Advanced Placement and Rural Schools

Abstract

Completing Advanced Placement (AP) coursework is an important part of the selective college admissions process, and access to AP coursework can be viewed as a measure of equal opportunity. Relatively little research has fully examined how access to AP coursework is mediated by school characteristics. Rural schools are at a particular disadvantage in promoting AP success due to a lack of sufficiently prepared students, teaching constraints, and other logistical challenges. This study uses regression to analyze AP access, enrollment, and success across a rich data set of U.S. school districts. We find that remote, small, and poor rural schools are considerably less likely to offer any AP courses. Even for rural schools that do offer AP coursework, enrollment and success rates lag well behind more urban and affluent schools. We explore potential policy solutions and examine alternative advanced coursework options for those rural schools that do not offer AP.

Keywords

equity advanced coursework rural schools gifted students

Advanced Placement (AP) offers a series of college-level courses and assessments for which students may receive college credit while still in high school. The AP program was established by the College Board, and this entity is responsible for certifying AP courses throughout the nation. Schools that decide to offer AP coursework must assign an AP coordinator to handle logistical aspects of using curricular and assessment materials, and school-designed AP curricula must pass an audit process to receive the “AP” designation. Although there are no strict requirements for teaching AP courses, many teachers attend AP-specific workshops, conferences, and/or summer institutes—often time fulfilling required professional learning hours in the process—to prepare for teaching an AP course.

The recent push for high, common standards and college-readiness has led to an exponential growth in the adoption of AP in public high schools: In the 10 years leading up to 2012, the number of students taking an AP course more than doubled to over 2 million, the number of schools offering an AP course increased by nearly a third to include roughly half of all U.S. public and private secondary schools, and the number of colleges that accepted AP scores increased by nearly a fifth (College Board, 2012).¹ Because of the growth in AP offerings and the level of rigor AP courses convey, AP coursework is viewed by many to be an important part of the selective college admissions process (Geiser & Santelices, 2004; Iatarola, Conger, & Long, 2011; Schneider, 2009). A recent report, which was endorsed by 80 key stakeholders in the college admissions process, cautions against “overloading” on AP courses (Weissbourd, 2016), perhaps signaling that the zenith of AP’s influence has already been reached. However, even this report acknowledges the importance of increasing access to challenging coursework to all schools, and it is likely that AP will continue to play a sizable role in competitive college admissions for the foreseeable future.

Despite the increasing importance of AP courses for student competitiveness and college preparation, disparities in who has access to AP coursework and who succeeds in it are evident. A study using 2000 data from the state of Texas found that rural districts are less likely to offer AP courses (Klopfenstein, 2004), although it is unclear how these gaps have persisted up to the present, or across the entire country. Furthermore, some have noted a shortage of research that asks why some schools offer AP courses while others do not (Iatarola et al., 2011). And, although well-documented disparities exist in rates of AP enrollment and success between poor and minority students on one hand, and their more affluent, non-minority peers on the other, less is known about the relative success of rural students that do engage in AP coursework. For gifted students in rural schools, a lack of AP coursework offerings would present a significant threat to equal opportunity given they could not demonstrate this college-readiness in the same manner as their more urban peers.² There is a need for a more comprehensive study of AP access and success, and across a wider range of school characteristics. This study fills this gap in the literature by presenting thorough analyses on trends in AP access across U.S. school districts. In addition, the present study expands the literature by also considering enrollment, and success, in districts where at least one AP course is offered. Finally, this study ends by exploring initiatives aimed at increasing advanced coursework options for gifted rural students.

Benefits of AP

Studies from the College Board, as well as the Educational Testing Service³ (ETS) support the effectiveness of AP coursework in improving college outcomes, specifically. Taking AP courses in high school may affect the type of college courses a student will take, as one study found that completing only one AP course increases the chance that a student will take courses in that subject area in college (Morgan & Maneckshana, 2000). A report from the Center for Public Education found that AP participation was also related to persistence in college, especially for at-risk students (Klepfer & Hull, 2012). However, although these studies do control for student-level demographics such as income and race, isolating the causal effect of taking AP coursework using such methodologies remains extremely difficult. Effect estimates of AP coursework may be biased due to unobserved student self-selection effects, peer effects, and differences in teacher and school quality (Dougherty, Mellor, & Jian, 2006), and therefore, it is difficult to conclude that completing a AP class, as opposed to other challenging coursework, improves student outcomes. Furthermore, other evidence suggests that taking an AP course may have deleterious effects for under-prepared students (Sadler & Tai, 2007). Despite these limitations, taken in concert, this and other research (Rose & Betts, 2001) offer suggestive evidence that prepared students likely benefit in college and beyond when they engage in advanced high school curriculum such as AP.

