Dually Noted: Examining the Implications of Dual Enrollment Course Structure for Students’ Course and College Enrollment Outcomes

Abstract

Dual enrollment (DE)—where students earn college credits during high school—is expanding rapidly. To facilitate DE, institutional actors across K-12 schools and colleges must build or repurpose structures across separate organizations to determine course offerings, assignments, modality, and composition. Yet the organization and implications of those structures remain a black box. Using statewide administrative data from Texas, we describe DE coursetaking and course characteristics for traditional Texas public high school students and examine how DE course characteristics predict students’ course performance and subsequent college enrollment. Our descriptive analyses illuminate striking differences between academic and career and technical education DE, both in students’ backgrounds and course structures. Our regression analyses illustrate how several malleable DE course characteristics predict student outcomes.

Keywords

dual enrollment dual credit community college high school course characteristics college enrollment

Dual enrollment (DE)—in which students take college courses during high school—presents an opportunity for students to enroll in and complete college requirements before high school graduation, accelerating progress toward postsecondary credentials. More than 80% of public high schools in the United States offer DE coursework (also referred to as dual credit), and about one third of high school students complete a DE course (Shivji & Wilson, 2019; Taie & Lewis, 2020). Most DE courses are provided through public 2‐year colleges in partnership with local school districts (Mehl et al., 2020). Using administrative data from a state with high rates of DE participation, we examine the organization of DE programming, highlighting common structures through which students experience DE and estimating the relationship between those course characteristics and student outcomes.

Growing evidence suggests that DE positively predicts college enrollment, persistence, and completion (An & Taylor, 2019; What Works Clearinghouse [WWC], 2017). Many colleges and their K–12 partners seek to leverage dual enrollment to improve postsecondary access and attainment, viewing it as—in contexts with subsidized DE—a more affordable alternative for students compared with other college acceleration models like Advanced Placement (AP) or International Baccalaureate (IB) (Belfield et al., 2023; Hemelt & Swiderski, 2022; Xu et al., 2021). At the same time, policymakers, education leaders, and instructors have little information about which DE course designs predict student success. DE requires distinct educational organizations to work together toward aligned goals, building new structures or repurposing existing structures across separate organizations based on collective decision‐making (DiMaggio & Powell, 1983; Levi Martin, 2000).

Because high schools and their partner colleges draw on the same “suppliers” (instructors) and “clients” (students) to enable high school students to earn college credit, DE constitutes an interorganizational field (Levi Martin, 2000). To facilitate DE, institutional actors often make decisions about DE course offerings, assignments, modality, and composition based on instructor availability, institutional resources, and convenience (Fink et al., 2022). To date, there is minimal evidence about how DE is structured and what those structures mean for students—that is, the organization of DE and “how DE works” is still a black box—much less the magnitude of students experiencing varied DE structures. Organizational contexts and structures shape how students experience schooling (Barr & Dreeben, 1983; Dallavis & Berends, 2023). DE structures include different DE course type (e.g., academic DE vs. career and technical education [CTE] DE), course location (college campus, high school, or elsewhere), instructional modality (online or face-to-face), and instructor affiliation (college faculty vs. qualified high school teachers).

In this study, we examine DE course structures and how they predict outcomes among students in the most common model of DE—the à la carte model where traditional public high school students select discrete college courses from offerings through their local community college (College Board, 2017; Mehl et al., 2020). To illuminate how DE coursework is organized, we constructed cohorts of recent high school entrants using statewide administrative data in Texas to examine DE course enrollments, structural elements, and outcomes. We differentiate between students enrolled in academic and CTE DE courses in our analyses, addressing the following research questions (RQs):

What are the demographic and academic backgrounds of public high school students enrolled in academic and CTE DE courses through Texas community colleges? What are their coursetaking and college enrollment patterns?

How are academic and CTE DE courses structured, including instructional modality, course location, subject, and instructor assignments and characteristics?

How do course characteristics for academic and CTE DE courses predict student outcomes?

Our findings contribute to the literature in several ways. First, although several studies in the DE literature used state administrative data to examine student outcomes (e.g., Giani et al., 2014; Hemelt & Swiderski, 2022; Liu et al., 2020; Miller et al., 2017), prior research captured both students taking DE courses in traditional comprehensive public schools and in more prescriptive DE curricula through Early College High Schools (ECHSs) or charter schools. The approach makes it hard to understand the DE coursetaking patterns of students under the common à la carte DE model. In contrast, we focus explicitly on students engaged in the most common DE partnerships: traditional public high schools with public 2‐year colleges. Second, by focusing explicitly on DE participants, we describe common DE structures, mapping the DE field. We then analyze how those DE course structures predict students’ course outcomes and subsequent college enrollment behavior. This differs from prior literature, which typically compared either the course outcomes of DE students to current college students (i.e., those taking college courses not through DE) or the college enrollment outcomes of DE students to nonparticipants (e.g., An, 2013; Britton et al., 2019; Karp et al., 2007; Miller et al., 2017). This is the first study, to our knowledge, to examine the associations between a comprehensive set of DE course characteristics and student outcomes. Finally, we examine differences in student characteristics and coursetaking patterns, course and instructor characteristics, and student outcomes across DE course type—or curricular tracks—by running separate descriptive and inferential analyses for academic and CTE DE. We illustrate variation in DE course structures and outcomes across DE course types, showing that academic DE is more exclusive and linked to stronger college outcomes.

Combined, our results can inform policies, as well as programmatic decisions and practices, and forge new directions for research on DE, which more frequently focuses on whether exposure to any DE predicts student outcomes rather than acknowledging variation in DE course types and structures. Our study lays the groundwork for understanding the organization of DE coursework and how DE course structures predict student success.

The Organization of DE Programming

Most students gain access to DE courses through K–12 and college partnerships whereby students select college courses from available offerings through their high school and a local college (College Board, 2017). The à la carte approach to DE stands in contrast to ECHS programs, which are immersive and provide coherent college course sequences and student support services (Edmunds et al., 2020). Quasi-experimental evidence confirms positive effects on college‐going and completion of college‐level courses among all DE participants, not just those concentrated in ECHS programming (An, 2013; Britton et al., 2019; Giani et al., 2014). However, DE course offerings and course structures vary greatly across—and within—institutions (Fink et al., 2017). The way DE is implemented, including the subject offered, course location and modality, and instructor assigned, likely shapes student experiences and, potentially, outcomes.

Organizational contexts of schooling, including how classes are structured, have been linked to student outcomes for decades (e.g., Barr & Dreeben, 1983; Dallavis & Berends, 2023; Gamoran, Nystrand, et al., 1995; Gamoran, Secada, et al., 2000). Within-school differences, including distinct curricular pathways and access to resources, have consequences for student learning and performance (Domina et al., 2023). In the case of an interorganizational field like DE, there is potential variation both across and within DE course type. In our review of the literature, we explore (a) curricular distinction in DE, in terms of academic and CTE DE, course subjects, and timing (and therefore intensity) of DE credit accrual; (b) how DE courses are typically structured through course locations, instructional modalities, and course composition (mix of students enrolled); and (c) the role instructors, including their institutional affiliations, play in DE.

Curricular Distinction in DE: DE Course Type, Subject, and Timing

Students taking á la carte DE may experience various DE curricular pathways, diverging across DE course types (academic DE vs. CTE DE), subject areas, and timing and intensity of DE coursetaking throughout the high school career. Most DE course enrollments are in academic DE courses rather than in CTE DE, although half of public high schools in the country offer CTE DE courses (Thomas et al., 2013). Noncollege CTE coursework likely serves as a common alternative to CTE DE, but high school students in CTE DE experience an increased probability of high school graduation and college matriculation compared to their peers in noncollege CTE coursework, even after controlling for their background (Karp et al., 2007). Academic DE courses, on the other hand, are more likely to overlap and compete with college acceleration offerings like AP or IB; in fact, the number of AP courses available at a school is inversely correlated with the number of DE participants (Xu et al., 2021). Recent evidence from Tennessee suggests that expanding CTE DE offerings rather than academic DE offerings increased DE participation among underserved student populations but did not translate to increases DE course completion (Hemelt & Swiderski, 2022). Given variation in student backgrounds, course structures, and student outcomes across academic and CTE DE courses, research examining how DE implementation shapes student outcomes needs to differentiate between DE course types.

The majority of students in Texas who enroll in DE take academic general education courses, which should transfer across institutions, particularly for students planning to enroll in a public college after high school (Miller et al., 2017). More so than electives, DE courses taken in core academic subjects—especially in math and science—appear to positively predict students’ postsecondary outcomes (Giani et al., 2014). For example, taking DE college algebra, a common prerequisite for STEM math coursework, is positively associated with college success (e.g., Heavin, 2020; Hemelt et al., 2020; Minaya, 2021; Speroni, 2011).

