

Approximately 50% remarry within three years (London &

Wilson, 1988). And father-absence effects could be lessened

if the divorced mother has male figures such as boyfriends

or fathers who influence the children' development. Thus,

the father-absence hypothesis may be overemphasized in the


The second major theoretical position comes from

sociological role theory and has been labeled the role

restructuring hypothesis (Brown & Foye, 1984). Due to the

divorce there is a different family structure which requires

a reassignment of duties and responsibilities. Children as

well as the custodial adult will be required to take on more

household tasks. The result is that the children see as

well as participate in a variety of tasks which do not fit

the traditional sex role stereotypes for behavior. As such,

the parent serves as a role model displaying nontraditional

behaviors which the children will incorporate into their own

sex roles. Hence the custodial parent who undertakes both

masculine and feminine tasks should serve to lessen the

traditional sex role stereotypes of the children.

An examination of the available evidence on the effects

of divorce on sex role development should help to clarify

which of these two theories is more applicable.

Unfortunately, there is very little evidence from which to

draw any firm conclusion in this area. In general, the

evidence to date appears to point to a weak but generally

positive correlation between divorce and sex role

egalitarianism. A very small but positive relationship was

found by Thornton and Freedman (1979) between divorce and

sex role egalitarianism. Unfortunately, a longitudinal


follow-up of this data (Thornton et al., 1983) completed 18

years after the original study was unable to distinguish

whether "divorce had no influence on sex-role attitudes or

if its effects were countervailing" (p. 224).

In addition to this methodological problem, it was

suggested by MacKinnon et al. (1984) that the personal sex

role change to egalitarian was not because of the divorce

per se, but rather because of the additional roles that the

mother must take on, most dramatically that of family

breadwinner. Their research then attempted to sort out the

effects of divorce from the effects of maternal employment.

Findings from this work indicated that the most important

factor for children ages three to six was marital status.

Children from divorced homes with working mothers had less

awareness of traditional sex role behaviors for both their

own sex and for the other sex, when compared to children

with still married working mothers. Hence it appears that

for young children divorce does play a factor in sex role


Adolescents were studied by Kiecolt and Acock (1988) in

order to determine the effects of family structure on sex

role attitudes. Their results showed a mixture of effects.

Adolescents living with mothers following a divorce were

found to have more egalitarian attitudes toward women in

politics. Females in stepfamilies also hold these beliefs,

while males in stepfamilies have more traditional attitudes.

A study by Richmond-Abbott (1984) found that both

mothers and children in divorced homes held very liberal sex

role attitudes. However, there was no comparison group to

detail the magnitude of these liberal attitudes. It is

interesting to note, however, that while attitudes were

labeled as liberal, in-home behaviors were typically sex

role stereotyped. Richmond-Abbott found that 90% of the

toys given to the children were sex typed, and the chores

given to the children were also typically sex typed. Thus,

while attitudes are becoming liberalized, behaviors remain


Overall conclusions for the area of divorce are

equivocal. However, there does seem to be a weak but

positive relationship between sex role egalitarianism and

the divorce experience. In terms of theoretical

interpretations it is likely that neither of the previously

discussed theories is sufficient alone, but rather that both

are correct since they are not mutually exclusive. The work

of Kiecolt and Acock (1988) appears to support the

father-absence hypothesis since males and females were

differentially affected depending on whether their family

structure was divorced and mother-headed or step-parented.

The results of MacKinnon et al. (1984) also supported the

father-absence hypothesis based on the fact that the

children were less aware of appropriate sex role

stereotypes. Boys were less aware of their own sex roles

and the girls were less aware of male sex roles, thus

implying the necessity of a father to indoctrinate the

children into the appropriate or at least stereotypical


The work of Richmond-Abbott (1984) supported the role

restructuring hypothesis as the sex roles of both mothers

and children were quite liberal, especially in the area of

family roles. Also, no detrimental effects for the male

children were noted which is a main feature of the

father-absence hypothesis. Therefore, support for both

theoretical viewpoints can be found even though the

literature in this area is scant. Due to the shortage of

studies in the area of divorce and the tremendous increase

in divorcing families it seems imperative that researchers

continue to explore the effects of this phenomenon on the

personal sex roles of both children and adults.