Understanding the benefits of succeeding in AP coursework, as opposed to merely participating in an AP course, is somewhat more straightforward. A score of 3 out of 5 is considered a passing grade on an AP examination, with a 5 out of 5 representing the highest possible score. One study found that passing an AP examination was related to better grades in college, even when accounting for SAT/ACT scores (Keng & Dodd, 2008). Many colleges offer credit for passing AP examinations, although policies differ tremendously between participating colleges as to the required passing scores and number of credits earned across AP subjects. These earned college credits may fulfill graduation requirements, and thus grant students considerably more flexibility in course taking while enrolled in college, enabling opportunities for study abroad, internships, electives, or more advanced college coursework that otherwise would not be feasible for students. Furthermore, students who earn enough AP credits may reduce tuition expenses through early graduation, although this may be more of an exception than a rule (Challenge Success, 2013). Even if colleges offer limited credit options for AP coursework, most would argue that a high school transcript is looked at more favorably when it includes strong AP results, as one study found that 91% of colleges and universities consider AP experience in the application process (Sathre & Blanco, 2006). Ultimately, whether or not a district offers AP courses is an important indicator of equality of educational opportunity: In districts without AP access, even the most gifted students might not have the opportunity to earn college credit in high school, and could face a disadvantage in applying to elite colleges.

Gaps in AP Access and Success

Given the likely benefits of AP course enrollment for gifted and well-prepared students and its increasing connection to college success, gaps in AP access and success serve as important markers of the equality of opportunity in our public schools. Unfortunately, well-documented disparities exist in AP enrollment and success between poor and minority students and their more affluent, non-minority peers. Low-income students represent only a quarter of AP test takers in 2013 despite making up roughly half of the public high school population (College Board, 2014). Similarly, Black and American Indian students participate at roughly half the average rate, although Hispanic students have only somewhat lower AP enrollment and success rates (College Board, 2014). Gaps in AP success are even more pronounced, especially for Black students, who make up 14.5% of high school graduates yet only 4.6% of students with a score of 3 or higher on AP examinations. Overall, it is unclear how rural students compare with their peers on these measures, as such data are not made readily available by the College Board.

Considering the overwhelming importance that home and community factors play in determining one’s likelihood of enrolling and succeeding in AP coursework, gaps in these metrics may be more a reflection of broader inequities then they are a source of its perpetuation. We argue that simply having access to an AP course at one’s school is a better measure of equality of opportunity, for lacking AP options entirely could put gifted students at a significant disadvantage compared with their peers with access. We find limited (and now, given the exponential increase in AP course taking, somewhat dated) research that examines AP access gaps across poverty and place in schools. A report by the National Center for Education Statistics (NCES) examined AP access gaps, finding that only 55% of rural schools offer an AP or International Baccalaureate (IB) course compared with 69% of all schools (Thomas, Marken, Gray, & Lewis, 2013). However, this study uses descriptive analyses alone on a limited sample, and ultimately it does not examine the remoteness of rural schools or the compounding effects of school size, student composition, and rurality. Iatarola et al. (2011) use Florida panel data to study determinants of AP course offerings, finding a dramatic relationship between school size and likelihood to offer AP or IB courses, although this relationship attenuates considerably when controlling for the number of students whose prior achievement is far above average; the authors suggest having a critical mass of high achieving students is the main reason for why larger schools offer more advanced courses. Klopfenstein (2004) finds that only 44% of rural Texas public high schools supported an AP program in the year 2000, compared with 73% of non-rural schools. Interestingly, this study found that schools with high Black and Hispanic compositions were actually more likely than the average school in the sample to support an AP program. Klopfenstein (2004) also finds higher school enrollment, fewer low-income students, and non-rural status to be associated with greater gains over time in the number of AP course offerings at a school.

Overall, it is unclear how rurality, remoteness, school size, and student demographic compositions are related to AP access throughout the entire nation. Disparities in AP access and achievement for rural students are not thoroughly documented, as no national study examines how contextual aspects of rurality affect AP access, enrollment, or success. A wide range of circumstances, including proximity to urban areas, industrial composition, racial/ethnic makeup, and level of poverty, vary widely across areas that share a “rural” designation (Lichter & Graefe, 2011). If one were to generalize current trends in AP access to all rural schools, one would most likely overstate levels of access for those rural schools that face more challenges, including small, poor, and remote ones. This study develops a richer understanding of AP access and success across poverty and place in the United States, using a recent, nationally representative data set to do so. A number of barriers exist to offering AP coursework, including the lack of a critical mass of gifted and sufficiently prepared students, the lack of teachers prepared to teach an AP course, and the financial costs associated with establishing an AP program. We assume that rural schools (especially remote ones), low-enrollment schools, high-minority, and poor schools face some or all of these barriers, and therefore, we hypothesize that such schools will be less likely to provide access to AP coursework.

Data and Method

This study assesses trends in access to, enrollment in, and success in AP coursework in relation to school district characteristics such as urbanicity, enrollment, remoteness, poverty, and racial composition. The data used here are from three sources: the 2011-2012 Civil Rights Data Collection (CRDC), the 2012 Small Area Income and Poverty Estimates (SAIPE), and data on urban-local codes conducted by the U.S. Census (with support from NCES) that are available through the Common Core of Data (CCD). The CRDC is a mandatory data collection that provides information on the rates of AP enrollment and success among all schools in the United States. SAIPE provides information on the number of students living in poverty in a district. The CCD supplies information on urbanicity as well as other district-level demographics. Because data on poverty and urbanicity are available at the district level, it is not possible to construct a national data set that would allow for one to draw conclusions about individual students or school-level trends from this analysis. Therefore, we aggregated CRDC data to the district level, and then merged these data with SAIPE and CCD data using the NCES district ID code. Any district not found in all three data sets, or any district that contained 10 or fewer secondary students,⁴ was dropped. This resulted in 11,111 school districts being examined for access to AP coursework, which together enroll the overwhelming majority of students in U.S. public schools. All binary variables that are derived from population distributions are done so using this data set. When examining AP enrollment and success, however, we consider only those districts that offer some access to AP coursework, because this makes visible the trends in districts that have broken the threshold in offering AP coursework. After dropping districts without AP access, 6,765 districts remained from which to analyze trends in enrollment and success.