School contexts likely shape DE coursetaking patterns beyond participation, given variation in DE course offerings (including breadth and number of courses) across DE partnerships. For example, when Tennessee schools worked to implement a statewide DE initiative, most public high schools offered only one DE course, and students subsequently took one DE course on average (Hemelt & Swiderski, 2022). Intensity of DE credit accrual appears to predict college outcomes (e.g., Miller et al., 2017; Troutman et al., 2018), but timing of DE coursetaking may also predict subsequent college enrollment and credential attainment. Earlier and more intensive exposure to college credits—for adequately prepared students—may increase students’ probability of earning a credential by increasing academic momentum (An & Taylor, 2019). At the same time, students can only take DE coursework that is made available to them (i.e., it must be available at their school, and they must have met eligibility requirements). To date, we are not aware of research that has explored patterns of DE coursetaking related to timing, particularly in the á la carte DE context.

Course Location, Modality, and Composition

Instructional conditions like location, modality, and course composition serve as “organizational resources” (broadly defined, in the tradition of organizational theory) essential to how DE students—and instructors—experience the course (Fink et al., 2023; Gamoran et al., 2000). Nationally, approximately three quarters of high school entrants in 2009 who participated in DE took those courses at a high school, with 17% taking them on a college campus and 8% through online education (Shivji & Wilson, 2019). Exposure to community college faculty and campuses during DE may improve students’ acclimation to and readiness for college (Karp, 2012; Speroni, 2011). In a mixed‐methods study, Lile et al. (2018) found that DE students who took DE courses on the college campus were better able to describe the college‐student role and responsibilities and more likely to self‐identify as a college student. The link between DE course location and student outcomes, however, is unclear, with mixed evidence across contexts (D’Amico et al., 2013; Speroni, 2011). Recent estimates using nationally representative data leveraged an inverse probability weighting technique to address self‐selection into courses and found no difference, on average, in immediate outcomes (such as enrolling in college) for students who took a DE course at a college campus compared with those at a high school, though more affluent students benefitted more from taking DE courses at a college campus than their lower‐socioeconomic‐status (SES) peers (Hu & Chan, 2021). Extant research does not explore, to our knowledge, whether taking a DE course at the college campus predicts subsequent enrollment at a community college (particularly the DE host college), a practical concern that may be of interest to colleges, as DE may serve to recruit subsequent college enrollees (Fink et al., 2023).

DE courses may also vary in instructional modality, with online DE offerings increasing rapidly both due to growth of online education in K–12 and higher education (Barnett & Stamm, 2010). Evidence regarding the consequences of instructional modality for DE students’ course outcomes is limited in scope; most research in this area relies on data from a small number of campuses and/or unadjusted means to evaluate the effects of instructional modality on DE student outcomes (e.g., Arnold et al., 2017; Holian et al., 2014; Lochmiller et al., 2016). Prior research on instructional modality for the broader population of community college students suggests that taking college courses online, compared with face‐to‐face, is associated with negative short‐term outcomes, like lower course persistence and grades (e.g., Xu & Jaggars, 2011), with some positive longer‐term associations with associate degree attainment and transfer to a 4‐year college (Ortagus, 2018). In the only large‐scale study to date examining DE course modality and student outcomes, Liu et al. (2020) found that Florida DE students who took the majority of their DE courses face-to-face—whether on or off campus—compared with online experienced slightly larger improvements in high school graduation, immediate college enrollment, and subsequent persistence in the 1st year of college, though results varied across race/ethnicity. Because the analyses were at the student level, with instructional modality captured as a proportion of DE coursework taken, the study cannot estimate the relationship between instructional modality of a given course and course‐specific outcomes.

In implementing DE coursework, practitioners must decide about mixed‐course composition, or whether to include both DE students and college‐only students in the same course, decisions that may be linked to other factors like instructor and space availability and preferred class size. In the only study (to our knowledge) examining the role of peer effects on course outcomes at community colleges, Liu and Xu (2022) found that the percentage of DE students enrolled in a community college course negatively predicts academic performance among non-DE students. The study did not, however, examine how mixed‐composition classrooms influence DE students’ outcomes.

Instructors of DE Courses

Staff are also essential organizational resources for implementing DE. Due to the complexity of finding staff coverage for DE coursework (i.e., working across two organizations), instructor assignments can be shaped by instructor availability and accreditation requirements rather than student needs or alignment with high‐demand pathways (Fink et al., 2022). Both across and within states, some community colleges rely primarily on their own faculty to teach DE, and others rely on qualified high school teachers (Mehl et al., 2020). The existing research on how instructors’ institutional affiliation predicts DE students’ outcomes is largely descriptive. Results range from no relationship between instructor affiliation and course grade or subsequent college enrollment (Dixon & Slate, 2014; Taylor & Yan, 2018) to a positive correlation between taking DE math with a high school instructor—compared with college instructor—and GPA (Hébert, 2001).

Qualitative research portrays high school instructors as more lenient than college faculty. Students describe high school DE instructors as offering them additional time and opportunities to redo assignments or to earn extra credit; college instructors, on the other hand, treat students more like adults, imposing fewer rules with less handholding (Duncheon, 2020; Edwards et al., 2011). High school teachers who implement DE coursework often receive instruction from both the college (which is accountable for the college‐level coursework) and their school, but their school's expectations may contradict those of the college and thus undermine implementation and course rigor (Duncheon & Relles, 2020). Anticipating differing standards across instructor affiliation, postsecondary faculty have voiced concern over the rigor of DE coursework taught by high school teachers—a common tension in states seeking to increase DE course availability while maintaining course rigor (Troutman et al., 2018). Such concerns evoke the tensions between “signaling” and skill development, where DE coursework may signal preparedness for college or employment, build skills toward such preparedness, or both (Spence, 1973).

Research on traditional‐age community college students (i.e., those who attend soon after high school) demonstrates that faculty characteristics, including faculty contract type, predict student performance in introductory college courses (e.g., Ran & Sanders, 2020; Ran & Xu, 2019). Taking a course with non‐tenure‐track (NTT) faculty—compared with tenure‐track (TT) faculty— positively predicts course grades but negatively predicts subsequent milestones like enrollment and performance in the next course; students also benefit more from full‐time faculty than from part‐time faculty (Ran & Sanders, 2020; Ran & Xu, 2019). It is unclear how these findings translate to DE students, though they may be particularly applicable to those experiencing DE on college campuses and taught by college instructors, further demonstrating the need to also consider location and instructor affiliation.

To better build an understanding of the DE interorganizational field and how DE coursework is structured, we describe DE coursetaking and DE course characteristics for traditional Texas public high school students who took DE courses through offerings at public 2‐year colleges. We illustrate how students participate in academic and CTE DE (e.g., the types of courses taken and when in their high school career they took these courses) and describe typical course structures. We then examine how DE course and instructor characteristics predict course completion, grades, and subsequent college enrollment.

Study Contexts

Texas leads the nation in terms of DE participation; in fall 2021, Texans comprised 16% of all DE students in the country, whereas they comprise 10% of all K-12 and college students (De Brey et al., 2021). Under Texas law, all districts are required to offer students the opportunity to earn a minimum of 12 college credits, but these credits can be offered through AP, IB, or DE (Texas Education Code Sec. 28.009, 2015). Students are eligible to participate in academic DE courses if they meet the minimum requirements for college‐level coursework; these standards can be met through achieving minimum proficiency scores on the Texas Success Initiative (TSI) assessment, high school state assessments (e.g. a passing score on their algebra and English II STAAR end‐of‐course test), and standardized tests like the SAT or ACT testing scores (Dual Credit Requirements, 2003). For entry‐level CTE DE courses (i.e., those specific to a short‐term certificate), students do not need to demonstrate college readiness to enroll, but they must meet college‐readiness standards for courses specific to higher credentials (Dual Credit Requirements, 2003).

DE course design and partnerships are largely free from government oversight; school districts and partnering colleges must, per state law, define how they will administer DE instruction and support services, including course location, class composition, faculty assignments and evaluation, and funding (Dual Credit Requirements, 2003). Although colleges must offer approved general education (i.e., “core”) college‐level coursework, which courses they offer and how those are structured is up to institutional partners (Miller et al., 2017). State policy dictates that colleges can receive state subsidies (since current funding is enrollment based) for students earning college credit in core curriculum, CTE, and foreign language DE courses, but leaves it to local control to determine how to cover the costs to students; colleges are allowed—but not required—to waive some or all tuition and fees for DE courses (Dual Credit Requirements, 2003). As a result, there is variation across DE partnerships, where some colleges discount students’ tuition and others require students to pay full cost (Texas Higher Education Coordinating Board [THECB], 2018).

In Texas, all DE instructors teaching academic courses must meet the same standards as college faculty: They must have a bachelor's degree plus 18 hours of graduate credits in the discipline of the subject they teach (Southern Association of Colleges and Schools Commission on Colleges [SACSCOC], 2018). For CTE DE, colleges have some flexibility and can substitute instructors’ relevant work experience for years of education; the college is responsible for justifying and documenting instructor qualifications and for evaluating whether DE instructors meet the institutions’ instructional standards (SACSCOC, 2018).

