science article - University of Southern Indiana

vainclamInternet and Web Development

Dec 14, 2013 (3 years and 6 months ago)

104 views







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

1


Running head: GENDER ROLES, SELF
-
ESTEEM, AND
PROBLEM
-
SOLVING






Problem
-
Solving in
the
Natural
Scienc
e
s

and Early Adolescent Females'

Gender Roles


and
Self
-
Esteem
:
A
Qualitative and Q
uantitative

Analysis from an Ecological Perspectiv
e


Michael Slavkin



University of Southern Indiana


2000



























Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

2


For further information on this study, please contact Michael Slavkin, School of Education and
Human Services, University of Southern Indiana, 8600 Universit
y Boulevard, Evansville, IN
47712, or contact at slavkin@prodigy.net.




























Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

3


Abstract


The purpose of this study was to examine the relations between gender roles, self
-
esteem, and
the
problem
-
solving

abilities
of early adolescent females whi
le participating in
natural
science
-
related activities. Bronfenbrenner’s human ecology model (1979) and Barker’s behavior setting
theory (Barker and associates, 1978; Schoggen, 1989) were used to assess how environmental
contexts related to
problem solving

in scientific contexts. These models also provided improved
methodology and increased understanding of these constructs when compared to prior research.
Early adolescent girls' gender roles and self
-
esteem were found to relate to differences in
problem solving

in science
-
related groups. Specifically, early adolescent girls’ gender roles were
associated with levels of verbal expression, expression of positive affect, dominance, and
supportive behavior during science experiments. Also, levels of
early adolescent girls’ self
-
esteem were related to verbal expression and dominance in peer groups. Girls with high self
-
esteem also were more verbally expressive and had higher levels of dominance during science
experiments. However, group behavior of ear
ly adolescent females' also was defined by
environmental contexts, and not governed solely by the personality of participants. A discussion
for the examination of environmental factors when assessing early adolescent females’ gender
roles and self
-
esteem f
ollows this discussion.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

4


Problem
-
Solving in
the
Natural
Scienc
e
s

and Early Adolescent Females'

Gender Roles and
Self
-
Esteem
:
A
Qualitative an
d Q
uantitative

Analysis from an Ecological Perspectiv
e

How are the gender roles, self
-
esteem and
problem solving

of early adolescent females
related to participation with
natural
science experiments? The purpose of this investig
ation was
to
qualitative and quantitative
ly examine how early adolescent females’ perceptions of self
(gender roles and self
-
est
eem) related to
problem solving

while participating in scientific
endeavors. While aspects of personality, such as gender roles and self
-
esteem, have been found
to impact how juveniles behave in peer groups (Bancroft, 1990; Huston, 1983; Huston &

Alvarez, 1990; Eccles, Flanagan, Lord, & Midgley, 1996), few researchers have actually
observed the interplay between self
-
concepts,
problem
-
solving
, and the surrounding
environment. Without an examination of the environment in which
problem solving

occur
s
,
researchers might have a limited understanding of the relationship between personal
characteristics and group phenomena.

Theoretical Framework

Bronfenbrenner’s Ecological Theory



Bronfenbrenner’s human ecology asserts that the physic
al environment and psychological
environment affect both individual and
problem solving
. Bronfenbrenner (1972, 1979) asserts
that human behaviors and belief systems are organized as facets of the environment, which are
further divided into separa
te subunits, such as cultures, social groups, and peer systems. To
understand subunits like
problem
-
solving
, researchers should identify characteristics of these
units, as well as the environmental systems with which they interact (Bechtel, 1977;

Bechtel,
1997; Bechtel, Marans, & Michelson, 1987; Bronfenbrenner, 1972; Bronfenbrenner, 1979;






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

5


Wohlwill & van Vliet, 1985). Human ecologists divide the psychological environment into four
levels: the microsystem, the mesosystem, the exosystem, and the mac
rosystem.


The microsystem is defined as the everyday environment with which the individual
interacts. The microsystem could include an individual interaction with a community member, a
family unit, a peer group, or a specific neighborhood. The mesosystem
constitutes the
relationships between units of the environment (e.g.
, students working on a science
activity
,
citizens within a community, citizens and established community groups). The exosystem is
represented by social institutions, such as media, organized rel
igious institutions, and local
governments. The exosystem is the level at which society influences what occurs within a
particular group. The macrosystem refers to the cultural and societal influences on an individual,
and is believed to be the most global

level of the environment (Bronfenbrenner, 1979).

Environmental psychology asserts that as an individual’s behavior changes, changes
occur simultaneously within the surrounding environmental systems. According to
Bronfenbrenner and other ecologists (see Br
onfenbrenner, 1972;
Bechtel et al., 1987 for a
review), failure to examine these environmental units leads to a limited understanding of the
organism (Bronfenbrenner, 1979). As such it is critical to refer to the physical and psychological
environment when

examining the roles and identities of individuals, groups, communities, and
cultures (Bechtel, 1977; Wohwill & van Vliet, 1985).

The Microsystem Examined: Barker’s Behavior Setting Theory


Roger Barker’s behavior setting theory divides the microsystem int
o environmental
subunits, known as behavior settings. Behavior settings are the smallest units of the physical
environment that can be used to predict behavior. An examination of these subunits is required to
better comprehend the microsystem’s effect on i
ndividual behaviors (Barker, Kounin, & Wright,






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

6


1943; Barker and Associates, 1978; Schoggen, 1989). Barker asserts that human action is
organized into specific behaviors based on traditions subsumed within these settings. Behavior
setting theory emphasizes
the transactional relationship between a setting, its participants, and
the standard patterns of behavior for that environment (Barker & Wright, 1955; Bechtel et al.,
1987).


Examinations of behavior settings are performed at the molar level (the level of

interaction between the individual and the environment, similar to Bronfenbrenner’s
microsystem), which is divided into three units: time, place, and object props (Barker & Wright,
1955; Bechtel et al., 1987; Schoggen, 1989). The time that an interaction

occurs and the
environmental place of its occurrence are as important as the object props (such as people,
furniture,
science
equipment
,
sources of illumination, etc.) found in the setting (Barker and
associates, 1978; Barker and Wright, 1955; Schoggen, 1989). Most theories

of ecological or
environmental psychology differentiate between a setting’s participants (e.g., students, teachers,
administrators) and objects (e.g., chairs, desks,
Bunsen burner
s
,
che
micals,
temperature), but
behavior setting theory treats them as equal facets of the en
vironment. Barker suggests that each
object prop is essential in defining both the behavior setting and its standing patterns of behavior
(see Schoggen, 1989).

In order to examine psychological factors at any level of the environmental system,
researchers
must analyze both the behavior and the environmental influences on these behaviors
(Bechtel et al., 1987). To understand behavior it should not be divorced from its environment.
The behavior setting serves as an effective way of examining environmental uni
ts of behavior at
the microsystemic (molar) level.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

7


To understand how gender roles, self
-
esteem, and
problem
-
solving

in
natural
scientific
experiments
are related, this investigation q
uantitatively
reviewed environmental characteristi
cs
at the levels of the microsystem (time, place, and object props), mesosystem (behavior setting),
and exosystem (culture of participants). The research was performed so that a clearer
understanding of the relationships between early adolescent females’ p
erceptions of self and
problem solving

could be identified. It is believed that early adolescent females' beliefs about
gender and self
-
esteem (at the microsystemic level) are displayed by
problem solving

(at the
mesosystemic level) as
parts of the community culture (at the exosystemic level).

The Microsystem: Personal Gender Roles

Personal gender roles are defined as part of an individual's belief system that importantly
influences perceptions people have of themselves and others. This

cognitive system affects how
individuals perceives themselves, family members, peers, and society (Bem, 1974, 1975, 1981;
Williams, Radin, & Allegro, 1992). Gender roles traditionally are defined as either masculine or
feminine. Whereas masculinity has be
en stereotyped as being independent and dominant,
femininity has been characterized as being emotive and sensitive (Bem, 1974; 1975).
It is
believed that personal gender roles, as one facet of the microsystem, have an impact on
behavioral patterns at the m
esosystem. Such belief systems have been identified as leading to the
sex
-
typing of fields relating to science, mathematics, and technology (Eccles, 1989; Greenfield,
1997).

