GENDER ROLE CONFLICT AND PSYCHOSOCIAL CONCERNS ACROSS RACE AND
SCHOOL TYPE AS INFLUENCES ON ADOLESCENT GIRLS’ SPORT PARTICIPATION
AND WITHDRAWAL
By
Elizabeth A. Wright

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Kinesiology-Doctor of Philosophy
2013

!

ABSTRACT
GENDER ROLE CONFLICT AND PSYCHOSOCIAL CONCERNS ACROSS RACE AND
SCHOOL TYPE AS INFLUENCES ON ADOLESCENT GIRLS’ SPORT PARTICIPATION
AND WITHDRAWAL
By
Elizabeth A. Wright
One of the most popular activities for both boys and girls to participate in are sports, with
over 44 million youth involved in youth sports (National Council of Youth Sports, 2008). While
girls are participating at higher rates than they were before the passage of Title IX in 1972, there
still exists a discrepancy between the number of boys and girls who participate in sports. Further,
there is an even larger discrepancy when looking at the demographics of the adolescent girls who
are participating at the lowest (girls of color and girls from low-income urban areas) and highest
(girls who are white and reside in suburban areas) rates.
One way of studying these discrepancies in sport participation is examining if gender role
conflict or psychosocial concerns have a significant impact on adolescent girls’ sport
participation. With there being a lack of research that studies the sport experiences of girls of
various racial and economic backgrounds, this study filled this void by examining not only girls’
gender role conflict and psychosocial concerns, but also their overall levels of sport participation
and their views on gender appropriateness of sports.
Two hundred and thirty-six girls who attended middle SES suburban (n = 130) and low
SES urban (n = 106) high schools participated in this study. It was found that adolescent girls
reported low amounts of gender role conflict and psychosocial concerns. Related, girls also
reported that gender role conflict and psychosocial concerns played a small role in any decreased
middle school sport participation. Some differences emerged when examining girls’ experiences

!

with these two measures across race, with African American girls having significantly more
concerns with their image and African American girls from middle SES suburban schools
reporting the highest amount of impact of gender role conflict.
There were similar mixed findings when examining current sport participation. Girls who
were current athletes in high school actually had higher gender role conflict scores than those
girls who were non-current sport participants. There were limited differences between current
and non-current sport participants’ psychosocial concerns, with non-current sport participants
reporting significantly higher image concerns. Similar results were found with girls who had
played and quit a feminine sport, as they had higher image concerns than girls who had played a
quit a masculine sport. When looking at middle school and high school sport participation, white
girls from middle SES suburban schools were typically the girls that participated the most in
sport. What played an especially significant role was the type of school a girl attended, with
African American girls particularly negatively impacted with their high school sport
participation if they attended a low SES urban school.
While more research is needed to confirm these patterns, what these findings suggest is
that there is a significant decrease in sport participation between middle school and high school,
but only for girls that attend low SES urban schools. Noting the small amount of impact that
psychosocial concerns and gender role conflict had with girls in this study, it can be tentatively
concluded that there are other reasons that impact adolescent girls’ decreasing or quitting their
sport participation.

!

Copyright by
ELIZABETH A. WRIGHT
2013!

!

ACKNOWLEDGEMENTS

I would like to acknowledge the College of Education for the Dissertation Completion
Fellowship that was provided to fund this research. I would like to thank Dr. Dan Gould, my
advisor and dissertation chair. Coming back to do my PhD at Michigan State was one of the best
decisions I have made and I consider myself a fortunate person to have had you as my advisor
and mentor for this. I thank you for the encouragement and guidance; not only with the
dissertation process but also throughout my graduate academic career. Your positivity and belief
in me kept me going through the tough times of a PhD program. I would also like to thank Dr.
Marty Ewing. I will never forget the Institute for the Study of Youth Sport staff meeting early on
in my first year where you wondered out loud, “where are all the girls and why are they not
participating? Someone needs to look into this!” This question truly informed my research
agenda at Michigan State. Thank you for posing this question and even more so for the support
you provided along the way with my research in this area. Your open door policy and
accessibility in always having time to sit, talk, and provide such excellent guidance and feedback
to me during this process was unbelievably appreciated. Thank you to Dr. Larry Lauer for
providing a supportive voice and encouraging me to do this research. Thank you for your time
with this dissertation and for being a great friend along the way. A big thank you to Dr. Toby
Ten Eyck for a voice who challenged me to look further into the “so what?” of my research. You
have pushed me in my preconceptions and have made me a better researcher. I am grateful of the
feedback you have provided along the way and for the supportive ways in which you do so. A
thank you to Dr. Al Smith for your guidance in my undergraduate career, and for the

!

v

recommendation to continue my studies with Dan; I am honored (and lucky) to have had both of
you as mentors as I made my way on this path.
A thank you to my parents who have provided unending support over the years,
especially to my mom who encouraged me to persevere when the days were long through the
words of wisdom, “well, if it was easy, everyone would do it!” Also, a formal thank you in here
is also deserved for all of the support you both provided me as a youth athlete, as I know how
much my sport background has influenced me in my research. LJ – a lot of this dissertation was
written in coffee shops and at my home, most times with you working nearby. Thank you for
constantly believing in me, being my sounding board, and an unbelievable source of emotional
support through the whole process. Also, thank you for modeling balance in your approach with
life; I continue to learn from you in this regard. A big thanks to my writing group – Ramona and
Evie, who provided insightful feedback, accountability, and a great balance of support and
challenge. I look forward to our work continuing on well past our meetings in the conference
room. I also have much appreciation for the time and helpful information Ryan, Teri, and Eric
provided with their statistical knowledge throughout this study.
I am indebted to all the teachers, principals, and contacts who were so extremely helpful
and accommodating in helping me gain access and coordinate with me for my data collections,
especially Lara Slee for being unbelievably giving in her advice, time, and connections in
helping me collect data at many schools. Thanks also to Leanne for being a great research
assistant in the schools! Lastly, I would like to thank all the girls and program directors that I
have worked with through programming and research.

!

vi

TABLE OF CONTENTS
LIST OF TABLES ..........................................................................................................................x
LIST OF FIGURES ..................................................................................................................... xiii
CHAPTER 1
INTRODUCTION ...........................................................................................................................1
Background of the Study .....................................................................................................1
Statement of the Problem.....................................................................................................2
Study Purposes ....................................................................................................................5
Significance .........................................................................................................................6
Gender role conflict .................................................................................................6
Quantitative measure of psychosocial concerns ......................................................7
Views on gender appropriateness of sports .............................................................7
Examination across race and school type ................................................................8
Research Questions and Hypotheses ..................................................................................8
Limitations and Assumptions ...........................................................................................12
Operational Definitions......................................................................................................13
CHAPTER 2
REVIEW OF LITERATURE ........................................................................................................14
Boys and Girls’ Sport Participation Trends .......................................................................14
Girls’ Sport Participation Trends .......................................................................................15
Benefits Girls Attain in Sport ............................................................................................17
Boys and Girls Sport Participation and Withdrawal..........................................................17
Psychosocial Concerns for Girls in Sport ..........................................................................21
Female/Athlete Paradox.....................................................................................................24
Gender Appropriateness of Sport ......................................................................................27
Gender Role Conflict ........................................................................................................32
Differences Across Race....................................................................................................40
African American females’ concern with sport participation................................41
African American females’ views on gender appropriateness of sport .................43
Synthesis of Research Findings .........................................................................................45
Restatement of Purpose of Study.......................................................................................46
CHAPTER 3
METHOD ......................................................................................................................................48
Research Design ................................................................................................................48
Participant Target Population ............................................................................................48
Participant Selection ..........................................................................................................49
Procedure ...........................................................................................................................50
Measures ............................................................................................................................52
Gender role conflict ...............................................................................................53
Psychosocial concerns ...........................................................................................56
!

vii

Gender appropriateness of sport ............................................................................59
Sport participation type..........................................................................................60
Psychosocial concerns sport participation type .....................................................60
Total middle school sport participation .................................................................60
Number of sports participated in middle school ....................................................61
Middle school sport involvement ..........................................................................61
Current sport participation .....................................................................................61
Number of sports currently participating in...........................................................61
Demographic information......................................................................................61
CHAPTER 4
RESULTS ......................................................................................................................................62
Treatment of the Data ........................................................................................................62
Psychosocial concerns ...........................................................................................64
Gender role conflict ...............................................................................................67
Demographic Profile..........................................................................................................68
Evaluation of Research Questions for Purpose 1...............................................................68
Research question 1a. ............................................................................................68
Research question 1b. ............................................................................................71
Research question 1c. ............................................................................................73
Research question 1d. ............................................................................................75
Research question 1e. ............................................................................................75
Research question 1f. .............................................................................................75
Research question 1g. ............................................................................................80
Research question 1h. ............................................................................................83
Evaluation of Research Questions for Purpose 2...............................................................84
Research question 2a. ............................................................................................84
Research question 2b. ............................................................................................93
Research question 2c. ............................................................................................99
Research question 2d. ..........................................................................................101
Evaluation of Research Questions for Purpose 3.............................................................102
Research question 3a. ..........................................................................................102
Research question 3b. ..........................................................................................104
Research question 3c. ..........................................................................................106
Research question 3d. ..........................................................................................112
CHAPTER 5
DISCUSSION..............................................................................................................................115
Area 1: Gender Role Conflict and Psychosocial Concerns .............................................115
Area 2: Current Sport Participation .................................................................................121
Area 3: Gender Appropriateness of Sports .....................................................................128
Strengths and Limitations ................................................................................................133
Recommendations for Future Research ...........................................................................134
Practical Implications ......................................................................................................138
FOOTNOTE ................................................................................................................................140
!

viii

APPENDICES .............................................................................................................................142
APPENDIX A. Informed Consent/Participant Assent Form...........................................143
APPENDIX B. Questionnaire..........................................................................................146
APPENDIX C. Factor Analysis of Psychosocial Concerns Scale from Pilot Data
Collection with 100-level Kinesiology students (n = 221) ..............................................161
APPENDIX D. Number and Percentages for the Gender Role Conflict 10 items, Average
Gender Role Conflict by Category, and Impact – Gender Role Conflict (n = 236) ........162
APPENDIX E. Number and Percentages for the Psychosocial Concerns 23 items and
Psychosocial Concerns Total (n = 236) ...........................................................................163
REFERENCES ............................................................................................................................165

!

ix

LIST OF TABLES

Table 1. Low SES Urban and Middle SES Suburban School Demographics ...............................50
Table 2. Comparison of Fit Indices for 4-Factor and 2-Factor Model from Pilot Test CFA (n =
206) ................................................................................................................................................58
Table 3. Descriptive Statistics for Dependent Variables of Interest (n = 236) ..............................62
Table 4. Descriptive Statistics for Collapsed Dependent Variables of Interest (n = 236) .............63
Table 5. Factor Analysis Results of Psychosocial Concerns Scale (n = 236) ...............................65
Table 6. Number and Percentages for the Gender Role Conflict 10 items and Average Gender
Role Conflict by Category with Collapsed Categories (n = 236) ..................................................69
Table 7. Impact – Gender Role Conflict with Collapsed Categories (n = 236).............................70
Table 8. Original and Collapsed Means and Standard Deviations for Total – Gender Role
Conflict and Impact – Gender Role Conflict .................................................................................70
Table 9. Median and Mean Ranks for Total – Gender Role Conflict by Race and Within School
Type ..............................................................................................................................................72
Table 10. Median and Mean Ranks for Impact – Gender Role Conflict by Race and Within
School Type ...................................................................................................................................73
Table 11. Median and Mean Ranks for Total – Gender Role Conflict by School Type and Within
Race ...............................................................................................................................................73
Table 12. Median and Mean Ranks for Impact – Gender Role Conflict by School Type and
Within Race ...................................................................................................................................74
Table 13. Number and Percentages for the Psychosocial Concerns 23 items and Average
Psychosocial Concerns by Category with Collapsed Categories (n = 236)...................................76
Table 14. Impact – Psychosocial Concerns with Collapsed Categories (n = 234) ........................77
Table 15. Spearman Correlations between Psychosocial Concerns Subscales and Impact –
Psychosocial Concerns .................................................................................................................77
Table 16. Collapsed Means and Standard Deviations for Psychosocial Concerns Subscales .......78
Table 17. Original Means and Standard Deviations for Impact – Psychosocial Concerns ...........79
!

x

Table 18. Median and Mean Ranks by Race for Physicality of Sport and Image Concerns .........80
Table 19. Median and Mean Ranks by Race and Within School Type for Image Concerns ........81
Table 20. Median and Mean Ranks by Race and Within School Type for Item #3 Messing up
Hair ................................................................................................................................................82
Table 21. Median and Mean Ranks by School Type for Physicality of Sport and Image Concerns
.......................................................................................................................................................83
Table 22. Top 10 Sports Participated In by Race and School Type ..............................................85
Table 23. Middle School Total Sport Participation Score and Average Number of Sports
Participated In ................................................................................................................................86
Table 24. Middle School Total Sport Participation Score and Average Number of Sports
Participated In by Race and School Type ......................................................................................87
Table 25. Current Sport Participation ............................................................................................90
Table 26. Average Number of Current Sports Participating In .....................................................91
Table 27. Current Sport Participation by Race and School Type ..................................................92
Table 28. Average Number of Current Sports Participating In by Participant Group...................93
Table 29. Logistic Regression Predicting Likelihood of Current Sport Participation...................95
Table 30. Logistic Regression Predicting Likelihood of Current Sport Participation for African
American Girls (n = 63).................................................................................................................97
Table 31. Logistic Regression Predicting Likelihood of Current Sport Participation for White
Girls (n = 102)................................................................................................................................98
Table 32. Original and Collapsed Means and Standard Deviations for Total – Gender Role
Conflict and Impact – Gender Role Conflict by Participant Status .............................................100
Table 33. Median and Mean Ranks for Total – Gender Role Conflict by Participant Status......100
Table 34. Median and Mean Ranks for Impact – Gender Role Conflict by Participant Status ...100
Table 35. Median and Mean Ranks for Image Concerns by Participant Status ..........................101
Table 36. Median and Mean Ranks for Impact – Psychosocial Concerns by Participant Status 102

!

xi

Table 37. Views on Masculinity and Femininity of Sports Overall (n = 236) ............................103
Table 38. Views on Masculinity and Femininity of Sports by Race and School Type (n = 166)105
Table 39. Number and Percentages of Participants who listed a Masculine, Neutral, or Feminine
Sport on their Psychosocial Concerns Survey (n = 232) .............................................................107
Table 40. Views on Masculinity and Femininity for Current Study and Parsons & Betz (2001)
Interpreted Results .......................................................................................................................108
Table 41. Psychosocial Concerns Sport Participation Type Classification using Parsons & Betz
(2001)...........................................................................................................................................109
Table 42. Classification of Sport Participation Type using Participant Ratings .........................113
Table 43. Collapsed Categorization using Parsons & Betz (2001) for Masculine-Type, FeminineType, and Combination/Neutral of Sport Participation ...............................................................113
Table 44. Research Questions, Hypotheses, and Findings for Area 1.........................................116
Table 45. Research Questions, Hypotheses, and Findings for Area 2.........................................121
Table 46. Research Questions, Hypotheses, and Findings for Area 3.........................................128
Table 47. Factor Analysis of Psychosocial Concerns Scale from Pilot Data Collection with 100level Kinesiology Students (n = 221) ..........................................................................................161
Table 48. Number and Percentages for the Gender Role Conflict 10 items, Average Gender Role
Conflict by Category, and Impact – Gender Role Conflict (n = 236) .........................................162
Table 49. Number and Percentages for the Psychosocial Concerns 23 items, Average
Psychosocial Concerns by Category, and Impact – Psychosocial Concerns (n = 236) ...............163

!

xii

LIST OF FIGURES

Figure 1. Image Concerns median scores by race and school type ...............................................81
Figure 2. Interaction between race and school type for Total Middle School Sport Participation
scores .............................................................................................................................................89
Figure 3. Interaction between race and school type for Number of Middle School Sports
Participated in ................................................................................................................................89
Figure 4. Views on Sports survey used in questionnaire .............................................................149
!

!

xiii

CHAPTER 1
INTRODUCTION
Background of the Study
Participating in sport is one of the most popular activities that youth take part in, with
over 44 million boys and girls having played an organized sport in 2008 (National Council of
Youth Sports, 2008). This participation in organized youth sport has consistently been dominated
by males, with males constituting 66% of the participants in 2008. This is also the case in high
school sports in the United States, with males making up 59% of this demographic (National
Federation of State High School Associations, 2012). While female participation in sport has for
the most part steadily increased since the passage of Title IX, youth sport is still an arena where
males make up the majority of participants.
There are two specific trends that can be identified by looking within girls’ participation
in sport. First, girls have been found to start sport later and to quit earlier than their male
counterparts (Sabo & Veliz, 2008), and while males also leave sport during their high school
years, females do so at a much higher rate (Gould, 1987). Second, not all girls participate at the
same rate in sports; there are differences when considering a girl’s race and the community that
she comes from. Girls from urban areas have been found to be the lowest participating sector in
youth sport and physical activity (Sabo & Veliz, 2008). Girls from urban areas also start sport
later and quit sport earlier than their urban male and suburban girl counterparts (Sabo, 2009) with
African American girls’ sport participation particularly negatively affected (Troutman & Dufur,
2007). These trends are especially disheartening when considering the various benefits girls can
achieve through sport participation. Staurowsky and colleagues (2009) summarized this research
and found that benefits included decreased chances of teenage pregnancy, decreased incidences

!

1!

of substance use and abuse, increased academic success, and improved mental health and wellbeing.
Knowing the disparities that exist both with and within girls’ sport participation, and the
benefits that can be attained through their sport involvement, it is imperative to examine the
factors that contribute to girls’ lower participation rates. Numerous factors have been examined
to shed light on why girls do not participate at the same rate as boys, including lower perceived
athletic competence (Butcher, Linder, & Jones, 2002), accessibility to sport and lack of
knowledge (Casper, Bocarro, Kanters, & Floyd, 2011), and absence of role models (Arient,
2006). There has also been some literature that specifically highlights some psychosocial
concerns that girls have with their sport and physical activity participation. Qualitative studies,
for example, have yielded a rich picture of a variety of psychosocial concerns that adolescent
girls experience with their sport participation. These include concerns over: (a) body-centered
issues (e.g., concerns with muscles) (Biscomb, Matheson, Beckerman, Tungatt, & Jarrett, 2000;
McCallister, Blinde, & Phillips, 2003); (b) what others think (Slater & Tiggemann, 2010; Vu,
Murrie, Gonzalez, & Jobe, 2006); (c) sweating (Biscomb et al., 2000; Vu et al., 2006) or messing
up one’s hair (Malcom, 2003); and (d) with being negatively stereotyped as a lesbian or tomboy
(Adams, Schmitke, & Franklin, 2005; Cockburn & Clarke, 2002; T. Taylor, Legrand, & Newton,
1999).
Statement of the Problem
Not only are females participating at lower rates than their male counterparts, but girls of
color, particularly African American girls from low-income communities, seem to be specifically
limited in their sport participation. With sport still being considered a highly male domain, many
researchers have examined female sport participation through the lens of the female/athlete

!

2!

paradox (e.g., Clasen, 2001; Krane, Choi, Baird, Aimar, & Kauer, 2004; Larabee, 2011), with
females having to negotiate both a feminine and an athletic role. Further, the type of sport that a
woman participates in also has been shown to influence her experiences in sport. Sports defined
as socially inappropriate (Kane, 1988), or socially unacceptable (Anthrop & Allison, 1983; Sage
& Loudermilk, 1979), or having negative or masculine stereotypes associated with them (Engel,
1994; T. Taylor et al., 1999), are usually viewed as less gender appropriate and socially
acceptable. In these studies, girls typically participate less in and/or experience more conflict in
these sports.
Many researchers have studied the female/athlete paradox and gender sport
appropriateness through analyzing the gender role conflict experienced by females, positing that
females experience role conflict with being both an athlete and female. Role conflict received
much attention in the late 1970’s through the 1990’s, when many more girls started to participate
in what was always considered a masculine activity – sport (Czisma, Wittig, & Schurr, 1988).
Numerous quantitative studies were done during this time (Allison & Butler, 1984; Anthrop &
Allison, 1983; Desertrain & Weiss, 1988; Sage & Loudermilk, 1979), resulting in mixed findings
about the prevalence of role conflict with female athletes. While it could perhaps be true that role
conflict has “outlived its usefulness” (Allison, 1991, p. 50), there are some methodological
concerns in how role conflict has been studied that need to be considered before reaching this
conclusion. Because most of this research has been done with high school and college athletes,
some researchers contend that many of the girls who have experienced conflict might have
dropped out already (Sage & Loudermilk, 1979), or have already made sense of each of these
roles and therefore ascribed more importance to their sport participation to reduce the conflict
(Goldberg & Chandler, 1991). It has also been suggested that women in society can now more

!

3!

easily navigate between their roles as an athlete and as female, thereby decreasing possible
conflict (Krane et al., 2004; Ross & Shinew, 2008). While there is this evidence that competitive,
elite, and collegiate female athletes have found ways to negotiate their feminine and athletic
roles, thereby diminishing role conflict, little is known about this ability to do so with a younger
population of adolescent girls, with this age being a time when “gender-role stereotypes and
expectations are likely to become particularly influential” (Guillet, Sarrazin, Fontayne, &
Brustad, 2006, p. 359). There is also a small body of qualitative research that supports that
females, and particularly adolescent girls, experience unique psychosocial concerns with their
sport participation in comparison to their male counterparts, with evidence of these concerns
leading to decreased participation in sport. With no quantitative measure of these concerns, there
is still a lack of knowledge on the prevalence of these psychosocial dilemmas, and how these
concerns impact girls’ participation in and/or withdrawal from sport.
The vast majority of the research in the areas outlined above have also represented a
1

predominantly white population. In the small number of studies that have had diverse
representation of races or have focused on African American girls, some differences emerged
among girls in their reported concerns with sport and physical activity when considering race and
ethnicity. For example, African American girls have discussed a “beauty cost” with sport or
physical activity participation, especially in regard to their hair and other cosmetic concerns
(Boyington et al., 2008; Wright, Griffes, & Gould, n.d.) and also being less concerned about
body size in comparison to white girls (Mabry et al., 2003). To the best of the investigators’
knowledge, however, there has been no research conducted that looks directly at gender role
conflict or gender appropriateness of sport with African American females, and only a few
studies that have looked at African American females’ views on and experiences with sports (e.g.,

!

4!

Collins, 2011; Grieser et al., 2006; Harrison, Lee, & Belcher, 1999). Therefore, research looking
across race with adolescent girls’ sport participation is quite rare. This type of research could be
particularly enriching in illuminating any differences that might exist between girls from lowincome urban areas (the least participating sector in sports and physical activity) and white girls
from suburban areas (on whom research has primarily focused) in their psychosocial concerns,
gender role conflict, views of the gender appropriateness of sport, and how these might impact
their sport participation.
Study Purposes
There were three major purposes of this study. The first purpose of this study was to
examine girls’ experiences with gender role conflict and psychosocial concerns associated with
their middle school sport participation and determine if there were differences when considering
a girl’s race and type of school that she attends. Namely, this area focused on whether girls: (a)
experienced gender role conflict and/or psychosocial concerns with their sport participation; (b)
differed significantly on the amount and type of gender role conflict and/or psychosocial
concerns they reported having experienced with their middle school sport participation; and (c)
differed significantly on the impact that gender role conflict and/or psychosocial concerns have
had on their past and current sport participation. Related to this purpose was the use of an
instrument that the investigator developed that allowed for the psychosocial concerns that have
arose from previous literature to be quantitatively measured and examined.
The second purpose of the study was to examine current sport participation. This was
done first by using current sport participation as a dependent variable against which significant
predictors for girls’ current involvement in sport in high school were examined. Current sport
participation was also used as an independent variable by comparing gender role conflict and

!

5!

psychosocial concerns among current sport and non-sport participants. Lastly, the third purpose
of this study was to ascertain what sports girls perceived as masculine, feminine, and neutral and
examine if there were differences when considering a girl’s race and type of school that she
attends. It was then determined if girls who have participated in what are viewed by their
respective group as masculine sports differed in their gender role conflict and psychosocial
concerns scores in comparison to girls who have participated primarily in what are perceived by
their respective group as feminine sports.
Significance
Gender role conflict. This study furthers the research in this area by examining if gender
role conflict exists as a concern for females, through using different methodologies that previous
research studies have not used. Some researchers have posited that girls who experience a great
deal of conflict have already dropped out of sport, hence the low reports of gender role conflict.
Therefore, this study had high school girls who both currently participated in and did not
participate in sport reflect back on their previous sport participation, and it allowed girls who had
dropped out of sport to report on their gender role conflict. Related to this, studies of gender role
conflict have predominantly used high school and college athletes; with only two known studies
using middle school aged participants (Arient, 2006; Miller & Heinrich, 2001). However, this
study is the only known one that was able to target girls who had played a sport in middle school,
and to look at their subsequent high school sport participation. This fills a gap in the literature by
providing information from girls who have dropped out during the late middle school or high
school years. Lastly, examining individual items on the role conflict inventory allowed for a test
of Allison’s (1991) hypothesis that there are two types of role conflict (internal and external),
with girls experiencing higher amounts of external conflict. By approaching the study of gender

!

6!

role conflict in this way, this study addressed not only if conflict does exist, but whether there are
differences between internal and external conflict, and what influence such differences have on
girls’ past and current sport involvement.
Quantitative measure of psychosocial concerns. An instrument was developed by the
investigator that was used in this study to address these concerns. In addressing the relatively low
amounts of role conflict that have been reported in previous studies, Allison (1991) suggested it
would be fruitful for researchers to think past the idea of conflicting roles and not assume that
females, by default, experience conflict with their sport participation. Therefore, this quantitative
measure of psychosocial concerns phrases questions differently by simply inquiring about a
variety of psychosocial concerns that have arisen from the literature about girls’ sport
participation, without assuming that all girls necessarily experience conflict. Also, this measure
includes concerns that have arisen more recently in the literature and that were not asked on the
Role Conflict – Experienced (role conflict inventory used in aforementioned studies), thereby
shedding light on these new concerns. This instrument is needed as the qualitative literature
continues to highlight a variety of psychosocial concerns that some girls report experiencing with
their sport involvement, yet little is known about the scale and scope of these concerns, and the
impact these concerns have on girls’ past and current involvement with sport. Therefore, this
study adds to the literature by filling this void in the previous research.
Views on gender appropriateness of sports. This study also provides information on
each group’s views on the gender appropriateness (as measured by one’s view on how masculine,
neutral, or feminine they perceive a sport to be) of a variety of sports. Namely, it allows for two
questions to be answered: (a) what sports are now seen by girls as masculine, neutral, or
feminine; and (b) what relationship is there between participation in sports that are viewed as

!

7!

masculine or feminine and the gender role conflict and psychosocial concerns that girls report
experiencing in these respective sports.
Examination across race and type of school. As Cooky (2009) stated, “In popular
culture, the sport experiences of white middle-class girls and women have come to serve as the
representation of all girls and women” (p. 266). Cooky further contended that this leads to an
inaccurate representation of how gender, race, and class intersect to comprise girls’ sport
experiences. One of the overarching goals of this study was (a) to gain more information on
African American adolescent girls’ views of the gender appropriateness of sport participation
and the psychosocial concerns and gender role conflict they experienced, thereby adding to the
scant literature; and (b) to compare these results with a population that is widely studied and also
seen as the highest participators in female youth sport, predominantly white girls who reside in
suburban areas. This is also in line with Wiese-Bjornstal and Lavoi’s (2007) suggestions for the
need of studies that examine the sociocultural factors that influence girls’ sport and physical
activity participation, and that specifically address the “intersections among race, class, and
gender as a means to understand the barriers preventing girls from participating in physical
activity” (p. 89). Lastly, the investigator was interested to see if there were differences between
these groups of girls in terms of sports participated in, current and past sport participation rates,
and their respective views on the gender appropriateness of sports.
Research Questions and Hypotheses
Listed below are the research questions and related hypotheses for this study. Some
research questions do not have hypotheses due to a limited amount of research that has addressed
these areas or the mixed findings that have arisen from research in these areas.

!

8!

Research question 1a. What is the prevalence and impact of gender role conflict with
adolescent girls’ middle school sport participation?
Research question 1b. Do girls differ in their retrospectively reported Total – Gender Role
Conflict and Impact – Gender Role Conflict from their middle school sport participation based
on their race?
Research question 1c. Do girls differ in their retrospectively reported Total – Gender Role
Conflict and Impact – Gender Role Conflict from their middle school sport participation based
on the type of school they attend?
Research question 1d. Do girls differ in their retrospectively reported external and internal
gender role conflict scores from their middle school sport participation based on their race?
Research question 1e. Do girls differ in their retrospectively reported external and internal
gender role conflict scores from their middle school sport participation based on the type of
school they attend?
Research question 1f. What is the prevalence and impact of psychosocial concerns with
adolescent girls’ middle school sport participation?
H1: Psychosocial concerns measures will have higher scores in comparison to the Gender
Role Conflict measures.
Research question 1g. Do girls differ in their retrospectively reported psychosocial concerns
scores from their middle school sport participation based on their race?
H1: Both African American girls and white girls will have significantly higher
retrospectively reported scores on the Physicality of Sport subscale compared to their scores on
the Image Concerns subscale.

!

9!

H2: African American girls will have significantly higher Image Concerns scores than
white girls.
Research question 1h. Do girls differ in their retrospectively reported psychosocial concerns
scores from their middle school sport participation based on school type?
Exploratory H1: Girls from both low SES urban and middle SES suburban schools will
have significantly higher retrospectively reported scores on the Physicality of Sport subscale
compared to their scores on the Image Concerns subscale.
Research question 2a. Are there differences in girls’ middle school and current sport
participation based on race or school type?
H1a: White girls from middle SES suburban schools will have the highest involvement
in middle school sports when compared with African American girls from both school types and
white girls from low SES urban schools.
H1b: White girls from middle SES suburban schools will have the highest current rates
of participation in sports when compared with African American girls from both school types
and white girls from low SES urban schools.
Research question 2b. What variables serve as the best predictors for girls’ current sport
involvement?
H1: Total - Gender Role Conflict (negative relationship), Psychosocial Concerns Scores
(negative relationship) and Total Middle School Sport Participation will serve as
significant predictors for current sport participation.
H2: Race and school type will also predict current sport participation in that being white
and being from a middle SES suburban school will serve as significant predictors for
current sport participation.

!

10!

Research question 2c. Do current sport participants and girls who have discontinued their sport
participation differ on their retrospectively reported Total – Gender Role Conflict or Impact –
Gender Role Conflict from their middle school sport participation?
H1: Girls who have withdrawn from sport will have significantly higher retrospectively
reported gender role conflict and impact of gender role conflict scores from their middle school
sport participation, compared to girls still participating in sport.
Research question 2d. Do current sport participants and girls who have discontinued their sport
participation differ on their retrospectively reported psychosocial concerns scores or impact of
psychosocial concerns from their middle school sport participation?
H1: Girls who have withdrawn from sport will have significantly higher retrospectively
reported psychosocial concerns and impact of psychosocial concerns scores from their middle
school sport participation, compared to girls still participating in sport.
Research question 3a. Do girls’ perceptions’ of what sports they view as masculine, feminine,
and neutral differ from those reported in past research (e.g., Matteo, 1986)?
Research question 3b. Do girls differ in their perceptions of what sports they view as masculine,
feminine, and neutral based on their race, school type, or the interaction of race and school type?
Research question 3c. Do girls differ on their retrospectively reported psychosocial concerns
scores or impact of psychosocial concerns from their middle school sport participation based on
their Psychosocial Concerns Sport Participation Type (masculine, feminine, or neutral rating on
the sport they listed on the psychosocial concerns survey) and based on their current involvement
in this sport (i.e., (a) quit or (b) still participating)?
H1: Girls who listed having participated in a masculine sport will have significantly
higher retrospectively reported scores on the subscales Impact of Others, Negative Stereotypes,

!

11!

and Physicality of Sport from their middle school participation than girls who participated in
feminine sports. These differences will be even more pronounced when comparing between girls
who have quit and those who are still participating in their sport.
H2: Girls who listed having participated in a feminine sport will have significantly higher
retrospectively reported scores on the subscales Impact of Uniforms and Image Concerns than
girls who participated in masculine or neutral sport types. These differences will be even more
pronounced when comparing between girls who have quit and those who are still participating in
their sport.
Research question 3d. Looking at girls’ overall sport participation, do girls who participated in
(a) masculine-type sports differ in their Total – Gender Role Conflict scores from those who
participated in (b) feminine-type sports, and are there differences based on middle school sport
level of level of involvement?
H1: Girls who are low middle school sport participators and who participated in
masculine-type sports will have the highest retrospectively reported gender role conflict scores
from their middle school participation than all other girls.
Limitations and Assumptions
This study is based on the assumption that sport participation is beneficial in the lives of
girls and boys and that equal sport participation between the genders is something that should be
strived for at the youth and high school levels. Another assumption is that the high school
students in this study attended the same type of school in middle school.
A limitation of this study is that it is self-report in nature, and the investigator could only
request that participants answer the items on the questionnaire openly and honestly and that
participants fully comprehended all of the questions. This was addressed through a pilot study

!

12!

with only four participants. One of the strengths of the design of this study is that it includes
current non-sport participants. However, by doing so, high school girls are being asked to reflect
back to their middle school sport participation, with recall and memory being a limitation
because of this retrospective part of the investigation.
Operational Definitions
th

th

Middle school: While some middle schools are only 7 and 8 grade, for the purpose of this
study to be consistent with other research that has looked at grades 6, 7, and 8, middle school
will be defined to include grades 6, 7, and 8.

!

13!

CHAPTER 2
REVIEW OF LITERATURE
Boys and Girls’ Sport Participation Trends
Participation in organized youth sport for both boys and girls has been steadily increasing
over the last 15 years, with over 44 million youth participating in an organized sport in 2008,
compared to roughly 38 million in 1997 (National Council of Youth Sports, 2008). This increase
in youth sport participation mirrors a similar pattern seen in high school sports, with participation
also steadily increasing in each of the last 15 years (National Federation of State High School
Associations, 2012). The number of girls participating in organized sport has increased
tremendously and consistently over the last 40 years since the passage of Title IX (Monaghan,
2012). There still exists, however, a large discrepancy between boys’ and girls’ sport
participation, with a recent report illustrating how the growth in girls’ high school sport
participation has even reversed in the 2000’s (Sabo & Veliz, 2012). When Title IX was passed in
1972, there was a staggering difference between boys and girls in high school sport participation,
with over 3 million boys and just fewer than 300,000 females participating in high school sports.
Since then, female participation in sports increased immensely, with 2011-2012 having the
highest level of girls’ participation achieved (just over 3.2 million), placing female sport
participation at an all-time high. However, these statistics are somewhat misleading. With total
overall high school participants at 7,692,520 in 2012, this puts female participation at only 41%
of the total participants. Unfortunately, the future does not look much brighter, as data from the
National Council on Youth Sports’ 2008 report show that, while their overall participation
increased from 2000 to 2007, in comparison to boys, girls are still participating much less (37%
of overall participants in 2000 compared to 34% in 2007). Therefore, while some may argue that

!