To understand trends in access to AP courses in U.S. school districts, a taxonomy of logistical regressions is fitted. The following groups of variables are added one at a time as predictor variables: urbanicity, rural remoteness, low student enrollment, school poverty, and high-minority status; we then refit the final model while including state fixed effects. This approach allows one to identify the association of each variable and AP access independent of the other variables included, giving insight into the unique effects of each variable. We do not control for total enrollment, as population is an inherent characteristic of urbanicity, and ultimately the two variables are highly collinear. However, we do use a binary indicator of low enrollment, which although related to urbanicity, would not conflate its effect to the same degree as a continuous enrollment variable. Furthermore, we see conceptual advantages to including this binary variable: Districts of very low enrollment face acute challenges to offering AP due to staff size, with many teachers being likely to instruct multiple subjects. When examining AP enrollment and success across U.S. school districts, we examine only those districts that offer at least one AP course. The same taxonomy of regressions is fitted to first enrollment and then success rates in school districts, this time using ordinary least squares (OLS) regression because these outcomes are continuous.⁵ The results of each taxonomy of regressions are interpreted and discussed.

Outcome Variables

AP access

A binary variable equals to 1 if a school district offers any AP courses, 0 otherwise. We consider AP access to exist if at least one secondary student (grades 9-12) is enrolled in at least one AP course anywhere in the district.⁶

AP enrollment

AP enrollment is the percentage of secondary students who are enrolled in at least one AP course. This study examines enrollment trends for school districts with AP access, only.

AP success

AP success is the percentage of secondary students enrolled in at least one AP course who take and pass (score of 3 or higher) at least one end-of-year AP exam. This study examines trends in AP success for school districts with AP access, only.

Predictor Variables

Urbanicity

Urbanicity is a vector of four dummy variables. The U.S. Census generates urbanicity variables using four major types: city, suburb, town, and rural.⁷ Rural serves as the excluded category in regression analyses.

Low enrollment

A binary variable equals to 1 if a school district is in the lowest quartile of secondary student enrollment, 0 otherwise. A low-enrollment district corresponds to an enrollment of 170 secondary students or fewer.

Rural remoteness

An ordinal variable equals to 0 if a district is within an urban cluster (i.e., non-rural), 1 if the district is less than 5 miles from an urbanized area or 2.5 miles from an urban cluster (fringe rural district); 2 if the district is more than 5 miles but less than or equal to 25 miles from an urbanized area, or more than 2.5 miles but less than or equal to 10 miles from an urban cluster (distant rural district); 3 if the district is more than 25 miles from an urbanized area and is also more than 10 miles from an urban cluster (remote rural district).

District poverty

This is defined as the percentage of school-aged youth in a school district who live in a family with income below the official poverty threshold.

High minority

A binary variable equals to 1 if a school district is in the highest quartile of non-White/non-Asian⁸ population, 0 otherwise. A high-minority district corresponds to a non-White/non-Asian racial composition of 28.6% or greater.

Findings

Descriptive analyses suggest that rural districts are less likely to support AP programs: Only 51.4% of rural school districts enroll at least one student in an AP course, compared with 78.3%, 93.8%, and 97.3% of town, suburban, and urban districts, respectively.⁹ Fitting a taxonomy of logistic regressions helps to illuminate how urbanicity, remoteness, low enrollment, poverty, and high-minority status combine to influence a district’s likelihood of offering AP coursework. Table 1 shows the odds ratio estimates for Models 1 through 7. A pseudo r² value of .36 for the full model (Model 7) suggests that a fair amount of variability in district-level AP access may be explained by relatively few variables. We find that the estimated effects of predictor variables are robust to the inclusion of other covariates, as is evidenced by fairly consistent values across models. It is important to note that the vector of urbanicity variables—city, suburb, and town, with rural serving as the excluded category—must be examined in conjunction with remoteness, and when doing so, we see that the relative difference between city and rural is robust to the inclusion of poverty, high-minority status, and state fixed effects.¹⁰ Table 1 illustrates that cities, suburbs, and high-minority populations are associated with a higher likelihood of AP access at the district level (an odds ratio higher than 1.0), while low enrollment, greater rural remoteness, and higher district poverty are associated with lower likelihoods. Town districts are statistically less likely to offer AP than cities and suburbs, and more likely to do so than rural districts.¹¹ Perhaps surprisingly, the signs, magnitudes, and significances of the urbanicity, remoteness, poverty, and high-minority status estimates remain similar even when fitting Model 5 to only those districts that fall within the interquartile range of secondary student population (results not shown in Table 1). Thus, these findings are not merely an artifact of higher enrollment districts having very high AP access and vice versa for low-enrollment districts. Also surprising is the estimated effect of high-minority status; descriptive analyses not shown here suggest that high-minority districts have AP access rates equal to or greater than non-high-minority districts in all urbanicity categories. It may be the case that higher average enrollment in high-minority districts leads to this measured effect, even though low-enrollment status is controlled for. Again, collinearity with place type precludes us from controlling for the full range of enrollment numbers.