The empirical work on the relationship between personal

sex roles and the impact of religion is also scant.

Frequently studies refer to the impact of religion as a side

issue, only one of many variables that are investigated

within the research framework. Yet, recent work in the area

has shown that religion is of substantial importance in

accounting for the variation in sex role attitude scores

(Brinkerhoff & MacKie, 1985). In fact, both Wilson (1978)


and Ruether (1974) claim that religion is probably the most

important influence in shaping and in enforcing sex roles,

especially female sex roles. Therefore, a review of the

available current research concerning the impact of religion

on personal sex roles will be undertaken.

In order to understand the potential impact of religion

on personal sex roles religion must be defined. A general

definition is needed as well as an examination of the

operational definitions employed in different studies.

Religion has been defined as "a combination of beliefs,

values, and behaviors which provides an overall worldview"

(Brinkerhoff & MacKie, 1985, p. 416). Employing this rather

broad definition it becomes possible to see how religion may

be drawn into several aspects of social life. Since

religion affects values, beliefs, and behaviors, it has been

suggested by Mason and Bumpass (1975) that membership in

groups espousing social values is likely to affect sex roles

more than purely demographic variables such as marital

status. Thus, the value structure of the religious

denomination can be expected to affect the sex role of the


It is also necessary to examine operational definitions

of religion in order to comprehend the research and its

various findings. One of the functions of operational

definitions is to untangle apparently contradictory research

results. When investigations conclude "yes," "no," and

"sometimes, sort of" it is often helpful to go back and

examine how the variable was operationalized. Within the

realm of religion there appears to be two commonly used

operational definitions 1) religion as defined by the

denomination to which one belongs, and 2) religion defined

as religiosity indicating the amount of religious practice

in which one engages, or self-perceived religiosity defined

by such questions as "I consider myself to be a very

religious person and I attempt to live my religious

conviction daily," or "I do not believe that there is any

truth in religion." With such differing aspects of religion

being analyzed it is not surprising that the literature is

not in complete agreement as to the role that religion plays

in sex role formation.

One of the earliest empirical investigations of sex

roles and religion was completed by Meier (1972). Meier

employed two measures of religion, denomination and

religiosity (church attendance). Overall, subjects who

reported "None" as religious denomination were rated as more

egalitarian in their sex roles than those who reported a

denomination. Further, subjects that attended church

frequently scored considerably lower and were labeled as

more traditional in their sex roles, especially female


Since Meier's study (1972), several other studies have

incorporated the use of denomination with varying results.

No differences at all between religious denominations were

found in the works of Mason and Bumpass (1975) and Tomeh

(1979). Subjects indicating "None" were found to be the

most egalitarian in studies by Brogan and Kutner (1976) and

Brinkerhoff and MacKie (1985). The most revealing work in

this area, however, indicated that there may be a shift in

the differential impact of religious denomination over the

last 20 years. Thornton et al. (1983) conducted a

longitudinal study of sex roles and found that while

denomination was not considered an important determinant in

1962, the retesting of subjects in 1977 and 1980 indicated

that specific denominations have moved in separate

directions. Those espousing Catholicism have become more

egalitarian, while those belonging to more Fundamental

Protestant denominations have become more traditional in

their sex roles. Thus, the impact of religious denomination

on sex roles is not static. This is most likely due to

changing church doctrine which in turn influences the values

and beliefs of the church members. Therefore, the area of

religious denomination continues to be an area of interest

for sex role researchers due to its changing influence

through the years.

Another major operational definition of religion has

been religiosity or the amount of time spent in religious

activities. All of the studies that have employed this

definition of religion have found significant results. As

previously stated, Meier (1972) found that increased

attendance decreased egalitarian scores. Swatos and

McCauley (1984) found the most pronounced effect for female

subjects that attended at least weekly. Scores for these

subjects indicated a much more traditional sex role, whereas

male subjects appeared unaffected by the amount of religious

practice. Finally, Thornton et al. (1983) found that over

18 years as church attendance increased respondents showed

less movement towards egalitarianism, while respondents

reporting infrequent attendance showed more movement towards

egalitarianism. There appears to be a consensus in the

literature concerning religiosity such that increased

religious activity is related to increased traditionalism.

