Relational health and disordered eating in black, latina, and white female college students

MISSING IMAGE

Material Information

Title:
Relational health and disordered eating in black, latina, and white female college students
Physical Description:
vi, 123 leaves : ; 29 cm.
Language:
English
Creator:
Goldman, Lynne G., 1947-
Publication Date:

Subjects

Subjects / Keywords:
Eating disorders in women   ( lcsh )
Women college students -- United States   ( lcsh )
Hispanic American women   ( lcsh )
Women, White -- United States   ( lcsh )
Women, Black -- United States   ( lcsh )
Counselor Education thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Counselor Education -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 2001.
Bibliography:
Includes bibliographical references (leaves 102-122).
Statement of Responsibility:
by Lynne G. Goldman.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 026986034
oclc - 47260653
System ID:
AA00017701:00001


This item is only available as the following downloads:


Full Text











RELATIONAL HEALTH AND DISORDERED EATING IN
BLACK, LATINA, AND WHITE FEMALE COLLEGE STUDENTS














By

LYNNE G. GOLDMAN


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

UNIVERSITY OF FLORIDA


2001














ACKNOWLEDGMENTS

In recognition of my growth within connection, I acknowledge the people who

have been an integral part of my support network during the many years of my academic.

journey.

First and foremost, I thank my best friend and husband, Dr. Richard Goldman,

whose unfailing faith in my ability to accomplish my goal constantly buoyed my spirits

and helped me 'stay the course' even during my most arduous struggles. I also thank my

two wonderful sons, Brian and Michael, who believed in me and offered their support

openly and lovingly.

Special thanks go to my mother and sister. Though it is currently difficult for my

mother, Ruth Greenstein, to understand my accomplishment, she has always been one of

my biggest fans and an ardent believer in the benefits of education throughout our

lifetime. My sister, Sue Kurtz, has offered unconditional love and support even when her

own road has been quite rocky.

I would also like to thank my friends, who from near and far, have always

encouraged and guided me to remain positive and connected. Gail, my dearest friend of

28 years, has always been a beacon of light through some very dark times. Tovah, Suni

and Mary Ann, my grad school buds, have been constantly supportive and understanding

during my dissertation process. And last but not least, the laughter and encouragement

from my colleagues, Ann, Sarah, Berta, Sandy, Jeannie, Bev, Linda, Julie, and Paula has

been priceless.








I would like to give my gratitude to my doctoral committee: Dr. James Archer,

Chair; Dr. David Miller; Dr. Silvia Echevarria-Doan; and Dr. Joe Wittmer. Without their

guidance, the completion of this research project would not have been possible.














TABLE OF CONTENTS

page

ACKNOW LEDGMENTS......................................................................................ii

ABSTRACT ...................................................................................................... vi

CHAPTERS

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

Overview...................................................................................................... 1
Theoretical Framework.................................................................................. 4
Statement of the Problem............................................................................... 6
Need for the Study......................................................................................... 9
Purpose of the Study................................................................................. 12
Rationale for the Approach........................................................................ 13
Research Questions ................................................................................ 15
Definition of Terms .................................................................................. 16

2 REVIEW OF THE LITERATURE ........................................................... 19

Introduction.................... .......................................................................... 19
Theoretical Model .................................................................................... 19
Risk Factors for Disordered Eating in College W omen............................ ... 27
Sociocultural Environment................................................................... 28
Adjustment to College .......................................................................... 33
Race and Ethnicity........................................... .................................... 36
Protective Factors Against Disordered Eating in College Women ............... 40
Peer Relationships .............................................................................. 41
M entor Relationships......................................................................... 47
Community Relationships......................... .................................... 52
Chapter Summary......................................... ......................... ............ 58

3 METHODOLOGY ..................................... ................................................ 60

Overview............................................................................... 60
Population ........... ........................................................................ ......... 60
Sampling and Sampling Procedures............................... ......................... 61
Design ............................................................. ............................. ....... 61









Instrumentation ......................................................................................... 62
Relational Health Indices................................................. ......... ..... 62
Eating Attitudes Test-26 .................................................. ............... 64
Data Analysis .......................................................... ............................... 67
Hypothesis and Research Questions......................................... .............. 67
Hypothesis............................................................................................. 67
Research Questions............................................................................. 67

4 RESULTS ................................................................................................. 70

Data Collection........................................................ ................................. 70
Data Analyses........................................ .................................................. 71
Hypothesis............................................................................................. 71
Research Questions................................................................................ 73

5 DISCUSSION ........................................................................................... 81

Hypothesis Summary and Explanation of Finding ....................................... 81
Research Questions and Explanations of Findings.......................................... 83
Limitations of the Study ........................................................................... 90
Implications of the Findings and Recommendations..................................... 91
Implications for Theory ....................................................................... 92
Implications for Practice ...................................................................... 93
Implications for Research .................................................................... 94
Summary ..................................................................................... ....... 95

APPENDICES

A COVER LETTER...................................................................................... 97

B THE RELATIONAL HEALTH INDICES............... ................................. 98

C THE EATING ATTITUDES TEST-26................................................... 101

REFERENCES ............................................................................................. 102

BIOGRAPHICAL SKETCH ........................................................................... 123














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

RELATIONAL HEALTH AND DISORDERED EATING IN
BLACK, LATINA, AND WHITE FEMALE COLLEGE STUDENTS

By

Lynne G. Goldman

May 2001

Chair: James Archer, Jr.
Major Department: Counselor Education

The purpose of this study was to examine the relationship between Black, Latina,

and White female college students' relational health, specifically peer, mentor, and

community relationships, and disordered eating attitudes and behaviors. Relational

health was defined according to the constructs of mutuality, authenticity, and

empowerment or zest within the Stone Center's Relational Model, a theory of women's

psychological development. Participants in the study were 237 Black, Latina, and White

female undergraduate students who completed the Relational Health Indices and the

Eating Attitudes Test-26. Results of this study showed a significant correlation between

peer and community relationships and disordered eating. No association between mentor

relationships and disordered eating was detected. Between group differences among the

three ethnic groups was not a significant research result. The implications of these

findings for theory, practice, and research are discussed.














CHAPTER 1
INTRODUCTION


Overview

Researchers consistently report that between 4 and 18 percent of female college

students meet the criteria for the clinical eating disorders of anorexia nervosa, bulimia

nervosa, and binge-eating disorder (Drewnowski, Yee, & Krahn, 1988; Hesse-Biber,

Marino, & Watts-Roy, 1999; Mintz & Betz, 1988; Pope, Hudson, Yurglen-Todd, &

Hudson, 1984; Pyle, Neuman, Halvorson, & Mitchell, 1991). Furthermore, between 60

and 80 percent of college women engage in subclinical eating disorders which are

characterized by excessive dieting and exercising, fasting, and other harmful behaviors

that fall short of the criteria set by clinical eating disorder scales (Hesse-Biber et al.,

1999) yet represent characteristics which are a part of a continuum of disordered eating

(Scarano & Kalodner-Martin, 1994; Tylka & Subich, 1999). Many of these weight and

body image concerns or even obsessions are so common on college campuses that they

have become normalized by our current sociocultural environment (Hesse-Biber, 1996).

The negative impact of disordered eating on physical and psychological health

and on social and vocational functioning has been documented extensively (Kaplan &

Woodside, 1987; Katzman & Wolchik, 1984; Mizes, 1988; Sharp & Freeman, 1993;

Williamson, Kelley, Davis, Ruggiero, & Bloudin, 1985). Women who have an eating

disorder report a wide variety of physical and psychological symptoms, including

dizziness, sleep disturbance, gastrointestinal complaints, anxiety, depression, and






2

substance abuse (Mitchell, 1984). Furthermore, these girls and women often experience

interpersonal difficulties, including conflicted relationships with family members, social

alienation, interpersonal distrust, and impaired sexual functioning (Coovert, Kinder, &

Thompson, 1989; Evans & Wertheim, 1998; Grissett & Norvell, 1992; Thelen, Farmer,

Mann, & Pruitt, 1990).

Most of the eating disorder research to date has focused primarily on White

women, often to the exclusion of other racial or ethnic groups (Lester & Petrie, 1998).

Only a small number of studies in the United States have specifically observed eating

disorder symptomatology in women of color (e.g., Abrams, Allen, & Gray, 1993; Akan

& Grilo, 1995; Altabe, 1998; Anderson & Hay, 1985; Caldwell, Brownell, & Wilfrey,

1997; Fitzgibbon, Spring, Avellone, Blackman, Pingitore, & Stolley, 1998; Grange,

Telch, & Agras, 1997; Gray, Ford, & Kelly, 1987; Hiebert, Felice, Wingard, Munoz, &

Ferguson, 1988; Lester & Petrie, 1998; Lester & Petrie, 1995; Nevo, 1985; Pumariega,

Edwards, & Mitchell, 1984; Robinson & Andersen, 1985; Rucker & Cash, 1992; Siber,

1986). In addition, only a few of these studies specifically examined disordered eating in

female college students of color. Root (1990) has suggested that even though certain

factors in minority cultures, such as an appreciation of a healthier body size and a stable

extended family and social structure, may protect some minority women from disordered

eating, the reality of within-group individual differences has been largely ignored.

Smolak and Levine (1996) report that disordered eating begins most commonly

during adolescence, the transitional period of growth between puberty and adulthood.

This is a time when a young girl needs to meet major developmental tasks and master the

developmental issues of separating from the family (Chatoor, 1999). Concurrently, a








young girl feels pressure to adapt to puberty when her body proportions change from

those of a child to those of a young adult (Surrey, 1991 a). She needs to make the

transition between loosening the ties with her parents and increasing her dependency on

her peers. In order to find her place in her peer group, she needs to deal with personal

and cultural values regarding body image, sexuality, and achievement (Chatoor, 1999).

In addition to all of these life challenges, the stressful event of leaving home for college

intensifies these developmental difficulties for young women and sets the stage for

struggles with disordered eating (Martz & Bazzini, 1999). Young women of color face

additional hardships in coping with the vagaries of college life as they discover the lack

of available connection to minority women role models and mentors in the university

environment (Turner, 1997).

Many traditional theories of human psychological development focus on the

consolidation of an autonomous identity through separation and individuation from

parental figures (Tantillo, 1998). Relational theory posits that the construction of the self

for women occurs through psychological connection and mutual sharing (Gilligan, 1982;

Jordan, 1986, 1995, 1997; Jordan, Kaplan, Miller, Stiver, & Surrey, 1991; Kaplan &

Klein, 1985; Mikel-Brown & Gilligan, 1992; Miller, 1976, 1984, 1986, 1988; Stem,

1990; Surrey, 1985). Miller & Stiver (1997) state that a female's sense of self and of

worth is most often grounded in the ability to make and maintain relationships.

Furthermore, it has been proposed that women who fail to recognize and meet their needs

for interpersonal connectedness struggle with higher levels of eating pathology than those

who value their relationships (Steiner-Adair, 1990; Surrey, 1991 a).








Although all students contend with difficulties related to adjustment to college,

this transition is generally more difficult for racial/ethnic minority students than for

White students (Gloria & Rodriguez, 2000). Ponterotto (1990) reviewed demographic

trends, enrollments, and attrition and graduation rates for Black and Latino students. He

indicated that minority students were more likely than White students to report feelings

of isolation and often feel unwelcomed and unappreciated on predominately White

college campuses. It is possible that these adaptation struggles and feelings of isolation

affect the mental and physical well-being of college women of color and may be

manifested in the attitudes and behaviors of disordered eating.

Currently, there is little research focused specifically on college women who

engage in disordered eating and their relationships, and research on the relationships of

eating disordered minority college women is virtually nonexistent. Looking more closely

at the dynamics between eating disordered behaviors and relationships may help increase

the understanding of the struggles of college women and potentially provide inroads for

effective education, prevention, and intervention strategies.


Theoretical Framework

The importance of social support and relationships in girls' and women's lives has

been studied extensively from various perspectives (Buehler & Legg, 1993; Evans &

Wertheim, 1998; Fuhrer, Stansfeld, Chemali, & Shipley, 1999). Researchers have

indicated that the nature and quality of girls' and women's relationships appear to be

more meaningful to their psychological health than their overall number or specific

interpersonal structure (Gilligan, Lyons, & Hammer, 1990; Hobfoll, 1986; Lu & Argyle,

1992). In particular, those growth-fostering relationships that are intimate and mutual






5

have the ability to foster self-disclosure, coping strategies, emotional resiliency, and

additional social support (Genero, Miller, Surrey, & Baldwin, 1992; Miller & Stiver,

1997).

In accordance with these principles, feminist scholars and researchers at the Stone

Center at Wellesley College have developed a theory called the Relational Model, which

is a paradigm for the understanding of female psychological development and well-being

(Jordan et al., 1991). Whereas Erickson's model of psychosocial development focuses on

the task of separation-individuation for healthy adolescent development (Erickson, 1963),

Relational Model theorists conceptualize ongoing, growth-fostering connection as critical

to women's healthy psychological development (Jordan, 1997; Surrey, 1985). The

Relational Model proposes that as relationships grow, so grows the individual.

Participating in growth-fostering relationships is both the source and goal of girls' and

women's psychological development (Miller & Stiver, 1997).

The relational qualities of mutual engagement, authenticity, and empowerment or

zest have been shown to be important in the arenas of both intrapersonal and

interpersonal growth. The Relational Model defines engagement as perceived mutual

involvement, commitment, and attunement (Surrey, 1991b). Engagement may have a

beneficial impact on individuals as well as relationships as indicated in studies on

empathy and closeness. These qualities of engagement mediate stress and depression and

are associated with self-actualization, self-esteem, low interpersonal distress, and

relationship satisfaction (Beeber, 1998; Burnett & Demnar, 1996; Gawronski & Privette,

1997; Schreurs & Buunk, 1996; Sheffield, Carey, Patenaude, & Lambert, 1995).

Authenticity is not a static state that is achieved at a discrete moment in time, but rather a









girl's or woman's ongoing ability to represent herself in a relationship with increasing

truth and fullness and thereby acquire knowledge of self and other (Miller & Stiver,

1997). Research on authentic self-disclosure and openness appears to be related to being

liked, increased liking of others, and motivation in relationships (Collins & Miller, 1994;

Kay & Christophel, 1995). Empowerment or zest, the experience of feeling personally

strengthened, encouraged, and thereby motivated to take action (Liang, Tracy, Taylor,

Williams, Jordan, & Miller, 2000), has been shown to have a direct impact on positive

affect, meaningful activity, and creativity (Hall & Nelson, 1996; Spreitzer, 1995).

When considering a female's important growth-fostering relationships outside of

her immediate family, close ties with peers and adult mentors and affiliation with some

type of community are three of the most significant types of connections in later

adolescence and young adulthood (Gilligan et al., 1990; Hagerty, Williams, Coyne, &

Early, 1996; Leadbeater & Way, 1996). Though the importance of these three types of

relationships in female psychological development has been mentioned in the literature,

there has been little empirical research within these sociocultural arenas especially related

to the connections of college age woman. In addition to which, there is even less

research specifically focused on relational dynamics and disordered eating in the female

college population (Rorty, Yager, Buckwalter, & Rossotto, 1999).


Statement of the Problem

Colleges and universities across the nation are reporting dramatic increases in the

past two decades in disordered eating attitudes and behaviors among their female students

(Hesse-Biber, 1996). Many college women who are at normal weights regularly express

a strong desire to be thinner and to hold beliefs about food and body image that are








similar to those of women who have clinical eating disorders (Hesse-Biber et al., 1999).

A number of researchers have concluded that instead of viewing eating disorders as

discrete categories, anorexia nervosa, bulimia nervosa, and binge eating disorder may

actually be at the extreme end of a complex continuum of a person's relationship to food

and her body (Hart, 1985; Scarano & Kalodner-Martin, 1994). At one end of the

continuum, individuals express satisfaction with their body image and a desire to practice

healthy eating and lifestyle habits. The other end of the continuum is marked by

excessive weight loss and/or cycles of binge eating with or without purging via fasting,

vomiting, laxative abuse, diet pills, and/or excessive exercise. A wide range of other

attitudes and behaviors related to food and body image otherwise known as subclinical

eating disorders exist between the opposing poles of this continuum (Hesse-Biber et al.,

1999; Shisslak & Crago, 1994; Tylka & Subich, 1999). The potential negative physical

and psychological ramifications of these varied attitudes and behaviors are numerous.