14!

girls’ participation in sports continues to flourish at encouraging rates, there is evidence to
suggest that much progress can still be made.
Girls’ Sport Participation Trends
Looking within girls’ sport participation, a few trends stand out. One pattern that is
especially concerning is girls’ age of entry and exit from sport. Girls’ sport participation shows a
major decline in the early adolescent years. Girls are typically participating the most in sixth to
the eighth grades then having their participation decrease throughout high school. Interestingly,
high school is a time when sport participation is starting to be seen as opposing femininity (Engel,
1994).While boys also experience a decrease in participation at this time, decreases are markedly
more pronounced with girls (Gould, 1987). Similar patterns in decreased participation from
middle school to high school were also represented in research by Butcher and colleagues (2002).
Specifically, they found that there were peaks in withdrawal for both males and females in
grades 7 and 10 (when youth are transitioning to a new school), and that girls in the ninth and
tenth grades were quitting sport at much higher rates than their male counterparts. Lastly, when
looking at girls’ sport participation between middle school and high school, girls from both
suburban and urban areas experience decreases in their participation. However, these decreases
are much larger with urban girls.
The percent of urban girls playing an organized sport declines from 78% in middle school
to 59% in high school, while girls from suburban areas experience only a 1% decline, from 70%
to 69% in their transition (Sabo & Veliz, 2008). Also, girls in urban areas experience the highest
sport dropout rates in middle school (24% compared to suburban girls at 6%). In making sense of
this, Sabo (2009) discussed how girls from low-income communities and families have the
lowest rates of participation in sports; these girls “enter sports later in childhood than rural or

!

15!

suburban girls and boys, they have the highest dropout rates during the middle school years, and
they are the "have-nots" of American physical education" (Sabo, 2009, p. 40). Sabo and Veliz
(2012) also point out the trend that urban schools tend to have a higher percentage of girls of
color. Not only do girls of color participate less than their white counterparts (Sabo & Veliz,
2008), girls of color in urban schools are also less likely to participate in sport outside of their
school. Sabo (2009) contends that girls of color from low-income areas are particularly
disadvantaged also in the shortened length of their sport participation (i.e., start sport later and
quit sport sooner).
These statistics highlight the large variability in participation when considering various
demographics. Race and social class have played a critical role in how girls are socialized into
sport, with race and socioeconomic status (SES) representing “critical components in the
formation of values and cultural practices that orient individuals to particular patterns of thinking
and feeling about sport, leisure, gender, and the body” (Greendorfer, 1993, p. 18). This complex
intersection between geography, gender, class, and race can make it especially challenging for
some girls to be active in sports and physical activity (Lavoi, 2011). Girls from low-income
families experience challenges with their access to sport and physical activity (Casper et al.,
2011; Dwyer et al., 2006), which include the elimination of physical education classes (Ewing,
Gano-Overway, Branta, & Seefeldt, 2002) and athletic programs at much higher rates than what
is seen with schools in suburban, town, and rural areas (Sabo & Veliz, 2012).
African American girls have been particularly affected because their sport involvement
has not traditionally been supported in their community (Acosta, 1993) and African American
females tend to participate less frequently in both physical activity (Kimm, Glynn, & Kriska,
2002; Vu et al., 2006) and, school sports (Sabo & Snyder, 2013), and other sports (Troutman &

!

16!

Dufur, 2007) compared to their white counterparts. Also, African American girls are more likely
to participate in school-sponsored sports than sports offered through private organizations
(Women's Sports Foundation, 1988), making them particularly vulnerable because of declines in
public school funding over the last decade and resulting eliminations of sport programs. All of
these trends highlight that there is a cause for concern, especially when considering the
numerous benefits that are afforded to girls through sport participation.
Benefits Girls Attain in Sport
There are a plethora of benefits that youth can attain through sport participation, such as
decreases in mental and general health problems (Steiner, McQuivey, Pavelski, Pitts, & Kraemer,
2000), increases in healthy behaviors (Pate, Trost, Levin, & Dowda, 2000), and gains in life
skills such as goal-setting and improved social skills (Eccles & Barber, 1999; Hansen, Larson, &
Dworkin, 2003; Marsh & Kleitman, 2003). There have also been numerous studies
demonstrating that girls experience specific and/or increased benefits in some areas. Research
summarized by the Women’s Sport Foundation (2009) highlighted a variety of benefits including
a decreased chance of teenage pregnancy, decreased incidences of substance use and abuse,
improved mental health and well-being, and increased academic success (Staurowsky et al.,
2009). Female athletes have higher grade point averages than their non-athlete counterparts (Fox,
Barr-Anderson, Neumark-Sztainer, & Wall, 2009), and there is also evidence that girls who play
sport in high school have lower dropout rates and are more likely to graduate from college
(Troutman & Dufur, 2007). In other research, girls reported benefits of participation that
included improved fitness, increased skill level, and opportunities to socialize (T. Taylor et al.,
1999), as well as feeling better and experiencing gains in confidence (Biscomb et al., 2000).
Boys and Girls Sport Participation and Withdrawal

!

17!

Given the discrepancies in girls’ versus boys’ sport participation, it is important to
understand why children in general and girls in particular withdraw from sport. Ewing and
Seefeldt (1989) found there are numerous reasons youth get involved with sport, with both boys
and girls viewing fun as the primary reason for participating. Girls differed slightly from boys in
that girls’ other top motives were to stay in shape and get exercise, whereas the boys noted skill
improvement and doing something they are good at. Although the order differs, the list of
reasons both boys and girls cite for participation are similar: making new friends was the only
participation motivation girls reported that the boys did not. In a more recent study that examined
motives for sport program participation with middle school aged children, Sirard, Pfeiffer, and
Pate (2006) also found differences between genders. Boys’ top participation motives were
competition, social benefits, and fitness, while girls’ participation motives were ranked
differently, with social and skill benefits being more important and competition and fitness
reasons less so. Looking at self-report research, three studies in particular have been conducted
to learn more about why girls participate in sport. Researchers in Australia found factors that
influenced girls’ sport participation were: they enjoyed it (77.5%), that sport helped with their
fitness (59%), that their friends also played (51.5%), and that sport improved their body image
(42.0%) (T. Taylor et al., 1999). Similar results were also found by Biscomb and colleagues
(Biscomb et al., 2000) in their focus group study with adolescent girls about their views on sports
in England. When asked why sport is desirable, the girls reported sport was fun, it was social and
gave them something to talk about with friends, and that sport was something that was away
from the family. In the US, Stewart and Taylor (2000) used questionnaires to examine adolescent
females’ rationales for participating in sport and, in line with the previous research, found that
girls in high school largely participate because sport is fun, to stay in shape, to exercise, for the

!

18!

competition, to be with friends and meet people, and for something to do. This brief synopsis of
the literature on reasons why boys and girls participate in sport also provides some insight into
their withdrawal from sport, as oftentimes the reasons reflected in the withdrawal literature are
the absence of these characteristics.
In Ewing and Seefeldt’s (1989) comprehensive study, 5,800 athletes in grades 7-12 who
had recently stopped playing a school or nonschool sport were surveyed to learn more about why
they quit. Reasons reported included lack of interest and fun, conflicts with time, poor coaching,
and an overemphasis on winning. Only one significant difference between genders was found:
girls reported experiencing too much pressure whereas pressure was not a factor in boys’ sport
withdrawal. In two other large-scale studies that have examined boys’ and girls’ sport
participation, some interesting gender differences have emerged. In Butcher, Linder, and Jones’
(2002) study that examined why youth (1,387 tenth grade students) dropped out of competitive
sports, they found the most salient withdrawal reasons were lack of enjoyment, wanting time for
other non-sport activities, and other sports taking too much time. Within their results, the
researchers noted differences between genders in that males had higher mean scores for the need
to take a job, while females reported more pressure to perform well, felt they were not good
enough (with this gaining importance as girls get into high school), needed time for studying,
and the negative impact of injuries. The Women’s Sport Foundation (2008) published similar
findings in a report examining youth sport in America and the reasons why girls and boys from
third through twelfth grade dropped out or stopped playing organized/team sport (Sabo & Veliz,
2008). The authors noted that, while both genders reported lack of fun as the number one reason
they quit, there were also numerous differences between boys and girls. Girls, in greater numbers
than boys, reported wanting to focus more on academics and other activities or clubs (with the

!

19!

latter reason not reported by boys). Results from follow-up focus groups provided other
interesting reasons for withdrawal including that girls have family and money concerns and other
activities such as babysitting and jobs.
To better understand some of the barriers that youth face with sport participation, Casper
and colleagues (Casper et al., 2011) used a 25-item instrument that was constructed based on the
previous literature to examine the three conceptual areas in which youth have faced constraints
with their sport participation: intrapersonal (e.g., interest, knowledge); interpersonal (e.g., social
support); and structural (e.g., accessibility, facilities). In their study of 2,465 sixth to eighth grade
students who had withdrawn from sport, they found girls rated knowledge, psychological
constraints, and partners significantly higher than boys, and concluded that “while opportunities
may be equal, the girls in this study viewed constraints as more of a limiting factor toward
continued participation” (Casper et al., 2011 p. S36). The authors posited girls might have more
restrictions placed on them in terms of family and household duties. Additionally, they suggested
girls reported lower self-confidence and less social approval with their sport participation. The
constraints that Casper and colleagues identified provide context for many of the reasons girls
have for leaving sport.
In summary, the extant literature suggests that there are many similarities between
reasons why boys and girls play and drop out of sport. Specifically, boys and girls both report
psychological and logistical reasons for why they stop sport: time conflicts, injuries, lack of fun
and enjoyment, and lack of perceived competence. What can also be seen is that adolescent girls
have additional unique reasons for ceasing participating in sport. In the last decade, there has
been an increased focus on learning more about the discrepancies between boys’ and girls’ sport
participation, with several studies having been conducted to inquire about why girls are quitting

!

20!

or not participating in sport at the same rate as boys. What has emerged from these findings is
that some girls encounter specific psychosocial concerns with their sport participation, with the
impact of society playing a significant role in their sport experience.
Psychosocial Concerns for Girls in Sport
When girls have been given the opportunity to discuss their sport participation in more
depth, various psychosocial concerns related to their participation have emerged. For example,
while cost, work and study commitments, and the competitive environment were the reasons
adolescent girls cited for leaving sport, in follow-up interviews it was found that girls
experienced numerous concerns with their participation, including being embarrassed to play in
front of boys, fear of being teased, and a hesitancy to participate in what are considered “butch”
sports (T. Taylor et al., 1999). Similar concerns have been reported in other studies that have
explored adolescent girls’ sport and physical activity participation. These studies have illustrated
that girls have concern with their sport participation as it relates to (a) participating with boys
(Cockburn & Clarke, 2002; T. Taylor et al., 1999) and boys’ views on girls’ sport participation
(McCallister et al., 2003; Vu et al., 2006); (b) the time that sport takes which keeps girls from
being with their friends and boyfriends (Biscomb et al., 2000; Coakley & White, 1992); (c) one’s
body in terms of how muscular she is and how she looks in a sport uniform (Biscomb et al.,
2000; Cockburn & Clarke, 2002; Slater & Tiggemann, 2010; 2011); and (d) maintaining an
overall feminine and heterosexual image (Biscomb et al., 2000; Cockburn & Clarke, 2002;
Dwyer et al., 2006; McCallister et al., 2003; Slater & Tiggemann, 2010).
Adolescent girls’ sport participation has not always been viewed positively by their male
counterparts. In Vu and colleagues’ (2006) study that looked at boys’ and girls’ perceptions of
girls’ physical activity and sport participation, they found that boys had negative views of active

!

21!

girls, describing these girls as “girls who wanted to be like boys” (p. 87), and that boys would
not want to be friends with these girls. This is also reflected in the discouraging comments that
girls receive regarding their sport participation: Leaper and Brown (2008) found that three
fourths of the girls in their study had been the recipient of discouraging comments about their
sport involvement or athletic abilities. Some girls have detailed that boys ridicule them for
participating in sport, (McCallister et al., 2003), while other girls have reported being pulled by
their friends and boyfriends not to participate in sport and instead to do other social activities or
support their boyfriends in their own athletic endeavors (Coakley & White, 1992). Biscomb and
colleagues (2000) discussed that some adolescent girls experienced a “cult of femininity” as they
were highly focused on being social and feminine. The cult of femininity is characterized by the
importance of getting and keeping a boyfriend, with this being “so significantly important to
many teenage girls that participation in physical activity would be sacrificed for this sole
purpose” (p. 3).
When girls participate in sport, some report experiencing pushback from their friends and
other boys. Girls have discussed how their embarrassment with sport stemmed from them not
only feeling like they were inadequate in skill, but also from the subsequent teasing that they
experienced from their friends and especially from boys (Malcom, 2003; T. Taylor et al., 1999).
Specific concerns have focused around getting sweaty or dirty (McCallister et al., 2003), with
participants from Dwyer and colleagues’ (Dwyer et al., 2006) focus group stating “what stops a
lot of girls from doing sports is that they're worried about how they look, their make-up, and
everything." (p. 84). Other girls have discussed the “need” to look good and the concern that
with sport, this will not typically be the case, with one girl stating “you’re completely windswept,
completely sweaty and out of breath…you can’t look perfect.” (Biscomb et al., 2000, p. 660).

!

22!

Malcom (2003) outlined how girls in her study exaggerated their displays of traditionally
feminine behavior, wearing jewelry to both practices and games (even though this practice was
illegal for games) and tending to their hair (with many girls having a hair brush with them during
practices and games). Girls who do not conform to a feminine, well-put together image (i.e.,
through wearing sweats or not wearing make-up) and who are sweaty and dirty from their sport
participation were viewed negatively by boys, with these girl athletes not seen as a “turn-on” to
the boys (Vu et al., 2006).
Apprehension to play sports by some adolescent girls has also centered on sport
participation not being seen as feminine and also related to fear of being called or thought to be a
lesbian. Participants in Slater and Tiggemann’s (2010) study outlined how it was not seen as
“cool” or feminine for girls to play sports, with girls feeling that they have to choose between
being feminine or being physically active and having to “demonstrate masculine qualities, such
as being a tomboy, to play sports” (Dwyer et al., 2006, p. 84). Concerns of being perceived as
too masculine are also related to how muscular girls are, with girls expressing concern over
becoming too muscular (Biscomb et al., 2000), and feeling that they, as a female athlete, need to
be “sort of strong, not weak, but not strong, but sort of in the middle” (McCallister et al., 2003, p.
98). Not only are some girls concerned about their muscular body, they also focus attention on
attaining and maintaining heterosexual attractiveness and managing the fear of having their
heterosexuality challenged through people thinking that they might be a lesbian (Adams et al.,
2005; Guillet, Sarrazin, & Fontayne, 2000). Girls have stated how it is often equated in society
that “to be a good athlete means you’re gay” (Adams et al., 2005, p. 21), with other girls having
felt the need to actively dispel the lesbian stereotypes through making a point to talk about their
boyfriend or boys they are interested in (Cockburn & Clarke, 2002; Malcom, 2003).

!

23!

One of the main concerns girls seem to have with their sport participation is with their
uniform, especially in wearing their uniform around their friends and other boys (Biscomb et al.,
2000; Cockburn & Clarke, 2002). The adolescent girls in Biscomb and colleagues’ study further
stated that they were always aware of how they looked, and while perhaps it did not stop them
from participating in sport, their uniforms made them feel uncomfortable. In the focus groups
with adolescent girls in Australia, many girls discussed their concern about their body and their
uniform, with some girls discussing how a girls’ uniform could actually be a reason why they
leave a sport, with the uniform being ugly or with girls’ concerned that they might be too big for
their tight-fitting uniform (Slater & Tiggemann, 2010).
While there are a number of similarities, significant evidence exists to show that there are
some important differences between genders with their sport participation. Some adolescent girls,
in comparison to their male counterparts, report reasons for ceasing sport participation either
more frequently or as more salient, or, as the research using qualitative methodology has shown,
have unique psychosocial barriers and concerns that impact their sport participation that have not
been seen in the literature with boys. These barriers and concerns that affect girls’ sport
participation exist and function in a social context where, traditionally, male sport participation
has been rewarded in society and female sport participation has been challenged (Hargreaves,
1994). The psychosocial concerns that have arisen from these qualitative studies can be better
understood through examining the existing literature that has analyzed females’ unique
experiences in sport, and specifically looking at the paradox that exists for female athletes.
Female/Athlete Paradox
Sport has long been thought to be a masculine domain and activity (Czisma et al., 1988;
Matteo, 1986; Messner, 1988), with females consistently being located on the fringes. The

!

24!

essence of the female/athlete paradox is that for athletes to be successful they need to possess
masculine characteristics, thereby presenting females with the choice (or paradox) of behaving in
line with their gender or taking on masculine behaviors and values that are associated with
athletes. Explanations for this paradox can be found by looking at the behaviors and values
associated with both masculinity and femininity and how these behaviors are reinforced or
challenged with female sport participation.
According to gender schema theory, the differentiations between genders are believed to
start at birth (Bem, 1974), and through the socialization process, girls and boys learn at young
ages that certain behaviors and values are expected of them. For example, girls are expected and
encouraged to be passive and submissive while boys are encouraged to be autonomous and
active (Bem, 1974). Sport participation is an arena where these behaviors are particularly
rewarded, with sport participation considered not only ‘natural’ for males and a place where
males fulfill their social roles (Messner & Sabo, 1990), but also an avenue where boys learn to
become men through competing, developing strength, and asserting mastery over others
(Boutilier & SanGiovanni, 1983).
On the other hand, because sport has been a place where men’s participation has been
reinforced and women’s involvement has been challenged or even denigrated (Hargreaves, 1994),
women have had to follow a narrower path in navigating in their sport participation. Social
acceptance for female athletes requires women to negotiate the traditional notions of femininity
in an arena that requires characteristics that often contradict this (Krane et al., 2004; Meân &
Kassing, 2008). Appropriate female behavior and appearance in sport have been described as
being soft, girly, dainty, and clean, with these feminine behaviors being accepted and rewarded,
and with athletic behaviors such as sweating, being a tomboy, and going to the weight room

!

25!

being viewed as the opposite of femininity and therefore socially unacceptable (Krane et al.,
2004). Krane (2001) termed the negotiation that the female athlete engages in with her sport
participation the femininity balancing act. This balancing requires females who participate in
sport to continually assess their presentation and (feminine) image in light of their involvement
in a stereotypically male realm. As Krane and colleagues (2004) put forth, female athletes who
perform femininity reap many benefits, while the female athletes who are viewed as masculine
experience discrimination. While there has been a wealth of research that has examined how
women navigate this female/athlete paradox (e.g., Krane, 2001; Ross & Shinew, 2008; Royce,
Gebelt, & Duff, 2003), there have also been a number of studies that have brought to light how
adolescent girls negotiate their involvement with sport in relation to their perceptions of athletics
and femininity, with much similarity between these two age groups.
The time period of early adolescence for girls is one where their displays of traditional
femininity behaviors peak (Malcom, 2003): Girls at this age feel particularly compelled to
exhibit feminine behaviors; it is therefore unsurprising that they too experience this femininitybalancing act with their sport involvement. Earlier research has shown that adolescent girls view
their sport participation to actually be incompatible with becoming a woman and as such, they
ascribe a lower priority to sport participation (Coakley & White, 1992). Adolescent girls both
perceive and experience a wealth of outside pressure from society to enact feminine behaviors.
Cockburn and Clarke’s (2002) study that looked at adolescent girls’ experiences with physical
education highlights how these high school girls feel they are constricted in their ability to be
both feminine and a sport participant:
A girl can identify herself as a masculinized “doer” of PE (a “tomboy”), or a feminized
(“girlie”) “non-doer” of sport and physical activity. It is highly unlikely that girls can

!

26!

achieve being both physically active and (heterosexually) desirable, so they are often
obliged to choose between these images. The result is a paradox, a double standard to
which teenage girls and young women are subjected. (p. 661)
Girls experience this paradox through their participation in sport, and are often criticized, with
Adams and colleagues stating how society has been “troubled by the athletic, muscular, makeupfree, jewellery-free, sweaty female body” (Adams et al., 2005, p. 20). While it is expected and
sought after that adolescent boys are muscular and in shape, an ideal feminine body for
adolescent girls has been described as “thin, slender but fit, nice ‘boobs’, ‘tight butt’, sexy, and
pretty face” (Klomsten, Marsh, & Skaalvik, 2005, p. 631). Some contend that as long as sport
and physical education arenas continue to be “proving grounds for (hegemonic) masculinity”,
sport will remain an “inappropriate territory” for most teenage girls and young women
(Cockburn & Clarke, 2002, p. 662), with this balancing act that many adolescent girls negotiate
with their sport participation most likely to continue.
Gender Appropriateness of Sport
One of the ways that female athletes are scrutinized in their sport participation is through
their choice of sport, with some sports being seen as more appropriate for women than others,
based on how a sport conforms to traditional notions of femininity and feminine behavior. The
characteristics that define a sport as masculine or feminine are socially constructed and are based
on how males and females are thought to differ, not on any actual differences (Klomsten et al.,
2005). Sports are therefore sex-typed as appropriate or not appropriate for males or females to
participate in, based on the stereotypically held views on how masculine or feminine the sport is,
and on how one’s participation might challenge traditional sex role stereotypes. Considering this,
the type of sport a female plays not only impacts the amount of stigma she might experience, but

!

27!

also in how long she plays this sport (or if she plays it at all) (Colley, Nash, O'Donnell, &
Restorick, 1987; Engel, 1994). Gauging the social appropriateness or acceptability of female
sport participation can trace its roots back to the 1800’s. In the Victorian era, rigorous
participation in sport was frowned upon for females because of the “female frailty” belief that
women could damage their reproductive system (Theberge, 2000). Sports that were seen to
require minimal energy, motion, and exercise (e.g., tennis, bicycling) were viewed as more
acceptable, appropriate, and in line with the ideal image of a woman (Boutilier & SanGiovanni,
1983). In considering the relationship between sport participation and femininity, it is not
surprising there have been many studies that have focused on this area of sport acceptability and
appropriateness for females.
A wealth of research has been conducted in the last 50 years to examine the sex typing of
sports with males and females. In these studies, a variety of similarly termed areas have been
looked at. Specifically, researchers have examined the social acceptance of sport for females
(Anthrop & Allison, 1983; Sage & Loudermilk, 1979), while others have asked participants to
rate certain sports on their masculinity-femininity (Matteo, 1986; Snyder & Spreitzer, 1983) or
their appropriateness for males and females (Colley et al., 1987; Koivula, 1995). Sports are then
considered appropriate for one’s gender based on the stereotypical perceptions of sports on how
feminine, gender neutral, or masculine they are.
One of the earliest studies looking at sport appropriateness was done by Metheny in 1967.
She posited that there were three levels of sex appropriateness for sports, with sports being either
“not appropriate,” “may be appropriate,” or “wholly appropriate” for women. Sports categorized
as not appropriate had characteristics of (a) subduing an opponent by body contact; (b) projecting
a body through space over long distances; and (c) face to face competition where body contact

!

28!

may occur (e.g., pole vaulting, boxing, all team sports except volleyball). This is in comparison
to wholly appropriate sports that had characteristics of (a) projecting the body in an aesthetically
pleasing way; (b) applying force through a light implement; (c) overcoming resistance of a light
object; and (d) using manufactured devices to facilitate movement of the body (e.g., volleyball,
figure skating, swimming, diving, skiing, bowling, golf, and tennis). Lastly, Metheny defined
sports that may be appropriate as those sports that had characteristics between these two levels,
and included sports such as long jump, shot put, gymnastics, and shorter running races (as cited
in Kane, 1988). Using this classification, almost all team sports (except volleyball) were
classified as not appropriate for women.
Research that followed Metheny’s (1967) seminal study has shown that still, relatively
few sports are considered fully appropriate for females, with few of these being team sports. For
example, Colley and colleagues (1987) found only a small number of sports are sex-typed for
females (e.g., netball, yoga), while there were numerous sports sex-typed for males (e.g., soccer,
baseball, basketball), with sports such as volleyball, tennis, athletics, running, ice skating, and
gymnastics not being sex-typed for either males or females. Other research at this time found
somewhat similar results, with adults perceiving basketball, track, and softball as less feminine
sports, while gymnastics, tennis, and swimming were actually found to enhance femininity
(Snyder & Spreitzer, 1983).
In Matteo’s (1986) study that asked 80 male and female undergraduate students to rate 68
different sports on a scale between 1 and 9 based on how feminine, masculine, or neutral they are
viewed in society, team sports were still not being considered feminine, while sports such as
gymnastics, dance, cheerleading, synchronized swimming, and figure skating were perceived as
feminine. There were, however, a larger number of sports thought to be neutral, including sports

!

29!

such as golf, volleyball, softball, tennis, swimming, and running considered. Soccer, baseball,
and basketball were still considered masculine sports, as well as lacrosse (a sport that currently
experiences popularity with both genders).
Research conducted in Sweden with male and female undergraduate students
demonstrated that many of Matteo’s (1986) findings held when compared across cultures. The
participants in Koivula’s (1995) study in Sweden rated their views on 60 different sports on a
scale of 1 to 7, based on how appropriate they believed the sports are for men, women, or both
genders to play. Again, similar findings arose with a small number of individual sports such as
gymnastics, dance, synchronized swimming, and figure skating being considered feminine.
However, there were a few more sports considered neutral, namely basketball and softball, sports
that have traditionally been viewed as masculine. Volleyball, softball, tennis, swimming, and
running were also considered neutral, with soccer, baseball, rugby, and football considered
masculine.
In considering Metheny’s (1967), Matteo’s (1986), and Koivula’s (1995) research, some
conclusions can be drawn. Namely, the individual sports of figure skating, dance, and
gymnastics are consistently seen as sports that are appropriate for females to play. All of these
sports possess characteristics of providing beauty and visual pleasure, with these characteristics
associated with stereotyped notions of femininity (Koivula, 2001). While tennis, volleyball, and
golf have typically been seen as sports appropriate for females, there does seem to be some
variability. For example, Kane (1988) found that participants in her study rated volleyball as
similar to the other non-appropriate sports (basketball and softball) for females. This highlights
the fact that there is a large amount of flexibility in terms of what sports are considered neutral or
may be appropriate for women to play. In looking at the team sports of softball, soccer, and

!

30!

basketball – team sports that have a high amount of participation of adolescent females in the US
– soccer has consistently been described as a masculine sport, with there being evidence that
basketball and softball are now being viewed as more acceptable sports for females.
Statistics of sport participation around the time of these studies illustrate how boys
largely participate in the sports of basketball, football, and soccer, much more than girls (62% vs.
30%) (Physical Activity and Health: A Report of the Surgeon General, 1996). More recent
statistics (National Federation of State High School Associations, 2012) show how the gap
between these sports is decreasing. While males still outnumber females in high school sport
participation in basketball (535,289 males and 435,885 females) and soccer (411,757 males and
370,975 females), these numbers are quite different from those of 15 years ago. It is therefore
somewhat surprising to see that females in softball, basketball, and soccer still experience the
stigma of participating in gender non-appropriate sports, with female athletes who play these
sports being seen as less likeable than their female athlete counterparts who participate in skating,
track, and tennis (Hoiness, Weathington, & Cotrell, 2008). Overall, while there have been some
changes in the perceptions of sport appropriateness for females, it can be argued that little
progress has been attained, with Ross and Shinew (2008) suggesting from their findings that the
notions of sport appropriateness still “remain linked to traditional femininity” (p. 48). However,
given the lack of current research (e.g., Koivula, 1995; Matteo, 1986) that has quantitatively
examined the perceptions of gender appropriateness of sports, there is a need for more
contemporary data to verify and extend these findings.
Perhaps what is even more valuable than examining males’ and females’ perceptions of
sport appropriateness is seeing how these perceptions influence females’, specifically adolescent
girls’, participation in and experiences with sport. As Colley and colleagues’ (1987) concluded,

!

31!

females are more accepting overall of female sport participation, with many men still holding
narrow views of what sports are appropriate for women. It is no surprise then that some girls
downplay or decrease their participation in sport in response to the challenge they receive from
boys (Daniels & Leaper, 2006; Guillet et al., 2000), with participation in sports that are seen in
society as masculine specifically alienating girls from becoming involved due to the fear of
having their femininity challenged through being teased for playing a “guys’” sport (McCallister
et al., 2003; Slater & Tiggemann, 2010) and being called a “butch” or a boy (T. Taylor et al.,
1999). While it has been found that girls experience harassment with their sport participation in
sports that are considered feminine, such as gymnastics (Cockburn & Clarke, 2002), the majority
of concerns seem to typically reside with girls’ participation in what are considered “boys’”
sports, such as softball (e.g., T. Taylor et al., 1999). Some have concluded that girls’ decreased
participation in these masculine-typed sports during their adolescent years supports the fact that
girl’s are experiencing role conflict due to the societal pressure to conform to stereotypical
feminine roles (Engel, 1994).
Gender Role Conflict
Roles have been defined by Goffman as “a bundle of activities visibly performed before a
set of others and visibly enmeshed into the activities these others perform” (Goffman, 1961, p.
96). When individuals enact these sets of behaviors and attitudes that are thought to belong
together, they are acting consistently with their roles, with there being accompanying
expectations for appropriate behaviors that coincide with their roles. Everyone enacts roles that
are both assigned to and assumed by them, with various societal expectations attached to these
roles. Following from this, role conflict theory suggests that “individuals experience role conflict
when others’ expectations for their behavior are inconsistent with the individuals’ own

!

32!

expectations” (Chusmir, 1986, p. 397). As previously outlined, there are many differences that
exist between male and female sex roles, with the female sex role including qualities such as
weakness, subordination, and passivity (Miller & Levy, 1996), and male qualities including
active, self-reliant, competitive, and tough (Worell, 2001). Gender role conflict is thus
experienced when individuals engage in behaviors that are inconsistent with society’s
expectations for that gender.
The application of gender role conflict as an explanation for females’ decreased
participation in sport has been applied by many researchers in the examination of females’
participation in athletics, with perhaps one of the best examples of conflicting roles being
demonstrated with the female athlete paradox. Highlighting this, Allison (1991) has defined
gender role conflict as the paradox between female athleticism and femininity, with this conflict
stemming from the conflicting and discrepant expectations of masculine (athlete) and feminine
(gender-role) behavior.
In one of the earliest studies that examined role conflict with the female athlete, Sage and
Loudermilk (1979) constructed a gender role conflict inventory in order to measure the amount
of conflict collegiate female athletes perceived and experienced with their sport participation. By
having these two separate constructs, the researchers were able to distinguish between one’s
awareness of society’s gender-role expectations and one’s internal experience of conflict that
occurs when one violates their ascribed gender-role behavior (Fallon & Jome, 2007). Their
results showed that 20% of the 268 participants experienced role conflict to a great or very great
extent, with 56% having experienced either none or little role conflict. The average mean score
for role conflict experienced was 2.38 (on a 1 to 5 scale), with this score representing very little
to moderate amounts of role conflict. This score was lower than the investigators expected. In

!

33!

subsequent research that looked at role conflict in collegiate female athletes and non-athletes,
Miller and Levy (1996) found that athletes and non-athletes experienced similar levels of conflict,
(2.81 for athlete, 2.87 for non-athlete), with these exceeding the findings from Sage and
Loudermilk (1979).
In noting the early research conducted by Sage and Loudermilk (1979) that exhibited that
college females experienced somewhat low role conflict, Anthrop and Allison (1983) were
interested in seeing if perhaps more role conflict exists with high school female athletes. They
found that only 11.5% of the 133 participants experienced role conflict to a great or very great
extent, with 49% of the participants having experienced either none or very little role conflict.
The amount of role conflict experienced by these high school girls was lower than with the
college age women (average mean score for role conflict in this study was 2.18); however, the
researchers pointed out that more of the high school girls had encountered role conflict. Similar
results were also attained in research by Desertrain and Weiss (1988). In their study with female
high school athletes and non-athletes, the athletes’ average amount of role conflict experienced
was 21.9 (out of 50), while the non-athletes’ were lower (18.1). The researchers concluded that
while some girls experienced high amounts of conflict (e.g., 47) and a vast majority of girls
experienced role conflict to some degree (e.g., A Little Problem to A Great Problem), overall,
role conflict did not seem to play a major role with high school female athletes.
Lastly, two studies have been done that have examined the prevalence of role conflict
with middle school students. Miller and Heinrich’s (2001) study examined role conflict with both
college athletes (n = 34) and non-athletes (n = 46) and middle school athletes (n = 44) and nonathletes (n = 65). The results from this study showed that the middle school female non-athletes
experienced more gender role conflict (28.88 out of 50) than the middle school female athletes

!

34!

(24.68). No significant differences were found, perhaps due to the low number of participants.
However, when non-athlete groups were combined across age, the female non-athletes
experienced significantly more gender role conflict (28.78) than the athletes (23.68). In the other
known study that examined gender role conflict with middle school populations, Arient (2006)
found that girls aged 10-14 experienced gender role conflict in their softball participation,
specifically citing that girls who did not have a role model or a mother who played sports had
significantly higher amounts of conflict. Arient used a different measure in measuring role
conflict (Children’s Attitudes Toward Female Involvement in Sports), therefore not allowing for
accurate comparisons to be made with the other studies.
As previously mentioned, the research with gender role conflict with female athletes has
often addressed the gender appropriateness or acceptability of sport. There have been mixed
findings in this area. Sage and Loudermilk (1979) found that athletes who participated in the
socially approved sports (tennis, golf, swimming, and gymnastics) had significantly lower
experienced role conflict scores (2.17) than the athletes who competed in the non-socially
approved sports (softball, basketball, volleyball, track and field, and field hockey) who had
experienced role conflict scores of 2.43. Results from Anthrop and Allison’s (1983) research that
looked at three categories of sport; socially approved (golf, tennis, swimming, and gymnastics),
non-socially approved (basketball, volleyball, and track) and combination showed that the high
school females who participated in both socially approved and non approved sports (i.e., the
combination group) experienced significantly higher levels of conflict compared to the other two
groups. While there was not a significant difference between the socially approved sports and the
non-socially approved sports, there was a tendency for girls to experience more conflict with
non-socially approved sports, supporting the findings from Sage and Loudermilk (1979). In

!

