Will We Ever Close the Gender Gap Among Top Mathematics Achievers? Analysis of Recent Trends by Race in Advanced Placement (AP) Exams

Abstract

Analyzing the test scores of more than 10,000,000 students who participated in the Advanced Placement (AP) math exams from 1997 to 2019, this study examined the direction and magnitude of the trend in gender disparity by race in participation in and top achievement on AP Calculus AB, Calculus BC, and Statistics exams. The results of this study indicated that, in general, females’ representation in all three AP exams increased significantly. Although the findings indicated that the female-to-male ratios (FMRs) in participation in the AP math exams increased significantly from 1997 to 2019 and favored females for all races, the gender disparities among top achievers for all math exams are still substantial. The relationships between the FMRs in participation and top achievement for all AP math exams were also analyzed within races, and the possible impacts of these findings within the context of the underrepresentation of women in science, technology, engineering, and mathematics (STEM) fields were also discussed.

Keywords

Advanced Placement gender issues mathematics underrepresentation STEM

Moving through the 21st century, the daily lives of human beings evolve with the influence of science and technology at a dizzying speed. In 2011, at the Hannover Fair, German government and industry leaders announced “Industry 4.0,” referring primarily to developments in automation and digitalization of industry with the integration of technological advances such as Big Data and analytics, augmented reality, autonomous robots, and cloud computing (Tay et al., 2018). As nations strive to integrate these technological advances, they also seek opportunities to cultivate talented individuals equipped with strong skills and competencies in science, technology, engineering, and mathematics (STEM) to sustain their advancements in these fields (Bahar & Maker, 2020).

Many developed countries continue to experience wide gender gaps in STEM fields. For example, in the United States, females make up only 15% of the engineering occupations and 26% of the computer workforce (National Science Board, 2018). According to a report published by The American Association of University Women, “More than ever before in history, girls are studying and excelling in science and mathematics” (Corbett & Hill, 2015, p. 2). Although the number of girls and women studying STEM subjects has increased dramatically at all levels (Corbett & Hill, 2015), women’s share of degrees at the bachelor’s, master’s, and doctoral levels continues to lag behind that of men. For example, engineering degrees earned by women in 2016 only represented 12%, 18%, and 18% of degrees at the bachelor’s, master’s, and doctoral levels, respectively, in the United States (National Science Foundation & National Center for Science and Engineering Statistics, 2019).

Understanding the roots of gender disparities in STEM careers has been a long-investigated journey among researchers from economics to psychology. A variety of factors (e.g., PK–12 school experiences, occupational discriminations, gender stereotypes, cultural settings, psychosocial factors, education policies, and biological traits) associate with these disparities (Morris, 2013). Contemporary educational research explains this complicated issue of occupational disparities via differences in educational attainments and opportunities, particularly in PK–12 level. For example, a nationwide call from the National Council of Supervisors of Mathematics (NCSM) and the National Council of Teachers of Mathematics (NCTM) highlighted that gender differences in school mathematics account for significant disparities in many STEM-related careers (NCSM & NCTM, 2018). Furthermore, several researchers have argued that the gender-related occupational differences might have resulted from disparities among top performing students who flowed into the field from major STEM pipelines, including but not limited to advanced programs (AP, IB, and dual-enrollment classes), talent development programs, and gifted education programs (Lazear & Rosen, 1990; Makel et al., 2016; Wai et al., 2010). The researchers argued that many of the current STEM workers were prepared through these pipelines and suggested that the earlier disparities among the students possibly reflected the later gender disparities in STEM-related occupations.

Reis and Park (2001) suggested that gender disparities in PK–12 settings were due to a range of factors and helped to explain the underrepresentation in later STEM pipelines. These factors included, but are not limited to, gender differences of talented females in mathematical self-efficacy; self-confidence in STEM subjects; teacher and parent perceptions about STEM aptitudes, and gender stereotyping (Callahan et al., 1994; Ceci & Williams, 2007; Dweck, 2007; Fennema et al., 1990; Halpern, 2007; Hyde, 2007; Lubinski & Benbow, 2007; Reis & Callahan, 1989; Stumpf & Stanley, 1996).

Gender Disparities Among High Achievers in Mathematics

Large-scale meta-analytic studies investigating gender-related disparities in mathematical achievement provided researchers and educators with trend data over the span of 30 years. Hyde et al. (1990) analyzed more than 100 studies between 1960 and 1990 on gender differences in math performance. The authors found these differences became nonexistent during the following three decades. Later studies (Hyde, 2005; Hyde et al., 2008; Hyde & Linn, 2006; Lindberg et al., 2010; J. P. Robinson & Lubienski, 2011) confirmed the findings of Hyde et al. (1990) that gender differences in overall K–12 math achievement have been relatively small since the 1990s.

Contrary to the minimal gender differences in mathematics performance among the general population, many studies have reported significant disparities among high-achieving students that favor males across all grades (Andreescu et al., 2008; Benbow & Stanley, 1980, 1983; Desjarlais, 2009; Ellison & Swanson, 2010; Hedges & Nowell, 1995; Hyde et al., 1990; Makel et al., 2016; Niederle & Vesterlund, 2010; Olszewski-Kubilius & Lee, 2011; Reis & Park, 2001; Wai et al., 2010, 2018; Xie & Shauman, 2003). In particular, studies that have investigated gender disparities in mathematics among high achievers suggest that the males significantly outnumbered females significantly in the high-score ranges. Furthermore, the gender gap was dramatically widened in the upper limits of the achievement distribution (Ellison & Swanson, 2010). For example, Hedges and Nowell’s (1995) analysis of high-achieving individuals’ test scores recorded in large nationwide databases found that the female-to-male ratio (FMR) in the top 3% participants was around 0.20 to 1 and even smaller in highest levels of the achievement distribution.

Studies that explored the gender disparity in math achievement have consistently documented more males than females among top achievers (Benbow & Stanley, 1980; Ceci & Williams, 2010; Makel et al., 2016; Olszewski-Kubilius & Lee, 2011; Wai et al., 2010, 2012). In one of the earliest studies investigating gender disparities among mathematically talented students, Benbow and Stanley (1980) reported that the FMR on the SAT-Math was approximately 1 to 13 in the 700 to 800 score range. In a follow-up study of a similar student population, Makel et al. (2016) found that there were approximately 2.5 males for every female in the same score range among 320,000 talented students in 2015. These findings illustrate that the FMRs in math achievement among top achievers have increased rapidly over the past four decades.

What Is the Advanced Placement Program?

The College Entrance Examination Board (College Board), a nonprofit education organization, created the Advanced Placement (AP) program in the mid-1950s to provide college-level coursework for high-achieving high school students (Christiansen, 2009). In the first year of the program, 12 AP courses, including AP Calculus, were offered to more than 1,000 students across 110 schools (Demaree, 2016). As of 2019, 2.8 million students at 20,000 high schools across the United States took more than 5 million AP exams (College Board, 2020a). In 2010, approximately 90% of U.S. high schools offer at least one AP course for their students (College Board, 2020c).

The AP program currently offers 38 courses in 22 different subjects ranging from mathematics to foreign languages (College Board, 2019). Although schools and teachers are responsible for designing their own AP course materials, College Board requires schools to fulfill some curricular requirements before a school can label a course “AP” (College Board, 2019). Thus, schools wishing to offer AP courses must go through a formal audit process conducted by College Board to review course syllabi and resources to ensure that each AP course content meets the equivalency of a college-level course (College Board, 2019).

Almost all AP courses are year-long courses followed by end-of-year exams, designed by teams of AP teachers and college professors in a standardized format, to assess the content and skill mastery of the course (College Board, 2020a). AP exams are criterion-referenced tests, with students’ exam scores reported on a 5-point scale, with 5 representing the highest possible score (Shaw et al., 2013). According to College Board (2019), guidelines recommend that a score of 5 is a college-grade equivalent of “A,” while a score of 4 is an “A−, B+, or B,” and a score of 3 equates to a “B−, C+, or C.” In 2019, students reported their exam scores to more than 3,500 colleges to seek course credit. Although each college or university has a credit policy, most award course credit for AP scores of 3 or higher (College Board, 2019).

Potential Benefits of AP Participation and Performance

The AP program has been one of the most popular programs in high schools for exposing students to advanced academic content (Warne, 2017). Rapid growth and program prevalence have attracted interest from many researchers. Most researchers who investigated potential benefits of the program have noted that the program has essential college-related outcomes. One benefit of the program is that the successful completion of an AP course increases the possibility that a student will enroll in college (Ackerman et al., 2013; Prong, 2018). For example, many high schools assign more weight to honors and advanced classes, including dual-enrollment or AP courses than they do for typical high school courses, so completing these advanced courses successfully provides students with an opportunity to achieve higher grade point averages (GPAs).

Furthermore, successful completion of an AP course typically provides students with an opportunity to reduce college costs by earning college credit while in high school, as well as completing a college degree in a shorter time (Callahan, 2003). A passing AP exam score does not guarantee that every college or university will grant credit. However, students who earn a 3 or higher may receive college credit at many universities or colleges (Hertberg-Davis et al., 2006; Moore & Slate, 2011; Prong, 2018; Shaw et al., 2013). In addition to college admission benefits, students’ AP exam performances influence their college performance as well. Regardless of gender and ethnicity, students who earn a 3 or higher on an AP exam earn higher grades in the associated college course (Hargrove et al., 2008; Murphy & Dodd, 2009; Shaw et al., 2013).

Participation in the AP program can increase participants’ college degree persistence and STEM career interest (Adelman, 1999; Bahar & Adiguzel, 2016; Kokkelenberg & Sinha, 2010; Lubinski & Benbow, 2007; Morgan & Klaric, 2007; Sadler et al., 2014). Students’ who participated in AP program courses were more likely to graduate within 4 years when compared to students who did not participate in the AP program (Mattern et al., 2009, 2013; Shaw et al., 2013). Sadler et al. (2014) examined the relationship between taking advanced high school courses and students’ interest in pursuing a STEM career. The authors used data from 4,691 college students and found that taking calculus, 2 years of chemistry or 1 or 2 years of physics in high school significantly predicted increases in STEM career interest when compared to students who do not take these courses (Sadler et al., 2014).

Advanced classes offered at high schools, such as AP courses, have also served as a natural curricular option for gifted and talented students because most secondary schools cannot offer beyond these options (Sytsma, 2000). In the field of gifted education, AP courses have been well regarded by researchers and educators as an appropriate option to accelerate talented students through advanced resources (Feldhusen, 1995; Hertberg-Davis et al., 2006; Kettler & Hurst, 2017; A. Robinson et al., 2007). Previous research recommended that rigorous curriculum is essential to talent development (Bahar, 2013; Bahar & Maker, 2015; Bloom, 1985; Callahan, 2006; Hertberg-Davis & Callahan, 2008; Kettler & Hurst, 2017; Subotnik et al., 2011; VanTassel-Baska, 2000).

Gender Disparities for AP Mathematics Exams

Although AP programs continue to serve as a major pipeline that provides students with advanced academic opportunities, there are questions as to whether these opportunities extend to all students equitably. Researchers investigating gender differences in mathematics have raised concerns about disparities in participation to and performance in AP math exams (Amelink, 2009; Brookhart, 2009; Buck et al., 2002; Moore et al., 2012; Morris, 2013; Morris et al., 2015; Morris & Slate, 2012; Prong, 2018; Stumpf & Stanley, 1996). In one of the earliest studies investigating gender disparities in AP math exams, Stumpf and Stanley (1996) explored differences in participation in and performances on AP exams (Calculus AB and Calculus BC) and compared these differences from 1984 to 1992. Over 8 years, female student participation increased from 40% to 45% for Calculus AB and from 30% to 35% for Calculus BC. Despite the increases in female representation in both exams, the FMRs among the top-achieving students who earned a score of five continued to decline and favored males for both exams. For the Calculus AB exam, the FMR dropped from 0.72:1 to 0.67:1 and dropped from 0.76:1 to 0.69:1 for Calculus BC.