Androgyny is the incorporation of both masculine and feminine characteristics (Bem
,
1978). Androgynous individuals define their gender roles as being masculine or feminine
depending on the role that is needed. In contrast, persons with undifferentiated gender roles do
not perceive themselves to be high in either traditionally masculine
or feminine characteristics.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

8


Personal gender roles and behavior at the mesosystem
. Masculine and feminine personal
gender roles, as subsets of individuals' larger belief systems, are believed to impact behaviors

and attitudes, particularly when environment
al systems are segregated based on gender

(
Abell,
1990;
Arrow, 1997; Bem, 1974; Bem, 1981
; Greenfield, 1997
). Those
subscribing to masculine
roles tend to be
organized,
assertive, dominant, a
nd focused on controlling task or group
situations
; such ch
aracteristics have been subscribed t
o the scientific model and its fields

(Bem,
1981
; Eccles, 1989
). Individuals defined as masculine typically are categorized by other
community members as being organized, dominant, and leadership
-
orien
ted, and also are thought
to be on
-
task throughout an activity more than other group members (Jordan, 1992).

Bem (1974) hypothesized that feminine
-
typed individuals are more likely than
masculine
-
typed or androgynous individuals to be concerned with the r
elationships between
themselves and others. These individuals are categorized by community members as showing a
commitment to peacekeeping and healing in groups, rather than asserting themselves (Bem,
1978; 1981).
Such characteristics have been attributed to
natural fields of science or the social
sciences, though little research supports such cla
ims (Abell, 1990
; Bailey, 1996
).
Feminine
-
typed
individuals also are though
t to be more likely than masculine
-
typed individuals to retreat from
group processes and resign from positions of dominance (Bem, 1974, 1981). However,
androgynous individuals are thought to be more flexible in their behaviors, behaving in a manner
that i
s most appropriate for that setting (Bem, 1981).


Though the literature on personal gender roles asserts that masculine
-
typed individuals
are dominant and feminine
-
typed individuals are submissive, it would seem likely that these
behaviors would be depende
nt upon the context defined by the exosystem. Such conditions






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

9


might impact whether masculine
-
typed, feminine
-
typed, or androgynous individuals involve
themselves in the group process.

The Exosystem: Society and Ideal Gender Roles


Ideal gender roles are d
efined as roles and cultural belief systems that are organized by
the divisions of gender (Grimmell & Stern, 1992; Slavkin & Stright, 2000). Ideal gender roles
are believed to impact behaviors at the exosystemic level of the environment, as they are part o
f
society's culture. Similar to personal gender roles, ideal gender roles also are defined as either
masculine or feminine. Ideal gender roles provide a cultural norm and social expectation for the
behaviors of female early adolescents. It is believed that

ideal gender roles, as a facet of the
community exosystem, have an impact on behaviors at the mesosystem and on individual beliefs
about self
at the microsystem.

As such, ideal values (such as scientists ar
e Caucasian or men, and
that women are not li
kely to enjoy the sciences) have an impact on

behaviors in schools and
attitudes of early
adolescents.


It is often thought that western societies are organized and defined by masculine values
and roles. Fem
inine roles are marginalized, and are seen as weaker subsets of our cultural belief
system. This dichotomy between masculine roles as strong and feminine roles as weak has been
shown to impact people’s behaviors, with both males and females often attemptin
g to subscribe
their behaviors to masculine
-
typed standards (Arrow, 1997; Bem, 1974, 1981). Though these
assertive, dominant, and controlling attitudes have been valued as needed characteristics for
leadership
,

group management
, and science
-
related fields

(
Eccles, 1997; Greenfield, 1996;
Jordan, 1992), it might
be that
som
e
environments exist where feminine
-
typed and androgynous
characteristics are effective for early adolescent females, if not more effective, than masculine
-
typed characteristics (Edwards, 1994; Krichevskii, 1977).







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

10



Ideal gender roles and personal gen
der roles at the microsystem: Issues with self
-
esteem.


It is widely believed that dissonance between personal values and societal ideals can
impact the self
-
evaluation of an individual (Lamke, 1982a; Muuss, 1974; Rose & Montemayor,
1994; Rosenberg, 1965).

Such dissonance is believed to affect self
-
esteem: the greater the
difference between personal and ideal roles, the lower the esteem of the individual. Self
-
esteem
is defined as an evaluative system of approval or disapproval (Burnett, Anderson, & Heppne
r,
1995; Coopersmith, 1981; Newcomb & Bagwell, 1995; Newcomb & Bagwell, 1998; Rosenberg,
1965). Rosenberg (1965) suggests that disparity between interpersonal competence (self
-
esteem,
self
-
efficacy) and identity can have implications for an individual’s in
teractions with peers.
When an early adolescent feels inferior to peers or feels incompetent in social situations, she will
be less likely to interact with peers than others with higher self
-
esteem (Cate & Sugawara, 1986;
Kurdek & Krile, 1982; Muuss, 1974)
.

Self
-
esteem and gender roles: Disparity between personal and ideal roles
. Burnett and
colleagues (1995) examined how the personal gender roles and self
-
esteem of college students
were impacted by the presence of feminine or masculine models, who served
as examples of the
community ideal gender role. The authors found that disparity between individuals’ personal
gender roles and the ideal gender roles evidenced by peers placed individuals at risk for lower
self
-
esteem (Burnett et al., 1995; also see Muuss

& Porton, 1998).

Similar results have been found for early adolescents and adolescents. In a study of 229
Caucasian early adolescents and adolescents, Cate and Sugawara (1986) found that participants
whose personal gender roles met the ideal for their c
ommunity (masculine) were more likely to
have higher levels of self
-
esteem than feminine
-
typed or androgynous participants. Lamke
(1982a, 1982b) found similar results in two samples of early adolescent and adolescent students,






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

11


with higher levels of self
-
es
teem relating to lower levels of dissonance between personal gender
roles and ideal gender roles.

However, limited discussion of environmental circum
stances reduces the generalizability
and the reliability of these results (also see Bechtel, 1977; Krasner, 1980). Also, these differences
might be the result of sampling differences, the presence of mixed
-
sex groups, or due to
mesosystem differences withi
n group activities. Environmental differences could alter the
examination of the relationships between personal gender roles and self
-
esteem.

T
he Mesosystem: Problem
-
S
olving
in

Early Adolescents

Individuals
might

alter their behaviors and attitudes by juxtaposing their actions with
those presented in peer contexts
, particularly when faced with changing ideal

values or gender
stereotyp
ed sett
i
ngs (su
c
h a
s when
young

ad
olescent

f
emales p
arti
cip
ate

in
scienc
e classes)
.
Environmental press, the pressure to conform to the contexts defined by a group, occurs when
early adolescents interact within
gender divided
group contexts. It is be
lieved that the
environmental press can impact group participants' behaviors as they interact with others (
Bailey,
1996;
Eccles, 1989;
Montemayor, Adams, & Gulotta, 1994; Muuss & Porton, 1998; Newcomb &
Bagwell, 1995; Newcomb & Bagwell, 1998; Newcomb, Bukowski, & Bagwell, 1999). Whil
e
other studies have examined environmental presses on young adolescent behaviors, none have
examined how gender roles an
d self
-
esteem

relate to
problem
-
solving

within the con
texts defined
by the
natural
science experiment
.