Data and Methods

To answer our research questions, we used statewide administrative data provided by the Texas Education Research Center (ERC), a research center and data clearinghouse at the University of Texas at Austin, and descriptive and regression analyses. Focusing on cohorts of traditional public high schoolers, we identified dual enrolled students—those enrolled in degree‐bearing college-level courses during high school—who took DE courses through public 2‐year colleges, which account for over 96% of DE students in the state.

Data

To create the analytic sample, we first identified public high school students who entered ninth grade in 2015 or 2016 and took at least one DE course through a Texas community college within 4 years of entry, by 2019 and 2020, respectively (N = 160,493). We then focused on students who attended traditional high schools (N = 125,315), excluding students who attended charter schools, ECHSs, or other schools that integrate high-school- and college‐level courses throughout their course sequences. We also restricted the analytic sample to students with scores on required state tests for Algebra I and English II (N = 120,812), which enabled us to include the test scores as proxies for academic readiness in our regression models.¹ In order to capture instructor characteristics in our models, we had to exclude DE courses from summer terms, since the ERC data only included instructor information for fall and spring. Lastly, we further restricted the analytic sample to students who attended high school in school districts with 10 or more DE students, maintaining a threshold of students in each district in order to control for variation across schools in our analytic models using fixed effects. The final analytic sample captured 497,399 DE course enrollments among 108,256 public high school students between fall 2015 and spring 2020.

Our analytic data set is structured at the student–course level, with separate observations by student–course. A given student is present in the data more than once if they took more than one DE course, which allows us to distinguish between DE course characteristics in our main regression analyses, as described later in the analytic strategy section. Given the overlap in students across DE courses, we present descriptive statistics in two distinct ways: (a) for students enrolled in DE courses (separately by DE course type (academic or CTE) and (b) for unique DE course sections. Cutting the data in these two different ways offers insights about the characteristics of students taking DE (the student‐level descriptives) and the structure of DE courses (the course‐level descriptives).

Variables

Our main independent variables of interest include DE course characteristics and instructor characteristics. For a full description of the variables used in our analyses, along with descriptive statistics for the full analytic sample, see Appendix Table A1 in the online version of the journal. In each of our analyses, we delineated two DE course types—academic and CTE. We captured DE class size, semester credit hours, broad course subject areas, class composition (indicating whether course enrollments include a mix of high school and college students), and whether the course was lecture‐based (as opposed to a lab or independent study). We also captured DE course location—at a high school, community college, or another location² (including at a multi‐institution teaching center or system center)—and instructional modality. In addition to DE course characteristics, we captured the grade in which students took DE courses, enabling us to understand their DE coursetaking patterns during high school.

We included characteristics of DE course instructors, including gender, race/ethnicity, age, instructor type (which captures institutional affiliation, as well as instructor rank and employment intensity), educational attainment, and 9‐month salary. When a course listed more than one instructor for a given DE course, we determined instructor of record by considering educational degree attainment, faculty teaching time for the course, and percentage of appointment related to instruction. For instructor type, we captured instructor affiliation by matching employee identification numbers from the postsecondary data with those from the K–12 data. We considered an instructor to be a high school teacher if they were concurrently employed at a school and college (following precedent of Miller et al., 2017). We also included a host of student demographic and academic background measures, which we used as descriptors of the DE students and as statistical controls in our regressions.

To capture student performance in a given DE course, we created two course outcome measures: (a) passing the DE course (receiving a final grade of A, B, C, or P, for pass) and (b) course grade (numeric grade captured on a 4‐point scale). To measure subsequent success, we captured three college enrollment outcomes within 1 year of high school graduation: (a) enrollment in any Texas postsecondary institution, including both public and private institutions; (b) enrollment in a Texas public 4‐year university; and (c) enrollment in a Texas public 2‐year college.³

In the online appendices, we present and interpret results for several additional outcomes. Appendix B presents findings for two other college enrollment outcomes that represent subsets of the main outcomes: returning to the DE host college of the given DE course and enrolling in a Texas private college. Appendix C provides supplemental analyses predicting longer‐term outcomes, including developmental education (dev-ed) participation and college GPA from the 1st year of college, and—among students who did not enroll in college—employment within 1 year after high school.

Analytic Strategy

To understand which traditional public high school students participate in DE through community colleges (RQ1) and what those DE courses look like (RQ2), we leveraged descriptive statistics. To examine which variables predict DE student outcomes (RQ3), we performed a series of logistic regressions for our dichotomous dependent variables, which include passing the DE course and enrolling in college post–high school. We used ordinary least squares (OLS) regression to predict numerical course grade. We ran separate regression models for each DE course type (academic or CTE) to estimate the relationships between course and instructor characteristics and student course outcomes.

For our main analytic models, to address RQ3, we had to grapple with the cross‐classified nature of the data. In analyzing our student‐course data set, we acknowledge that analytic variables can be classified at two distinct and hierarchical levels (i.e., student level and course level), where students are nested in DE courses, but courses are also nested in students. To address this issue, we analyzed the student‐course data primarily at the course level— where we included the structure and characteristics of DE coursework as the focal independent variables in our analytic models—but also employed robust clustered standard errors at the individual student level to account for correlation between repeated observations (i.e., multiple courses) within each student.

For our dichotomous outcomes, we used the following logit model for student i in cohort j at district k in semester t:

Logit (p_{ijkt}) = b_{0} + b_{1} X_{1} + b_{2} X_{2} + \dots . + b_{n} X_{n} + ξ_{j} + θ_{k} + λ_{t},

where p_ijkt is the probability of a discrete outcome occurring, b₀ is the intercept, X₁ –X_n are the independent variables, b₁– b_n are the associated regression weights, ξ_j is cohort fixed effects, θ_k is district fixed effects, and λ_t is semester fixed effects. Independent variables include student demographic and academic measures, DE course characteristics, and DE instructor characteristics. For college enrollment outcomes, we also included passing the DE course as an independent variable. Given that DE offerings and experiences may vary across schools and time, we included fixed effects for school district, cohort, and semester in all models to address this endogeneity (Cameron & Miller, 2015). We used robust cluster‐adjusted standard errors with individuals as the clustering variable to further account for within‐individual error correlation and heteroskedasticity, given the nesting of courses within individuals (Angrist & Pischke, 2009; Cameron & Miller, 2015). In the analysis for numerical course grade, we leveraged OLS regression with the same independent variables.

Limitations and Robustness Checks

Given the observational nature of our data, a regression with rich covariates is our strongest analytic strategy for examining which course features predict student success. We included a variety of statistical control variables and performed several robustness checks; nevertheless, the estimated relationships could still partially be explained by unobserved factors (e.g., motivation, social networks, instructional quality). Thus, the results are correlations that partially reflect the sorting of students into specific DE courses; that is, some students are more inclined to enter a specific type of course than others, and those unobserved characteristics may also predict subsequent academic outcomes. However, since we are comparing within a sample of students enrolled in different DE course types (separate analytic samples of academic DE students and CTE DE students), we reduce selection bias related to selection into DE more generally, compared with studies using nonparticipants as a counterfactual to DE participants (as is common in most of the research on DE). At the same time, we do acknowledge that there are important selection mechanisms for enrollment in academic DE versus CTE DE, and our separate regression models do not allow us to compare across those subsets of students. Despite these limitations, the results stand to inform the literature on DE implementation and the state of knowledge about DE more generally.

We ran a series of supplementary analyses to assess the sensitivity of our results to modeling choices. First, to test whether our main results are sensitive to the students with a higher number of DE course observations, we ran supplemental analyses—described and presented in Appendix D in the online version of the journal—where we assigned each students’ overall weight in the regression equal to one (so each student‐course observation ultimately adds up to one, lowering the weight of each individual course observation for students with multiple DE courses). We also ran supplemental models, presented in online Appendix E, with high school fixed effects to account for variation across individual schools (this was our initial preference but leads to a considerable loss in the number of schools—results do not appear sensitive to using school campus instead of district fixed effects). Next, as described in online Appendix F, we ran supplementary regression models without the indicator of passing the DE course to consider whether passing the course inadvertently explained away the association between DE course characteristics and outcomes of interest, but we only observed minor differences in results.

Another challenge related to our data availability is the difficulty of disaggregating between location and instructor type, which are captured through separate measures. Despite overlap between some conditions—for example, taking a course on a college campus and having an instructor affiliated with the college rather than high school, we could not condense the measure into one variable in our models due to considerable variation in the sample. To further explore the relationship between instructor type and course outcomes, we present the interaction between instructor type and course location in supplemental analyses in Appendix Table A2 in the online version of the journal. We show that CTE DE students experienced a decrease in their probability of passing the course and in their course grade when enrolled in a DE course with college faculty on a college campus (or another location) compared with a course with a high school teacher on a high school campus, whereas students in academic DE experienced a slight increase in their probability of passing the course (but a somewhat lower final course grade) in conditions with college faculty at a college (or other non-high-school) location compared with a course taught by a high school teacher at a high school campus.