Eating disorders are associated with serious and even fatal medical complications

(Lemberg, 1999; Pomeroy, 1996). Mortality in anorexia nervosa can range from 6% to

20% in the clinical population (Crisp, Callender, Halek, & Hsu, 1992). Recent studies

have shown that even with treatment, only about one half of the affected clinical

population recover and up to one half of this remaining group are severely disabled by

chronic sequela of the disorder (Cavanaugh & Lemberg, 1999). Death is most often a

result of starvation, fluid and electrolyte abnormalities, or suicide (Beumont, Russell, &

Touyz, 1993). A distressing number of patients with bulimia nervosa also die, usually

from cardiac arrhythmias related to electrolyte imbalances (Cavanaugh & Lemberg,

1999). Although death secondary to binge eating is uncommon, fatalities do occur from








complications such as gastric rupture or tearing (Lemberg, 1999). Many other chronic

physical maladies are directly related to the continuum of disordered eating such as

cardiac complications, bowel dysfunction, renal electrolyte abnormalities, alterations of

endocrine function, pulmonary complications, and dental problems (Pomeroy, 1996).

The psychological features of anorexia nervosa, bulimia nervosa, and binge eating

disorder are often similar. They may include depression, anxiety, low self-esteem, the

need for approval and acceptance by others, difficulty expressing anger and frustration,

and feelings of disgust and guilt associated with their eating disorder (Evans &

Wertheim, 1998; Surrey, 1991a). These feelings may become magnified as the disorder

progresses and the affected girl or woman may become more socially isolated,

withdrawn, and obsessed with losing weight (Lemberg, 1999). Concentration difficulties

related to malnutrition, preoccupation with food and exercise, and depression are also

common and oftentimes interfere with academic performance (Thurstin, 1999).

A critical developmental challenge for adolescent girls in our culture is to come to

terms with the biological changes accompanying pubertal development (Striegel-Moore

& Cachelin, 1999). The accompanying adolescent growth spurt, the normal tendency to

gain weight, and the significant increase in body fat relative to overall weight are

important factors in girls' developing preoccupation with culturally mandated thinness

(Wooley & Wooley, 1980). Certain facets of the college social environment also

exacerbate this focus on thinness. Because college is a time in which dating serves an

important social function and appearance is a critical determinant of partner attraction,

the pressure on young women to be attractive which directly equates with low body

weight is especially salient during these years (Martz & Bazzini, 1999).








Traditional theories of psychodynamic object relations and family systems ascribe

the fundamental problems of eating disorders as a failure to separate and individuate from

familial connections and a failure to gain a sense of independence and autonomy in

relationships (Fishman, 1995; Friedlander & Siegal, 1990; Rhodes & Kroger, 1992).

Contemporary researchers and theorists have challenged the validity and applicability of

these constructs to girls and women (e.g., Gilligan, 1982; Gilligan et al., 1990; Guisinger

& Blatt, 1995; Jordan, 1991; Lang-Takac & Osterweil, 1992; Rude & Burnham, 1995;

Steiner-Adair, 1990; Surrey, 1991a) and these researchers and theorists now state that the

formation and maintenance of relationships and connections to others is critical to female

identity development (Gilligan, Rogers, & Tolman, 1991; Kenny, 1987; Lapsley, Rice, &

Shadid, 1989; McDermott, Robillard, Char, Hsu. Tseng, & Ashton, 1983). In addition,

the importance that girls and women attach to maintaining relationship connections with

others has been overlooked in the majority of previous eating disorder research

(Friedlander & Siegal, 1990; Steiner-Adair, 1991).


Need for the Study

Numerous theoretical models have been proposed to explain the etiology of eating

disorders. Several domains including the socio-cultural context, the familial context,

constitutional vulnerability, and adverse life events have been suggested as potential risk

factors (Striegel-Moore, 1993). Even though the theoretical models differ in the

emphasis placed on given risk domains, there is considerable agreement that the etiology

of eating disorders is multifactorial (Striegel-Moore & Cachelin, 1999). Risk for eating

problems derives from a combination of specific risk factors unique to eating disorders

and general risk factors that are associated with other mental health disorders. Exposure








to these risk factors occurs in diverse settings such as family, school, and peer group and

the salience or potency of risk factors derives in part from the point in development at

which they occur. Risk is also cumulative in that the greater the number of risk factors

experienced, the greater the chance a girl or woman will develop an eating disorder

(Striegel-Moore & Steiner-Adair, 1998).

Until recently, disordered eating was thought to be rare among girls and/or

women of color. As a consequence, only a few studies have included minority

populations (Striegel-Moore, Schreiber, Lo, Crawford, Obarzanek, & Rodin, 2000).

However, girls and women of color have higher rates of eating disorders than has

previously been described or documented in the research literature (Crago, Shisslak, &

Estes, 1996; Striegel-Moore & Smolak, 1996). A major problem with these limited

studies is that the results are oftentimes inconsistent and/or contradictory.

For example, some of the research that has examined Latina and White

differences in eating disorders has been equivocal, with several studies finding similar

incidence rates (Jane, Hunter, & Lozzi, 1999). Other researchers have found that Latinas

are both heavier and less concerned with their weight than Caucasian females (Harris &

Koehler, 1992), while Fitzgibbon et al. (1998) and Smith and Krejci (1991) reported that

certain eating disorder symptoms were more severe in their sample of Hispanic girls and

women compared to White participants. In a study of college females, low self-esteem

was associated with eating disordered behaviors and attitudes among Black and White

students (Akan & Grilo, 1995). In another study, Black female college students were less

likely than White female students to feel depressed after a binge, reported less family

emphasis on food and weight, and felt that a five pound weight gain would not make a








difference in their attractiveness (Gray et al., 1987). Abrams et al. (1993) found that

Black women who were less enculturated into the Black culture scored higher on dieting

and weight concern measures, while Akan and Grilo (1995) found no association

between enculturation and eating attitudes and behaviors. Inconsistencies in eating

disorder research studies with women of color abound and may be due in part to the

inherent diversity with respect to sociocultural background and acculturation and

assimilation within ethnic categories (French, Story, Neumark-Sztainer, Downes,

Resnick, & Blum, 1997).

Research on the etiology of eating disorders is not as advanced as research on

other mental health disorders such as depression. In addition to which, factors that

contribute to resilience against eating disorders have not yet been investigated in detail

(Striegel-Moore & Cachelin, 1999). One possible model for explaining resilience may be

the Stone Center's Relational Model of female psychological development.

Qualitatively, researchers have examined female development through the lens of the

Relational Model and have concluded that mutually empathic relationships are essential

for a sense of overall well-being and for promoting healthy growth and development in

girls and women (Surrey, 1991 a). However, empirical examination of this model has

been limited by a lack of validated instruments designed to measure the specific

constructs of the Relational Model (Liang et al., 2000).

The only published measure that explicitly reflects the constructs of the Relational

Model of female development is the Mutual Psychological Development Questionnaire

(MPDQ; Genero et al., 1992). The MPDQ is an assessment based specifically on

impressions during verbal interactions with a spouse, partner, and/or friend. This








instrument assesses the unitary concept of perceived mutuality within dyadic

relationships, which is only one aspect of the Relational Model (Liang et al., 2000). A

new measure, the Relational Health Indices, has been developed and validated to examine

the growth-fostering qualities of peer, mentor, and community relationships based on the

three types of relational constructs (authenticity, engagement, and empowerment/zest)

which comprise the Stone Center's Relational Model of female development (Liang et

al., 2000).

Researchers have stated that peer, mentor, and community relationships are

important factors in the psychological and physical well-being of Black and Latina

adolescents (e.g., Falicov, 1998; Gloria & Rodriguez, 2000; Leadbeater & Way, 1996;

Way & Chen, 2000). However, very little empirical research has been done which looks

at the possible relationship of disordered eating in older female adolescents of color and

their specific relationship connections. This type of research may assist in the future

development of successful programs for eating disorder prevention, education, and

intervention on college campuses.


Purpose of the Study

This research was undertaken to determine the relationships among White, Black,

and Latina college women's peer, mentor, and community relationships and their

disordered eating attitudes and behaviors. This study empirically used the theoretical

constructs of the Stone Center's Relational Model of female psychological development.

As noted, such an application cannot be found to date in the research literature.

One of the primary purposes of this research was to empirically examine eating

disorders in a diverse female college population. It has been recognized that the college








years are an especially vulnerable time for young women in our sociocultural

environment to develop disordered eating attitudes and/or behaviors (Martz & Bazzini,

1999). As American college campuses become more racially diverse, it is important to

study the dynamics of disordered eating in varied female college populations. Eating

disorders in Black female college students have been examined in a limited number of

studies with inconsistent results. Research on Latina college women and disordered

eating is almost nonexistent. Therefore, this research seeks to expand the empirical

knowledge base of college women of color and disordered eating attitudes and behaviors.

Another purpose of this research is to examine specific types of relationships of

college women. Chatoor (1999) stated that developmentally the older adolescent female

needs to begin loosening the ties with her parents and increase her relational involvement

with others. College women typically develop important relationships with peers,

mentors, and various types of communities. The value of social support to both physical

and psychological health has been well documented (Hobfoll, 1986), but the possible

relationship of eating problems to the peer, mentor, and community relationships of a

diverse population of female college students has not been studied. There is a possibility

that these types of relationships, which are integral to the constructs of the Stone Center's

Relational Model, may prove to be a factor of resilience for female college students

vulnerable to the struggles of disordered eating.


Rationale for the Approach

This research is unique in its focus on Relational Health as a potential factor of

resilience in college women's struggle with the behaviors and attitudes of disordered

eating. Whereas existing measures of social support have tended to assess the structure,








quantity, and general functions of support, Relational Health represents more nuanced

aspects of a broad range of interpersonal connections that are believed to be fundamental

to female psychological development (Liang et al., 1998). In addition, almost all of the

previous studies on girls and women and relational theory have been qualitative in design

and rarely look at the dynamics of disordered eating.

This research also sought to expand the knowledge base on disordered eating by

directly addressing the role that peer, mentor, and community relationships of a diverse

female college population have on the continuum of eating problems. The recognition of

the Relational Health of White, Black, and Latina college women as a variable in the

complex structures of disordered eating may provide insights for university mental health

providers and college administrators as they attempt to make inroads within this

nationwide female student health epidemic.

In this study, Relational Health was assessed using the three scales of the

Relational Health Indices (RHI). Liang et al. (2000) developed these scales in response

to research that validates that among growth-fostering relationships, close ties with peers

and adult mentors and belonging to a supportive community are three of the most

significant types of connections in later adolescence and young adulthood (Gilligan et al.,

1990; Hagerty et al., 1996; Leadbeater & Way, 1996). The RHI was developed for a

female college student population.

Disordered eating was assessed using the Eating Attitudes Test-26 (EAT-26)

(Garner, Olmstead, Bohr, & Garfinkel, 1982), a 26-item self-report measure of eating

attitudes and pathology. This instrument has been useful in identifying eating

disturbances which interfere with normal psychosocial functioning in non-clinical








samples in high risk populations such as female college students (Alexander, 1998;

Button & Whitehouse, 1981; Gross, Rosen, Leitenberg, & Willmuth, 1986; Heesacker &

Neimeyer, 1990; Mazzeo, 1999; Thompson & Schwartz, 1982).


Research Questions

The following research questions were examined in this study:

1. What is the relationship between peer relationships and disordered eating?

2. Does the relationship between peer relationships and disordered eating differ

for White, Black, and Latina female college students?

3. What is the relationship between mentor relationships and disordered eating?

4. Does the relationship between mentor relationships and disordered eating

differ for White, Black, and Latina female college students?

5. What is the relationship between community relationships and disordered

eating?

6. Does the relationship between community relationships and disordered eating

differ for White, Black, and Latina female college students?

7. What is the relationship among White, Black, and Latina college women's

disordered eating?

8. What is the relationship among White, Black, and Latina college women's

peer relationships?

9. What is the relationship among White, Black, and Latina college women's

mentor relationships?

10. What is the relationship among White, Black, and Latina college women's

community relationships?








Definition of Terms

The following list of terms are operationally defined according to their meanings

in this study.

Anorexia nervosa is an eating disorder that is often life-endangering and is

characterized by a distorted body image, excessively low weight, and a relentless pursuit

of thinness (Lemberg, 1999).

Authenticity is the process of acquiring knowledge of self and others and feeling

free to be genuine in the context of the relationship (Liang et al., 2000).

Binge eating disorder is an eating disorder characterized by recurrent episodes of

binge eating in the absence of regular use of inappropriate compensatory behaviors such

as vomiting, excessive exercise, and laxative abuse (Lemberg, 1999).

Black is the term used to refer to girls and women of both African American and

Caribbean American heritage (Comas-Diaz & Greene, 1994).

Body image refers to feelings and attitudes toward one's own body (Hsu &

Sobkiewicz, 1991).

Bulimia nervosa is an eating disorder characterized by recurrent episodes of binge

eating accompanied by inappropriate compensatory behavior in order to prevent weight

gain such as self-induced vomiting and/or laxative abuse (Lemberg, 1999).

Connection is an interaction between two or more people that is mutually

empathic and mutually empowering (Miller & Stiver, 1997).

Continuum of disordered eating places unrestrained eating (asymptomatic group)

at one end of the continuum, clinical eating disorders (eating disordered group) at the








other end of the continuum, and the milder forms of disturbed eating (symptomatic

group) at intermediate points (Tylka & Subich, 1999).

Disconnections are the experiences of feeling cut off from those with whom we

share a relationship. This cutting off is experienced as the pain of not being understood

and of not understanding the other person. It is an encounter that works against mutual

empathy and mutual empowerment (Miller & Stiver, 1997).

Empathy is the capacity that exists in all people to feel and think something

similar to the feelings and thoughts of another person (Miller & Stiver, 1997).

Engagement is a relational quality that is defined by perceived mutual

involvement, commitment, and attunement to the relationship (Liang et al., 2000).

Latina is the term used to refer to girls and women of Mexican, Caribbean, and

South and Central American heritage (Gloria & Rodriguez, 2000).

Mentor is an adult person other than a primary caretaker, peer, or romantic partner

who is available for support and guidance and is a positive role model (Leadbeater &

Way, 1996).

Mutual empathy is a joining together based on the authentic thoughts and feelings

of all the participants in a relationship. It is different from one-way empathy and out of it

flows mutual empowerment (Miller & Stiver, 1997).

Mutual empowerment is composed of five essential components: zest, action,

knowledge, worth, and a desire for more connection (Miller & Stiver, 1997).

Mutuality is a way of relating, a shared experience in which all of the people

involved are participating as fully as possible (Miller & Stiver, 1997).








Relational Model/Theory is a paradigm for the assessment of women's

psychological development and well-being. The goal is not for the individual to grow and

mature out of relationships, but to grow into them. Growth-fostering relationships are

both the source and the goal of women's development (Miller & Stiver, 1997).

Relationship is the set of interpersonal interactions that occur over a length of

time. A relationship is usually a mixture of both connections and disconnections (Miller

& Stiver, 1997).

Subclinical eating disorders are disordered eating behaviors that do not meet

diagnostic criteria for a clinical eating disorder, yet are problematic for the individual and

may be a precursor to a more serious eating disorder at a later time (Lemberg, 1999).

Zest/empowerment is the experience of feeling personally strengthened,

encouraged, and inspired to take action (Liang et al., 2000).


Organization of the Remainder of the Study

The remainder of this study is organized into four chapters. Chapter 2 presents a

review and analysis of relevant, related literature. Chapter 3 presents the research

methodology including a description of the population and sample, sampling procedures,

research hypotheses, and instrumentation. Data collection procedures, data analyses, and

the results of the study are presented in Chapter 4. Chapter 5 concludes the study with a

summary of the dissertation research, discussion of results, limitations, implications, and

recommendations for future research.














CHAPTER 2
REVIEW OF THE LITERATURE


Introduction

The purpose of this chapter is to summarize the professional literature relevant to

this study of relational health and disordered eating in Black, Latina, and White female

college students. This literature review includes the following topics: (a) traditional

models of human psychological development and the Wellesley College Stone Center's

Relational Model of female psychological development and well-being; (b) risk factors

for disordered eating in college women which include sociocultural context, college

adjustment, and race and ethnicity; and (c) protective factors against disordered eating in

college women which include peer relationships, mentor relationships, and community

relationships.