Previous research has examined the r
ole that peers play in impacting the behaviors of
group members (Arrow, 1997; Burnett et al., 1995; Fishbein & Imai, 1993; Huston, 1983;
Huston & Alvarez, 1990; Maccoby, 1994; Maccoby, 1996; Newcomb et al., 1999). However,
none of these studies examined th
e impact of the environment on early adolescent involvement






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

12


within groups. As a result of environmental contexts, changes in gender roles have been
evidenced in early adolescent groups (Jordan, 1992; Lamke & Bell, 1982). Using differences in
gender role pa
rticipant pairs to alter mesosystemic contexts, Lamke and Bell (1982), found that
dyads made with androgynous early adolescent females showed greater interaction between
participants and improved activity participation, when compared with sex
-
typed or mixe
d
-
typed
early adolescent female dyads. Ridley, Lamke, Avery, and Harrell (1982) altered the behavior
settings of participants in an attempt to show how environmental differences have an impact on
microsystem characteristics: the personal gender roles of ob
ject props. The authors found that
when adolescent participants in mixed
-
sex groups were involved with a feminine
-
oriented
environmental setting, the gender roles of grouped participants were significantly higher on
femininity following participation. The
authors assert that the importance of the environment on
personal gender roles has been largely unexplored, and needs to be considered when examining
gender roles of grouped individuals (Ridley et al., 1982).

Purpose of the Study

Because microsystem differ
ences, such as same
-
sex versus mixed
-
sex sampling or
differences within group activities, might affect the resultant behaviors of individuals and
groups, methods are needed to better assess these setting characteristics

within the
natural
science curriculum
. Moreover, it is important that this

measurement system is able to identify
characteristics at each level of the environment. The ability to investigate these relationships has
serious implications for researchers
or teachers working with
early adolescent females, who are
th
ought to have reductions in affinity toward
science and tec
hnology
, not to mention issues with

self
-
esteem
following early adolescence (
Eccles, 1989;
Muuss & Porton, 1998; Rose &
Montemayor, 1994).







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

13


The purpose of this study was to examine the relations between gender rol
es, self
-
estee
m
and the problem
-
solving

of early ad
olescent females. The study us
ed
qualitative and quantitative

methodologies to supplement the findings of previous studies by creating a controlled
experimental
setting in which activities could be standardized and environm
ental characteristics
assessed.

Five hypothe
ses were identified in the present study:

1.

Microsystem
-
Me
sosystem examination: Personal g
ender r
oles.

Personal gender roles
will be related to the
problem
-
solving

of early adolescent females. Masculine
-
typed
girls are expected to participate more

within group
science
activities than those of
other gender types. It also is expected that feminine
-
typed girls will show lower levels
of dominance and higher levels of verbal expression and supportive behavior during
group activities.

2.

Microsystem
-
Mesosystem exam
ination: Self
-
Esteem.

Self
-
esteem will be related to
the
problem
-
solving

of early adolescent females

in science experiments
. Girls with
higher self
-
esteem are expected to participate more within group activities than girls
with lower self
-
esteem. It also is expected th
at girls with higher self
-
esteem will be
more verbally expressive and dominant than individuals with lower self
-
esteem (see
Edwards, 1994; Lamke, 1982b; Krichevskii, 1977).

3.

Microsystem
-
Exosy
stem examination: Ideal g
ender r
oles.

The dissonance between
perso
nal gender roles and ideal gender roles will be related to the self
-
esteem of early
adolescent females. In other words the greater the difference between personal gender
roles and ideal gender roles, the lower the self
-
esteem of the individual (see also






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

14


Bu
rnett, et al., 1995; Newcomb & Bagwell, 1995; Newcomb & Bagwell, 1998;
Rosenberg, 1965).

4.

Microsystem
-
Mesosystem examination.

An interaction between personal gender roles
and self
-
esteem will be related to the
problem
-
solving

of early adolescent f
emales

in
group based science experiments
. Androgynous girls with high self
-
esteem will show
greater group participation, higher verbal expression, and higher levels of positive
affect than other gender
-
typed girls or girls with lower self
-
esteem (see Cate &
Sugawara, 1986; Rose & Montemayo
r, 1994). Whereas masculine
-
typed girls with
high self
-
esteem will show higher levels of dominance, feminine
-
typed girls with low
self
-
esteem will show higher levels of supportive behavior.

5.

Mesosystem
-
Microsystem examination.

Differences in the
problem
-
solving

of early
adolescent females will be associated with differences in the environments with
which participants interact. It is believed that
scientific
behavior settings with few
object props or obvious patterns of behavior will require masculine
-
typed fo
rms of
group behavior (independence, dominance). Also, it is believed that behavior settings
with one or two identifiable patterns of behavior will require feminine
-
typed forms of
group behavior (peacekeeping, egalitarianism). Finally, it is believed that
behavior
settings that reveal the need for multiple behaviors to occur simultaneously will
require androgynous forms of group behavior (reflexive, adaptable).

Because differences at the microsystem could affect the behaviors of individuals,
methods are nee
ded which could better assess the relation
ship between gender roles, self
-
esteem,
and
problem
-
solving
. The ability to investigate these relationships has serious implic
ations for

researchers or teachers working wi
th

early adolescent females. As such, this study

will attempt to






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

15


create both
qualitative and quantitative

and quantitative methodologies that can assist in
understanding the relationship between gender roles, self
-
esteem,
problem
-
solving
, and behavior
settings.

Method

Participants

Participants were 48 early adolescent females enrolled for summer programming at a
Midwestern Girl Scout camp. The girls ranged in ages from 9 years to 13 years old. Participant
groups were organized based on membership in

camp cabins, which each housed six early
adolescent girls. Each six
-
member group had been living together in the same cabin for two
weeks. No differences in age, race, or socioeconomic status were found between those girls who
participated and those who d
id not. In total, eight groups of early adolescent girls participated,
with four groups of 9 to 11.5 year
-
olds and four groups of 11.6 to 13 year
-
olds. 46 of the girls
were of European
-
American descent, with two early adolescents describing themselves as
A
frican
-
American. The majority of the girls came from middle
-
class families.

Procedure

All data collection took place during the final days of the two
-
week camp experience to
ensure that the girls were familiar with one another. Participants were observed a
s they
completed a hands
-
on
natural
science
activity that required individual and group involvement.
Completion of the questionnaires, group activity, and study involvement took approximately
three hours. Participants completed this activity as fulfillment of optional pro
gram requirements.

Early adolescent females completed 2 forms of the Bem Sex
-
Role Inventory (BSRI, Bem,
1978), one to assess personal gender roles and one to assess ideal gender roles, and then
completed the Coopersmith Self
-
Esteem Inventory (CSEI, Coopers
mith, 1981). Each group of






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

16


six then participated in an activity that required participation at three behavior settings with
differing microsystemic standing patterns of behavior: behavior setting one required the
exhuming of a set of horse bones, behavior
setting two required the organizing of the horse
skeleton, and behavior setting three required the reconstructing of a horse skeleton on a steel
stand.

Completion of questionnaires.

First, the girls completed the following questionnaires: the
Bem Sex
-
Role
-
Personal Inventory (BSRI, Bem, 1978), the Bem Sex
-
Role
-
Ideal Inventory
(BSRI, Bem, 1978; Grimmell & Stern, 1992) and the Coopersmith’s Self
-
Esteem Scale (CSEI,
Coopersmith, 1981). The questionnaires were filled out at picnic benches in a field located nex
t
to the behavior settings, and were completed before the group activity. The questionnaires
required approximately 30 minutes to complete.

Behavior settings.

Three distinct behavior settings were created and were defined in
accordance with Barker, Ko
unin, & Wright (1955) and Barker and Associates (1978). Each
behavior setting was created with specific time, place, and object props. It was the goal of the
researcher to create three separate environments that could be investigated as to their relationsh
ip
with
problem
-
solving
, gender roles, and self
-
esteem. In other words, three separate behavior
settings were created that could assist in the investigation of mesosystem
-
microsystem and
mesosystem
-
exosystem relationships. Participants were direc
ted that they were to (1) uncover the
contents of behavior setting one, (2) organize the items to create something identifiable at
behavior setting two, and (3) reconstruct the items at behavior setting three.