Results

Characteristics and Coursetaking Patterns of Academic and CTE DE Students

To address the first research question, Table 1 presents means and standard deviations of student characteristics, coursetaking patterns, and college enrollment outcomes for DE students at traditional public high schools, broken down by type of DE courses taken (academic DE, CTE DE, or both). The bulk of students in our analytic sample—81%—took only academic DE courses (whom we refer to as academic DE students), with 12% taking only CTE DE (CTE DE students) and 7% taking a mix of academic and CTE DE courses (academic-CTE DE students). We observed clear differences in the demographic composition of students across DE course type. Women comprised 60% of academic DE students and only 39% of CTE DE students. Hispanic students comprised the majority of CTE and academic-CTE DE students. Black students appear overrepresented among CTE DE students and underrepresented among academic and academic-CTE DE students, whereas White and Asian students were disproportionately enrolled in academic and academic-CTE DE coursework. Two thirds of CTE DE students were from low‐income families, compared with 32% of academic DE students.

Table 1

Descriptive Statistics of DE Students by Coursetaking Type

	DE Student by Coursetaking Type
Variable	Academic DE (% or M)	CTE DE (% or M)	Academic-CTE DE (% or M)	All High School Students (% or M)
Student characteristics
Female	59.8%	38.8%***	58.3%*	48.8%
White	48.7%	27.5%***	39.3%***	30.0%
Black	6.6%	9.6%***	5.8%*	12.8%
Hispanic	37.3%	59.9%***	50.4%***	50.7%
Asian	4.8%	1.4%***	2.7%***	3.7%
Other	2.6%	1.6%***	1.8%***	2.8%
Low-income student	32.1%	62.6%***	45.4%***	54.6%
AP/IB participant	68.8%	30.2%***	57.7%***	44.4%
Algebra I test score	4,385	3,977***	4,314***	4,072
English II test score	4,453	3,997***	4,352***	4,112
DE coursetaking patterns
Number of courses taken	4.6	2.8***	7.7***
Took first DE course in 9th grade	1.7%	3.6%***	2.9%***
Took first DE course in 10th grade	3.7%	13.2%***	7.9%***
Took first DE course in 11th grade	37.8%	38.1%	38.0%
Took first DE course in 12th grade	56.8%	45.1%***	51.3%***
Passed course	91.6%	86.6%***	91.3%**
DE course grade	3.2	3.1***	3.3***
College enrollment after high school
Enrolled in any college in Texas	77.0%	41.4%***	76.1%	43.5%
Enrolled in a Texas public university	43.2%	10.2%***	39.4%***	19.3%
Enrolled in a Texas public two‐year college	34.2%	30.9%***	38.1%***	23.5%
Student N	87,669	13,054	7,533	724,825

Note. Total N (DE student) = 108,256. The first three columns of the table outline student characteristics, DE coursetaking patterns, and college enrollment outcomes of DE students, reported at the student level, with p-values comparing CTE DE and academic-CTE DE students to academic DE students (column 1, reference). The final column shows the demographic characteristics and college‐enrollment outcomes for all high school students (N = 724,825) in the same timeframe. We provide means for continuous variables and percentages for categorical measures.

p < .05, **p < .01, ***p < .001.

Academic backgrounds and coursetaking patterns also varied across DE course types. Average STAAR (State of Texas Assessments of Academic Readiness) scores—measures of academic readiness—were lower among CTE DE students than academic or academic-CTE DE students. Over two thirds of academic DE students participated in AP or IB (with academic-CTE DE students close behind), whereas fewer than a third of CTE DE students did. Academic-CTE DE students, although the smallest subgroup, took substantially more DE courses, with almost eight on average, than academic and CTE DE students, who took an average of approximately five and three DE courses, respectively. We also illustrate the timing of DE coursetaking during students’ 4 years in high school. In all three groups, students predominantly experienced their first DE course in 11th and 12th grade, but once again we see differences across DE course types. Approximately 17% of CTE DE students took the first DE course in 9th or 10th grade compared with 5% and 11% of academic and academic-CTE DE students, respectively.

In terms of DE course outcomes, we see high passing rates in all three groups of students, though it was slightly lower among CTE DE students than among students enrolled in any academic DE courses. The average DE course grade looked fairly similar across DE course types, with students earning just above 3.0 (equivalent to a B). College enrollment outcomes, however, differed across DE course type: Students in academic DE courses were much more likely to subsequently enroll in college than students in only CTE DE, though they were not much more likely than CTE DE students to enroll in a 2‐year college.

How Are DE Courses Structured, and Who Teaches the DE Courses?

To address RQ2, we present averages and percentages for course and instructor characteristics of all unique DE sections in Table 2, shown separately for academic and CTE courses. Instructional modality appears to differ across DE course types; although both academic and CTE courses were more likely to be offered face‐to‐face (vs. online or hybrid), far fewer CTE DE sections (only 10%) were offered online than academic DE sections (34%). More than half of CTE DE sections (56%) were offered on college campuses, with a third on high school campuses (33%) and 12% at other locations. The course location for academic DE courses was more varied, with 46% offered on college campuses, 32% on high school campuses, and 22% at other locations. The majority of DE sections for both course types were comprised exclusively of high school students, but a greater portion of academic DE courses than CTE DE courses were mixed composition. Most academic DE course sections were in humanities and liberal arts, whereas the bulk of CTE DE sections were in industry/agriculture/manufacturing/construction.

Table 2

Descriptive Statistics of DE Courses by Course Type

	DE Course Type
Variable	Academic DE (% or M)	CTE DE (% or M)
Course characteristics
Class size	22.2	12.6
Number of credits	3.1	3.2
Lecture section	86.7%	85.1%
Instructional modality
Face-to-face	64.0%	87.6%
Online	34.2%	10.2%
Hybrid	1.8%	2.2%
Course location
On high school campus	32.2%	32.5%
On college campus	46.3%	55.7%
At other location	21.5%	11.8%
Mixed course composition	45.4%	29.8%
Broad course subject
Humanities, liberal arts, and general studies	50.8%	2.1%
Social and behavioral sciences	24.6%	3.3%
STEM (science, technology, engineering, and mathematics)	21.7%	20.1%
Education	0.6%	0.3%
Business	0.5%	8.5%
Health	0.2%	19.2%
Industry/agriculture/manufacturing/construction	0.7%	33.7%
Service-oriented	0.9%	12.8%
Instructor characteristics
Female	56.4%	41.4%
Race/ethnicity
White	70.1%	57.5%
Black	5.9%	7.3%
Hispanic	15.3%	26.9%
Asian	3.4%	0.9%
Other	5.3%	7.5%
Age	47.8	47.8
Instructor type
High school teacher	37.0%	46.2%
TT/tenured	11.9%	8.0%
Full-time NTT	25.2%	32.1%
Part-time NTT	19.7%	11.2%
Unknown	6.4%	2.4%
Highest education level
Associate degree or less	3.7%	45.5%
Bachelor’s degree	1.9%	32.0%
Master’s degree	77.9%	20.5%
Doctoral degree	16.5%	2.0%
Calculated 9‐month salary	$31,355	$25,318
Unique course section N	60,060	11,401

Note. N (course) = 71,461. The table describes characteristics of DE courses and instructors, reported at the course level. We provide means for continuous variables and percentages for categorical measures. Columns 1 and 2 show results for academic DE courses and CTE DE courses, respectively.

Table 2 also presents instructor characteristics and instructor type, which captures institutional affiliation and contract type, for academic and CTE DE courses. The descriptive statistics highlight differences in instructors across DE course type. Fifty-six percent of academic DE course sections were taught by female instructors, compared with only 41% of CTE DE sections. White faculty taught approximately 70% of academic DE course sections and 58% of CTE DE course sections. Instructor type differed across DE course types. Forty-five percent of academic DE course sections were taught by NTT community college faculty (most of whom worked full-time), 37% were taught by high school teachers, and 12% were taught by TT or tenured faculty. A larger proportion—46%—of CTE DE courses were taught by high school teachers, with 43% taught by NTT college faculty (again, primarily those working full-time) and only 8% by TT/tenured faculty. Most academic DE instructors held a graduate degree, whereas the most prevalent level of educational attainment among CTE DE instructors was an associate degree or less.

Regression Results: Course Outcomes and Subsequent College Enrollment

We next turn to results for a series of regression models predicting DE course and college enrollment outcomes, run separately for academic and CTE DE course enrollments, in order to address RQ3. For ease of interpretation, we present results for logistic regressions as average marginal effects (AMEs), which can be interpreted as the change in the predicted probability of the outcome for each additional unit of the independent variable.