Theoretical Model

Many theorists of human development (e.g., Blos, 1962; Erickson, 1968;

Levinson, 1978) have proposed that the processes of individuating oneself from others

define various stages of adolescent development. From this perspective, development of

a sense of self is believed to be attained through a series of crises by which an individual

adolescent accomplishes independence and autonomy via separation from others

(Johnson, Roberts, & Worell, 1999).








Blos (1962), a psychoanalytic theorist of adolescent development, was an ardent

believer in the necessity of separation and individuation. He proposed that the

developmental stage of adolescence could be considered "the second individuation

process," whereas the first process of individuation occurs towards the third year of life

with the attainment of object constancy. Blos (1967) further suggested that failure to

individuate fully in adolescence leads to the probable consequences of deviant

development or psychopathology.

In Erikson's (1963) developmental schema, following the initial stage of "trust

versus distrust" in the first year of life, every subsequent stage until young adulthood

involves some variation of separation from others. It is not until the early 20's when an

individual reaches the stage of "intimacy versus isolation" that close relationships with

others are emphasized (Erikson, 1968).

Levinson (1978) considered the ages from 17 to 22 as the stage of "early adult

transition." During this stage, he proposed that separation was the primary theme

especially separation from adolescent groups and family of origin. Levinson (1978)

believed that this stage was characterized by an entrance into new, more autonomous

roles with a significant increase in psychological distance from the family.

Such traditional theories of lifespan development have become influential in

Western culture because they represent prescriptions for what 'should' happen in the

process of human development (Miller, 1984). From this perspective, according to

McGoldrick (1989), the reality of continuing interpersonal connection is often lost or

relegated to the background. Miller (1991) contended that the prevalent definition of a








mature autonomous "self" is not congruent with female experience which is more aptly

described by a mutually interacting self, informed by mutual empathic experience.

During the last three decades of the twentieth century, writers on the psychology

of women have questioned the priority of valuing individuation and autonomy over

relational connection in female psychological development and well-being (Gilligan,

1982; Jordan et al., 1991; Miller, 1976). Substantial research has validated the fact that

women are healthier both physically and psychologically when they develop in the

context of relationships from infancy through later adulthood and that this connectedness

to others helps not hinders the development of a solid sense of self (Ainsworth, 1989;

Beeber, 1998; Buehler & Legg, 1993; Burnett & Demnar, 1996; Evans & Wertheim,

1998; Kenny, 1991; Lu & Argyle, 1992; Rhodes & Kroger, 1992).

Chodorow (1978) stated that the most important feature of early infantile

development is that it occurs in relation to another person or persons, usually the mother.

She proposed that a female affiliative or relational self emerges from a parenting

structure in which mothers interact differently with their sons and daughters. Daughters

are treated as maternal projections and never fully separate from her, and thereby come to

define themselves as connected to or continuous with others with more permeable ego

boundaries. Boys, on the other hand, in order to develop a sense of male gender identity

are treated as separate from their mothers and come to identify themselves as

differentiated from others with more rigid ego boundaries.

This sense of developing 'within' rather than 'away from' relationships with

others seems to follow females as they progress through the evolving stages of infant,

child, adolescent, and adult development (Ainsworth, 1989; Gilligan, 1982; Gilligan et








al., 1991; Kenny, 1987, 1991; Miller, 1991). This construct of connection and the

importance of social support and relationships in girls' and women's lives have been

studied extensively from various perspectives (e.g., Boyce, Harris, Silove, Morgan,

Wilhelm, Hadzi-Pavlovic, 1998; Buehler & Legg, 1993; Fuhrer et al., 1999; Harris,

Blum, & Resnick, 1991; Warren, 1997).

Boyce et al. (1998) examined the mediating effect of social support in the

development of depression among 193 high-risk, low socioeconomic females between

the ages of 16 and 50 with dependent children. Major depression was associated with

perceptions of low parental care in childhood, low care with current partner, and an

unsatisfactory social support network. Buehler and Legg (1993) examined the effects of

social support on the relationship between stressful life change and the psychological

well-being of 144 separated women aged 20 to 45 with dependent children. The

relationship between life change and psychological well-being was positively associated

with various aspects of social support. In a large cohort of 1,877 middle-aged female

British Civil Servants, Fuhrer et al. (1999) found that regardless of the source of social

support, social relationships within and outside the workplace appeared to be negatively

associated with psychological distress. In a study of 36,284 Minnesota adolescents in the

7th through 12th grades, girls were found to express more emotional distress than boys

when feeling disconnected in intimate relationships. Those girls who experienced lower

levels of connectedness with others evidenced higher levels of emotional stress, negative

body image, suicide attempts, and pregnancy risks (Harris et al., 1991). Warren (1997)

examined relationships between depression, stressful life events, social support, and self

esteem in 100 middle class African American women aged 20 to 35. Statistical analysis








revealed a positive relationship between depression and stressful life events and a

negative relationship between depression and social support.

Research findings from such studies have indicated that the quality and nature of

women's relationships are probably more meaningful than their specific quantity or

structure (Liang et al., 2000). Some large etiological studies have found significance for

the value of network size and health for men, but not for women (House, Robbins, &

Metzner, 1982; Schoenbach, Kaplan, Fredman, & Kleinbaum, 1986). Women's physical

and mental health tends to benefit most from relationships with female friends and

relatives who are nurturing and providers of emotional support (Berkman & Syme, 1979;

Wheeler, Reis, & Nzlek, 1983). In addition, Wheeler et al. (1983) found that high

numbers of social contacts do not ward off loneliness. It is only when these relationships

involve emotional intimacy and disclosure that women are able to combat feelings of

loneliness and alienation. VanderVoort (1999) examined 280 male and female

undergraduate students and found that the women in the study reported significantly more

satisfaction with their social support systems and less isolation and depression. For these

college women, the emotional support given by their various relationships met their

emotional needs by enabling them to feel valued as well as process or work through their

emotional difficulties (Vandervoort, 1999).

Those relationships that are intimate and mutual can facilitate self-disclosure,

emotional resiliency, coping strategies, and additional social support for girls and women

(Genero, Miller, Surrey, & Baldwin, 1992; Jordan, 1986; Miller & Stiver, 1997). From

this basic construct, researchers at the Wellesley College Stone Center have developed

the Relational Model as a theoretical paradigm for the assessment of women's








psychological development and well-being (Liang et al., 2000). Relational Model

theorists have focused on the "self-in-relation" and consider their theory to be a relational

approach to the understanding of female psychological development and the importance

of relationships in the lives of girls and women (Miller & Stiver, 1997). They have

identified three major growth-fostering characteristics of relationships, which have been

supported by previous research. They are mutual engagement, authenticity, and

empowerment or zest (Liang et al., 2000).

Mutual engagement is defined as perceived mutual involvement, commitment,

and attunement to the relationship (Surrey, 1991b). Mutual engagement may have a

beneficial impact on individuals as well as relationships as indicated in studies on

empathy and closeness (Liang et al., 2000). These qualities mediate stress and depression

and are associated with self-actualization, self-esteem, low interpersonal distress, and

relationship satisfaction (Beeber, 1998; Burnett & Demnar, 1996; Gawronski & Privette,

1997; Schreurs & Buunk, 1996; Sheffield, Carey, Patenaude, & Lambert, 1995).

Authenticity is not a static state that is achieved at a discrete moment in time, but rather a

girl's or woman's ongoing ability to represent herself in a relationship with increasing

truth and fullness and thereby acquire knowledge of self and other (Miller & Stiver,

1997). Research on authentic self-disclosure and openness appears to be related to being

liked, increased liking of others, and motivation in relationships (Collins & Miller, 1994;

Kay & Christophel, 1995). Empowerment or zest, the experience of feeling personally

strengthened, encouraged, and thereby motivated to take action (Liang et al., 2000), has

been shown to have a direct impact on positive affect, meaningful activity, and creativity

(Hall & Nelson, 1996; Spreitzer, 1995).








The Relational Model was initially conceptualized by Surrey (1985) as the

mutually interacting self or the "self-in-relation." She posited a theory of female

psychological development that rejected the notion that connectedness and differentiation

were dichotomous and viewed the overall maturation process as the development of a

complex, defined self within a structure of connected relationships. Surrey (1985)

contended that a young woman's self-esteem was dependent on her capacity to develop

relational competence, which included the capacity to experience and communicate

accurate empathy. The traditional definition of empathy is a temporary blurring of ego

boundaries which allows an individual to experience the affect of another followed by a

distancing then return to an objective position (Nelson, 1996). Jordan (1991), another

Stone Center theorist, contended that this perspective on empathy inaccurately

perpetuates seeing the self as either distinctly autonomous or "merged and imbedded."

Jordan (1991) proposed that it is possible for an individual to feel connected to another or

be effectively joined while at the same time appreciate and be fully aware of her own

separateness. She also contended that the skillful use of empathy requires well-defined

ego boundaries.

Relational competence as defined by the Relational Model is the ability to attend

to the affect and experience of another individual and then respond in an appropriate

manner that compromises neither that individual self nor another (Nelson, 1996).

Mutuality, authenticity, and empowerment or zest are the key components of this

relational skill (Liang et al., 2000). Relational competence leads to mutual

empowerment, a state in which each person can receive and then respond to the feelings

and thoughts of the other, each is able to enlarge both her own and another person's








feelings and thoughts, and simultaneously each person enlarges the relationship (Miller &

Stiver, 1997). Psychological health is therefore the outgrowth of connection with others

while psychological distress develops in response to repeated and chronic patterns of

disconnection (Miller, 1988). According to the principles of the Relational Model, the

goal of healthy psychological development for girls and women is attained via the

increasing ability to build and enlarge mutually enhancing relationships in which each

individual can feel an increased sense of well-being through being in touch with others

and finding ways to act on individual thoughts and feelings (Surrey, 1985).

Pollack and Gilligan (1982), in a study using Thematic Apperception Test (TAT)

cards, found that male subjects reported seeing more violent images when people were

physically brought closer together in TAT pictures. Women, on the other hand, related

more violent stories when people were set further apart. Pollack and Gilligan (1982)

concluded that closeness and relatedness form the context for girls' and women's

psychological development and exclusion and isolation are purposefully avoided.

In this research study, peer, mentor, and community relationships were

hypothesized as potential protective factors in Black, Latina, and White college women's

struggles with disordered eating. Among growth-fostering relationships, close ties with

peers, mentors, and belonging to a community are three of the most significant types of

connections for young adults (Liang et al., 2000). Rosen & Neumark-Sztainer (1998)

reviewed options for primary prevention of eating disturbances and stated that almost

nothing is known or written about the protective factors that may increase resilience to

the development of disordered eating. Protective factors are conceptualized as

characteristics in an individual's world that mitigate against the development of








behavioral and psychological problems despite the existence of risk factors (Rutter,

1979). For this population, relevant risk factors for the problems associated with

disordered eating are the sociocultural environment (Striegel-Moore, 1993) and the

stresses involved with adjusting to and successfully navigating the challenges of college

life (Martz & Bazzini, 1999). Women of color face additional hardships in college

adjustment especially in predominantly White university settings (Gloria & Rodriguez,

2000), therefore race and ethnicity was also considered a relevant risk factor.


Risk Factors for Disordered Eating in College Women

The attitudes and behaviors of the continuum of disordered eating generally

develop in adolescence and increase in prevalence as young women transition through

their college years (Alexander, 1998). Bulimia nervosa is virtually unheard of prior to

adolescence and the vast majority of women clinically diagnosed with bulimia nervosa

have symptom onset before the age of 25. Similarly, in clinical samples the modal age of

onset of binge eating is 18 and it is rarely seen in children. Anorexia nervosa does occur

in prepubertal children but the incidence increases dramatically after puberty, with the

majority of cases beginning before the age of 25 (Woodside & Garfinkel, 1992).

Researchers have estimated that 3% to 19% of college women have bulimia, 1%

to 2% have anorexia, and as many as 61% display subclinical eating disordered attitudes

and behaviors such as extreme body consciousness, chronic dieting, overexercising,

bingeing without purging, and/or purging without bingeing (Mintz & Betz, 1988). These

unhealthy symptoms may actually be considered relatively normative among

undergraduate women (Mazzeo, 1999). Striegel-Moore and Cachelin (1999) stated that

the etiology of disordered eating is multifactorial and varies for each individual girl








and/or young woman. If inroads are to be made in the area of primary prevention, it is

important to understand the risk variables for disordered eating in the female college

population.


Sociocultural Environment

Concerns with food and weight have become such a major sociocultural

preoccupation for girls and women in Western culture that it is almost impossible to grow

up female today without ever feeling fat, worrying about weight, and developing

ambivalent feelings towards food (Friedman, 1998). At a very young age, girls are

socialized to hate obesity and to accept a cultural standard of thinness that is close to or

below the minimum required weight for reproduction (Wooley & Wooley, 1982).

Adolescent girls receive the message to be thin no matter what the costs, to deny their

needs and appetites to achieve this goal, and to deny their selves and their bodies to

please others (Mirkin, 1990).

Collins (1991) examined Black and White elementary school girls and found that

a majority of the subjects expressed a bias toward thinner child and adult figures and 42%

preferred body figures that were different and thinner than their own perceived body

shape. This finding was true across all age, weight, and ethnic groups. Collins (1991)

concluded that the onset of disparate figure perceptions and expectations regarding

thinness among females might be evident as young as 6 or 7 years of age.

Puberty and adolescence appear to be critical developmental stages for the

evolving preoccupation with body shape and weight (Surrey, 1991). Adolescent girls'

growth spurt, which involves the normal tendency to gain weight and increase their body

fat, seems to be an important factor in this growing preoccupation (Wooley & Wooley,








1982). Among many young women in a Western culture that values thinness, weight

gain and the personal experience of 'getting fatter' seem to initiate psychological

disturbances in body image and attempts toward weight reduction (Surrey, 1991).

Although girls tend to naturally gain weight and body fat during puberty, Western culture

encourages girls to strive for an extremely thin body ideal and then judges them harshly

according to this unrealistic standard (Steiner-Adair, 1990).

Clausen (1975) stated that there is a correlation between female adolescent body

build and actual positive and negative evaluations, prestige, and success in relationships.

While boys were given social approval for academic success and achievement, girls were

praised for their physical appearance and more specifically for being slim (Clausen,

1975). Rosenbaum (1993), in a study of adolescent girls, found all subjects preferred

small, unobtrusive body parts with the exception of a preference for large breasts. In

addition to which, given three wishes about what they would change about their bodies,

the girls consistently chose these priorities: (a) to lose weight and keep it off, (b) blond

hair and blue eyes, (c) a clear complexion and a perfect figure (Rosenbaum, 1993).

Robinson, Killen, Litt, and Hammer (1996) surveyed 939 6th and 7th grade girls

aged 10 to 14 years old in order to assess body dissatisfaction. Latina girls reported

significantly greater body dissatisfaction than White girls even among the leanest 25% of

the subjects. Striegel-Moore, Schreiber, Pike, and Wilfley (1995) examined 311 Black

and 302 White girls aged 9 to 20 years old for racial and food intake differences looking

at the dynamic of the drive for thinness, a variable linked to the etiology of eating

disorders. Black girls in the study reported significantly greater drive for thinness than

White girls. In an Essence magazine study of 600 randomly selected Black females aged








17-62 years old, results showed that surveyed subjects had adopted attitudes toward body

image, weight, and eating, and suffered from levels of depression that were comparable

to their White counterparts. These factors also served to increase their risk for eating

disorders (Pumariega, Gustavson, Gustavson, & Motes, 1994). Root (1990) states that

racial/cultural context may afford protection from disordered eating to an ethnic group,

but it does not necessarily protect individuals who are subject to the standards of the

dominant Western culture. Levels of acculturation and assimilation are also important

variables to be considered in the relationship of disordered eating and women of color.

Thompson (1992) labeled our current sociocultural environment as a 'culture of

thinness,' an environment which glorifies thinness as the ticket to happiness and success

and denigrates overweight bodies by linking them to such negative characteristics as

laziness, ugliness, and failure. Antifat prejudice is so acceptable in our society that

Crandall (1991) stated that it is a better element for studying the dynamics of

discrimination and prejudice than is the study of racism or sexism since the latter are

influenced by social desirability factors.