Behavior setting one
. The first behavior setti
ng was a 10
-
foot x 10
-
foot x 2
-
foot
wooden sandbox that was divided into 6 sections, one section for each participant. Rope
was used to mark off the six sections. A horse skeleton was buried under loose wood






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

17


chips. At the beginning of the activity, each of

the six girls was positioned toward one of
the six sections of the sandbox. The girls were instructed to search through the wood
chips and examine the contents. Then, group members were to remove parts of the
skeleton and place them at behavior setting tw
o, a grid of the same dimensions located
next to behavior setting one.

Distinct place, time, object props, and standing patterns of behavior separated
behavior setting one from the other two behavior settings. The
place

of the first behavior
setting was t
he wooden sandbox. The
time

of the first behavior setting occurred during
the first 15
-
minutes of the group activity.
Object props

of the first behavior setting were
the wooden sandbox, wood chips, the separated horse skeleton, separating rope, and
partici
pants. It was believed that the
standing patterns of behavior

would be (1) digging
individually for objects, (2) expressing excitement over findings, and (3) removing of
objects to behavior setting two.

Behavior setting two
. The second behavior setting c
onsisted of rope laid in a
graph in the same shape and size as behavior setting one. A steel rod that could be used to
connect the vertebrae of the spinal column lay in between behavior settings one and two.
At behavior setting two the girls were asked to
reorganize the bones and to determine the
identity of the skeleton. The participants gravitated naturally from behavior setting one to
behavior setting two as they moved the bones between behavior settings.

Distinct place, time, object props, and standing

patterns of behavior separated
behavior setting two from the other two behavior settings. The
place

of the second
behavior setting was a 10
-
foot x 10
-
foot grid of rope placed on the grass. The
time

of the
second behavior setting was approximately between
15
-
minutes and 45
-
minutes of the






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

18


group activity.
Object props

at the second behavior setting were the rope grid, the
separated horse skeleton, a steel rod that could be used to hold together the horse spinal
column, and the participants. It was believed th
at the
standing patterns of behavior

would
be (1) organizing the skeleton, (2) talking to review appropriate placement of bones,

(3) reorganizing of the bones into a definable skeleton, and (4) moving the skeleton to
behavior setting three.

Behavior setti
ng three
. Behavior setting three was a steel stand that measured 4
-
feet x 10
-
feet x 8
-
feet. Four steel rods the lengths of the horse legs were held together by
a wooden base. At behavior setting three participants were asked to reconstruct the
skeleton on
a steel frame, upon which the skeleton would stand with the use of steel
hooks and pins. By the time the activities at behavior setting three were complete, a horse
skeleton would be standing on the four steel rods, a defined spinal column would be held
to
gether with the steel rod moved from behavior setting two, and a horse skull would be
attached to the spinal column. The participants were observed until the group had
determined that the skeleton was complete.

Distinct place, time, object props, and stand
ing patterns of behavior separated
behavior setting three from the other two behavior settings. The
place

of the third
behavior setting was the wood and steel frame used to stand the skeleton upright. The
time

of the third behavior setting approximately oc
curred during the final 30
-
minutes to
45
-
minutes of the 100
-
minute group activity.
Object props

at the third behavior setting
were the wood and steel frame, the organized parts of the horse skeleton, six
-
inch steel
pins to hold the bones against the steel
frame, and the participants. It was believed that
the
standing patterns of behavior

would be (1) organizing sections of skeleton, (2)






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

19


placing and attaching of the skeleton to the stand, and (3) confirming that the skeleton
was complete.

Group participants
were to work together to identify the most appropriate methods for
completion of the activity. The girls enjoyed interacting in the group activities, and appeared
oblivious to the observers.
1


Three graduate
-
student researchers, two women and a man, observ
ed the activities from a
distance of approximately 15
-
feet. Each sat equidistant from the three settings. The researchers
had each been trained in using the rating scales in a similar study that served as a pilot test. Inter
-
rater reliabilities also were
performed on a pilot group at the three behavior settings under study.

Measures

In order to test the hypotheses under consideration, information was obtained from three
levels of the environment: the microsystem, the mesosystem, and the exosystem. Informat
ion
obtained on the microsystem included participant demographic information, the personal gender
roles of participants, and the self
-
esteem of participants, all taken from written questionnaires
(BSRI, Bem, 1978; CSEI, Coopersmith, 1981). Information obta
ined on the mesosystem
included behavior rating scales of early adolescents’
pr
oblem
-
solving
, behavior setting analysis,
individual interviews with group participants, and group interviews with participants.
Information obtained at the exosystem
included participants’ perceptions of ideal gender roles
(BSRI, Bem, 1978), individual interviews with group participants, group interviews with
participants, document analysis of the Girl Scout Handbook (Girl Scout Handbook, 1995), and




1

To ensure that the setting was as naturalistic as possible, the groups' camp counselors
observed the actions of the group. Following the activities, the counselor was asked “How
typical was the behavior of the girls during this exercise?”
Counselors agreed that the girls were
behaving in a typical fashion.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

20


document analysis o
f journals, letters sent home to parents, and camp counselor documentation
of participants.

Measures at the Microsystem

Information obtained on the microsystem included participant demographic information,
personal gender roles, and self
-
esteem, all taken
from written questionnaires (BSRI, Bem, 1978;
CSEI, Coopersmith, 1981).

Demographic factors.

Information on the girls’ age, race, and years of education was
obtained from self
-
reports.

Bem Sex
-
Role Inventory (BSRI): Personal Gender Roles.

Participants we
re asked to
complete the Bem Sex
-
Role Inventory (BSRI, Bem, 1978) to gain information on participants'
personal gender roles. Using a 7
-
point Likert scale that ranged from Never or Almost Never to
Always or Almost Always, respondents were asked to rate whi
ch of 60 adjectives were
descriptive of self. Of the 60 personality characteristics, 20 were stereotypically feminine (e.g.,
gentle, submissive, affectionate, sensitive to the needs of others), 20 were stereotypically
masculine (e.g., ambitious, independen
t, assertive, self
-
reliant), and 20 were neutral items (e.g.
,
conceited, happy, truthful).

The masculinity scale and femininit
y scale of the BSRI
-
Personal were coded separately.
Following the coding of both scales, an examination of the feminine and masculine scales would
determine the personal gender role of the participant: feminine
-
typed (feminine
-
typed = high
feminine, low ma
sculine), masculine
-
typed (masculine
-
typed = high masculine, low feminine),
androgynous (androgynous = high masculine, high feminine), or undifferentiated
(undifferentiated = low masculine, low feminine)(Bem, 1978).







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

21


Coefficient alpha was found to range fro
m .84 to .90 on both the femininity and
masculinity scales (Bem, 1978). For this study, coefficient alpha was .93 for the masculinity
scale and .81 for the femininity scale. Internal consistency for the femininity scale was .78, while
internal consistency

for the masculinity scale was .87.

Coopersmith’s Self
-
Esteem Scale (CSEI)
. Self
-
esteem was measured using
Coopersmith’s Self
-
Esteem Inventory (CSEI, Coopersmith, 1981). This 58
-
item school form
contains items with short statements that are answered “Lik
e me” or “Unlike me.” The
instrument was designed to measure individual attitudes of self in four areas: general self
-
esteem,
family, social, and academic areas. Since participants were staying at a summer camp, questions
on the general self
-
esteem scale a
nd soci
al self
-
esteem scale were used.

General self subscale
. The General Self Subscale measured the variance in self
-
esteem re
sulting from perceptions of self. This 26
-
item forced choice scale contained
such statements as “There are lots of things about myself I’d change” or “I’m a failure.”
Each of these forced
-
choice items (1 = Like me, 2 = Unlike me) was computed by
summing up

the number of positive items. A high score indicated high self
-
esteem.
Internal consistency was acceptable, with a Cronbach’s alpha of .77.