35!

explaining their findings with the combination group, Anthrop and Allison (1983) put forth that
these participants “experience the best and worst from the world of sport” (p. 110), as they are
accepted in one realm but receive negative reinforcement in another.
Other research has suggested that the type of sport is not related to conflict. Surprisingly,
researchers with 44 elite female power lifters (a traditionally masculine sport as defined by
Metheny (1965)) found low levels of experienced and perceived role conflict, with participants
reporting an average score of 2.5 for perceived role conflict and 2.3 for experienced role conflict.
Related, Desertrain and Weiss (1988) found that there was no difference on the amount of role
conflict experienced by team or individual athletes, suggesting that perhaps participating in team
sports has become more acceptable.
One other area that has been measured in a select number of studies with role conflict and
female athletes was the participants’ self-report of their gender role orientation. There are mixed
findings when examining the relationship between feminine gender role orientation and athlete
status. Miller and Levy (1996) found that non-athletes had slightly lower femininity scores than
the athletes, while other researchers found that middle school female athletes had higher amounts
of feminine attributes than their non-athlete counterparts, suggesting they have not yet
“discarded their expressive characteristics” (Miller & Heinrich, 2001, p. 128). This provides
support for Malcom’s (2003) research that found that this early adolescent age of 12 and 13 is an
age where females are displaying higher levels of traditionally feminine behaviors. Also,
Desertrain and Weiss’s results showed there were no differences in the amount of role conflict
experienced for the feminine, masculine, androgynous, and undifferentiated athletes, with the
authors concluding that gender role orientation was not a factor that related to conflict for these
women in their roles of athlete and female (Desertrain & Weiss, 1988). More recent research has

!

36!

also shown that there is no relationship between a female’s athletic identity and her gender
identity (Larabee, 2011).
Researchers who have used Sage and Loudermilk’s (1979) role conflict inventory with
female athletes have suggested that this inventory is actually assessing more than one concept.
As Anthrop and Allison (1983) point out, there seems to exist a difference between internal role
conflict (self-perceptions of one’s personal worth and one’s perceived self-competence with
athletic skills), and external role conflict (pressures imposed on an individual by outside sources
such as family, community, media, and friends). They put forth that more attention should be
focused on the dimensionality of this inventory. Other differences in responses have emerged
when looking at the participant status. For example, Desertrain and Weiss (1988) found that
female athletes reported higher scores on two questions (questions eight and nine) that addressed
the incompatibility of expectations of friends, parents, and others in regards to the female gender
role (e.g., time should be allowed for social activities) and also with conflict concerning the high
amount of skill needed for a female athlete and the low level of skill expected of girls (question
two). This dimensionality concept was further supported in research by Allison and Butler
(1984). They conducted a factor analysis to examine if differences existed on the participants’
(female power lifters) scores on each of the 10 items. They found that four factors emerged;
emotion, attraction, sport-performance, and significant other, again suggesting that there are
differences between external and internal conflict (e.g., what society thinks of the female athlete
and what she thinks of herself). Allison and Butler concluded by stating that the female athlete
has a high level of awareness of the negative images and stereotypes that exist for them in sport,
with society stereotyping the female athlete, but that one should not assume that automatically
the female athlete internalizes these messages.

!

37!

Results from Fallon and Jome’s (2007) qualitative study with collegiate female rugby
players provide support for this idea of internal and external conflict, as they found that conflict
manifested itself in various ways (e.g., conflict with not being feminine enough, conflict about
their sexual orientation). Their results highlighted that women may not experience high levels of
internal conflict, as “apparent gender-role conflict might perhaps be better understood as the
external environment projecting stereotypes upon them” (Fallon & Jome, 2007, p. 319), with
Goldberg and Chandler (1991) also providing support that external conflict is most likely to be
experienced due to the roles that society and others expect from them.
In addressing the mixed and lower than expected results that have been found in the
previous research, many explanations have been put forth. Allison (1991) has suggested that
perhaps role conflict has outlived its utility, and the idea of the female athlete, as a conflicted
being does not truly exist. There are other explanations for these low reports of conflict. Some
have posited that females are actively decreasing the dissonance they experience in their roles as
an athlete and role as a female through placing more emphasis on areas that are important to
them (e.g., sport and the role of athletics), thereby serving to decrease experienced conflict
(Goldberg & Chandler, 1991; Miller & Heinrich, 2001). Sage and Loudermilk (1979) have also
put forth that girls might feel more competent in their role as an athlete, with this relating to
society being more accepting now of female athletes in the post-Title IX era (Anthrop & Allison,
1983; Sage & Loudermilk, 1979).
However, there also have been methodological concerns that have been put forth as
explanations for lower conflict scores that need to be addressed. It is a realistic idea that many
girls would have withdrawn from sport before college or high school; therefore, many of these
studies perhaps did not capture participants with higher levels of conflict, for example, girls who

!

38!

experienced the most conflict either have dropped out or did not participate in sport to begin with
(Anthrop & Allison, 1983; Sage & Loudermilk, 1979). As previously outlined, there are higher
levels of external conflict reported (Anthrop & Allison, 1983), but relatively little accounts of
internal, therefore the current inventory might not accurately measure a singular concept of
gender role conflict. Allison and Butler (1984) contend that different approaches are needed in
the study of role conflict with female athletes, especially in how it is analyzed and in the overall
methodology. Specifically, they suggested cohort sampling, as this would allow for comparisons
between dropouts and athletes and active and inactive participants, thereby providing a deeper
understanding “of the experiences of the female involved in all types, levels, and degrees of
physical activity” (p. 165). More recently, Miller and Heinrich (2001) have concluded that there
are many questions left unanswered in examining gender role conflict with the female athlete.
They suggest further research that looks at the impact of girls’ involvement with team or
individual sports, as well as examining if there is a developmental aspect to role conflict.
One of the critiques of gender role conflict theory is that it hinges on the dualistic and
contrasting views of gender (i.e., feminine) roles and athletic roles and also assumes (as
evidenced in the wording of the survey questions) that conflict is inherent in the experiences of
the female athlete (Allison, 1991). While there is evidence of the psychosocial concerns that girls
and females encounter with their sport participation that lead to conflict, there also have been
accounts where female athletes are able to occupy and successfully navigate between numerous
roles, with Cox and Thompson’s (2000) research demonstrating how females effectively occupy
various roles such as soccer player and girlfriend and is in line with what Birrel (2004, citing
Hall, 1981) contends, in that women athletes have multiple selves or roles and can employ role
distancing strategies to navigate their different role expectations. This concept of sportswomen’s

!

39!

experiences as “multiple-bodied” has been put forth in numerous studies with highly involved
collegiate and professional female athletes (Cox & Thompson, 2000; Krane et al., 2004; Royce
et al., 2003). Being able to negotiate numerous roles suggests that, while females experience
concerns and might have conflict with some aspects (e.g., external conflict) of their sport
participation, some are able to navigate this conflict, thereby decreasing their experience of
conflict. These studies provide evidence that these female athletes have understandings of both
traditional definitions of femininity, but have also worked on creating their own version. This
allows for the female to, at certain times, display traditional feminine characteristics and
behaviors and at other times not. This is demonstrated in the research with the collegiate female
athlete population in how female athletes cope with conflict and persist with their sport
participation (Krane et al., 2004; Ross & Shinew, 2008). However, there is little knowledge on
middle school female sport participants’ ability to maneuver between roles or utilize distancing
techniques to diminish any role conflict (and therefore sustain their sport participation). What is
known is that during this adolescent stage of the middle school years, the feminine role is
becoming increasingly salient (Guillet et al., 2000), with adolescent girls’ gender role identity
still evolving, thereby adolescent girls could be more susceptible to gender role conflict (Fallon
& Jome, 2007). With only two known studies involving middle school-aged females, there is still
uncertainty if middle school aged girls experience role conflict and if so, are they able to find
ways “to persist and succeed within sport” (Ross & Shinew, 2008, p. 53), like some collegiate
and high school athletes have, or does this conflict lead to decreased involvement in future sport
participation?
Differences Across Race

!

40!

African American females’ concern with sport participation. It is critical to highlight
that the vast majority of the research reviewed so far has predominantly sampled a white
population, supporting the view that the experiences of African American female athletes have
been silenced (Bruening, Armstrong, & Pastore, 2005). More needs to be known about the
“possible differences in attitudes and beliefs among adolescent girls of various ethnic and racial
backgrounds” (Grieser et al., 2006, p. 41), as well as the prevalence of role conflict among girls
of different ethnicities and geographic locations (Goldberg & Chandler, 1991). In one of the few
studies that examined African American and white girls’ experiences with sport, African
American girls differed from white girls in their perceptions of parental support and influence,
issues with transportation, and their beliefs in how boys views girls who participate in sport.
African American girls had higher incidences of boys teasing them for their sport participation
and also faced beliefs from their parents that sports were more important for boys to participate
in (Women's Sports Foundation, 1988).
Given the concern about obesity and the declines in physical activity, especially with
minority populations (e.g., Kimm et al., 2002), other studies have focused on girls’ physical
activity beliefs, that in some cases include sport participation. In these studies, similarities arose
between African American and white girls, with these girls expressing concern over their
hygiene and image (having to walk around being “nasty” for the rest of the day after they had
gym class in the morning) and issues of masculinity, with girls expressing how boys do not like
tomboys and that when girls are muscular, they get made fun of (Mabry et al., 2003). However,
Mabry and colleagues also noted that African American girls were seen to be more accepting of
different body shapes and sizes and had higher self-esteem about their body compared to white
girls, who were more concerned about their body image and wanting a smaller body size, with

!

41!

this relating to them wanting to attract boys and gain the acceptance of their female friends. In
similar research with African American girls, other differences arose, with girls reporting that
they lack support and have negative experiences in their physical education classes, and are
concerned about their appearance after participating in a sport or physical activity (Grieser et al.,
2008; W. C. Taylor et al., 1999). This concern for appearance with sport participation has been
termed a beauty cost by the authors of one study with African American girls, with the
participants discussing a concern that some girls have in that “girls, they don’t like to sweat and
get their hair messed up…They think like they do the wrong thing, they break their nails, it’s a
crisis” (Boyington et al., 2008, p. 5).
This investigator and her colleagues conducted a qualitative study with African
American middle school and high school girls to examine their perceptions of girls’ experiences
in sport; namely, what barriers and benefits they perceive girls to experience, as well as reasons
girls do and do not participate (Wright et al., N.d.). Results from this study showed that these
girls perceived numerous barriers to sport participation, including physical challenges (e.g., lack
of skill, concern with injuries), lack of support from others (e.g., boys, parents, friends), and lack
of available resources (e.g., transportation and financial concerns). Some of the participants also
discussed their concerns with sweating, the need to be a “girly girl”, and concerns about their
hair and skin. One participant stated that when she gets her hair done, she would not play sports,
as “then it get messed up. There’s no point in me getting my hair done so. I like to look pretty all
of the time. And I can’t just, you know, look nasty, it’s not me.”
This unique concern with African American female athlete’s hair has been rarely
discussed in the literature, but seems to be a considerable concern. Noting these reasons, Collin’s
dissertation (2011) examined African American female athlete’s views on their hair and found

!

42!

that black female athletes face unique and different challenges in caring for their hair, with these
challenges having “a sizeable impact on the way in which black female athletes prepare, train
and experience athletics” (p. 78). The women in this study also discussed how they felt a
pressure to be feminine and to avoid being thought of as a lesbian, a common concern that seems
to span across race.
African American females’ views on gender appropriateness of sport. As the
majority of this research on the gender appropriateness of sport was conducted with a primarily
homogenous, white population, again little is known about the experiences of other races. Wright
and colleagues (N.d.) asked girls about their barriers to participation. Girls in this study
discussed many of the same barriers that girls in other studies (reviewed above) have talked
about. For example, they discussed the hesitation for some girls to play sports because they do
not want to develop muscles, and that there is the perception that sports are seen as a boys’
activity. One participant from this study stated, “Like they (girls) kinda feel that it’s kinda boyish
to even play a sport, even if it’s like something, cheerleading”, with her viewing that girls are
probably lazier than boys and that girls “don’t really see sport as something they want to do.”
There have been a few studies that have looked at sport participation and views on sports
for females. From one study, the team sports of basketball and volleyball were more popular
with African American girls than with white girls (Grieser et al., 2006), supporting this notion of
participation in what has been considered a more masculine sport (basketball). However, the
authors go on to state that there are not many similarities across race in terms of sport and
physical activity preference and that “factors other than the activity itself likely affect
participating in physical activity programs (i.e., accessibility, social networks, cultural sensitivity
of leaders and participants, etc.)” should be considered (Grieser et al., 2006, p. 49). Others point

!

43!

out the prevalence of “race effects” that factor into what sports African American and white
females see as culturally appropriate for them, with sports such as track and field and basketball
being sports that African American females have in-group status with, while sports such as
volleyball, tennis, and golf not being as popular due to financial limitations, parental
involvement, and lack of available organizations (Bruening, Borland, & Burton, 2008).
In one study that looked specifically at the relation between gender and race schemas,
socioeconomic status (SES), and sport participation, numerous interesting findings arose.
African American adolescent females were more likely to view track sprinting as something they
participate in, compared to a relatively low amount of European American females. For both
African American males and females, basketball was the sport they felt most competent in, with
competency in basketball reported at a 4:1 ratio when compared with European American males
and females. This ratio is reversed, however, when looking at baseball and softball participation,
with European American males and females reporting much higher competence ratings in these
sports. Also of note, when looking at SES, soccer, golf, and tennis were sports that participants
of higher SES participated in, with these opportunities many times afforded to European
Americans, while football, basketball, and weightlifting are more popular with low SES
participants, with the authors suggesting that role models and the low cost of participation were
significant (Harrison et al., 1999).
In summary, while there are differences between African American and white females’
views on sports and experiences in sport, few conclusions can be drawn. In acknowledging the
little research that is comparison in nature, much can be gained in studies that examine sport
participation across race. In noting weaknesses of generalizability of their research, Richman and
Shaffer (2000) suggested that future studies with female athletes would benefit from the

!

44!

recruitment of a more diverse sample that would “treat the unique racial, cultural, and
socioeconomic backgrounds of study participants as integral…rather than simply assume that a
middle-class cultural context represents the norm for all young women” (p. 198). This type of
research would also follow the suggestion of Mabry and colleagues (2003), who found that their
initial qualitative study helped identify some important themes that could be benefited by further
quantitative examination to help shed more light and “quantify racial differences in attitudes
toward physical activity and determine whether these attitudes help explain racial differences in
physical activity levels” (Mabry et al., 2003, p. 315).
Synthesis of Research Findings
In research that has addressed the sport participation discrepancy between boys and girls,
differences have arose that highlight some girls’ unique concerns in sport. Recent qualitative
studies have shown that some adolescent girls are concerned with such things as their uniform,
being stereotyped as a lesbian, and developing muscles. There also exists a relationship with
these concerns and sport participation in appropriate and non-appropriate sports. However, there
is presently no information on the scale and scope of these concerns with adolescent girls. This
leaves a gap in the literature that needs to be addressed. One measure that has traditionally been
used in examining female sport participation as it relates to the female athlete paradox is the role
conflict inventory as used in previous research (Allison, 1991; Allison & Butler, 1984;
Desertrain & Weiss, 1988; Sage & Loudermilk, 1979). Results from studies with high school and
collegiate females have found lower than anticipated conflict scores. Before it can be concluded
that role conflict has outlived its utility (Allison, 1991), some methodological considerations
should be taken into account. From the scant number of studies that have been done with middle
school females, there is some evidence that suggests middle school girls experience gender role

!

45!

conflict in their sport participation (e.g., Arient, 2006); however, little is known on how this
affects their participation. Lastly, there is a dearth of literature that looks across race with these
areas of girls’ sport participation, leaving an overrepresented account of white, middle class
girls’ experiences in sport, with little information on others girls’ experiences.
Restatement of Purpose of Study
There were three major purposes of this study. The first purpose of this study was to
examine girls’ experiences with gender role conflict and psychosocial concerns associated with
their middle school sport participation and determine if there were differences when considering
a girl’s race and type of school that she attends. Namely, this area focused on whether girls: (a)
experienced gender role conflict and/or psychosocial concerns with their sport participation; (b)
differed significantly on the amount and type of gender role conflict and/or psychosocial
concerns they reported having experienced with their middle school sport participation; and (c)
differed significantly on the impact that gender role conflict and/or psychosocial concerns have
had on their past and current sport participation. Related to this purpose was the use of an
instrument that the investigator developed that allowed for the psychosocial concerns that have
arose from previous literature to be quantitatively measured and examined.
The second purpose of the study was to examine current sport participation. This was
done first by using current sport participation as a dependent variable against which significant
predictors for girls’ current involvement in sport in high school were examined. Current sport
participation was also used as an independent variable by comparing gender role conflict and
psychosocial concerns in current sport and non-sport participants. Lastly, the third purpose of
this study was to ascertain what sports girls perceived as masculine, feminine, and neutral and
examine if there were differences when considering a girl’s race and type of school that she

!

46!

attends. It was then determined if girls who have participated in what are viewed by their
respective group as masculine sports differed in their gender role conflict and psychosocial
concerns scores in comparison to girls who have participated primarily in what are perceived by
their respective group as feminine sports.!

!

47!

CHAPTER 3
METHOD
Research Design
The research questions in this study were addressed quantitatively through the use of
paper and pen questionnaires. The dependent and/or criterion variables of interest were gender
role conflict, psychosocial concerns, current sport participation, and views on feminine and
masculine sports. The independent and/or predictor variables included race (African American
and white), type of school attended (middle SES suburban or low SES urban), type of sport
participated in, and past and current sport involvement. Gender role conflict and psychosocial
concerns also served as predictor variables for current sport involvement.
Participant Target Population
The two specific populations of interest were ninth and tenth grade African American and
white girls who participated in at least one sport for one season when they were in middle school.
Having participants who were predominantly in the ninth and tenth grade allowed for more
th

th

accurate recall of their middle school sport experiences when compared with girls in 11 or 12
grade. Participants were drawn from middle SES suburban and low SES urban schools. This
combination of age, race, and type of school was chosen in response to four important

considerations that drove the purpose of this study. First, a wealth of previous research has been
collected on predominantly white, suburban girls (thereby only allowing for results to be
generalized to this population) who have also been shown to be the highest participating sector in
girls’ sports. Second, girls from urban low-income communities are the lowest participating
sector in girls’ sports. Third, the literature demonstrates that African American girls have
challenges that negatively impact their sport participation and also have particular concerns with

!

48!

sport participation (e.g., hair concerns). Finally, girls at the late middle school/early high school
age are the most susceptible to leaving sport and have been shown to have high levels of
traditional feminine behaviors.
Participant Selection
The investigator was interested in recruiting participants from two specific locations: (a)
suburban schools that are close to a large university and therefore represent a population of girls
that are typically studied in this type of research (middle SES suburban schools) and; (b) low
SES urban schools, as research has shown that girls participate significantly less in sports in
these schools. To attain a representative sample for the four target populations (African
American girls in middle SES suburban schools, African American girls in low SES urban
schools, white girls in middle SES suburban schools, white girls in low SES urban schools), the
investigator conducted this data collection in selected middle SES suburban and low SES urban
schools that meet certain criteria as outlined below. The investigator used two characteristics or
“indicators” of schools’ economic conditions in this study. These included whether or not the
school receives Title 1 funds and the number of students eligible for free and reduced lunches.
These two indicators are commonly used when assessing school SES (National Center for
Education Statistics, 2012). The Title 1 program (fully titled “Title 1 – Improving the Academic
Achievement of the Disadvantaged”) is a federal program in the US that provides financial
assistance to public schools that have a high percentage of children from low SES backgrounds.
In measuring poverty concentration, the proportion of students eligible for free and reduced
lunches is broken down into four levels; 0-5 percent (low poverty schools), 6-20 percent, 21-40
percent, and greater than 40 percent (high poverty schools) (Lippman, Burns, & McArthur, 1996).
Therefore, for this study, low SES urban schools were identified for participation if (a) they

!

49!

receive Title 1 funding; (b) have over 40% of their student body eligible for free and reduced
lunches; and (c) are located in a city. Middle SES suburban schools were chosen for data
collection if (a) they do not receive Title 1 funding; (b) have less than 40% of their student body
eligible for free and reduced lunches; and (c) are located in a suburban area. Information for
these three stipulations was attained using the National Center for Educational Statistics website
(http://nces.ed.gov/ccd/schoolsearch/). Data collection took place at a total of 10 schools; six low
SES urban, 4 middle SES suburban (See Table 1). Two of these 10 schools had met the two
economic criteria but not the “city” criteria, but were still included in this study. One school is
located in a small city (included as middle SES suburban) and one is located in a large suburb
(included as low SES urban), because these two schools were very similar to the other schools in
the respective categories and were representative of the target population.
Table 1
Low SES Urban and Middle SES Suburban School Demographics
School Type
Low SES urban (n = 6)
Middle SES suburban (n
= 4)
City designation
4 midsize, city; 1 large, city; 1 3 large, suburb; 1 small,
large, suburb
city
Average % of Free and Reduced
70.1%
27.3%
Lunches
Average % of white students
22.7%
64.0%
Average % of African American
60.3%
17.2%
students
Type of Classroom Questionnaire
19 health, 9 English, 2 all-girls 14 health, 2 all-girls PE,
Completed In
PE, 2 Spanish
1 nutrition
Questionnaire Completion Rate
39.6%
65.3%
Note. Questionnaire completion rate was calculated by the number of total students on the class
roster divided by the number of students who completed a questionnaire.
Procedure

!

50!

Using contacts from colleagues and the state high school athletic association, the
investigator reached out to principals, athletic directors, teachers, and administrators from
schools that fit the participation criteria, to inquire if schools were available and interested in
having the investigator conduct this research at their school. After approval was gained from
each school’s principal, the investigator worked with a contact (i.e., the principal, an athletic
director, or a teacher) at each school to ascertain what teachers might be interested in having this
research conducted in their classrooms. While some data collections were done in ninth and tenth
grade English and Spanish classes, the vast majority of the data collections took place in health
classrooms. This is in part due to some schools not allowing ninth and tenth grade core academic
classes to be involved in the research (as not to disrupt learning in those subjects), and that the
curriculum in health and physical education classes linked well with this research. Because the
majority of high schools in Michigan require a semester of health and physical education, and the
majority of high school students fulfill these requirements in ninth and tenth grade, this helped
ensure that a majority of the participants would be from the ninth grade.
The investigator sent an email to each teacher, providing an overview of the procedure
for collecting data and also outlining how they would be compensated (with a $20 Amazon gift
card) for their time and as a token of appreciation for their cooperation. It should be noted that in
accordance with recommendations from Michigan State University’s Institutional Review Board,
the teachers were not involved in the recruitment of participants. The gift card incentive was
used to acknowledge the teachers’ cooperation in the study, and was not dependent on or related
to their students participating.
The investigator made two visits to each classroom. A brief initial visit was done with
each classroom to explain the study and to distribute an informed consent/participant assent form

!

51!

(See Appendix A) to each student. At this time, the investigator provided each teacher with a
sealed box for the classroom in which students were instructed to place their signed consent
forms. There was a delay between initial visit and survey administration day to allow for the
signing and returning of consent forms. An incentive was used to aid in the returning of parental
consent forms; students received $5.00 if they had returned a parent/guardian-signed consent
form by the time the questionnaire was administered. Students received the $5.00 incentive as
long as they returned a signed form, regardless of whether a parent/guardian gave or withheld
consent. This was done in an effort to maximize consent form return rates and also to ensure
students were not coerced financially into engaging in research in which they did not want to
participate. The investigator then returned to the classroom roughly one week after consent forms
were distributed to administer the appropriate questionnaire to those students who had a signed
parent/guardian consent form and who also provided assent. Instructions were given to students
about which questionnaire to complete: If they were female and had participated in at least one
th

th

th

organized sport for at least one season when they were in 6 , 7 , or 8 grade, they were
instructed to complete the questionnaire for this study. If they did not fit those requirements (i.e.,
students who were boys and girls that did not play a sport in middle school), they completed an
alternate questionnaire that was similar in length and asked about their perceptions of girls’
experiences in sport. Data from the alternate questionnaires were not included in this current
study. Questionnaires were completed in the classroom or another quiet area (e.g., library), with
completion time taking roughly 15 minutes.
Measures
The participant questionnaire (See Appendix B) included measures for gender role
conflict, psychosocial concerns, views on gender appropriateness of sport, past and current sport

!

52!

participation and level of involvement, items addressing reasons why youth dropout or decrease
their sport participation in middle school (used to decrease response bias/increase validity of
questionnaire – not included in analyses), and other demographic information (year in school and
race/ethnicity). The questionnaire was constructed so that demographic and sport participation
questions were first, followed by questions that were more personal in nature (e.g., questions of
psychosocial concerns they might have encountered with past sport participation). This was done
following common practices in the layout of questionnaires, because it allows for the participants
to gain comfort with the questionnaire before more personal questions are posed. Pilot tests were
conducted with three eighth grade girls and one tenth grade girl. Each of the four girls completed
the questionnaire in 13-15 minutes. The investigator asked for feedback after they had finished
the questionnaire to assess the questionnaires’ readability and clarity, and small changes were
made based on these suggestions.
Gender role conflict. The Role Conflict – Experienced (RCE) was the inventory used to
measure gender role conflict for this study. This role conflict inventory was developed by Sage
and Loudermilk (1979) and was adapted from a model of role conflict that Massengale and
Locke (1976) had put forth. There are two parts to this inventory: (1) Role Conflict – Perceived;
and, (2) Role Conflict – Experienced. Each has the same 10 questions but uses different question
stems. The questions on the perceived part asks respondents about their perception of each item
posing a role conflict problem, while the questions on the experienced part asks respondents
about their personal experience with “role incompatibilities in enacting the roles of
female/female athlete” (Sage & Loudermilk, 1979, p. 91). For example, Question 1 reads,
“Because American society traditionally places little value on girls’ participation in sports, the
female athlete receives little recognition for her skills and accomplishments.” For the role

!

53!

conflict – perceived part, the wording that follows is “this seems to me to be”, with the subject
having five options to choose from 1 (Not a Problem at All) to 5 (A Problem of Very Great
Importance), while for the role conflict – experienced, the wording that follows is “I have
personally experienced this problem”, and the subject has five options to choose from 1 (Not at
All) to 5 (To a Very Great Extent). For the purposes of this study, participants were asked to
think only of their sport participation in middle school when completing this inventory.
Three small changes were made in the wording of three questions to make the questions
more “current.” One original question read “Girls should be available for dates, parties and other
social activities, but an athlete needs to devote much time to practices and evening games” and
the investigator changed it to read “Girls should be available for dating, hanging out with friends,
and other social activities, but an athlete needs to devote much time to practices and evening
games.” For the second adapted question, the original version reads “Girls are expected to look
attractive and dress well, but participation in athletic activities often results in sweat, tossled hair,
and broken fingernails;” the investigator changed it to read “Girls are expected to look attractive
and dress well, but participation in athletic activities often results in sweating, messing up one’s
hair, and broken fingernails.” For the third adapted question, the original version reads “Because
American society traditionally places little value on girls’ participation in sports, the female
athlete receives little recognition for her skills and accomplishments;” the investigator changed
“little value” to “less value”.
Construct validation for this measure was originally performed by Sage and Loudermilk
(1979), who found that items on both the perceived and experienced role conflict parts correlated
positively with the total scores of perceived and experienced role conflict, with all items on both
parts having correlation scores greater than .60 with their respective total score. All 10 items on

!

54!

the scale for both role conflict perceived and role conflict experienced are significantly positively
correlated. Reliability was established through a test-retest procedure, with reliability scores
ranging from .54 to .79 for Role Conflict – Perceived, and .58 to .83 for Role Conflict –
Experienced.
Other researchers studying role conflict have used this inventory (Allison & Butler, 1984;
Anthrop & Allison, 1983; Desertrain & Weiss, 1988; Miller & Levy, 1996; Miller & Heinrich,
2001), with the Miller and colleagues’ studies (1996, 2001) only using parts of this inventory in
combination with another general sex role conflict inventory. In addressing the purposes of this
study, only the 10 questions that measure the role conflict experienced part of the inventory were
used. Participants also answered one question that examined the impact of their role conflict on
any decreased participation they had with any sports in middle school. Participants were
instructed to rate how much of an impact that the “problems” they rated higher (e.g., 3, 4, or 5)
had on any decreased participation they experienced with sports in middle school.
Researchers have analyzed the Role Conflict inventory in numerous ways. Desertrain and
Weiss (1988) summed each of the participants’ scores to produce a total score between 10-50
and then used these averages to compare between athletes and non-athletes. Both Anthrop and
Allison (1983) and Sage and Loudermilk (1979) calculated the percentage of respondents who
responded to each of the five categories (Not at all, Very little, Moderate, Great, Very Great),
with Anthrop and Allison (1983) also calculating the average sum of responses, thereby
producing a score for each participant between 1-5. Also, the investigator was interested if there
are differences between what has been considered internal conflict (represented by questions five
and six) and external conflict (represented by questions one, three, and nine) (Anthrop & Allison,
1983). Hence, there are several options for calculating scores on this instrument.

!

55!

Psychosocial concerns. Although there has been much recent literature that has
illuminated the psychosocial concerns some girls face with their sport participation, there was no
existing quantitative measure to examine these concerns. Therefore, the investigator constructed
an instrument to be able to measure these concerns. There were a number of steps completed in
constructing and validating this instrument. First, the investigator performed a review of the
literature that has highlighted concerns that some girls have with their sport and physical activity
participation, and proceeded to identify themes that emerged from extant literature. These
included areas of concern with image (sweat, jewelry, hair, nails, makeup, uniform), with
stereotypes (tomboy and lesbian), with the physicality of sport (muscles, injuries, perception of
this sport in society as really physical), and with the impact of peers and boys (teasing, sport
requiring time away from friends).
From there, 19 items were created that addressed these possible areas of concern for
girls’ sport participation. These 19 items were then verified with two youth sport experts who
have published extensively in the field of youth sport participation, and with two PhD candidates
who have both conducted research with girls’ sport participation and have coached middle
school and high school girls. They were asked if they believed all the critical items were
included in the scale, and if any other items should be added. They confirmed that all 19 items
looked appropriate, and they could not think of any other items to add. A 5-point scale was used
with the same answer stems; 1 (Not at All) to 5 (To a Very Great Extent) utilized with Sage and
Loudermilk’s (1979) role conflict inventory. Instead of phrasing the questions to assume the
existence of conflict, subjects in this inventory are asked to respond to the question, “When you
played this sport, how concerned were you…” Instructions were provided for the participant to
choose one sport to use when completing this part of the questionnaire (due to the nature of the

!

56!

questions and how the questions could be answered differently based on what sport they
participated in).
This instrument was then pilot tested with a small group (n = 4) of teenage girls (ages 1418) to ascertain the readability and clarity of the questions, as well as to attain their feedback if
any items should be added. From this pilot, changes were made including the addition of three
questions. This iteration of the survey was then reviewed by one of the youth sport experts to
confirm the changes. This 22-item instrument was then pilot tested with 221 females from a
variety of 100-level Kinesiology classes at the investigator’s university. The stipulations for
completing this survey were that, as a female, they had played a sport (at any level: school, elite
or recreational) for at least one season in middle school and/or high school and had stopped
playing this sport before attending college. They were instructed to complete the survey with this
one sport in mind (if they had played and quit more than one sport, they were instructed to
choose the sport that they most identified with, as defined by which they played the most or liked
the best).
These data were entered into SPSS v. 20 and the results were analyzed first by running
descriptive statistics to ensure that there were no large, erroneous scores entered and to examine
the means, and then through running an exploratory factor analysis. Because it was expected no
more than four factors would emerge from these 22 items, a 4-factor analysis (PCA, direct
oblimin rotation with Kaiser Normalization) was conducted, eliminating items with factor
loadings less than .5. Eighteen of the 22 items loaded cleanly on four factors. Four factors that
emerged were then labeled: 1 - “Image Concerns” (7 items), 2 - “Fear of Negative Stereotypes”
(4 items), 3 - “Physicality of Sport” (3 items), and 4 - “Impact of Others” (4 items). The pattern
matrix from this four-factor analysis is provided in the appendix (Appendix C).

!

57!

At the end of the large pilot survey, there was space for participants to provide feedback
if any items were confusing. When possible, verbal feedback was also requested from
participants after the survey was completed. Changes were made to the instrument and a second
data collection was completed with 206 participants. Using the statistical software package
AmosTM, a confirmatory factor analysis (CFA) was conducted on the four factors that had
emerged in the exploratory analysis. The original four-factor model yielded a poor fit (See Table
2). In this model, two of the four hypothesized constructs, “Image Concerns” and “Physicality of
Sport,” all had standardized coefficients above an acceptable level of 1.00 (with item
standardized coefficients ranging between 1.00 and 3.14). The two other hypothesized constructs,
“Fear of Negative Stereotypes” and “Impact of Others,” had five of eight standardized
coefficients below 1.00 (with item standardized coefficients ranging between .56 to 1.06).
Therefore, another model was run, adapting the “piecewise approach” that Martens and
colleagues (2002) used, allowing for this poorly fitting model to be reassessed by removing
constructs. Therefore, a new model was run with the two strongest path coefficients from the
original model; “Image Concerns” and “Physicality of Sport,” with “Fear of Negative
Stereotypes” and “Impact of Others” eliminated based on their lower standardized coefficients
on the paths. This model yielded a much better fit.
Table 2
Comparison of Fit Indices for 4-Factor and 2-Factor Model from Pilot Test CFA (n = 206)
2
CFI RMSEA
X
4-Factor
483.11, p = .000, df = 129 .744 .116
2-Factor (Image Concerns, Physicality of Sport) 76.78, p = .000, df = 34
.847 .078

Model

!
The CFA only allowed the reporting of two subscales that represent 10 of the 23 items
from the original inventory. However, in considering the purposes of this investigation of

!