Images of supposedly attractive women such as beauty pageant winners and

Playboy centerfolds have become progressively thinner since the 1960's (Garner,

Garfinkel, Schwartz, & Thompson, 1980; Wiseman, Gray, Mosimann, & Ahrens, 1992).

This contemporary sociocultural phenomenon of thinner and thinner female bodies has

also been prevalent in female fashion models employed by leading modeling agencies

(Morris, Cooper, & Cooper, 1989).

American women report a preoccupation with body weight that begins before

puberty and intensifies in adolescence and young adulthood (Brumberg, 1988). In a 1984








Glamour magazine survey of 33,000 women, when asked to choose among potential

sources of happiness, the majority of respondents chose weight loss over success at work

or in interpersonal relationships. Eighty percent of these surveyed women also believed

that they had to be thin in order to attract men (Brumberg, 1988). This 'perfect female

weight' represented by media image ideals has progressively decreased to that of the

thinnest 5-10% of American women and consequently, 90-95% of American women feel

as if they'll never be able to 'measure up' physically (Seid, 1994).

Individuals who adhere to popular stereotypes of female beauty are more likely to

exhibit thoughts and behaviors associated with disordered eating (Hesse-Biber, 1991;

Mintz & Betz, 1988). Garfinkel and Garner (1982) found that women who were dancers

and fashion models and therefore experienced a heightened pressure to conform to a thin

body shape were at greater risk for developing eating disorders. Heinberg & Thompson

(1995) determined that direct exposure to media-communicated images of thin, attractive

women via print or film materials produced an immediate increase in women's body

dissatisfaction and dysphoric mood.

Steiner-Adair (1990) interviewed 32 White girls aged 14 to 18 years old in order

to examine their perceptions of cultural values and cultural and individual images of

women. Three weeks after the interviews were finished, girls completed the Eating

Attitudes Test (EAT), which assessed eating disordered behavior. Interpretation of the

interviews revealed two different patterns of responses: the wise woman pattern and the

superwoman pattern. Sixty percent of the girls fit into the wise woman response pattern,

indicating a personal awareness of the cultural expectations and values of a woman's

autonomy and independent achievement in career and appearance, yet differentiating








their own individual ideal from this societal image. Forty percent of the girls fell into the

superwoman response pattern, identifying with the independent and autonomous

superwoman as the societal ideal as well as their own. All of the wise women scored in

the noneating disorder range of the EAT in contrast to the 11 of 12 superwomen who

scored in the eating disorder range. Steiner-Adair (1990) concluded that those girls who

uncritically accept a societal image that conflicts with their own female developmental

history of connection are at a greater risk for developing eating disorders.

In another research study looking at body image, Schwartz and Thompson (1982)

examined the effects of the thin sociocultural ideal on "normal" college women and

found that a high percentage of these young women engaged in disordered eating

behaviors including both binge eating and purging. They concluded that there is a large

percentage of functioning female college students who engage in eating disordered

behaviors in order to stay thin. Although only a minority of adolescent girls develops

clinical eating disorders, a large majority of these young women diet and suffer from

subclinical eating concerns, body dissatisfaction, and what has been called a "normative

discontent" (Rodin, Silberstein, Striegel-Moore, 1985).

The prevalence of the continuum of disordered eating attitudes and behaviors in a

diverse population of female college students across the nation has been documented in

various research studies (e.g., Abrams et al., 1993; Akan & Grilo, 1995; Alexander, 1998;

Drewnowski et al., 1988; Gray et al., 1987; Hesse-Biber et al., 1999; Pyle et al., 1991).

The exact numbers of affected college females varies widely but Riley (1991) stated that

eating disorders are among the most rapidly increasing diseases of our time. In response

to this pervasive problem, Striegel-Moore & Cachelin (1999) stated that intervention








studies are clearly needed that are based on models of risk and resilience in the arena of

disordered eating. Within the current body of eating disorder research, the sociocultural

environment as a possible risk factor is well documented. In contrast, the variables of

adjustment to college for young women have not been empirically explored as potential

risk factors for disordered eating.


Adjustment to College

Attending college away from home for the first time entails a major transition for

late adolescents which is fraught with a great deal of stress (Wintre & Yaffe, 2000).

Students envision a new life free of parental control, filled with interesting and novel

experiences, with new people to meet and stimulating academic activities (Pancer,

Hunsberger, Pratt, & Alisat, 2000). The reality of students' experiences at a college or

university is actually harsher and more stressful than many late adolescents ever

anticipate (Compas, Wagner, Slavin, & Vannatta, 1986).

Students need to adjust to the new responsibilities of young adult independence

and cope with an environment that is very different from the one they have been used to

at home during their high school years (Pancer et al., 2000). Usually students are moving

away from home for the first time and are therefore cut off from their existing support

network made up of extended family and close friends (Rice, 1992). In addition, they are

forced to perform many tasks that were formerly done by their parents such as doing their

own laundry and managing their finances (Koplik & Devito, 1986). On top of all of these

challenges, their collegiate academic requirements are usually more difficult than the

work they did in high school (Pancer et al., 2000).








Most of the work in establishing an independent identity for young adults is

completed during college rather than during high school. The majority of college

students spend at least part of their college years in a state of moratorium. During this

period of moratorium, a student actively explores options for values and beliefs

concerning politics, religion, career, etc. and women also consider family and gender

roles. This developmental stage is both exciting and challenging but can also be

psychologically stressful for most college students (Smolak & Levine, 1996).

Chickering's (1969) model of college students' psychosocial development has

been widely researched and is a structured way of conceptualizing this developmental

period. It was recently updated (Chickering & Reisser, 1993) and includes the following

seven vectors: (a) developing competence, (b) managing emotions, (c) moving through

autonomy toward interdependence, (d) developing mature interpersonal relationships, (e)

establishing identity, (e) developing purpose, and (f) developing integrity. This model

defines the development of mature interpersonal relationships via tolerance and

appreciation of individual differences and a capacity for intimacy. These dynamics are

similar to the growth-fostering characteristic of mutual engagement within the Relational

Model. Chickering's fifth dimension of establishing identity includes the need to find

comfort with body and appearance and self-acceptance and self-esteem (Chickering &

Reisser, 1993). This developmental task directly relates to college women's struggles

with their sociocultural environment and the continuum of disordered eating. Chickering

and Reisser's (1993) updated version of student developmental theory recognizes the

importance of students' experiences with relationships in the formation of their core

sense of self.








Although some students adjust fairly well to these major developmental

challenges, others are overwhelmed and are unable to make use of positive coping skills.

Many students experience physical and/or emotional problems related to the struggles of

college adjustment (Wintre & Yaffe, 2000). Striegel-Moore and Steiner-Adair (1998)

stated that adolescent young women in response to a wide range of difficulties and

stressors in life might often use disordered eating as a coping mechanism. The multiple

challenges of adjusting to college life and feeling successful in the university

environment would be considered a major life stressor for many young adult women and

therefore could be a potential risk factor for disordered eating.

Certain facets of the college sociocultural environment may also contribute to

college females' vulnerability to a continuum of attitudes and behaviors related to

disordered eating (Martz & Bazzini, 1999). Hesse-Biber (1996) stated that there are

several reasons why the college environment may be a breeding ground for weight

obsession and eating problems. College provides a "semi-closed" environment that tends

to amplify sociocultural messages. Disordered eating is often spread through imitation,

competition, and/or solidarity, sociocultural experiences, which are commonplace on

university campuses. In addition, weight gained during college is especially detrimental

in a climate primed to value thinness and often begins the vicious cycle of disordered

eating. Stress leads to overeating, which leads to weight gain, which leads to restricted

eating, binge/purge cycles, and/or overexercising, which may lead to more stress if these

compensatory behaviors fail to work (Hesse-Biber, 1996).

Because college is a time in which dating serves an important social function,

the constant pressure on young women to be attractive may be particularly salient during









these years (Martz & Bazzini, 1999). Low body weight has become critical in defining

attractiveness for these young women and their appearance is an important determinant of

attracting and dating young men (Rodin et al., 1985). Therefore, this period of a young

woman's life may present stress-related pressures to look "perfect" in order to appeal to

and attract friends and potential romantic partners (Martz & Bazzini, 1999).

Epidemiology suggests an increased incidence of disordered eating since the 1960's for

women in their 20's and 30's. Many of the struggles associated with the entire arena of

problem eating for these women oftentimes began in their college years (Pawluck &

Gorey, 1998).

In summary, the transition involved in the academic and social adjustment to

college for residential students is oftentimes very challenging. Many students are unable

to make use of positive coping mechanisms and as a result end up dealing with physical

and/or psychological problems. This transition is especially difficult for young women as

they face on-going pressures to be attractive in order to attract both friends and dating

partners. Many young college women become involved in the continuum of disordered

eating as they strive for the "perfect" body. Adjustment to college life on predominantly

White university campuses may be even more difficult for minority women than for

White female students as they face additional challenges associated with their race and/or

ethnicity.


Race and Ethnicity

The number of both Black and Latina female college students enrolling in

colleges is growing and most of these young women attend predominantly White

institutions of higher education. Their minority status in these settings usually translates








into different academic and personal experiences than those encountered by their White

female student counterparts (Gloria & Rodriguez, 2000; Schwitzer, Griffin, Ancis, &

Thomas, 1999). Although all students must contend with academic stresses and

adjustment difficulties, coping with the challenges of college life is generally more

difficult for racial/ethnic minority students than for White students (Gloria & Rodriguez,

2000). Ponterotto (1990) stated that many Black and Latina students often feel

unwelcomed and unappreciated on predominantly White university campuses. There is

little written in the literature specifically on Black and Latina female college student

adjustment, therefore the majority of the following studies are based on samples of both

male and female students of color.

Using qualitative research methods, Schwitzer et al. (1999) described a model that

identified four key features that tend to constitute African American students' social

adjustment to college experiences. Focus group participants in this study were 22

traditionally aged 4th year seniors, a majority of which were female, at a predominantly

White university. The first two features in the model relate to aspects of adjustment to

the institutional climate as a whole and are labeled "sense of underrepresentedness" and

"direct perceptions of racism." The next two features refer to specific influences on

academic relationships with faculty. In this study, respondents reported feeling

"unsupported" and "different" and stated that the transition to the institution's social

climate had been "hard," "difficult," "a struggle," or "unhappy." They also felt

"isolated," "frustrated," "overlooked," or "misunderstood" by others on campus because

of their race.








This four-feature descriptive model for Black students described by Schwitzer et

al. (1999) is consistent with the literature on other models of successful college

matriculation and persistence. These models focus on two central tasks involved in

college adjustment which are establishing successful interpersonal relationships in the

campus environment also known as social adjustment and effectively interacting with

faculty in and outside the classroom which is a component of academic adjustment

(Baker, McNeil, & Siryk, 1985; Baker& Siryk, 1984).

Gossett, Cuyjet, and Cockriel (1998) reported that 324 African American students

attending four midwestern universities felt marginalized in their campus environment.

This marginalization occurred in diverse situations involving administrative staff,

academic advising, classroom activities, faculty interactions, peer interactions, and

student services. The college environment, particularly when perceived as

discriminatory, hostile, alienating, or isolating, can be a major factor in the impediment

of African American students' participation and persistence in higher education (Allen,

1992).

Cervantes (1988) found that many African American and Latino students felt a

need to assimilate into the White university culture in order to be accepted. 'Blending in'

required them to hide and/or disown their ethnic background, which resulted in feelings

of isolation, cultural alienation, and an overall sense of being unwelcomed in higher

education because of their cultural, racial, or ethnic differences.

Sedlacek's (1999) meta-analysis of twenty years of research on Black students on

White college campuses found that racism, both individual and institutional, remains a

major problem for Black students across the country. Institutional racism involves








university policies and procedures, either formal or informal, which result in negative

outcomes for Black students. This type of racism is often more of a problem for Black

students than is racism expressed and/or acted out by individuals (Sedlacek, 1999). A

common example of institutional racism is the separate African American and White

fraternity and sorority systems in which African American organizations are stereotyped

and must compete with the larger White Greek system for university funding (Schwitzer

et al., 1999).

Smedley, Myers, and Harrell's (1993) study, which included a diverse sample of

Hispanic and Latino students, reported that students on predominantly White college

campuses faced stressors associated with their minority status. They exhibited

considerable psychological sensitivity to the campus social environment including

interpersonal tensions with White students and faculty and actual or perceived

experiences of racism and discrimination. Among Latino sophomores and juniors

attending several different universities, perceptions of racial-ethnic tensions and

experiences of discrimination affected numerous dimensions of adjustment to their

college experiences (Hurtado, Carter, & Spuler, 1996).

In addition to the stresses of racism, many Latino students also struggle with the

psychological and social impact of functioning within two different cultures (Torres,

1999). Feeling isolated, alienated, or "nonentitled," Latino students are often faced with

the need to adopt a bicultural understanding of themselves as Latino young adults

attempting to navigate the White cultural world of academia (DeFreece, 1987).

Bicultural adaptation is an example of a healthy coping strategy, but one that can be very

stressful as Latino students are challenged to maintain their cultural values and identity








while adjusting to the university culture and environment (Gloria & Robinson Kurpius,

1996).

In summary, the transition to college, which involves social, emotional, and

academic adjustments, is a normal but often very stressful life event. The stressors

associated with that transition might be greatest for students attending a college or

university where the predominant racial and/or ethnic culture differs from their own. The

aforementioned research studies predominantly examined the variables involved in

academic persistence for Black and Latino college students without looking specifically

at the unique adjustment problems of minority women. The pressures and strain of

college, from academics to social life, magnify female students' problems with body

image and weight because women often use food or restriction of food as a means of

calming and/or coping (Hesse-Biber, 1996). A Black or Latina female college student

may be especially vulnerable to the continuum of disordered eating as she copes with the

additional stressors associated with her minority status within a White college or

university environment.


Protective Factors Against Disordered Eating in College Women

Epidemiologists, after studying thousands of lives through time, have consistently

found that close, intact relationships predict health. Compared with those with few social

ties, people supported by friends, family, fellow members at school, work, church or

other support groups are less vulnerable to ill health and premature death (Myers, 2000).

The problem of disordered eating in female college students is a serious health concern

but there has been a paucity of research specifically focused on eating pathology and

relationship connections (Steiner-Adair, 1990). The Stone Center's Relational Model








proposes that relational health within peer, mentor, and community relationships is

associated with mental health and adjustment in college-aged women (Liang et al., 2000).

This research study hypothesized that relational health is a protective factor in the

development of attitudes and behaviors associated with the continuum of disordered

eating in Black, Latina, and White female college students.


Peer Relationships

Positive experiences of friendship contribute significantly to cognitive, social, and

moral development as well as to psychological adjustment and socioemotional health for

adolescents (Brown, Way, & Duff, 1999). In a study of 1,131 fifteen to sixteen-year-old

teenagers, Vilhjalmsson (1994) found that parental support had the largest effect on self-

assessed health, closely followed by friend support, and support by other adult figures.

However, Eder (1985) more than a decade ago, stated that very few studies have focused

on female peer relationships during adolescence and even less research has looked at

girls' friendships with other girls. This statement remains true even today especially in

the field of psychological research. Despite evidence that peer relationships are critically

important for adolescents and young adults, little is known about the intricacies, nuances,

and contexts of female friendships especially in young women of color (Brown et al.,

1999).

An increasing number of theorists have noted the different 'cultures' that are

manifest in male and female peer relationships (Bukowski, Newcomb, & Hartup, 1996).

Maccoby (1990) stated that interactions between females place priority on the building of

interpersonal connections, whereas interactions between males are more directed toward

the enhancement of individual status. In addition, the socializing influences of peer








relationships rival, and in some respects surpass, the socializing influence of parental

figures (Maccoby, 1990). Young females tend to evaluate close friendships with other

females equally as intimate or even more intimate than they rate their relationships with

their parents (Blyth & Foster-Clark, 1987).

Beginning in early adolescence, girls report more frequent interactions of an

intimate and supportive nature with female friends than boys do with their male friends

(Johnson et al, 1999). Wright (1982) characterized female friendships as "face-to-face"

with an emphasis on talking and male friendships as having a "side-by-side" orientation

focused on doing things together such as sports and competitive games. Furman and

Buhrmester (1992) studied 200 12- to 15-year-old male and female adolescents and asked

them via telephone interviews to recount the social events of the preceding 24-hour

period. They found that female friendships provided more opportunity than male

friendships for the fulfillment of interpersonal or communal needs. Females reported

more frequent interactions with friends, and also substantially higher levels of self-

disclosure and emotional support in daily interactions (Furman & Buhrmester, 1992).