Amelink (2009) explored math participation and performance over a variety of high-stake tests, including AP Calculus AB, Calculus BC, and Statistics exams. The findings revealed that roughly equal proportions of male and female high school students took the exams in Calculus AB and Statistics in 2007 (the FMRs were 0.95 and 1.01 for Calculus AB and Statistics, respectively). However, male students outnumbered female test-takers for the Calculus BC tests (the FMR was 0.70). In terms of the number of students who scored a 5, males outperformed females on all exams. While the percentages of males who scored a 5 on Calculus AB, Calculus BC, and Statistics exams were 24%, 47%, and 15%, respectively; the percentage of females were 17%, 37%, and 9%. In a similar study, Morris and Slate (2012) investigated the relationship between gender and student performance on the AP Calculus AB, Calculus BC, and Statistics exams over three examinations years 2000, 2005, and 2010. The findings revealed a significant relationship between gender and exam scores, with males tending to score higher than females.

Gender Disparities by Race Among High Achievers in Mathematics

The literature investigating gender disparities among high achievers in mathematics is well-established. Concerning disparities, gender gaps and race, there are a limited number of studies. One possible reason for the scarcity of the literature on this topic might be due to researchers’ assumption that gender disparity would be same among racial groups (Baird, 2019).

In one of the few existing studies, Catsambis (1994) compared the development of gender gaps in learning opportunities and achievement in mathematics among White, Black, and Hispanic students. Using data from a nationally representative sample of approximately 24,500 eighth-grade students who were again surveyed in 10th grade, female students were found to perform equally well, or even better, in mathematics tests than do male students in eighth grade across all racial groups. Concerning learning opportunities, female and male students were equally likely to enroll in average-ability mathematics classes and a higher proportion of girls were enrolled in high-ability classes. However, Catsambis found that all female students tended to have less interest in mathematics and less confidence in their mathematics abilities when they were resurveyed in 10th grade. She also found that gender gaps in interest and confidence in their mathematics abilities were the largest among Hispanic and the smallest among African Americans (Catsambis, 1994).

Examining the results of the Education Longitudinal Study (ELS), Baird and Keene (2019) investigated gender disparities in mathematical ability high-achieving students’ confidence in all races. The study sample included students who recently graduated from high school with 4.0 GPAs for their high school math coursework. Using ordinary least squares (OLS) regression analysis, females of all races had lower confidence in their mathematical abilities than males, even when they had the same level of mathematical achievement. Moreover, female Asian students were found to have far less math confidence than did all other students.

Present Study

Gaps in participation and achievement in STEM pipelines not only threaten educational opportunities (Darling-Hammond, 2010; Prong, 2018), but may also explain later gender disparities in STEM careers. The following research questions guided this study:

What is the trend in gender disparity by race in participation in AP math exams?

What is the trend in gender disparity by race in top achievement in AP math exams?

To what extend are the trends in gender disparity by race in participation and top achievement related?

Methods

Participants and Setting

Data were obtained from the annually published College Board’s AP National Summary reports. These reports consist of detailed data on AP Exam participation, volume, and performance separated into categories, including exam type, grade level, and gender (College Board, 2020c). From the 38 different AP exams, only data related to three mathematics exams (Calculus AB, Calculus BC, and Statistics) were processed for the analyses. The data analyzed in this study included test scores of more than 10,000,000 students who participated in the AP mathematics exams from 1997 to 2019. The gender and racial profiles of participants in each of these three exams appear in Tables 1 to 3, respectively. Likewise, Tables 4 to 6 show the gender and racial breakdown of top achievers who scored five on the exams during this same time period.

Table 1.

The Number of Participants in the AP Calculus AB Exam and the FMRs Across Races.

Year	Native American			Black			Hispanic			Asian American			White			Total
Year	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR
1997	184	179	1.03	2,385	1,634	1.46	2,424	2,720	0.89	7,964	8,219	0.97	33,810	39,409	0.86	46,767	52,161	0.90
1998	198	193	1.03	2,532	1,804	1.40	2,634	2,996	0.88	8,271	8,588	0.96	35,407	41,360	0.86	49,042	54,941	0.89
1999	208	256	0.81	2,998	2,002	1.50	3,126	3,403	0.92	8,993	9,369	0.96	38,300	45,163	0.85	53,625	60,193	0.89
2000	206	289	0.71	3,169	2,311	1.37	3,587	4,030	0.89	10,076	10,204	0.99	42,917	49,931	0.86	59,955	66,765	0.90
2001	212	289	0.73	3,325	2,399	1.39	4,213	4,484	0.94	10,662	10,448	1.02	45,920	53,458	0.86	64,332	71,078	0.91
2002	238	330	0.72	3,542	2,629	1.35	4,717	4,936	0.96	11,653	11,486	1.01	49,472	56,450	0.88	69,622	75,831	0.92
2003	300	328	0.91	3,912	2,766	1.41	5,599	5,641	0.99	11,760	11,963	0.98	51,950	60,402	0.86	73,521	81,100	0.91
2004	298	372	0.80	4,055	2,875	1.41	6,097	6,087	1.00	12,695	12,416	1.02	54,244	62,460	0.87	77,389	84,210	0.92
2005	348	398	0.87	4,615	3,227	1.43	6,844	6,955	0.98	13,521	13,613	0.99	56,611	64,799	0.87	81,939	88,992	0.92
2006	344	412	0.83	4,937	3,516	1.40	7,535	7,505	1.00	14,754	14,295	1.03	58,690	65,500	0.90	86,260	91,228	0.95
2007	389	434	0.90	5,558	3,771	1.47	8,743	8,586	1.02	15,947	15,975	1.00	63,454	70,519	0.90	94,091	99,285	0.95
2008	361	444	0.81	5,922	4,368	1.36	10,024	9,806	1.02	16,660	16,487	1.01	66,358	73,443	0.90	99,325	104,548	0.95
2009	402	473	0.85	6,576	4,749	1.38	10,736	10,975	0.98	17,740	17,610	1.01	67,257	74,491	0.90	102,711	108,298	0.95
2010	440	473	0.93	6,928	5,277	1.31	11,915	12,138	0.98	18,925	18,683	1.01	69,150	77,133	0.90	107,358	113,704	0.94
2011	497	529	0.94	7,560	5,907	1.28	13,482	13,733	0.98	19,950	20,175	0.99	71,431	79,821	0.89	112,920	120,165	0.94
2012	508	587	0.87	7,791	6,061	1.29	14,346	15,160	0.95	20,987	21,533	0.97	72,940	81,467	0.90	116,572	124,808	0.93
2013	580	707	0.82	8,200	6,708	1.22	16,847	17,879	0.94	22,990	23,546	0.98	76,652	86,192	0.89	125,269	135,032	0.93
2014	633	666	0.95	8,962	7,012	1.28	18,813	19,242	0.98	25,085	24,079	1.04	78,311	87,905	0.89	131,804	138,904	0.95
2015	618	673	0.92	9,419	7,443	1.27	20,314	20,501	0.99	25,890	25,027	1.03	78,791	87,542	0.90	135,032	141,186	0.96
2016	537	635	0.85	8,854	6,525	1.36	25,147	24,600	1.02	25,216	24,203	1.04	78,339	85,721	0.91	138,093	141,684	0.97
2017	588	622	0.95	8,821	6,622	1.33	26,896	26,651	1.01	26,270	25,307	1.04	78,192	85,369	0.92	140,767	144,571	0.97
2018	526	589	0.89	8,339	6,196	1.35	25,983	25,462	1.02	26,393	25,177	1.05	75,611	81,856	0.92	136,852	139,280	0.98
2019	487	558	0.87	7,983	6,054	1.32	25,786	25,579	1.01	25,979	25,249	1.03	72,613	78,716	0.92	132,848	136,156	0.98

Note. AP = advanced placement; FMR = female-to-male ratio; F = female; M = male.

Table 2.

The Number of Participants in the AP Calculus BC Exam and the FMRs Across Races.

Year	Native American			Black			Hispanic			Asian American			White			Total
Year	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR
1997	12	24	0.50	182	212	0.86	218	412	0.53	2,479	3,548	0.70	4,591	8,441	0.54	8,327	14,022	0.59
1998	14	32	0.44	226	232	0.97	324	448	0.72	2,874	3,812	0.75	5,751	10,202	0.56	10,256	16,381	0.63
1999	24	45	0.53	267	283	0.94	375	569	0.66	3,099	4,460	0.69	6,669	11,712	0.57	11,525	18,762	0.61
2000	35	62	0.56	291	323	0.90	416	720	0.58	3,594	4,912	0.73	7,831	13,528	0.58	12,899	20,769	0.62
2001	24	54	0.44	331	347	0.95	485	864	0.56	4,127	5,412	0.76	8,713	14,987	0.58	14,513	23,044	0.63
2002	37	59	0.63	369	414	0.89	571	913	0.63	4,425	5,652	0.78	9,851	16,455	0.60	16,103	24,975	0.64
2003	41	85	0.48	455	497	0.92	718	1,169	0.61	4,919	6,362	0.77	10,814	17,783	0.61	17,939	27,312	0.66
2004	55	82	0.67	478	546	0.88	916	1,316	0.70	5,367	6,760	0.79	11,898	19,171	0.62	19,765	29,567	0.67
2005	54	85	0.64	554	571	0.97	1,051	1,478	0.71	6,026	7,671	0.79	12,826	19,938	0.64	21,760	31,597	0.69
2006	57	114	0.50	595	632	0.94	1,076	1,568	0.69	6,519	8,202	0.79	13,154	21,037	0.63	23,038	34,281	0.67
2007	53	101	0.52	723	681	1.06	1,250	1,946	0.64	7,331	9,463	0.77	14,834	22,529	0.66	25,724	36,890	0.70
2008	72	115	0.63	790	781	1.01	1,536	2,182	0.70	7,980	10,086	0.79	15,741	23,863	0.66	27,618	39,167	0.71
2009	78	138	0.57	863	838	1.03	1,706	2,496	0.68	8,518	10,774	0.79	16,183	24,636	0.66	28,974	41,114	0.70
2010	95	138	0.69	917	948	0.97	1,873	2,774	0.68	8,889	11,682	0.76	16,667	26,122	0.64	30,502	44,621	0.68
2011	83	152	0.55	928	997	0.93	2,212	3,220	0.69	9,870	12,839	0.77	17,686	27,341	0.65	32,457	46,904	0.69
2012	86	160	0.54	1,026	1,083	0.95	2,447	3,560	0.69	10,769	14,025	0.77	18,983	29,111	0.65	35,391	51,175	0.69
2013	134	220	0.61	1,201	1,448	0.83	2,869	4,456	0.64	12,009	15,866	0.76	20,533	32,526	0.63	38,308	56,839	0.67
2014	127	206	0.62	1,436	1,632	0.88	3,347	5,069	0.66	13,048	16,894	0.77	22,257	33,742	0.66	42,052	60,083	0.70
2015	118	220	0.54	1,531	1,650	0.93	3,718	5,585	0.67	14,017	17,183	0.82	23,272	35,317	0.66	44,723	62,648	0.71
2016	99	194	0.51	1,442	1,527	0.94	4,959	7,083	0.70	14,409	17,484	0.82	23,751	35,746	0.66	47,161	65,534	0.72
2017	124	206	0.60	1,617	1,525	1.06	5,513	7,635	0.72	15,485	18,806	0.82	25,141	36,553	0.69	50,977	68,613	0.74
2018	112	176	0.64	1,687	1,630	1.03	5,952	8,307	0.72	16,430	20,117	0.82	25,471	37,305	0.68	53,027	72,218	0.73
2019	116	165	0.70	1,565	1,614	0.97	5,758	8,218	0.70	16,794	21,293	0.79	24,860	36,648	0.68	52,581	72,686	0.72

Note. AP = advanced placement; FMR = female-to-male ratio; F = female; M = male.

Table 3.

The Number of Participants in the AP Statistics Exam and the FMRs Across Races.