Table 1.

Predictors of AP Access, U.S. School Districts With AP Access (Odds Ratio Estimates From Logistic Regression).

	1	2	3	4	5	6	7
	No FE						State FE
Constant	2.05 (0.041)***	1.06 (0.027)*	2.53 (0.093)***	7.18 (0.64)***	9.03 (0.89)***	9.58 (0.96)***	NA
Urbanicity (rural excluded):
City		33.79 (8.35)***	14.77 (3.56)***	5.14 (1.31)***	5.89 (1.50)***	4.91 (1.26)***	7.30 (1.97)***
Suburb		14.38 (1.36)***	6.30 (0.61)***	2.19 (0.28)***	2.19 (0.28)***	2.02 (0.26)***	2.04 (0.28)***
Town		3.42 (0.27)***	1.57 (0.10)***	0.55 (0.056)***	0.60 (0.063)***	0.60 (0.063)***	0.88 (0.105)
Low enrollment			0.13 (0.007)***	0.18 (0.011)***	0.17 (0.011)***	0.17 (0.011)***	0.21 (0.016)***
Rural remoteness				0.57 (0.025)***	0.60 (0.027)***	0.60 (0.027)***	0.65 (0.033)***
Percent poverty					0.98 (0.0026)***	0.98 (0.0029)***	0.95 (0.0034)***
High minority						1.69 (0.12)***	1.69 (0.14)***
Psuedo R²	0	.144	.252	.263	.266	.270	.355

Note. Robust standard errors in parentheses. AP = Advanced Placement; FE = fixed effects.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Examining odds ratios allows one to interpret the relative effect size of included variables. For instance, we estimate that suburban districts have 6.3 times the odds of rural districts of offering an AP class, controlling for low-enrollment status (Model 3). Interpreting a continuous variable is slightly different: One percentage point difference in student poverty is associated with a 5% decrease in the likelihood of offering an AP course (as an odds ratio of .95 is .05 less than 1; Model 7) for a district of average poverty. Such an odds ratio is meaningfully large, given the modesty of a one percentage point difference in the poverty rate. Interpretation in this manner becomes difficult if one wishes to examine the impact of multiple variables simultaneously, especially when collinear variables (i.e., urbanicity and remoteness) are included at the same time. Therefore, we compute the predicted probabilities of a number of hypothetical school districts in the United States (see Table 2). This analysis shows that for a suburban school district of average poverty, not having low enrollment, and not in the highest quartile of minority students, we estimate a 93% chance of offering AP coursework. This contrasts considerably for the estimate of a remote, poor rural district of low enrollment: a mere 14%.

Table 2.

Predicted Probabilities of AP Access, Select District Characteristics.

Type of district	Poor (10th percentile)	Median poverty	Affluent (90th percentile)
Urban, high minority	0.97	0.98	0.99
Urban	0.96	0.97	0.98
Suburban	0.90	0.93	0.94
Suburban, low enrollment	0.61	0.69	0.74
Town	0.72	0.79	0.83
Town, low enrollment	0.31	0.40	0.46
Fringe rural	0.73	0.79	0.83
Fringe rural, low enrollment	0.32	0.40	0.46
Distant rural	0.61	0.70	0.74
Remote rural	0.49	0.58	0.64
Remote rural, low enrollment	0.14	0.19	0.23

Note. AP = Advanced Placement.

Enrollment and Success

Descriptive analyses show that even in districts that have AP access—those that have found a way to offer AP to at least one student, thus making it easier for additional students to take AP courses—disparities in enrollments still exist. The average urban or suburban district enrolls 6.1% of secondary students in an AP class, which is a rate 30% higher than in town or rural districts. OLS regression shows that this positive effect persists even when including a host of controls, and that disparities widen with increased rural remoteness (see Table 3). Estimates from Model 6, for instance, suggest that remote rural schools are associated with AP enrollment rates 1.4 percentage points lower than those of city and suburb districts, controlling for low-enrollment status, poverty, and high-minority status. The effect of school poverty is also meaningful: A 10 percentage point difference in student poverty rates is negatively associated with between a 1.1 and 1.3 percentage point difference in enrollment rates. Note that such differences in AP enrollment is rather large, given the fact that the district average AP enrollment rate, at 5.2%, is relatively low. High-minority status is not a significant predictor of AP enrollment. A somewhat surprising finding from this analysis is the rather large, positive estimated effect of low enrollment: Schools of low enrollment are associated with upward of a 2 percentage point higher enrollment rate (depending on the controls included). One explanation for this finding is that those small districts that do commit to offering AP coursework likely view AP as a big investment, and therefore undertake efforts to promote AP course offerings. Furthermore, the addition of even one AP class likely has a greater impact on the percentage of AP students in smaller districts than it would in larger districts, simply due to the difference in overall number of high school students. When controlling for state fixed effects (Model 7), we find that estimated magnitudes remain relatively stable. Overall, these analyses show that rural students are meaningfully less likely to take AP coursework than their urban and suburban peers, even when AP courses are available at their schools.