Waldrop and Halverson (1975) found that young girls who had "intense and intimate"

friendships were rated as more socially mature than other girls, whereas young boys who

were rated as more socially mature had a "greater number" of friends.

The results of several large prospective epidemiological studies such as the

Alameda County Population Monitoring Study (Berkman & Syme, 1979), the Tecumseh

Community Health Study (House et al., 1982), and the Evans County, Georgia Study

(Schoenbach et al., 1986) highlight the significance in the connection of physical health

and social support network size for men but not for women. Women tended to benefit








more from relationships with other female friends and relatives who are more intimate

and nurturing thus providing better emotional support. Wheeler et al. (1983) stated that

high numbers of social contacts do not ward off feelings of loneliness for an individual.

Only when these relationships involve emotional intimacy and disclosure, a person no

longer experiences loneliness.

VanderVoort (1999) studied 280 college undergraduate students in order to

examine the relationship between social support and mental and physical health. In this

study, poor functional support or quality of support was related to mental and physical

health problems while structural support or social network size was not. Women in this

study reported significantly more satisfaction with their social support systems and less

feelings of isolation. VanderVoort (1999) concluded that intimate, emotional support

meets our emotional needs by enabling individuals to feel valued as well as process or

work through their emotional difficulties. Processing feelings and meeting one's

emotional needs also helps ward off chronic negative affects such as depression, anxiety,

and hostility which have been shown to be related to poor physical health (VanderVoort,

1995).

A consistent finding of a variety of research studies is that the integrating

experiences of involvement, engagement, and affiliation are central to students' healthy

development and progress in college (Hurtado & Carter, 1997). Schwitzer et al. (1999)

stated that a critical factor in the retention and success of African American students at

predominantly White universities is the individual student's experience of the campus

social environment. Academic, institutional, personal-emotional, and social adjustments

are the major demands facing all college students. Of these four, adjusting to the social








environment seems to be central to the success of many Black students in White

institutional settings (Schwitzer et al., 1999).

Gloria, Robinson, Kurpius, Hamilton, and Wilson (1999) examined the influence

of social support, university comfort, and self-beliefs on the persistence decisions of 98

African American undergraduates enrolled in a predominantly White university. The

sample included nearly three times as many females as males. Each of the three

constructs significantly predicted persistence, with social support and university comfort

as being the strongest predictors. Watson and Kuh (1996) found that the quality of

African American students' relationships with peers, faculty, and administrators tended to

be as important as individual academic effort in their scholastic achievement.

Additionally, descriptions of ethnic/racial minority families as cohesive and

interdependent contribute to an expectation that positive interpersonal attachments are

salient to the psychological adjustment of Black college students (Kenny & Perez, 1996).

Conversely, negative interpersonal experiences in predominantly White university

settings can limit or mediate the ability of some Black students to engage in learning,

developmental programs, and other valuable opportunities that are an integral part of

campus life (Schwitzer et al., 1999). A common example of negative interpersonal

experiences are African American students living on campus and being confronted with

unwelcoming residence hall environments, less friendly peers, and racial problems which

were undetected by other White college students in similar residential situations

(Johnson-Durgans, 1994).

A Latino teenager's experience of growing up in America has been described as

entremundos meaning between two worlds (Falicov, 1998). This is often an uneasy








period of coexistence between two very different cultural orientations, languages, sets of

values, and philosophies of life. The development of a coherent ethnic identity, which

encompasses knowing and valuing who one is socially and ethnically, is critical to

effective coping and a healthy outlook on life (Falicov, 1998). However, the degree to

which Latino college students adhere to cultural prescription varies by age, proximity to

culture and family, and interactions with friends and others of similar ethnic background

(Sodowsky, Lai, & Plake, 1991). Higher education, with its inherent exposure to

differing cultures and new experiences, may either solidify or threaten Latino students'

abilities to function in a healthy and effective way (Ethier & Deaux, 1990).

Gloria and Rodriguez (2000) stated that in struggling with their cultural identity

and formulating their role and function in the collegiate environment, Latino students

have found that different forms of social support are very helpful. Social support refers

to the helpfulness of social relationships and is a recognizable buffer to the negative

influences of stressful events and depression in Latinos (Briones, Heller, Chalfant,

Roberts, Aguirre-Hauchbaum, & Farr, 1990). For Latinos, family is a primary means of

social support and adherence tofamilismo, a strong sense of family centrality and

importance, is a core value. Latinos in general do not value individuation from one's

parents and family to the same extent as those of the dominant culture. Greater value is

placed instead on relationships in which individual needs are secondary to the welfare of

the family or group (Gloria & Rodriguez, 2000). For young adults in college who are

away from home and family for the first time, peer relationships take on additional

importance as a primary social support system that facilitates psychological and physical

well-being. Close friendships, considered by many social scientists to be the most








satisfying and rewarding of all human relationships, are very important for the social,

emotional, and physical health of all adolescents and young adults regardless of ethnicity

or race (Brown et al., 1999).

In looking at this important developmental stage of later adolescence, Striegel-

Moore and Smolak (1996) stated that several studies have reported a link between

stressful life events such as college adjustment and the onset of eating disorders. Strober

(1984) found that the magnitude of life stress experienced by eating disordered patients

18 months prior to the onset of their disorder was 2.5 times greater than that of a

normative sample of female adolescents. Smolak, Levine, and Gralen (1993) found that

girls whose transition to middle school coincided with other stressors such as the onset of

dating and menarche were significantly more likely to have elevated scores on the EAT-

26 (Garner et al., 1982) when compared with girls whose school transition was not

accompanied by other stressors. Smolak and Levine (1996) stated that the availability of

social support as an adolescent enters middle school or moves away to college may either

buffer or intensify the stresses of these pivotal adolescent transitions and thereby affect

potential eating pathology.

In summary, there is a relationship between physical and psychological health and

peer relationships for adolescents and young adults. Especially for young women,

support that is intimate and nurturing is most beneficial and in addition, reflects the

Relational Model characteristics of mutual engagement, authenticity, and empowerment

or zest. This research finding is also valid for Black and Latina women who struggle

with adjustment to predominantly White university environments. Since there is a

paucity of research on peer relationships and disordered eating, this study hypothesized








that peer relationships may also act as a protective factor in disordered eating in ways that

are similar to the positive effects of social support on social, emotional, and physical

health. In addition to peer relationships, mentoring relationships may also have a

significant influence in the well-being of female college students.


Mentor Relationships

Adolescents consistently identify non-parental adults as playing a very important

role in their psychological development (Blyth, Hill, & Smith, 1982). Despite this fact,

there is a paucity of research on the impact and nature of mentoring relationships for

adolescents and young adults and even less research on female mentoring relationships.

Most of the literature on mentoring has focused exclusively on adult professional

development and career advancement (Carden, 1990).

The word "mentor" is derived from the classical Greek character Mentor, an old

and trusted friend of Ulysses who was charged with the care of Telemachus, Ulysses'

son. A definition and model for mentoring evolved from this first mentor as the

development of a relationship that is characterized by an individually delivered and

intentional process that is supportive, nurturing, insightful, and protective (Scott, 1992).

Over the past several decades, mentoring in America has become an integral part of the

business world. Senior executives in big business who had a mentoring relationship

reached their positions at a younger age and earned a higher income (Roche, 1979).

Businesses continue to be firmly entrenched in the benefits of mentoring and support

formal programs to develop mentoring relationships (Scott, 1992).

In his research study based on longitudinal interviews with 40 men in varied

career pursuits, Yale developmental psychologist Levinson (1978) described mentoring








as a form of a love relationship. Functioning as guides, teachers, and sponsors, mentors

help their protegds realize their goals and dreams in life. Levinson (1978) maintained

that the mentoring relationship is one of the most important developmental relationships a

person can have in early adulthood. This psychosocial view of mentoring from the

prospective of adult growth and development differs from the concept of mentoring in the

business world (Beck, 1989) and is more closely aligned with mentoring in an

educational setting. In the field of education, the main focus of the mentor-prot6g6

relationship is experiential learning. Mentors in education facilitate learning by acting as

teachers, guides, counselors, role models, and friends (Beck, 1989).

A review of the literature describing beneficial mentoring relationships in

educational settings identified three general categories of components that are helpful to

female college students: (a) psychosocial support, (b) role modeling, and (c) professional

development (Jacobi, 1991). Within the psychosocial arena, supportive mentors have

offered opportunity for growth of self-awareness and identities, which have resulted in,

increased assertiveness, positive presentation of self, and high career expectations (Bruce,

1995). Role modeling affords young women the opportunity to see other women in a

variety of situations successfully balancing career and personal goals and thereby

challenging sociocultural beliefs and attitudes, which are self-limiting (Shakeshaft,

Gilligan, Pierce, 1984). Finally, mentors have assisted in professional development by

offering visibility, protection, and sponsorship as well as facilitating student interaction

with a variety of people in career promoting endeavors (Bruce, 1995).

Soucy and Larose (2000) studied 158 academically at-risk adolescents (63 males,

95 females, 16-20 years old) in order to determine the value of mentoring contexts as








determinants of adolescent adjustment. The students completed questionnaires twice

during their first semester of college, once before and once again after they participated

in a mentoring program. Soucy and Larose (2000) found that the perception of a secure

relationship with a mentor was predictive of adolescent adjustment to college. Nora,

Cabrera, Hagedorn, and Pascarella (1996) in a national study of 3,900 freshman college

students found that for females only, the most significant positive effect on college

persistence came from mentoring experiences in the form of nonclassroom interactions

with faculty.

Despite the widely acknowledged benefits to both prot6g6 and mentor, mentoring

has been focused primarily within the population of White men. European American

women and ethnic minorities of both genders have historically been underrepresented in

the mentoring process (Atkinson, Casas, & Neville, 1994). According to Blackwell

(1989), mentors tend to select prot6g6s who are of the same gender and who share social

and cultural attributes such as race, ethnicity, religion, and social class. This tendency of

mentors to select same sex and ethnicity prot6g6s is problematic in educational settings

because women and ethnic minorities are so underrepresented among university faculty

and staff positions (Atkinson et al., 1994).

Having ethnically similar mentors who have successfully traversed the academic

environment may create a sense of vicarious self-efficacy or a belief in one's ability to

persist in their pursuit of higher education especially for minority students on

predominantly White college campuses (Gloria et al., 1999). Absence of powerful Black

figures as role models has strong effects on the feelings of loneliness and isolation of

Black students. Because Black students are dealing with racism and face a difficult








adjustment to a White university, they are particularly in need of a person that they can

turn to for advice and guidance (Sedlacek, 1999).

Latino students in the university environment face a similar dilemma. Gloria and

Rodriguez (2000) stated that role models and mentors not only serve as primary social

support for Latino students, they also provide help in increasing academic persistence.

An analysis by the Hispanic Association of Colleges and Universities (1995) reported a

faculty to student ratio of 1 to 76 for Latinos, compared with a ratio of 1 to 54 for African

Americans, and 1 to 24 for White students. Although empirical research with Latino

students in mentor-protdg6 relationships is limited, Fiske (1988) suggested that Latino

students and other racial/ethnic minority students who attend predominantly White

institutions can be positively guided through their experience of "culture shock"

stemming from being on their own, overt and covert discrimination, and the loneliness

and tensions inherent in finding their way within an alien culture. Gloria and Rodriguez

(2000) stated that Latino students who have a mentor who takes personal and academic

interest in their educational experiences are more likely to succeed in the university

environment.

Whereas the traditional male-to-male mentoring models within the business world

rely on an acceptance of hierarchy and focus on task activities, females appear to desire

more psychosocial and emotional support in their mentoring relationships (Kalbfleisch &

Keyton, 1995). In support of this premise, Ball (1989) suggested that a good mentor for a

woman (and maybe a man) is more than a good role model, a mentor is also a teacher, a

sounding board, a cheerleader, and a friend. Kalbfleisch and Keyton (1995) stated that if

gender plays a role in accounting for differences in friendships, it is also likely that the








mentoring experience will be different as well. In support of this contention, Reich

(1986) found that more women than men noted that their relationships with mentors

(67% versus 42%) and prot6g6s (63% versus 44%) developed into close friendships.

In comparing male-to-female and female-to-female mentoring relationships,

Jeruchim and Shapiro (1992) stated that affective, or emotional quality is more vital for

women that for men. More importantly, developing intimacy in a female mentoring

relationship yields increased levels of productivity and development for the relationship

while avoiding the negative effects of possible sexual overtones. Worell and Remer

(1992) underscored the therapeutic nature of female-to-female mentoring relationships

and stated that women may receive ancillary benefits beyond those normally accrued

through the mentoring process. Kalbfleisch and Keyton (1995) stated that greater

intimacy, based on sharing, self-disclosure, listening, and building rapport is more likely

to build stronger mentor-prot6g6 relationships. These are some of the same qualities

identified as growth-producing in Relational Model relationships.

Kalbfleisch and Keyton (1995) studied 56 mentor-prot6g6 relationship pairs in an

attempt to examine the dynamics of female mentoring relationships. They found that the

nature of these relationships closely reflected models of female friendship that are

characterized by emotional intimacy and differ from the more hierarchical and task-

oriented male mentorships and friendships. In another study of urban adolescent girls,

Leadbeater and Way (1996) found that those girls who had mentors that listened,

understood, and validated their experiences and feelings evidenced transformations in

personal confidence and ability and were encouraged to develop strategies of resistance








that maintained health. These mentoring relationships appeared to serve as a buffer

against a variety of social stressors.

Mentoring relationships in educational settings are helpful in the psychosocial

support of college women in addition to being influential in role modeling and

professional development. There has been little empirical research that has specifically

examined the effects of mentoring relationships on struggles with disordered eating. It is

possible that mentoring relationships that involve mutual engagement, authenticity, and

empowerment/zest may be a potential protective factor for female college students in

their struggles with the continuum of disordered eating.

Both peer relationships and mentoring relationships are defined by the dynamics

of dyadic communication. Another type of relationship structure, group or community

affiliation, has also been shown to have a beneficial impact on an individual's social and

psychological functioning.


Community Relationships

In addition to dyadic relationships such as close peer and mentor relationships,

community or group affiliation has been shown to have a significant impact on social,

psychological, and physical functioning. Community relationships contribute to an

individual's sense of belonging (Liang et al., 2000). Maslow (1954) stated that human

behavior could be explained as motivation to satisfy needs. He identified belonging as a

basic human need, ranking it third in his hierarchy. Anant (1966) posited that belonging

is the missing conceptual link in understanding mental health and mental illness from a

perspective of relationships and interactions. Anant's (1967) early empirical research on

belonging, described initially as the recognition and acceptance of a member by other








members in a group, suggested that there was an inverse relationship between the

construct of belonging and anxiety. Anant (1967) however, questioned the validity of his

belonging measure and stated that it may have tapped dependence rather than the

construct of belonging.

Narrative accounts of a sense of belonging depict its importance for psychological

and physical well-being (Hagerty & Patusky, 1995). Dasberg (1976) interviewed battle-

fatigued Israeli soldiers and reported descriptions of feelings of loss of belonging, of

being cut-off and uprooted, abandoned, rejected, and psychologically severed. He stated

that the lack of a sense of belonging was the common denominator in the soldiers' mental

breakdowns during war. World War II Holocaust child survivors reported that they felt

they did not belong anywhere in terms of country, social group, or age after being

rescued from the Nazi concentration camps (Kestenberg & Kestenberg, 1988).

Kestenberg and Kestenberg (1988) stated that belonging is an important component of

identity and object relations and that a child grows and develops a sense of belonging not

only to family, but also to community, cultural group, and nation.

Researchers who have studied the role of social relations in health promotion have

suggested that social network ties and social integration influence mortality (Hagerty &

Patusky, 1995). Between 1979 and 1994, there have been eight longitudinal community-

based prospective studies that reveal an association between social ties and mortality

rates from a broad range of diseases. Although there were substantial variations among

these studies in measurement of social relationships, in types of communities

investigated, and length of follow-up, the results were remarkably consistent. In almost

all cases, those individuals who were most socially isolated and disconnected were at








increased mortality risk (Berkman, 1995). Berkman (1995) stated that for social support

to be health promoting, it must provide not only a sense of belonging and intimacy, but it

must also help people to be more competent and self-efficacious. Similarly, Relational

Model theorists have also stated that the characteristics of growth-fostering relationships

empower individuals by increasing a sense of self-worth, vitality, validation, knowledge

of self and others, and a desire for further connection (Liang, 2000).