Year	Native American			Black			Hispanic			Asian American			White			Total
Year	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR
1997	12	14	0.86	185	128	1.45	200	190	1.05	574	760	0.76	2,139	2,710	0.79	3,388	4,163	0.81
1998	26	30	0.87	352	234	1.50	429	390	1.10	1,184	1,302	0.91	4,559	5,272	0.86	7,263	7,959	0.91
1999	42	38	1.11	499	346	1.44	637	586	1.09	1,878	2,121	0.89	7,993	8,470	0.94	12,088	12,717	0.95
2000	60	76	0.79	748	496	1.51	878	845	1.04	2,613	3,209	0.81	11,253	11,603	0.97	16,462	17,189	0.96
2001	74	66	1.12	957	578	1.66	1,199	1,092	1.10	3,318	3,596	0.92	13,769	14,017	0.98	20,474	20,560	1.00
2002	98	104	0.94	1,158	792	1.46	1,509	1,370	1.10	3,954	4,294	0.92	16,603	16,765	0.99	24,605	24,636	1.00
2003	88	93	0.95	1,416	847	1.67	1,862	1,622	1.15	4,482	4,835	0.93	19,518	19,723	0.99	28,809	28,687	1.00
2004	125	120	1.04	1,648	993	1.66	2,302	1,991	1.16	4,906	5,510	0.89	21,788	22,158	0.98	32,525	32,538	1.00
2005	134	133	1.01	1,990	1,293	1.54	2,758	2,372	1.16	5,807	6,377	0.91	25,446	25,070	1.01	38,240	37,428	1.02
2006	145	180	0.81	2,400	1,540	1.56	3,207	2,809	1.14	6,785	7,357	0.92	28,035	27,806	1.01	43,718	43,164	1.01
2007	178	195	0.91	2,737	1,859	1.47	3,678	3,250	1.13	7,281	8,129	0.90	31,669	31,651	1.00	48,339	47,943	1.01
2008	203	202	1.00	3,432	2,171	1.58	4,644	4,061	1.14	8,274	9,206	0.90	34,717	33,983	1.02	54,082	52,452	1.03
2009	231	221	1.05	3,697	2,471	1.50	5,155	4,674	1.10	9,104	9,897	0.92	36,374	36,375	1.00	57,763	56,735	1.02
2010	263	224	1.17	4,276	2,770	1.54	6,089	5,119	1.19	10,092	10,925	0.92	39,287	39,425	1.00	64,204	62,779	1.02
2011	278	275	1.01	4,757	3,198	1.49	7,046	6,333	1.11	11,258	12,317	0.91	43,167	42,815	1.01	70,314	68,677	1.02
2012	311	295	1.05	5,243	3,666	1.43	7,976	6,847	1.16	12,155	13,182	0.92	45,337	45,061	1.01	75,712	73,453	1.03
2013	359	387	0.93	5,876	4,121	1.43	9,538	8,269	1.15	14,067	14,949	0.94	50,303	49,334	1.02	83,530	80,334	1.04
2014	422	422	1.00	6,400	4,535	1.41	11,398	9,719	1.17	15,963	15,778	1.01	53,799	52,164	1.03	91,802	86,212	1.06
2015	445	397	1.12	6,823	4,467	1.53	13,181	10,416	1.27	16,853	16,723	1.01	56,536	53,570	1.06	98,424	90,057	1.09
2016	411	426	0.96	6,585	4,276	1.54	16,635	13,603	1.22	17,059	16,712	1.02	57,925	54,506	1.06	104,053	94,500	1.10
2017	467	374	1.25	6,829	4,454	1.53	18,464	14,636	1.26	18,269	17,690	1.03	59,412	55,075	1.08	109,496	97,581	1.12
2018	416	373	1.12	6,681	4,324	1.55	18,495	14,933	1.24	19,446	19,304	1.01	60,018	55,588	1.08	111,754	100,872	1.11
2019	353	377	0.94	6,517	4,241	1.54	18,764	14,815	1.27	20,039	19,906	1.01	58,168	53,615	1.08	110,381	99,188	1.11

Note. AP = advanced placement; FMR = female-to-male ratio; F = female; M = male.

Table 4.

The Number of Top Scorers in the AP Calculus AB Exam and the FMRs Across Races.

Year	Native American			Black			Hispanic			Asian American			White			Total
Year	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR
1997	8	15	0.53	58	68	0.85	106	232	0.46	1,051	1,443	0.73	3,006	5,449	0.55	4,680	7,980	0.59
1998	7	25	0.28	77	127	0.61	128	314	0.41	1,270	1,807	0.70	4,223	8,191	0.52	6,368	11,619	0.55
1999	14	38	0.37	110	118	0.93	106	359	0.30	1,406	1,981	0.71	5,014	8,719	0.58	7,427	12,236	0.61
2000	14	35	0.40	132	145	0.91	220	448	0.49	1,572	2,202	0.71	5,899	10,265	0.57	8,292	13,877	0.60
2001	9	30	0.30	104	148	0.70	197	452	0.44	1,474	2,154	0.68	5,812	10,976	0.53	8,010	14,470	0.55
2002	21	45	0.47	142	184	0.77	283	489	0.58	2,094	2,681	0.78	7,662	12,261	0.62	10,770	16,491	0.65
2003	25	37	0.68	152	189	0.80	318	640	0.50	2,172	2,931	0.74	7,858	13,714	0.57	11,050	18,424	0.60
2004	22	43	0.51	187	230	0.81	408	737	0.55	2,644	3,251	0.81	9,712	15,224	0.64	13,670	20,556	0.67
2005	20	66	0.30	221	251	0.88	451	832	0.54	2,978	3,668	0.81	10,484	15,767	0.66	14,967	21,752	0.69
2006	33	54	0.61	237	307	0.77	610	960	0.64	3,614	4,248	0.85	11,622	17,225	0.67	17,277	24,577	0.70
2007	18	66	0.27	231	281	0.82	578	1,055	0.55	3,524	4,484	0.79	11,325	18,183	0.62	16,543	25,366	0.65
2008	28	64	0.44	301	365	0.82	707	1,250	0.57	4,013	4,963	0.81	12,736	19,538	0.65	18,803	27,560	0.68
2009	45	43	1.05	335	380	0.88	775	1,380	0.56	4,640	5,727	0.81	13,396	20,526	0.65	20,267	29,442	0.69
2010	28	73	0.38	285	381	0.75	735	1,539	0.48	4,199	5,507	0.76	12,174	20,820	0.58	18,607	30,145	0.62
2011	40	82	0.49	351	416	0.84	956	1,666	0.57	4,994	6,179	0.81	13,554	20,295	0.67	20,939	30,056	0.70
2012	61	89	0.69	479	616	0.78	1,237	2,310	0.54	6,116	7,524	0.81	16,758	24,417	0.69	26,145	37,039	0.71
2013	49	118	0.42	565	670	0.84	1,533	2,623	0.58	6,316	7,766	0.81	16,689	24,713	0.68	26,208	37,279	0.70
2014	77	122	0.63	623	734	0.85	1,748	2,895	0.60	7,575	8,339	0.91	17,860	24,946	0.72	29,125	38,539	0.76
2015	51	121	0.42	528	631	0.84	1,500	2,592	0.58	6,915	7,870	0.88	15,283	23,038	0.66	25,476	35,800	0.71
2016	51	114	0.45	700	705	0.99	2,671	3,761	0.71	8,076	8,880	0.91	18,392	25,356	0.73	31,619	39,784	0.79
2017	42	66	0.64	422	478	0.88	1,685	3,093	0.54	6,391	7,484	0.85	12,777	19,302	0.66	22,697	32,248	0.70
2018	32	63	0.51	435	472	0.92	1,878	3,009	0.62	6,915	7,908	0.87	12,598	18,061	0.70	23,345	31,600	0.74
2019	20	49	0.41	429	467	0.92	1,804	2,901	0.62	7,008	7,866	0.89	12,366	16,478	0.75	23,206	29,556	0.79

Note. AP = advanced placement; FMR = female-to-male ratio; F = female; M = male.

Table 5.

The Number of Top Scorers in the AP Calculus BC Exam and the FMRs Across Races.

Year	Native American			Black			Hispanic			Asian American			White			Total
Year	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR
1997	0	9	0.00	16	34	0.47	36	99	0.36	678	1,288	0.53	1,152	2,900	0.40	2,106	4,804	0.44
1998	3	11	0.27	30	47	0.64	50	118	0.42	964	1,635	0.59	1,631	4,269	0.38	3,034	6,802	0.45
1999	1	19	0.05	35	51	0.69	62	183	0.34	1,051	1,911	0.55	2,042	5,168	0.40	3,546	8,070	0.44
2000	5	24	0.21	42	59	0.71	81	213	0.38	1,275	2,190	0.58	2,516	5,724	0.44	4,174	8,733	0.48
2001	5	18	0.28	40	87	0.46	85	249	0.34	1,455	2,487	0.59	2,724	6,505	0.42	4,565	9,940	0.46
2002	5	16	0.31	57	126	0.45	137	279	0.49	1,861	2,840	0.66	3,636	7,785	0.47	6,026	11,720	0.51
2003	7	22	0.32	57	101	0.56	114	338	0.34	1,900	3,064	0.62	3,501	8,157	0.43	5,875	12,289	0.48
2004	11	21	0.52	75	108	0.69	172	364	0.47	2,158	3,175	0.68	4,059	8,346	0.49	6,857	12,735	0.54
2005	7	37	0.19	86	158	0.54	206	469	0.44	2,668	4,013	0.66	4,820	9,503	0.51	8,231	15,047	0.55
2006	16	21	0.76	112	170	0.66	205	436	0.47	2,821	4,269	0.66	4,622	9,365	0.49	8,355	15,478	0.54
2007	14	39	0.36	99	148	0.67	239	599	0.40	3,146	4,899	0.64	5,321	10,827	0.49	9,396	17,507	0.54
2008	11	35	0.31	102	193	0.53	296	660	0.45	3,292	5,161	0.64	5,687	11,571	0.49	9,925	18,636	0.53
2009	18	43	0.42	114	207	0.55	284	823	0.35	3,711	5,727	0.65	5,691	11,577	0.49	10,392	19,415	0.54
2010	20	50	0.40	192	273	0.70	430	757	0.57	4,384	6,862	0.64	6,871	14,077	0.49	12,767	23,808	0.54
2011	13	52	0.25	217	277	0.78	506	1,143	0.44	4,823	7,363	0.66	7,115	13,841	0.51	13,366	23,838	0.56
2012	25	60	0.42	223	326	0.68	593	1,433	0.41	5,701	8,541	0.67	8,451	15,800	0.53	15,871	27,782	0.57
2013	27	79	0.34	210	401	0.52	669	1,479	0.45	5,817	8,856	0.66	8,001	16,205	0.49	15,339	28,128	0.55
2014	39	71	0.55	318	477	0.67	807	1,779	0.45	6,897	10,061	0.69	9,441	17,256	0.55	18,331	30,992	0.59
2015	21	85	0.25	286	453	0.63	740	1,784	0.41	6,869	9,734	0.71	8,860	17,160	0.52	17,594	30,529	0.58
2016	29	62	0.47	312	452	0.69	1,264	2,485	0.51	7,760	10,594	0.73	10,129	18,611	0.54	20,640	34,020	0.61
2017	24	66	0.36	261	368	0.71	1,057	2,297	0.46	7,429	10,515	0.71	8,912	16,528	0.54	18,948	31,590	0.60
2018	27	52	0.52	255	372	0.69	1,026	2,434	0.42	7,401	10,898	0.68	8,350	16,035	0.52	18,352	31,955	0.57
2019	19	50	0.38	269	425	0.63	1,152	2,605	0.44	8,073	12,123	0.67	8,808	16,170	0.54	19,757	33,726	0.59

Note. AP = advanced placement; FMR = female-to-male ratio; F = female; M = male.

Table 6.

The Number of Top Scorers in the AP Statistics Exam and the FMRs Across Races.