Table 3.

Determinants of District AP Enrollment Rates, U.S. School Districts With AP Access (Ordinary Least Squares).

	1	2	3	4	5	6	7
	No FE						State FE
Intercept	5.2 (0.06)***	4.9 (0.10)***	4.6 (0.10)***	5.6 (0.28)***	7.1 (0.28)***	7.2 (0.28)***	NA
Urbanicity (rural excluded)
City		0.44 (0.21)*	0.65 (0.21)**	−0.32 (0.33)	0.74 (0.33)*	0.66 (0.34)^†	0.55 (0.33)
Suburb		1.45 (0.16)***	1.65 (0.16)***	0.68 (0.30)*	0.66 (0.30)*	0.61 (0.30)*	0.47 (0.30)
Town		−0.49 (0.15)**	−0.30 (0.15)*	−1.28 (0.30)***	−0.48 (0.30)	−0.48 (0.30)	−0.62 (0.30)*
Low enrollment			1.59 (0.35)***	2.01 (0.38)***	2.00 (0.38)***	2.00 (0.38)***	1.68 (0.39)***
Rural remoteness				−0.58 (0.14)***	−0.27 (0.14)^†	−0.26 (0.14)^†	−0.36 (0.14)*
Percent poverty					−0.11 (0.006)***	−0.11 (0.007)***	−0.13 (0.008)***
High minority						0.31 (0.17)^†	−0.10 (0.19)
R ²	0	.021	.026	.028	.068	.068	.140

Note. Robust standard errors in parentheses. AP = Advanced Placement; FE = fixed effects.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Suburban districts exhibit by far the highest rates of AP success. In suburban districts, the average percentage of AP-enrolled students who have passed at least one AP exam is 45.3%, compared with 36.0%, 32.0%, and 31.8% for students in urban, town, and rural districts, respectively. Such disparity in success is even greater across lines of poverty, as school districts in the most affluent quartile of the United States exhibit an average success rate (52.1%) more than double that of districts in the poorest quartile (24.0%). OLS regression results (see Table 4) confirm that both urbanicity and poverty remain powerful predictors even when controls are added. Again examining Model 6, we see that remote rural school districts have rates of AP success roughly 11 percentage points lower than urban or suburban school districts, controlling for low-enrollment status, poverty, and high-minority status. This finding should be tempered by the fact that urban and suburban AP students likely take more AP courses per year, and therefore have more opportunities to be successful in at least one exam. These regression results also highlight the oversized effect of poverty on chances of success: School district poverty rates alone explain 14% of the variation in AP success rates (analysis not shown), and a 10 percentage point difference in poverty rate is roughly associated with AP success rates about 8 percentile points lower, depending on the controls used. We find that although low district enrollment is positively associated with AP enrollment rates, low-enrollment districts are associated with success rates 3.9 percentage points lower, controlling for urbanicity, remoteness, poverty, and minority (Model 6). Some of this effect may be due to the average AP taking student in such districts being less prepared than the average AP student, and/or because students in such schools are less likely to take multiple AP exams in a year. Estimates are generally robust to the inclusion of state fixed effects, although the estimated magnitude of remoteness and low-enrollment decrease considerably, suggesting that states with more small and remote rural schools might have other contributing factors to low AP success rates.

Table 4.

Determinants of District AP Success Rates, U.S. School Districts With AP Access (Ordinary Least Squares Regression With Robust Standard Errors).

	1	2	3	4	5	6	7
	No FE						State FE
Intercept	36.0 (0.33)***	31.8 (0.50)***	36.0 (0.33)***	39.9 (1.39)***	53.1 (1.41)***	36.0 (0.33)***	NA
Urbanicity (rural excluded)
City		4.25 (1.12)***	3.34 (1.14)**	−3.90 (1.72)*	5.41 (1.64)**	6.08 (1.66)***	6.65 (1.63)***
Suburb		13.51 (0.80)***	12.63 (0.82)***	5.38 (1.53)***	5.17 (1.45)***	5.51 (1.46)***	4.16 (1.41)**
Town		0.19 (0.83)	−0.63 (0.84)	−7.91 (1.54)***	−1.00 (1.50)	−1.05 (1.50)	1.07 (1.46)
Low enrollment			−6.93 (1.37)***	−3.75 (1.47)*	−3.89 (1.46)**	−3.89 (1.45)**	−1.76 (1.43)
Rural remoteness				−4.36 (0.76)***	−1.62 (0.74)*	−1.73 (0.74)*	−1.08 (0.72)
Percent poverty					−0.94 (0.031)***	−0.89 (0.035)***	−0.79 (0.039)***
High minority						−2.34 (0.78)**	−2.36 (0.78)**
R ²	0	.048	.052	.056	.164	.165	.254

Note. Robust standard errors in parentheses. AP = Advanced Placement; FE = fixed effects.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Discussion

We find that rural school districts are much less likely to offer any AP courses, especially when a rural district is small, remote, and/or poor. These findings clearly have worrisome implications regarding equal access to educational opportunity, as many gifted rural students cannot use AP to experience college-level coursework, earn college credit while in high school, or develop an advantage in the selective admissions process. Furthermore, even among rural districts that do offer AP coursework, AP enrollment and success rates lag behind their more metropolitan peers. This is true even when poverty is controlled for, and such disparities in enrollment and success widen as remoteness increases. Considering that rural students lack access to the kinds of rigorous coursework in earlier grades that might prepare students for AP success (Graham, 2009), this is unsurprising. Affluent urban and suburban schools might have better developed AP cultures or more selective requirements for enrolling in AP courses, leading to higher rates of success. A culture of AP success may be engendered by providing teachers with AP-specific professional development, or covering the AP exam costs for students, although this likely requires a critical mass of interested and prepared students. Many rural districts may deem AP programs unsuitable to their context due to insufficient numbers of capable students, lack of appropriate teacher staffing, and/or other logistical concerns owing to small, isolated populations.