Hagerty, Lynch-Sauer, Patusky, Bouwsema, and Collier (1992) posited that sense

of belonging represents a unique relational phenomenon that is different from the singular

constructs of loneliness and alienation. They defined sense of belonging as the

experience of personal involvement within a particular environment or system so that the

persons feel themselves to be an integral part of that environment or system. Sense of

belonging was proposed to have two defining attributes: (1) valued involvement or the

experience of feeling valued, needed, or accepted, and (2) fit, the perception that the

individual's characteristics articulate with or compliment the environment or system

(Hagerty et al., 1992). Hagerty et al. (1996) in a study of 379 community college

students (59% female) postulated that a lower sense of belonging is related to poorer

psychological functioning, which can be represented by depression, loneliness, anxiety,

history of psychiatric treatment, and suicidality. They also stated that sense of belonging

seems to be more strongly related to both social and psychological functioning for

women that for men. In addition, women when compared to men were more likely to

report a sense of belonging due to their community relationships and involvement.

In a study of 31 clients diagnosed with and in treatment for major depression and

379 students in a midwestern community college, Hagerty and Williams (1999) examined








the effects of sense of belonging, social support, conflict, and loneliness on depression.

According to results of this study, a sense of belonging was a better predictor of

depression than perceived social support. Hagerty and Williams (1999) posited that

perceived social support refers to the perceived presence or absence of potentially

supportive relationships, while sense of belonging is related to the perception of self as

integrated within an interpersonal system. This experience of integration involves

multiple components of cognition, emotion, and behavior that speak to the quality and

specific characteristics of interpersonal relationships (Hagerty & Williams, 1999). This

construct of sense of belonging shares many of the defining characteristics of the Stone

Center's Relational Model of relational health.

Hurtado and Carter (1997) reported that a consistent finding of a variety of

research studies is that the integrating experiences of involvement, engagement, and

affiliation are central to students' development and progress in college. In their study of

272 Latino students (58.1% female and 41.9% male), Hurtado and Carter (1997) found

that membership in religious and social-community organizations were strongly

associated with students' sense of belonging within the university environment. In

addition, these organizational memberships seem to have strong external-to-campus

affiliations that helped students maintain some link with the communities that they were

familiar with before they entered college.

Sedlacek (1999) suggested that Black students need to have identification with

and be active in a community as part of their support system. This community may be on

or off campus, large or small, but will commonly be based on components of race and/or

culture. Because of racism, Black students have historically been excluded from being








full participants in many White organizations within the educational system even though

Blacks seem to be more community oriented that Whites (Sedlacek, 1999). Bohn (1973)

found that a high score on the Community scale of the California Psychological

Inventory was associated with Black student success in grades and college retention.

Davis (1991) studied 888 Black college undergraduates and stated that Black students'

college experiences were more favorable when there was involvement and participation

in extracurricular activities sponsored by campus organizations such as clubs, fraternities,

sororities, and interest groups. Attinasi (1989) stated that minority students who made

sense of their environments through group memberships that also helped them acquire

needed college skills, also benefited by being linked to the larger whole of campus life.

Harris (1992), in a qualitative study of 54 African American young women (14 to

25 years old), stated that a sense of belonging for the African American female is defined

by the cultural construct of interdependence which starts during childhood and involves a

familial type attachment to both family and community. This sense of belonging has

served as a strength for African Americans and has helped to mitigate feelings of

isolation and detachment (Harris, 1992).

Swift and O'Dougherty Wright (2000) studied whether specific functions of

social support buffered the relationship between different types of stressful life events

and anxiety and depression in 60 college women. They found that belonging support and

self-esteem support were the most significant buffers of specific stressors. Belonging

support, which was indicated by feelings that one has others with whom to engage in

activities, decreases a female's chances of experiencing symptoms of anxiety and

depression. Self-esteem support also buffered the relationship between interpersonal








events and anxiety and depression (Swift & O'Dougherty Wright, 2000). Relational

theory suggests that a woman's self-esteem is strongly affected by her ability to maintain

mutually intimate emotional connections within relationships (Miller & Stiver, 1997).

Within a community perspective, Spencer (1998) implemented a hospital-based

outpatient psychiatric treatment program for women based on the Relational Model. In

this group setting, healing occurred in the context of connected relationships and

therefore building connections among the participants was central to the success of the

program. In another group research study, Tantillo (1998) developed a relational

approach to group therapy for women with bulimia nervosa. Through promoting

validation, self-empathy, mutuality, and empowerment, the group members learned to

identify and change relational patterns that have kept them connected with food and

disconnected from themselves and others. The goal of treatment was to help women

move toward mutually empathic and empowering relationships inside and outside the

group (Tantillo, 1998).

Resilient adolescents and young adults are sociable and able to seek and garner

social support from a variety of sources (Johnson et al., 1999). In contrast, females who

struggle with disordered eating often report a sense of isolation, social anxiety,

impoverished relationships, public self-consciousness, and a failure to seek social support

(Szmukler, Dare, & Treasure, 1995). Research studies have shown that community

relationships that promote a sense of belonging and are defined by the growth-producing

qualities of the Relational Model have had a beneficial influence on the physical and

psychological health of women. Therefore, this research study hypothesized that








community relationships may be a potential protective factor for Black, Latina, and

White female college students in their struggles with disordered eating.


Chapter Summary

The research literature reviewed herein provided supporting evidence for a study

of relational health and disordered eating in Black, Latino, and White female college

students. The beneficial impact of different types of social support on the physical and

psychological health of young women was validated by various research studies.

Consistent with this research, social support via the growth-producing qualities of the

Relational Model was hypothesized to have a positive influence on college women's

struggles with the continuum of disordered eating.

This chapter included information about traditional theories of human

psychological development and the Wellesley College Stone Center's Relational Model

of female psychological development and well-being. The nationwide epidemic of

disordered eating on college campuses was described in addition to the salient risk

factors, which included sociocultural context, college adjustment, and racial and ethnic

factors. Possible protective factor against the struggles with disordered eating in female

college students were hypothesized as peer, mentor, and community relationships.

Disordered eating in the female college population has been researched

extensively, but there is a paucity of research focused specifically on these young women

and their relationships. In addition, research on the relationships of eating disordered

minority college women is virtually nonexistent. The importance of social support and

relationships to the physical and psychological health of girls and young women has been

well documented in the research literature. Therefore, this study hypothesized that






59

relationships also play a part in the dynamics of disordered eating for female college

students. This study provides researchers, health educators, counselors, and other health

care providers information about potential protective factors in the struggles with

disordered eating and can also lead to the development of more effective programs of

prevention, education, and intervention with eating disordered college women.














CHAPTER 3
METHODOLOGY


Overview

The purpose of this study was to examine the relationship between White, Black,

and Latina college women's peer, mentor, and community relationships and their

disordered eating attitudes and behaviors. A correlational design for survey research was

utilized. Data were gathered on peer, mentor, and community relationships, as measured

by the Relational Health Indices (RHI) and on eating disordered attitudes and behaviors

as measured by the Eating Attitudes Test-26 (EAT-26).

This chapter describes the methodology employed in this study. It includes

descriptions of the population, sample, sampling procedures, research design,

instrumentation, and data analysis. The chapter concludes with a discussion of the

methodological limitations of this study.


Population

The population from which the sample for this study was drawn is undergraduate

White, Black, and Latina female students from a large co-ed residential university in the

southeastern United States. Total university enrollment for the fall 1999 semester was

44,276 students with 75% of this number being undergraduate students. The ratio of

women to men is 51:49. Black student enrollment is approximately 6.5% or 2,900








students and Latino/a students number approximately 4,200 or 9.4% of the total student

population (University of Florida, 2000).


Sampling and Sampling Procedures

The mail survey packet was sent to a stratified random sample of 480 female

undergraduate students. This sample consisted of three groups, 160 in each group, of

White, Black, and Latina female undergraduate students. Separate lists of current female

White, Black, and Latina undergraduate students were acquired from the university

Registrar's office. Freshmen were not included since data collection occurred in the fall

semester and it was proposed that these students would not have had enough time at the

university to establish collegiate relationships. Recent community college transfers were

also eliminated for this same reason.

After obtaining IRB approval for the sample and the methodology, each randomly

selected student was mailed a packet which contained a cover-request to participate letter,

two assessment instruments with specific instructions for completion, and a return self-

addressed stamped envelope One week later a follow-up postcard was sent to each

randomly selected student. Follow-up emails and phone calls were also made to students

that had not yet returned the mail survey packets.


Design

The design for this study was a correlational design for survey research. The

independent variables were White female peer, mentor, and community relationships,

Black female peer, mentor, and community relationships, and Latina female peer, mentor,

and community relationships. The dependent variable was problem eating or more








specifically disordered eating attitudes and behaviors. The independent variables were

operationalized by the three scales (peer, mentor, and community) of the Relational

Health Indices (RHI) in addition to the variable of ethnicity. The Eating Attitudes-26

(EAT-26) operationalized the dependent variable.


Instrumentation

Relational Health Indices

The Relational Health Indices (RHI; Liang et al., 2000) is a new instrument that

measures women's relationships. It is comprised of three scales that assess growth-

fostering connections with peers, mentors, and communities. The RHI was developed

using the Stone Center's Relational Model, a theory of girls' and women's psychological

development (Jordan et al., 1991; Miller & Stiver, 1997). Within each of the three

relationship domain scales, there are three subscales that have been identified as key

aspects of growth-fostering relationships. These subscales are empathy/engagement,

authenticity, and empowerment/zest.

The Peer Relationship scale (RHI-P) contains 12 items, the Mentor Relationship

scale (RHI-M) contains 11 items, and the Community Relationship scale (RHI-C)

contains 14 items. Subjects rated these relationship domains according to a five-point

Likert-type scale with the responses "never, seldom, sometimes, often, and always" with

corresponding values from one to five. A high mean composite score on each of these

scales corresponds to a high degree of relational health in the context of peer, mentor, and

community relationships. The subscale composite Cronbach's alpha coefficients for

internal consistency are peer = .85, mentor = .86, and community = .90 (Liang et al.,

2000). The RHI was developed for a female college student population.








Convergent validity (r = .69) has been assessed by the correlation of the RHI-P

scale and the Mutual Psychological Development Questionnaire (MPDQ; Genero et al.,

1992). The MPDQ is a 22-item instrument that measures Relational Model concepts in

close dyadic relationships. The RHI-P was also very similar to two of the scales of the

Quality of Relationship Questionnaire (QRI; Pierce, Sarason, Sarason, & Solky-Butzel,

1997), the Support scale (r = .61) and the Depth of Relationship scale (r = .64). The QRI

also assesses aspects of a dyadic relationship. Moderately high positive correlations (r =

.50) were also found between the RHI-P and the Friend Support subscale of the

Multidimensional Scale of Perceived Social Support (MSPSS; Zimet, Dahlem, Zimet, &

Farley, 1988). This 4-item scale measures perceived social support from friends (Liang

et al., 2000).

Convergent validity for the RHI-M is similar to that of the RHI-P. When

correlated to the MPDQ, the relationship is r = .68. The correlations with the QRI

Support scale (r = .58) and Depth of Relationship scale (r = .51) are both moderately

strong (Liang et al., 2000). Both the MPDQ and the QRI were designed for assessing

dyadic relationships and there is no equivalent measure for assessing the convergent

validity of the community relationship scale (Liang et al., 2000).

Concurrent validity was measured using several psychological outcome scales.

The RHI-P, RHI-M, and the RHI-C relationship scales were all negatively correlated (r =

-.35, -.14, -.47) with the UCLA Loneliness Scale (RULS; Russell, Peplau, & Cutrona,

(1980). The RHI-C scale was also negatively correlated (r = -.39) with the Center for

Epidemiological Studies Depression Scale (CES-D; Radloff, 1977) and negatively

correlated (r = -.32) with the Perceived Stress Scale (PS; Cohen, Kamarck, &








Mermelstein, 1983). The Relational Health Indices (RHI) has good reliability and

validity and has utility as a new quantitative instrument that can be used for theory

development of the Stone Center's Relational Model of female psychological

development.


Eating Attitudes Test-26

The Eating Attitudes Test-26 (EAT-26; Garner et al., 1982) is a 26-item self-

report questionnaire designed to measure the degree to which respondents possess a

variety of behaviors and attitudes associated with disordered eating (Heesacker &

Neimeyer, 1990). The original version of the Eating Attitudes Test (EAT; Garner &

Garfinkel, 1979) consisted of 40 forced-choice items rated on a six-point Likert-type

scale. Respondents rated whether each item applies "always, very often, often,

sometimes, rarely, or never." The higher the score, the more symptomatic the

respondent. The 26-item version of the EAT was subsequently developed from factor

analyses when items not loading on any of the three factors (dieting, bulimia and food

preoccupation, and oral control) were eliminated (Garner et al., 1982). The total score of

the EAT-26 is highly correlated (r = .97 for female college students) with the 40-item

EAT total score (Hersen & Bellack, 1988).

The EAT-26 consists of three factors. Factor I, dieting, reflects a pathological

avoidance of fattening foods and shape preoccupations. Respondents who score highly

on Factor I are overestimators of their body size and may be dissatisfied by their body

size and shape but are not bulimic. Factor II, bulimia and food preoccupation, is

positively related to bulimia and a heavier body weight. Factor III, oral control, reflects

self-control about food as well as an acknowledgement of social pressure to gain weight








(Garner et al., 1982). All three factors are combined in an overall total score that reflects

a variety of behaviors and attitudes associated with disordered eating.

When scoring the EAT-26, Garner et al. (1982) recommend that the responses

"never, rarely, and sometimes" receive a score of 0, while the responses "often, very

often, and always" receive scores of 1, 2, and 3, respectively. Total scores may range

from 0-78. Higher scores signify more extreme pathology on each of the three factors.

Higher total scores indicate more pathological eating symptomatology overall. A score

equal to or greater than 20 reflects eating patterns, dieting, and preoccupations with body

weight that interfere with normal psychosocial functioning (Garfinkel & Garner, 1982).

The EAT-26 can also be used as a continuous measure of eating disturbances in a

nonclinical population (Koslowsky, Scheinberg, Bleich, Mark, Apter, Danon, &

Solomon, 1992). In the statistical analyses in this study, EAT-26 scores were treated as a

continuous variable and the participants' total score was equal to the sum of all of the

coded responses.

The EAT-26 has been used to assess the presence of eating pathology in both

clinical and non-clinical settings (Hersen & Bellack, 1988). While this instrument may

indicate the presence of disordered eating attitudes and behaviors, it does not reveal the

motivation or possible psychopathology underlying the disturbed eating patterns (Garner

et al., 1982). Therefore, the EAT-26 may be most suitable as either a treatment outcome

measure in clinical groups or as a screening instrument in non-clinical settings to identify

individuals who have disturbed eating patterns (Hersen & Bellack, 1988).

Internal consistency (alpha = .94) and test-retest reliability (r = .84) of the EAT-

26 are excellent (Garfinkel & Garner, 1988). Gross et al. (1986) found total scores of the








EAT to be moderately correlated with the three eating disorder symptom scales (Drive

for Thinness r = .81; Body Dissatisfaction r = .50; Bulimia r = .42) of the Eating Disorder

Inventory (EDI; Garner, Olmsted, & Polivy, 1983). Williamson (1990) reported

unpublished data from a sample of both clinical and nonclinical subjects and found the

EAT to be moderately correlated (r = .67) with the Bulimia Test (Smith & Thelan, 1984)

and the Eating Questionnaire-Revised (r = .59; EQ-R; Williamson, Davis, Goreczny,

Bennett, & Watkins, 1989).

Button and Whitehouse (1981) administered the EAT to a large sample of

students at a technical college and reported that this assessment instrument was useful in

identifying study participants with abnormal concerns regarding eating and weight.