The other approach operationalizing religion as

religiosity is concerned with finding out how important

religion is to the subject. For example, Swatos and

McCauley (1984) labeled their religiosity variable as

"self-perceived religiosity." Subjects answered four

questions regarding their own perceptions of the importance

of religion in their own life. Swatos and McCauley's

results showed that for women, higher self-perceived

religiosity was related to higher sex role traditionalism.

Mason and Bumpass (1975) tested the same concept employing

only one question "Quite apart from attending religious

services, how important would you say religion is to


you---very important, fairly important, or unimportant?" (p.

1217). These authors concluded that religiosity had a

modest effect on sex role traditionalism. Again,

religiosity defined as self-perceived religiosity affects

personal sex roles, with those that indicate religion as

more important showing more traditional sex roles.

There is also a miscellaneous category of religiosity.

Larsen and Long (1988) operationalized religion in terms of

religious orthodoxy and religious fanaticism. Both of these

definitions proved to be positively related to traditional

sex roles. The most hybrid design of religiosity came from

the recent work of Brinkerhoff and MacKie (1985). These

authors combined the ideas of religious belief and religious

behavior into one category of religiosity. Brinkerhoff and

MacKie found a very strong relationship between religiosity

and sex role traditionalism with religiosity accounting for

22.3% of the variance in their sex role measure. Hence

again, religion operationalized as religiosity even in a

variety of ways seems to exhibit a very robust effect on

personal sex roles.

Overall, the findings of the relationship between

religion and personal sex roles appears to be very strong.

This relationship holds even though the subject populations

vary between studies, the measures of sex role vary, and the

time period of the studies varies from the early 1970's to

the present. Religiosity appears to be the most powerful of

the potential operational definitions of religion, while

denomination is the most troublesome.

It has been suggested that there are several

explanations for the difficulty involved with the use of a

denominational definition (Brinkerhoff & MacKie, 1985).

Studies will generally vary in their specificity of

denomination from broad categories such as Protestant,

Catholic, Jewish, to very specific categories such as

Baptist, Orthodox Jewish, Greek Catholic. The lumping

together of several different denominations into a few

categories invites substantial variation into the results

which may render the categorizations useless. Also, it is

important to clarify whether the response to denomination is

based on current practices or childhood rearing. One may

have been raised a Catholic for 18 years, but may not be

currently identifying with the Catholic Church.

Finally it should be noted that church doctrine is not

static. New interpretations that affect member's beliefs

and values are mandated by the church hierarchy

continuously. One of the most controversial doctrine

changes has a direct effect on personal sex roles. Some

denominations have decreed that women may hold positions of

authority in the church and may even become ministers or

priests, while other denominations have tenaciously held to

their positions of women as subordinate to men in all

aspects of religion (as well as other social roles). Thus

the effect of denomination offers many difficulties for

researchers, especially in light of the changing church

doctrines. Even with the difficulties of, and many

definitions for religion, it appears that the area of

religion and its effect on personal sex roles is still an

important research domain due to the robust effects the

research has noted to date.


The family is the major socialization agent in a

child's life. The family background will delineate the

roles for each of the family's members. Therefore, family

roles will be affected based on whether the family is

intact, divorced, or blended. Additionally, the mother's

employment status will affect not only her domestic roles,

but those of other family members as well. Family members'

roles are also a function of other socializing experiences

such as level of educational achievement and religious

activity. The variation in roles displayed by family

members serves as models of different personal sex roles.

It is not surprising that children from homes with

nontraditional families are found to acquire more

egalitarian personal sex roles. While the results from the

areas of education, maternal employment, divorce, and

religion are not completely consistent, there is a

considerable body of evidence that each area affects the sex

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs