Inequalities in Girls’ High School Sports Participation: How Social Class, Race/Ethnicity, and Gender Route Opportunities to Play and Persist in Athletics

Figures & data

Table 1. Participants’ Characteristics from the Women College Athlete Study

Participant	Sport	Race	Home State	Class	Median Family Income	Hometown demographics % White	Hometown demographics % Black	Caregivers’ Highest Ed Level	Athletic Participation	High School	# of Sports High School
Amanda	Crew	White	WA	Middle	$99,597	77.10 percent	13.30 percent	Advanced	Recreational, private club, school	Public	22
Brittany	Track	Black	CA	Middle	$74,869	30.80 percent	20.70 percent	BA	Recreational, private club, camps, private coach, school	Public	24
Camilla	Crew	White	CA	Upper Middle	$109,592	67.80 percent	1.20 percent	Professional	Recreational, private club, camps, school	Public	23
Captain America	Crew	White	CO	Upper Middle	$111,901	97.30 percent	0.00 percent	Professional	Recreational, private club, camps, private coach, school	Private	23
Casey	Crew	White	CA	Upper Middle	$167,561	89.80 percent	0.60 percent	BA	Recreational, private club, school	Public	24
Chantae	Track	Black	CA	Lower	$60,665	67.40 percent	8.20 percent	HS	Recreational, private club, private coach, school	Public	19
Chelsea	Crew	White	CA	Middle	N/A*	N/A*	N/A	BA	Recreational, private club, camps, school	Private	30
CM	Track	White	CA	Upper Middle	$109,762	50.80 percent	2.10 percent	BA	Recreational, private club, camps, school	Public	24
Cooper	Crew	White	WA	Upper Middle	$91,149	92.30 percent	0.80 percent	BA	Recreational, private club, camps, school	Private	16
Ella	Crew	White	Germany	Middle	N/A	N/A	N/A	Professional	Recreational, private club, camps, school	Public	N/A**
Imani	Track	Black/Spanish Japanese	CA	Middle	$81,498	62.10 percent	10.80 percent	HS	Recreational, private club, camps, private coach, school	Private	24
Josephine	Track	Black	CA	Upper Middle	$81,498	62.10 percent	10.80 percent	BA	Recreational, private club, camps, private coach, school	Private	6
Joy	Crew	White	MI	Upper Middle	96,210	72.40 percent	7.60 percent	Professional	Recreational, private club, camps, school	Private	27
Kalie	Crew	White	WA	Upper Middle	$99,597	77.10 percent	13.30 percent	BA	Recreational, private club, camps, school	Private	21
Kayla	Track	Chicana White	CA	Middle	$97,628	59.40 percent	5.90 percent	HS	Recreational, school	Public	23
Laura	Crew	White	CT	Upper Middle	$211,313	93 percent	0.40 percent	BA	Recreational, private club, camps, school	Private	34
Lisa	Crew	White	Germany	Middle	N/A	N/A	N/A	BA	Recreational, private club, camps, school	Public	N/A
London	Track	White/Middle Eastern	CA	Upper Middle	$136,150	85.20 percent	0.50 percent	BA	Recreational, private club, camps, school	Private	24
Monique	Crew	White	CA	Upper Middle	$167,561	89.80 percent	0.60 percent	HS	Recreational, private club, camps, school	Public	24
Morgan	Crew	White	NY	Middle	$120,860	87.60 percent	2.70 percent	Advanced	Recreational, private club, camps, school	Public	26
Noelle	Crew	White	CA	Upper Middle	$69,944	90.10 percent	0.20 percent	BA	Recreational, private club, school	Public	22
Sanya	Track	White	CA	Lower	$114,764	82.90 percent	1.80 percent	BA	Recreational, school	Public	24
Savannah	Crew	White	CA	Lower	$92,192	79.90 percent	0.70 percent	HS	Recreational, private club, school	Public	24
Sophia	Crew	White	NJ	Upper Middle	117,727	39.80 percent	6.60 percent	Advanced	Recreational, private club, school	Private	31
Stella	Crew	White	HI	Upper Middle	85,837	21.50 percent	2.50 percent	Advanced	Recreational, private club, school	Private	34
Taylor	Track	White	CA	Upper Middle	$109,693	81.60 percent	1.60 percent	BA	Recreational, private club, camps, school	Public	26
Vera	Track	Black/Haitian/German	Germany	Middle	N/A	N/A	N/A	HS	Recreational, private club, camps, school	Public	N/A
Victoria	Crew	White	Canada	Upper Middle	N/A	N/A	N/A	BA	Recreational, private club, camps, school	Public	27

*Data for hometown median income and racial demographics came from the U.S. Census Bureau (2016) data. No adequate comparison exists for countries represented in this study. As a result, no median income or demographics are listed for international student athletes in the study.