There are, however, a number of policy initiatives that may promote AP access. Dedicated funding for poor and remote rural schools could help to overcome some of the aforementioned barriers by covering the costs of books, distance technology, and AP-specific teacher professional development. The preparation problem may be somewhat addressed by identifying promising students, and offering them pre-AP curricula in middle and early high school years to develop the prerequisite skills for success in AP. State-sponsored summer AP programs, which allow students to engage in face-to-face instruction in courses not offered at their own schools, offer particular promise for gifted rural students. However, AP examinations are only offered in the spring, so such summer programs likely require coordination with schools to ensure that summer AP students also have the opportunity to take and pass the AP examination. Further research should endeavor to understand the effectiveness of these policies that are already in place, and policy makers should continue to find creative solutions to address the gaps in AP access and success.

Alternatives to AP

We end our discussion with a brief survey of some alternatives to AP, including dual enrollment, distance learning, and district-developed curricula,¹² to better situate AP in the full landscape of possibilities. Note that our data cannot speak to trends in these alternatives, and we do not present these alternatives as “novel” ideas, necessarily. Rather, our motivation for here is twofold. First, it is important to leave the reader with the understanding that AP, although widespread and highly influential, is not the only means to promote advanced offerings to gifted students. Second, given the striking findings in AP access found in this article, it is important to consider the relative advantages and disadvantages of each alternative to understand how those areas with low AP access may or may not have suitable advanced coursework options other than AP.

Dual enrollment (sometimes known as early college) enables high school students to take class at local colleges and earn college credit if successful. Unlike AP, dual enrollment entails taking actual college classes, with syllabi and credit criteria set by the college. Classes may be offered at the participating college, or at a high school, with the teacher acting as an adjunct faculty member; they may be academic or vocational in nature. Dual enrollment is well-established alternative to AP, and a greater proportion of rural schools offer dual enrollment options than do more metropolitan locales (Thomas et al., 2013). There are number of distinct advantages to dual enrollment, perhaps most notably that a more seamless transition between secondary and tertiary education may be established. For instance, some have argued that exposure to the non-academic aspects of college may serve as an important opportunity for students to develop social capital (Bailey, Hughes, & Karp, 2002). In addition, dual enrollment may bypass concerns of inadequate teacher preparation for delivering advanced instruction by relying on college instructors.

However, successfully implementing a dual enrollment program likely faces more impediments than starting an AP program, and it is important that factors such as funding arrangements, communication between levels, and community awareness are addressed (Howley, Howley, Howley, & Duncan, 2013; Stephenson, 2014). Furthermore, some rural districts that are unsuitable for AP programs may also lack a nearby tertiary institution with which to establish a dual enrollment program. Another possible disadvantage of dual enrollment compared with AP is that because dual enrollment lacks the standardization of AP, it may be difficult for college admissions to judge the rigor of some dual enrollment options. Indeed, one could easily envision cases where a dual enrollment class offers inferior instruction to an AP course staffed by a trained educator. Although the connection between dual enrollment and increased college attendance and completion is rather robust (An, 2012; Cowan & Goldhaber, 2015; Giani, Alexander, & Reyes, 2014), it remains difficult to establish a causal connection, and other evidence suggests that it performs no better than AP in this regard (Speroni, 2011).

In the case of especially small and isolated rural schools that may lack internal capacity and the proximity to colleges, distance (online) learning may offer promise for gifted students. Here we use distance learning to describe any means of taking highly rigorous coursework without the need for students and educators to be in physical proximity. Distance learning has been used in AP coursework (known as Virtual AP) as well as in dual enrollment programs. The CRDC requires districts to report on AP enrollment and success of students within their district, and as Virtual AP is a form of AP, the statistics presented here should include Virtual AP. According to the Education Commission of the States (2015), 28 states currently offer some form of Virtual AP, and a number of these states also provide technical and financial support to students who take such courses. Thus, Virtual AP can be a cost-effective way to provide challenging courses to isolated students; virtual dual enrollment is similar in this regard. Distance education is especially attractive in hard-to-staff subject areas, and some research suggests models that offer particular promise for small rural schools wishing to extend their advanced offerings (Irvin, Hannum, Farmer, de la Varre, & Keane, 2009). Furthermore, some research suggests that online delivery may be a credible means to deliver AP content when compared with brick and mortar instruction (Johnston & Barbour, 2013), although the effectiveness of online formats is heavily dependent on how such instruction is designed, delivered, and supported (Barbour, 2014).