Predictors of DE Course Passing and Grade

Table 3 shows the estimated relationships between DE course characteristics and course outcomes for both DE course types. We begin by describing common predictors of DE course outcomes for both academic and CTE DE coursetakers. Taking the DE course before 12th grade, for example, is negatively associated with course outcomes in both academic and CTE DE courses; students who took a DE course in 9th, 10th, or 11th grade experienced a decrease in course grade and in their probability of passing the course compared with students who took it in 12th grade. Mixed-course composition (compared with DE with only DE students) and taking DE online (compared with face-to-face) are negatively correlated with course outcomes in both DE course types. For example, academic DE students who enrolled in an online or a hybrid section, compared with face‐to‐face, experienced a 4.2- and 1.9-percentage-point decrease in the probability of passing the course, respectively (AME = –.042, SE = .002, p < .001; AME = –.019, SE = .004, p < .001).

Table 3

Results for Regression Models Predicting DE Course Outcomes

	Passed the Course		Course Grade
Variable	Academic DE AME (SE)	CTE DE AME (SE)	Academic DE Coefficient (SE)	CTE DE Coefficient (SE)
Timing of DE coursetaking
School grade (ref. = in 12th grade)
In 9th grade	−.093***(.009)	−.219***(.031)	−.299***(.045)	−.709***(.115)
In 10th grade	−.075***(.008)	−.128***(.019)	−.238***(.033)	−.459***(.078)
In 11th grade	−.038***(.003)	−.049***(.008)	−.191***(.017)	−.199***(.041)
DE course characteristics
Class size	.000***(.000)	.000(.000)	.001***(.000)	−.004***(.001)
Number of credits	−.002(.001)	−.001(.002)	−.050***(.005)	−.059***(.007)
Lecture section	−.005(.002)	.030***(.007)	−.036***(.009)	−.111***(.021)
Instructional modality (ref. = face-to-face)
Online	−.042***(.002)	−.038***(.011)	−.110***(.008)	−.392***(.035)
Hybrid	−.019***(.004)	.007(.012)	−.067***(.013)	−.110*(.050)
Course location (ref. = on high school campus)
On college campus	−.009***(.002)	.004(.006)	−.011***(.007)	−.055**(.017)
At other location	−.008***(.002)	.022*(.009)	−.034***(.009)	.103**(.037)
Mixed-course composition	−.029***(.002)	−.035***(.005)	−.036***(.005)	−.055**(.018)
Broad course subject (ref. = humanities, liberal arts, general studies)
Social and behavioral sciences	−.001(.001)	−.047**(.016)	−.016***(.004)	−.070(.045)
STEM	−.043***(.002)	−.012(.011)	−.153***(.005)	−.010(.034)
Education	.002(.007)	.082**(.017)	.201***(.036)	.479***(.104)
Business	−.014*(.009)	−.007(.013)	.000(.038)	−.043(.039)
Health	−.057***(.014)	−.065***(.013)	−.039(.052)	−.121**(.038)
Industry/agriculture/manufacturing/construction	.004(.007)	.019(.011)	.055(.028)	.016(.035)
Service-oriented	.016*(.005)	.007(.012)	.172***(.030)	−.021(.038)
DE course instructor characteristics
Female	−.001(.001)	−.005(.005)	−.002(.003)	−.069***(.015)
Race/ethnicity (ref. = White)
Asian	−.011***(.003)	.030(.017)	−.088***(.010)	.090(.073)
Black	−.007***(.002)	.018*(.007)	−.009(.008)	−.043(.027)
Hispanic	−.001(.002)	.006(.006)	.055***(.006)	.044*(.019)
Other	−.007***(.002)	.018*(.007)	.009(.007)	−.042(.025)
Age	.000***(.000)	.000*(.000)	.000(.000)	.001*(.001)
Instructor type (ref. = high school teacher)
TT/tenured	−.011***(.002)	−.086***(.011)	−.011(.009)	−.391***(.033)
Full-time NTT	−.016***(.002)	−.063***(.006)	−.036***(.007)	−.273***(.022)
Part-time NTT	−.018***(.001)	−.057***(.008)	.025***(.005)	−.267***(.025)
Unknown	−.012***(.003)	−.007(.012)	−.011(.013)	−.032(.056)
Highest education level (ref. = associate or less)
Bachelor’s degree	.014***(.004)	.013**(.004)	.059***(.014)	−.029(.015)
Master’s degree	.004(.003)	.004(.005)	.015(.009)	−.042*(.018)
Doctoral degree	−.002(.003)	.011(.013)	−.029**(.011)	.012(.050)
Calculated 9‐month salary	.000***(.000)	.000(.000)	.000***(.000)	.000(.000)
Sample size	440,571	53,956	426,788	52,563

Note. The table presents regression results, and each column represents a separate regression model. We used logistic regression for course passing and OLS regression for numerical letter grades captured on a 4‐point scale. The top grade is an A, which equals 4; the lowest grade is an F, which equals 0; and the other grades are B, C, and D. All models included the following student characteristics: gender, race/ethnicity, low‐income status, AP or IB participation, a z-score for their Algebra I test score, and a z-score for their English II test score. All models also included cohort, semester, and district fixed effects and used robust standard errors clustered by individual students. We present AMEs and standard errors (SEs) for each covariate included in the binary logistic regression models. The first two analyses included the entire sample, and the subsequent analyses included students who earned numerical course grades. The sample size across outcomes varies slightly due to the inclusion of both semester and district fixed effects, where some districts with no variation in a given outcome during a given term were dropped from those analyses. For ease of interpretation of the sample, the means for the outcomes of interest in each of the four regressions are as follows: passed the academic DE course: 0.915; passed the CTE DE course: 0.884; grade in the academic DE course: 3.164; grade in the CTE DE course: 3.196.

p < .05, **p < .01, ***p < .001.

Course location also predicts DE course outcomes, but the patterns are mixed across the two DE course types. In academic DE courses, taking the course somewhere other than the high school (at the college or another location) is negatively associated with passing the course and with course grade. For students in CTE DE courses, taking the course at the college rather than at the high school is also negatively correlated with course grade; however, taking the CTE DE course at another location (not high school or college) positively predicts both passing the course and course grade compared with taking it at the high school.

DE instructor affiliation and educational attainment are also correlated with students’ course outcomes. For both DE course types, taking a DE course with an instructor from the college—compared with a high school teacher—is generally associated with a lower probability of passing the course and a lower course grade. Students in academic DE course sections with full‐time NTT instructors experienced a 1.6-percentage-point decrease in their probability of passing the course and a 0.04-unit decrease in final grade, compared with students in sections with high school teachers (AME = –.016, SE = .002, p < .001; B = –.036, SE = .007, p < .001). The only exception is that, for academic DE courses, taking DE with a part‐time NTT faculty—rather than a high school teacher—is positively associated with course grade, although it negatively predicts passing the course. We also observe mixed relationships between the instructor's degree attainment and course outcomes. Taking a DE course with an instructor with a bachelor's degree, compared with an associate degree or below, positively predicts passing the course for both academic and CTE DE. However, our results suggest that taking a DE course with an instructor with a graduate degree negatively predicts course grade.

Predictors of Subsequent Postsecondary Education After High School

Table 4 presents regression results for college enrollment outcomes, including enrolling in any Texas postsecondary institution, in a Texas public university, and in a Texas public 2‐year college after high school. (See supplemental results for enrolling in a private college or at the DE host college in Appendix B in the online version of the journal.) The models are the same as those run for DE course outcomes, except that we added a measure of whether students passed the DE course as a predictor.

Table 4

Results for Regression Models Predicting College Enrollment in Texas After High School