58!

addressing various psychosocial concerns that have arisen from the literature, and considering
that the investigator would conduct a factor analysis with the current data, all 23 items were
included on the questionnaire.
There were two sets of directions for this part of the questionnaire. The first set had the
participants answer Yes/No questions to direct them to what sport to use when answering the 23
questions. The participants were instructed to choose a sport that they had played in middle
school and since (a) quit; or (b) decreased their participation in. If neither of these situations
applied to them, they were instructed to choose a sport that they played the least in middle school.
The purpose of these instructions was to have the participant select a sport that they had
decreased their involvement with, as this addressed the purpose of the study. The second set of
instructions directed the participants to, “With this sport in mind, please indicate for each
question below whether you experienced any concerns (i.e., that you experienced this situation
and were to some extent troubled by it) by circling the appropriate number for each question. If a
situation does not apply for you, please circle ‘1 – Not at all’”. The last set of directions
instructed the participants to rate how much of an impact that the concerns that they rated higher
(e.g., 3, 4, or 5) had on any decreased participation they experienced with this sport. If they had
not decreased their sport involvement, they were instructed to circle “1 – No impact at all.”
Gender appropriateness of sport. Participants were asked to rate 22 different sports on
their gender appropriateness, as measured by how masculine, neutral, or feminine they perceive
each sport. Participants were provided with the directions of “Please rate how masculine, neutral,
or feminine you believe each of these sports are.” A 9-point scale was used, ranging from 1
(Very Masculine) to 9 (Very Feminine), with a 5 representing sports viewed as Neutral – Equally
Masculine/Feminine. This questionnaire was adapted from Matteo (1986) in her study that had

!

59!

participants’ rate 68 sports on how stereotypically feminine, masculine, or neutral these sports
are perceived in society. For the purposes of this study, two changes were made to this
instrument. The investigator changed the wording of the question so as to inquire about the
participants’ own perceptions of each sport, not how they think the sports are viewed in society.
The investigator also decided to only use 22 sports instead of 68, with the sports chosen
reflecting the sports participated in by middle school and high school aged youth in Michigan.
There were some small differences in that the investigator considered track and field as one sport
and cross country as another sport, whereas Matteo (1986) had numerous categories for these
sports; jogging, running, hurdles, discus, javelin, pole vaulting, and high jump. Matteo (1986)
also had numerous categorizations for dance (e.g., modern, ballet, jazz), while the investigator
decided to not further subdivide dance. Consistent with Matteo’s (1986) groupings, sports that
have mean ratings between 1 and 3.5 are considered as masculine, between 3.6 and 6.5 as neutral,
and 6.6 and 9 as feminine.
Sport participation type. A variable was created that assessed what type(s) of sport(s) a
participant was involved with in middle school (masculine, feminine, neutral, or combination of
one or more sports from two or more categories).
Psychosocial concerns sport participation type. Another variable was created that
assessed the type of sport a participant listed (masculine, feminine, or neutral) for their
psychosocial concerns survey.
Total middle school sport participation. Participants checked off from a list of 12
sports, any sport that they participated in for at least one season anytime in grades 6, 7, or 8.
Participants also rated their level of involvement on any sports that they played in middle school,
between 1 (Not Very Involved) to 5 (Very Involved), as has been done in previous research (e.g.,

!

60!

Boyer, 2007; Richman & Shaffer, 2000). A score was attained for each sport a participant played.
This was done by multiplying the number of years participated in sport in middle school (1, 2, or
3) by one’s level of involvement (1, 2, 3, 4, or 5). The resulting participation scores (between 1
to 15) from each sport were then summed in SPSS to produce an overall middle school sport
participation score (Total Middle School Sport Participation).
Number of sports participated in middle school. The total number of sports that
participants listed participating in for at least one season anytime in grades 6, 7, or 8.
Middle school sport involvement. Based on previous research with sport participation
(Boyer, 2007; Parsons & Betz, 2001), participants were categorized as low involved if they
participated in one sport or high involved if they participated in more than one sport.
Current sport participation. Participants were considered current athletes if they
answered “yes” to the question “in the past school year (including this past summer), have you
played an organized sport for at least 1 season?” and were considered non-athletes if they
answered “no”.
Number of sports currently participating in. Participants listed what sports they had
participated in/were currently participating in, in the current year (previous summer included).
Demographic information. Year in school and race/ethnicity were two demographic
pieces of information that were asked on the questionnaire.

!

61!

CHAPTER 4
RESULTS
Treatment of the Data
Data were entered into SPSS v. 20 and descriptive statistics were run to find and correct
any data entry errors. A factor analysis was performed on the 23 items of the Psychosocial
Concerns (PSC) scale and five subscales emerged (see page 65). Other variables of interest were
calculated, including: Total – Gender Role Conflict (GRC), Impact – Gender Role Conflict,
Total Middle School Sport Participation, Middle School Sport Involvement, Number of Sports
Participated in Middle School, Number of Sports Currently Participating in,
Masculine/Neutral/Feminine Classification for 22 Sports by Race, Sport Participation Type
(Participation in Masculine/Neutral/Feminine/Combination Sports in Middle School), and
Psychosocial Concerns Sport Participation Type (Masculine/Neutral/Feminine Rating for Sport
listed for Psychosocial Concerns survey). Descriptive statistics were run on the dependent
variables of interest to attain the means, standard deviations, internal reliability, and distribution
statistics.
Table 3
Descriptive Statistics for Dependent Variables of Interest (n =236)
Measure
M
SD Internal
Skewness Skewness
Reliability
SE
Total – GRC
2.12 .87 .883
.87
.159
Impact – GRC
1.70 .96 N/A
1.31
.159
Impact – PSC
1.82 1.06 N/A
1.22
.159
PSC 1, Impact of
1.60 .70 .694
1.89
.159
Others
PSC 2, Negative
1.27 .67 .835
3.16
.159
Stereotypes
PSC 3, Impact of
1.39 .59 .677
2.46
.159
Uniforms

!

62!

Kurtosis Kurtosis
SE
.23
.317
1.03
.317
.72
.317
3.94
.317
10.30

.317

7.33

.317

Table 3 (cont’d)
Measure
PSC 4, Image
Concerns
PSC 5, Physicality
of Sport

M

SD Internal
Reliability
1.43 .74 .750

Skewness Skewness
SE
2.57
.159

Kurtosis Kurtosis
SE
7.13
.317

1.63 .63 .592

1.57

3.74

.159

.317

All dependent variables listed in Table 3 were positively skewed and non-normally
distributed. Transforming these data (e.g., by taking the natural log) would still not allow for an
accurate representation of the data given the heavy response rate in the 1 (Not at All) or 2 (To a
Small Extent) categories. After examining the response rates for each item within Gender Role
Conflict and Psychosocial Concerns (See Appendix D and E, respectively) the investigator
collapsed Gender Role Conflict variables into four response options and Psychosocial Concerns
into three response options. Collapsing variables decreased the number of outliers, helped
increase power, and aided in making the results more reliable. Therefore, all of the Gender Role
Conflict variables have a new range of 1 (Not at All) to 4 (To a Great or Very Great Extent) and
the Psychosocial Concerns variables have a new range of 1 (Not at All) to 3 (To a Moderate,
Great, or Very Great Extent).
Table 4
Descriptive Statistics for Collapsed Dependent Variables of Interest (n =236)
Measure
M
SD Internal
Skewness Skewness
Kurtosis
Reliability
SE
Total – GRC
2.05 .77 .88
.58
.159
-.52
Impact – GRC
1.69 .92 N/A
1.12
.159
.12
Impact – PSC
1.71 .82 N/A
.59
.159
-1.26
PSC 1, Impact of
1.50 .50 .618
1.12
.159
.82
Others
PSC 2, Negative
1.20 .45 .809
2.54
.159
6.22
Stereotypes
PSC 3, Impact of
1.37 .49 .610
1.30
.159
.85
Uniforms

!

63!

Kurtosis
SE
.317
.317
.317
.317
.317
.317

Table 4 (cont’d)
Measure
PSC 4, Image
Concerns
PSC 5, Physicality
of Sport

M

SD Internal
Reliability
1.35 .51 .669

Skewness Skewness
SE
1.68
.159

Kurtosis Kurtosis
SE
2.23
.317

1.54 .47 .549

.79

.04

.159

.317

Given the ordinal nature and non-normality of the dependent variables outlined above,
non-parametric statistical tests were employed. Mann-Whitney U Tests, Wilcoxon Signed Rank
Tests, and Spearman’s Rho were used for analyses with these dependent variables because nonnormal data does not influence the accuracy of the tests. As Mann-Whitney U Tests and
Wilcoxon Signed Rank Tests compare median scores (by converting the scores to ranks), median
scores and mean ranks, rather than mean scores, are presented. Tests of significance are
performed on the mean ranks, not on median scores. This is important to point out, as in some
analyses where differences were significant between two groups; the groups actually have the
same median scores. For all statistical tests, significance was set at p ≤ .05, except where noted.
Psychosocial concerns. Due to the previous knowledge that the 23 items comprising the
Psychosocial Concerns scale coalesced onto a smaller number of subscales, a factor analysis
(PCA, direct oblimin rotation with Kaiser Normalization) was performed on the original 23 items
(See Table 5). Eight factors with eigenvalues over one emerged. However, some of these factors
only contained two items that had values factor loadings over .40, and thus could not be
considered a true factor. When the factor analysis was run as a five-factor model, five clean
factors emerged and were labeled: PSC 1. Impact of Others, PSC 2. Negative Stereotypes, PSC 3.
Impact of Uniforms, PSC 4. Image Concerns, and PSC 5. Physicality of Sport. Of the original 23
items, 19 loaded on these five factors and these five factors explained 55% of the variance.

!

64!

Cronbach’s alpha was performed to check for internal consistency on each of the five factors,
with values ranging from .549 to .809 (See Table 4). Alpha values less than .6 are typically
considered unacceptable (Hair, Black, Babin, Anderson, & Tatham, 2006). Because of the small
number of items in the factors and the nature of the psychosocial constructs they are assessing,
these lower alpha values were not a surprise. In considering this, although PSC 5, Physicality of
Sport had a Cronbach’s coeffecient of .549, it will be included and measured as a subscale.
One item called “messing up your makeup” cross loaded evenly at .486 on Impact of
Others and Image Concerns and therefore was discarded. The item “that your friends teased you
for playing this sport” loaded on PSC 3, Impact of Uniforms. When Cronbach’s alpha was
performed on PSC 3, Impact of Uniforms, the removal of this item increased the alpha from .584
to .610. Therefore, “that your friends teased you for playing this sport” was removed from the
scale leaving the three items for this factor. Three items did not load on any factor: “uniform was
too baggy”, “friends didn’t understand why you played this sport”, and “getting sweaty”.
Therefore, these three items, along with “that your friends teased you for playing this sport” and
“messing up your makeup” were not included in any analyses.
Table 5
Factor Analysis Results of Psychosocial Concerns Scale (n = 236)
PSC 1 PSC 2
Psychosocial Concerns Items
PSC 1, Impact of Others
8. That spending time with this sport kept you
away from doing things that other girls at school
do (e.g. go to the mall, hang out with friends)
13. Boys seeing you sweat
19. That boys teased you for playing this sport
23. That spending time with this sport kept you
away from spending time with someone you were
dating
PSC 2, Negative Stereotypes
17. Being thought to be a tomboy

!

65!

PSC 3

0.680
0.490

1

0.404

0.412
0.548
-0.765

PSC 4

PSC 5

Table 5 (cont’d)
Psychosocial Concerns Items
18. Being called a tomboy
20. Being thought to be a lesbian
21. Being called a lesbian
PSC 3, Impact of Uniforms
4. That your friends teased you for playing this
sport
5. That your uniform was too tight
7. That your uniform made you look too muscular
9. That your uniform was too short
PSC 4, Image Concerns
3. Messing up your hair
6. Not being able to wear your jewelry
11. Messing up your nails/breaking a nail
PSC 5, Physicality of Sport
2. Getting injured because of how physical this
sport is
12. That some people in society view this sport as
a really physical sport
15. That to be good at this sport, you have to have
muscles/be strong
16. That it is not considered "cool" for a girl to
play this sport
Eigenvalue
Percent of Variance
Cumulative Percent

PSC 1

PSC 4

PSC 5

2

0.639
0.743
0.495
0.774
0.712
0.774
0.808
0.623
0.649
0.639
0.758

Items that did not load/loaded at same value
1. Getting sweaty
10. That your friends did not understand why you
played this sport
14. That your uniform was too baggy

1

0.486

1

PSC 3

-0.806
-0.812
-0.849

5.15
22.40
22.40

22. Messing up your makeup

PSC 2

2

2.52
10.95
33.34

1.90
8.25
41.60

1.77
7.70
49.29

1.33
5.76
55.06

1

0.486

n = values > .4, but not included on respective factor. n = item removed after Cronbach’s
conducted due to increasing Cronbach’s coeffecient above .6.
To address the research questions of interest, the variables for psychosocial concerns
were computed and reported in three ways; (1) A score for each of the five Psychosocial

!

66!

Concerns subscales (between 1 and 3); (2) an Impact – Psychosocial Concerns score that reflects
the impact that Psychosocial Concerns had on their middle school sport participation (between 1
and 3 for the 1-item question); and (3) Number and Percentage of Responses by category for
each of the 23 items and for the overall average of the 23 items.
Gender role conflict. The Cronbach alpha coefficient for the 10-item gender role
conflict scale in this study was .88, suggesting very good internal consistency. All of the 10
items on the scale were significantly related to each other at p = .01, with values ranging
from .27 to .57. After the 10 items were collapsed and new variables created, the Cronbach’s
alpha did not change (.88) and all of the 10 items on the scale were significantly related to each
other at p = .01, with values ranging from .20 to .55. To examine the dimensionality of Role
Conflict – Experienced, a factor analysis (PCA, direct oblimin rotation with Kaiser
Normalization) was run to determine if the internal dimensions (questions 5 and 6) and external
dimensions (questions 1, 3, and 9) (Anthrop & Allison, 1983) emerged. Two factors emerged
with items 1, 2, 3, 4, 5, and 7 loading on one factor and items 8, 9, and 10 loading on another.
These factors were not consistent with factors or dimensions that had emerged from previous
research (e.g., Allison & Anthrop, 1983; Allison & Butler, 1984) and the investigator could not
accurately name these two factors.
Therefore, the variables for gender role conflict were computed and reported in four
ways; (1) a Total – Gender Role Conflict score that reflects the participants’ responses on the 10
item Role Conflict – Experienced (between 1 and 4); (2) an Impact – Gender Role Conflict score
that reflects the impact that Gender Role Conflict had on their middle school sport participation
(between a low of 1 and high of 4) for the 1-item question); (3) Number and Percentage of
Responses by category for each of the 10 items and for the overall average of the 10 items; and

!

67!

(4) Original and collapsed mean scores for Total Gender Role Conflict and Impact – Gender
Role Conflict. These mean scores were provided to allow for general comparisons with previous
research, however, no statistical tests were conducted on these mean scores due to the data being
non-normally distributed.
Demographic Profile
A total of 236 girls (who had confirmed that they had played at least one organized sport
for one season during middle school) completed a questionnaire. The majority (74.6%) of the
participants were freshmen (n = 176), 14.0% were sophomores (n = 33), and 5.5% were juniors
(n = 13) and seniors (n = 13). One hundred and six participants (44.9%) were from low SES
urban schools and 130 participants (55.1%) were from middle SES suburban schools. This
sample was diverse in their racial/ethnic background. Girls who identified as white (n = 103
[43.6%]), African American (n = 63 [26.7%]), white/African American (n = 17 [7.2%]),
Hispanic/Latina (n = 12 [5.1%]), white/Hispanic/Latina (n = 9 [3.8%]), and Asian (n = 6 [2.5%])
comprised the majority of the participants (n = 210 [89.4%]). Demographic information related
to specific research questions are presented in those respective sections.
Evaluation of Research Questions for Purpose 1
The following eight research questions were posed in addressing the first purpose of the
study. This purpose was to examine girls’ experiences with gender role conflict and psychosocial
concerns with their middle school participation and determine if there were any differences
based on the participants’ race or the type of school they attended.
Research question 1a. What is the prevalence and impact of gender role conflict with
adolescent girls’ middle school sport participation?

!

68!

Overall, the amount of gender role conflict respondents experienced with their sport
participation was lower than anticipated. When averaging the responses to the 10 items by
category, exactly two-thirds of participants (66.6%) in this study experienced no or a small
amount of conflict with their sport participation and only 16.1% participants reported
experiencing either great or very great amounts of conflict with their sport participation (See
Table 6).
Table 6
Number and Percentages for the Gender Role Conflict 10 items and Average Gender Role
Conflict by Category with Collapsed Categories (n = 236)
To a Moderate
To a Great Extent and
Not at All (1) To a Small
Extent (2)
Extent (3)
Very Great Extent (4 –
5)
Item #
#
%
#
%
#
%
#
%
#1* Little
120
51.1% 54
23.0%
33
14.0%
28
11.9%
recognition for
skills
#2 Low level of 97
41.1% 63
26.7%
40
16.9%
36
15.3%
skill
#3 Girls lose to
89
37.7% 49
20.8%
41
17.4%
57
24.2%
boys
#4 Express
92
39.0% 56
23.7%
48
20.3%
40
16.9%
emotions
#5 Risk jock
156
66.1% 31
13.1%
30
12.7%
19
8.1%
image
#6 Risk looking 111
47.0% 54
22.9%
39
16.5%
32
13.6%
attractive
#7 Athletes
109
46.2% 53
22.5%
41
17.4%
33
14.0%
aggressive
#8* Date or
87
37.0% 56
23.8%
47
20.0%
45
19.1%
practice
#9 Parent role
74
31.4% 59
25.0%
52
22.0%
51
21.6%
#10 Knowledge 104
44.1% 56
23.7%
39
16.5%
37
15.7%
of sport
Average GRC
103.9 44.1% 53.1 22.5% 41.0 17.4%
37.8
16.1%
per category
n* = 235

!

69!

Consistent with experiencing low gender role conflict, respondents also reported that the
impact of gender role conflict played a small role with any decrease in their middle school sport
participation (See Table 7). Spearman Rho correlations were conducted to look at the
relationship between Total – Gender Role Conflict and Impact – Gender Role Conflict. There
was a moderate correlation between the two variables (r = .415, n = 236, p = .000). This
moderate relationship existed for girls from middle SES suburban schools (r = .471, n = 130, p
= .000), girls from low SES urban schools (r = .351, n = 106, p = .000), and for white girls (r
= .379, n = 103, p = .000). For African American girls, there was a large correlation between
these two variables (r = .545, n = 63, p = .000).
Table 7
Impact – Gender Role Conflict with Collapsed Categories (n = 236)
No Impact
Small Impact
Moderate Impact Great and Very Great
(1)
(2)
(3)
Impact (4 – 5)
Measure
#
%
#
%
#
%
#
%
Impact –
134 56.8% 55 23.3%
33
14.0%
14
5.9%
GRC
Total – Gender Role Conflict and Impact – Gender Role Conflict varied across race and
across school type, as outlined in Table 8. African American girls reported the highest amount of
Total – Gender Role Conflict and Impact – Gender Role Conflict, while girls from low SES
urban schools and white girls had the lowest mean scores for Total – Gender Role Conflict and
Impact – Gender Role Conflict.
Table 8
Original and Collapsed Means and Standard Deviations for Total – Gender Role Conflict and
Impact – Gender Role Conflict
Total –
Impact –
Total – GRC
Impact – GRC
GRC
GRC
(Collapsed)
(Collapsed)
Participants
n
M
SD M
SD
M
SD
M
SD
Overall
236 2.12 .87 1.70 .96
2.05
.77
1.69
.92

!

70!

Table 8 (cont’d)

Participants
Middle SES suburban
African American and
white
Low SES urban African
American and white
White
African American

n
100

Total –
GRC
M
SD
2.16 .88

Impact –
GRC
M
SD
1.70 .90

Total – GRC
(Collapsed)
M
SD
2.11
.80

Impact – GRC
(Collapsed)
M
SD
1.69
.87

66

2.18

.91

1.79

1.05

2.07

.76

1.76

.96

103
63

2.09
2.29

.90
.87

1.66
1.86

.91
1.03

2.02
2.21

.78
.78

1.65
1.82

.88
.94

Overall, low levels of Total – Gender Role Conflict and Impact – Gender Role Conflict
were reported by the adolescent girls in this study. As outlined above, these items were
moderately correlated for all participants, suggesting that if girls had lower scores on Total –
GRC, they would have corresponding low score on Impact – GRC, as anticipated. The only
exception to this was for African American girls (who also had the highest mean Total – GRC
and Impact – GRC scores), where the correlation between Total – GRC and Impact – GRC was
strong. Tests for significant differences on these two measures when looking across race and
school type were performed and are presented in research questions 1b and 1c.
Research question 1b. Do girls differ in their retrospectively reported Total – Gender Role
Conflict and Impact – Gender Role Conflict from their middle school sport participation based
on their race?
A Mann-Whitney U test revealed no significant differences on Total – Gender Role
Conflict between African American girls (Md = 2.10, n = 63) and white girls (Md = 1.90, n =
103), U = 2723.5, z = -1.736, p = .083, r = .13. To see if there were any differences when
accounting for school type, two separate Mann-Whitney tests were run for girls from middle SES
suburban schools and girls from low SES urban schools. A Mann-Whitney U test revealed

!

71!

differences that approached significance on Total – Gender Role Conflict between African
American girls in middle SES suburban schools (Md = 2.50, n = 23) and white girls in middle
SES suburban schools (Md = 1.90, n = 77), U = 673.0, z = -1.743, p = .081, r = .17. A post-hoc
power analysis was conducted with this finding, with power calculated to be .147, which is low
(one explanation for this low power is the small number of participants (i.e., n = 23)). A MannWhitney U test revealed no significant differences on Total – Gender Role Conflict between
African American girls in low SES urban schools (Md = 2.00, n = 40) and white girls in low SES
urban schools (Md = 1.85, n = 26), U = 450.0, z = -.920, p = .357, r = .11. These results are
presented in Table 9.
Table 9
Median and Mean Ranks for Total – Gender Role Conflict by Race and Within School Type
White
African American
Participants
n
Md Mean Rank n Md Mean Rank
Overall
103 1.90 78.44
63 2.10 91.77
Middle SES suburban 77 1.90 47.74
23 2.50 59.74
Low SES urban
26 1.85 30.81
40 2.00 35.25
A Mann-Whitney U test revealed no significant differences on Impact – Gender Role
Conflict between African American girls (Md = 2.00, n = 63) and white girls (Md = 1.00, n =
103), U = 2884.5, z = -1.320, p = .187, r = .10. To determine if there were differences when
accounting for school type, two separate Mann-Whitney U tests were conducted for girls from
middle SES suburban schools and girls from low SES urban schools (See Table 10). A MannWhitney U test revealed significant differences on Impact – Gender Role Conflict between
African American girls in middle SES suburban schools (Md = 2.00, n = 23) and white girls in
middle SES suburban schools (Md = 1.00, n = 77), U = 600.5, z = -2.587, p = .010, r = .26. There
was no significant difference on Impact – Gender Role Conflict between African American girls

!

72!

in low SES urban schools (Md = 1.00, n = 40) and white girls in low SES urban schools (Md =
1.50, n = 26), U = 463.0, z = -.819, p = .413, r = .06.
Table 10
Median and Mean Ranks for Impact – Gender Role Conflict by Race and Within School Type
White
African American
Participants
n
Md Mean Rank n Md Mean Rank
Overall
103 1.00 80.00
63 2.00 89.21
Middle SES suburban 77 1.00 46.80*
23 2.00 62.89*
Low SES urban
26 1.50 35.69
40 1.00 32.08
*p = .01.
Research question 1c. Do girls differ in their retrospectively reported Total – Gender Role
Conflict and Impact – Gender Role Conflict from their middle school sport participation based
on the type of school they attend?
A Mann-Whitney U test revealed no significant differences on Total – Gender Role
Conflict between African American and white girls who attend low SES urban (Md = 2.00, n =
66) and middle SES suburban schools (Md = 2.00, n = 100), U = 3247.5, z = -.173, p = .862, r
= .01. To determine if there were any differences when accounting for race, two separate MannWhitney U tests were conducted for white girls and for African American girls (See Table 11). A
Mann-Whitney U test revealed no significant differences on Total – Gender Role Conflict
between white girls who attend low SES urban (Md = 1.85, n = 26) and white girls who attend
middle SES suburban schools (Md = 1.90, n = 77), U = 962.5, z = -.293, p = .770, r = .03.
African American girls who attend low SES urban schools had lower Total – Gender Role
Conflict median scores (Md = 2.00, n = 40) than African American girls who attend middle SES
suburban schools (Md = 2.50, n = 23), however, this was not a significant difference, U = 379.5,
z = -1.151, p = .250, r = .14.
Table 11

!

73!

Table 11 (cont’d)
Median and Mean Ranks for Total – Gender Role Conflict by School Type and Within Race
Middle SES suburban
Low SES urban
Participants
n
Md Mean Rank n Md Mean Rank
Overall
100 2.00 84.03
66 2.00 82.70
White
77 1.90 52.50
26 1.85 50.52
African American 23 2.50 35.50
40 2.00 29.99
A Mann-Whitney U test revealed no significant differences on Impact – Gender Role
Conflict between African American and white girls who attend low SES urban (Md = 1.00, n =
66) and middle SES suburban schools (Md = 1.00, n = 100), U = 3206.0, z = -.342, p = .733, r
= .03. To determine if there were any differences when accounting for race, two separate MannWhitney U tests were conducted for white girls and for African American girls (See Table 12). A
Mann-Whitney U test revealed no significant differences on Impact – Gender Role Conflict
between white girls who attend low SES urban schools (Md = 1.50, n = 26) and white girls who
attend middle SES suburban schools (Md = 1.00, n = 77), U = 834.5, z = -1.424, p = .154, r = .14.
African American girls who attend low SES urban schools had much lower Impact – Gender
Role Conflict (Md = 1.00, n = 40) than African American girls who attend middle SES suburban
schools (Md = 2.00, n = 23). This difference was significant, U = 332.5, z = -1.959, p = .050, r
= .25.
Table 12
Median and Mean Ranks for Impact – Gender Role Conflict by School Type and Within Race
Middle SES suburban
Low SES urban
Participants
n
Md Mean Rank n Md Mean Rank
Overall
100 1.00 82.56
66 1.00 84.92
White
77 1.00 49.84
26 1.50 58.40
African American 23 2.00 37.54*
40 1.00 28.81*
*p = .05.

!

74!

When considering research question 1b and 1c, significant differences exist between (a)
African American girls who attend middle SES suburban schools in that they have much higher
Impact – GRC scores than African American girls who attend low SES urban schools; and (b)
African American girls who attend middle SES suburban schools in that they have much higher
Impact – GRC scores than their white counterparts. Taken together, African American girls who
attend middle SES suburban schools experienced Impact – Gender Role Conflict at the highest
level when compared to all other groups.
Research question 1d. Do girls differ in their retrospectively reported external and internal
gender role conflict scores from their middle school sport participation based on their race?
Research question 1e. Do girls differ in their retrospectively reported external and internal
gender role conflict scores from their middle school sport participation based on the type of
school they attend?
Analyses to address these questions were not conducted because factors did not emerge
from the factor analysis.
Research question 1f. What is the prevalence and impact of psychosocial concerns with
adolescent girls’ middle school sport participation?
H1: Psychosocial concerns measures will have higher scores in comparison to the Gender
Role Conflict measures.
In comparison to Gender Role Conflict, scores on the Psychosocial Concerns scale were,
on average, lower. When averaging the responses to the 23 items by category, there were a larger
percent of responses (87.2% vs. 66.6% with Gender Role Conflict) that were in the 1 (Not at All)
and 2 (To a Small Extent) categories (See Table 13). Also, only 12.8% of participants on average
reported experiencing Moderate, Great, or Very Great amounts of psychosocial concern with

!

75!

their sport participation in comparison with 16.1% of participants who experienced Great or
Very Great amounts of Gender Role Conflict.
Table 13
Number and Percentages for the Psychosocial Concerns 23 items and Average Psychosocial
Concerns by Category with Collapsed Categories (n = 236)
Not at All (1) To a Small Extent Moderate Extent to a Very Great
(2)
Extent (3-5)
Item #
#
%
#
%
#
%
#1 Getting sweaty
115
48.7% 61
25.8%
60
25.4%
#2 Getting injured
90
38.1% 83
35.2%
63
26.7%
#3 Messing up hair
157
66.5% 39
16.5%
40
16.9%
#4 Friends tease
209
88.6% 12
5.1%
15
6.4%
#5 Uniform too tight
147
62.3% 48
20.3%
41
17.4%
#6 Not wear jewelry
188
79.7% 29
12.3%
19
8.1%
#7 Look muscular
203
86.0% 26
11.0%
7
2.9%
#8 Unable to do other
106
44.9% 75
31.8%
55
23.3%
things
#9 Uniform too short
170
72.0% 37
15.7%
29
12.3%
#10 Friends not
194
82.2% 24
10.2%
18
7.6%
understand
#11 Break nails
191
80.9% 26
11.0%
19
8.0%
#12 Sport too physical 183
77.5% 30
12.7%
23
18.6%
#13 Boys see me sweat 146
61.9% 48
20.3%
42
17.8%
#14 Uniform too baggy 155
65.7% 45
19.1%
36
15.3%
#15 Have muscles to
117
49.6% 56
23.7%
61
25.8%
be good
#16 Not cool to play
206
87.3% 16
6.8%
14
5.9%
sport
#17 Thought to be
202
85.6% 18
7.6%
16
6.7%
tomboy
#18 Called tomboy
211
89.4% 7
3.0%
18
7.7%
#19 Teased for playing 208
88.1% 13
5.5%
15
6.3%
#20 Thought to be
205
86.9% 12
5.1%
19
8.1%
lesbian
#21 Called lesbian
207
87.7% 9
3.8%
20
8.4%
#22 Messing up
178
75.4% 37
15.7%
21
8.9%
makeup
#23 Time away from
162
68.6% 33
14.0%
41
17.4%
someone I date
Average PSC per
171.7 72.8% 34.1
14.4%
30.2
12.8%
category

!

76!

As outlined in Table 14 below, just over three quarters of participants (77.0%) reported
that psychosocial concerns had no impact or a small impact (compared with 80.1% with Impact –
Gender Role Conflict) on their decision to discontinue participation and 22.9% of participants
reported that psychosocial concerns had a moderate to very great impact (compared with 19.9%
with the same three categories for Impact – Gender Role Conflict).
Table 14
Impact – Psychosocial Concerns with Collapsed Categories (n = 234)
No Impact (1) Small Impact (2) Moderate to Very Great Impact (3-5)
Measure
#
Impact – PSC 123

%
52.6%

#
57

%
24.4%

#
54

%
22.9%

Spearman Rho correlations were conducted to look at the relationship between the five
Psychosocial Concerns subscales and Impact – Psychosocial Concerns (Table 15). There were
weak to moderate relationships between all subscales with Impact – Psychosocial Concerns, with
all relationships significant at p = .05 and p = .01 level. However, when looking across race and
school type, some PSC subscales did not correlate with Impact – Psychosocial Concerns (See
Table 15). There were no significant correlations between Impact – Psychosocial Concerns and
Image Concerns when looking across race and school type, and Negative Stereotypes had only
one significant correlation (with white girls).
Table 15
Spearman Correlations between Psychosocial Concerns Subscales and Impact – Psychosocial
Concerns
Impact
Negative
Impact of
Image
Physicality
of
Stereotypes
Uniforms
Concerns of Sport
Others
Participants
n
r
r
r
r
r
Overall
234 .455**
.152*
.252**
.182**
.352**

!

77!

Table 15 (cont’d)

Participants
Middle SES
suburban African
American and white
Low SES urban
African American
and white
White
African American
*p = .05. **p = .01.

Negative
Stereotypes

Impact of
Uniforms

Image
Concerns

Physicality
of Sport

n
99

Impact
of
Others
r
.385**

r
.290**

r
.363**

r
.078

r
.412**

65

.519**

.174

.223

.073

.216

.245*
.235

.223*
.436**

-.018
.212

.477**
.042

101 .421**
63 .483**

Collapsed mean scores for each of the five subscales of Psychosocial Concerns were also
quite low, both overall and when looking across race and school type (See Table 16). Physicality
of Sport had the highest mean scores, followed by Impact of Others. This trend was consistent
across race and school type, except with African American girls, who reported higher scores on
Impact of Others than on Physicality of Sport.
Table 16
Collapsed Means and Standard Deviations for Psychosocial Concerns Subscales
Impact of
Negative
Impact of
Image
Others
Stereotypes
Uniforms
Concerns
Participants
n
M
SD M
SD
M
SD M
SD
Overall
236 1.50 .50 1.20
.45
1.37 .49 1.35 .51
Middle SES
100 1.52 .48 1.20
.45
1.37 .48 1.31 .45
suburban African
American and
white
Low SES urban
66 1.56 .55 1.22
.49
1.48 .58 1.38 .61
African American
and white
White
103 1.52 .51 1.18
.41
1.36 .46 1.22 .41
African American
63 1.57 .51 1.24
.55
1.51 .61 1.52 .62

!

78!

Physicality
of Sport
M
SD
1.54
.47
1.57
.49

1.56

.46

1.59
1.53

.52
.42

The collapsed mean score of Impact – Psychosocial Concerns (M = 1.71) was similar to
the collapsed mean score of Impact – Gender Role Conflict (M = 1.69). Girls from low SES
urban schools reported the highest scores on Impact – Psychosocial Concerns (M = 1.77), while
girls from middle SES suburban schools had the lowest scores (M = 1.66) (See Table 17).
Table 17
Original Means and Standard Deviations for Impact – Psychosocial Concerns
Impact – PSC (Collapsed)
Participants
n
M
SD
Overall
234 1.71
.82
Middle SES suburban African American and white 99 1.66
.77
Low SES urban African American and white
65 1.77
.86
White
101 1.71
.82
African American
63 1.68
.80
Summarizing research question 1f, adolescent girls in this study reported low levels of
Impact – Psychosocial Concerns and low levels on the individual subscales of Psychosocial
Concerns. These levels were even lower (Total) or similar (Impact) to the scores on the Gender
Role Conflict variables and therefore this hypothesis that PSC would have higher scores than
GRC was not supported. The five subscales of PSC were weak to moderately correlated with
Impact – PSC for the overall participants (n = 236); however, there were many differences on
these correlations when considering race and school type, with girls who attend low SES urban
schools and girls that are African American only having two of the five subscales correlated with
Impact – PSC. Therefore, one should be aware of this when making any conclusions about the
impact PSC has had on girls when relating it to their scores on a certain subscale. Subscale
scores overall were highest for Physicality of Sport and lowest for Negative Stereotypes. There
appeared to be some differences across race and school type when tested in research questions 1g
and 1h.

!

79!