However, many critics believe that online learning is not a replacement for traditional face-to-face classroom settings where students can engage more readily and deeply with their instructor and peers. A meta-analysis of distance learning research suggests that online strategies are generally as effective as face-to-face instruction (Nguyen, 2015), although most of this research fails to examine socio-emotional learning or the long-term impacts of such instructional practice. Overall, a conservative conclusion here is that distance learning offers promise in cases where face-to-face instruction is not feasible. In cases where the promise of distance learning lies largely in its cost savings, it should be implemented cautiously: Not all forms of distance learning are equally effective, nor can all the benefits of face-to-face instruction be easily replaced by technology.

Perhaps the least discussed alternative to adopting an AP program is for districts to create their own advanced curricula for gifted students. Clearly, challenges exist in taking such an approach—most notably a lack of capacity to devise and deliver advanced instruction, especially in a variety of content areas—but a creative, student-centered approach may allow some rural schools to provide a challenging environment to even their most advanced students. Conley (2007) outlines a number of college-readiness approaches which could be adapted by high schools to support gifted learners, such as conducting school-wide audits of syllabi and creating challenging senior seminars. One drawback to district-developed advanced curricula is that AP and dual enrollment have nearly become synonymous with rigor; students attending schools without these sanctioned bastions of advanced offerings may be placed at a disadvantage in selective college admissions even if students do pursue challenging coursework. Regardless, given the scant evidence to support the effectiveness of AP programs in particular over many other forms of rigorous coursework, school districts without AP programs should thoroughly explore all available options.

We caution against letting the tail wag the dog: Although we argue that AP access is an important metric through which to better understand the equality of opportunity for rural students, the creation of an AP program is not an ends unto itself. For instance, smaller rural districts have viewed the creation of an AP program as a major benefit to school district consolidation (Nitta, Holley, & Wrobel, 2010), even though there may be opposition to consolidation, and distance learning or dual enrollment may be more viable policy options (Howley, Rhodes, & Beall, 2009). Whatever solution is pursued, it is important that administrators and policy makers consider carefully the plight of gifted rural students. As Azano, Callahan, Missett, and Brunner (2014) note, although federal policies ensure that special education services are delivered to those that qualify, no similar protections are offered for gifted students. In metropolitan and highly resourced areas, gifted students are seldom left behind due to the critical mass of well-prepared students, engaged parents, and proximity to enriching opportunities. All too often, however, gifted rural students are not so lucky. Ultimately, the onus falls squarely on states and localities to ensure that their gifted rural students have an opportunity to be challenged in school.

Footnotes

Authors’ Note

This study builds upon a research brief that the authors recently published (http://scholars.unh.edu/cgi/viewcontent.cgi?article=1234&context=carsey). Although no additional data are used in this study, it uses multivariate regression analysis, whereas the associated brief only presented descriptive results. In addition, policy implications are much more thoroughly explored, including the alternatives to Advanced Placement.

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported in part by grants from the Annie E. Casey Foundation, The W. K. Kellogg Foundation, and an anonymous donor.

Notes

About the Authors

Douglas J. Gagnon is a research associate at the University of New Hampshire’s Carsey School of Public Policy. Doug’s research has focused primarily on education policy as it relates to the equality of opportunity, teacher quality and school staffing, and trends in student discipline.

Marybeth J. Mattingly is director of research on vulnerable families at the Carsey School of Public Policy. Beth manages all work relating to family well-being, including such topics as poverty trends, safety net policy, and access to education.

References

B. P.

(2012). The impact of dual enrollment on college degree attainment: Do low-SES students benefit? Educational Evaluation and Policy Analysis, 35, 57-75. doi:10.3102/0162373712461933

Azano

A. P.

Callahan

C. M.

Missett

T. C.

Brunner

(2014). Understanding the experiences of gifted education teachers and fidelity of implementation in rural schools. Journal of Advanced Academics, 25, 88-100. doi:10.1177/1932202X14524405

Bailey

Hughes

Karp

(2002). What role can dual enrollment programs play in easing the transition between high school and postsecondary education? (Publication No. ED-99-CO-0160). Washington, DC: Office of Vocational and Adult Education, U.S. Department of Education.

Barbour

M. K.

(2014). Review of virtual schooling and student learning. Boulder: National Policy Education Center, University of Colorado Boulder.

Callahan

C. M.

(2003). Advanced Placement and International Baccalaureate programs for talented students in American high schools: A focus on science and mathematics. Charlottesville, VA: National Research Center on the Gifted & Talented.

Challenge Success. (2013). The Advanced Placement program: Living up to its promise? Stanford, CA: Author. Retrieved from http://www.challengesuccess.org/wp-content/uploads/2015/07/ChallengeSuccess-AdvancedPlacement-WP.pdf

College Board. (2012). Annual AP program participation 1956-2012. Retrieved from http://media.collegeboard.com/digitalServices/pdf/research/2012_annual_participation.pdf

College Board. (2014). The 10th annual AP report to the nation. Retrieved from http://media.collegeboard.com/digitalServices/pdf/ap/rtn/10th-annual/10th-annual-ap-report-to-the-nation-single-page.pdf

Conley

D. T.

(2007). The challenge of college readiness. Educational Leadership, 64(7), 23-29.

10.