Social self subscale
. The Social Self Subscale focused on the individual’s self
-
esteem resulting from social influ
ences. This 8
-
item scale contained such items as “I’m a
lot of fun to be with” or “I’m popular with kids my own age.” Each of these forced
-
choice items (1 = Like me, 2 = Unlike me) was computed by summing up the number of
positive items. A high score indic
ated high self
-
esteem. Internal consistency was
adequate, with a Cronbach’s alpha of .74.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

22


Lie Subscale
. Information for the lie scale was obtained by summing up the points
on eight items. A high score on the lie scale indicated defensiveness in a particip
ant’s
responses. The lie scale indicated low defensiveness for all participants, and was not
included in the results of the study.

Measures at the Mesosystem

Information obtained at the level of the mesosystem level included (1) observations of
early adol
escents’
problem
-
solving
, (2) behavior setting analysis, (3) individual interviews with
group participants, and (4) group interviews with participants. Measures obtained at the
mesosystem were made in accordance with the transactional models of B
ronfenbrenner’s
ecological theory (1972, 1979) and Barker’s behavior setting theory (Barker and associates,
1978; Barker et al., 1943; Schoggen, 1989). These theories suggest that behavior modifies and is
modified by the environment in a continuous sequen
ce. Observations were performed in
accordance with assessment methods defined by Bechtel (1977), Bechtel et al. (1987), and
Krasner (1980).

Behavior rating scales of early adolescent
problem
-
solving
.

Throughout the group
activity naturalistic r
ecording methods were utilized to rate
problem
-
solving

(Barker and
associates, 1978; Barker and Wright, 1955; Schoggen, 1989). Rating scales assessed behaviors at
the molar level (the level between the individual and the environment). Three gradu
ate students
rated two members of each six
-
member group. Inter
-
rater agreement for each rating scale held an
alpha level of .85. Raters were blind to the other variables in the study. Field notes were taken in
an attempt to identify phenomena between group

members as it occurred. Observations were
performed to capture descriptions of group members interacting and sharing their beliefs about
the group’s activity, the roles of group members, information about girls their age, and so on.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

23


Behavior rating scale
s used at the mesosystem identified information on the following
behaviors: levels of participation, levels of verbal expression, levels of positive affect, levels of
dominance, levels of support, and levels of intrusions on other group member’s behaviors.

Participation
. Participation in the activity was measured by summing behaviors,
which included digging, placing of bones on a grid, arranging the bones in skeletal form,
sharing of skeletal parts, and interlocking of skeletal pieces. Participation was bas
ed on a
five
-
point continuum (1 = observation with no action, 2 = primarily observation with
little to no action, 3 = moderate action and observation, 4 = strong action with some
observation, 5 = advanced levels of action with group with no observation).

V
erbal expression
. All instances in which a girl was engaged in reciprocal
communication with another group member were coded as verbal expressions.
Communication included talking, yelling, or whispering. Based on a five
-
point
continuum, a participant could

range from deferent to expressive (1 = not expressing
opinions verbally, 2 = mild verbalization, 3 = moderate verbalization, 4 = open
verbalization, 5 = verbal expressiveness with openness and assurance).

Positive affect
. Positive affect was assessed base
d on whether a participant was
emotionally involved with other group members. Positive affect included verbal praise
towards others, helping other participants with activities, and smiling and joking with
group members. Based on a three
-
point continuum, an

individual could range from low to
high expression (1 = does not express affect, passive, cold, and removed from group
members, 2 = some expression of affect with moderate interaction, 3 = generally open
and expressive, friendly, warm, and personable).







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

24


D
ominance
. Dominance was coded if the participant’s behaviors were found to
direct or influence other group member’s behaviors or the actions of the group.
Challenging the actions of another group member, stating one’s opinions, influencing the
behaviors of

the group, or refusing to listen to the demands of another group member
were all examples. Based on a three
-
point scale, participants ranged from low to high
dominance (1 = deferent with behavior, allowing another to change their behaviors,

2 = somewhat
submissive and withdrawn, but some assertion of beliefs, 3 = assertive
with behavior, not easily accepting of another’s behaviors, involved, active responses).

Support
. Support was coded if a participant was found giving encouragement,
assisting with activ
ities, gesturing, working with the demands made by others, and
showing behaviors that moved with the actions of the group. Examples would be handing
others a needed piece of the skeleton, helping attach a bone, or holding a section while it
was reconstruct
ed by another group member. Based on a three
-
point scale, participants
ranged from low to high support (1 = ignoring of other’s behaviors, no communication
with others, individualized participation, 2 = moderate communication with other
members, moderate i
nteraction with actions of the group, 3 = responding to other’s
behaviors, strong communication between members, involvement with others in a
specific activity).

Intrusions
. Intrusions were defined as one individual’s physical attempts at
restraining the
behaviors of another. Intrusions included any form of touching, body
language, gesturing, or behavior that was meant to detract another group member from
behaviors they were displaying. Based on a three
-
point scale, participants ranged from
low to high int
rusions (1 = low intrusive behavior, little to no detracting behavior,







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

25


2 = moderate intrusive behavior, some detracting behavior, 3 = high intrusive behavior,
consistent detracting behavior).

Behavior setting analysis
. Throughout the group activity natura
listic recording methods
were used to record aspects of the behavior setting, including time, place, and object props, as
well as standing patterns of behavior (Barker and associates, 1978; Barker and Wright, 1955;
Schoggen, 1989). Observations were perfor
med at the molar level of behavior (the level between
the individual and the environment), in accordance with Barker’s behavior setting theory (Barker
& Wright, 1955). Observations were divided and coded using rating scales that were taken from
the observa
tion records (Barker & Wright, 1955).

Field notes were taken to identify phenomena between group members and the
environment as the transactions occurred. Observations of these phenomena were performed in
their environment to create a running description o
f group members interacting with the
environment. Field notes also were used to gain information on the interactions between
members, as well as observe participants as they shared their beliefs about the roles of group
members, girls their age, girls thei
r race, and so on.

Individual interviews with participants
. Following the group activity some of the girls
participated in individual interviews. Individual interviews were of a semi
-
structured nature, and
lasted approximately 15 to 20 minutes. Participati
on with the individual interviews were based
on random selection of group members. Interpretation was essential in identifying how and
whether these girls’ perceptions agreed with or differed from their
problem
-
solving
.

Group interviews with part
icipants
. Following the group activity the girls participated in
group interviews. Group interviews were of a semi
-
structured nature, and lasted approximately






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

26


15 to 20 minutes. Interpretation was essential in identifying how and whether these girls’
percep
tions agreed with or differed from their
problem
-
solving
.

Measures at the Exosystem

Information obtained at the exosystem included (1) participant perceptions of ideal
gender roles, which were taken from written questionnaires (BSRI, Bem, 1978);
(2) individual
interviews with group participants, (3) group interviews with group participants,
(4) document
analysis of the Girl Scout Handbook (1995), and (5) document analysis of journals, letters sent
home to parents, and camp counselor documentation
of participants.

Bem Sex
-
Role Inventory (BSRI): Ideal Gender Roles.

Participants were asked to
complete a second copy of the Bem Sex
-
Role Inventory (Bem Sex
-
Role Inventory, Bem, 1978),
which was used to identify participant perceptions of cultural beliefs
regarding gender roles:
ideal gender roles. Using a 7
-
point Likert scale that ranged from Never or Almost Never to
Always or Almost Always, respondents were asked to rate which of 60 adjectives were
descriptive of “a hero of theirs or a role model,” as use
d by Grimmell & Stern (1992) and
Slavkin and Stright (2000). The 60 adjectives used were the same 60 items used to identify
participants’ personal gender roles. Of the 60 personality characteristics, 20 were stereotypically
feminine (e.g., gentle, submissi
ve, affectionate, sensitive to the needs of others), 20 were
stereotypically masculine (e.g., ambitious, independent, assertive, self
-
reliant), and 20 were
neutral items (e.g., conceited, happy, truthful).