Year	Native American			Black			Hispanic			Asian American			White			Total
Year	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR	F	M	FMR
1997	0	2	0.00	3	10	0.30	7	11	0.64	69	189	0.37	202	593	0.34	304	883	0.34
1998	0	2	0.00	5	14	0.36	12	32	0.38	149	279	0.53	418	969	0.43	659	1,430	0.46
1999	0	2	0.00	8	15	0.53	16	33	0.48	182	371	0.49	582	1,273	0.46	868	1,871	0.46
2000	2	6	0.33	9	10	0.90	18	36	0.50	243	493	0.49	701	1,543	0.45	1,044	2,219	0.47
2001	0	2	0.00	18	25	0.72	32	65	0.49	374	656	0.57	1,170	2,099	0.56	1,703	3,012	0.57
2002	2	4	0.50	21	29	0.72	33	66	0.50	548	792	0.69	1,404	2,300	0.61	2,111	3,370	0.63
2003	3	9	0.33	28	41	0.68	63	107	0.59	681	1,044	0.65	1,877	3,277	0.57	2,796	4,726	0.59
2004	6	10	0.60	29	42	0.69	52	127	0.41	631	1,126	0.56	2,035	3,689	0.55	2,904	5,253	0.55
2005	6	10	0.60	22	46	0.48	77	145	0.53	833	1,334	0.62	2,533	3,934	0.64	3,651	5,788	0.63
2006	7	17	0.41	28	61	0.46	74	181	0.41	954	1,628	0.59	2,488	4,672	0.53	3,797	7,084	0.54
2007	5	11	0.45	41	51	0.80	100	181	0.55	1,035	1,640	0.63	2,833	4,828	0.59	4,232	7,099	0.60
2008	10	11	0.91	52	54	0.96	118	271	0.44	1,347	2,156	0.62	3,258	5,641	0.58	5,025	8,534	0.59
2009	8	17	0.47	66	69	0.96	143	292	0.49	1,545	2,163	0.71	3,403	5,394	0.63	5,444	8,321	0.65
2010	7	16	0.44	60	87	0.69	192	324	0.59	1,726	2,584	0.67	3,671	6,245	0.59	6,010	9,846	0.61
2011	11	13	0.85	83	89	0.93	197	339	0.58	1,950	2,865	0.68	4,167	6,244	0.67	6,700	10,031	0.67
2012	14	20	0.70	98	99	0.99	242	400	0.61	2,274	3,046	0.75	4,617	6,504	0.71	7,641	10,628	0.72
2013	17	35	0.49	97	128	0.76	259	540	0.48	2,470	3,532	0.70	4,974	7,774	0.64	8,101	12,444	0.65
2014	21	31	0.68	146	156	0.94	308	653	0.47	3,177	4,137	0.77	5,848	9,189	0.64	9,877	14,724	0.67
2015	24	28	0.86	123	171	0.72	356	632	0.56	3,293	4,396	0.75	5,839	8,924	0.65	10,023	14,734	0.68
2016	7	44	0.16	162	170	0.95	495	957	0.52	3,645	4,533	0.80	6,594	9,528	0.69	11,519	14,470	0.80
2017	16	25	0.64	116	119	0.97	559	938	0.60	3,722	4,940	0.75	6,156	9,118	0.68	11,208	16,014	0.70
2018	11	26	0.42	152	165	0.92	579	1,133	0.51	4,109	5,550	0.74	6,374	9,724	0.66	11,996	17,778	0.67
2019	19	33	0.58	173	185	0.94	649	1,084	0.60	4,445	5,992	0.74	6,471	8,795	0.74	12,630	17,229	0.73

Note. AP = advanced placement; FMR = female-to-male ratio; F = female; M = male.

Testing Procedures

Like many other AP exams, all three of AP mathematics exams, Calculus AB, Calculus BC, and Statistics, are proctored on specific dates, usually in the first 2 weeks of May at thousands of high schools and designated exam centers across the country. Although students can select their testing locations, most students choose to take these exams at their school during regular school hours. After the designated school personnel proctor the exams, they are mailed back to the College Board centers.

All three AP mathematics exams, Calculus AB, Calculus BC, and Statistics, are similar in terms of testing procedures and structure. The exams include a multiple-choice section and a free-response section. These two sections are graded in separate forms due to their different structures. For example, the multiple-choice section is scored by computer by scanning the answer sheet. In contrast, the free-response section is scored by appointed educators, including experienced high school AP teachers and college professors during the summer at a large convention (Sundquist, 2016). The raw scores from each section are added to form a composite score, which is later translated into a 5-point scale using statistical processes to hold consistency across years. Like many other criterion-referenced tests, AP scores are not graded on a curve but instead calculated based on mastery and of the content and skills of a specific AP course (College Board, 2020a; Sundquist, 2016). Research that investigated the predictive validity of AP mathematics exams have found significant positive relationship between AP Exam performance and college course placement (Mattern et al., 2009; Patterson & Ewing, 2013). The interrater reliability statistics for different sections of Calculus AB and BC exams were found to range from high 0.70s to low 0.90s (Bridgeman et al., 1996). Course-specific testing procedures are described below.

AP Calculus AB

More than 300,000 students take the AP Calculus AB exam annually, and fewer than 20% achieve a score of 5. Students traditionally take the AP Calculus AB course after Precalculus. It is approximately the equivalent of a first-semester college calculus course, which covers topics in differential and integral calculus (College Board, 2020a). The examination lasts 3 hours and 15 minutes and consists of a section of 45 multiple-choice items and another section of six free-response questions, with each section having equal weight. A graphing calculator is permitted for some parts of the exam (College Board, 2020a). The gender and racial breakdown of participants in the AP Calculus AB exam, from 1997 to 2019, appears in Table 1. Table 4 shows the gender and racial breakdown of top achievers who scored a 5.

AP Calculus BC

Approximately 130,000 students take the AP Calculus BC exam every year, and roughly 40% of the participants can score a 5. Although successful completion of a Calculus AB course is not a prerequisite for Calculus BC course enrollment, most students take BC after completing AB, and they are not allowed to take both exams within the same year. Similar to AP Calculus AB, the examination lasts 3 hours and 15 minutes and consists of a section of 45 multiple-choice items and another section of six free-response questions, with each section having equal weight. Likewise, a graphing calculator is permitted for some parts of the exam. The gender and racial breakdown of participants in the AP Calculus BC exam, from 1997 to 2019, is given in Table 2. Table 5 shows the gender and racial breakdown of top achievers who scored a 5 out of 5 possible points.

AP Statistics

The AP Statistics program is the newest of the AP program’s math offerings. The first exam of the program was administered in 1997. Because the course readiness requires less advanced course content than do Calculus AB and BC, the number of participants in AP Statistics exams grew faster than other AP exams. In 2019, more than 200,000 students took the exam, and almost 15% of the examinees scored a 5. Similar to the other two AP math exams, students have 40 multiple-choice and six free-response questions in 3 hours, and they can use a graphing calculator for some parts of the exam. The gender and racial breakdown of participants in the AP Calculus AB exam, from 1997 to 2019, appear in Table 3. Table 6 shows the gender and racial breakdown of top achievers who scored a 5 on the exam.

Data Analysis

Operational definitions of some significant concepts covered in this study and related index calculations are listed below. The variables were defined to identify trends in gender disparity in participation and top achievement in AP mathematics exams by race. Furthermore, it was possible to examine the relationship between the trends in gender disparities in AP participation and top performance.

Disparity index calculations

Many prior studies have used disparity indexes to examine gender gaps in mathematics performances (Bahar, 2021; Ellison & Swanson, 2010; Hyde et al., 1990, 2008; Lindberg et al., 2010; Olszewski-Kubilius & Lee, 2011; J. P. Robinson & Lubienski, 2011). In this study, a simple FMR was reported as a measure of gender disparity and denoted with the FMR index. The FMR index is calculated as

FMR = \frac{no . of females}{no . of males} .

When the numbers of male and female students are equal, the value of the FMR function should be equal to 1, which indicates parity. As the number of female students exceeds that of males, the value of the function becomes larger than 1, and the disparity favors female students. As the number of male students exceeds that of females, the value of the FMR function becomes smaller than 1, and disparity favors male students (Bahar, 2021).

Index of gender disparity in participation

AP participation was represented by the number of students who participated in the corresponding AP exam. Gender disparity in exam participation is denoted with the FMR-P index and is calculated as

FMR_P = \frac{no . of females who participated in the exam}{no . of males who participated in the exam} .

Index of gender disparity in top achievement

In prior studies (Bahar, 2021; Benbow & Stanley, 1983; Makel et al., 2016; Wai et al., 2010, 2018), which investigated the gender disparities among top-achieving students, researchers examined the top performances in the extreme right tail of curves including the top 0.01%, 0.1%, 1%, and 5% (Bahar, 2021). However, different from many other norm-referenced high-stake tests, AP exams are criterion-referenced measures and students’ scores range from 1 to 5, with 5 being the highest possible score. According to AP exam score distributions published on College Board (College Board, 2020b), varying from exam to exam, 10% to 45% of the participants can make the top score. In this study, a score of 5 is the top achievement. The AP Score scale table published by College Board also recommends a score of 5 as exceptionally well qualified to receive college credit and placement, which is a college course grade equivalent of A+ or A (College Board, 2020b). Gender disparity among top achievers is denoted with the FMR-P index, and it is calculated as

FMR_TA = \frac{no . of females who scored 5}{no . of males who scored 5} .

Analysis of research questions

Upon obtaining the data from College Board’s database that includes the number of participants and top scorers, the FMRs for participation and top achievement was calculated for all three of AP exams across races for each year from 1997 to 2019. The calculated FMRs were later employed in the analyses described below to answer the research questions.

Research questions 1 and 2

A Mann–Kendall (MK) trend test was employed to explore the trends in gender disparity in participation and top achievement in AP mathematics exams across races. As a nonparametric test, MK is a powerful exploratory analysis method to identify the presence of monotonic trends in data collected over time (Kendall, 1975; Mann, 1945). The MK test has been commonly used to evaluate trends in hydrological and climatic data (Jaiswal et al., 2015), though it is not prevalent in educational research (Bahar, 2021). However, MK trend tests are useful, especially for researchers exploring the flow of changes in educational settings and outcomes. Moreover, because the MK is a non-parametric procedure, no assumption of normality is required, and the test can be used with even small data sets.

In addition to the MK trend analysis, Sen’s nonparametric procedures were used. Sen’s slope was calculated for each AP exam across races to identify the magnitude of the slope of the trend over time (Sen, 1968). Sen’s slope has been used widely to identify the slope of trend lines (Bahar, 2021), and according to Yu et al. (2002), it is a robust measure of a trend’s magnitude. Both the MK trend test and Sen’s slope were performed using R software.

Research question 3

Correlation analysis measured the relationship between the trends in gender disparity in participation and top achievement across races. Using SPSS 25, Pearson product-moment correlation coefficients among female-to-male ratios in participation (FMR-P) and female-to-male ratios in top achievement (FMR-TA) were calculated for each AP exam.

Results

Trends in Gender Disparity in AP Math Exam Participation

Tables 1 to 3 depict the number of students who participated in each AP math exam; Calculus AB, Calculus BC, and Statistics, as well as the FMRs among the participants, respectively, for years between 1997 and 2019, separated by race and gender. As seen in Table 1, the number of female and male students who took the AP Calculus AB exam increased rapidly over the years for all students across races. MK trend analyses (see Table 7) determined whether a trend was present in gender parity/disparity in participation. For the entire student population, the FMR values showed an upward trend across the years. They increased from 0.90:1 to 0.98:1, meaning that the percentage of female students who enrolled in the Calculus AB exam grew and reached parity as of 2019. The MK test indicated that the trend was statistically significant and favored females (p < .001). According to Sen’s slope analysis, the magnitude of the change in the FMR values was 0.004 per year (see Figure 1). Similar to the entire population, the FMR values for Hispanic, Asian American, and White students showed upward trends, which were statistically significant and favored female participants. However, the FMR values for Black students showed a downward trend, which was statistically significant and favored male participants. According to the MK test, no trend existed for Native American students.

Table 7.

Mann–Kendall Two-Tailed Trend Test and Sen’s Slope Values for Female to Male Ratios in Participation (FMR-P).