	Enrolled in Any College		Enrolled in Public University		Enrolled in 2–Year College
Variable	Academic DE AME (SE)	CTE DE AME (SE)	Academic DE AME (SE)	CTE DE AME (SE)	Academic DE AME (SE)	CTE DE AME (SE)
Passed the DE course^a	.119***(.003)	.100***(.008)	.152***(.004)	.037***(.006)	−.015***(.003)	.083***(.008)
Timing of DE coursetaking
School grade (ref. = in 12th grade)
In 9th grade	−.191***(.054)	−.311***(.078)	−.064(.052)	−.191**(.054)	−.174***(.044)	−.258**(.067)
In 10th grade	−.106***(.031)	−.211***(.055)	.001(.035)	−.142**(.048)	−.138***(.032)	−.195***(.053)
In 11th grade	−.056***(.014)	−.124***(.028)	−.007(.018)	−.076*(.031)	−.073***(.019)	−.125***(.031)
DE course characteristics
Class size	.000(.000)	.000(.000)	.000(.000)	.000(.000)	.000(.000)	.000(.000)
Number of credits	.004(.002)	−.010**(.003)	.010***(.003)	−.003(.002)	−.007**(.003)	−.010**(.003)
Lecture section	−.003(.004)	.032**(.011)	−.005(.005)	.001(.008)	.007(.005)	.044***(.011)
Instructional modality (ref. = face-to-face)
Online	.012***(.003)	.011(.017)	.022***(.004)	−.005(.012)	.000(.004)	−.005(.017)
Hybrid	.004(.006)	.005(.023)	−.003(.007)	−.010(.019)	.000(.007)	.034(.025)
Course location (ref. = high school campus)
On college campus	.009*(.004)	.009(.010)	−.004(.004)	−.005(.008)	.015***(.004)	.030**(.011)
At other location	.002(.004)	.022(.021)	−.007(.005)	.023(.016)	.011*(.005)	.009(.021)
Mixed course composition	.006*(.002)	.047***(.009)	−.002(.003)	.005(.007)	.011***(.003)	.052***(.010)
Broad course subject (ref. = humanities, liberal arts, general studies)
Social and behavioral sci	−.003(.001)	.006(.025)	.001(.002)	−.009(.018)	−.006***(.002)	.027(.026)
STEM	.000(.002)	−.032(.019)	.017***(.002)	.006(.013)	−.012***(.002)	−.019(.020)
Education	−.036*(.015)	.128*(.058)	−.040*(.016)	.081(.051)	−.014(.015)	−.071(.063)
Business	−.026(.018)	−.029(.023)	.037(.020)	.014(.016)	−.037(.019)	−.036(.023)
Health	−.005(.017)	.032(.022)	.013(.020)	.034*(.016)	.003(.020)	.063**(.023)
Industry/agriculture/manufacturing/construction	−.018(.011)	−.102***(.021)	.001(.014)	−.055***(.014)	.003(.014)	−.053**(.021)
Service-oriented	−.098***(.014)	−.109***(.023)	−.079***(.016)	−.028(.018)	−.028(.016)	−.035(.024)
DE course instructor characteristics
Female	−.001(.001)	.005(.009)	−.005**(.002)	−.003(.007)	.002(.002)	−.005(.009)
Race (ref. = White)
Asian	.009*(.004)	−.040(.038)	.018***(.005)	−.064*(.022)	−.006(.005)	.015(.042)
Black	−.007*(.003)	.024(.015)	.003(.004)	.001(.013)	−.009*(.004)	.046**(.016)
Hispanic	.001(.003)	.010(.011)	−.001(.003)	.006(.008)	.001(.003)	.000(.011)
Other	−.001(.003)	.000(.014)	.003(.004)	−.029*(.011)	−.003(.004)	.005(.014)
Age	.000(.000)	.000(.000)	.000(.000)	−.001**(.000)	.000(.000)	.000(.000)
Instructor type (ref. = high school teacher)
TT/tenured	.006(.004)	.030(.019)	.007(.004)	.003(.015)	.000(.004)	.035(.019)
Full-time NTT	.004(.003)	.011(.013)	.002(.003)	−.009(.009)	−.001(.003)	.021(.013)
Part-time NTT	.006*(.002)	.019(.013)	.009***(.003)	−.011(.010)	−.001(.003)	.051***(.014)
Unknown	.012*(.006)	.025(.028)	.016*(.007)	−.011(.020)	.009(.007)	.052(.031)
Highest education level (ref. = associate degree or less)
Bachelor’s degree	−.010(.006)	.032***(.009)	.002(.007)	.033***(.007)	.002(.007)	.004(.010)
Master’s degree	−.005(.004)	.043***(.010)	.000(.005)	.036***(.008)	.001(.005)	.006(.011)
Doctoral degree	−.007(.004)	.016(.027)	.004(.005)	.042*(.022)	−.006(.005)	−.025(.024)
Calculated 9‐month salary	.000(.000)	.000(.000)	.000**(.000)	.000(.000)	.000(.000)	.000(.000)
Sample Size	432,818	53,221	433,531	51,367	433,655	53,338

Note. The table presents logistic regression results, and each column represents a separate regression model. All models included the following student characteristics: gender, race/ethnicity, low‐income status, AP or IB participation, a z-score for their Algebra I test score, and a z-score for their English II test score. All models also included cohort, semester, and district fixed effects and used robust standard errors clustered by individual students. We present AMEs and SEs for each covariate included in the binary logistic regression models. The six analyses included high school graduates from the entire sample. The sample size across outcomes varies slightly due to the inclusion of both semester and district fixed effects, where some districts with no variation in a given outcome during a given term were dropped from those analyses. For ease of interpretation of the sample, the means for the post‐high‐school outcomes of interest in each of the six regressions are as follows: any Texas college enrollment for academic DE: 0.817; any Texas college enrollment for CTE DE: 0.552; Texas public university enrollment for academic DE: 0.492; Texas public university enrollment for CTE DE: 0.201; Texas public 2‐year college enrollment for academic DE: 0.338; Texas public 2‐year college enrollment for CTE DE: 0.359.

“Passed the DE course” was included as an independent variable in regressions on college enrollment.

p < .05, **p < .01, ***p < .001.

Whether students passed the DE course and the timing of that course are correlated with college enrollment outcomes. Passing the DE course is associated with an 11.9- and 10.0-percentage-point increase in the probability of enrolling in any Texas postsecondary institution after high school for students in both academic and CTE DE courses, respectively (AME = .119, SE = .003, p < .001; AME = .100, SE = .008, p < .001). Course passing also appears consequential for subsequent enrollment at Texas public universities by students in both academic and CTE DE courses, while passing an academic DE course—compared with passing a CTE DE course—has a stronger relationship with the public‐university enrollment outcome. In terms of predicting enrollment at Texas public 2‐year colleges after high school, passing an academic DE course negatively predicts enrolling, whereas passing a CTE DE course positively predicts enrolling. The timing of coursetaking is also a significant predictor of college enrollment after high school. Taking a course in 9th, 10th, or 11th grade, compared with 12th grade, is negatively correlated with enrolling in any Texas college and enrolling in a Texas public 2‐year college for both DE course types; however, the relationship between the timing of coursetaking and public university enrollment appears significant for CTE DE courses only, suggesting that the negative relationship observed between early academic DE coursetaking and subsequent college enrollment is driven primarily by 2‐year college enrollment.

DE course structures such as composition of enrollees, location, and modality also predict college enrollment. Mixed course composition positively predicts enrolling in any Texas college and enrolling in a public 2‐year college, where students taking both academic and CTE DE courses with college‐only students—compared with only with other DE students—experienced increased probabilities of enrollment. Taking the DE course on the college campus—compared with at high school—also positively predicts any college and public 2‐year college enrollment, with up to a 3.0-percentage-point increase in the probability of public 2‐year college enrollment among CTE DE students (AME = .030, SE = .011, p < .004). DE instructional modality predicts a small change in college enrollment among students in academic DE courses, wherein taking the course online, compared with face‐to‐face, is associated with a 1.2- and 2.2-percentage-point increase in the probability of any college enrollment and public university enrollment, respectively (AME = .012, SE = .003, p < .001; AME = .022, SE = .004, p < .001).

DE course instructors’ characteristics, particularly instructor type and education background, are also correlated with students’ college enrollment. Students taking a DE course taught by college faculty, compared with high school teachers, generally experienced increased probabilities of college enrollment, though the positive relationships are only present for some subgroups of college instructors, primarily predicting small shifts in any college enrollment and public university enrollment for academic DE students and a larger shift in enrollment at public 2‐year colleges for CTE DE students. For example, taking an academic DE course with a part‐time NTT or unknown type of instructor (note that the “unknown” designation typically occurred at colleges with no faculty ranks) is associated with a 0.6- and 1.2-percentage-point increase, respectively, in the probability of enrolling in any college (AME = .006, SE = .002, p = .012; AME = .012, SE = .006, p = .049). Taking a CTE DE course with a part‐time NTT college instructor, compared with a high school instructor, predicts a 5.1-percentage-point increase in the probability of enrolling in a public two‐year college (AME = .051, SE = .014, p < .001). We also observe positive relationships between CTE DE instructors’ education background and enrolling in any postsecondary institution and enrolling in a public university after high school, suggesting students may benefit from CTE DE courses taught by instructors with higher educational attainment, even though CTE courses do not require instructors to have a baccalaureate.

Discussion

Participation in DE—compared with nonparticipation—appears to increase immediate entry into college and early accrual of college credits, improving students’ academic momentum and degree attainment (An & Taylor, 2019; WWC, 2017). In light of positive findings, high schools and colleges across the country have ramped up efforts to provide DE coursework, but, to do so, they often must repurpose high school and college resources and structures for this new interorganizational context. To understand the organization of DE coursework and the implications of DE course structures for student outcomes, we used student‐level administrative data following recent DE participants in Texas. We examined how students at traditional public high schools experience DE, how DE courses are commonly structured, and which DE course characteristics predict students’ course outcomes and subsequent college enrollment. The results highlight the salience of curricular distinction in DE, with striking differences in participants and outcomes across academic and CTE DE, and surprising findings regarding the timing of DE course taking in the á la carte DE context. We also find evidence for varied outcomes across DE course structures like location, modality, and course composition and across instructor type.