**Data unavailable for some international college athletes.

Table 2. Summary of Findings from the Women College Athlete Study

Codes, Definitions, and Themes	Representative Quotes and Evidence
THEME 1: Intensive Parenting (Fixed code: Habitus | Lower-order codes: Components of habitus)
Parental expectations and socialization – Parents value and expect sports participation; believe sports develop dispositions and skills needed for academic and career success	– “My family really, really values [sports] because I think it teaches you a lot. It teaches you competitiveness, determination, it teaches you how to compete, and how to be mentally tough, and it also teaches you how to work with other people in different environments.” (Brittany) – “Sports were encouraged … I was definitely pushed into sports … my parents pushed me into sports because they saw that I could excel. But also, just because they wanted me to be doing stuff. Have extracurriculars, have interests … . But yeah, but I don’t think I was pushed into sports for the sake of sports, but for the sake of being a well-rounded person. Having interests and drive.” (Noelle)
Parental investments – Parents invest time, money, and their emotions in child’s sport career	– “My mom really supported me, in my rowing, and she would drive me to practice every day … . And [rowing] is a very expensive sport. I’m really lucky that my parents could afford to send me [to a private rowing club].” (Casey) – “To get me through a year of ski racing at the semi-lower level was about the same as going to private school, so like 25–30 thousand dollars a year. And you only race for half of the year.” (Victoria) – “I could [play sports] because of my family’s financial situation. [Money] didn’t impact which sport I did.” (Brittany) – “My Dad is one of those people who doesn’t want to know how your friends are doing. He wants to know how you’re doing in your sport.” (CM)
Durable training – Investments and expectations carried across sport levels (club through high school), to different sports (basketball, swimming, etc.), and settings (club, school sports, etc.)	– “Besides basketball, soccer, ballet … a year of softball, a year of lacrosse, a year of volleyball … to row I had to quit basketball, our middle school team. I started [rowing] the spring of my eighth grade. Which is kinda like the winter/springtime when you start. So I quit basketball. But I did end up trying that next fall to do rowing and soccer. But that was physically exhausting. So I stopped soccer and I just 100 percent did rowing.” (Cooper) – “Probably grade four, I dropped down to three sports. I did figure skating, I did soccer in the summer, and I did ski racing in the winter. And ski racing has always been my main sport until … I quit ski racing to be a rower.” (Victoria)
THEME 2: Racial/Ethnic and Gender Segregation (Fixed codes: Structural racism and sexism | Lower-order codes: Segregation mechanisms—law, policy, rules, organization, resources)
Racial/ethnic segregation: Neighborhoods	Neighborhood characteristics—Race: 25 lived in communities 50 percent or more white population; 21 lived in communities with 67 percent or more white population; 16 of 28 lived in communities with 80 percent or more white population Descriptions of their neighborhood – “I kinda grew up in the “bubble” that’s what we call it. [There’s no] “homelessness” or people struggling.” (Camilla) – “I had the most sheltered up brining … it’s literally as suburbia as you can get … And, a lot of, money-centered behavior and pressure to ‘keep up that perfect family image.’” (CM) – “[Q]uintessential small-town America.” (Joy) – “I just grew up in this community that wasn’t diverse at all … [an] affluent part of [the East Coast] … there’s a lot of people in my neighboring towns who are in the one percent. It was very wealthy. And people dress like it, and they drove Jeeps. It was all about how you were perceived by other people.” (Laura)