AP exam	Variable	MK trend test values								Sen’s slope values
AP exam	Variable	Minimum	Maximum	M	SD	K’s tau	S	Var(S)	p	Slope	SLB	SUB	IV	ILB	IUB
Calculus AB	Native American	0.71	1.03	0.87	0.08					0.00	0.00	0.01	−7.32	−17.77	9.17
	Black	1.22	1.50	1.36	0.07	−0.49	−125.00	1,433.67	.001	−0.01	−0.01	0.00	14.59	8.88	24.35
	Hispanic	0.88	1.02	0.97	0.04	0.46	117.00	1,433.67	.002	0.00	0.00	0.01	−8.10	−12.62	−2.68
	Asian American	0.96	1.05	1.01	0.03	0.45	115.00	1,433.67	.002	0.00	0.00	0.00	−4.97	−7.50	−1.36
	White	0.85	0.92	0.89	0.02	0.70	177.00	1,433.67	<.001	0.00	0.00	0.00	−5.40	−6.65	−3.72
	Total	0.89	0.98	0.93	0.03	0.76	193.00	1,433.67	<.001	0.00	0.00	0.00	−6.94	−8.30	−5.21
Calculus BC	Native American	0.44	0.70	0.57	0.07	0.32	82.00	1,432.67	.030	0.01	0.00	0.01	−10.39	−20.06	−0.08
	Black	0.83	1.06	0.95	0.06	0.18	45.00	1,433.67	.249	0.00	0.00	0.01	−4.71	−12.61	3.97
	Hispanic	0.53	0.72	0.66	0.05	0.36	91.00	1,433.67	.016	0.00	0.00	0.01	−7.13	−14.63	−0.13
	Asian American	0.69	0.82	0.77	0.03	0.44	111.00	1,433.67	.003	0.00	0.00	0.00	−5.52	−8.83	−1.63
	White	0.54	0.69	0.63	0.04	0.79	201.00	1,433.67	<.001	0.01	0.00	0.01	−10.60	−12.58	−8.20
	Total	0.59	0.74	0.68	0.04	0.79	199.00	1,433.67	<.001	0.01	0.00	0.01	−10.79	−12.97	−8.49
Statistics	Native American	0.79	1.25	1.00	0.12	0.26	67.00	1,433.67	.082	0.01	0.00	0.01	−15.00	−28.52	2.66
	Black	1.41	1.67	1.52	0.07	−0.01	−3.00	1,433.67	.958	0.00	−0.01	0.00	2.51	−7.66	14.22
	Hispanic	1.04	1.27	1.15	0.06	0.70	177.00	1,433.67	<.001	0.01	0.01	0.01	−15.13	−19.36	−10.41
	Asian American	0.76	1.03	0.93	0.07	0.60	153.00	1,433.67	<.001	0.01	0.00	0.01	−13.31	−20.50	−7.05
	White	0.79	1.08	1.00	0.07	0.82	207.00	1,433.67	<.001	0.01	0.01	0.01	−11.85	−17.15	−9.45
	Total	0.81	1.12	1.02	0.07	0.91	231.00	1,433.67	<.001	0.01	0.01	0.01	−15.27	−18.25	−12.03

Note. FMR-P = female-to-male ratios in participation; AP = advanced placement; M = mean; SD = standard deviation; K’s Tau = Kendall’s Tau; SLB = slope lower bound (95%); SUB = slope upper bound (95%); IV = intercept value; ILB = intercept lower bound (95%); IUP = intercept upper bound (95%).

Figure 1.

Trends in participation in AP exams.

Table 2 depicts the number of female and male students who participated in the AP Calculus BC exam. The number of participants grew almost six times, exceeding 120,000 as of 2019. From 1997 to 2019, female students taking the exam increased faster than it did for males; accordingly, the FMRs moved from 0.59:1 to 0.72:1 (see Table 7). The MK test indicated that the FMR trend was upward and statistically significant and favored females (p < .001). Sen’s slope analysis indicated a 0.004-unit change in the magnitude of the FMR trend per year (see Figure 1). In addition, the FMR values for all races showed upward trends and favored female participants. According to the MK test, the trends were statistically significant, except for the Black population (p = .249).

Similar to AP Calculus AB and BC exams, the number of female and male students who participated in the AP Statistics exam increased for all races from 1997 to 2019 (see Table 3). Among all exam participants, the FMR values increased from 0.81:1 to 1.11:1 during the stated period. The MK test indicated that the FMR trend was upward, statistically significant, and favored females (p = .000). Thus, the percentage of female students who enrolled in the AP Statistics exam increased rapidly and surpassed that of male students. According to Sen’s slope analysis, the magnitude of the change in the FMR trend was 0.008 per year (see Figure 1). Likewise, the FMR values for Hispanic, Asian American, and White students showed upward trends that were statistically significant and favored female participants. The FMR trend for the Native American population increased, favoring females, though not significantly (p = .082). Different from other groups, the FMR trend for Black students was downward and favored male participants; however, it was not statistically significant (p = .958).

Trends in Gender Disparity Among Top Achievers on AP Math Exams

Tables 4 to 6 depict the number of students who scored highest (a 5) on each AP math exam; Calculus AB, Calculus BC, and Statistics, as well as the FMRs among the top scorers, respectively, for years between 1997 and 2019, separated by race and gender. As seen in Table 4, the number of female and male students who scored a five on the AP Calculus AB exam increased for all students from 1997 to 2017. MK trend analyses (see Table 8) determined whether a trend was present in gender parity/disparity among top scorers. For the entire student population who scored a 5, the FMR values showed an upward trend over the years and increased from 0.59:1 to 0.79:1. The MK test indicated that the trend was statistically significant and favored females (p < .001). Sen’s slope analysis indicated a 0.008-unit change in the magnitude of the FMR trend per year (see Figure 2). In addition, the FMR values for all races showed upward trends that favored top-scoring females. The trends were also statistically significant, except for the Native American students (p = .208).

Table 8.

Mann–Kendall Two-Tailed Trend Test and Sen’s Slope Values for Female to Male Ratios Among Top Scorers (FMR-TA).

AP exam	Variable	MK trend test values								Sen’s slope values
AP exam	Variable	Minimum	Maximum	M	SD	K’s tau	S	Var(S)	p	Slope	SLB	SUB	IV	ILB	IUB
Calculus AB	Native American	0.27	1.05	0.49	0.17	0.19	49.00	1,433.67	.208	0.01	0.00	0.02	−11.59	−30.78	7.36
	Black	0.61	0.99	0.83	0.08	0.33	83.00	1,433.67	.030	0.01	0.00	0.01	−9.95	−21.80	0.29
	Hispanic	0.30	0.71	0.54	0.09	0.54	137.00	1,433.67	.000	0.01	0.01	0.01	−15.14	−23.13	−9.52
	Asian American	0.68	0.91	0.80	0.07	0.67	169.00	1,433.67	<.001	0.01	0.01	0.01	−15.77	−20.51	−10.76
	White	0.52	0.75	0.64	0.06	0.68	171.00	1,433.67	<.001	0.01	0.01	0.01	−15.98	−20.71	−10.38
	Total	0.55	0.79	0.67	0.07	0.75	191.00	1,433.67	<.001	0.01	0.01	0.01	−16.38	−21.87	−11.04
Calculus BC	Native American	0.00	0.76	0.35	0.16	0.42	105.00	1,433.67	.005	0.01	0.00	0.02	−23.90	−38.18	−8.91
	Black	0.45	0.78	0.62	0.09	0.20	51.00	1,433.67	.190	0.00	0.00	0.01	−7.45	−22.62	3.37
	Hispanic	0.34	0.57	0.43	0.06	0.26	67.00	1,433.67	.082	0.00	0.00	0.01	−5.27	−12.81	1.58
	Asian American	0.53	0.73	0.65	0.05	0.62	157.00	1,433.67	<.001	0.01	0.00	0.01	−11.37	−15.31	−6.43
	White	0.38	0.55	0.48	0.05	0.76	193.00	1,433.67	<.001	0.01	0.00	0.01	−12.79	−16.27	−8.68
	Total	0.44	0.61	0.53	0.05	0.75	191.00	1,433.67	<.001	0.01	0.01	0.01	−13.20	−17.01	−10.00
Statistics	Native American	0.00	0.91	0.45	0.28	0.41	101.00	1,423.00	.007	0.02	0.01	0.04	−44.26	−81.36	−12.46
	Black	0.30	0.99	0.76	0.21	0.46	117.00	1,433.67	.002	0.02	0.01	0.03	−40.21	−64.53	−13.36
	Hispanic	0.38	0.64	0.52	0.07	0.23	58.00	1,432.67	.125	0.00	0.00	0.01	−9.25	−16.46	2.81
	Asian American	0.37	0.80	0.65	0.11	0.72	183.00	1,433.67	<.001	0.01	0.01	0.02	−25.76	−33.66	−20.38
	White	0.34	0.74	0.59	0.10	0.72	183.00	1,433.67	<.001	0.01	0.01	0.01	−23.81	−29.53	−16.39
	Total	0.34	0.80	0.61	0.10	0.76	193.00	1,433.67	<.001	0.01	0.01	0.02	−25.44	−31.00	−18.09

Note. AP = advanced placement; FMR-TA = female-to-male ratio in top achievement; M = mean; SD = standard deviation; K’s Tau = Kendall’s Tau; SLB = slope lower bound (95%); SUB = slope upper bound (95%); IV = intercept value; ILB = intercept lower bound (95%); IUP = intercept upper bound (95%).

Figure 2.

Trends in top achievement in AP mathematics exams.

Table 5 presents the number of female and male students who scored a five on the AP Calculus BC exam between 1997 and 2019 over the years. The MK test indicated that the FMR trend in the entire top scorers was upward and statistically significant (p < .001). From 1997 to 2019, the percentage of female students continued to increase, and the FMR value moved from 0.44 to 0.59. Likewise, the FMR values for each race showed upward trends that favored female participants. According to Sen’s slope analysis, the magnitude of the change in the FMR trend was 0.007 per year (see Figure 2). The trends were also statistically significant, except for Hispanic (p = .190) and Black (p = .082) students (see Table 8).

Table 6 depicts the number of top-scoring students and the FMRs on the AP Statistics exams. Similar to Calculus AB and BC, the number of top scorers increased rapidly for all races on the AP Statistics exam. The male-to-female ratio among top scorers continued to grow over the years, increasing from 0.34 to 0.73. The MK test indicated that the FMR trend was upward and statistically significant for all populations (p < .001). Sen’s slope analysis indicated a 0.013-unit change in the magnitude of the FMR trend per year (see Figure 2). Likewise, the FMR values for each racial group showed upward trends that favored female participants. As seen in Table 8, the trends were also statistically significant except for Hispanic students (p = .125).

Relation of Trends in Participation and Top Achievement

Basic descriptive statistics and Pearson product–moment correlation coefficients that show the relationship between the FMR-P and FMR-TA across races are presented in Table 9. As seen in the table, the results of Pearson correlation analysis indicated that there was a significant positive association between the FMRs in participation and top achievement on the AP Calculus AB exam for the entire population (r = .83, p < .001). The correlation coefficients were also significantly positive for Hispanic (r = .69, p < .001), White (r = .77, p < .001), and Asian American (r = .72, p < .001) students. However, there were no associations between the FMRs in participation and top achievement on AP Calculus AB exam for Native American (r = .08, p = .696) and Black (r = −.001, p = .995) students.

Table 9.

Correlation Values Between FMR-P and FMR-TA Across Races for Each AP Exams.

Exam	Group	Pearson correlation	Sig. (2-tailed)	Sum of squares cross-products	Covariance
AP Calculus AB	Native American	.086	.696	.028	.001
	Black	−.001	.995	.000	.000
	Hispanic	.693	.000	.057	.003
	Asian American	.717	.000	.029	.001
	White	.769	.000	.024	.001
	Total	.834	.000	.034	.002
AP Calculus BC	Native American	.219	.315	.059	.003
	Black	.241	.267	.030	.001
	Hispanic	.511	.013	.034	.002
	Asian American	.901	.000	.034	.002
	White	.920	.000	.042	.002
	Total	.914	.000	.042	.002
AP Statistic	Native American	.243	.264	.171	.008
	Black	−.037	.868	−.012	−.001
	Hispanic	.261	.228	.027	.001
	Asian American	.876	.000	.136	.006
	White	.885	.000	.127	.006
	Total	.906	.000	.144	.007

Note. FMR-P = female-to-male ratio in exam participation; FMR-TA = female-to-male ratio in top achievement; AP = advanced placement.