Research question 1g. Do girls differ in their retrospectively reported psychosocial concerns
scores from their middle school sport participation based on their race?
H1: Both African American girls and white girls will have significantly higher
retrospectively reported scores on the Physicality of Sport subscale compared to their scores on
the Image Concerns subscale.
To address hypothesis 1, a Wilcoxon Signed Rank Test was performed. With both
African American girls and white girls, the Physicality of Sport subscale scores were higher than
Image Concerns subscale scores (See Table 18). However, there was only a significant
difference for white girls (n = 103); their Physicality of Sport subscale score (Md = 1.50) and
their Image Concerns subscale score (Md = 1.00) were significantly different at z = -5.438, p
= .000, with a large effect size of .54. As mentioned, there was no significant difference for
African American girls (n = 63) on their Physicality of Sport subscale (Md = 1.50) and their
Image Concerns subscale (Md = 1.33), z = -.840, p = .401, with a very small effect size of .11.
Table 18
Median and Mean Ranks by Race for Physicality of Sport and Image Concerns
White (n = 103)
African American (n = 63)
Measure
Md Mean Rank Md
Mean Rank
Physicality of Sport 1.50 32.11*
1.50
36.05
Image Concerns
1.00 46.73*
1.33
25.19
*p = .000.
H2: African American girls will have significantly higher Image Concerns scores than
white girls.
For Image Concerns, there was a significant difference between African American girls
(Md = 1.33, n = 63) and white girls (Md = 1.00, n = 103), U = 2285.5, z = -3.533, p = .000, r
= .27, which is approaching a medium effect. To determine if there were differences on Image
Concerns when accounting for school type, a Mann-Whitney U test was conducted for girls from

!

80!

middle SES suburban schools and girls from low SES urban schools. There was a significant
difference between African American girls in middle SES suburban schools (Md = 1.67, n = 23)
and white girls in middle SES suburban schools (Md = 1.00, n = 77), U = 498.5, z = -3.495, p
= .000, r = .35. To see if there were differences when examining Image Concerns with African
American and white girls from low SES urban schools, a Mann-Whitney U test was conducted.
Results approached significance, with African American girls in low SES urban schools (Md =
1.33, n = 40) having slightly higher scores than white girls in low SES urban schools (Md = 1.00,
n = 26), U = 394.0, z = -1.845, p = .065, r = .23. These data are represented in Figure 1 and Table
19 below.

Figure 1. Image Concerns median scores by race and school type.
Note. For interpretation of the reference to color in this and all other figures the reader is
referred to the electronic version of this dissertation.
Table 19
Median and Mean Ranks by Race and Within School Type for Image Concerns
White
African American
!

81!

Table 19 (cont’d)
White
Participants
n
Md
Overall
103 1.00*
Middle SES suburban 77 1.00*
Low SES urban
26 1.00
*p = .000.

African American
Mean Rank n Md
Mean Rank
74.19
63 1.33* 98.72
45.47
23 1.67* 67.33
28.65
40 1.33 36.65

Results from the reviewed literature suggested African American girls have specific
concerns with their hair that seem to differ from other races/ethnicities. To explore these
concerns, a Mann-Whitney U test was conducted to examine the relationship between African
American girls and white girls on the item messing up your hair. Concerns about hair for African
American girls (Md = 1.00, n = 63) and white girls (Md = 1.00, n = 103) were significantly
different, U = 2723.5, z = -2.073, p = .038, r = .16. (See Table 20 for differences on mean ranks).
To determine if the same pattern existed when examined by school type, a Mann-Whitney U test
was conducted to examine the relationship between African American girls in middle SES
suburban schools and white girls in middle SES suburban schools. Concerns about hair for
African American girls who attend middle SES suburban schools (Md = 2.00, n = 23) and white
girls who attend middle SES suburban schools (Md = 1.00, n = 77) were significantly different,
U = 565.0, z = -3.015, p = .003, r = .30, which is a moderate effect size. There was no significant
difference with girls in low SES urban schools.
Table 20
Median and Mean Ranks by Race and Within School Type for Item #3 Messing up Hair
White
African American
Participants
n
Md
Mean Rank n Md
Mean Rank
Overall
103 1.00* 78.44
63 1.00* 91.77
Middle SES suburban 77 1.00** 46.34
23 2.00** 64.43
Low SES urban
26 1.00
31.27
40 1.00
34.95
*p = .038. **p = .003.

!

82!

Research question 1h. Do girls differ in their retrospectively reported psychosocial concerns
scores from their middle school sport participation based on school type?
This question was examined in an exploratory manner to determine if a trend would
emerge with the results in comparing girls from middle SES suburban schools and girls from low
SES urban schools.
Exploratory H1: Girls from both low SES urban and middle SES suburban schools will
have significantly higher retrospectively reported scores on the Physicality of Sport
subscale compared to their scores on the Image Concerns subscale.
To address this hypothesis, a Wilcoxon Signed Rank Test was performed (See Table 21).
For African American and white girls from both low SES urban schools and from middle SES
suburban schools, the Physicality of Sport subscale scores were higher than Image Concerns
subscale scores. For African American and white girls from low SES urban schools (n = 66),
Physicality of Sport subscale scores (Md = 1.50) and their Image Concerns subscale scores (Md
= 1.00) were significantly different at z = -2.795, p = .005, with a moderate effect size of .34. For
girls from middle SES suburban schools (n = 100), there was also a significant difference
between their Physicality of Sport subscale (Md = 1.50) and their Image Concerns subscale (Md
= 1.00), z = -4.030, p = .000, with a moderate effect size of .40.
Table 21
Median and Mean Ranks by School Type for Physicality of Sport and Image Concerns
Middle SES suburban African American Low SES urban African American
and white (n = 100)
and white (n = 66)
Measure
Md
Mean Rank
Md
Mean Rank
Physicality of
1.50
39.64**
1.50
31.21*
Sport
Image
1.00
46.43**
1.00
26.90*
Concerns
*p = .005. **p = .000.

!

83!

The exploratory hypothesis in research question 1h was supported; girls from both low
SES urban and middle SES suburban schools had higher scores on Physicality of Sport than on
Image Concerns. However, when the same comparison was made by race in research question 1g,
H1 was just partially supported; only white girls had significantly higher Physicality of Sport
scores than Image Concerns scores. One possible reason for this was, as found in H2, African
American girls have significantly higher Image Concerns scores than do white girls (thereby
supporting H2). When examined by school type, it was found that it is the African American
girls in middle SES suburban schools who have the highest Image Concerns scores, with there
being no differences in low SES urban schools. A similar pattern exists when the specific
question of messing up one’s hair was examined; African American girls were significantly more
concerned about messing up their hair than white girls are and the difference was even more
pronounced for African American girls in middle SES suburban schools compared to white girls
in middle SES suburban schools.
Evaluation of Research Questions for Purpose 2
The following four research questions were posed in addressing the second purpose of
the study. These research questions focused on girls’ current sport participation. Namely, on the
demographic information about current sport participants, what predicts being a current sport
participant, and current sport participants’ experiences with gender role conflict and psychosocial
concerns.
Research question 2a. Are there differences in girls’ middle school and current sport
participation based on race or school type?
The most popular sports (measured by the sports participants cited playing for at least one
season in middle school) were volleyball (40.3%), basketball (34.0%), cheer (28.9%), track and

!

84!

field (27.7%), dance (26.4%), soccer (24.2%), softball (19.5%), swimming (17.4%), cross
country (10.6%), tennis (8.5%), lacrosse (3.8%), and golf (3.0%). Less than 3% of participants
had participated in horseback riding, ice hockey, fencing, gymnastics, karate, baseball, figure
skating, diving, and boxing.
Chi-square tests for independence (with Yates Continuity Correction) were conducted to
examine sport participation across race and school type. Because 10 tests were conducted, the
significance value was set at p < .01. In terms of race, a 2 (white/African American) x 2
(participant/non-participant) Chi-square test revealed that African American girls participated at
2

significantly higher rates in basketball (χ (1, n = 165) = 7.11, p = .008, phi = -.22) than white
girls. In terms of school type, a 2 (low SES urban/ middle SES suburban) x 2 (participant/nonparticipant) Chi-square test revealed that girls from low SES urban schools (n = 66) also
2

participated at significantly higher rates in cheer (χ (1, n = 165) = 6.94, p = .008, phi = -.22)
than did girls from middle SES suburban schools (n = 99). These results are presented in Table
22.!
Table 22
Top 10 Sports Participated In by Race and School Type
Overall (n African
White (n = Low SES urban
= 236)
American (n 103)
African
= 63)
American and
white (n = 66)
Sport
#
%
#
%
#
%
#
%
Volleyball 95
40.3 20
31.7
46
44.7 28
42.4
3* 27.5 30
Basketball 801 34.0 31* 49.2
45.5
28
Cheer
68
28.9 26
41.3
24
23.3 28*
42.4
Track and
field
Dance
Soccer
Softball
!

1

27.7

18

29.0

26

25.2

15

22.7

29

29.3

1

26.4

17

27.0

25

24.3

15

22.7

27

27.0

24.2
19.5

9
9

14.3
14.3

26
24

25.2
23.3

10
12

15.2
18.2

25
21

25.0
21.0

65
62
57
46

2

Middle SES
suburban African
American and
white (n = 100)
#
%
38
38.0
4
29.3
29
4*
22.0
22

85!

Table 22 (cont’d)
Overall (n
= 236)

African
White (n =
American (n 103)
= 63)

Sport
#
%
#
%
Swimming 41 17.4 6
9.5
Cross
25 10.6 4
6.3
country
Tennis
20 8.5
4
6.3
1
2
3
4
n = 235, n = 62, n = 102, n = 99.
*p < .01.

#
20
14

%
19.4
13.6

Low SES urban
African American
and white (n =
66)
#
%
12
18.2
3
4.5

8

7.8

4

6.1

Middle SES
suburban African
American and
white (n = 100)
#
%
14
14.0
15
15.0
8

8.0

H1a: White girls from middle SES suburban schools will have the highest involvement
in middle school sports when compared with African American girls from both school types and
white girls from low SES urban schools.
Because both of these variables were non-normally distributed, the planned factorial
ANOVA could not be performed to test the hypothesis. Data presented below examine the
relationships that exist by race and school type on middle school sport participation. As
illustrated in Table 23, participants (n = 233) were involved on average in 2.54 sports in middle
school (ranging from one sport to 11 sports, SD = 1.50). Total Middle School Sport Participation
score (see page 60 for review of calculation) for participants (n = 233) was 19.72 and ranged
from one (n = 2) to 97 (n = 1). Mann-Whitney U tests revealed no significant differences
between race or school type on either of these variables.
Table 23
Middle School Total Sport Participation Score and Average Number of Sports Participated In
Overall (n = African
White (n =
Low SES urban Middle SES
233)
American (n 102)
African
suburban African
= 62)
American and
American and
white (n = 66)
white (n = 100)
Measure
M
SD
M
SD
M
SD
M
SD
M
SD

!

86!

Table 23 (cont’d)
Overall (n =
233)
Measure
Total
Middle
School
Sport
Part.
Number
of Middle
School
Sports
Part.

M
SD
M
19.72 13.87 17.39

SD
12.28

Low SES urban
African
American and
white (n = 66)
M
SD
M
SD
19.83 11.96 18.02 12.35

2.54

1.40

2.55

1.50

African
American (n
= 62)

2.35

White (n =
102)

1.36

2.47

1.41

Middle SES
suburban African
American and
white (n = 100)
M
SD
1
11.96
19.51

2.46

1.36

To address the hypothesis that white girls from middle SES suburban schools will have
the highest involvement in middle school sports, data were examined by race and school type
together. Table 24 illustrates that white girls from middle SES suburban schools were in fact the
highest in their middle school sport participation and the number of middle school sports
participated in.
Table 24
Middle School Total Sport Participation Score and Average Number of Sports Participated In by
Race and School Type
Overall (n = African
White low
African
White middle
236)
American low SES urban (n American
SES suburban
SES urban (n
= 26)
middle SES
(n = 77)
= 40)
suburban (n =
23)
Measure M
SD
M
SD
M
SD
M
SD
M
SD
Total
19.72 13.87 18.30 12.79 17.58 11.86 15.73* 11.40 20.61* 11.97
Middle
School
Sport
Part.

!

87!

Table 24 (cont’d)
Overall (n =
236)
Measure
M
Number
2.54
of Middle
School
Sports
Part.
*p ≤ .05.

SD
1.50

African
American low
SES urban (n
= 40)
M
SD
2.55 1.45

White low
SES urban (n
= 26)
M
2.35

SD
1.35

African
American middle
SES suburban (n
= 23)
M
SD
1.96*
1.26

White middle
SES suburban
(n = 77)
M
2.61*

SD
1.36

On both of these variables, the largest differences existed between white girls from
middle SES suburban schools and African American girls from middle SES suburban schools
(Illustrated in Figure 2 and 3). These differences were both significant with white girls from
middle SES suburban schools (Md = 19.0, Mean Rank = 52.53, n = 76) having significantly
higher Total Middle School Participation scores than African American girls from middle SES
suburban schools (Md = 12.0, Mean Rank = 39.02, n = 22), U = 605.5, z = -1.964, p = .050, r
= .20 and white girls from middle SES suburban schools (Md = 2.0, Mean Rank = 54.16, n = 77)
having significantly higher Number of Middle School Sports Participated in than African
American girls from middle SES suburban schools (Md = 2.0, Mean Rank = 38.26, n = 23), U =
604.0, z = -2.377, p = .017, r = .24. Mann-Whitney U tests were conducted to examine if there
were significant differences on either measure between white girls from middle SES suburban
schools and white girls from low SES urban schools; none were found. Therefore, this
hypothesis was partially supported in that white girls from middle SES suburban schools were
the most involved in middle school sports; however, they only significantly differed from
African American girls in middle SES suburban schools (the least involved in middle school
sports).

!

88!

Figure 2. Interaction between race and school type for Total Middle School Sport
Participation scores.

Figure 3. Interaction between race and school type for Number of Middle School Sports
Participated in.

!

89!

In examining the participation rates from their middle school years to their current
participation in high school, participation in sports decreased significantly from middle school
(M = 2.54 sports, ranging from 1 sport to 11 sports, SD = 1.50) to high school (M = 1.20 sports,
ranging from 1 sport to 5 sports, SD = 1.06). Interestingly, when looking only at freshman girls
(n = 176), the average number of sports participated in was also low (M = 1.22 sports, SD = 1.07),
suggesting a decrease in sport participation shortly after girls began high school.
H1b: White girls from middle SES suburban schools will have the highest current rates
of participation in sports when compared with African American girls from both school types
and white girls from low SES urban schools.
Because the variables of interest are categorical (Current Sport Participation) or nonnormally distributed (Average Number of Current Sports Participating in), this hypothesis could
not be tested directly. When looking at Current Sport Participation and Average Number of
Current Sports Participating in, there were some differences across race and school type that
arose. Two Chi-square tests for independence (with Yates Continuity Correction) were
conducted to examine the association between current sport participation (yes/no) and race
(white/African American), and between current sport participation (yes/no) and school type
(middle SES suburban/low SES urban). There was no difference between current sport
2

participation and race, χ (1, n = 166) = .411, p = .522, phi = .064. However, there was a
2

significant difference between current sport participation and school type, χ (1, n = 166) = 6.85,
p = .009, phi = .22, with girls from middle SES suburban schools having much higher current
participation rates than girls from low SES urban schools. See Table 25.
Table 25
Current Sport Participation

!

90!

Table 25 (cont’d)
Overall (n =
236)
Measure
Current
Sport
Participant
*p = .000.!

African
American (n
= 63)

White (n =
103)

#
174

#
46

# %
81 78.6%

%
73.7%

%
73.0%

Low SES urban
African
American and
white (n = 66)
#
%
43*
65.2%

Middle SES
suburban African
American and
white (n = 100)
#
%
84*
84.0%

Two separate Mann-Whitney U tests were conducted to determine if there were
significant differences across race or across school type for Average Number of Current Sports
Participating in (See Table 26). There was not a significant difference between African American
(Md = 1.0, Mean Rank = 77.71, n = 63), and white participants (Md = 1.0, Mean Rank = 87.04, n
= 103), U = 2879.5, z = -1.302, p = .193, r = .08. There was a significant difference when
considering school type, U = 2681.5, z = -2.188, p = .029, r = .17, with girls from middle SES
suburban schools (Md = 1.0, Mean Rank = 89.69, n = 100), currently participating in
significantly more sports than girls from low SES urban schools (Md = 1.0, Mean Rank = 74.13,
n = 66).
Table 26
Average Number of Current Sports Participating In
Overall (n
African
White (n
= 236)
American (n = 103)
= 63)
Measure
Average
Number of
Current Sports
Participating in
*p < .05.
!

M
1.20

SD
1.06

M
1.13

SD
1.13

M
1.26

91!

SD
.99

Low SES urban
African
American and
white (n = 66)
M
1.09*

SD
1.21

Middle SES
suburban
African
American and
white (n = 100)
M
SD
1.29*
.91

To address the hypothesis that white girls from middle SES suburban schools will have
the highest current rates of participation in sports, data were examined by race and school type
together (See Table 27). For current sport participation (yes/no), the major indicator for being a
current sport participant seemed to be the type of school one attends (middle SES suburban),
with both African American girls from middle SES suburban schools (87.0% current
participation) and white girls from middle SES suburban schools (83.1%) having similarly high
current sport participation rates. Chi-square tests revealed that there were no significant
differences in the scores between these two groups on Current Sport Participation, and
differences only approached significance between white girls who attend low SES schools and
2

white girls who attend middle SES suburban schools, χ (1, n = 103) = 2.66, p = .056, phi = .188.
Table 27
Current Sport Participation by Race and School Type
Overall (n = African
White low
236)
American low SES urban (n
SES urban (n = = 26)
40)
Measure
Current
Sport
Participant

#
174

%
73.7%

#
26

%
65.0%

#
17

%
65.4%

African
American
middle SES
suburban (n =
23)
#
%
20
87.0%

White middle
SES suburban
(n = 77)
#
64

%
83.1%

In examining the Average Number of Current Sports Participating in, white girls from
middle SES suburban schools participated in the largest number of sports (M = 1.35), with white
girls from low SES urban schools playing less (M = 1.00). Because this variable was nonnormally distributed, a Mann-Whitney U test was performed to examine the influence of school
type for white girls. White girls from low SES schools (n = 26, Mean Rank = 43.90, Md = 1.00)

!

92!

and white girls from middle SES suburban schools (n = 77, Mean Rank = 54.73, Md = 1.00) had
differences that only approached significance, U = 790.5, z = -1.702, p = .089, r = .17. These
results are presented in Table 28 below.
Table 28
Average Number of Current Sports Participating In by Participant Group
Overall (n
African
White low
African
= 236)
American low SES urban
American
SES urban (n
(n = 26)
middle SES
= 40)
suburban (n =
23)
Measure
M
SD
M
SD
M
SD M
SD
Average Number 1.20 1.06 1.15
1.35
1.00
.98 1.09
.60
of Current Sports
Participating in

White
middle SES
suburban (n
= 77)
M
1.35

SD
.98

In sum, H1b was not supported. It seems that school type plays a dominant role in girls’
high school sport participation. While white girls from middle SES suburban schools had high
current rates of participation, they were actually behind those of African American girls in
middle SES suburban schools. White girls from middle SES suburban schools had the highest
number of sports currently participating in, but they did not differ significantly from their middle
SES suburban counterparts.
Research question 2b. What variables serve as the best predictors for girls’ current sport
involvement?
H1: Total - Gender Role Conflict (negative relationship), Psychosocial Concerns Scores
(negative relationship) and Total Middle School Sport Participation will serve as
significant predictors for current sport participation.

!

93!

H2: Race and school type will also predict current sport participation in that being white
and being from a middle SES suburban school will serve as significant predictors for
current sport participation.
The predictor variables (Impact of Others, Negative Stereotypes, Impact of Uniform,
Image Concerns, Physicality of Sport, Impact of PSC, Total GRC, Impact of GRC, School Type,
Race, and Total Middle School Participation) were checked for multicollinearity. There were no
variables significantly correlated > .7, with correlations ranging from .016 to .643 (for Impact –
Psychosocial Concerns and Impact – Gender Role Conflict). A direct logistic regression was
performed to assess the impact of these 11 variables on predicting the likelihood of respondents
being a current sport participant. A backwards Wald method was used which resulted in seven
2

steps. The full model containing all 11 variables at step 1 was statistically significant, χ (11, n =
2

162) = 43.88, p = .000 and at step 7, the model was statistically significant, χ (5, n = 162) =
39.52, p = .000, suggesting that both models were able to distinguish between participants who
are current sport participants and those who are not currently sport participants. The initial model
explained between 23.7% (Cox & Snell R square) and 35.8% (Nagelkerke R square) of the
variance in current sport participation, and the final model explained between 21.6% (Cox &
Snell R square) and 32.6% (Nagelkerke R square) of the variance in current sport participation.
The final model correctly classified 82.7% of cases, in comparison to the base model that
correctly classified 76.5% of cases. In the final model (step 7), only three variables (Total Sport
Participation, GRC Total, and School Type) made unique, statistically significant contributions
(see Table 29). The strongest predictor for being a current sport participant was being from a
middle SES suburban school, with an odds ratio of 2.61, indicating that participants who attend
middle SES suburban schools are 2.61 times more likely to be a current sport participant than

!

94!

girls who attend a low SES urban school. Interestingly, race was the variable that had the
smallest (non-statistically significant) contribution and was the first variable to be dropped from
the model. Therefore, hypothesis two was partially supported. Hypothesis one was also partially
supported. While Impact of Others (p = .060) and Image Concerns (p = .081) approached
significance, no PSC subscales were significant predictors of current participation. GRC Total
was positively related to current sport participation: the opposite of what was hypothesized. Total
Middle School Sport Participation was a significant predictor of being a current sport participant,
as hypothesized.
Table 29
Logistic Regression Predicting Likelihood of Current Sport Participation
At Elimination
Final Model
Odds
Predictor
B
Wald p
B
Wald p
Odds Ratio
Ratio
Total Middle School
--------.09
13.04 .000 1.09
Sport Participation
--------Total – GRC
.88
5.35 .021 2.41
---------!
School Type
.96
5.04 .025 2.61
---!
---!
---!
--PSC Impact of Others
-.87 3.54 .060 .42
--------PSC Image Concerns
-.70 3.04 .081 .50
PSC Negative
.60
1.10 .294 1.82
Stereotypes
Impact – PSC
.25
.80
.372 1.29
Impact – PSC
-.33
.84
.360 .72
PSC Impact of Uniform
.45
.82
.364 1.57
PSC Physicality of Sport -.46
.67
.413 .63
Race
-.005
.00
.992 1.00
Note. Elimination signifies the values at the step before the variable was eliminated from the
model.
The investigator was interested if there were differences between African American girls
and white girls in terms of what are the best predictors for their respective sport participation. To
address this question, two separate logistic regressions were conducted to predict current sport
participation for each group. The predictor variables for each regression were checked for

!

95!

multicollinearity. There were no variables significantly correlated > .7, with correlations in the
African American model ranging from -.001 to .670 (for Impact – Psychosocial Concerns and
Impact – Gender Role Conflict), and correlations in the model with white girls ranging from .006
to .634 (also for Impact – Psychosocial Concerns and Impact – Gender Role Conflict). Then, two
separate logistic regressions were performed.
As there were initially 10 predictors in the model, the investigator chose to eliminate four
predictors for the regression with African American girls, in order to have roughly a 10:1
participant/predictor ratio. To accomplish this, the investigator ran a logistic regression with all
10 predictor variables and then eliminated the four variables that had the lowest significance
values. When this was done, Total – Gender Role Conflict (p = .871), Physicality of Sport (p
= .846), Impact – Psychosocial Concerns (p = .617), and Impact – Gender Role Conflict (p
= .526) were eliminated.
For African American girls, this model resulted in two steps. The full model containing
2

all six variables at step 1 was statistically significant, χ (6, n = 62) = 22.21, p = .001 and at step
2

2, the model was statistically significant, χ (5, n = 62) = 20.07, p = .001, suggesting that both
models were able to distinguish between participants who are current sport participants and those
who are not currently sport participants. The initial model explained between 30.1% (Cox &
Snell R square) and 43.6% (Nagelkerke R square) of the variance in current sport participation,
and the final model explained between 27.7% (Cox & Snell R square) and 40.0% (Nagelkerke R
square) of the variance in African American girls’ current sport participation. This final model
correctly classified 82.3% of cases, in comparison to the base model that correctly classified
72.6% of cases. In the final model (step 2), three variables (Total Sport Participation, Impact of
Others, and School Type) made unique, statistically significant contributions (see Table 30). Of

!

96!

the three variables, the strongest predictor for being a current sport participant was being from a
middle SES suburban school, with an odds ratio of 6.91, indicating that African American girls
who attend middle SES suburban schools are almost seven times more likely to be a current sport
participant than African American girls who attend a low SES urban school.
Table 30
Logistic Regression Predicting Likelihood of Current Sport Participation for African American
Girls (n = 62)
At Elimination
Final Model
Odds
Predictor
B
Wald p
B
Wald p
Odds Ratio
Ratio
Total Middle School
---!
---!
---!
---!
.113
6.65 .010 1.12
Sport Participation
--------School Type
1.93
5.30 .021 6.91
--------PSC Impact of Others
-2.17 4.43 .035 .11
--------PSC Impact of Uniform
1.29
3.51 .061 3.64
PSC Negative
--------1.94
3.50 .062 6.93
Stereotypes
PSC Image Concerns
-.94 1.95
.163 .37
Note. Elimination signifies the values at the step before the variable was eliminated from the
model.
For white girls, this model was run with all 10 variables and resulted in 10 steps. The full
2

model containing all 10 variables at step 1 was statistically significant, χ (10, n = 100) = 25.49,
2

p = .004 and at step 10, the model was statistically significant, χ (1, n = 100) = 19.67, p = .000,
suggesting that both models were able to distinguish between participants who are current sport
participants and those who are not currently sport participants. The initial model explained
between 22.5% (Cox & Snell R square) and 35.0% (Nagelkerke R square) of the variance in
current sport participation, and the final model explained less of the variance, between 17.9%
(Cox & Snell R square) and 27.8% (Nagelkerke R square), in white girls’ current sport
participation. This final model correctly classified 81.0% of cases, in comparison to the base
model that correctly classified 79.0% of cases. Throughout all 10 stages of the model, only one
!

97!

variable, Total Sport Participation, was significant (See Table 31). This was the only variable
remaining in the final model (step 10). This suggests that the only significant predictor for
participating in sports in high school for white girls is their participation in sports in middle
school (i.e., the more middle school sport participation that a white girl has, the more likely she
will be a current sport participant while in high school). Interestingly, this odds ratio for white
girls (1.14) was almost exactly the same for African American girls (1.12).
Table 31
Logistic Regression Predicting Likelihood of Current Sport Participation for White Girls (n =
100)
At Elimination
Final Model
Odds
Predictor
B
Wald p
B
Wald p
Odds Ratio
Ratio
Total Middle School Sport
---!
---!
---!
---!
.13
13.28 .000 1.14
Participation
PSC Image Concerns
-.88
2.01 .156 .42
Total – GRC
.36
.78
.377 1.44
PSC Physicality of Sport
-.80
1.34 .248 .45
School Type
.59
.90
.342 .53
Impact – PSC
-.22
.30
.585 .81
Impact – PSC
.22
.25
.619 1.24
PSC Negative Stereotypes -.06
.01
.944 1.06
PSC Impact of Others
-.09
.01
.913 .91
PSC Impact of Uniform
-.04
.00
.959 .96
Note. Elimination signifies the values at the step before the variable was eliminated from the
model.
In summary, when examining what predicts current sport participation for girls overall in
this study, H1 was partially supported. While some psychosocial concerns subscales approached
significance as predictors, none predicted current sport participation. Interestingly, Total –
Gender Role Conflict predicted the opposite of what was hypothesized (i.e., higher scores
predicted current participation). Race did not serve as a predictor, while type of school that a girl
attends (i.e., middle SES suburban) played the largest role, thus partially supporting H2. When

!

98!

conducting separate analyses by race, it was found that attending a middle SES suburban school
played an even larger role in predicting African American girls’ current sport participation, and
played no role in predicting white girls’ current sport participation. The only significant predictor
for current sport participation for white girls was their Total Middle School Sport Participation,
with this variable serving as a predictor in all three analyses.
Research question 2c. Do current sport participants and girls who have discontinued their sport
participation differ on their retrospectively reported Total – Gender Role Conflict or Impact –
Gender Role Conflict from their middle school sport participation?
H1: Girls who have withdrawn from sport will have significantly higher retrospectively
reported gender role conflict and impact of gender role conflict scores from their middle school
sport participation, compared to girls still participating in sport.
First, Spearman Rho correlations were conducted to ascertain the relationship between
Total – Gender Role Conflict and Impact – Gender Role Conflict for current and non-current
participants. The strongest relationship existed between these two variables for non-current
participants (girls who reported not having played a sport in the last year), than with girls who
were still playing a sport. There was a moderate correlation between the two variables, r = .386,
n = 174, p = .000 for current participants and a strong, positive relationship for non-participants,
r = .559, n = 62, p = .000. The correlation between these two variables with the total population
was r = .415, n = 236, p = .000.
The mean scores for Total – Gender Role Conflict were higher with current participants
than with non-current sport participants. The mean scores of Impact – Gender Role Conflict were
slightly higher for non-current sport participants. These data are presented in Table 32 and tests
for significant differences on these two measures were performed and are presented in Table 33.

!

99!

Table 32
Original and Collapsed Means and Standard Deviations for Total – Gender Role Conflict and
Impact – Gender Role Conflict by Participant Status
Total –
Impact –
Total – GRC
Impact – GRC
GRC
GRC
(Collapsed)
(Collapsed)
Participant
n
M
SD M
SD M
SD
M
SD
Overall
236 2.12 .87 1.70 .96 2.05
.76
1.69
.92
Current Sport
174 2.20 .87 1.68 .96 2.13
.77
1.67
.91
Participants
Non – Current Sport 62 1.92 .85 1.76 .97 1.85
.72
1.76
.97
Participants
For the Total – Gender Role Conflict score, there was a significant difference that arose
between current (Md = 2.00, n = 174) and non-current sport participants (Md = 1.80, n = 62), U =
4213.0, z = -2.562, p = .010, r = .17. This result was opposite of what was hypothesized (See
Table 33).
Table 33
Median and Mean Ranks for Total – Gender Role Conflict by Participant Status
Current Athlete
Non-Current Athlete
Participant n
Md Mean Rank n Md Mean Rank
Overall
174 2.00 125.29*
62 1.80 99.45*
*p = .01.
When looking at the impact of Gender Role Conflict, there was no significant difference
between current athletes (Md = 1.00, n = 174) and non-current athletes (Md = 1.00, n = 62), U =
5126.0, z = -.649, p = .517, r = .04 (See Table 34).
Table 34
Median and Mean Ranks for Impact – Gender Role Conflict by Participant Status
Current Athlete
Non-Current Athlete
Participant n
Md Mean Rank n Md Mean Rank
Overall
174 1.00 116.96
62 1.00 122.82

!

100!

Overall, H1 was not supported. Non-current athletes had a lower mean rank Total –
Gender Role Conflict score than current athletes. For Impact – Gender Role Conflict, noncurrent athletes overall had slightly higher scores than current athletes. However, no significant
differences arose between these two groups.
Research question 2d. Do current sport participants and girls who have discontinued their sport
participation differ on their retrospectively reported psychosocial concerns scores or impact of
psychosocial concerns from their middle school sport participation?
H1: Girls who have withdrawn from sport will have significantly higher retrospectively
reported psychosocial concerns and impact of psychosocial concerns scores from their middle
school sport participation, compared to girls still participating in sport.
To address hypothesis one, five consecutive Mann-Whitney U tests were conducted on
the five subscales of Psychosocial Concerns. There were no significant differences on four
subscales, with only one subscale (Image Concerns) having significant differences between
current participants and non-current participants (See Table 35). Girls who have stopped their
participation (Md = 1.33, n = 62) had significantly higher Image Concerns scores than girls who
were current sport participants (Md = 1.00, n = 174), U = 4346.0, z = -2.480, p = .013, r = .16.
Table 35
Median and Mean Ranks for Image Concerns by Participant Status
Current Athlete
Non-Current Athlete
Participant n
Md Mean Rank n Md Mean Rank
Overall
174 1.00 112.48*
62 1.33 135.40*
*p = .013.
To examine if there were differences between current and non-current sport participants
on Impact – Psychosocial Concerns, a Mann Whitney U test was conducted (See Table 36). Girls
who have stopped their participation (Md = 2.00, n = 61) had higher Impact – Psychosocial

!

101!

Concerns scores than girls were current sport participants (Md = 1.00, n = 173); however, this
difference was not significant, U = 4651.0, z = -1.512, p = .131, r = .10.
Table 36
Median and Mean Ranks for Impact – Psychosocial Concerns by Participant Status
Current Athlete
Non-Current Athlete
Participant n
Md Mean Rank n Md Mean Rank
Overall
173 1.00 113.88
61 2.00 127.75
In summary, non-current athletes had a significantly higher mean rank score on Image
Concerns than current athletes. There were no significant differences found between non-current
and current sport participants on any of the other four subscales. There also were no differences
when examining Impact – Psychosocial Concerns scores between current and non-current sport
participants. Therefore, there was only isolated partial support for this hypothesis.
Evaluation of Research Questions for Purpose 3
The following four research questions were posed in addressing the third purpose of the
study. This purpose addressed girls’ views on how masculine, feminine, or neutral they viewed
certain sports and if there were differences based on one’s race or school type. Analyses based
on girls’ participation in masculine or feminine sports were then conducted, examining girls’
psychosocial concerns and gender role conflict from their middle school sport participation.
Research question 3a. Do girls’ perceptions’ of what sports they view as masculine, feminine,
and neutral differ from those reported in past research (e.g., Matteo, 1986)?
Using the masculine (1 to 3.5), neutral (3.6 to 6.5), and feminine (6.6 to 9) categories
that Matteo (1986) implemented, participants (n = 236) in this study rated four sports as
masculine (football, wrestling, baseball, and ice hockey), 12 sports as neutral (golf, field hockey,
basketball, rowing, lacrosse, water polo, soccer, track and field, cross country, diving, swimming,

!