Curricular Tracks and Timing in DE

The demographic and academic backgrounds of students who take academic and CTE DE differ dramatically. Academic DE students were more likely to identify as White and female and much less likely to receive free or reduced‐price lunch than their CTE DE counterparts; they were also more likely to participate in other college acceleration programs, such as AP or IB. Although these differences confirm prior evidence about the privilege that exists in DE course access (e.g., Brown et al., 2018; Shivji & Wilson, 2019), they also emphasize varied access to academic DE and CTE DE. In Texas, students who participate in academic DE must meet college‐readiness standards (typically achieved through scores on end‐of‐course STAAR exams or placement tests). Students who took both CTE and academic DE courses appear similar to academic DE-only students, suggesting that meeting eligibility requirements for academic DE courses serves as a primary driver of stratification across DE course type.

Although DE students, on average, took most of their DE courses in 11th and 12th grade, more CTE DE students (17%) took their first DE courses in 9th and 10th grade, which suggests that sorting into DE course type occurs early in high school, if not before (indeed, recent research suggests that high schools begin emphasizing DE pathways as early as middle school [Fink et al., 2023]). This tracking likely aligns with high schools’ delineation between college preparation—traditional academic coursework pathways—and vocational pathways. Although course outcomes (passing and grades) look similar across DE course types, students in academic DE courses were much more likely to subsequently enroll in college (particularly at universities) than those in CTE DE courses, though direct comparisons are complicated by differential student selection into DE course type. In examining equity in DE access and outcomes, scholars and practitioners should consider stratification across DE type (and the outcomes within each type) rather than prioritizing increased access to DE overall. Future research may also evaluate additional measures of student success, given divergent pathways across DE course types (e.g., among noncollege entrants, CTE DE students were more likely to gain employment in the year after high school than academic DE students; see Appendix Table C2 in the online version of the journal).

Our regression results suggest that students who take DE courses— whether academic DE or CTE DE—earlier in high school, compared with those who take DE in their senior year, experience worse course outcomes. However, we hypothesize that stronger course outcomes in terms of both grades and passing among high school seniors may be related to programmatic support such as better guidance for college‐level courses as students progress through high school. Although we also observed lowered probabilities of enrolling in any college after high school for students taking DE before 12th grade, we did not find a significant relationship between academic DE course timing and public university enrollment, whereas the relationship with community college enrollment (and DE host college enrollment, as shown in Appendix B in the online version of the journal) is significant and negative. This suggests that, at least for academic DE, the correlation between course timing and any college enrollment is driven by lower community college enrollment; the sooner students started to take DE courses, the less likely they were to attend a public 2‐year college after high school. It is possible that the observed relationship could also be partially due to additional selection mechanisms into DE courses that are unobservable in the data. For instance, high school seniors may know which college courses are of interest to them (and thus improve their performance) and be more likely to take DE courses if planning to go to college.

To further understand these patterns, we ran descriptive statistics for DE students, broken down by when they took their first DE course (see Appendix Table A3 in the online version of the journal). Students who started taking DE later in high school were more likely to have taken academic DE courses, which further suggests that the selection of DE course type taken in later grades may be strategic—with an eye toward college plans—though it also may be related to academic preparation (students must meet college‐readiness standards to take academic DE but not CTE DE), and doing so may be more plausible later in high school. Our findings appear to be the first empirical evidence about how DE coursework timing correlates with student outcomes. However, it is important to consider the context: Our analytic sample is focused on students attending public high schools that were not using prescribed DE course sequences, as students might experience under an ECHS-style model. This area merits additional inquiry to better understand the mechanisms at play in both à la carte DE and more structured DE pathways. Without additional evidence, programs focused on expanding opportunities for college coursework early in the high school trajectory do so with little information on how it could influence student outcomes.

Combining Organizational Resources to Structure DE Courses

Institutional partners draw on organizational resources including technology, campus space, student peers, and staff when creating DE course structures. Our regression results illustrate correlations between several of those key DE course structures and students’ course and subsequent college‐enrollment outcomes. Across the board, course outcomes in DE sections that were online (and, for the most part, hybrid) are worse than in sections that were face-to-face—that is, they received lower grades and were less likely to pass. This is consistent with findings for online courses in community college settings (Xu & Jaggars, 2011). In contrast to Liu et al.’s (2020) findings based on the proportion of DE coursework students took online, we found that participating in an online academic DE course is positively associated with going to college after high school, though the relationship is small. If we extrapolate from the mean college enrollment rates of academic DE coursetakers, the estimates translate to a shift from 81.7% to 82.9% of students enrolling in any Texas college. Given the strong evidence—with much larger effect sizes—for the positive link between DE participation compared with non‐participation (see WWC, 2017),⁴ our results suggest that various modalities of DE may improve students’ subsequent enrollment in college compared with no DE at all, though that question is beyond the scope of our study. At the same time, the negative relationships with course passing and grades suggest that continued investment in face‐to‐face DE, where possible, may improve the rate of DE course completion. In some contexts, such as rural areas where students or personnel must travel a sizeable distance for in‐person courses, it may be hard to overcome hurdles to offer face‐to‐face opportunities; Fink et al. (2023) describe how some colleges use in‐person facilitators at high schools for online DE courses to mimic the accountability of face‐to‐face coursework. The effectiveness of that approach has not yet been tested.

We also observed a relationship between DE course location and outcomes. In terms of immediate course outcomes, taking DE coursework on the college campus, compared with at the high school, is associated with lower grades and a lower probability of passing. Despite the negative associations with course outcomes, taking a DE course on the college campus is positively associated with enrolling at a community college after high school. This is a particularly important insight for community college stakeholders who describe building a pipeline into the college as an incentive for increasing DE course offerings (Fink et al., 2023). Additionally, although prior research found no difference in postsecondary enrollment between students who take DE coursework at the high school versus the college using nationally representative data (Hu & Chan, 2021), we found a small positive correlation between taking an academic DE course on the college campus and enrolling in any Texas postsecondary institution, which appears to be largely driven by enrollment in community colleges. For CTE DE courses, there is a significant association only between taking the course at the college campus and public 2‐year college enrollment. It is possible that the differences between our results and Hu and Chan's (2021) findings are explained by our differentiating between academic and CTE DE courses, but the differences could also stem from something specific to the Texas context.

Our regression results similarly suggest that DE instructor affiliation predicts course outcomes and subsequent college enrollment. Taking DE with college faculty—compared with taking DE with a high school teacher—is negatively associated with course passing and final course grade. Our supplemental analyses (see Appendix Table A2 in the online version of the journal) of the interaction between instructor type and course location offer further evidence that, compared with students who took DE courses at the high school with a high school instructor, students who took CTE DE courses on the college campus with a college instructor received lower grades and were less likely to pass, and those who took academic DE on the college campus with a college instructor were somewhat more likely to pass but still experienced a small decrease in final grade among some college faculty types. Qualitative research suggests that high school DE instructors offer students more flexibility and have different standards than college faculty, which may ultimately translate to better final course grades (Duncheon, 2020; Edwards et al., 2011). Our findings may further fuel the debate over the rigor of DE courses, particularly when taught by high school instructors (Hemelt & Swiderski, 2022; Troutman et al., 2018), though it is important to acknowledge that course grades and passing are imperfect measures of student learning. In our study, despite negative course outcomes, taking DE with college faculty—particularly NTT faculty—is positively associated with college enrollment outcomes, though the results differ across academic DE and CTE DE courses. For academic DE students, taking DE with part‐time NTT faculty, compared with a high school instructor, positively predicts enrolling in any college after high school, whereas for CTE DE students, it primarily predicts enrolling at a public 2‐year college.

For CTE courses, we also observed a link between educational attainment of instructors and college‐going behavior among students. Although CTE DE courses have more flexibility in terms of the educational attainment of instructors—in Texas, years of experience in the field can substitute for postsecondary education—taking a CTE DE course with an instructor who holds a baccalaureate increased students’ probability of enrolling in college after high school. Exposure to an instructor with a 4‐year degree (or beyond) may boost students’ educational aspirations, though additional research is necessary to understand how those processes might work in CTE DE settings. Almost half of all CTE DE courses were taught by instructors who had not attained at least a bachelor's degree. In CTE fields in which a college credential is particularly important for entry into a career in that field, colleges might benefit from further emphasizing educational attainment in their hiring of CTE DE instructors, as feasible.

Building DE coursework that exposes DE students to college students may also make use of an organizational resource—a new peer group—available at the community college. For academic DE courses, students in mixed‐composition DE courses—those including both high school and college‐only students in the same section—were less likely to pass the course but more likely to enroll at a community college. Mixed-composition DE courses may be more similar to college‐only courses, so although DE students were less likely to pass them compared to when taking the course with only other high school students, their exposure to college‐only students—and the shift in the course atmosphere that exposure creates—may have encouraged students to come back to a community college (mixed course composition did not significantly predict enrollment at Texas public universities or private institutions).

Conclusion

Declining numbers in high school graduating classes, combined with increased competition from public, regional 4‐year institutions and online universities, create strong pressures for community colleges to expand the pool of eligible students. Community colleges will have to partner more intentionally with high schools to recruit, support, and engage DE students as a means to both boost current enrollment numbers and build a pipeline of subsequent enrollees. K–12 leaders also have strong incentives to improve college access, especially among underserved students, and DE is one way to achieve this goal. Our findings provide evidence about which DE course characteristics K–12 and college stakeholders can leverage to improve course outcomes and subsequent college enrollment among DE students.