The correlation analysis for the AP Calculus BC exam showed a significant positive association between the FMRs in participation and top achievement for the entire population (r = .91, p < .001). The correlation coefficients were also significantly positive for Hispanic (r = .51, p = .013), White (r = .92, p < .001), and Asian American (r = .90, p < .001) students. Similar to Calculus AB exam, the findings indicated no associations between the FMRs in participation and top achievement on Calculus BC exam for Native American (r = .22, p = .315) and Black (r = .24, p = .267) students (see Table 9). Finally, results of Pearson correlation analysis for AP Statistics exam indicated that there was a significant positive association between the FMRs in participation and top achievement for the entire population (r = .91, p < .001). The correlation coefficients were also significantly positive for White (r = .89, p < .001) and Asian American (r = .88, p < .001) students. However, the findings indicated no associations between the FMRs in participation and top achievement on AP Statistics exam for Native American (r = .24, p = .264), Black (r = −.04, p = .868), and Hispanic (r = .26, p < .228) students.

Discussion

Trends in Gender Disparity in AP Math Exam Participation

One of the goals of this study was to explore the trends in gender parity participation in AP mathematics exams across races. In general, the findings indicated that females’ representation in all three AP exams increased significantly from 1997 to 2019, as the MK tests and Sen’s slope indicated significant and upward trends for all exams that favored females.

The findings related to the AP Calculus AB exam showed that females reached parity in participation in the entire population over the years, as the FMR among all participants increased from 0.90:1 to 0.98:1 between 1997 and 2019. Furthermore, if the trend continues, Sen’s slope analysis indicates that males most likely will be minority in exam classrooms year by year. When the data are segregated by race, significant findings appeared as well. Similar to the entire population, the FMRs in participation for Hispanic and Asian American students was already in parity. Furthermore, the significant and upward trend that favored females might emerge as a sign for Hispanic and Asian American males’ underrepresentation in AP Calculus AB exams over the years. For White students, although the FMR trend was upward and close to parity, Sen’s slope shows that it might take a few more decades for females to reach parity with male students. Among all AP Calculus AB exam participants, Black males were the only underrepresented male group among all participants. As seen in Table 7, the mean Black FMR over the years was 1.36:1, while the next highest FMR that belonged to Asian American students was only 1.00:1. Although the FMR trend for Black students was significantly downward and favored male students, the Sen’s slope magnitude showed that it might take several decades to reach parity between male and female participants. For the Native American students, the mean FMR value was close to parity, but the MK test failed to identify a trend for the population, which might be a result of large deviations and outliers in the FMR data that were led by a small number of Native American participants over the years.

Similar to the Calculus AB exam, the findings related to participation in the AP Calculus BC exam showed that the FMR trend was significantly upward and favored females in the entire population over the years. The MK trend analysis showed that the FMR among Calculus BC participants increased from 0.59:1 to 0.72:1 between 1997 and 2019. However, different from the Calculus AB exam, the female participants were largely underrepresented in the Calculus BC exam (see Table 2). According to Sen’s slope analysis, reaching parity in the number of female and male participants would take up to a half-century if the trend continues. When the data are segregated by race, the FMR trend in participation for all races was similar to the trend in the entire population, except Black students. The FMR values were in parity for Black students, while females from Hispanic, White, Asian American, and Native American students were widely underrepresented among the Calculus BC exam participants. Reaching parity in the number of female and male participants might be expected to take several decades for each race, other than Black students, as the MK test indicated a stagnant FMR trend for Black students in AP Calculus BC participation. Namely, without implementing any further culturally and gender-responsive policies and practices, the present trends suggest that closing the gaps in AP Calculus BC participation would not be happening soon.

The AP Statistics exam was the only AP math exam for which female participants were overrepresented. When the first AP Statistics exam became available in 1997, the FMR was 0.81:1 among the participants. However, the FMR value continued to increase and finally reached 1:11:1 as of 2019. Furthermore, the MK test analysis showed that the FMR trend was upward, which means the gap keeps widening in favor of females. For all racial groups, the female participants were in parity with males or majority among the participants. For Black and Native American students, the trends were stagnant and increased for Asian American, Hispanic, and White students. Taken together, these findings mean that, except Native American students, the male students will remain unrepresented, and their share will decline year by year among the AP Statistics exam participants.

Different from the studies (Stumpf & Stanley, 1996) that found significant disparities among AP Calculus AB and BC participants, in favor of male students, the current findings indicate parity in AP Calculus AB and even overrepresentation of females in AP Statistics exams. These results, in general, might explain the increase in females’ interest in AP math exams, and such an increase in interest might positively influence parity among top achievement in these exams. Besides, although the findings indicated that females are still underrepresented among AP Calculus BC exam participants, their share among the participants increased significantly over the years.

Given that mathematics and related fields have long been identified as masculine disciplines (Fennema & Sherman, 1977), the increase in female students’ interests in all AP math exams over the years is welcome. However, underrepresentation of males in recent AP Statistics exams and future AP Calculus AB exams could also be considered precarious. The findings indicate that males from certain races, specifically Black males, are largely underrepresented in AP Statistics and AP Calculus AB exam rooms. Although investigating the factors that contributed to the Black male students’ participation in advanced mathematics programs and courses is beyond the scope of this study, it is crucial to identify and remove the barriers that prevent Black male students from participating in advanced mathematics courses. As the decision or ability to enroll in any AP math course usually was determined by the level of mathematics that students take in Grade 8 (Finkelstein et al., 2012; Lubienski et al., 2004; Stevenson et al., 1994), Black males’ underrepresentation in AP courses might result from their underrepresentation in advanced middle school math courses and gifted programs. This assumption might hold as many prior studies already identified Black boys underrepresented in elementary and middle school gifted programs (Berry, 2008; Davis, 2014; Wright & Counsell, 2018). Policies and research-based practices that increase Black students’ representation in advanced elementary and middle school math courses and gifted programs would also support Black male students’ participation in advanced high school mathematics courses.

Trends in Gender Disparity Among Top Achievers on AP Math Exams

Another primary goal of this study was to explore trends by gender among top achievers on AP mathematics exams. Disparities in top achievement can account for subsequent disparities in STEM achievement and outcomes. These disparities include earning higher degrees, publications, and patents (Makel et al., 2016; Wai et al., 2010), the findings of this study might partially shed light on predicting future gender disparities in STEM fields.

The findings related to the AP Calculus AB exam indicated that females were underrepresented among the students who scored a 5. As of 2019, roughly 43% of the top scorers in the Calculus AB exam were female students. However, the findings related to the AP Calculus AB exam indicated that the share of females among the students who scored a 5 increased significantly over the years. The MK trend analysis showed that the FMR among top scorers increased significantly from 0.59:1 to 0.79:1 over the years. However, according to Sen’s slope analysis, reaching parity in the number of female and male top scorers would take at least three decades if the trends remained similar.

When the data are segregated by race, the trend in FMR among the top-scoring students for all races was upward, significant, and favored females in the entire population, except Native American students as the MK test indicated a stagnant FMR trend. The FMR values were closest to parity for Black and Asian American students with a mean FMR value of 0.83:1 and 0.80:1, respectively. FMR ratios among top achievers were lowest for Native American and Hispanic students (0.49:1 and 0.54:1, respectively). As the trend in FMR was also stagnant for the Native American students, the small number of representations among the top achievers is a significant disparity issue for female Native American students.

The findings related to top achievement in AP Calculus BC showed that the female students were widely underrepresented among the students who scored a 5. Despite vast differences in the numbers of female and male top scorers, FMR value trends indicated the gap was diminishing in favor of females between 1997 and 2019. The MK trend analysis showed that the FMR among top scorers on the Calculus BC exam increased significantly from 0.44:1 to 0.59:1 over the years. Likewise, when the data are segregated by race, the trend in FMR among the top-scoring students for all races was upward, significant, and favored females in the entire population, except Black and Hispanic students; the MK test indicated stagnant FMR trends for both groups. The FMR values for Black and Asian students were the highest among all racial groups with a mean FMR value of 0.62:1 and 0.65:1, respectively, while the lowest FMRs were for Native American and Hispanic students (0.34:1 and 0.42:1, respectively).

AP Statistics exam was another AP math exam that female participants were widely underrepresented among the students who scored a 5. However, similar to the other two exams, the trend in FMR values among the students who scored a 5 was significantly upward and favored females. The MK trend analysis showed that the FMR among top scorers increased from 0.34:1 to 0.73:1 over the years. Compared to the other two math exams, the magnitude of the change in the FMR values was more significant. As Sen’s slope analysis indicated in Table 8, females could reach parity among top achievers in AP Statistics in less than two decades. However, this rapid change was not consistent for Hispanic students as the MK test indicated a stagnant FMR trend for the group.

In general, different from the findings that indicated a high representation of female students in participation in all AP math exams, especially in Calculus AB and Statistics, the findings showed lower female representation in top achievement for all exams. This finding has been consistent with prior studies that have documented higher male-to-female ratios in the highest score ranges (Andreescu et al., 2008; Benbow & Stanley, 1980, 1983; Desjarlais, 2009; Ellison & Swanson, 2010; Hedges & Nowell, 1995; Hyde et al., 1990; Linn, 2010; Makel et al., 2016; Niederle & Vesterlund, 2010; Olszewski-Kubilius & Lee, 2011; Reis & Park, 2001; Wai et al., 2010, 2018; Xie & Shauman, 2003).

Relation of Trends in Participation and Top Achievement

As seen in Table 9, the results of Pearson correlation analysis indicated that a significant positive correlation exists between the FMRs in participation and top achievement for all AP math exams for the entire population. The author interprets this finding as a strong association between participation and performance, which means when we achieve to lower the gender disparities in participation, we might get closer to increasing parities among top-achieving female and male students. Although reaching parity in participation positively benefits parity in top achievement, a 1:1 FMR value in participation does not necessarily lead to a 1:1 FMR in top achievement. For example, the FMR value in participation in AP Statistics was 1.11:1 for the entire population in 2019 while it was 0.73:1 among top achievers in the same year. This situation also holds for most mathematics exams and tests, including the ACT/SAT, AP Calculus test, Graduate Record Exam (GRE), American Mathematics Competitions (AMC), and National Assessment of Educational Progress (NAEP). Prior studies report disparities among high achievers favoring male students despite parity in participation (Niederle & Vesterlund, 2010).

When the data are segregated by race, while the relationship between FMRs in participation and top achievement was significantly positive for Asian American, Hispanic, and White students for all exams, the findings indicated no associations between the FMRs in participation and top achievement for Native American and Black students in any of the exams. This result is most probably due to the small number of top scorers from Native American and Black students that led to significant variations in the mean FMR values. Although this study only focused on gender disparities as disaggregated by race, analyzing the racial disparities was beyond the goal of this study. However, the findings showed significant racial disparities not only in participation in the exams but also in top achievement between racial groups. For example, only 694 Black students scored a 5 on the AP Calculus BC exam in 2019 while the number of Asian American students was more than 20,000. Thus, Asian American students were overrepresented almost 100 times more than Black students among the top scorers when considering their share in the total population. Significant racial disparities in precollege achievement are already perilous. Disparities will also lead to enormous differences in later outcomes and achievements; future research should explore the trend in racial disparities in AP math exams as an important area.

Among all three AP math exams, Calculus BC is a platform for which gender disparities in both participation and top achievement are most apparent. This finding is consistent with the findings of Amelink (2009), who identified the AP Calculus BC as the exam in which females were underrepresented most. Generally, students enroll in AP Calculus BC courses after successful completion of AP Calculus AB exam. In other words, most of the participants of Calculus BC are the students who already scored a 3, 4, or 5 on the Calculus AB exam, which is also evident when the mean FMR value (0.678:1) in top achievement for AP Calculus AB participation is compared with the mean FMR (0.671:1) in participation for AP Calculus BC. This fact may imply that the smaller number of females appear in the more advanced level of mathematics, which is consistent with the findings of Bahar (2021), who found higher proportions of males in the higher levels of the AMC competitions.