Racial/ethnic segregation: Sports	– “I think, rowing is such a White-washed sport. I mean, our entire team [is White]. Just looking at my team. I’m like, ‘Wow. We’re also 75 percent Blonde.’ I think that unfortunately, sports are very racialized.” (Morgan) – “Obviously there are certain sports that are more racially geared, like basketball is more African American. I can’t think of anyone on the team who isn’t White. Crew is very Caucasian-based. And then gymnastics may have more Asians. [Race] is not really something I’ve noticed.” (Sophia)
Racial/ethnic segregation: Schools	Majority-white schools had plentiful sport opportunities. Sports sponsored at their high school: average, 24 sports. Diverse and niche sports opportunities like Badminton, squash, curling, rugby, equestrian activities, ultimate frisbee, lacrosse, field hockey, and ice hockey.
Gender segregation: Separate and unequal/fewer sports, different sports, and less resources for girls	– “My [track] coach … he was also a football coach, so, I had asked him, we were supposed to have practice, and like, he wasn’t there, and I’d seen him in the weight room with the football players and I had asked him, ‘Are we having practice?’ And he’s like, ‘Oh you go here?’ And I was like ‘OK, seriously? I don’t have a coach, what do you want me to do? You want me to just sprint five days a week and never have any jump practice? What do you expect me to do?’” (Brittany) – “There was more emphasis on the male sports in my high school … And sports like women’s rowing or women’s field hockey, we don’t really get the credit that football and basketball do. Like we put in the same amount of effort, in practice, I mean in different ways. Both for the gender, we’re both equally trying as hard, but men just get that extra advantage for being a dude than the females.” (Captain America)
THEME 3: Individualizing Inequality: Choice (Fixed codes: White habitus | feminine athletic habitus)
White habitus: Universalism, race neutral, denial of racism, comfortable in most settings	– “Where I’m from is very much just White people. But [race] never been a problem for me, I don’t have a problem with anything. People are people are people … I don’t think [race], plays a factor.” (Cooper) – “[Race is] definitely something I notice in rowing. There’s no diversity at all … . I don’t think [my race has] been to my advantage, but it’s definitely been something I’ve noticed.” (Amanda) – “I am on a team with all White people. So [race] doesn’t really come up. Because we’re all White … Everyone looks exactly the same, in my entire boat … . It’s very homogenous” (Casey)
Feminine athletic habitus: Accept gender segregated, different, and unequal athletic opportunities; sense of inferiority to men’s athletic performance	– “There’s this one guy on the team … .who would say some things about women’s rowers that weren’t so great. Something about girls steady-state being really slow. And they shouldn’t be rowing … . . I understand we’re slower, but, it doesn’t make it less hard, or less worthy … . men are faster than women, but that doesn’t make them better … . . Because it is a time. So it’s easy to say faster is better.” (Camilla) – “It’s like the idea of a girl doing something that—would be considered better than probably 99 percent of people, in general. And I think I like that. It’s annoying to me still that there’s always boys, that no matter what, the boys are going to be faster. The ones on my team and stuff. That bothers me. But, it’s just the way it is. But I think as a female athlete, I’m still able to say that I’m better than most people at this.” (Kayla) – “When we had to do track in high school and we had to race and then guys were like, ‘Oh [Vera] runs track let’s see if we can race her.’ They just think they’re better just because they’re male. And then when I jump further or when I run faster they’re surprised … that women can be very athletic and good at what they’re doing.” (Vera)