Cowan

Goldhaber

(2015). How much of a “running start” do dual enrollment programs provide students? The Review of Higher Education, 38, 425-460.

11.

Dougherty

Mellor

Jian

(2006). The relationship between Advanced Placement and college graduation (2005 AP Study Series, Report 1). Austin, TX: National Center for Educational Accountability.

12.

Education Commission of the States. (2015). Advanced Placement: State support for encouraging access to AP. Retrieved from http://ecs.force.com/mbdata/MBQuestSNR?Rep=AP08

13.

Geiser

Santelices

(2004). The role of Advanced Placement and honors courses in college admissions (Research & Occasional Paper Series: CSHE.4.04). Berkeley, CA: Center for Studies in Higher Education.

14.

Giani

Alexander

Reyes

(2014). Exploring variation in the impact of dual-credit coursework on postsecondary outcomes: A quasi-experimental analysis of Texas students. The High School Journal, 97, 200-218.

15.

Graham

(2009, December). Students in rural schools have limited access to advanced mathematics courses (Issue Brief No 7). Durham: Carsey Institute, University of New Hampshire.

16.

Howley

M. D.

Howley

C. B.

Duncan

(2013). Early college and dual enrollment challenges: Inroads and impediments to access. Journal of Advanced Academics, 24, 77-107. doi:10.1177/1932202X13476289

17.

Howley

Rhodes

Beall

(2009). Challenges facing rural schools: Implications for gifted students. Journal for the Education of the Gifted, 32, 515-536.

18.

Iatarola

Conger

Long

M. C.

(2011). Determinants of high schools’ advanced course offerings. Educational Evaluation and Policy Analysis, 33, 340-359.

19.

Irvin

M. J.

Hannum

W. H.

Farmer

T. W.

de la Varre

Keane

(2009). Supporting online learning for Advanced Placement students in small rural schools: Conceptual foundations and intervention components of the facilitator preparation program. Rural Educator, 31, 29-37.

20.

Johnston

Barbour

M. K.

(2013). Measuring success: Examining achievement and perceptions of online Advanced Placement students. American Journal of Distance Education, 27, 16-28.

21.

Keng

Dodd

B. G.

(2008). A comparison of college performances of AP and non-AP student groups in 10 subject areas. Princeton, NJ: The College Board.

22.

Klepfer

Hull

(2012). High school rigor and good advice: Setting up students to succeed. Alexandria, VA: The Center for Public Education.

23.

Klopfenstein

(2004). The advanced placement expansion of the 1990s: How did traditionally underserved students fare? Education Policy Analysis Archives, 12. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.14507/epaa.v12n68.2004

24.

Lichter

D. T.

Graefe

D. R.

(2011). Rural economic restructuring: Implications for children, youth, and families. In Smith

K. E.

Tickamyer

A. R.

(Eds.), Economic restructuring and family well-being in rural America (pp. 25-39). University Park: Pennsylvania State University Press.

25.

Morgan

Maneckshana

(2000). AP students in college: An investigation of their course-taking patterns and college majors (ETS Statistical Report No. 2000-09). Princeton, NJ: Educational Testing Service.

26.

Nguyen

(2015). The effectiveness of online learning: Beyond no significant difference and future horizons. Journal of Online Learning and Teaching, 11, 309-319.

27.

Nitta

Holley

M. J.

Wrobel

S. L.

(2010). A phenomenological study of rural school consolidation. Journal of Research in Rural Education, 25(2), 1-19.

28.

Rose

Betts

J. R.

(2001). Math matters: The links between high school curriculum, college graduation, and earnings. San Francisco: Public Policy Institute of California.

29.

Sadler

P. M.

Tai

R. H.

(2007). Advanced Placement exam scores as a predictor of performance in introductory college biology, chemistry and physics courses. Science Educator, 16(2), 1-19.

30.

Sathre

C. O.

Blanco

C. D.

(2006). Policies and practices at postsecondary institutions. In Western Interstate Commission for Higher Education (Ed.), Accelerated learning options: Moving the needle on access and success (pp. 25-36). Boulder, CO: Western Interstate Commission for Higher Education.

31.

Schneider

(2009). Privilege, equity, and the Advanced Placement program: Tug of war. Journal of Curriculum Studies, 41, 813-831.

32.

Speroni

(2011). Determinants of students’ success: The role of advanced placement and dual enrollment programs (An NCPR Working Paper). National Center for Postsecondary Research. Retreived from http://ccrc.tc.columbia.edu/media/k2/attachments/role-advanced-placement-dual-enrollment.pdf

33.

Stephenson

L. G.

(2014). College to high school: Kentucky’s dual enrollment alternative. New Directions for Community Colleges, 2014(165), 7-16. doi:10.1002/cc.20086

34.

Thomas

Marken

Gray

Lewis

(2013). Dual credit and exam-based courses in U.S. public high schools: 2010-11 (first look). Washington, DC: Institution of Education Sciences, National Center for Education Statistics.

35.

Weissbourd

(2016). Turning the tide: Inspiring concern for others and the common good through college admissions (Making Caring Common Project). Harvard Graduate School of Education. Retrieved from http://mcc.gse.harvard.edu/files/gse-mcc/files/20160120_mcc_ttt_report_interactive.pdf?m=1453303517