This study presented encouraging findings in terms of reaching parity between females and males. However, there is no evidence as to whether these advancements are due to effective policies and practices that encourage females and males for STEM fields or some unobtrusive factors such as ceiling effect of the exams. Especially given that 43% of the entire Calculus BC participants scored a 5 in 2019, this suspicion is not rootless. Although the present findings did not test the significance of the change over the years, the number of students who scored five on the Calculus BC exam was only 30% in 1997. As AP exams are designed as criterion-referenced tests, it is not possible to investigate disparities beyond a score of five. However, we could have expected to see lower FMRs if we had the chance to move through the upper percentages in the achievement distribution of the AP Calculus BC exam. As prior studies that examined gender disparities in other high-stake tests, such as the SAT, ACT, GRE, and AMC, found larger underrepresentation of females in top achievement levels, this assumption might also hold but needs validation through further investigations.

The number of female and male participants stopped increasing for all AP math exams recently. For example, as depicted in Table 1, the number of female and male participants in Calculus AB started to decline as of 2017 for the entire population. This picture is almost similar to AP Calculus BC and Statistics (see Tables 2 and 3). Because this study focused on the trend analysis in gender disparities rather than general trends in participation in AP math exams, analyzing the factors that contribute to the decline in the number of the participants is beyond the scope of this study. However, a further investigation would be needed to understand the reasons leading to the decline in the interest in AP math exams. More specifically, if such a decline stems from a decline in students’ interest in STEM fields or subjects, then this might be a significant threat to future STEM labor outcomes. Perhaps such an exploration might also reveal whether the AP program is currently at a breakout point that just started to lose its popularity or reached a saturation point in terms of student participation. Investigation of the matter would be an additional area for future research.

A limitation of this study was that, every year, a substantial number of participants either chose not to report their racial background or belonged to two or more than two races. Although these students were included in the total number of participants, they were not represented under any racial groups. For example, in 2019, the total number of students who chose not to report their racial background or belonged to two or more than two races exceeded 30,000 in total for all three AP math exams. The number of students who participated in one of these exams exceeded 600,000; the percentage of such a student population constituted only 5% of the entire population. However, because Native American and Black students are represented with a smaller size of participants and top achievers on these exams, such unreported numbers might be impacting their trends more than that were of overrepresented groups such as White and Asian American students. As no further information about these participants was accessible, it is reasonable to assume that the number of students who chose not to report their racial background or belonged to two or more than two races was represented equally weighted in all racial groups. However, this assumption needs further validation.

Conclusion

This study investigated the trends in gender disparities among participants and top achievers in AP math exams. Although the findings indicated that the FMRs in participation in the AP math exams increased significantly from 1997 to 2019 and favored females, the gender disparities among top achievers for all math exams are still substantial. Furthermore, the MK test and Sen’s slope analysis indicated that reaching parity in high achievement may take up to 60 years for the Calculus BC exam and roughly 30 years for Calculus AB and Statistics exams. Monitoring these differences through robust and scientifically sound analysis tools is crucial because the lack of parity in participation in and top achievement at advanced academics will not only lead to later gender disparities in STEM outcomes and achievements (Eccles, 2007; Makel et al., 2016; Wai et al., 2010), but they also threaten social justice initiatives and capitalizing on human potential (Darling-Hammond, 2010; Ma & Johnson, 2008; Prong, 2018; Sells, 1980; Shaw & Barbuti, 2010; Trusty, 2002).

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research and/or authorship of this article.

ORCID iD

A. Kadir Bahar

Author Biography

A. Kadir Bahar, PhD, is an assistant professor in the Department of Educational Psychology at the University of Georgia. Dr. Bahar serves as the associate editor for Gifted and Talented International, and he is on the editorial board of the Gifted Child Quarterly and Gifted Education International journals. His research interests include mathematical talent, problem solving, and excellence gaps in STEM. Dr. Bahar is the recipient of numerous awards including Math Hero of the Year Award and Edith May Sliffe - Distinguished Teaching Award.

References

Ackerman

P. L.

Kanfer

Calderwood

(2013). High school advanced placement and student performance in college: STEM majors, non-STEM majors, and gender differences. Teachers College Record, 115(10), 1–43.

Adelman

(1999). Answers in the tool box: Academic intensity, attendance patterns, and bachelor’s degree attainment. http://www.ed.gov/pubs/Toolbox/toolbox.html

Amelink

C. T.

(2009). Literature overview: Gender differences in math performance. https://www.researchgate.net/publication/242515470_Literature_Overview_Gender_Differences_in_Math_Performance/related

Andreescu

Gallian

Kane

J. M.

Mertz

J. E.

(2008). Cross-cultural analysis of students with exceptional talent in mathematical problem solving. Notices of the American Mathematical Society, 55(10), 1248–1260.

Bahar

A. K.

(2013). The influence of cognitive abilities on mathematical problem-solving performance [Doctoral dissertation, University of Arizona]. http://arizona.openrepository.com/arizona/handle/10150/293594

Bahar

A. K.

(2021). Trends in gender disparities among high-achieving students in mathematics: An analysis of the American Mathematics Competition (AMC). Gifted Child Quarterly, 65(2), 167–184. https://doi.org/10.1177/0016986220960453

Bahar

A. K.

Adiguzel

(2016). Analysis of factors influencing interest in STEM career: Comparison between high ability and motivated American and Turkish high school students. Journal of STEM Education, 17, 64–69.

Bahar

A. K.

Maker

C. J

. (2015). Cognitive backgrounds of problem solving: A comparison of open-ended vs. closed mathematics problems. Eurasia Journal of Mathematics, Science & Technology Education, 11(6), 1531–1546. https://doi.org/10.12973/eurasia.2015.1410a

Bahar

A. K.

Maker

C. J.

(2020). Culturally responsive assessments of mathematical skills and abilities: Development, field testing, and implementation. Journal of Advanced Academics, 31(3), 211–233. https://doi.org/10.1177/1932202X20906130

10.

Baird

K. L.

(2019). Gender justice and the health care system. Routledge.

11.

Baird

Keene

(2019). Closing the gender gap in math confidence: Gender and race/ethnic similarities and differences. International Journal of Gender, Science, and Technology, 10(3), 378–410.

12.

Benbow

C. P.

Stanley

J. C.

(1980). Sex differences in mathematical ability: Facts or Artifact? Science, 210(12), 1262–1264.

13.

Benbow

C. P.

Stanley

J. C.

(1983). Sex differences in mathematical reasoning ability: More facts. Science, 222(4627), 1029–1031. https://doi.org/10.1126/science.6648516

14.

Berry

R. Q.

III . (2008). Access to upper-level mathematics: The stories of successful African American middle school boys. Journal for Research in Mathematics Education, 39, 464–488.

15.

Bloom

B. S.

(1985). Developing talent in young people. Ballantine Books.

16.

Bridgeman

Morgan

Wang

(1996). Reliability of advanced placement examinations [ETS research report series]. https://www.ets.org/Media/Research/pdf/RR-96-03.pdf

17.

Brookhart

S. M.

(2009). Assessment, gender and in/equity. In Wyatt-Smith

Cumming

(Eds.), Educational assessment in the 21st century: Connecting theory and practice (pp. 119–136). Springer.

18.

Buck

Kostin

Morgan

(2002). Examining the relationship of content to gender-based performance differences in Advanced Placement exams (2002-02 ETS RR-02-25). College Board.

19.

Callahan

C. M.

(2003). Advanced Placement and International Baccalaureate programs for talented students in American high schools: A focus on science and math (RM03176). The National Research Center on the Gifted and Talented, University of Connecticut.

20.

Callahan

C. M.

(2006). Secondary program models and the evaluation of secondary programs. In Dixon

F. A.

Moon

S. M.

(Eds.), The handbook of secondary gifted education (pp. 505–526). Prufrock Press.

21.

Callahan

C. M.

Cunningham

C. M.

Plucker

J. A.

(1994). Foundations for the future: The socioemotional development of gifted, adolescent women. Roeper Review, 17, 99–105.

22.

Catsambis

(1994). The path to math: Gender and racial-ethnic differences in mathematics participation from middle school to high school. Sociology of Education, 67(3), 199–215. https://doi.org/10.2307/2112791

23.

Ceci

S. J.

Williams

W. M.

(2007). Are we moving closer and closer apart? Shared evidence leads to conflicting views. In Ceci

S. J.

Williams

W. M.

(Eds.), Why aren’t more women in science: Top researchers debate the evidence (pp. 213–236). American Psychological Association. https://doi.org/10.1037/11546-004

24.

Ceci

S. J.

Williams

W. M.

(2010). The mathematics of sex: How biology and society conspire to limit talented women and girls. Oxford University Press.

25.

Christiansen

(2009). The impact of Advanced Placement (AP) participation and success on school-wide student achievement [Doctoral dissertation, University of Central Florida]. https://stars.library.ucf.edu/cgi/viewcontent.cgi?article=4874&context=etd

26.

College Board. (2019). Course and exam description: AP Calculus AB and AP Calculus BC, including the curriculum framework. https://apstudents.collegeboard.org/ap/pdf/ap-calculus-ab-bc-course-and-exam-description.pdf

27.

College Board. (2020a). AP exams overview. https://apstudents.collegeboard.org/about-ap-exams

28.

College Board. (2020b). AP score distributions. https://apstudents.collegeboard.org/about-ap-scores/score-distributions

29.

College Board. (2020c). National reports. https://research.collegeboard.org/programs/ap/data/participation/ap-2020

30.

Corbett

Hill

(2015). Solving the equation: The variables for women’s success in engineering and computing. American Association of University Women.

31.

Darling-Hammond

(2010). Teacher education and the American future. Journal of Teacher Education, 61(1–2), 35–47.

32.

Davis

(2014). The mathematical experiences of Black males in a predominantly Black urban middle school and community. International Journal of Education in Mathematics, Science and Technology, 2, 206–222.

33.

Demaree

R. D.

(2016). A quantitative analysis of the association between advanced placement access and equity at high schools in a Mid-Atlantic state [Doctoral dissertation, George Washington University]. https://media.proquest.com/media/hms/ORIG/2/fYneI?_s=JBPev14pvbgaqGTJXKfW%2BMi1q%2BA%3D

34.

Desjarlais

M. A.

(2009). Gender Differences on the American Mathematics Competition AMC 8 Contest (Doctoral dissertation). http://digitalcommons.unl.edu/cehsdiss/39

35.

Dweck

C. S.

(2007). Is math a gift? Beliefs that put females at risk. In Ceci

S. J.

Williams

W. M.

(Eds.), Why aren’t more women in science: Top researchers debate the evidence (pp. 47–55). American Psychological Association. https://doi.org/10.1037/11546-004

36.

Eccles

J. S.

(2007). Where are all the women? Gender differences in participation in physical science and engineering. In Ceci

S. J.

Williams

W. M.

(Eds.), Why aren’t more women in science? Top researchers debate the evidence (pp. 199–210). American Psychological Association.

37.

Ellison

Swanson

(2010). The gender gap in secondary school mathematics at high achievement levels: Evidence from the American mathematics competitions. The Journal of Economic Perspectives, 24(2), 109–128.

38.

Feldhusen

J. F.

(1995). Talent development during the high school years. Gifted Education International, 10(2), 60–64.

39.

Fennema

Peterson

P. L.

Carpenter

T. P.

(1990). Teachers’ attributions and beliefs about girls, boys, and mathematics. Educational Studies in Mathematics, 21(1), 55–69. https://doi.org/10.1007/BF00311015

40.

Fennema

Sherman

(1977). Sex-related differences in mathematics achievement, spatial visualization, and affective factors. American Educational Research Journal, 14(1), 51–71.

41.

Finkelstein

Fong

Tiffany-Morales

Shields

Huang

(2012). College bound in middle school & high school? How math course sequences matter. Center for the Future of Teaching and Learning at WestEd.

42.

Halpern

D. F.

(2007). Science, sex, and good sense: Why women are underrepresented in some areas of science and math. In Ceci

S. J.

Williams

W. M.

(Eds.), Why aren’t more women in science: Top researchers debate the evidence (pp. 121–130). American Psychological Association. https://doi.org/10.1037/11546-010

43.