Thompson and Schwartz (1982) used the EAT to identify college women with abnormal

eating concerns who were later distinguished from anorexic patients with regard to social

adjustment. Gross et al. (1986) demonstrated criterion validity for the EAT by

discriminating bulimia nervosa subjects from normal subjects. The EAT-26 was used to

compare eating disturbances between Japanese and American college women (Mukai,

Kambara, & Sasaki, 1998) and in another study, EAT scores in both Asian and Caucasian

college women were significantly positively correlated with fear of fat (Sanders & Heiss,

1998). Jane, Hunter, and Lozzi (1999) administered the EAT-26 to Cuban American

women in order to discern whether continuing identification with and participation in

aspects of Cuban Hispanic culture may serve as a mitigating factor in the predisposition

and development of eating disorders. Due to wide and varied use over time, the EAT and

the EAT-26 have been found to have very good concurrent, predictive, and discriminant








validity (Williamson, Anderson, & Gleaves, 1996). This measure of eating attitudes and

behaviors was chosen for its good psychometric properties.


Data Analysis

One regression equation was used where the EAT-26 is the dependent measure

with the three scales of the RHI (RHI-P, RHI-M, RHI-C) and ethnicity, and the

interactions of ethnicity with the RHI-P, RHI-M, and RHI-C are the independent

variables. In addition, three ANOVA's were used with the RHI-P, RHI-M, and RHI-C as

dependent variables and ethnicity as the independent variable.


Hypothesis and Research Questions

Hypothesis

There is a negative relationship between the relational health of White, Black, and

Latina college women's peer, mentor, and community relationships and the attitudes and

behaviors related to disordered eating. One regression equation was used where the

EAT-26 is the dependent measure with the three scales of the RHI (RHI-P, RHI-M, RHI-

C) and ethnicity, and the interactions of ethnicity with the RHI-P, RHI-M, and RHI-C are

the independent variables.


Research Questions

1. What is the relationship between peer relationships and disordered eating? A

regression equation was used where the EAT-26 is the dependent measure and the RHI-P

is the independent variable.

2. Does the relationship between peer relationships and disordered eating differ

for White, Black, and Latina female college students? A regression equation was used








where the EAT-26 is the dependent measure and the interaction of ethnicity with the

RHI-P is the independent variable.

3. What is the relationship between mentor relationships and disordered eating?

A regression equation was used where the EAT-26 is the dependent measure and the

RHI-M is the independent variable.

4. Does the relationship between mentor relationships and disordered eating

differ for White, Black, and Latina female college students? A regression equation was

used where the EAT-26 is the dependent measure and the interaction of ethnicity with the

RHI-M is the independent variable.

5. What is the relationship between community relationships and disordered

eating? A regression equation was used where the EAT-26 is the dependent measure and

the RHI-C is the independent variable.

6. Does the relationship between community relationships and disordered eating

differ for White, Black, and Latina female college students? A regression equation was

used where the EAT-26 is the dependent measure and the interaction of ethnicity with the

RHI-C is the independent variable.

7. What is the relationship among White, Black, and Latina college women's

disordered eating? A regression equation was used where the EAT-26 is the dependent

measure and ethnicity is the independent variable.

8. What is the relationship among White, Black, and Latina college women's

peer relationships? An ANOVA was used with the RHI-P as the dependent variable and

ethnicity as the independent variable.





69


9. What is the relationship among White, Black, and Latina college women's

mentor relationships? An ANOVA was used with the RHI-M as the dependent variable

and ethnicity as the independent variable.

10. What is the relationship among White, Black, and Latina college women's

community relationships? An ANOVA was used with the RHI-C as the dependent

variable and ethnicity as the independent variable.














CHAPTER 4
RESULTS

The purpose of this study was to examine the relationship between Black, Latina,

and White college females' relational health, specifically peer, mentor, and community

relationships, and disordered eating attitudes and behaviors. The data collection

procedures, data analyses, and results of this study are presented in this chapter.


Data Collection

One hundred-sixty research packets were sent to a stratified randomized sample

of Black, Latina, and White sophomore, junior, and senior undergraduate college females

for a total sample of 480 students. Each research packet contained a cover-request to

participate letter, the Relational Health Indices (Peer, Mentor, and Community Scales)

and the Eating Attitudes Test-26, and a self-addressed stamped envelope. Approximately

one week after the packets were mailed, a follow-up postcard was sent to each student in

the study. Subsequent follow-up emails were also sent.

A total of 237 (49.3%) packets were returned which included 68 (42.5%)

undergraduate female Black respondents, 77 (48.1%) undergraduate female Latina

respondents, and 92 (58.1%) undergraduate female White respondents. All 68 Black

students completed the Relational Health Indices Peer Scale (RHI-P) and the Eating

Attitudes Test-26 (EAT-26), 59 Black students completed the Relational Health Indices

Mentor Scale (RHI-M), and 66 Black students completed the Relational Health Indices

Community Scale (RHI-C). All 77 Latina students completed the RHI-P and the EAT-









26, 72 Latina students completed the RHI-M, and 74 Latina students completed the RHI-

C. All 92 White students completed the RHI-P and the EAT-26 and 90 White students

completed the RHI-M and RHI-C. Summary statistics are presented in Table 1.

Table 1

Summary Statistics

Variable N Mean Standard Deviation


RHI-P
RHI-M
RHI-C
EAT-26



RHI-P
RHI-M
RHI-C
EAT-26



RHI-P
RHI-M
RHI-C
EAT-26


92
90
90
92


Black

46.12
44.24
44.56
55.24

Latina

49.05
43.75
46.27
63.09

White

49.79
44.88
49.11
63.40


6.95
7.55
10.93
16.14


5.66
7.74
10.15
20.09


5.59
7.07
10.04
19.15


Data Analyses


Hypothesis

There is a negative relationship between the relational health of White, Black, and

Latina college women's peer, mentor, and community relationships and the attitudes and

behaviors related to disordered eating. One regression equation was used where the

EAT-26 is the dependent measure with the three scales of the RHI (RHI-P, RHI-M,








RHI-C) and ethnicity, and the interactions of ethnicity with the RHI-P, RHI-M, and RHI-

C are the independent variables.

The regression equation included 215 observations and resulted in an R-square of

0.1433 and an adjusted R-square of 0.1143. Therefore, ethnicity, RHI-P, RHI-M, RHI-C,

and the interactions of RHI-P, RHI-M, and RHI-C with ethnicity jointly accounted for

11.43% of the variation in the EAT-26 scores. Only RHI-P and RHI-C were significant

predictors, at nominal alpha 0.05, of EAT-26 with p-values of 0.0103 and 0.0308,

respectively. Results are presented in Table 2.


Table 2

Regression Equation: Model 1

Source DF F p-value

Ethnicity 2 1.15 0.3177
RHI-P 1 6.71 0.0103*
RHI-P*Ethnicity 2 0.39 0.6790
RHI-M 1 0.00 0.9545
RHI-M*Ethnicity 2 2.46 0.0880
RHI-C 1 4.73 0.0308*
RHI-C*Ethnicity 2 2.12 0.1232

Note. "*" denotes significance at a=0.05.


Since all 2-way interactions were non-significant, a reduced regression model was

run with the main effects (Ethnicity, RHI-P, RHI-M, RHI-C) separated out. The

regression equation also included 215 observations and resulted in an R-square of 0.1139

and an adjusted R-square of 0.0970. This second regression model accounted for 9.7% of

the variation in the EAT-26 scores. Results are presented in Tables 3.








Table 3

Regression Equation with Main Effects: Model 2

Source B SE B Beta
Ethnicity
RHP -0.5296 0.2256 -3.2855
RHM -0.0684 0.1790 -0.5067
RHC -0.2755 0.1340 -2.8838

Intercept
White 105.76
Black -11.51
Latino -0.56
Note. "*" denotes significance at a=0.05.


F
7.60
5.51
0.15
4.23



13.10
0.04


p-value
0.0007*
0.0198*
0.7027
0.0410*



0.0004*
0.8498


Multicollinearity was tested with these regression equations: RHI-P = RHI-M +

RHI-C; RHI-M = RHI-C + RHI-P; and RHI-C = RHI-P + RHI-M. The R-squares for

these equations are 0.148, 0.125, and 0.187 respectively. Therefore, since each result is

not greater than .9, the regression equation has no serious multicollinearity problem.


Research Questions

1. What is the relationship between peer relationships and disordered eating? A

regression equation was used where the EAT-26 is the dependent measure and the RHI-P

is the independent variable.

The equation included 237 observations and resulted in a squared correlation of

0.0258. The results indicated RHI-P is a significant predictor of EAT-26, F = 6.23 and p-

value = 0.0132. The regression equation is EAT-26 = 92.562 0.491(RHI-P) + error.

The results indicated that an increase of one point on the RHI-P scale predicted a

decrease of 0.5 in the EAT-26 score.








2. Does the relationship between peer relationships and disordered eating differ

for White, Black, and Latina female college students? A regression equation was used

where the EAT-26 is the dependent measure and the interaction of ethnicity with the

RHI-P is the independent variable.

The equation included 237 observations. The interaction of ethnicity and RHI-P

is not significant, F = 0.66 and p-value = 0.5202. Interaction results including the main

effects of ethnicity and the RHI-P are presented in Table 4.


Table 4

Interaction of RHI-P and Ethnicity

Source Degrees of Freedom F p-value

Ethnicity 2 0.512 0.5962
RHI-P 1 10.71 0.0012*
RHI-P*Ethnicity 2 0.66 0.5202

Note. "*" denotes significance at a=0.05.


3. What is the relationship between mentor relationships and disordered eating?

A regression equation was used where the EAT-26 is the dependent measure and the

RHI-M is the independent variable.

RHI-M by itself is not a significant predictor of EAT-26. There were 221

observations included in the equation, F = 2.93 and p-value = 0.0885.

4. Does the relationship between mentor relationships and disordered eating

differ for White, Black, and Latina female college students? A regression equation was








used where the EAT-26 is the dependent measure and the interaction of ethnicity with the

RHI-M is the independent variable.

The equation included 221 observations. The interaction of ethnicity and RHI-M

is not significant, F = 1.23 and p-value = 0.2937. Interaction results including the main

effects of ethnicity and the RHI-M are presented in Table 5.


Table 5

Interaction of RHI-M and Ethnicity

Source Degrees of Freedom F p-value

Ethnicity 2 1.28 0.02798
RHI-M 1 3.06 0.0818
RHI-M*Ethnicity 2 1.23 0.2937


5. What is the relationship between community relationships and disordered

eating? A regression equation was used where the EAT-26 is the dependent measure and

the RHI-C is the independent variable.

The equation included 230 observations and resulted in a squared correlation of

0.0283. The results indicated RHI-C is a significant predictor of EAT-26, F = 6.64 and

p-value = 0.0106. The regression equation is EAT-26 = 74.933 0.3042(RHI-C) + error.

The results indicated that an increase of one point on the RHI-C scale predicted a

decrease of 0.3 in the EAT-26 score.

6. Does the relationship between community relationships and disordered eating

differ for White, Black, and Latina female college students? A regression equation was








used where the EAT-26 is the dependent measure and the interaction of ethnicity with the

RHI-C is the independent variable

The equation included 230 observations. The interaction of ethnicity and RHI-C

is not significant, F = 0.38 and p-value = 0.6870. Interaction results including the main

effects of ethnicity and the RHI-C are presented in Table 6.


Table 6

Interaction of RHI-C and Ethnicity

Source Degrees of Freedom F p-value

Ethnicity 2 1.12 0.3293

RHI-C 1 9.46 0.0024*

RHI-C*Ethnicity 2 0.38 0.6870

Note. "*" denotes significance at ax=0.05.



7. What is the relationship among White, Black, and Latina college women's

disordered eating? A regression equation was used where the EAT-26 is the dependent

measure and ethnicity is the independent variable.

This equation included 237 observations and indicated that there are ethnic group

differences in the mean scores of the EAT-26 scale, F = 4.47 and p-value = 0.0124.

Multiple comparison procedures indicated Black scores differed from both White and

Latina scores, but that White and Latina scores did not significantly differ from each

other. Contrasts of the EAT-26 mean scores are presented in Table 7.








Table 7

Contrasts of the EAT-26 Mean Scores


Contrast F p-value

Black vs. Latina 6.40 0.0121*
Black vs. White 7.43 0.0069*
Latina vs. White 0.01 0.9231

Note. "*" denotes significance at a=0.05.


8. What is the relationship among White, Black, and Latina college women's

peer relationships? An ANOVA was used with the RHI-P as the dependent variable and

ethnicity as the independent variable.

This ANOVA included 237 observations. There were ethnic group differences in

the mean scores of the RHI-P variable, F=7.74, p-value=0.0006. Multiple comparison

procedures indicated Black scores differed from both Latina and White scores, but White

and Latina scores did not significantly differ. Contrasts of RHI-P mean scores are

presented in Table 8.


Table 8

Contrasts of RHI-P Mean Scores

Contrast F p-value

Black vs. Latina 8.54 0.0038*
Black vs. White 14.51 0.0002*
Latina vs. White 0.63 0.4270

Note. "*" denotes significance at a=0.05.








9. What is the relationship among White, Black, and Latina college women's

mentor relationships? An ANOVA was used with the RHI-M as the dependent variable

and ethnicity as the independent variable.

This ANOVA included 221 observations. There were no ethnic group differences

in the mean scores of the RHI-M variable, F=0.47, p-value=0.6260.

10. What is the relationship among White, Black, and Latina college women's

community relationships? An ANOVA was used with the RHI-C as the dependent

variable and ethnicity as the independent variable.

This ANOVA included 230 observations. There were ethnic group differences in

the mean scores of the RHI-P variable, F=3.89, p-value=0.0219. Multiple comparison

procedures indicated Black scores differed from White scores, but Latina scores did not

significantly differ from either Black or White scores. Contrasts of RHI-C mean scores

are presented in Table 9.


Table 9

Contrasts of RHI-C Mean Scores

Contrast F p-value

Black vs. Latina 0.95 0.3297
Black vs. White 7.38 0.0071*
Latina vs. White 3.07 0.0812

Note. "*" denotes significance at cr=0.05.


In summary, the overall hypothesis of this research study was partially

substantiated by results indicating that participants who reported higher levels of

relational health, specifically peer and community relational health, also reported lower








levels of disordered eating. The first research question indicated that participants in this

study who reported higher levels of peer relational health also reported lower levels of

disordered eating. The second research question was not substantiated, as participants in

the study did not show an ethnic difference in reported peer relational health and

disordered eating. The third research question was not substantiated, as participants in

the study did not report a relationship between mentor relational health and disordered

eating. The fourth research question was not substantiated, as participants in this study

did not show an ethnic difference in reported mentor relational health and disordered

eating. The fifth research question indicated that participants in this study who reported

higher levels of community relational health also reported lower levels of disordered

eating. The sixth research question was not substantiated, as participants in this study did

not show an ethnic difference in reported community relational health and disordered

eating. The seventh research question indicated that Black participants in this study

differed from both Latina and White participants in levels of disordered eating but Latina

and White participants in this study did not differ significantly in levels of disordered

eating. The eighth research question indicated that Black participants in this study

differed from both Latina and White participants in levels of peer relational health but

Latina and White participants in this study did not differ significantly in levels of peer

relational health. The ninth research question was not substantiated, as participants in

this study did not show an ethnic difference in reported mentor relational health. The

tenth research question indicated that Black participants in this study differed from White

participants in levels of community relational health but Latina participants in this study






80

did not differ significantly from either Black or White participants in levels of

community relational health.














CHAPTER 5
DISCUSSION

The purpose of this study was to examine the relationship between Black, Latina,

and White college females' relational health, specifically peer, mentor, and community

relationships, and disordered eating attitudes and behaviors. The research findings

related to the hypothesis and individual research questions of this study will be presented

in this chapter followed by a discussion of the limitations of the study and implications of

the findings and recommendations for the future.


Hypothesis Summary and Explanation of Finding

The overall hypothesis of this study was that there is a negative relationship

between the relational health of Black, Latina, and White college females' peer, mentor,

and community relationships and the attitudes and behaviors related to disordered eating.

This hypothesis was partially substantiated by results indicating that participants

in this study who reported higher levels of both peer and community relational health also

reported lower levels of disordered eating. Mentor relationships however did not

significantly relate to the measured attitudes and behaviors of disordered eating.