102!

and tennis), and six sports as feminine (gymnastics, figure skating, dance, volleyball, competitive
cheer, and softball). In comparison, Matteo’s participants (n = 40 females and 40 males), rated
more (eight) sports as masculine (football, wrestling, baseball, ice hockey, basketball, rowing,
lacrosse, and soccer), less (eight) sports as neutral (golf, water polo, cross country, diving,
swimming, tennis, volleyball, and softball) and less (five) sports as feminine (field hockey,
gymnastics, figure skating, dance, and cheer).
Seven sports changed ratings between Matteo’s (1986) study and this one. Four sports
(basketball, rowing, lacrosse, and soccer) are now viewed as neutral that were previously viewed
as masculine. Two sports, volleyball and softball, are now viewed as more feminine, while field
hockey is now viewed as more neutral and less feminine. It is important to note that participants
differed from this study and Matteo’s (1986) study; half of Matteo’s participants were males and
all were college students. See Table 37 for the means and categorizations (depicted by shading)
for all 22 sports.
Table 37
Views on Masculinity and Femininity of Sports Overall (n = 236)
Current Study (n = 236)
Matteo (1986) (n = 80)
Sport
M (SD)
M
Football
2.19 (1.69)
1.27
Wrestling
2.39 (1.63)
1.27
Baseball
2.40 (1.70)
2.27
Ice hockey
3.02 (1.67)
2.37
Golf
4.10 (1.55)
3.97
2
4.19 (2.31)
Field hockey
6.72
2

Basketball

4.50 (1.16)

Rowing

4.63 (1.19)

Lacrosse

4.70 (1.36)

Water polo
Soccer
Track and field

4.77 (1.13)
4.87 (.85)
4.96 (.79)
2
4.97 (.91)

Cross country

!

2.70

2

2.30

2

2.22

3

3.82
3.00
N/A
4.62 (Running)

103!

Table 37 (cont’d)
Sport
Diving
Swimming
Tennis
Gymnastics

Current Study (n = 236)
M (SD)
2
5.09 (1.14)
5.14 (.93)
5.86 (1.29)
3
6.64 (1.70)
3

6.87 (1.84)
6.97 (1.66)
2
7.35 (1.53)
Volleyball
Competitive cheer
7.44 (1.70)
Softball
7.47 (1.81)
1
2
3
n = 233, n = 234, n = 235.
Figure skating
Dance

Matteo (1986) (n = 80)
M
4.17
5.50
4.87
6.80
7.40
7.37 (modern dance)
6.00
8.65
5.87

Research question 3b. Do girls differ in their perceptions of what sports they view as masculine,
feminine, and neutral based on their race, school type, or the interaction of race and school type?
Twenty-two subsequent 2 (white/African American) x 2 (low SES urban/middle SES
suburban) factorial ANOVA’s were conducted, (with significance set at p < .01 to account for
the large number of tests conducted). Four sports significantly differed across school type (See
Table 38). Girls from low SES urban schools viewed cheer as significantly less feminine than
girls from middle SES suburban schools, and also saw dance as significantly less feminine
(actually rated as a neutral sport) compared with girls from middle SES suburban schools.
Wrestling was viewed as significantly more masculine by girls from middle SES suburban
schools, in comparison to girls from low SES urban schools. Field hockey was viewed as more
masculine by girls from low SES urban schools, while girls from middle SES suburban schools
viewed it as neutral. Field hockey was the only sport that differed across race, with African
American girls viewing field hockey as considerably more masculine than white girls did. There
were no significant interactions across race and school type.

!

104!

White (n = 103) and middle SES suburban (n = 100) participants were similar in that
these two groups both rated four sports as masculine, 12 sports as neutral, and six sports as
feminine. African American (n = 63) and low SES urban (n = 66) participants both rated more
(five) sports as masculine, but differed on their neutral and feminine sport ratings. African
American participants rated 13 sports as neutral, and four sports as feminine while low SES
urban participants rated 14 sports as neutral, and three sports as feminine.
Table 38
Views on Masculinity and Femininity of Sports by Race and School Type (n = 166)
Middle SES
African
suburban African
Low SES urban
White (n =
American (n American and white African American and
Sport
103)
= 63)
(n = 100)
white (n = 66)
Football
2.10 (1.39)
2.35 (2.03)
1.86 (1.44)
2.70 (1.85)
Wrestling
2.15 (1.41)
2.71 (1.91)
1.91 (1.31)*
3.05 (1.84)*
Baseball
2.20 (1.46)
2.48 (1.86)
2.03 (1.37)
2.73 (1.88)
Ice hockey
2.91 (1.53)
2.98 (1.60)
2.70 (1.41)
3.30 (1.68)
Golf
4.32 (1.37)
3.67 (1.72)
4.34 (1.27)
3.67 (1.83)
a1
d1
3.20 (1.72)*
Field hockey 5.06 (2.30)*
3.05 (1.87)*
5.00 (2.46)*
Basketball
Rowing
Lacrosse
Water polo
Soccer
Track and
field
Cross
country
Diving
Swimming
Tennis
Gymnastics
Figure
skating
Dance
Volleyball
Competitive
cheer

!

b2

4.39 (.97)
4.75 (.96)

c1

4.34 (1.18)
4.56 (1.27)
a2
4.58 (1.17)
4.69 (1.24)
4.78 (.79)

4.77 (.84)
4.76 (.86)

4.63 (1.52)
4.83 (.87)
4.76 (.70)

4.72 (1.18)
4.80 (1.13)
4.85 (.63)

4.67 (1.22)
4.47 (1.35)
d2
4.42 (1.48)
4.65 (1.09)
4.64 (.91)

4.98 (.70)

4.89 (.85)

4.96 (.45)

4.92 (1.07)

5.13 (.75)
a2
5.25 (1.16)
5.17 (1.03)
5.85 (1.12)

4.82 (.80)

5.15 (.50)

4.80 (1.06)

5.30 (1.08)
5.05 (.93)
5.87 (.95)
6.94 (1.50)

4.73 (1.24)
5.11 (.81)
5.85 (1.50)
d1
6.317(1.88)

b2

d2
d2

6.88 (1.52)

4.81 (1.15)
4.92 (.52)
5.87 (1.33)
b2
6.23 (1.89)

7.15 (1.81)
7.23 (1.59)
a1
7.61 (1.21)

6.45 (1.95)
6.79 (1.64)
b1
7.21 (1.79)

7.29 (1.59)
7.58 (1.39)*
c1
7.63 (1.27)

6.26 (2.15)
6.39 (1.79)*
d2
7.20 (1.70)

7.84 (1.36)

7.17 (1.89)

8.08 (1.16)*

6.85 (1.89)*

b2

105!

d1

Table 38 (cont’d)
African
Middle SES suburban
American (n African American and
White (n =
Sport
103)
= 63)
white (n = 100)
Softball
7.96 (1.41)
7.11 (2.19)
7.87 (1.48)
a1
a2
b1
b2
c1
d1
d2
n = 102, n = 101; n = 61, n = 62; n = 99; n = 65, n = 64.
*p < .01.

Low SES urban
African American
and white (n = 66)
7.29 (2.14)

In summary, some differences existed between race and school type for views on what
sports are seen as masculine, feminine, or neutral. There were four sports rated differently
between girls in low SES urban and girls in middle SES suburban schools. Notably, competitive
cheer and dance are sports that were seen as more neutral by girls in low SES urban schools and
more feminine by girls in middle SES suburban schools. Field hockey was the only sport that
had differences both between race and school type, with African American girls and girls from
low SES urban schools viewing field hockey as significantly more masculine when compared
with their respective counterparts.
Research question 3c. Do girls differ on their retrospectively reported psychosocial concerns
scores or impact of psychosocial concerns from their middle school sport participation based on
their Psychosocial Concerns Sport Participation Type (masculine, feminine, or neutral rating on
the sport they listed on the psychosocial concerns survey) and based on their current involvement
in this sport (i.e., (a) quit or (b) still participating)?
H1: Girls who listed having participated in a masculine sport will have significantly
higher retrospectively reported scores on the subscales Impact of Others, Negative Stereotypes,
and Physicality of Sport from their middle school participation than girls who participated in
feminine sports. These differences will be even more pronounced when comparing between girls
who have quit and those who are still participating in their sport.

!

106!

H2: Girls who listed having participated in a feminine sport will have significantly higher
retrospectively reported scores on the subscales Impact of Uniforms and Image Concerns than
girls who participated in masculine or neutral sport types. These differences will be even more
pronounced when comparing between girls who have quit and those who are still participating in
their sport.
Using the current study categorizations from the rating system illustrated in Table 38
(above), each sport listed by respondents on their Psychosocial Concerns survey was categorized
as masculine, feminine, or neutral. Using this categorization, only eight participants listed a
masculine sport for their Psychosocial Concerns survey; the overwhelming majority of
participants listed a neutral or feminine sport (See Table 39).
Table 39
Number and Percentages of Participants who listed a Masculine, Neutral, or Feminine Sport on
their Psychosocial Concerns Survey (n = 232)
PSC Sport Participation Type #
%
Masculine
8
3.4%
Neutral
120 50.8%
Feminine
104 44.1%
The original research question 3c was designed to use the gender appropriateness ratings
on the 22 sports that were presented in research question 3b, and then categorize the sport that
participants listed on their PSC survey into one of three categories; masculine, feminine, or
neutral and perform analyses using these categories. Although this categorization allowed for
valuable comparison with data from Matteo (1986) on the gender appropriateness of one’s sport,
it did not allow for the purpose of this research question to be addressed. The investigator has
realized that the optimal way to categorize a sport by masculine/neutral/feminine for this current
research question is not through the gender appropriateness of sport, but more so on how much a

!

107!

sport focuses on femininity. Therefore, this research question is explored using data from
Parsons and Betz (2001), whose study examined the relation of different variables to a woman’s
participation in a sport that focuses on physical appearance and femininity.
Parsons and Betz (2001) had 195 participants (95 males and 100 females) rate 17 sports
on two closely related (correlation of .93 was reported) measures; (a) focus on appearance (how
much emphasis is usually placed on the female athlete’s appearance and body, rated on a 1 to 5
Likert scale with 1 being little focus on appearance and 5 being strong focus on appearance); and
(b) femininity (1 being least feminine and 5 being most feminine). A combined mean (between 1
to 5) for each sport was calculated by averaging the scores of focus on appearance and femininity.
The combined mean scores for the 17 sports ranged from 2.09 (golf) to 4.71 (cheerleading).
The 17 sports in the current study were then categorized as masculine, neutral, and
feminine (using a similar categorization as Matteo, 1986) based on their mean scores (Parsons &
Betz, 2001). Sports that had combined mean scores of 1 to 2.5 were categorized by the
investigator as masculine (golf, lacrosse, crew, field hockey, soccer, and basketball), between
2.51 and 3.5 as neutral (cross country, softball, track, volleyball, diving, swimming, and tennis),
and between 3.51 and 5 as feminine (synchronized swimming, gymnastics, dance, and cheer)
(See Table 40). Lastly, there were three sports that Parsons and Betz (2001) did not include in
their list of sports that were rated and played by the current participants. Ice hockey and baseball
were subsequently included in the masculine sport category and figure skating was included in
the feminine sport category, based on perceptions of respondents in the current study (See Table
37, above). Analyses were then conducted with this new Parsons and Betz (2001) investigatorinterpreted Psychosocial Concerns Sport Participation Type variable (See Table 41).
Table 40

!

108!

Table 40 (cont’d)
Views on Masculinity and Femininity for Current Study and Parsons & Betz (2001) Interpreted
Results
Sport
Current Study (n = 234)* Parsons & Betz (2001) (n = 195)**
Football
2.19 (1.69)
Not rated
Wrestling
2.39 (1.63)
Not rated
Baseball
2.40 (1.70)
Not rated
Ice hockey
3.02 (1.67)
Not rated
Golf
4.10 (1.55)
2.09
Field hockey
4.19 (2.31)
2.25
Basketball
4.50 (1.16)
2.27
Rowing
4.63 (1.19)
2.17
Lacrosse
4.70 (1.36)
2.12
Water polo
4.77 (1.13)
Not rated
Soccer
4.87 (.85)
2.33
Track and field
4.96 (.79)
2.60 (track)
Cross country
4.97 (.91)
2.51
Diving
5.09 (1.14)
2.99
Swimming
5.14 (.93)
3.17
Tennis
5.86 (1.29)
3.29
Gymnastics
6.64 (1.70)
4.02
Figure skating
6.87 (1.84)
Not rated
Dance
6.97 (1.66)
4.59
Volleyball
7.35 (1.53)
2.93
Competitive cheer 7.44 (1.70)
4.71
Softball
7.47 (1.81)
2.58
Note. * Denotes scale from 1 to 9. ** Denotes scale from 1 to 5.
Table 41
Psychosocial Concerns Sport Participation Type Classification using Parsons & Betz (2001)
Overall
Current PSC Participant Non-Current PSC Participant
Sport Participation Type #
%
#
%
#
%
Masculine
74 31.4% 33
33.7%
41
31.3%
Neutral
105 44.5% 42
42.9%
62
47.3%
Feminine
52 22.0% 23
23.5%
28
21.4%
Total
231 100% 98
100%
131
100%
Exploratory H1: Girls who listed having participated in a masculine sport will have
significantly higher retrospectively reported scores on the subscales Impact of Others, Negative
Stereotypes, and Physicality of Sport from their middle school participation than girls who

!

109!

participated in feminine sports. These differences will be even more pronounced when
comparing between girls who have quit and those who are still participating in their sport.
A Mann-Whitney U test was performed on Negative Stereotypes, comparing girls who
had listed a masculine-rated sport and girls who had listed a feminine-rated sport. Girls who had
listed a masculine-rated sport had significantly higher scores on their Negative Stereotypes (Md
= 1.00, Mean Rank = 71.01, n = 74) than girls who had listed a feminine-rated sport (Md = 1.00,
Mean Rank = 52.82, n = 52), U = 1368.5, z = -3.735, p = .000, r = .33, which is a moderate effect
size. When this analysis was conducted only with girls who had quit their sport, there was a
difference approaching significance between the feminine-rated group (Md = 1.00, Mean Rank =
30.88, n = 28) and the masculine-rated group (Md = 1.00, Mean Rank = 37.82, n = 41), U =
458.5, z = -1.910, p = .056, r = .23.
Mann-Whitney U tests were also performed on Impact of Others and Physicality of Sport,
comparing girls who had listed a masculine-rated sport and girls who had listed a feminine-rated
sport on each of these subscales, with no significant differences found. For Impact of Others,
girls who had listed a masculine-rated sport (Md = 1.25, Mean Rank = 60.90, n = 74) did not
differ significantly from girls who had listed a feminine-rated sport (Md = 1.50, Mean Rank =
67.20, n = 52), U = 1731.5, z = -.975, p = .329, r = .09. For Physicality of Sport, girls who had
listed a masculine-rated sport (Md = 1.50, Mean Rank = 63.77, n = 74) did not differ
significantly from girls who had listed a feminine-rated sport (Md = 1.50, Mean Rank = 63.12, n
= 52), U = 1904.0, z = -.101, p = .920, r = .01.
Overall, the hypothesis was not supported; girls who listed having played a masculine
sport on their Psychosocial Concerns survey did have significantly greater Negative Stereotypes
scores than girls who played a feminine sport. However, it was hypothesized that this difference

!

110!

would exist when only examining between girls who had quit these respective sports. This
difference only approached significance (p = .056). The other two subscales, Impact of Others
and Physicality of Sport, had no significant differences for girls who had quit and listed having
played masculine-rated sports versus girls who listed having played feminine-rated sports.
Exploratory H2. Girls who listed having participated in a feminine sport will have
significantly higher retrospectively reported scores on the subscales Impact of Uniforms and
Image Concerns than girls who participated in masculine or neutral sport types. These
differences will be even more pronounced when comparing between girls who have quit and
those who are still participating in their sport.
A Mann-Whitney U test was performed on Impact of Uniforms, comparing girls who had
listed a masculine-rated sport and girls who had listed a feminine-rated sport. Girls who had
listed a feminine sport had the same median score (Md = 1.00, Mean Rank = 67.35, n = 52) with
girls that listed a masculine-rated sport (Md = 1.00, Mean Rank = 60.80, n = 74). There was no
significant difference between these two groups, U = 1724.0, z = -1.080, p = .280, r = .10. When
this analysis was conducted only with girls who had quit their sport, the result was similar in that
girls who quit a feminine sport (Md = 1.50, Mean Rank = 39.73, n = 28) did not significantly
differ on their Impact of Uniforms scores with girls who had quit a masculine sport (Md = 1.00,
Mean Rank = 31.77, n = 41), U = 441.5, z = -1.745, p = .081, r = .21.
A Mann-Whitney U test was performed on Image Concerns, comparing girls who had
listed a masculine-rated sport and girls who had listed a feminine-rated sport. Girls who had
listed a feminine-rated sport had significantly higher scores on their Image Concerns (Md = 1.33,
Mean Rank = 72.13, n = 52), than girls who had listed a masculine-rated sport (Md = 1.00, Mean
Rank = 57.43, n = 74) U = 1475.0, z = -2.437, p = .015 r = .22. When this analysis was

!

111!

conducted only with girls who had quit their sport, the result was similar in that girls who quit a
feminine sport (Md = 1.33, Mean Rank = 40.95, n = 28) significantly differed on their Image
Concerns scores with girls who had quit a masculine sport (Md = 1.00, Mean Rank = 30.94, n =
41), U = 407.5, z = -2.278, p = .023, r = .27. The effect size of .27 in this test was slightly larger
than the effect size of .22 that was found when comparing all participants.
Overall, when using the interpretation of Parsons and Betz’ (2001) data, only the scores
on Negative Stereotypes were found to differ significantly between girls who played masculinerated sports (higher scores) and girls who played feminine-rated sports (lower scores), with there
being no significant differences between the other two hypothesized subscales (Impact of Others
and Physicality of Sport). Thus, hypothesis one was not fully supported, as even with Negative
Stereotypes, there were only differences approaching significance between girls who had quit a
masculine sport and those who had quit a feminine sport. Hypothesis two was also just partially
supported. Only Image Concerns scores were significantly different between girls that had
played feminine sports (higher scores) with girls who had played masculine sports (lower). This
difference was also significant with a larger effect size when examining girls who had played a
feminine-rated sport and had quit this sport and girls who had played a masculine-rated sport and
had quit. This hypothesis was only partially supported, as Impact of Uniforms did not differ
significantly between girls who had played feminine-rated sports and girls that had played
masculine-rated sports.
Research question 3d. Looking at girls’ overall sport participation, do girls who participated in
(a) masculine-type sports differ in their Total – Gender Role Conflict scores from those who
participated in (b) feminine-type sports, and are there differences based on middle school sport
level of level of involvement?

!

112!

H1: Girls who were low middle school sport participators and who participated in
masculine-type sports will have the highest retrospectively reported gender role conflict scores
from their middle school participation than all other girls.
To address this question, the variable Sport Participation Type (participation in
Masculine/Neutral/Feminine/Combination sport(s) in middle school) was calculated for each
participant (See Table 42). When this was done, it was found that only one girl had participated
only in a masculine sport and no girls had participated in both a masculine and a neutral sport.
Table 42
Classification of Sport Participation Type using Participant Ratings
Classification
#
%
Masculine only
1
.4%
Neutral only
54 22.9%
Feminine only
57 24.2%
Masculine and Neutral
0
0.0%
Neutral and Feminine
113 47.9%
Masculine and Feminine
4
1.7%
Masculine and Feminine and Neutral 2
.8%
Total
231 100%
Because the investigator was interested in comparing between girls who participated in
more masculine sports and those who participated in more feminine sports, this research question
is explored using the investigator-interpreted ratings from Parsons and Betz (2001) (Table 40).
For the purposes of this research question, the investigator collapsed the seven categories of
sport participation (masculine only, neutral only, feminine only, masculine and neutral, neutral
and feminine, masculine and feminine, and masculine, feminine, and neutral) into three
categories: masculine-type (participation in only masculine and masculine and neutral sports),
feminine-type (participation in only feminine and feminine and neutral sports), and “other”
(neutral only, masculine and feminine, and masculine and feminine and neutral). The

!

113!

investigator was only interested in comparing between the masculine-type and feminine-type and
therefore the “other” category was not used in any analyses. See Table 43 for details on these
categories.
Table 43
Collapsed Categorization using Parsons & Betz (2001) for Masculine-Type, Feminine-Type, and
Combination/Neutral of Sport Participation
High Involvement (> 1 Low Involvement (1
Overall
sport)
sport)
Classification
#
%
#
%
#
%
Masculine-type category
Masculine only
14 5.9% 3
1.8%
10
15.6%
Masculine and Neutral
66 28.0% 66
39.8%
----80 34.3% 69
41.6%
10
15.6%
Feminine-type category
Feminine only
31 13.1% 7
4.2%
24
37.5%
Feminine and Neutral
26 11.0% 26
15.7%
----57 24.5% 33
19.9%
24
37.5%
Combination/neutral category
Neutral only
47 19.9% 15
9.0%
30
46.9%
Masculine and Feminine
9
3.8% 9
5.4%
----Masculine and Feminine and
40 16.9% 40
24.1%
----Neutral
96 41.2% 64
38.5%
30
46.9%
Total

233 100% 166

100%

64

100%

A Mann-Whitney U test showed that no differences existed between the masculine-type
category (Md = 2.00, Mean Rank = 70.45, n = 80) and the feminine-type category (Md = 2.00,
Mean Rank = 66.96, n = 57), U = 2164.0 z = -.507, p = .612, r = .04. This suggests that girls that
play only masculine sports or masculine and neutral sports do not experience any different
amounts of gender role conflict than those who participate in more feminine or feminine and
neutral sports. The second part of this research question could not be analyzed given the small
number of participants (n = 10) in masculine sports who had a low level of involvement (as
defined by having played only one sport in middle school).
!

114!

CHAPTER 5
DISCUSSION
The overarching purpose of this study was to examine the influence of gender role
conflict, psychosocial concerns, and gender appropriateness of sport with adolescent girls’ sport
participation. Overall, it was found that girls do not experience significant amounts of, or impact
with, gender role impact or psychosocial concerns, and also appear to find many sports as more
appropriate for both genders to participate in. There are three possible explanations for these
findings. First, it is possible that today, gender role conflict and psychosocial concerns do not
exist at any significant level with most girls. Although the qualitative literature with adolescent
girls’ sport participation has highlighted numerous psychosocial concerns, results from this study
suggest that girls’ concerns and conflict exist on a limited scale and scope. Two other reasons
that could have influenced the lower than expected results in this study were the measurement of
the constructs and the procedures used with this research. Both of these areas could be improved
upon in future research, with these improvements serving to increase the reliability and validity
of the data. Therefore, these three explanations are explored throughout this discussion.
Overall, this discussion is centered on the three areas that encompass the three specific
purposes of the study. For each area, the following are included; (a) the purpose is restated; (b) a
summary table of the research questions, hypotheses, and findings is presented; and (c) the
results are discussed in reference to the extant literature. Lastly, strengths and limitations,
recommendations for future research, and practical implications are provided.
Area 1: Gender Role Conflict and Psychosocial Concerns
The first purpose of this study was to examine girls’ experiences with gender role conflict
and psychosocial concerns associated with their middle school sport participation and determine

!

115!

if there were differences when considering a girl’s race and type of school that she attends.
Namely, this area focused on whether girls: (a) experienced gender role conflict and/or
psychosocial concerns with their sport participation; (b) differed significantly on the amount and
type of gender role conflict and/or psychosocial concerns they reported having experienced with
their middle school sport participation; and (c) differed significantly on the impact that gender
role conflict and/or psychosocial concerns have had on their past and current sport participation.
Findings related to the various research questions related to this purpose, as well as the tested
hypotheses are summarized in Table 44.
Table 44
Research Questions, Hypotheses, and Findings for Area 1
Research Question
Hypothesis
Findings
RQ 1a. What is the prevalence and impact of gender role conflict with adolescent girls’ middle
school sport participation? Less prevalent and lower impact than previous research
suggested
RQ 1b. Do girls differ in their retrospectively reported Total – Gender Role Conflict (GRC) and
Impact – Gender Role Conflict from their middle school sport participation based on their race?
African American girls had significantly higher Impact – GRC than white girls, but this
only existed in middle SES suburban schools
RQ 1c. Do girls differ in their retrospectively reported Total – Gender Role Conflict and Impact
– Gender Role Conflict from their middle school sport participation based on the type of school
they attend? No differences for Total – GRC. African American girls from middle SES
suburban schools had significantly higher Impact – GRC than African American girls
from low SES urban schools
RQ 1f. What is the prevalence and impact of psychosocial concerns with adolescent girls’ middle
school sport participation?
H1: Psychosocial concerns measures will have higher scores in comparison to the Gender Role
Conflict measures.
Not supported. PSC subscale scores were lower than Total – GRC and Impact – PSC score
was only slightly higher than Impact – GRC. Highest PSC scores overall on Physicality of
Sport and lowest scores on Negative Stereotypes

!

116!

Table 44 (cont’d)
Research Question
Hypothesis
Findings
RQ 1g. Do girls differ in their retrospectively reported psychosocial concerns scores from their
middle school sport participation based on their race?
H1: Both African American girls and white girls will have significantly higher retrospectively
reported scores on the Physicality of Sport subscale compared to their scores on the Image
Concerns subscale.
Only supported for white girls
H2: African American girls will have significantly higher Image Concerns scores than white
girls.
Supported. Significant differences found for African American versus white girls in middle
SES suburban schools. Findings approached significance with girls from low SES urban
schools. Specific examination of hair-related concerns scores exhibited the same findings
RQ 1h. Do girls differ in their retrospectively reported psychosocial concerns scores from their
middle school sport participation based on school type?
Exploratory H1: Girls from both low SES urban and middle SES suburban schools will have
significantly higher retrospectively reported scores on the Physicality of Sport subscale
compared to their scores on the Image Concerns subscale.
Supported
Note. Analysis for RQ 1d and RQ 1e not included, because these research questions could not be
analyzed.
One of the key aspects of this study was to examine middle school girls’ experiences with
gender role conflict because this stage of development is when the feminine role is becoming
more important to girls (Guillet et al., 2000), and therefore, they might experience more conflict
(Fallon & Jome, 2007). Only two studies focusing on gender role conflict had been conducted
with middle-school girls and Arient (2006) demonstrated that girls did experience gender role
conflict. Therefore, it was surprising that, when compared with past research with high school
and college-aged women, few differences were evident in gender role conflict. However, it was
encouraging to see that the amount of gender role conflict females experienced with their sport
participation has slightly decreased from the finding in the previous studies. When comparing the
overall, non-collapsed mean of 2.12 on the Role Conflict – Experienced inventory from all

!

117!

participants in this study against previous research, the current study found the lowest overall
role conflict score; participants from Sage and Loudermilk (1979), Anthrop and Allison (1983),
and Allison and Butler (1984) all had mean scores of 2.38, 2.18, and 2.30, respectively. In
examining the impact that gender role conflict had on girls’ decreasing their sport participation,
mean scores were even lower than the total gender role conflict scores. This suggests that not
only did girls report having low overall gender role conflict, but also that gender role conflict
plays an even smaller role than previously thought on any decreases in their middle school sport
participation.
Because the comparable research (with high school and college–aged women) was from
30 years ago, it could be that all girls and women’s gender role conflict has decreased, and it is
not simply that middle schools girls experience the lowest amounts of conflict. Because Arient
(2006) used a different gender role conflict inventory and because only middle school girls were
the focus of this study, the speculation that all girls and women’s conflict has decreased cannot
be examined. Regardless, it seems that adolescent girls are also able to navigate a female role
and an athletic role without much conflict. Suggestions as to why gender role conflict is lower
than expected in college and high school-aged girls include the notion that perhaps gender role
conflict has outlived its utility (Allison, 1991), that females now place much more importance on
athletics and by doing so, help decrease the experiences of conflict (as hypothesized by Goldberg
& Chandler, 1991), or that it is now much more acceptable for females to be athletes (Anthrop &
Allison, 1983; Sage & Loudermilk, 1979). These suggestions from extant literature and findings
from the current study can also now be thought to apply to the middle school girl population.
More recent research with college and professional female athletes could also inform this study’s
findings of low conflict. Notably, perhaps girls from this study were able to occupy numerous

!

118!

roles (Cox & Thompson, 2000; Krane et al., 2004; Royce et al., 2003), which helped decrease
experiences of conflict, or they have developed their own versions of femininity that are not tied
to traditional notions of femininity (Krane et al., 2004; Ross & Shinew, 2008), and therefore they
can cope and persist with any conflict that is externally brought onto them. Unfortunately, this is
only speculation and further investigation into the aforementioned ideas could be the focus of
future study.
A related purpose of the current study was the development of the Psychosocial Concerns
survey, because it would allow for an investigation of more current concerns emerging from the
literature, as well as frame questions without conflict inherent in the question. Therefore, it was
especially interesting to see that overall, scores on this PSC survey were lower than on the GRC
survey, suggesting that participants experienced few psychosocial concerns when it came to their
involvement in sport. Specifically, looking at the PSC subscales (there was no total PSC score),
Physicality of Sport, which had the highest non-collapsed mean score of 1.63, was still
considerably lower than the Total – Gender Role Conflict non-collapsed mean of 2.12. The
impact that PSC had on girls with the decrease in their sport participation was similar with the
impact with GRC (non-collapsed mean scores of 1.82 vs. 1.70, respectively). Findings from this
inquiry suggest that the psychosocial concerns that have been outlined in the qualitative literature
is limited in its scale and scope, and that many adolescent girls do not encounter significant
amounts of psychosocial concerns with their sport participation. Some girls in this study might
be negotiating a femininity balancing act (Krane, 2001), whereby they are considering their
(feminine) presentation in their sport participation, but this balancing act does not seem to have a
significant impact on their sport participation. Perhaps times have changed in the last 20 years
and adolescent girls are now ascribing a higher priority to their sport participation compared with

!

119!

previous studies (Coakley & White, 1992), and thus, are not experiencing psychosocial concerns
on a large scale. It also could be that female athletes are not internalizing the negative
stereotypes and images that are present in society (Allison & Butler, 1984), as participants in this
study had the least amount of concern with being called, or being thought to be, a tomboy or a
lesbian.
With gender role conflict, there were minimal differences that emerged when examined
by race and by school type. Only African American girls from middle SES suburban schools
reported significantly higher impact of gender role conflict on sport participation but this finding
could not be easily explained. With psychosocial concerns, many of the results across race and
school type were expected. The idea that the physical dimensions of sport participation for
females was more of a concern than image concerns was supported. However, when examining
across race, this difference only existed with white girls, most likely in part due to them having
significantly lower image concerns than African American girls. Previous research has
highlighted the idea of a beauty cost that some African American girls experience with their
sport participation (Boyington et al., 2008). Further, Collins’ (2011) findings from her research
on African American female college athletes’ and their hair also extends to adolescent girls, as
African American adolescent girls in this study reported significantly higher concerns with
messing up their hair through participating in their sport. It is important to note that the
correlation between African American girls’ image concerns and the impact of psychosocial
concerns was low and not significant (See Table 15). This suggests that while African American
girls did have these concerns, they did not seem to play a substantial role with any decreases in
their middle school sport participation. Lastly, when examining if any differences existed with
school type, it was African American girls from middle SES suburban schools who significantly

!

120!

differed from their counterparts, which could not be easily explained. Overall, results in this area
addressed Mabry and colleagues’ (2003) suggestion to examine if there are attitude differences
between girls of different races that could explain racial differences in physical activity. It can be
concluded from this investigation that, while differences in attitudes emerged, they had a
minimal impact on white and African American girls’ level of participation in sport.
Area 2: Current Sport Participation
The second purpose of the study was to examine current sport participation. This was
done first by using current sport participation as a dependent variable against which significant
predictors for girls’ current involvement in sport in high school were examined. Current sport
participation was also used as an independent variable by comparing gender role conflict and
psychosocial concerns in current sport and non-sport participants. Findings related to the various
research questions related to this purpose, as well as the tested hypotheses are summarized in
Table 45.
Table 45
Research Questions, Hypotheses, and Findings for Area 2
Research Question
Hypothesis
Findings
RQ 2a. Are there differences in girls’ middle school and current sport participation based on their
race or school type? African American girls participate significantly more in basketball than
white girls. Girls from low SES urban schools participate significantly more in competitive
cheer than girls from middle SES suburban schools.
H1a: White girls from middle SES suburban schools will have the highest involvement in middle
school sports when compared with African American girls from both school types and white girls
from low SES urban schools. Partially supported. White girls were the most involved in
middle school sports, but only differed significantly from African American girls in middle
SES suburban schools
H1b: White girls from middle SES suburban schools will have the highest current rates of
participation in sports when compared with African American girls from both school types and
white girls from low SES urban schools. Not supported. School type plays a very strong role,
with both white and African American girls from middle SES suburban schools having
similarly high current sport participation rates

!

121!

Table 45 (cont’d)
Research Question
Hypothesis
Findings
RQ 2b. What variables serve as the best predictors for girls’ current sport involvement?
H1: Total - Gender Role Conflict (negative relationship), Psychosocial Concerns Scores
(negative relationship) and Total Middle School Sport Participation will serve as significant
predictors for current sport participation. Limited isolated support. Only Total Middle
School Participation was a significant predictor. In contrast, Total – GRC had an
unexpected positive relationship and PSC subscales did not significantly contribute.
H2: Race and school type will also predict current sport participation in that being white and
being from a middle SES suburban school will serve as significant predictors for current sport
participation. Partially supported. Race did not contribute. School type was a significant
predictor contributing most to the relationship. School type had an interesting interaction
when examined by race, as it was even more pronounced for the model that predicted
African American girls’ participation and was not found to be a significant predictor for
white girls
RQ 2c. Do current sport participants and girls who have discontinued their sport participation
differ on their retrospectively reported Total – Gender Role Conflict or Impact – Gender Role
Conflict from their middle school sport participation?
H1: Girls who have withdrawn from sport will have significantly higher retrospectively reported
gender role conflict and impact of gender role conflict scores from their middle school sport
participation, compared to girls still participating in sport. Not supported. The opposite was
found. Girls who were current sport participants had significantly higher Total – GRC. No
significant differences on Impact – GRC
RQ 2d. Do current sport participants and girls who have discontinued their sport participation
differ on their retrospectively reported psychosocial concerns scores or impact of psychosocial
concerns from their middle school sport participation?
H1: Girls who have withdrawn from sport will have significantly higher retrospectively reported
psychosocial concerns and impact of psychosocial concerns scores from their middle school
sport participation, compared to girls still participating in sport. Limited isolated support.
Only one of the five subscales (Image Concerns) had significant differences as
hypothesized. No differences were found when examining Impact – PSC
Similar to what Grieser and colleagues (2006) found in their study, African American
girls reported participating significantly more in basketball than white girls. There are a number
of possible reasons for this including easier access or lower cost than other sports (Grieser et al.,
2006; Harrison et al., 1999), having in-group status or role models (Bruening et al., 2008;
Harrison et al., 1999), or increased feelings of competency resulting from participation in the

!