Hargrove

Godin

Dodd

(2008). College outcomes comparisons by AP and non-AP high school experiences (College Board Research Report No. 2008-3). College Board.

44.

Hedges

L. V.

Nowell

(1995). Sex differences in mental test scores, variability, and numbers of high-scoring individuals. Science, 269, 41–45. https://doi.org/10.1126/science.7604277

45.

Hertberg-Davis

Callahan

(2008). A narrow escape: Gifted students’ perceptions of Advanced Placement and International Baccalaureate programs. Gifted Child Quarterly, 52, 199–216.

46.

Hertberg-Davis

Callahan

C. M.

Kyburg

R. M.

(2006). Advanced Placement and International Baccalaureate programs: A “fit” for gifted learners? (RM06222). University of Connecticut, National Research Center on the Gifted and Talented.

47.

Hyde

J. S.

(2005). The gender similarities hypothesis. American Psychologist, 60, 581–592.

48.

Hyde

J. S.

(2007). Women in science: Gender similarities in abilities and sociocultural forces. In Ceci

S. J.

Williams

W. M.

(Eds.), Why aren’t more women in science? Top researchers debate the evidence (pp. 131–145). American Psychological Association. https://doi.org/10.1037/11546-011

49.

Hyde

J. S.

Fennema

Lamon

S. J.

(1990). Gender difference in mathematical performance: A metaanalysis. Psychological Bulletin, 107(2), 139–155.

50.

Hyde

J. S.

Lindberg

S. M.

Linn

M. C.

Ellis

A. B.

Williams

C. C.

(2008). Gender similarities characterize math performance. Science, 321(5888), 494–495.

51.

Hyde

J. S.

Linn

M. C.

(2006). Gender similarities in mathematics and science. Science, 314, 599–600.

52.

Jaiswal

Lohani

A. K.

Tiwari

H. L.

(2015). Statistical analysis for change detection and trend assessment in climatological parameters. Environmental Processes, 2, 729–749. https://doi.org/10.1007/s40710-015

53.

Kendall

M. G.

(1975). Rank correlation methods. Griffin.

54.

Kettler

Hurst

L. T.

(2017). Advanced academic participation: A longitudinal analysis of ethnicity gaps in suburban schools. Journal for the Education of the Gifted, 40(1), 3–19. https://doi.org/10.1177/0162353216686217

55.

Kokkelenberg

E. C.

Sinha

(2010). Who succeeds in STEM studies? An analysis of Binghamton University undergraduate students. Economics of Education Review, 29(6), 935–946.

56.

Lazear

E. P.

Rosen

(1990). Male-female wage differentials in job ladders. Journal of Labor Economics, 8(1), 106–123.

57.

Lindberg

S. M.

Hyde

J. S.

Petersen

J. L.

Linn

M. C.

(2010). New trends in gender and mathematics performance: A meta-analysis. Psychological Bulletin, 136, 1123–1135. https://doi.org/10.1037/a0021276

58.

Linn

M. C.

(2010). Designing standards for lifelong science learning. Journal of Engineering Education, 99(2), 103–105. https://doi.org/10.1002/j.2168-9830.2010.tb01047.x

59.

Lubienski

S. T.

McGraw

Strutchens

M. E.

(2004). NAEP findings regarding gender: Mathematics achievement, student affect, and learning practices. In Kloosterman

Lester

J. F. K.

(Eds.), Results and interpretations of the 1990 through 2000 mathematics assessments of the national assessment of educational progress (pp. 305–336). National Council of Teachers of Mathematics.

60.

Lubinski

D. S.

Benbow

C. P.

(2007). Sex differences in personal attributes for the development of scientific expertise. In Ceci

S. J.

Williams

W. M.

(Eds.), Why aren’t more women in science? Top researchers debate the evidence (pp. 79–100). American Psychological Association. https://doi.org/10.1037/11546-007

61.

Johnson

(2008). Mathematics as the critical filter: Curricular effects on gendered career choices. In Watt

H. M. G.

Eccles

J. S.

(Eds.), Gender and occupational outcomes: Longitudinal assessments of individual, social, and cultural influences (pp. 55–83). American Psychological Association.

62.

Makel

M. C.

Wai

Peairs

Putallaz

(2016). Sex differences in the right tail of cognitive abilities: An update and cross-cultural extension. Intelligence, 59, 8–15. https://doi.org/10.1016/j.intell.2016.09.003

63.

Mann

H. B.

(1945). Nonparametric tests against trend. Econometrica, 13, 245–259.

64.

Mattern

K. D.

Marini

Shaw

E. J.

(2013). Are AP students more likely to graduate from college on time? (College Board Research Report No. 2013-5). College Board.

65.

Mattern

K. D.

Shaw

E. J.

Xiong

(2009). The relationship between AP Exam performance and college outcomes (College Board Research Report No. 2009-4). College Board.

66.

Moore

G. W.

Combs

J. P.

Slate

J. R.

(2012). Advanced Placement exams participation and performance: A national study of gender differences. E-International Journal of Educational Research, 3(3), 18–32.

67.

Moore

G. W.

Slate

J. R.

(2011). A multi-year analysis of Asian gender differences on Advanced Placement exams. International Journal of Educational Leadership Preparation, 6(4). http://cnx.org/content/m41412/latest/

68.

Morgan

Klaric

(2007). AP students in college: An analysis of five-year academic careers (College Board Research Report No. 2007-4). College Board.

69.

Morris

J. B.

(2013). Gender differences in STEM related Advanced Placement exams. Sam Houston State University.

70.

Morris

J. B.

Slate

J. R.

(2012). Advanced Placement math scores from 2000 to 2010: Does gender still matter? E-International Journal of Educational Research, 3(1), 33–48.

71.

Morris

J. B.

Slate

J. R.

Moore

G. W.

(2015). Gender and STEM related Advanced Placement exams: A review of the literature. The Online Journal of New Horizons in Education, 5(1), 55–71.

72.

Murphy

Dodd

(2009). A comparison of college performance of matched AP and non-AP student groups (College Board Research Report No. 2009-6). College Board.

73.

National Council of Supervisors of Mathematics & National Council of Teachers of Mathematics. (2018). Joint position statement: Building STEM education on a sound mathematical foundation. https://www.mathedleadership.org/docs/resources/positionpapers/NCSMPositionPaper17.pdf

74.

National Science Board. (2018). Science and engineering indicators 2018 (NSB-2018-1). National Science Foundation.

75.

National Science Foundation & National Center for Science and Engineering Statistics. (2019). Women, minorities, and persons with disabilities in science and engineering: 2019 (Special Report NSF 19-304). https://www.nsf.gov/statistics/wmpd

76.

Niederle

Vesterlund

(2010). Explaining the gender gap in math test scores: The role of competition. Journal of Economic Perspectives, 24(2), 129–144.

77.

Olszewski-Kubilius

Lee

S. Y.

(2011). Gender and other group differences in performance on off-level tests: Changes in the 21st century. Gifted Child Quarterly, 55(1), 54–73. https://doi.org/10.1177/0016986210382574

78.

Patterson

B. F.

Ewing

(2013). Validating the use of AP exam scores for college course placement (Research Report No. 2013-2). College Board.

79.

Prong

C. S.

(2018). Addressing ethnic and gender differences in Advanced Placement Calculus (AB) exam scores [Doctoral thesis, Concordia University, St. Paul]. https://digitalcommons.csp.edu/cup_commons_grad_edd/193

80.

Reis

S. M.

Callahan

C. M.

(1989). Gifted females: They’ve come a long way—Or have they? Journal for the Education of the Gifted, 12(2), 99–117. https://doi.org/10.1177/016235328901200203

81.

Reis

S. M.

Park

(2001). Gender differences in high-achieving students in math and science. Journal for the Education of the Gifted, 25(1), 52–73.

82.

Robinson

Shore

B. M.

Enersen

D. L.

(2007). Best practices in gifted education: An evidenced based guide. Prufrock Press.

83.

Robinson

J. P.

Lubienski

S. T.

(2011). The development of gender achievement gaps in mathematics and reading during elementary and middle school: Examining direct cognitive assessments and teacher ratings. American Educational Research Journal, 48(2), 268–302.

84.

Sadler

P. M.

Sonnert

Hazari

Tai

(2014). The role of advanced high school coursework in increasing STEM career interest. Science Educator, 23, 1–13.

85.

Sells

L. W.

(1980). The mathematics filter and the education of women and minorities. In Fox

Brody

Tobin

(Eds.), Women and the mathematical mystique (pp. 65–75). Johns Hopkins University Press.

86.

Sen

P. K.

(1968). Estimates of the regression coefficient based on Kendall’s tau. Journal of the American Statistical Association, 63(1968), 1379–1389.

87.

Shaw

E. J.

Barbuti

(2010). Patterns of persistence in intended college major with a focus on STEM majors. NACADA Journal, 30(2), 19–34. https://doi.org/10.12930/0271-9517-30.2.19

88.

Shaw

E. J.

Marini

J. P.

Mattern

K. D.

(2013). Exploring the utility of Advanced Placement participation and performance in college admission decisions. Educational and Psychological Measurement, 73(2), 229–253. https://doi.org/10.1177/0013164412454291

89.

Stevenson

D. L.

Schiller

Schneider

(1994). Sequences of opportunities for learning. Sociology of Education, 67, 184–198.

90.

Stumpf

Stanley

J. C.

(1996). Gender-related differences on the College Board’s Advanced Placement and Achievement tests, 1982–1992. Journal of Educational Psychology, 88(2), 353–364. https://doi.org/10.1037/0022-0663.88.2.353

91.

Subotnik

R. F.

Olszewski-Kubilius

Worrell

F. C.

(2011). Rethinking giftedness and gifted education: A proposed direction forward based on psychological science. Psychological Science in the Public Interest, 12(1), 3–54.

92.

Sundquist

(2016). AP exam scores: All your questions answered. https://blog.collegevine.com/ap-exam-scores-all-your-questions-answered/

93.

Sytsma

(2000, Spring). Gifted and Talented Programs in America’s High Schools: A preliminary survey report. The National Research Center on the Gifted and Talented Newsletter.

94.

Tay

Te Chuan

Aziati

Ahmad

(2018). An overview of Industry 4.0: Definition, components, and government initiatives. Journal of Advanced Research in Dynamical and Control Systems, 10(14), 1379–1387.

95.

Trusty

(2002). Effects of high school course-taking and other variables on choice of science. Journal of Counseling and Development, 80(4), 464–474. https://doi.org:10.1002/j.1556-6678.2002.tb00213.x

96.

VanTassel-Baska

(2000). Curriculum policy development for secondary gifted programs: A prescription for reform coherence. NASSP Bulletin, 84(615), 14–29.

97.

Wai

Cacchio

Putallaz

Makel

M. C.

(2010). Sex differences in the right tail of cognitive abilities: A 30-year examination. Intelligence, 38, 412–423. https://doi.org/10.1016/j.intell.2010.04.006

98.

Wai

Hodges

Makel

M. C.

(2018). Sex differences in ability tilt in the right tail of cognitive abilities: A 35-year. Intelligence, 67, 76–83. https://doi.org/10.1016/j.intell.2018.02.003

99.

Wai

Putallaz

Makel

M. C.

(2012). Studying intellectual outliers: Are there sex differences, and are the smart getting smarter? Current Directions in Psychological Science, 21, 382–390. https://doi.org/10.1177/0963721412455052

100.

Warne

R. T.

(2017). Research on the academic benefits of the advanced placement program: Taking stock and looking forward. SAGE Open. https://doi.org/10.1177/2158244016682996

101.

Wright

B. L.

Counsell

S. L.

(2018). The brilliance of Black boys: Cultivating school success in the early grades. Teachers College Press.

102.

Xie

Shauman

K. A.

(2003). Women in science: Career processes and outcomes. Harvard University Press.

103.

P. S.

Yang

T. C.

Chou

C. C.

(2002). Effects of climate change on evapotranspiration from paddy fields in southern Taiwan. Climate Change, 54, 165–179.