Relational Model theorists conceptualize on-going, growth-fostering

interpersonal connections as critical to women's healthy psychological development

(Jordan, 1997; Surrey, 1985). Growth-fostering peer and community relationships, which

reflect the relational components of mutuality, authenticity, and empowerment or zest,

may act as potential protective factors in Black, Latina, and White college females'








struggles with disordered eating. This finding is consistent with many other research

studies that have found that close, intact relationships predict good health in a wide

variety of population samples.

Mentoring relationships, on the other hand, may not increase resilience to

disordered eating due to the fact that this type of relationship, unlike peer and community

relationships, is inherently defined by a power differential and is usually predicated on

foundational differences in areas such as skill level, age, expertise, and education (Liang

et al., 2000). One of the main theoretical tenets of the Relational Model is that growth-

fostering relationships that are intimate and mutual can facilitate self-disclosure,

emotional resiliency, and coping strategies (Miller & Stiver, 1997). The previously

mentioned foundational differences in mentoring relationships may hinder the

development of intimacy and mutuality and therefore this type of relationship may be

ineffective as a protective factor against disordered eating.

Another potential factor in the lack of correlational relationship between

mentoring relationships and disordered eating may be the specific characteristics of the

mentors described by each participant in the study. Whereas the traditional male-to-male

mentoring models within the business world rely on an acceptance of hierarchy and focus

on task activities, females appear to desire more psychosocial and emotional support in

their mentoring relationships (Kalbfleisch & Keyton, 1995). Worell and Remer (1992)

underscored the therapeutic nature of female-to-female mentoring relationships and

stated that women may receive ancillary benefits beyond those normally accrued through

the mentoring process. Gloria et al. (1999) emphasized the importance of having

ethnically similar mentors especially for minority students on predominantly White








college campuses. This research study did not specifically ask gender or ethnicity of

each participant's mentor. This unknown variable could have had an effect on the

significance of the relationship between mentoring relationships and disordered eating in

this study.

Supportive mentoring relationships can be an important part of a student's

adjustment to college by assisting with academic achievement, confidence, leadership

skills, and/or career direction but may not necessarily embody the growth-fostering

relational components of mutuality, authenticity, and empowerment or zest. It is

therefore possible that mentor relationships, unlike peer and community relationships,

might not significantly relate to the attitudes and behaviors of disordered eating in college

females in this study.


Research Questions and Explanations of Findings

Research Question 1 examined the relationship between peer relationships and

disordered eating. Results from this research indicate that peer relationships were a

significant predictor of disordered eating attitudes and behaviors. As the peer relational

health of college females increased, disordered eating decreased proportionately. An

increase of 1.0 point on the RHI-P scale predicted a decrease of 0.5 on the EAT-26 score.

This finding is consistent with previous research on adolescent peer relationships that has

shown that the positive experiences of friendship contribute significantly to cognitive,

social, and moral development as well as psychological adjustment and socioemotional

health (Brown et al., 1999).

Research Question 2 examined the differences in the relationship between peer

relationships and disordered eating among Black, Latina, and White female college








students. Results from this research did not indicate any significant differences among

ethnic groups in their relationship between peer relationships and disordered eating. This

research finding is consistent with previous empirical research that has shown the value

of mutually intimate and nurturing peer support in the physical and psychological health

of young women of varied ethnic backgrounds. Peer relational health may act as a

protective factor for disordered eating for Black, Latina, and White female college

students.

Research Question 3 examined the relationship between mentor relationships and

disordered eating. Results from this research did not indicate any significant relationship

between these two variables. Relational competence as defined by the Relational Model

is the ability to attend to the affect and experience of another individual and then respond

in an appropriate manner that compromises neither that individual self nor another

(Nelson, 1996). Relational competence leads to mutual empowerment, a state in which

each person can receive and then respond to the feelings and thoughts of the other, each

is able to enlarge both her own and another person's feelings and thoughts, and

simultaneously each person enlarges the relationship (Miller & Stiver, 1997). This

principle of mutuality within Relational Theory may not be valid for mentoring

relationships and therefore may contribute to the lack of significance in the relationship

between mentoring and disordered eating in this study.

As discussed previously, the unique interpersonal dynamics of mentoring

relationships can be very helpful to college students' success but may not directly affect

the psychosocial well-being of college females and more specifically their struggles with

disordered eating. The gender and ethnicity of the participants' selected mentors may








have also affected the potential significance of this relationship with disordered eating.

Many eating disordered young women have learned how to successfully

compartmentalize their academic pursuits and in the 'short run' may do well in their

classes and/or academic activities even though their attitudes and behaviors are congruent

with the disordered eating continuum.

Research Question 4 examined whether the relationship between mentor

relationships and disordered eating differed for Black, Latina, or White female college

students. Results from this research did not indicate any significant differences among

ethnic groups and their relationship between mentoring relationships and disordered

eating.

Research Question 5 examined the relationship between community relationships

and disordered eating. Results from this research indicate that community relationships

were a significant predictor of disordered eating attitudes and behaviors. As the

community relational health of college females increased, disordered eating decreased

proportionately. An increase in 1.0 point on the RHI-C scale predicted a decrease of 0.3

in the EAT-26 score. This research finding is consistent with previous research that

shows that resilient adolescents and young adults are sociable and able to seek and garner

social support from a variety of resources (Johnson et al., 1999). This interpersonal

strength is in contrast to females who struggle with disordered eating and also report a

sense of isolation, social anxiety, impoverished relationships, public self-consciousness,

and a failure to seek social support (Szmukler et al., 1995). Community relational health

may act as a protective factor for disordered eating for Black, Latina, and White female

college students.








Research Question 6 examined the differences in the relationship between

community relationships and disordered eating among Black, Latina, and White female

college students. Results from this research did not indicate any significant differences

among ethnic groups in their relationship between community relationships and

disordered eating. In other words, female college students who had higher community

relational health also had lower disordered eating regardless of their specific ethnic

background.

Researchers who have examined sense of belonging, a primary construct of

community relational health, for Black and Latina females state that the cultural construct

of interdependence which starts during childhood and involves a familial type of

attachment to community serves as a personal strength which can help to mitigate

feelings of isolation and detachment (Harris, 1992; Hurtado & Carter, 1997).

Community relationships that promote a sense of belonging and are defined by the

growth-producing qualities of the Relational Model have had a beneficial influence on

the physical and psychological health of women from various ethnic backgrounds in

numerous research studies. This benefit may also extend to helping in the prevention of

disordered eating in Black, Latina, and White female college students.

Research Question 7 examined the relationship among Black, Latina, and White

female college students' disordered eating. Results from this research indicated that there

were ethnic group differences. Black female mean scores on the EAT-26 were lower

than both Latina and White female mean scores, which did not significantly differ from

each other. Lower scores on the EAT-26 indicate lower levels of disordered eating

attitudes and behaviors.








Root (1990) stated that racial/cultural context may afford protection from

disordered eating to an ethnic group, but it does not necessarily protect individuals who

are subject to the standards of the dominant Western culture. This study did not examine

issues related to acculturation and assimilation that may have affected the survey results

of the different ethnic groups in this research question. It also did not include the

variable of actual body weight, which may also have affected the results of this question.

Research has shown that there is a positive association between extent of body

dissatisfaction and actual body weight in samples of Black female college students but

White female college students were likely to adopt disordered eating attitudes and

behaviors regardless of actual weight (Abrams et al., 1993). Similar studies that examine

the dynamics of actual body weight and disordered eating have not been done with Latina

college students. Other research studies have generally shown that Black women

compared to White and Latina women tend to be more satisfied with their bodies, are

more accepting of being overweight, and are less driven to achieve thinness (Abrams et

al., 1993; Fitzgibbon et al., 1998). These aforementioned factors may have played a part

in the lower levels of disordered eating in Black female participants in this study.

Research Question 8 examined the relationship among Black, Latina, and White

female college students' peer relationships. Results from this research indicated that

there were ethnic group differences. Black female mean scores on the RHI-P were lower

than both Latina and White female mean scores, which did not significantly differ from

each other. Lower scores on the RHI-P indicate lower levels of peer relational health in

accordance with the Stone Center's Relational Model.








Sedlacek's (1999) meta-analysis of twenty years of research on Black students on

predominantly White college campuses found that racism remains a major problem for

Black students across the country. This research study was done at a large co-ed

university in the southeastern United States where Black students comprise only 6.5%

(2,900) of the student population and Black female undergraduate students comprise less

than half of that total (University of Florida, 2000). The difficulty of 'fitting in' and

finding compatible peers may be difficult for many young Black women in this study.

Various research studies document the difficulty that Black students face in the arena of

social adjustment on predominantly White college campuses. Though Latina students are

also a minority on this research study campus (9.4%; 4,200; University of Florida, 2000),

and probably deal with racism on an individual and institutional level, it is possible that

they may assimilate more easily into the White college campus environment and are

more able to find mutually intimate and nurturing peers. The results of this research

question may also have been affected by the lower response rate of the Black female

participants.

Research Question 9 examined the relationship among Black, Latina, and White

college females' mentoring relationships. Results indicated that there were no significant

ethnic group differences in the mean scores of the RHI-M. This research result may be

attributed to the fact that the instructions for this survey scale were flexible and allowed

the participants to select any "adult who is often older than you, and is willing to listen,

share her or his experiences, and guide you through some part or area of your life."

Several participants indicated that they chose a parent or close relative as their mentor.

For minority students, this choice flexibility may have contributed to the lack of








significance between group differences by giving Black and Latina women the ability to

identify mentors outside of the university setting. This study result may have been

different if mentors had been required to be affiliated with the university research study

site. In addition, the university in this study has an active on-going mentoring program

for minority students, which may have decreased ethnic group differences. It is assumed

that White female students on their own could more easily find a same gender/ethnicity

mentor if that was a priority. Another factor that should be considered is the difference in

response numbers (Black, 59; Latina, 72; White, 90) for the RHI-M, which may have

affected the results of this research question. This was the smallest response sample of

the three RHI scales and this research question result may have been statistically

significant with a larger sample.

Research Question 10 examined the relationship among Black, Latina, and White

college females' community relationships. Research results indicated ethnic group

differences in the mean scores of the RHI-C. Black mean scores were lower than White

mean scores, but Latina scores were not significantly different from Black or White mean

scores. Lower mean scores on the RHI-C indicate lower levels of community relational

health. As previously discussed, lower levels of community relational health for Black

college females may be attributed to their struggles with racism and assimilation within a

predominantly White college campus. Latina mean scores, though not statistically

significantly different, were between Black and White mean scores and may also indicate

a difficulty with cultural assimilation but not to the same degree as for Black female

students.








In conclusion, the findings of the research questions in this study indicated that

both peer and community relationships significantly predicted disordered eating attitudes

and behaviors and may be protective factors in disordered eating for Black, Latina, and

White female college students. Mentor relationships were not significantly related to

disordered eating in college females as a whole or for any of the specific ethnic groups.

Black female participants had lower levels of disordered eating when compared with

Latina and White female students. Black female participants also had lower levels of

peer and community relational health. Ethnicity and the interactions of ethnicity with

relational health in this study did not prove to be significant variables in the relationship

between disordered eating and relational health.


Limitations of the Study

This study used a stratified randomized sample of sophomore, junior, and senior

undergraduate Black, Latina, and White college females from a large co-ed university in

the southeastern United States. Therefore, generalizability of the findings of the current

study would be limited primarily to undergraduate Black, Latina, and White female

college students. It is also important to recognize that the composition of Black and

Latina ethnic groups vary within the United States and therefore the generalizability of

this study may be limited to the southeastern region of the United States.

This study was limited by the exclusive reliance on self-report measures to assess

peer, mentor, and community relational health and disordered eating. Self-report

instruments are susceptible to respondent bias though they are also considered to be an

efficient method for obtaining data in a relatively nonobtrusive and confidential manner.








Respondent bias may have been a problem in this study due to the personal nature of the

questions related to interpersonal relationships and eating attitudes and behaviors.

Another limitation to the study is the self-selection of the participants receiving

the mail surveys. Overall the response rate was 49.3%, but among ethnic groups there

was a variable response rate (Black, 42.5%; Latina, 48.1%; White, 58.1%). Even though

ethnicity did not prove to be a significant variable, the unique ethnic composition of the

entire participant group could potentially be a limitation to generalizability. In addition,

questions arise about the unknown characteristics of the non-participating group and the

unknown reasons for response variability among ethnic groups.

A critical limitation inherent to the correlational nature of this research design

involves the inability to ascertain causal connections. The findings show a correlation

between relational health and disordered eating, but there is no way of knowing

whether poor relational health brings about disordered eating or vice versa, or even if

there were other unidentified factors at work. This study was only a first step in

understanding the role of relationships and disordered eating attitudes and behaviors in

college women.


Implications of the Findings and Recommendations

Findings from this research study have added to an increased understanding of the

relationship between peer, mentor, and community relationships and disordered eating in

Black, Latina, and White female college students. Implications for theory, practice, and

research are presented in the following chapter section.








Implications for Theory

The theoretical intent of this research study was to test a model of female

psychological development with disordered eating attitudes and behaviors within a

multicultural female college population. The results of the study offered mixed support

for the elements of the proposed model. Research findings showed that relational health,

specifically peer and community but not mentor relationships, were related to disordered

eating. Ethnicity and the interaction of ethnicity with relational health did not play a

significant role in this relationship.

According to the principles of the Stone Center's Relational Theory, the goal of

healthy psychological development for girls and women is attained via the increasing

ability to build and enlarge mutually enhancing relationships in which each individual

can feel an increased sense of well-being through being in touch with others and finding

ways to act on individual thoughts and feelings (Surrey, 1985). The results of this study

reinforced the tenets of this theory by showing that there is a negative relationship

between peer and community relational health and disordered eating. Disordered eating

being recognized as symptomatic of a decreased sense of well-being. The finding that

ethnicity was not a significant factor in the research model can be interpreted as a

strength of Relational Theory. According to the findings of this study, Black, Latina, and

White female college student participants all showed a negative relationship between

peer and community relational health and disordered eating and therefore, in this study,

the Relational Model is relevant for a multicultural population sample.

The vast majority of empirical research on Relational Theory has been conducted

primarily through qualitative studies. The results of this research study add to the body








of quantitative research literature that proposes that the formation and maintenance of

relationships and connections to others is critical to healthy female psychological

development. The results of this study are also consistent with numerous other studies on

the benefits of social support as being health promoting by providing a sense of

belonging and relational intimacy.


Implications for Practice

The immediate and practical implications of this research will be of interest to

university administrators, health educators, and campus mental health providers. It is

important to understand the struggles of college women with disordered eating and the

potential protective factors of peer and community relationships.

Most eating disorder inpatient programs provide a predominantly female

community for patients, with an emphasis on group therapy. These programs emphasize

the importance of the building of intimate connections as a treatment goal in the healing

of women struggling with disordered eating. Colleges can learn from this type of

intervention and make group experiences available for all women on campus as part of

their education, prevention, and eating disorder intervention strategies.

On-going campus groups that focus specifically on issues related to body image,

eating problems, healthy coping strategies, and personal growth should be easily and

routinely available for the female college population. These groups should be facilitated

by someone who is familiar with the Relational Model's principles of mutuality,

authenticity, and empowerment or zest and thereby create a beneficial group experience

within the framework of the community relational model. It would be important to target

in-coming freshman and transfer students who are dealing with the uncertainties of life








transition and college adjustment especially at large university campuses that may seem

unfriendly and alien to the new student. It would be imperative to also insure that all

group environments were culturally sensitive so that a diverse group of college women

would benefit.

College women who struggle with disordered eating would also benefit from

individual therapy with feminist therapists who believe that it is crucial to build an

egalitarian relationship between the client and the counselor (Worell & Remer, 1992).

This type of collaborative therapy model embraces the relational principles of mutuality,

authenticity, and empowerment or zest and potentially would be beneficial in helping the

female college student more effectively deal with her disordered eating issues along with

the other stressors that are prevalent during this developmental stage of her life.


Implications for Research

There are a number of directions that future research could take based upon the

results of this study. To increase the external validity of current results that correlate

relational health with disordered eating, it will be important to employ a similar research

design with different age groups, ethnic groups, and with girls and women from varied

socioeconomic and educational backgrounds. Most of the previous eating disorder

research has been done with groups of White females. It is important to continue to do

empirical research with multiculturally diverse groups of college females with large

ethnic research groups.

This study could be replicated and potentially strengthened by including measures

of ethnic acculturation and/or assimilation. With these added variables, ethnicity may