Table 3. Descriptive Statistics for All Variables Used in the Quantitative High School Cohort Analysis

Predictor Variables	M(SD)/%	Dependent Variables	M(SD)/%
Individual Level
(White)	62%	10^th grade athlete	46%
Black	11%	–basketball	11%
Latinx	14%	–soccer	9%
Asian	8%	–softball	14%
Other race/ethnicity	5%	–other team	28%
Family SES	.10 (.76)	–individual sport	12%
(Married)	63%	Persistent athlete	26%
Cohabit	16%
Single	21%
# of siblings	2.28 (1.52)
Physical disability	2%
Cognitive disability	7%
Played (another) sport^a	23–25%
Played (other) sports^a	12–20%
School Level
% White	.61 (.31)
% low SES families	1.56 (.83)
School size	3.25 (1.82)
# of school sports	8.29 (2.51)
Private	24%
(Urban)	32%
Suburban	48%
Rural	20%
(Northeast)	18.56%
Midwest	27%
South	37%
West	19%

N = 4,271; ^aThese variables are unique measures used for predicting each specific sport.

Table 4. Results from Multilevel Logistic Regressions Predicting the Likelihood of Playing a School Sport in 10^th Grade

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	−.39**	(.13)	.68	−.43**	(.13)	.65	−.41**	(.13)	.66	−.24	(.13)	.78
Latinx	−.48***	(.12)	.62	−.45***	(.12)	.64	−.45***	(.12)	.64	−.27*	(.12)	.76
Asian	−.35*	(.14)	.70	−.35**	(.14)	.70	−.34*	(.14)	.71	−.39**	(.14)	.68
Other race/ethnicity	−.46**	(.16)	.63	−.49**	(.16)	.61	−.49**	(.16)	.61	−.42*	(.16)	.66
Family SES										.41***	(.05)	1.50
Cohabit										−.33**	(.10)	.72
Single										−.24**	(.09)	.78
# of siblings										−.02	(.02)	.98
Physical disability										−.20	(.24)	.82
Cognitive disability										−.34*	(.14)	.71
School Level
% White	.76***	(.17)	2.13	.24	(.19)	1.27	.16	(.19)	1.17	.16	(.19)	1.17
% low SES families				−.19**	(.05)	.82	−.08	(.06)	.92	−.00	(.06)	1.00
School size				−.11***	(.02)	.90	−.12***	(.03)	.89	−.13***	(.03)	.88
# of school sports							.07***	(.02)	1.08	.06**	(.02)	1.06
Private							.46***	(.12)	1.58	.22	(.12)	1.24
Suburban	−.23*	(.09)	.80	−.20*	(.09)	.82	−.08	(.09)	.92	−.08	(.09)	.93
Rural	−.30*	(.12)	.74	−.31**	(.12)	.73	.01	(.13)	1.01	.03	(.13)	1.03
Midwest	−.08	(.12)	.93	−.02	(.12)	.98	−.02	(.11)	.98	.01	(.12)	1.01
South	−.33**	(.11)	.72	−.25*	(.11)	.78	−.25*	(.11)	.78	−.29**	(.11)	.75
West	−.14	(.13)	.87	−.08	(.13)	.92	−.09	(.13)	.92	−.12	(.13)	.89
Variance component	0.24			0.20			0.15			0.16

N = 4,271; OR signifies odds ratio; *p < .05; **p < .01; ***p < .001.

Table 5. Results from Multilevel Logistic Regressions Predicting the Likelihood of Playing Basketball in 10^th Grade

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	.82***	(.18)	2.26	.75***	(.19)	2.11	.74***	(.19)	2.10	1.06***	(.20)	2.90
Latinx	−.19	(.18)	.83	−.12	(.21)	.89	−.13	(.21)	.88	.09	(.22)	1.10
Asian	−.48	(.28)	.62	−.35	(.29)	.70	−.33	(.29)	.72	−.23	(.30)	.80
Other race/ethnicity	.49*	(.22)	1.63	.46*	(.22)	1.59	.46*	(.22)	1.58	.73**	(.23)	2.07
Family SES										.05	(.08)	1.05
Cohabit										−.43*	(.17)	.65
Single										−.22	(.14)	.80
# of siblings										−.00	(.04)	1.00
Physical disability										.10	(.40)	1.10
Cognitive disability										−.11	(.24)	.90
Played other sport										1.36***	(.13)	3.88
Played other sports										1.81***	(.14)	6.11
School Level
% White	1.10***	(.28)	3.00	.56	(.31)	1.74	.56	(.31)	1.75	.49	(.32)	1.63
% low SES families				−.03	(.09)	.97	−.02	(.09)	.98	.04	(.09)	1.04
School size				−.27***	(.04)	.76	−.20***	(.05)	.82	−.17**	(.05)	.84
# of school sports							−.03	(.03)	.97	−.07**	(.03)	.93
Private							.24	(.17)	1.28	.05	(.18)	1.05
Suburban	.19	(.15)	1.29	.24	(.15)	1.27	.31*	(.16)	1.36	.34*	(.16)	1.41
Rural	.59**	(.18)	1.81	.49**	(.17)	1.62	.60**	(.21)	1.83	.63**	(.21)	1.89
Midwest	.20	(.18)	1.22	.22	(.17)	1.25	.22	(.17)	1.25	.28	(.17)	1.32
South	−.20	(.18)	.82	−.14	(.17)	.87	−.15	(.17)	.86	−.04	(.18)	.96
West	.10	(.21)	1.11	.18	(.21)	1.20	.16	(.21)	1.18	.24	(.21)	1.27
Variance component	0.45			0.32			0.32			0.28

N = 4,271; OR signifies odds ratio; *p < .05; **p < .01; ***p < .001.