The masculinity scale and femininity scales of th
e BSRI
-
Ideal were coded separately.
Following the coding of both scales, an examination of the feminine and masculine scales would
determine the ideal gender role beliefs of the participant: feminine
-
typed (feminine
-
typed = high
feminine, low masculine), m
asculine
-
typed (masculine
-
typed = high masculine, low feminine),






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

27


androgynous (androgynous = high masculine, high feminine), or undifferentiated
(undifferentiated = low masculine, low feminine)(Bem, 1978).

Coefficient alpha was found to range from .84 to .9
0 on both the femininity and
masculinity scales (Grimmell & Stern, 1992). For this study, coefficient alpha was .91 for the
masculinity scale and .86 for the femininity scale. Internal consistency for the femininity scale
was .75, while internal consisten
cy for the masculinity scale was .84.


Individual interviews with participants
. Following the group activity some of the girls
participated in individual interviews. Individual interviews were of a semi
-
structured nature, and
lasted approximately 15 to 20

minutes. However, specific questions were asked on (1) the roles
of men and women and (2) participant perspectives on the roles of group members during group
activities. Participation with the individual interviews were based on random selection of group
members. Interpretation was essential in identifying participants’ perceptions of ideal gender
roles (Bechtel et al., 1987; Krasner, 1980).


Group interviews with participants
. Following the group activity the girls participated in
group interviews. Group
interviews were of a semi
-
structured nature, and lasted approximately
15 to 20 minutes. However, specific questions were asked about (1) their perceptions of their
behavior during the group activity and (2) the roles of men and women. Interpretation was
es
sential in identifying participants’ perceptions of ideal gender roles (Bechtel et al., 1987;
Krasner, 1980).

Document analysis of the Girl Scout Handbook
. Document analysis of the Girl Scout
Handbook (Girl Scout Handbook, 1995) was performed in order to g
ain information about the
cultural values that participants might hold as members of the Girl Scouts of America.
Information taken from the Girl Scout Handbook (1995) was used in confirming statements that






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

28


participants made regarding their participation at

camp, their participation with the group
activity, and their discussion of gender roles, self
-
esteem, and
problem
-
solving

(such as
participation, dominance, and positive affect).

Document analysis of journals, letters sent home to parents, and c
amp counselor
communications
. Document analysis of participants’ journals, letters sent home to parents, and
camp counselor communications were performed in order to gain information about the cultural
values that group members hold. Further, this informat
ion was used in confirming statements
that participants made regarding their participation at camp, their participation with the group
activity, and their discussion of gender roles, self
-
esteem, and
problem
-
solving

(such as
participation, domina
nce, and positive affect).

Two
-
step Member Checking Process



40 of the 48 girls who participated in the study were consulted as to the accuracy of field
observations, interview information, and the completed paper. 8 participants were unable to be
contact
ed for purposes of member
-
checking. The purpose of these consultations were performed
in order to assess the validity of records and the accuracy of the results in describing their culture
(exosystem), behaviors (mesosystem), and gender roles and self
-
este
em (microsystem). After
criticisms regarding information were documented and included, 32 of the 40 participants were
consulted again. No new information was included and no criticisms were offered.

Triangulation of Data Sources


Data collection included t
riangulation of data sources. Triangulation was designed to
verify the validity of data collected at each level of the environment. Triangulation included
triangulation by observer, as observers reviewed (1)
problem
-
solving

identified during fiel
d
observations, (2) behavior rating scales, (3) field observation notes, (4) comments made during






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

29


individual and group interviews, and (5) documents analysis of the Girl Scout Handbook (1995)
and participant journals and letters.

Results

(Quantitative)

I
t was an interest of this study to note whether gender roles and self
-
esteem are related to
differences in early adolescent females'
problem
-
solving
. It also was an interest to note whether
activities could be standardized and environmental chara
cteristics assessed.

Five hypotheses were identified in the present study: (1) personal gender roles will be
related to the grouped behaviors of early adolescent females, (2) self
-
esteem will be related to the
problem
-
solving

of early adolescent

females, (3) the dissonance between personal gender roles
and ideal gender roles will be related to the self
-
esteem of early adolescent females,

(4) interactions between gender roles and self
-
esteem will be related to the
problem
-
s
olving

of
ear
ly adolescent females, and (5) the
problem
-
solving

of early adolescent females will be
associated with differences in the environments with which participants interacted. Correlations
between dependent and independent variables can be found in Ta
ble one
.

_________________________

Insert Table One

About Here

___________________________

Microsystem
-
Mesosystem Examination: Personal Gender Roles and Self
-
Esteem

Personal gender roles and self
-
esteem were both hypothesized to be related to the
problem solving

of early adolescent females. To detect whether participation in a female early
adolescent group activity was related to participants’ personal gender roles or self
-
esteem, a 4
(personal gender roles typology) x 3 (level of self
-
esteem)

analysis of variance was performed on






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

30


each dependent variable involving
problem
-
solving
. Age was the between subject variable, while
personal gender roles typology and level of self
-
esteem were within
-
subjects’ variables.

Participation.

Partici
pation was not found to be related to either personal gender role
typology (
F

(3, 47) = 1.04,
p

= .40) or level of self
-
esteem (
F

= (2, 47) = .31,
p

= .74).

Verbal expression.

Verbal expression was related to individuals’ personal gender roles
(
F

=

(3, 47) = 4.14,
p

= .02), with feminine
-
typed individuals showing greater levels of
verbalization when compared with other personal gender role types. Individuals with higher
levels of self
-
esteem also were found to be more verbal during the group activi
ty than individuals
with lower levels of self
-
esteem (
F

= (2, 47) = 3.31,
p

= .02). A significant interaction between
personal gender roles and self
-
esteem also was identified, with androgynous early adolescent
females with high self
-
esteem showing increa
sed verbalization when compared with other group
members (
F

= (5, 47) = 3.18,
p

= .03).

Positive affect.

Positive affect was found to be related significantly to individuals'
perceptions of personal gender roles (
F

= (3, 47) = 5.06,
p

= .01), but not the
ir level of self
-
esteem (
F

= (2, 47) = .47,
p

= .63). Androgynous individuals were identified as being more
positive during group interactions than other gender
-
typed participants. Also, a significant
interaction effect for personal gender roles and self
-
esteem was identified, with feminine
-
typed
high self
-
esteem females showing greater positive affect (
F

= (5, 47) = 3.66,
p

= .02).

Dominance.

Dominance was related significantly to participants’ personal gender roles

(
F

= (3, 47) = 12.29,
p

= .001), with

masculine
-
typed girls showing greater dominance.
Individuals with higher levels of self
-
esteem also were found to be more dominant during the
group activity than individuals with lower levels of self
-
esteem (
F

= (2, 47) = 5.17,
p

= .01). A
significant in
teraction effect between gender roles and self
-
esteem also was identified for group






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

31


members’ dominance (
F

= (5, 47) = 8.96,
p

= .001).

Support.

Feminine
-
typed individuals were found to be more supportive during group
interactions than other gender
-
typed g
roup members (
F

= (3, 47) = 7.90,
p

= .001). No
significant relationship was identified between levels of supportive behaviors and level of self
-
esteem (
F

= (2, 47) = 1.81,
p

= .19), but an interaction between personal gender roles and self
-
esteem was id
entified, with feminine
-
typed early adolescents with low self
-
esteem supporting the
actions of others more often than other group members (
F

= (5, 47) = 5.09,
p

= .004).

Intrusive behaviors
. Level of intrusive behaviors was not found to be related to
part
icipants’ gender roles (
F

= (3, 47) = 1.781,
p

= .19) or level of self
-
esteem (
F

= (2, 47) =
.80,
p

= .47).