122!

sport (Harrison et al., 1999). Examining details about girls’ middle school and current sports
participation rates provided a more in-depth picture of the relationship between race and sport
participation.
Noting: (a) the decreases in African American girls sport participation in high school
(Troutman & Dufur, 2007); (b) that white girls in comparison to African American girls were
more likely to be highly involved athletes and less likely to be a non-athlete (Sabo & Veliz,
2008); and (c) the overall high rates of sport participation by girls in suburban areas that Sabo
and Veliz (2008) and others have highlighted, it was somewhat surprising (and encouraging) to
realize that it was not white girls from middle SES suburban schools who had the highest middle
school and current sport participation rates. For sport participation in middle school, it was white
girls from middle SES suburban schools who significantly differed from African American girls
from middle SES suburban schools. The significantly lower middle school sport participation
rates by African American girls in middle SES suburban schools could perhaps relate to African
American girls from these schools having higher impact with their gender role conflict and
higher amounts of image concerns.
When examining current sport participants in high school, it was interesting to observe
the significant difference that existed in rates between girls who attend middle SES suburban
high schools (83% are current participators) and girls that attend low SES urban high schools
(66% are current participators). This finding was even more pronounced than what Sabo and
Veliz (2008) found with urban girls having a 59% participation rate in high school compared
with suburban girls at 69%. Interestingly, race did not seem to play a role in terms of whether or
not a girl was currently participating in sport; both African American girls and white girls from
low SES urban schools had significantly lower current sport participation rates than African

!

123!

American and white girls who attend middle SES suburban schools. This difference was
examined more in-depth through a logistic regression to see what other predictors factored into
girls being current sport participants.
Girls’ involvement in sport at the middle school level was the only significant predictor
that appeared in all three regressions (overall, African American, white). Perhaps the most
interesting finding in the regression analyses was the influence school type had for African
American and for white girls. For African American girls, attending a middle SES suburban high
school had significant influence on their sport participation; they were almost seven times more
likely to be a current sport participant if they attended a middle SES suburban high school versus
a low SES urban high school. For white girls, the type of school they attended did not play a
significant role in whether they would be a current sport participant. Therefore, although there
are similarities in that: (a) the more a girl is involved in sports in middle school, the more likely
she will continue to play in high school; and (b) girls in urban schools are much more likely to be
non-sport participants in high school, there are different variables that factor into this depending
on if a girl is African American or white. Specifically, parents of these girls might have
constraints with time and money for their daughters’ sport participation, sport might not being
supported or valued at home and/or in their community, and there might be a lack of PE classes
in their middle schools – thereby impacting girls’ ability to develop requisite sport skills for
future involvement. Overall, this finding supports Sabo’s (2009) suggestion that girls of color
from low-income urban areas seem to face myriad barriers (i.e., gender, race, SES) that inhibit
their sport involvement. In the current study, this was evident only with girls in the high school
years; as African American girls from middle SES suburban schools were the lowest middle
school sport participators. However, it is critical to point out that the design of this study did not

!

124!

allow middle school participation rate to be measured (due to the fact that only girls who had
participated in a middle school sport completed a survey). Previous research has found that girls
in urban middle schools have drop out rates that are four times higher than those of suburban
middle school girls (Sabo & Veliz, 2008). The investigator believes that a similar trend existed in
the schools where she conducted this research, in that there were also differences in the number
of girls from low SES urban schools and middle SES suburban schools who were sport
participants in middle school. Evidence supporting this can be seen by examining the number of
girls in low SES urban schools (n = 55/83 [66.3%]) and middle SES suburban schools (n = 28/83
[33.7%]) who completed the alternate survey (the survey completed by boys and by girls that did
not play a sport in middle school).
What was consistent in these findings was that, while they did not always differ from
other groups of girls, white girls from middle SES suburban schools were the ones participating
the most in middle school and high school sports. This inequity that emerged with race and
school type is especially concerning given females’ participation in sport in the high school years
is related to numerous benefits, including the development of physical skills and body
satisfaction, as well as attributes like self-reliance and assertiveness (Boyer, 2007) – skills that
all girls should have an equal chance of developing.
One of the other hypotheses from this purpose area was that girls who were current sport
participators would have significantly lower gender role conflict scores than girls who had
played a sport in middle school and had since quit in high school. Previous researchers (e.g.,
Anthrop & Allison, 1983) have suggested that girls who experienced the highest amounts of
conflict with sport have since dropped out of sport, and are therefore not captured in these
studies, leading to an incomplete portrayal of gender role conflict with athletes. Thus, the design

!

125!

of the current study allowed for this question to be examined. Given that previous research
(Miller & Heinrich, 2001) found that middle school non-athletes had higher Role Conflict –
Perceived scores (n = 65, M = 28.88) than middle school athletes (n = 44, M = 24.68), it was
surprising that the sport participants in the current study had significantly higher (but still slight)
gender role conflict scores (M = 2.20) than those who have ceased their sport participation (M =
1.92).
The hypothesis put forth by the investigator was grounded in the idea that the amount of
gender role conflict a girl experienced with her middle school sport participation would have an
influence on her being a current athlete, specifically, that girls who had dropped out of sport
would have experienced more conflict which would play a role in their dropping out. However,
girls that had dropped out of sport had low amounts of conflict and low scores on the impact of
gender role conflict. One way to account for these findings is to consider that girls who have
dropped out of sports had lower conflict scores as they had already foreclosed on the role of
being an athlete and therefore ascribed little priority to it (thus experiencing little conflict).
Conversely, the athletes have assigned a much larger role to being an athlete and therefore
experienced more conflict with negotiating their feminine and athletic roles. These female
athletes were not negatively impacted by their experiences with conflict (as reflected by their low
impact scores), and seemed to demonstrate resiliency with their continued sport participation,
similar to the female athletes in previous studies (Fallon & Jome, 2007; Ross & Shinew, 2008).
The fact that female athletes have been found to have higher ratings on athletic competence and
body-image self-concepts than their non-athlete counterparts (Miller & Levy, 1996) also
provides support for this explanation.

!

126!

The investigator also hypothesized that non-current sport participants would have
significantly higher psychosocial concerns scores. Non-current sport participants did have higher
scores on all five of the PSC subscales; however, scores only differed significantly from current
athletes on their image concerns. Non-current sport participants also had higher scores on the
impact that psychosocial concerns had on their sport participation, but these scores were not
significantly higher than their current athlete counterparts. These findings, that non-current
athletes had higher scores on their image concerns with their middle school sport participation,
could suggest that the more image concerns a girl has with their middle school sport participation,
the less likely she will be a current sport participant in high school. However, the investigator is
cautious in drawing this conclusion given the low scores on the impact of psychosocial concerns
and when considering the low predictive ability of image concerns for current sport participation
from the regression analyses that were performed.
Overall, when examining the influence gender role conflict and psychosocial concerns
has on current sport participation from the analyses conducted, a few conclusions can be made.
Only two subscales (Impact of Others and Image Concerns) were predictors in the final model
for overall girls current sport participation; however, these predictors only approached
significance (p = .060 and p = .081, respectively) and had small odds ratios. Therefore, in
considering all the analyses with psychosocial concerns and current sport participation, it can be
concluded that girls’ retrospectively reported psychosocial concerns from their middle school
sport participation had little impact on their current involvement in sport in high school. A
different conclusion exists for gender role conflict. Current sport participants had significantly
higher (but still slight) gender role conflict scores and gender role conflict served as a significant

!

127!

predictor for girls’ current high school sport participation (in that higher amounts of conflict are
related to being a current participant).
Area 3: Gender Appropriateness of Sport
The third purpose of this study was to ascertain the gender appropriateness of sports and
examine if there are differences when considering a girl’s race and type of school that she attends.
It was then determined if girls who have participated in what are viewed by their respective
group as masculine versus feminine sports differed in their gender role conflict and psychosocial
concerns scores. Findings related to the various research questions related to this purpose, as well
as the tested hypotheses, are summarized in Table 46.
Table 46
Research Questions, Hypotheses, and Findings for Area 3
Research Question
Hypothesis
Findings
RQ 3a. Do girls’ perceptions’ of what sports they view as masculine, feminine, and neutral differ
from those reported in past research (e.g., Matteo, 1986)? Yes, girls in this study perceived
some sports (basketball, lacrosse, rowing, and soccer) to be more neutral and less
masculine, some (volleyball and softball) to be more feminine and less neutral, and field
hockey to be more neutral and less feminine.
RQ 3b. Do girls differ in their perceptions of what sports they view as masculine, feminine, and
neutral based on their race, school type, or the interaction of race and school type?
Yes, field hockey was viewed as more masculine by African American girls (the only
difference across race). Field hockey and wrestling were rated as more masculine by girls
from low SES urban schools, and cheer and dance were rated as less feminine by those
same participants.

!

128!

Table 46 (cont’d)
Research Question
Hypothesis
Findings
RQ 3c. Do girls differ on their retrospectively reported psychosocial concerns scores or impact
of psychosocial concerns from their middle school sport participation based on their
Psychosocial Concerns Sport Participation Type (masculine, feminine, or neutral rating on the
sport they listed on the psychosocial concerns survey) and based on their current involvement in
this sport (i.e., (a) quit or (b) still participating)?
H1: Girls who listed having participated in a masculine sport will have significantly higher
retrospectively reported Impact of Others, Negative Stereotypes, and Physicality of Sport scores
from their middle school participation than girls who participated in feminine sports. These
differences will be even more pronounced when comparing between girls who have quit and
those who are still participating in their sport. Not supported. While Negative Stereotype
scores were significantly higher for girls that played a masculine sport, differences only
approached significance when further compared with girls who had quit with girls who are
still participating. No significant differences for Impact of Others and Physicality of Sport
were found.
H2: Girls who listed having participated in a feminine sport will have significantly higher
Impact of Uniforms and Image Concerns scores than girls who participated in a masculine
sport. These differences will be even more pronounced when comparing between girls who have
quit and those who are still participating in their sport. Partially supported. Image Concerns
scores were significantly higher for girls that played a feminine sport. This existed for girls
who had quit their sport as well (with effect size slightly larger). No significant differences
on Impact of Uniforms
RQ 3d. Looking at girls’ overall sport participation, do girls who participated in (a) masculinetype sports differ in their Total – Gender Role Conflict scores from those who participated in (b)
feminine-type sports, and are there differences based on middle school sport level of level of
involvement?
H1: Girls who were low middle school sport participators and who participated in masculinetype sports will have the highest retrospectively reported gender role conflict scores from their
middle school participation than all other girls. Not supported. Girls who played in
masculine-type sports and feminine-type sports did not differ in their Total – GRC scores
Seven sports have changed in their gender appropriateness ratings since Matteo (1986)
conducted this research. The four sports of lacrosse, rowing, basketball, and soccer that are now
rated as neutral (vs. masculine by Matteo’s participants) are also all sports that have seen
increases in female sport participation in the last 30 years. Similarly, volleyball and softball are
now seen as more feminine, perhaps because only girls have volleyball and softball teams in high

!

129!

school in Michigan. Lastly, field hockey is now viewed as more neutral and less feminine. The
investigator hypothesizes that this was in part due to some African American girls and some girls
in low SES urban schools not knowing what field hockey is, and therefore assuming it to be very
similar to hockey, thus providing it with a more masculine rating. Therefore, this analysis of field
hockey might not truly reflect their views on this sport. This was not only demonstrated by a few
participants asking about what this sport was, but can also be seen by examining the percentage
of participants who viewed field hockey as 1 (Very Masculine). With African American girls (n
= 63), 33% rated field hockey as 1 (Very Masculine) and 27% of low SES urban girls (n = 106)
rated it as 1 (Very Masculine). This is in comparison to 10% of girls from middle SES suburban
schools (n = 130), and 8.7% of white girls (n = 102) rating field hockey as 1 (Very Masculine).
Other differences across school type included cheerleading and dance being viewed as less
feminine by girls from low SES urban schools. These differences could not easily be explained
but the investigator hypothesized perhaps dance is more popular with boys in low SES urban
schools and hence more appropriate for both genders. Overall, due to the differences in samples
from this study and Matteo’s (1986) study, (adolescent girls versus college-aged men and
women), the conclusion cannot be drawn that certain sports are now viewed in society as more
gender neutral. It is, however, promising that the general trend found in this research question
was that adolescent females are now seeing many sports as appropriate for both genders. This is
a positive finding in light of Guillet and colleagues’ (2000) conclusion that girls’ sustained sport
involvement will suffer due to stereotypically masculine views of sports.
The investigator hypothesized that there would be differences with girls’ psychosocial
concerns scores based on whether they participated in a masculine or feminine sport (as defined
by the participants in this study). Two subscales (Image Concerns and Impact of Uniforms) were

!

130!

hypothesized to have higher scores for girls that had played (and quit) a feminine sport compared
with girls who played (and quit) a masculine sport. Only the hypothesis for image concerns (that
girls who played and quit a feminine sport would have significantly higher scores on their image
concerns subscale than girls who had played a masculine sport and quit) was fully supported.
This suggests that girls who play and quit feminine sports are significantly more concerned about
portraying a stereotypical feminine image (concerns over their hair, nails, and/or makeup) than
girls who play masculine sports. This finding provides support for what girls in Dwyer et al.’s
(2006) study stated, in that many of the girls will stop participation in sport because of their
concern over how they look. The finding from the current study, however, extends the literature
in that image concerns were only found to exist for girls who play and quit feminine sports. One
possible reason could be that a higher percentage of girls who participate in feminine-type sports
are gender-typed females and have “stereotypic images of femininity and gender-defined
behavioral prescriptions” (Guillet et al., 2000, p. 420). Interestingly, this hypothesis was not
supported for the other subscale (Impact of Uniforms). Girls who played and quit a feminine
sport (gymnastics, dance, cheerleading, figure skating) did have higher scores for concerns about
their uniform when compared with girls who played and quit a masculine sport (e.g., basketball,
softball, and soccer); however, the difference in scores only approached significance. Therefore,
it can be tentatively concluded that concerns about one’s uniform being too tight, too short, or
making one look too muscular are universal across sport.
There were three subscales (physicality of sport, impact of others, and negative
stereotypes) that were hypothesized to have higher scores for girls that played and quit a
masculine sport compared with girls who had played and quit a feminine sport. The scores on
physicality of sport and impact of others were very similar for girls who had played and quit a

!

131!

masculine sport and those who had played and quit a feminine sport. Girls who played a
masculine sport had significantly higher scores on the negative stereotypes subscale; however,
this difference only approached significance when examining the stated hypothesis that this
difference would be even greater for girls who had quit their respective sports. This suggests that,
similar to the high school girls in Adams and colleagues’ (2005) study, there is a stigma
associated with participation in masculine-type sports such as soccer, basketball, and softball in
that girls who participate in these sports are thought to be or are concerned about being called
gay. However, this difference in scores decreased when examining girls who had quit a
masculine or a feminine sport, suggesting that concerns about being called or being thought to be
a tomboy or lesbian were similar for girls who quit a masculine sport and girls who quit a
feminine sport. One possible explanation for this is that girls who did not quit their masculine
sport had higher amounts of concern due to having more time to experience these concerns, and
quitting their sport would serve to lessen their concerns.
Lastly, numerous studies (e.g., Anthrop & Allison, 1983; Sage & Loudermilk, 1979)
found that girls who participated in traditionally more masculine or non-socially approved sports
experience higher amounts of gender role conflict. This finding was not replicated in the current
study; girls that played masculine and masculine and neutral sports had very similar gender role
conflict scores as girls that played feminine and feminine and neutral sports. This study provided
evidence that sports such as basketball and soccer (sports that used to be considered more
masculine) are now thought to be more neutral and appropriate for both genders. Therefore,
perhaps because there are now fewer sports that are seen as not appropriate for girls (e.g., hockey,
baseball, wrestling), and with girls participating at much higher numbers in sports such as soccer
and basketball (which are now seen as appropriate for both genders to play), this helps explain

!

132!

the decrease. As Miller and Heinrich (2001) suggested, perhaps the increased coverage of female
sports on TV has played a role, particularly with more masculine-type sports including basketball
(WNBA and the NCAA championships) and soccer (U.S. Women’s National Soccer) team
receiving much greater media coverage in recent years. This result lends support to the idea that
the Role Conflict Inventory is most likely outdated (Allison, 1991) and is measuring a concept
that is no longer a concern for many females, including adolescent girls.
Strengths and Limitations
One specific strength of this study was the design, in that the investigator was able to
compare and contrast past and current participation to fill in gaps of previous research.
Specifically, this design allowed for a comparison between girls that are current sport
participants and girls that have since discontinued their sport participation from middle school,
thus allowing for a better understanding of the experiences of adolescent girls’ sport involvement
(Allison & Butler, 1984). Further strengthening the design was the fact that data were gathered
from low SES urban schools and middle SES suburban schools and from a racially diverse
sample, which provided valuable comparisons examining where differences might exist with
adolescent girls’ sport participation. Additionally, this design addressed concerns from many
researchers (e.g., Cooky, 2009; Lavoi & Wiese-Bjornstal, 2007; Mabry et al., 2003; Richman &
Shaffer, 2000), that only middle class or suburban white girls’ experiences in sport were
consistently described in the literature and that research was needed to examine girls of different
racial and socioeconomic backgrounds’ experiences with sport. This strength also speaks to a
limitation of study.
It was much harder to collect data in urban schools and because of various challenges
encountered, the sample size from the low SES urban schools was not as large as initially hoped.

!

133!

In the low SES urban schools there was, on average, lower classroom attendance and fewer girls
eligible to take the survey because they had not participated in a sport during middle school. The
investigator also struggled to gain access to these schools and did not anticipate the amount of
classrooms (32 classrooms in low SES urban schools vs. 17 classrooms in middle SES suburban
schools) that she would need access to in order to collect a sufficient number of surveys. In these
schools, fewer parental consent forms were returned, and about twice the percent of students in
low SES urban schools returned a consent form withholding consent (thereby receiving the $5.00
incentive but not completing a survey). This was a frustrating aspect of the research, which made
it harder to get an acceptable number of participants from these schools. It also suggests that if
additional research of this nature is conducted investigators should seek (if research is no/low
risk to participants) full human subjects review to have the active parent/guardian consent
process waived so that a larger sample can be obtained.
Perhaps one of the biggest limitations was on how this survey was delivered. Because of
researcher resource limitations, students completed a paper rather than electronic survey. Despite
the survey having been pilot tested and corrected to enhance readability and comprehensibility,
there were still questions that challenged participants. Therefore, students were instructed to alert
the investigator or research assistant when the survey was complete so it could be reviewed for
completion and accuracy (e.g., no items were circled twice). The investigator stressed to the
participants that neither she nor the research assistant would be looking at how they answered.
This checking process could have led to girls answering in a more socially desirable way (i.e.,
that they experienced less concerns or less conflict). Related to this, both the investigator and the
research assistant are white women, which may have influenced how students answered.
Recommendations for Future Research

!

134!

The psychosocial concerns survey needs additional development. There seems to be
consistently four or five dimensions that emerge from a factor analysis. However, in the two
factor analyses that were conducted in the development of the survey for the current study, a few
of the items loaded on different factors, suggesting that works needs to be done on the individual
items. Related, the internal consistency of these factors was in the lower range of acceptability,
highlighting the variability in responses by participants. Therefore, a next step with the
development of this survey would be to conduct focus groups with girls to examine the items and
see what other items could be added or what current items should be adapted to help make this
survey a more reliable measure. The investigator believes it would be beneficial to continue this
research in low SES urban settings, as these are the schools with lowest girls’ sport participation
rates. One specific initiative with future research would be to attain larger samples and to look
within race, as there seemed to be some difference with how girls of different races responded to
questions.
Another suggestion for future research with the psychosocial concerns survey is to
deliver this survey on a tablet versus the paper and pen method that was used in this study. This
format for delivering the PSC survey will help in numerous ways: (a) aide in participants
correctly answering questions with skip logic (i.e., “if yes, go to…if no, go to…”); (b) eliminate
the need for the investigator to check their survey; (c) allow for presentation of information in a
medium that participants of this age typically prefer; and (d) provide for a better way to assess
Impact – PSC, for when this question appears on their screen, the items that the participant
answered either a 3, 4, or 5 on will also appear, making it much easier to complete and will also
increase the validity of this question.

!

135!

Future research that examines girls’ participation in masculine-type sports could benefit
from using Parsons and Betz’ (2001) method of assessing the focus on appearance and the
femininity of different sports. As evidenced in the current study, more and more sports are
becoming appropriate for both genders. However, as found in previous studies (e.g., Hoiness et
al., 2008), girls’ participation in what are seen in society as more masculine-type sports still
seems to be a somewhat contested terrain. Studies that have this latter concept as their aim would
be wise to incorporate this method and to include both genders in their analyses to provide for a
complete picture of this phenomenon.
In line with what Allison (1991) and others have put forth, the investigator believes that
there could be better ways than the Role Conflict Inventory to study how girls navigate their
feminine and athletic roles with their participation in sport. For future research in this area,
implementing some of Allison’s (1991) suggestions by adapting research questions and looking
at role conflict differently could be particularly useful in furthering this area of research. For
example, using focus groups with middle school girls could help generate information on how
girls experience and negotiate any conflict with their sport participation, and could be an initial
step in the development of a new and different type of inventory. This could also allow for
comparisons across ages, as rich information about how high school and college-aged women
negotiate the feminine and athletic roles has been gained in other research using qualitative
methods (e.g., Krane et al., 2004; Ross & Shinew, 2008). Also, it could be particularly
illuminating to conduct a study with African American girls in middle SES suburban schools, as
these were the girls who reported the highest amounts of impact with their gender role conflict.
For any research in this area of gender role conflict, inquiring about girls’ role models
and/or parental involvement in sport could also help further this research, as previous research

!

136!

has shown that parental involvement in sport can serve to mediate or decrease gender role
conflict with females (Arient, 2006; Miller & Levy, 1996) and thus could also be an important
factor to consider with girls’ experiences with both gender role conflict and possibly
psychosocial concerns with sport. Lastly, in line with Allison (1991), turning the focus of the
research to examining the beliefs of those who stereotype (e.g., boys in their schools) could
provide valuable information in understanding girls’ experiences with sport (this last suggestion
was actually carried out by the investigator in this current study, however, these data were not
presented because they were beyond the scope of the research questions in this dissertation).
Involving boys in this type of research can also have benefit to them by making sport a less
stigmatizing and heterosexist environment (Cockburn & Clarke, 2002)
Overall, future research in the area of adolescent girls’ sport participation could also
benefit from inquiry that takes a broad, comprehensive approach that examines a variety of areas
that could impact a girls’ sport involvement. Specifically, the development and use of a
questionnaire that asks about such areas as: (a) the value placed on sport and physical activity by
a girl and her family; (b) the number and gender of her siblings and what their sport involvement
is; (c) how accessible her home is to sport arenas/venues and sporting goods stores; (d) how
physically competent she feels in sports she has played; (e) her experience in prior physical
education classes (number and overall experience in class(es); and (f) if she has female athlete
role models, could help shed further light on constraints girls experience with sport. For any
research that aims to learn more about barriers girls encounter with their sport participation, a
helpful step to include would be to inquire about what aspects girls would like to see that would
serve to increase their sport involvement. This has been done with other research (e.g., Burk &
Shinew, 2013; Oliver, Hamzeh, & McCaughtry, 2009), and could be particularly beneficial in

!

137!

program development and using research in practical and productive ways. This could be an
especially fruitful endeavor in low SES urban schools where girls’ sport participation is
especially lacking.
Practical Implications
It can be concluded that gender role conflict, as measured through the Role Conflict –
Experienced inventory, was not a concern that factored into adolescent girls’ sport participation
in this study. If this finding were upheld in future studies (particularly with studies that had
implemented the suggestions listed above), this would have important implications in that girls in
middle school do not experience any significant amounts of conflict with being an athlete and
being female. This study also provided evidence that suggests middle school girls have relatively
low amounts of psychosocial concerns with their sport participation in middle school. There
were some important differences that emerged, such as African American girls having more
concern about their image in general and hair concerns in particular, with their sport participation.
It is important to note that while these differences were significant between white and African
American girls, the concerns for girls of both races were still relatively low. However, it would
still be beneficial for coaches and other stakeholders that are involved with girls’ sports teams
and programs to know that differences exist and that African American female athletes have
“different challenges when carrying out tasks such as styling, washing, maintaining, and relating
to their peers over hair” (Collins, 2011, p. 78).
Overall, while there are considerations that need to be addressed with this psychosocial
concerns measure, an initial conclusion from this research is that there are other variables that
influence girls’ decreasing their sport participation in their middle school and high school years.
One finding in particular from this study highlights this. Girls that attend low SES urban high

!

138!

schools are particularly negatively impacted with their sport participation, especially African
American girls. Knowing this, and knowing the benefits that are associated with girls’
participation in sport, it could be particularly valuable for low SES urban schools to target
specific efforts and initiatives in an attempt to foster and retain girls as sport participants
throughout their high school career. What is of equal or more importance is getting and keeping
girls involved in their middle school years, as increased involvement in sport during these years
was directly related to girls being involved in sport in high school. These efforts in middle
schools and high schools are of course easier said than done, especially when considering the
budget, time, and priority constraints that many administrators of low SES urban schools are
faced with. This is evidenced by Sabo and Veliz (2012) in the stark difference in the number of
female sports teams in lower economic resource schools (12.3 female teams) than in higher
economic resource schools (19.5 female teams). Unfortunately, this same research also outlined
that schools that have (a) the lowest economic resources; (b) are in urban areas; and (c) have the
highest percentage of students of color are also the schools that are cutting sports at the highest
rates (Sabo & Veliz, 2012). When noting the wealth of studies that highlight the gains that can be
accomplished by females through sport participation, perhaps initial efforts should be focused
on: (a) informing stakeholders of these benefits so that they are at least well informed when
making decisions about girls’ sport offerings; and (b) in line with Sabo and Veliz (2012),
ensuring that these schools are complying with Title IX and accommodating the interests of their
female student body.
!!

!

139!

FOOTNOTE

!

140!

FOOTNOTE

1

The investigator chose to use the terms white and African American as descriptors for race in
this study. The term white (not Caucasian) was used and purposely not capitalized.

!

141!

APPENDICES

!

142!

APPENDIX A
Informed Consent/Assent Form
!

!

Research(Participant(Information:(Parental(Consent/Participant(Assent(Form(
(
An(Examination(of(Influences(on(Girls’(Sport(Participation(and(Withdrawal((

*"Important!"Please"complete"and"have"your"child"return"*"

!
Dear%parent%or%guardian,%%
!

This%form%provides%information%for%you%about%a%research%study%that%your%child%has%the%
opportunity%to%be%a%part%of%in%one%of%their%classes%at%school.%Please%take%a%moment%to%read%this%
form%as%it%describes%this%research%study%that%is%looking%at%views%and/or%experiences%of%girls’%
sport%participation.%
%

Participation%of%your%child%in%this%study%is%not%required.%We%would%love%to%have%your%
son/daughter%participate,%but%we%respect%their%right%to%not%be%involved.%
!

Description(of(the(research(and(your(child’s(participation(
!

Your!child!will!be!asked!to!participate!in!a!research!study!that!is!being!conducted!by!Missy!
Wright!and!Dr!Daniel!Gould!from!the!Kinesiology!department!at!Michigan!State!University.!
Your!child!is!being!asked!to!participate!in!this!study!as!they!are!a!high!school!student,!and!
we!are!interested!in!learning!more!about!girls!sport!participation,!concerns!girls!might!
have!with!their!sport!participation,!and!boys’!views!on!girls’!sport!participation.!From!this!
study,!we!hope!to!gain!some!insights!on!the!type!of!concerns!girls!have!with!participating!
in!sport,!and!views!that!boys!have!about!girls’!sports.!
!

Your!child’s!participation!will!involve!completing!a!survey!in!one!of!their!classrooms!
during!school!time,!sometime!in!the!next!week!or!two.!The!survey!will!take!roughly!10P15!
minutes!to!complete.!
!

Risks(and(discomforts(
!

The!researchers!believe!there!are!little!to!no!risks!for!participating!in!this!research.!Both!
boys!and!girls!will!be!completing!surveys!that!are!examining!girls’!sport!participation!so!
perhaps!girls!might!feel!a!little!uncomfortable!knowing!boys!are!completing!surveys!about!
girls’!sport!participation,!but!we!don’t!believe!this!will!cause!much,!if!any,!discomfort!for!
any!of!the!participants.!
!

Potential(benefits(
!
The!benefits!that!will!be!attained!from!this!research!is!that!the!researchers!hope!to!be!able!
to!learn!from!this!information!and!improve!girl’s!sport!programming!in!the!future.!!
!

!

143!

(
(
(
Protection(of(confidentiality(
!

Your!child!will!not!put!their!name!on!their!survey,!therefore!all!surveys!will!be!anonymous.!
Some!demographic!information!will!be!collected,!but!participants!will!not!be!able!to!be!
identified!from!this!information.!Your!child’s!identity!will!not!be!revealed!in!any!
publication!that!might!result!from!this!research.!The!child’s!teacher!will!not!have!any!
access!to!these!surveys.!Your!child’s!confidentiality!will!be!protected!to!the!maximum!
extent!allowable!by!law.!
!
The!anonymous!surveys!and!consent!forms!will!be!stored!at!Michigan!State!University!for!a!
minimum!of!three!years,!with!only!the!researchers!involved!in!this!study!and!the!
Institutional!Review!Board!at!Michigan!State!University!having!access!to!this!data.!
!
Voluntary(participation(
!

Participation!in!this!research!study!is!totally!voluntary.!Refusal!of!your!child!to!participate!
in!this!study!will!involve!no!penalty!and!will!not!affect!their!grade!in!this!class!or!any!
relation!with!Michigan!State!University!whatsoever.!!Also,!if!you!provide!consent!for!your!
child,!your!child!has!the!right!to!not!provide!their!assent,!and/or!to!discontinue!
participating!at!any!time!or!to!skip/not!answer!any!questions!on!the!survey.!!
!

Compensation(for(being(in(the(study(
(

If!your!child!returns!their!signed!consent!form!on!or!before!the!day!and!time!the!survey!is!
given!in!their!class!(regardless!if!consent!is!provided!or!withheld),!your!child!will!receive!
$5!on!this!day.!!
!

Alternative(Options(
(

If!your!child!does!not!participate!in!this!research,!they!will!have!time!to!work!on!
coursework!or!activities!associated!with!this!class!that!they!are!in.!
!

Contact(Information(
!

If!you!have!any!concerns!or!questions!about!this!research!study,!such!as!scientific!issues,!
how!to!do!any!part!of!it,!or!to!report!an!injury,!you!or!your!child!may!contact!either!!Dr.!
Daniel!Gould!(210!IM!Circle;!drgould@msu.edu)!or!E.!Missy!Wright!(209!IM!Circle,!MSU,!
wrigh233@msu.edu).!
!
If!you!have!any!questions!about!the!role!and!rights!of!your!child!as!a!research!participant,!
would!like!to!obtain!information!or!offer!input,!or!would!like!to!register!a!complaint!about!
this!study,!you!may!contact,!anonymously!if!you!wish,!the!Michigan!State!University's!
Human!Research!Protection!Program!at!517P355P2180,!Fax!517P432P4503,!or!ePmail!
irb@msu.edu!or!regular!mail!regular!mail!at!Olds!Hall,!408!W.!Circle!Dr!#207,!MSU,!East!
Lansing,!MI!48824.!!

!

144!

(
I(I(I(DOCUMENTATION(OF(INFORMED(CONSENT(I(I(I(

!
PLEASE:!
1. Sign!the!form!below!
a. If!you!are!giving(consent,!please!sign!on!the!top!line!
b. If!you!are!not(giving(consent,!please!sign!the!line!on!the!bottom(of(the(page!
2. If!you!are!providing!consent,!your!child!will!also!sign!in!the!space!provided,!
documenting!that!they!voluntarily!assent!to!participate.!
3. Please(keep(the(first(page(and(give(this(last(page(to(your(child(so(they(can(
return(it(right(away(to(their(classroom!(
!
ThankByou%for%your%time%and%energy%in%considering%this%valuable%research%project!%
"
My"signature"below"indicates"I"voluntarily"agree"to"allow"my"child"participate"in"this"
study."
!
_____________________________________________________! !
!
_______________________!
Signature!of!Parent/Guardian! !
!
!
!
Date!
!
!
My"signature"below"indicates"I"voluntarily"agree"to"participate"in"this"study."
!
!
_________________________________________________!
!
!
_______________________!
Signature!of!Assenting!Participating!Child!!
!
!
Date!
%
!
!
!
!
!
!
!
!
!
!
!
P OR"""">"
!
I!have!read!the!consent!form!and!do(not(wish!to!have!my!child!participate.!!
!
!
_____________________________________________________! !
!
_______________________!
Signature!of!Parent/Guardian!
!
!
!
!
Date!

!

145!

APPENDIX B
Questionnaire

Girls’ Experiences with Sport Participation
Directions: Please answer the following questions.
1. To complete this survey, you:
a. Played in 1 or more organized sports (An organized sport is one with organized
practices and scheduled competitions)
b. For at least 1 season
th th
th
c. Sometime when you were in 6 , 7 , or 8 grade
Please place an “X” in this box to state that all of these statements are true (if
not true, please just see Missy to get a different survey)
2. Your current year in school:

FR: ____

SO: ____

JR: ____

SR: ____

3. Please place an “X” on the line(s) that best describe(s) your racial or ethnic background
Asian

_____

Black/African American

_____

White/Caucasian

_____

Native American

_____

Hispanic/Latina (may be any race)

_____

Prefer not to answer

_____

Other (Please specify: _______________________________________)

!

146!

4. There are lots of reasons why boys and girls drop out or decrease their participation in sport.
For the 8 reasons listed below, please circle the # that corresponds to how much these
reasons played a role in any of your decreased sport participation in middle school:

REASON

Didn’t play
a role

Didn’t like the coach
Wasn’t having fun
Needed time for school work
Wasn’t good enough
Wanted time for non-sport activities
Too much focus on winning
Got injured
Other sports took too much time

!

147!