Table 6. Results from Multilevel Logistic Regressions Predicting the Likelihood of Playing Soccer in 10^th Grade

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	−2.01***	(.38)	.13	−1.99***	(.38)	.14	−1.95***	(.38)	.14	−1.88***	(.39)	.15
Latinx	−.72**	(.23)	.48	.69**	(.23)	.50	−.67**	(.23)	.51	−.54*	(.23)	.58
Asian	−.67*	(.27)	.51	−.73**	(.27)	.48	−.73**	(.27)	.48	−.70*	(.27)	.50
Other race/ethnicity	−1.15**	(.38)	.32	−1.16**	(.38)	.31	−1.15**	(.38)	.32	−1.13**	(.38)	.32
Family SES										.23*	(.09)	1.26
Cohabit										−.12	(.18)	.89
Single										−.03	(.16)	.97
# of siblings										.06	(.04)	1.07
Physical disability										−.02	(.39)	.98
Cognitive disability										.36	(.22)	1.44
Played other sport										.56***	(.14)	1.76
Played other sports										.96***	(.14)	2.61
School Level
% White	.30	(.30)	1.35	−.09	(.35)	.92	−.13	(.35)	.88	−.20	(.35)	.82
% low SES families				−.30**	(.11)	.74	−.20	(.11)	.82	−.16	(.12)	.85
School size				.00	(.04)	1.00	−.04	(.05)	.97	−.01	(.05)	.99
# of school sports							.09**	(.03)	1.09	.07*	(.03)	1.07
Private							.28	(.20)	1.31	.13	(.20)	1.14
Suburban	−.37*	(.14)	.69	−.33*	(.14)	.72	−.25	(.16)	.78	−.24	(.16)	.78
Rural	−.99***	(.21)	.37	−.92***	(.21)	.40	−.68**	(.24)	.51	−.69**	(.24)	.50
Midwest	−.27	(.18)	.76	−.21	(.19)	.81	−.21	(.19)	.81	−.16	(.19)	.85
South	−.23	(.18)	.79	−.16	(.18)	.85	−.15	(.18)	.86	−.09	(.18)	.92
West	−.21	(.22)	.81	−.20	(.22)	.82	−.17	(.22)	.84	−.15	(.22)	.86
Variance component	0.35			0.34			0.34			0.34

N = 4,271; OR signifies odds ratio *p < .05; **p < .01; ***p < .001.

Table 7. Results from Multilevel Logistic Regressions Predicting the Likelihood of Playing Softball in 10^th Grade

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	−.43*	(.19)	.65	−.48*	(.19)	.62	−.48*	(.20)	.62	−.56**	(.21)	.57
Latinx	−.39*	(.18)	.68	−.37*	(.18)	.69	−.37*	(.18)	.69	−.38	(.20)	.69
Asian	−1.14***	(.27)	.32	−1.10***	(.28)	.33	−1.11***	(.28)	.33	−1.14***	(.29)	.32
Other race/ethnicity	−.19	(.22)	.82	−.22	(.23)	.80	−.22	(.23)	.80	−.20	(.24)	.82
Family SES										−.22**	(.08)	.80
Cohabit										.03	(.15)	1.03
Single										.18	(.13)	1.19
# of siblings										.02	(.03)	1.02
Physical disability										−.48	(.41)	.62
Cognitive disability										−.30	(.22)	.74
Played other sport										1.09***	(.13)	2.96
Played other sports										2.09***	(.12)	8.12
School Level
% White	.21	(.24)	1.23	.02	(.28)	1.02	.03	(.28)	1.03	−.16	(.29)	.86
% low SES families				.02	(.08)	1.02	.01	(.08)	1.01	−.01	(.09)	.99
School size				−.10**	(.03)	.91	−.11*	(.04)	.89	−.07	(.05)	.93
# of school sports							.00	(.02)	1.00	−.02	(.03)	.98
Private							−.09	(.16)	.92	−.16	(.17)	.86
Suburban	.08	(.13)	1.08	.09	(.13)	1.10	.07	(.14)	1.07	.08	(.14)	1.09
Rural	.26	(.15)	1.29	.21	(.16)	1.23	.16	(.18)	1.17	.13	(.19)	1.14
Midwest	−.19	(.15)	.82	−.19	(.15)	.83	−.19	(.15)	.83	−.19	(.15)	.82
South	−.32*	(.15)	.72	−.31*	(.15)	.73	−.31*	(.15)	.74	−.19	(.15)	.83
West	−.47*	(.19)	.63	−.44*	(.18)	.65	−.43*	(.19)	.65	−.39*	(.19)	.68
Variance component	0.28			0.25			0.26			0.25