Microsystem
-
Exosystem Examination: Ideal Gender Roles


It was believed that the dissonance between personal and ideal gender roles would be
relat
ed to the self
-
esteem of early adolescent females. In other words, the greater the difference
between personal and ideal gender roles, the lower the self
-
esteem of the participant. A score
was obtained by subtracting participants' personal and ideal scores

for both the masculine and
feminine scales, resulting in a dissonance
-
masculine and a dissonance
-
feminine score. To detect
whether self
-
esteem in female early adolescents was related to the dissonance between
participants' personal and ideal gender roles,

a
P
earson

product moment correlation was
performed. The correlation confirmed that the dissonance on the masculine scale was related to
the self
-
esteem of participants, with higher dissonance on the
masculine

personal and ideal scales
resulting in lower se
lf
-
esteem (
r

=
-

.35,
p

= .01). However, the correlation was not significant
between the dissonance on the feminine scale and the measure of self
-
esteem (
r

=
-

.05,
p

= 78).

Results (
Qualitative and quantitative
)







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

32


Mesosystem
-
Microsystem Examination


Differences in the
problem solving

of early adolescent females were believed to be
related to differences in the environments with which participants interacted. It is believed that
environments with few object props or obvious standing patterns of behavior would requ
ire
masculine
-
typed forms of group behavior (independence, dominance). Also, it was believed that
behavior settings with one or two identifiable standing patterns of behavior would require
feminine
-
typed forms of group behavior (peace
-
keeping, egalitariani
sm). Finally, it was believed
that behavior settings that reveal the need for multiple standing patterns of behavior to occur
simultaneously would require androgynous forms of group behavior (reflexive, adaptable).

Rather than follow the adult model of
problem
-
solving

(the managerial model with
masculine as dominant and actively engaged, and feminine as submissive and hesitant to
participate within group interactions, see Edwards, 1994), the author chose to allow the
categories of girls'
problem
-
solving

to arise from the data observed (Bechtel et al., 1987;
Krasner, 1980).


Behavior Setting One: Modeling Behaviors as the Standard Pattern of Behavior

Behavior setting one was based on a fairly simple, but initially ill
-
defined activity:
pa
rticipants were to find bones hidden under wood chips and soil and move them to behavior
setting two. Distinct place, time, object props, and standing patterns of behavior separated
behavior setting one from the other two behavior settings. The
place

of th
e first behavior setting
was the wooden sandbox. The
time

of the first behavior setting occurred during the first 15
-
minutes of the group activity.
Object props

of the first behavior setting were the wooden
sandbox, wood chips, the separated horse skeleton
, separating rope, and participants. It was






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

33


believed that the
standing patterns of behavior

would be (1) digging individually for objects, (2)
expressing of excitement over findings, and (3) removing of objects to behavior setting two.

It would appear that

the principal standing pattern of behavior identified at behavior
setting one was a modeling behavior. Modeling has been defined as feminine
-
typed or
androgynous individuals accepting or paralleling the behaviors of masculine
-
typed group
members (Edwards,

1994). Modeling was evidences as girls began digging and removing bones
from behavior setting one to behavior setting two. However, as the following passage from the
beginning of the activity shows, some girls were quick to show that the masculine
-
based
m
odeling dichotomy used to describe adult group and leadership behaviors would not be
adequate in describing early adolescent modeling behaviors:

The girls gaze at each other for several seconds, seemingly unsure of how to begin. Then
they watch, . . as She
ila (feminine
-
typed, high self
-
esteem) jumps feet
-
first into the wood
shavings and starts furiously digging through the chips.

“I found something!” she shrieks with apparent delight on her face.

With a disgusted and uncomfortable look, Amanda (feminine
-
ty
ped, average self
-
esteem)
backs away from the site where her cohort has just identified the first remains of a horse
skeleton. “Oh my God. It’s a bone,” she shares, apparently with no one. Amanda pauses
for several seconds, then enters the wooden box, cle
aring away wood chips with a
disgusted look on her face (field observations, 1998).


As a feminine
-
typed individual, Sheila identified herself as a person who “values warm,
friendly and comfortable interactions” (BSRI results, 1998). She stated, “I have al
ways liked
doing things myself. It’s hard to be a group member ‘cause you have to share. Sometimes I do,
but most of the time, I will let others help me” (Sheila, individual interview, 1998). Although the
literature suggests that the higher dominance and h
igher participation of masculine
-
typed group
members would be paralleled by other members (Edwards, 1994; Jordan, 1992), this does not
appear to be the case with this group of early adolescent females. In fact, at behavior setting one,
it appeared that nei
ther feminine
-
typed nor masculine
-
typed early adolescent girls were more






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

34


likely to enter the activity and serve as leaders. However, initial entrance into the activity did
seem to require a high sense of self
-
esteem, such as that evidenced by Sheila.

When

asked whether or not she would believe herself to have either dominant or
masculine leadership qualities (aspects of personality assumed to be valued as ideal traits by
early adolescent females), she responded “I don’t think so! I just wanted to figure ou
t what was
going on, and so I just [got involved]. That doesn’t mean that other people should do what I do”
(Sheila, group interview, 1998). Sheila shrugged off the label of a leader/ideal figure who should
be modeled, and did not believe it possible that
her way of engaging in the activity should be
followed by others. “They should just do it their way, that’s all,” she challenged (Sheila,
individual interview, 1998). If a personality characteristic could be easily identified as being
critical for entrance

and involvement into a site with ill
-
defined standing patterns of behavior, it
would be high levels of self
-
esteem.

In another group, Tia (masculine
-
typed, low self
-
esteem) quickly admitted that she
watched her peers before understanding what she should b
e doing. Though masculine
-
typed girls
have been thought to be more dominant within groups (Edwards, 1994), Tia's behaviors are not
indicative of such a standing pattern of behavior. Clearly self
-
esteem was critical as the girls
tried to make sense of an il
l
-
defined setting. Tia later shared that she didn’t mind the help from
her friends. “I didn’t know what I was doing, so I looked for help. They already started, so they
helped me out” (Tia, individual interview, 1998). Tia denied that she believed her idea
s about the
activity were not as advanced as those of Elizabeth or Megan


she believed that they had begun
first, and would be able to help her, or perhaps because they were more comfortable with the
setting. However, it was important that the person star
ting the activity have the esteem and
efficacy to initiate the task.







Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

35


Whether these girls had a higher stronger affinity for ill
-
defined behavior settings than
Tia or whether these girls were more assertive in environments with few object props or obvious
s
tanding patterns of behavior was neither identified in their inventory results or in the group or
individual discussions that followed the study participation (CSEI results, 1982; Elizabeth &
Megan, group interview, 1998). In fact, Tia strongly denied that

Elizabeth or Megan knew any
better than she did what to do (Tia, individual interview, 1998; document analysis of journal and
counselor letters, 1998). Tia rejected the viewpoint that dominant or masculine
-
oriented
leadership was present at behavior setti
ng one, in that no dominant leader or managerial
behaviors from one individual were present.

Some girls modeled the standing patterns of behavior for other group members at
behavior setting one during the activity. Though some girls were placed in a positi
on often
described as hierarchical and dominant, they were not necessarily aware of its existence. In fact,
they rejected the idea that one girl served as an ideal figure or leader. Even when girls didn’t
appear to want to be modeled, their behaviors conti
nued to guide other group members:

Sheila continues to guide the group, moving between the sandbox of bones and the rope
grid of behavior setting two set up next to it. She tromps between behavior setting one
and behavior setting two, beaming at her discov
eries, and placing the bones in behavior
setting two’s grid. Amanda (masculine
-
typed, average self
-
esteem) and Corrine
(masculine
-
typed, extremely high self
-
esteem) use their eye contact to follow what Sheila
is doing, sometimes glancing quickly and awkwar
dly at the other group members as if to
get a sense of what they should be doing, but never appearing comfortable.

Amanda furiously joins into Sheila’s activity, quickly removing bones, moving them to

behavior setting two, and running back to the sandbox
to continue the search. Corrine

states, “Oh, man! . . it looks like a skull.” She frees herself of the skeletal head and it

thuds at her feet. Corrine jumps back and avoids looking at what was once a horse’s head

(field observation notes, 1998).