1
1
1
1
1
1
1
1

2
2
2
2
2
2
2
2

Played a
major role

3
3
3
3
3
3
3
3

4
4
4
4
4
4
4
4

5
5
5
5
5
5
5
5

Your Sport Participation
Directions: I’m interested in knowing all the organized sports that you played for at least 1
th th
th
season when you were in 6 , 7 , and 8 grade. An organized sport is one with organized
practices and scheduled competitions.
1. Place an “X” in the box for the years that you played each sport
2. Circle the number for how involved you were in this sport
o If you have played a sport that isn’t listed, please use the space at the bottom to write it in
o For any sports that you played in the summer, please include them in the following
th
school year (*So if you played a sport in the summer after 5 grade, you would place an
th

X in 6 grade)
o For all the sports you haven’t played, just leave these boxes empty

Grade in School
6th

E.G. SOCCER
E.G. TENNIS

7th

X

8th

X

Involvement in Sport
Not very involved

Very involved

Soccer
Volleyball
Basketball
Track and field
Cross country
Softball
Tennis
Swimming
Diving
Golf
Cheerleading
Dance

2
2

3
3

4
4

5
5

1
1
1
1
1
1
1
1
1
1
1
1
1

X

1
1

2
2
2
2
2
2
2
2
2
2
2
2
2

3
3
3
3
3
3
3
3
3
3
3
3
3

4
4
4
4
4
4
4
4
4
4
4
4
4

5
5
5
5
5
5
5
5
5
5
5
5
5

1. In this past school year (including this past summer), have you
played an organized sport for at least 1 season?
_______ YES
________ NO
If yes, what sport(s)?: _______________________________________________

!

148!

Views on Sports
Directions: Please rate how masculine, neutral, or feminine you believe each of these sports are.
Very Masculine

Neutral

Very Feminine

Equally
masculine/feminine

1. Soccer
2. Volleyball
3. Basketball
4. Track and field
5. Cross country
6. Golf
7. Tennis
8. Swimming
9. Diving
10. Softball
11. Baseball
12. Dance
13. Competitive cheer
14. Football
15. Field hockey
16. Gymnastics
17. Lacrosse
18. Ice hockey
19. Wrestling
20. Figure skating
21. Rowing
22. Water polo

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2

3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3

Figure 4. Views on Sports survey used in questionnaire

!

149!

4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6

7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7

8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8

9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9

Concerns for Girls in Sports
This part of the survey asks about concerns girls might have about playing sports. Some girls
have concerns with their sport participation, while other girls don’t have as many or any at all.
So there are no right or wrong answers with this. I am interested in knowing of any concerns that
you might have experienced with a sport that you participated in, in middle school.
When answering the questions for this part of the survey, only answer with 1 sport in mind.
Please read the following instructions for how to choose what sport to use.
1. Is there a sport that you played in middle school and have since quit? _____ Yes _____ No
If you checked “yes”, please use this sport when answering questions for this part of the survey
and identify this sport here: *(If you quit more than 1 sport, choose the one that you played the
most).
Sport: _________________________________

* IF YOU ANSWERED YES, PLEASE SKIP QUESTION 2
& 3 AND TURN THE PAGE TO CONTINUE *
** Answer #2 only if you answered NO in question 1…
2. Is there a sport that you have decreased your amount of participation in? _____ Yes _____
No
If you checked “yes”, please use this sport when answering questions for this part of the survey
and identify this sport here: *(If you decreased your participation in more than 1 sport, choose
the one that you decreased your participation in the most).
Sport: __________________________________

* IF YOU ANSWERED YES, PLEASE SKIP QUESTION 3
AND TURN THE PAGE TO CONTINUE *
** Answer # 3 only if you answered NO in question 1 and 2…
3. If you haven’t quit or decreased your participation in any of your middle school sports,
please choose the sport that you played the least in middle school and use this sport when
answering questions for this part of the survey.
Sport: __________________________________

!

150!

Directions: With this 1 sport in mind, please indicate for each question below whether
you experienced any concerns in this sport (i.e. that you experienced this situation and were to
some extent troubled by it) by circling the appropriate number for each question.
* If a situation does not apply for you, please circle “1 – Not at all”
When you played this sport, how concerned were you with -1. Getting sweaty
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-2. Getting injured because of how physical this sport is
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-3. Messing up your hair
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-4. That your friends teased you for playing this sport
1.
2.
3.
4.
5.

!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

151!

When you played this sport, how concerned were you-5. That your uniform was too tight
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-6. Not being able to wear your jewelry
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-7. That your uniform made you look too muscular
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-8. That spending time with this sport kept you away from doing things that other girls
at school do (e.g. go to the mall, hang out with friends)
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-9. That your uniform was too short
1.
2.
3.
4.
5.

!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

152!

When you played this sport, how concerned were you-10. That your friends did not understand why you played this sport
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-11. Messing up your nails/breaking a nail
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-12. That some people in society view this sport as a really physical sport
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-13. Boys seeing you sweat
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-14. That your uniform was too baggy
1.
2.
3.
4.
5.

!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

153!

When you played this sport, how concerned were you-15. That to be good at this sport, you have to have muscles/be strong
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-16. That it is not considered “cool” for a girl to play this sport
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-17. Being thought to be a tomboy
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-18. Being called a tomboy
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-19. That boys teased you for playing this sport
1.
2.
3.
4.
5.

!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

154!

When you played this sport, how concerned were you with-20. Being thought to be a lesbian
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-21. Being called a lesbian
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you with-22. Messing up your makeup
1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

When you played this sport, how concerned were you-23. That spending time with this sport kept you away from spending time with someone
you were dating
1.
2.
3.
4.
5.

!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

155!

From the 23 questions that you just answered…
For the concerns that you rated higher (e.g. a 3, 4, or 5):
How much of an IMPACT did these concerns have with you
DECREASING your involvement with this sport in middle school?
*If you haven’t decreased your involvement, please just circle “1 – No impact at all”
1.
2.
3.
4.
5.

!

No impact at all on my decreased involvement in this sport
A small impact on my decreased involvement in this sport
A moderate impact on my decreased involvement in this sport
A great impact on my decreased involvement in this sport
A very great impact on my decreased involvement in this sport

156!

Your Experience Playing Sports in Middle School
Directions: Please indicate whether any of these problems CAUSED YOU ANY PERSONAL
CONCERN with ALL THE SPORT(S) that you played in middle school (i.e., that you
experienced this problem and were to some extent troubled by it). If so, please indicate to what
extent by circling the appropriate number for each question.
1. Because American society traditionally places less value on girls’ participation in sports,
the female athlete receives little recognition for her skills and accomplishments.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

2. Girls are usually expected to have low levels of sports skills, but athletes must develop
their skills to a high level.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

3. Girls are expected to lose consistently to boys when competing with them, but athletes
should strive to win over all opponents.
I have personally experienced this problem -1.
2.
3.
4.
5.

!

157!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

4. Girls generally are expected to express their emotions freely (giggling, cheering, crying),
but athletes are supposed to remain poised during competition.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

5. Because our society typically feels that sports are for men, the woman athlete must risk
the female “jock” image when she participates in sports.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

6. Girls are expected to look attractive and dress well, but participation in athletic activities
often results in sweating, messing up one’s hair, and broken fingernails.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

7. Girls have traditionally been characterized as passive and submissive, but athletic
competition often requires dominating and aggressive behavior.
I have personally experienced this problem -1.
2.
3.
4.
5.

!

158!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

8. Girls should be available for dating, hanging out with friends, and other social activities,
but an athlete needs to devote much time to practices and evening games.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

9. Parents may encourage their daughters to participate in many kinds of activities, but an
athlete often devotes a great deal of attention to sport.
I have personally experienced this problem -1.
2.
3.
4.
5.

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

10. Girls usually are expected to show very little knowledge of or interest in sports, but an
athlete must possess an in-depth knowledge of sport.
I have personally experienced this problem -1.
2.
3.
4.
5.

!

159!

Not at all
To a small extent
To a moderate extent
To a great extent
To a very great extent

From the 10 questions that you just answered…
For the questions that you rated higher (e.g. a 3, 4, or 5):
How much of an IMPACT did these “problems” have with you
DECREASING your sport participation in middle school?
*If you haven’t decreased your involvement, please just circle “1 – No impact at all”
1.
2.
3.
4.
5.

No impact at all on my decreased involvement in sport
A small impact on my decreased involvement in sport
A moderate impact on my decreased involvement in sport
A great impact on my decreased involvement in sport
A very great impact on my decreased involvement in sport

Thank you very much for your participation in this survey!

!

160!

APPENDIX C
Table 47
Factor Analysis of Psychosocial Concerns Scale from Pilot Data Collection with 100-level
Kinesiology Students (n = 221)
Psychosocial Concerns Items
PSC 1 PSC 2 PSC 3 PSC 4
PSC 1, Image Concerns
Getting sweaty due to running a lot
.752
Getting sweaty and boys seeing you sweat
.615
Messing up your hair
.783
Messing up your makeup
.745
Messing up your nails/breaking a nail
.632
Not being able to wear your jewelry
.728
That your uniform made you look too muscular
.733
PSC 2, Negative Stereotypes
Being thought to be a lesbian
.688
Being called a lesbian
.826
Being thought to be a tomboy
.817
Being called a tomboy
.744
PSC 3, Physicality of Sport
That to be good at this sport, you had to have muscles/be
.767
strong
That this sport is seen in society as a really physical sport
.790
Getting injured because this sport is very rough/physical
.733
PSC 4, Impact of Others
That it's not considered "cool" for a girl to play this sport
-.535
That your friends would not understand why you played
-.702
this sport
That your friends would tease you for playing this sport
-.846
That boys would tease you for playing this sport
-.770
Items that did not load
That spending time with this sport kept you away from
doing things that other girls at school do (e.g. go to the
mall, hang out with friends)
That spending time with this sport kept you away from
being with your boyfriend
That your uniform was too tight or too short
That your uniform was too baggy

!

161!

APPENDIX D
Table 48
Number and Percentages for the Gender Role Conflict 10 items, Average Gender Role Conflict
by Category, and Impact – Gender Role Conflict (n = 236)
To a Moderate To a Great
To a Very
Not at All (1) To a Small
Extent (2)
Extent (3)
Extent (4)
Great Extent
(5)
Item #
#
%
#
%
#
%
#
%
#
%
#1* Little
120
51.1% 54
23.0% 33
14.0% 18
7.7%
10
4.3%
recognition
for skills
#2 Low level 97
41.1% 63
26.7% 40
16.9% 25
10.6% 11
4.7%
of skill
#3 Girls lose 89
37.7% 49
20.8% 41
17.4% 34
14.4% 23
9.7%
to boys
#4 Express
92
39.0% 56
23.7% 48
20.3% 23
9.7%
17
7.2%
emotions
#5 Risk jock 156
66.1% 31
13.1% 30
12.7% 14
5.9%
5
2.1%
image
#6 Risk
111
47.0% 54
22.9% 39
16.5% 15
6.4%
17
7.2%
looking
attractive
#7 Athletes
109
46.2% 53
22.5% 41
17.4% 23
9.7%
10
4.2%
aggressive
#8* Date or
87
37.0% 56
23.8% 47
20.0% 21
8.9%
24
10.2%
practice
#9 Parent
74
31.4% 59
25.0% 52
22.0% 29
12.3% 22
9.3%
role
#10
104
44.1% 56
23.7% 39
16.5% 12
5.1%
25
10.6%
Knowledge
of sport
Average
103.9 44.1% 53.1 22.5% 41.0 17.4% 21.4 9.1% 16.4 7.0%
GRC per
category
Impact –
134
56.8% 55
23.3% 33
14.0% 11
4.7% 3
1.3%
GRC
n* = 235

!

162!

APPENDIX E
Table 49
Number and Percentages for the Psychosocial Concerns 23 items, Average Psychosocial
Concerns by Category, and Impact – Psychosocial Concerns (n = 236)
Not at All
To a Small
To a Moderate To a Great
To a Very
(1)
Extent (2)
Extent (3)
Extent (4)
Great Extent
(5)
Item #
#
%
#
%
#
%
#
%
#
%
#1 Getting
115 48.7% 61 25.8%
39 16.5%
16 6.8%
5
2.1%
sweaty
#2 Getting
90 38.1% 83 35.2%
41 17.4%
14 5.9%
8
3.4%
injured
#3 Messing up 157 66.5% 39 16.5%
25 10.6%
5
2.1%
10
4.2%
hair
#4 Friends
209 88.6% 12 5.1%
9
3.8%
3
1.3%
3
1.3%
tease
#5 Uniform too 147 62.3% 48 20.3%
24 10.2%
12 5.1%
5
2.1%
tight
#6 Not wear
188 79.7% 29 12.3%
10 4.2%
6
2.5%
3
1.3%
jewelry
#7 Look
203 86.0% 26 11.0%
4
1.7%
2
.8%
1
.4%
muscular
#8 Unable to do 106 44.9% 75 31.8%
39 16.5%
8
3.4%
8
3.4%
other things
#9 Uniform too 170 72%
37 15.7%
14 5.9%
7
3.0%
8
3.4%
short
#10 Friends not 194 82.2% 24 10.2%
13 5.5%
2
.8%
3
1.3%
understand
#11 Break nails 191 80.9% 26 11.0%
6
2.5%
7
3.0%
6
2.5%
#12 Sport too
183 77.5% 30 12.7%
17 7.2%
4
1.7%
2
.8%
physical
#13 Boys see
146 61.9% 48 20.3%
21 8.9%
12 5.1%
9
3.8%
me sweat
#14 Uniform
155 65.7% 45 19.1%
19 8.1%
9
3.8%
8
3.4%
too baggy
#15 Have
117 49.6% 56 23.7%
33 14.0%
22 9.3%
8
3.4%
muscles to be
good
#16 Not cool to 206 87.3% 16 6.8%
8
3.4%
2
.8%
4
1.7%
play sport
#17 Thought to 202 85.6% 18 7.6%
8
3.4%
6
2.5%
2
.8%
be tomboy
#18 Called
211 89.4% 7
3.0%
7
3.0%
7
3.0%
4
1.7%
tomboy

!

163!

APPENDIX E (cont’d)
Table 49 (cont’d)
Not at All
(1)
Item #
#19 Teased for
playing
#20 Thought to
be lesbian
#21 Called
lesbian
#22 Messing
up makeup
#23 Time away
from someone
I date
Average PSC
per category
Impact –
PSC*
n* = 234

!

To a Small
Extent (2)

To a Moderate
Extent (3)

To a Great
Extent (4)

#
208

%
88.1
%
86.9
%
87.7
%
75.4
%
68.6
%

#
13

%
5.5%

#
8

%
3.4%

#
5

%
2.1%

To a Very
Great Extent
(5)
#
%
2
.8%

12

5.1%

9

3.8%

3

1.3%

7

3.0%

9

3.8%

8

3.4%

2

.8%

10

4.2%

37

15.7%

14

5.9%

4

1.7%

3

1.3%

33

14.0%

24

10.2%

8

3.4%

9

3.8%

72.8
%
52.1
%

34.
1
57

14.4%

17. 7.4%
4
34 14.4%

7.2 3.1%

5.6

2.4%

14

6

2.5%

205
207
178
162
171
.7
123

24.2%

164!

5.9%

REFERENCES

!

165!

REFERENCES

Acosta, R. V. (1993). The minority experience in sport: Monochromatic or technicolor. Women
in sport: Issues and controversies, 204–213.
Adams, N., Schmitke, N., & Franklin, A. (2005). Tomboys, dykes, and girly girls: Interrogating
the subjectivities of adolescent female athletes. Women's Studies Quarterly, 33, 17–34.
Allison, M. T. (1991). Role conflict and the female athlete: Preoccupations with little grounding.
Journal of Applied Sport Psychology, 3(1), 49–60.
Allison, M. T., & Butler, B. (1984). Role conflict and the elite female athlete: Empirical findings
and conceptual dilemmas. International Review for the Sociology of Sport, 19(2), 157–
168. doi:10.1177/101269028401900205
Anthrop, J., & Allison, M. T. (1983). Role conflict and the high school female athlete. Research
Quarterly for Exercise and Sport, 54, 104–111.
Arient, J. M. (2006). The impact of role models on the perceived sport competence and gender
role conflict of young female athletes. (Doctoral dissertation). Retrieved from ProQuest.
(3234971).
Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and
Clinical Psychology, 42, 155–162.
Biscomb, K., Matheson, H., Beckerman, N., Tungatt, M., & Jarrett, H. (2000). Staying active
while still being you: Addressing the loss of interest in sport amongst adolescent girls.
Women in Sport and Physical Activity Journal, 9(2), 79–98.
Boutilier, M. A., & SanGiovanni, L. (1983). The American woman's sporting history. In M. A.
Boutilier & L. SanGiovanni (Eds.), The sporting woman (pp. 23–48). Champaign, IL:
Human Kinetics.
Boyer, E. M. (2007). Psychological benefits of sport participation and physical activity for
adolescent females. (Unpublished doctoral dissertation). University of North Texas,
Denton (TX).
Boyington, J. E. A., Carter-Edwards, L., Piehl, M., Hutson, J., Langdon, D., & McManus, S.
(2008). Cultural attitudes toward weight, diet, and physical activity among overweight
African American girls. Preventing Chronic Disease, 5, 1–9.
Bruening, J. E., Armstrong, K. L., & Pastore, D. L. (2005). Listening to the voices: The
experiences of African American female student athletes. Research Quarterly for
Exercise and Sport, 76(1), 82–100.

!

166!

Bruening, J. E., Borland, J. F., & Burton, L. J. (2008). The impact of influential others on the
sport participation patterns of African American female student-athletes. Journal for the
Study of Sports and Athletes in Education, 2(3), 379–417.
Burk, B., & Shinew, K. (2013). Factors That Impact African American Girls' Participation in
Health-Promoting Leisure Activities. Journal of Park and Recreation Administration, 31,
1–14.
Butcher, J., Linder, K., & Jones, D. (2002). Withdrawal from competitive youth sport: A
retrospective ten-year study. Journal of Sport Behavior, 25(2), 145–163.
Casper, J. M., Bocarro, J. N., Kanters, M., & Floyd, M. F. (2011). “Just Let Me Play!”—
Understanding constraints that limit adolescent sport participation. Journal of Physical
Activity & Health, 8(Suppl 1), S32–S39.
Chusmir, L. H. (1986). Development and Validation of the Sex Role Conflict Scale. The Journal
of Applied Behavioral Science, 22(4), 397–409. doi:10.1177/002188638602200404
Clasen, P. R. W. (2001). The female athlete: Dualisms and paradox in practice. Women and
Language, 24(2), 36–41.
Coakley, J., & White, A. (1992). Making Decisions. Sociology of Sport Journal, 9, 20–35.
Cockburn, C., & Clarke, G. (2002). "Everybody's looking at you!": Girls negotiating the
‘femininity deficit’ they incur in physical education. Women's Studies International
Forum, 25, 651–665.
Colley, A., Nash, J., O'Donnell, L., & Restorick, L. (1987). Attitudes to the female sex role and
sex-typing of physical activities. International Journal of Sport Psychology, 18(1), 19–29.
Collins, J. (2011). Beyond the beauty salon: Sport, women of color and their hair. Unpublished
masters thesis, University of Maryland.
Cooky, C. (2009). "Girls Just Aren‘t Interested": The social construction of interest in girls’ sport.
Sociological Perspectives, 52(2), 259–283. doi:10.1525/sop.2009.52.2.259
Cox, B., & Thompson, S. (2000). Multiple bodies: Sportswomen, soccer and sexuality.
International Review for the Sociology of Sport, 35(1), 5–20.
doi:10.1177/101269000035001001
Czisma, K. A., Wittig, A., & Schurr, K. (1988). Sport stereotypes and gender. Journal of Sport &
Exercise Psychology, 10(1), 1–13.
Daniels, E., & Leaper, C. (2006). A longitudinal investigation of sport participation, peer
acceptance, and self-esteem among adolescent girls and boys. Sex Roles, 55(11-12), 875–

!

167!

880. doi:10.1007/s11199-006-9138-4
Desertrain, G. S., & Weiss, M. R. (1988). Being female and athletic: A cause for conflict? Sex
Roles, 18(9), 567–582.
Dwyer, J., Allison, K. R., Goldenberg, E. R., Fein, A. J., Yoshida, K. K., & Boutilier, M. A.
(2006). Adolescent girls' perceived barriers to participation in physical activity.
Adolescence, 41(161), 75-89.
Eccles, J. S., & Barber, B. L. (1999). Student council, volunteering, basketball, or marching band.
Journal of Adolescent Research, 14(1), 10–43.
Engel, A. (1994). Sex roles and gender stereotyping in young women's participation in sport.
Feminism & Psychology, 4(3), 439–448. doi:10.1177/0959353594043016
Ewing, M. E., Gano-Overway, L. A., Branta, C. F., & Seefeldt, V. D. (2002). The role of sports
in youth development. Paradoxes of Youth and Sport, 31–47.
Fallon, M. A., & Jome, L. M. (2007). An exploration of gender-role expectations and conflict
among women rugby players. Psychology of Women Quarterly, 31(3), 311–321.
doi:10.1111/j.1471-6402.2007.00374.x
Fox, C. K., Barr-Anderson, D., Neumark-Sztainer, D., & Wall, M. (2009). Physical activity and
sports team participation: Associations with academic outcomes in middle school and
high school students. Journal of School Health, 80, 31–37.
Goffman, E. (1961). Asylyums: Essays on the social situation of mental patients and other
inmates. Anchor Books.
Goldberg, A. D., & Chandler, T. J. L. (1991). Sport participation among adolescent girls: Role
conflict or multiple roles? Sex Roles, 25(3), 213–224.
Gould, D. (1987). Understanding attrition in children’s sport. In D. Gould & M. Weiss (Eds.),
Advances in pediatric sport sciences (pp. 61–85). Champaign, IL: Human Kinetics.
Greendorfer, S. L. (1993). Gender role stereotypes and early childhood socialization. In G. L.
Cohen (Ed.), Women in sport: Issues and controversies (pp. 3–14). Sage Publications.
Grieser, M., Neumark-Sztainer, D., Saksvig, B. I., Lee, J. S., Felton, G. M., & Kubik, M. Y.
(2008). Black, Hispanic, and white girls’ perceptions of environmental and social support
and enjoyment of physical activity. Journal of School Health, 78(6), 314–320.
Grieser, M., Vu, M. B., Bedimo-Rung, A. L., Neumark-Sztainer, D., Moody, J., Young, D. R., &
Moe, S. G. (2006). Physical activity attitudes, preferences, and practices in African
American, Hispanic, and Caucasian Girls. Health Education & Behavior, 33(1), 40–51.
doi:10.1177/1090198105282416

!

168!

Guillet, E., Sarrazin, P., & Fontayne, P. (2000). “If it contradicts my gender role, I'll stop”:
Introducing survival analysis to study the effects of gender typing on the time of
withdrawal from sport practice: A 3-year study. European Review of Applied Psychology,
50(4), 417–421.
Guillet, E., Sarrazin, P., Fontayne, P., & Brustad, R. J. (2006). Understanding female sport
attrition in a stereotypical male sport within the framework of Eccles's expectancy-value
model. Psychology of Women Quarterly, 30(4), 358–368. doi:10.1111/j.14716402.2006.00311.x
Hair, J. J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate
data analysis (6th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.
Hansen, D. M., Larson, R. W., & Dworkin, J. B. (2003). What adolescents learn in organized
youth activities: A survey of self-reported developmental experiences. Journal of
Research on Adolescence, 13(1), 25–55.
Hargreaves, J. A. (1994). Sporting females: Critical issues in the history and sociology of
women's sports. New York: Routledge.
Harrison, L., Lee, A. M., & Belcher, D. (1999). Race and gender differences in sport
participation as a function of self-schema. Journal of Sport & Exercise Psychology, 23(3),
287–307. doi:10.1177/0193723599233004
Hoiness, A., Weathington, B., & Cotrell, A. (2008). Perceptions of female athletes based on
observer characteristics. Athletic Insight: The Online Journal of Sport Psychology, 10,
43–54.
Kane, M. J. (1988). The female athletic role as a status determinant within the social systems of
high school adolescents. Adolescence, 23, 253–264.
Kimm, S., Glynn, N. W., & Kriska, A. M. (2002). Decline in physical activity in Black girls and
White girls during adolescence. New England Journal of Medicine, 347, 709–715.
Klomsten, A. T., Marsh, H. W., & Skaalvik, E. M. (2005). Adolescents’ perceptions of
masculine and feminine values in sport and physical education: A study of gender
differences. Sex Roles, 52(9-10), 625–636. doi:10.1007/s11199-005-3730-x
Koivula, N. (1995). Ratings of gender appropriateness of sports participation: Effects of genderbased schematic processing. Sex Roles, 33(7), 543–557.
Koivula, N. (2001). Perceived characteristics of sports categorized as gender-neutral, feminine
and masculine. Journal of Sport Behavior, 24, 377–393.
Krane, V. (2001). We can be athletic and feminine, but do we want to? Challenging hegemonic
!

169!

femininity in women's sport. Quest, 53, 115–133.
Krane, V., Choi, P. Y. L., Baird, S. M., Aimar, C. M., & Kauer, K. J. (2004). Living the paradox:
Female athletes negotiate femininity and muscularity. Sex Roles, 50(5), 315–329.
Larabee, A. (2011). Negotiating the female-athlete paradox: Examining gender identity and body
image. (Doctoral dissertation). Retrieved from ProQuest. (3453929)
Lavoi, N. (2011). Trends in gender-related research in sport and exercise psychology. Revista de
Iberoamericana de Psicologia del Ejercicio y el Deporte, 6, 269–281.
Lavoi, N., & Wiese-Bjornstal, D. (2007). Physical activity behavior: What girls do. In M. J.
Kane & N. Lavoi (Eds.), The 2007 Tucker Center Research Report Developing
Physically Active Girls: An Evidence-based Multidiscliplinary Approach. Minneapolis,
MN: University of Minnesota.
Leaper, C., & Brown, C. S. (2008). Perceived experiences with sexism among adolescent girls.
Child Development, 79(3), 685–704.
Lippman, L., Burns, S., & McArthur, E. K. (1996). Urban schools: The challenge of location
and poverty. DIANE Publishing.
Mabry, I. R., Young, D. R., Cooper, L. A., Meyers, T., Joffe, A., & Duggan, A. K. (2003).
Physical activity attitudes of African American and white adolescent girls. Ambulatory
Pediatrics, 3(6), 312–316.
Malcom, N. L. (2003). Constructing female athleticism: A study of girl's recreational softball.
Qualitative Inquiry, 46(10), 1387–1404. doi:10.1177/0002764203046010007
Marsh, H. W., & Kleitman, S. (2003). School athletic participation: Mostly gain with little pain.
Journal of Sport & Exercise Psychology, 25, 205–228.
Martens, M. P., & Webber, S. N. (2002). Psychometric properties of the Sport Motivation Scale:
An evaluation with college varsity athletes from the US. Journal of Sport & Exercise
Psychology, 24(3), 254–270.
Massengale, J. D., & Locke, L. F. (1976). The teacher/coach in conflict: A selective analysis of
perceived and experienced dysfunction within the occupational role. Meeting of the
American Alliance for Health, Physical Education and Recreation, Milwaukee, WI.
Matteo, S. (1986). The effect of sex and gender-schematic processing on sport participation. Sex
Roles, 15, 417–432.
McCallister, S. G., Blinde, E. M., & Phillips, J. M. (2003). Prospects for change in a new
millennium: Gender beliefs of young girls in sport and physical activity. Women in Sport
and Physical Activity Journal, 12, 83–109.

!

170!

Meân, L. J., & Kassing, J. W. (2008). “I Would Just Like to be Known as an Athlete”: Managing
hegemony, femininity, and heterosexuality in female sport. Western Journal of
Communication, 72(2), 126–144. doi:10.1080/10570310802038564
Messner, M. (1988). Sports and male domination: The female athlete as contested ideological
terrain. Sociology of Sport Journal, 5, 197–211.
Messner, M., & Sabo, D. (1990). Toward a critical feminist reappraisal of sport, men, and the
gender order. In M. Messner & D. Sabo (Eds.), Sport, men and the gender order: Critical
feminist perspectives (pp. 1–15). Champaign, IL: Human Kinetics.
Miller, J. L., & Levy, G. D. (1996). Gender role conflict, gender-typed characteristics, selfconcepts, and sport socialization in female athletes and nonathletes. Sex Roles, 35(1),
111–122.
Miller, J., & Heinrich, M. (2001). The gender role conflict in middle school and college female
athletes and non-athletes. Physical Educator, 58(3), 124–133.
Monaghan, S. (2012). High school sports participation achieves all-time high. National
Federation of State High School Associations (NFHS). Retrieved from
http://www.nfhs.org/content.aspx?id=7495.
National Center for Education Statistics. (2012). Improving the measurement of socioeconomic
status for the National Assessment of Educational Progress: A theoretical foundation.
Retrieved from
http://nces.ed.gov/nationsreportcard/pdf/researchcenter/Socioeconomic_Factors.pdf
National Council of Youth Sports. (2008). National Council of Youth Sports: Report on trends
and participation in organized youth sports. Retrieved from
http://www.ncys.org/pdfs/2008/2008-ncys-market-research-report.pdf
National Federation of State High School Associations. (2012, August 23). 2011-12 High School
Athletics Participation Survey. National Federation of State High School Associations
(NFHS). Retrieved from http://www.nfhs.org/content.aspx?id=3282
Oliver, K. L., Hamzeh, M., & McCaughtry, N. (2009). Girly girls can play games/Las ninas
pueden jugar tambien: Co-creating a curriculum of possibilities with fifth-grade girls.
Journal of Teaching in Physical Education, 28(1), 90–110.
Parsons, E. M., & Betz, N. E. (2001). The relationship of participation in sports and physical
activity to body objectification, instrumentality, and locus of control among young
women. Psychology of Women Quarterly, 25(3), 209–222. doi:10.1111/1471-6402.00022
Pate, R. R., Trost, S. G., Levin, S., & Dowda, M. (2000). Sports participation and health-related
behaviors among US youth. Archives of Pediatrics and Adolescent Medicine, 154, 904-

!

171!

911.
Physical Activity and Health: A Report of the Surgeon General. (1996). Physical Activity and
Health: A Report of the Surgeon General. Atlanta, GA: Centers for Disease Control and
Prevention, National Center for Chronic Disease Prevention and Health Promotion.
Richman, E. L., & Shaffer, D. R. (2000). “If you let me play sports”: How might sport
participation influence the self-esteem of adolescent females? Psychology of Women
Quarterly, 24(2), 189–199.
Ross, S. R., & Shinew, K. J. (2008). Perspectives of women college athletes on sport and gender.
Sex Roles, 58(1-2), 40–57. doi:10.1007/s11199-007-9275-4
Royce, W. S., Gebelt, J. L., & Duff, R. W. (2003). Female athletes: Being both athletic and
feminine. Athletic Insight: The Online Journal of Sport Psychology, 5(1), 1–15.
Sabo, D. (2009). The Gender Gap in Youth Sports Too Many Urban Girls Are Being Left
Behind. Journal of Physical Education Recreation and Dance, 80, 35–40.
Sabo, D., & Veliz, P. (2008). Go out and play: Youth sports in America. East Meadow, NY:
Women's Sports Foundation.
Sabo, D., & Veliz, P. (2012). The decade of decline: Gender equity in high school sports. Ann
Arbor, MI: SHARP Center for Women and Girls.
Sage, G. H., & Loudermilk, S. (1979). The female athlete and role conflict. Research Quarterly,
50(1), 88–96.
Seefeldt, V., & Ewing, M. (1989). Participation and attrition patterns in American agencysponsored and interscholastic sport: An executive summary final report. North Palm
Beach, FL: Sporting Goods Manufacturer's Association.
Sirard, J., Pfeiffer, K., & Pate, R. (2006). Motivational factors associated with sports program
participation in middle school students. Journal of Adolescent Health, 38(6), 696–703.
doi:10.1016/j.jadohealth.2005.07.013
Slater, A., & Tiggemann, M. (2010). “Uncool to do sport”: A focus group study of adolescent
girls' reasons for withdrawing from physical activity. Psychology of Sport & Exercise,
11(6), 619–626. doi:10.1016/j.psychsport.2010.07.006
Slater, A., & Tiggemann, M. (2011). Gender differences in adolescent sport participation, teasing,
self-objectification and body image concerns. Journal of Adolescence, 34(3), 455–463.
doi:10.1016/j.adolescence.2010.06.007
Snyder, E. E., & Spreitzer, E. (1983). Change and variation in the social acceptance of female
participation in sports. Journal of Sport Behavior, 6, 3–8.

!

172!

Staurowsky, E. J., DeSousa, M., Ducher, G., Gentner, N., Miller, K., Shakib, S., & Williams, N.
(2009). Her life depends on it II: Sport, physical activity, and the health and well-being of
American girls and women. East Meadow, NY: Women's Sports Foundation.
Steiner, H., McQuivey, R. W., Pavelski, R., Pitts, T., & Kraemer, H. (2000). Adolescents and
sports: Risk or benefit? Clinical Pediatrics, 39(3), 161–166.
doi:10.1177/000992280003900304
Stewart, C., & Taylor, J. (2000). Why female athletes quit: Implications for coach education.
Physical Educator, 57(4), 170–177.
Taylor, T., Legrand, N., & Newton, J. (1999). Myth and reality: Exploring teenage schoolgirl
sport participation. Journal of the International Council for Health, Physical Education,
Recreation, Sport & Dance, 35(4), 32–36.
Taylor, W. C., Yancey, A. K., Leslie, J., Murray, N. G., Cummings, S. S., Sharkey, S. A., et al.
(1999). Physical activity among African American and Latino middle school girls:
Consistent beliefs, expectations, and experiences across two sites. Women & Health,
30(2), 67–82. doi:10.1300/J013v30n02_05
Theberge, N. (2000). Gender and sport. In J. Coakley & E. Dunning (Eds.), Handbook of sports
studies (pp. 322–333). Thousand Oaks, CA: Sage Publications.
Troutman, K. P., & Dufur, M. J. (2007). From high school jocks to college grads: Assessing the
long-term effects of high school sport participation on females' educational attainment.
Youth & Society, 38(4), 443–462. doi:10.1177/0044118X06290651
Vu, M. B., Murrie, D., Gonzalez, V., & Jobe, J. B. (2006). Listening to girls and boys talk about
girls' physical activity behaviors. Health Education & Behavior, 33(1), 81–96.
doi:10.1177/1090198105282443
Wiese-Bjornstal, D., & Lavoi, N. (2007). Girls' physical activity participation:
Recommendations for best practices, programs, policies, and future research. In M. J.
Kane & N. Lavoi (Eds.), The 2007 Tucker Center Research Report Developing
Physically Active Girls: An Evidence-based Multidiscliplinary Approach. Minneapolis,
MN: University of Minnesota.
Women's Sports Foundation. (1988). The Wilson report: Moms, dads, daughters, and sports.
River Grove, IL: Wilson Sporting Goods Co.
Worell, J. (2001). Encyclopedia of women and gender: Sex similarities and differences and the
impact of society on gender. San Diego, CA: Academic Press.
Wright, E. A., Griffes, K. R., & Gould, D. (N.d). A qualitative examination of girls’ sport
participation in an urban environment. Manuscript in preparation.

!

173!