Notes: N = 4,271; OR signifies odds ratio *p < .05; **p < .01; ***p < .001.

Table 8. Results from Multilevel Logistic Regressions Predicting the Likelihood of Playing Another Team Sport in 10^th Grade

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	−.22	(.14)	.80	−.24	(.14)	.79	−.21	(.14)	.81	.01	(.16)	1.01
Latinx	−.32*	(.14)	.72	−.30*	(.14)	.74	−.29*	(.14)	.75	−.02	(.15)	.98
Asian	−.29	(.16)	.75	−.31	(.16)	.73	−.31	(.16)	.73	−.23	(.17)	.80
Other race/ethnicity	−.13	(.17)	.88	−.15	(.17)	.86	−.15	(.17)	.86	.05	(.19)	1.05
Family SES										.25***	(.06)	1.28
Cohabit										−.24*	(.12)	.78
Single										−.15	(.10)	.86
# of siblings										−.03	(.03)	.97
Physical disability										−.07	(.28)	.94
Cognitive disability										−.06	(.16)	.94
Played other sport										1.30***	(.09)	3.66
Played other sports										2.24***	(.11)	9.41
School Level
% White	.75***	(.19)	2.11	.34	(.22)	1.40	.26	(.22)	1.30	.18	(.23)	1.19
% low SES families				−.19**	(.06)	.82	−.09	(.07)	.92	−.02	(.07)	.98
School size				−.06*	(.03)	.94	−.09**	(.03)	.91	−.06	(.04)	.94
# of school sports							.09***	(.02)	1.09	.07**	(.02)	1.07
Private							.35**	(.13)	1.42	.18	(.14)	1.20
Suburban	−.18	(.10)	.84	.15	(.10)	.86	−.05	(.10)	.95	−.06	(.11)	.94
Rural	−.17	(.13)	.85	−.15	(.13)	.86	.15	(.15)	1.16	.17	(.15)	1.19
Midwest	−.19	(.13)	.83	−.13	(.12)	.88	−.13	(.12)	.88	−.10	(.13)	.91
South	−.49***	(.12)	.61	−.42**	(.12)	.65	−.42**	(.12)	.66	−.41**	(.13)	.66
West	−.29*	(.15)	.75	−.26	(.14)	.77	−.25	(.14)	.78	−.27	(.15)	.77
Variance component	0.26			0.24			0.20			0.19

N = 4,271; OR signifies odds ratio *p < .05; **p < .01; ***p < .001.

Table 9. Results from Multilevel Logistic Regressions Predicting the Likelihood of Playing an Individual Sport in 10^th Grade