Such beha
vior explains why gender roles are weak predictors of
problem solving
. First,
leaders are not self
-
defied. Second, the setting under examination was just as strong as






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

36


personality characteristics in defining the group behavior. Even though Sheila
did not believe her
behavior would affect the behaviors of her peers, Allison and Corinne claimed that they were
following her lead. Allison shared, “Sure I do what Sheila did, even if she didn’t think I was. She
always knows what to do, and she is always
right . . . so I just do it [like she does]” (Allison,
individual interview, 1998). Even though early adolescent females might not be aware that they
serve as an ideal representation for peers, they can in fact be considered a leader, a mentor, or a
guide.

However, it would be ill
-
advised to state that the time, place, and object props that
organized behavior setting one were solely based on the standing patterns of modeling behavior.
For the most part, the girls showed little difference in how they began t
he activities at behavior
setting one. Kay, one of the oldest girls participating in the activity, shared “It didn’t matter who
started in first


we just all had to start. If someone hadn’t gotten involved, then things wouldn’t
work and we wouldn’t figure

out what was going on” (Kay, letter sent home to parents, 1998).
Unlike the masculine
-
typed style of placing the emphasis on the activity or creating a hierarchy
from which to work, these Girl Scouts seemed comfortable beginning the experience by working
simultaneously on the activity. Little emphasis was found of a hierarchy, those who lead and
those who followed, or whether there was a need for one specific leader to emerge.

The standing pattern of behavior, modeling, appeared to be dominant/masculine i
n that a
model emerged, but it also appeared feminine/empathetic in that emphasis was on cooperation
and group involvement. Feminine
-
typed girls could show high leadership capabilities, since
cooperation and group involvement was necessary in some situatio
ns. Whereas the quantitative
results suggested that masculine
-
typed girls would be dominant and feminine
-
typed girls would
display more positive affect and verbalization, the
qualitative and quantitative

results indicated






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

37


that self
-
esteem was critical in determining who be
came involved at behavior setting one.
Quantitative measures do not necessarily explain or predict the individual
-
environmental
interactions that lead to such behaviors.
Problem
-
solving

that resulted from the interplay between
individual characte
ristics and environmental circumstances were also apparent at behavior
setting two.

Behavior Setting Two: Dyadic Interaction as the Standard Pattern of Behavior

Distinct place, time, object props, and standing patterns of behavior separated behavior
setti
ng two from the other two behavior settings. The
place

of the second behavior setting was a
10 x 10 foot grid of rope placed on the grass. The
time

of the second behavior setting was
approximately between 15
-
minutes and 45
-
minutes of the group activity.
Ob
ject props

at the
second behavior setting were the rope grid, the separated horse skeleton, a steel rod that could be
used to hold together the horse spinal column, and the participants. It was believed that the
standing patterns of behavior

would be (1) o
rganizing the skeleton, (2) talking to review
appropriate placement of bones, (3) reorganizing of the bones into a definable skeleton, and (4)
moving the skeleton to behavior setting three. It was believed that as an environment with
several object props a
nd patterns of behavior, more feminine
-
typed forms of group behavior
(egalitarianism and collaboration) would be present.

During the activities at behavior setting two, dyadic interactions were found to occur in
two situations. First, collaborative dyadic
interactions occurred if same gender
-
typed dyads
worked together. Second, positive and collaborative dyadic interactions were more likely to
occur if masculine
-
typed girls were not managerial or dominant with feminine
-
typed girls.


What occurred when a ma
sculine
-
typed girl entered a feminine
-
oriented environment at






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

38


behavior setting two? Amanda asserted that she is a “nice person, warm, and funny” (BSRI
results, 1998; document analysis of journal and letters sent home to parents, 1998). She believed
that s
he could never “give attitude” to another group member, which also could be evidenced by
her fairly average level of self
-
esteem (CSEI results, 1998). Though Amanda entered the digging
at behavior setting one hesitantly, she revealed a dominant style as sh
e involved herself at
behavior setting two.

The activity at behavior setting two was to reconstruct the skeleton and identify what type
of skeleton the girls had discovered. Behavior setting two required a standing pattern of behavior
where several activi
ties could be performed at once, which appeared to make a dominant style
less effective than a collaborative one. Such a setting initiated a more collaborative style of group
work, where the standing pattern of dyadic interaction could be seen. Although th
e girls were
more apt to participate in dyads, limitations to their interactions were discovered:

Clara, a somewhat talkative girl who was quick to get involved at behavior setting two,
begins to organize the leg bones of the horse. Amanda is careful about

carrying bones
between the sites, but quickly involves herself in the activities at behavior setting two
when she spots Clara.

“Look at this one!” Clara exclaims, as she organizes the remnants of the horse’s leg.

Amanda walks over and shouts, “This one’s

mine!,” pushing her, while grabbing it from
Clara’s hands and making a hateful face (field observations, 1998).


Amanda, a feminine
-
typed girl in a collaborative
-
oriented environment, was identified
scolding Clara, a masculine
-
typed girl who was beginning

to show some dominance in the
organizing of the bones at behavior setting two. It would appear that girls who define themselves
as feminine with average levels of self
-
esteem, when involved with activities that are
collaborative and group
-
oriented, are no
netheless able to show forms of dominance and reject the
masculine behaviors of masculine
-
typed girls with equally
-
average levels of self
-
esteem. These
girls did not engage each other in conversations as a group, as Edwards (1994) has suggested that






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

39


empat
hetic and feminine
-
typed leaders would display.

Interaction between masculine
-
typed and feminine
-
typed girls only occurred successfully
when the behavior setting’s activities necessitated it, as in the case when girls needed to align the
bones of the spin
al cord:

Devon (feminine
-
typed, high self
-
esteem) and Jessica (masculine
-
typed, high self
-

esteem) sit cross
-
legged between a huge pile of small, cylindrical bones. Focusing

intently on a sheet of paper that outlines their organization into the spinal colu
mn of the

horse, Devon states, “This one needs to go next . . . no, then this one."

Small bursts of time follow each statement. The girls seem more focused on the hierarchy

of numbers that follow in the pile in front of them and on the sheet beside them.

J
essica, who has remained fairly uninvolved until this time, states, “I don’t think these go

this way. Move that one before this one.” She continues to fiddle with the bones, while

sliding them one by one onto a steel rod that is to serve as a support for
the backbone

(field notes, 1998).


Jessica, a masculine
-
typed girl (BSRI results, 1998), was fairly uninvolved throughout
most of behavior setting two’s activities. However, she responded to Devon asking for
assistance. Jessica was able to collaborate acce
ptably as long as she controlled her dominance.
Devon stated, “Jessica is okay to be with, if she isn’t being bossy and always trying to tell you
what to do … sometimes she just needs to listen and help out (Devon, individual interview,
1998). When in the
course of behavior setting two’s activities it was time to place the bones
together, they both joined in quickly. The other girls moved away from the activity, and began to
organize behavior setting three, preparing for the work Jessica and Devon were doin
g. Though
neither girl classified themselves as active participants throughout the activity (document
analysis of journals, 1998; group interview, 1998), clearly their behaviors guided the finalization
of behavior setting two’s activities, in that the girl
s were successfully able to connect the spinal
column of the horse together (field observation notes, 1998).

During the interviews, Devon shared, “I didn’t see the need to get involved until they got






Gender R
oles, Self
-
Esteem, and Problem
-
S
olving

40


to the spine


I could see the bones fitting together, a
nd thought it would be fun to do that”
(Devon, group interview, 1998). Jessica agreed, saying “It just made sense to me” (Jessica,
personal communication, 1998). Jessica was self
-
classified as masculine on the BSRI (BSRI
results, 1998), while Devon was cla
ssified as feminine (BSRI results, 1998), yet they supported
each other through the activity. Though collaboration is traditionally not identified as a