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	−.52*	(.22)	.59	−.51*	(.22)	.60	−.48*	(.22)	.62	−.23	(.23)	.80
Latinx	−.52*	(.20)	.59	−.48*	(.20)	.62	−.44*	(.20)	.64	−.12	(.21)	.88
Asian	.40*	(.19)	1.49	.32	(.19)	1.38	.30	(.19)	1.35	.47*	(.20)	1.59
Other race/ethnicity	−.59*	(.28)	.55	−.60*	(.28)	.55	−.60*	(.28)	.55	−.43	(.28)	.65
Family SES										.37***	(.08)	1.45
Cohabit										−.20	(.17)	.82
Single										−.14	(.14)	.87
# of siblings										−.06	(.04)	.94
Physical disability										.29	(.37)	1.33
Cognitive disability										−.46	(.25)	.63
Played other sport										1.30***	(.12)	3.68
Played other sports										1.51***	(.13)	4.52
School Level
% White	.74**	(.26)	2.10	.18	(.29)	1.20	.16	(.29)	1.17	.08	(.30)	1.09
% low SES families				−.42***	(.09)	.66	−.30**	(.10)	.74	−.19	(.10)	.83
School size				−.02	(.03)	.98	−.11*	(.04)	.90	−.07	(.05)	.93
# of school sports							.13***	(.03)	1.14	.10***	(.03)	1.11
Private							.17	(.17)	1.19	−.12	(.17)	.89
Suburban	−.06	(.13)	.94	−.02	(.13)	.98	.03	(.13)	1.03	.06	(.14)	1.07
Rural	−.33*	(.17)	.72	−.24	(.17)	.79	−.01	(.19)	.99	.01	(.20)	1.01
Midwest	.08	(.17)	1.08	.16	(.17)	1.18	.18	(.16)	1.20	.27	(.17)	1.32
South	.08	(.16)	1.09	.19	(.16)	1.20	.22	(.16)	1.25	.33*	(.17)	1.39
West	.36	(.19)	1.43	.38*	(.19)	1.46	.43*	(.19)	1.53	.51**	(.19)	1.66
Variance component	0.27			0.23			0.19			0.16

N = 4,271; OR signifies odds ratio *p < .05; **p < .01; ***p < .001.

Table 10. Results from Multilevel Logistic Regressions Predicting the Likelihood of Persisting in Playing School Sports

	Model 1			Model 2			Model 3			Model 4
Variables	b	SE	OR	b	SE	OR	b	SE	OR	b	SE	OR
Individual Level
Black	−.45**	(.15)	.64	−.50**	(.16)	.61	−.48**	(.16)	.62	−.28	(.18)	.75
Latinx	−.53**	(.14)	.59	−.49**	(.14)	.62	−.49**	(.14)	.62	−.21	(.16)	.81
Asian	−.17	(.16)	.84	−.17	(.16)	.85	−.16	(.16)	.85	−.13	(.19)	.88
Other race/ethnicity	−.25	(.18)	.78	−.29	(.18)	.75	−.30	(.18)	.74	−.21	(.20)	.81
Family SES										.48***	(.07)	1.61
Cohabit										.19	(.13)	.83
Single										−.29*	(.11)	.75
# of siblings										−.04	(.03)	.96
Physical disability										−.01	(.31)	.99
Cognitive disability										−.73***	(.20)	.48
Played other sport										2.40***	(.11)	11.06
Played other sports										2.25***	(.13)	9.50
School Level
% White	.98***	(.20)	2.66	.31	(.22)	1.37	.23	(.22)	1.26	−.02	(.24)	.98
% low SES families				−.25***	(.06)	.78	−.14*	(.07)	.87	−.05	(.07)	.96
School size				−.14***	(.03)	.87	−.14***	(.03)	.87	−.12**	(.04)	.89
# of school sports							.06**	(.02)	1.06	.03	(.02)	1.03
Private							.44**	(.12)	1.55	.23	(.14)	1.26
Suburban	−.24*	(.10)	.79	−.19	(.10)	.83	−.06	(.10)	.94	−.08	(.11)	.92
Rural	−.18	(.13)	.84	−.18	(.13)	.84	.15	(.15)	1.17	.21	(.16)	1.23
Midwest	−.11	(.13)	.90	−.03	(.12)	.97	−.02	(.12)	.98	.11	(.13)	1.12
South	−.32*	(.12)	.73	−.22	(.12)	.80	−.22	(.12)	.80	−.12	(.13)	.89
West	.03	(.15)	1.03	.08	(.14)	1.09	.09	(.14)	1.09	.18	(.15)	1.20
Variance component	0.24			0.19			0.11			0.10

N = 4,271; OR signifies odds ratio *p < .05; **p < .01; ***p < .001.

U.S. Census Bureau. 2016. Quickfacts. United States: Median Household Income. https://www.census.gov/quickfacts/map/INC110214/00.

 Google Scholar