The relationships we illuminate offer insights for policymakers, administrators, and practitioners seeking evidence for how to design DE courses for students and colleges. We also highlight additional areas for inquiry. Our descriptive statistics show variation in how DE coursework is offered, illustrating the complexity of designing DE courses, given that colleges and their K–12 partners work to overcome challenges of instructor availability, geographic proximity and feasibility of bringing students to the college campus, and resource constraints. Overall, our findings related to course location, instructor affiliation, and course composition suggest that when DE courses look more like college courses—that is, they are taught on the college campus and/or by a college professor and taken with at least some college‐only students—DE students appear to experience lower course‐specific grades but an increased probability of subsequent college enrollment. But additional inquiry may be necessary to determine how to set up similar conditions in the face of resource constraints (e.g., lack of college faculty, transportation or distance to college, misalignment of schedules across institutions).

We hope that by describing DE organizational structures, we set up potential comparison groups for subsequent research to examine the role of DE course design. Rather than continuing to compare DE participants to nonparticipants, future research can ideally also compare across different DE conditions, using statistical controls for student background. This approach can inform evidence‐based practices, through which educational leaders must balance the interests of institutional partners and weigh the potential influence on short- and long‐term outcomes.

Footnotes

ORCID iDs

Wonsun Ryu

Lauren Schudde

Kimberly Pack-Cosme

Notes

Wonsun Ryu is a research associate at the Population Research Center at the University of Texas at Austin. His research investigates how higher education policies work and examines their impact on student outcomes. He holds a PhD in Educational Leadership and Policy and an MS in Statistics from the University of Texas at Austin.

Lauren Schudde is an associate professor of Educational Leadership and Policy at the University of Texas at Austin. Her research examines the impact of educational policies and practices on college student outcomes, with a primary interest in how higher education can be better leveraged to ameliorate socioeconomic inequality in the United States. She earned her PhD in sociology from the University of Wisconsin-Madison, 1912 Speedway D5400, Austin, TX 78712; e-mail: schudde@austin.utexas.edu.

Kimberly Pack-Cosme is a doctoral student in Educational Leadership and Policy at the University of Texas at Austin. Her research primarily uses quantitative methods to explore policies and practice that impact student success outcomes and alleviate socioeconomic disparities in postsecondary education, stratification within college and its effects on social mobility, and the educational pipeline from high school through graduate education.

References

B. P.

(2013). The impact of dual enrollment on college degree attainment: Do low-SES students benefit? Educational Evaluation and Policy Analysis, 35(1), 57–75. https://doi.org/10.3102/0162373712461933

B. P.

Taylor

J. L.

(2019). A review of empirical studies on dual enrollment: Assessing educational outcomes. In Paulsen

Perna

(Eds.), Higher education: Handbook of theory and research (Vol. 34, pp. 99–151). Springer. https://doi.org/10.1007/978-3-030-03457-3_3

Angrist

J. D.

Pischke

J.-S.

(2009). Mostly harmless econometrics: An empiricist's companion. Princeton University Press.

Arnold

Knight

Flora

(2017). Dual enrollment student achievement in various learning environments. Journal of Learning in Higher Education, 13(1), 25–32. https://files.eric.ed.gov/fulltext/EJ1139694.pdf

Barnett

Stamm

(2010). Dual enrollment: A strategy for educational advancement of all students. Blackboard Institute. https://doi.org/10.7916/D81G0KNQ

Barr

Dreeben

(1983). How schools work. University of Chicago Press.

Belfield

Jenkins

Fink

(2023). How can community colleges afford to offer dual enrollment college courses to high school students at a discount?Community College Research Center. https://ccrc.tc.columbia.edu/publications/community-colleges-afford-dual-enrollment-discount.html

Britton

Chelliah

Symns

Campbell

(2019). College now . . . or later: Measuring the effects of dual enrollment on postsecondary access and success (EdWorkingPaper No. 19-118). Annenberg Institute. https://doi.org/10.26300/2wb0-ka92

Brown

Dalton

Laird

Ifill

(2018). Paths through mathematics and science—Patterns and relationships in high school coursetaking (NCES 2018-118). National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2018118

10.

Cameron

A. C.

Miller

D. L.

(2015). A practitioner's guide to cluster‐robust inference. Journal of Human Resources, 50(2), 317–372. https://doi.org/10.3368/jhr.50.2.317

11.

College Board. (2017). College credit in high school. https://secure-media.collegeboard.org/pdf/research/college-credit-high-school-working-group-report.pdf

12.

Dallavis

J. W.

Berends

(2023). Charter schools after three decades: Reviewing the research on school organizational and instructional conditions. Education Policy Analysis Archives, 31 (1), 1–30. https://doi.org/10.14507/epaa.31.7364

13.

D’Amico

M. M.

Morgan

G. B.

Robertson

Rivers

H. E.

(2013). Dual enrollment variables and college student persistence. Community College Journal of Research and Practice, 37(10), 769–779. https://doi.org/10.1080/10668921003723334

14.

De Brey

Snyder

T. D.

Zhang

Dillow

S. A

. (2021). Digest of education statistics 2019 (NCES 2021-009). National Center for Education Statistics. https://eric.ed.gov/?id=ED611019

15.

DiMaggio

P. J.

Powell

W. W.

(1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147–160. https://doi-org.ezproxy.lib.utexas.edu/10.2307/2095101

16.

Dixon

D. B.

Slate

J. R.

(2014). Differences in selected dual credit courses’ success rates by location for Texas college students. International Journal of University Teaching and Faculty Development, 5(3), 175–185.

17.

Domina

Penner

A. M.

Penner

E. K.

(2023). Schooled and sorted: How educational categories create inequality. Russell Sage Foundation.

18.

Dual Credit Requirements, 19 Tex. Admin. Code § 4.85. (2003). https://texreg.sos.sta te.tx.us/public/readtac$ext.TacPage?sl=R&app=9&p_dir=&p_rloc=&p_tloc=&p_ ploc=&pg=1&p_tac=&ti=19&pt=1&ch=4&rl=85

19.

Duncheon

J. C.

(2020). “We are exposed to that college environment”: Exploring the socialization of early college high school students. Community College Review, 48(2), 173–194. https://doi.org/10.1177/0091552119898880

20.

Duncheon

J. C.

Relles

S. R.

(2020). “We’re caught in between two systems”: Exploring the complexity of dual credit implementation. Review of Higher Education, 43(4), 989–1016. https://doi.org/10.1353/rhe.2020.0028

21.

Edmunds

J. A.

Unlu

Furey

Glennie

Arshavsky

(2020). What happens when you combine high school and college? The impact of the early college model on postsecondary performance and completion. Educational Evaluation and Policy Analysis, 42(2), 257–278. https://doi.org/10.3102/0162373720912249

22.

Edwards

Hughes

K. L.

Weisberg

(2011). Different approaches to dual enrollment: Understanding program features and their implications. The James Irvine Foundation. https://ccrc.tc.columbia.edu/publications/different-approaches-to-dual-enrollment.html

23.

Fink

Fay

Gilliard

Griffin

Jenkins

Schudde

(2022, April 4). From “random acts” and “programs of privilege” to dual enrollment equity pathways. The CCRC Mixed Methods Blog. Community College Research Center. https://ccrc.tc.columbia.edu/easyblog/introducing-dual-enrollment-equity-pathways .html

24.

Fink

Griffin

Tulloch

A. G.

Jenkins

Fay

Ramirez

Schudde

Steiger

(2023). Deep insights: Redesigning dual enrollment as a purposeful pathway to college and career opportunity. Community College Research Center. https://files.eric.ed.gov/fulltext/ED631297.pdf

25.

Fink

Jenkins

Yanagiura

(2017). What happens to students who take community college “dual enrollment” courses in high school?Community College Research Center and National Student Clearinghouse. https://ccrc.tc.columbia.edu/publications/what-happens-community-college-dual-enrollment-students.html

26.

Gamoran

Nystrand

Berends

LePore

P. C.

(1995). An organizational analysis of the effects of ability grouping. American Educational Research Journal, 32(4), 687–715. https://doi.org/10.3102/00028312032004687

27.

Gamoran

Secada

W. G.

Marrett

C. B.

(2000). The organizational context of teaching and learning: Changing theoretical perspectives. In Hallinan’s

(Ed.), Handbook of the sociology of education (pp. 37–63). Springer.

28.

Giani

Alexander

Reyes

(2014). Exploring variation in the impact of dual‐credit coursework on postsecondary outcomes: A quasi‐experimental analysis of Texas students. High School Journal, 97(4), 200–218. http://doi.org/10.1353/hsj.2014.0007

29.

Heavin

K. S.

(2020). Bridging the gap: Effects of dual credit college algebra on postsecondary education outcomes [Doctoral dissertation]. University of Kentucky. https://doi.org/10.13023/etd.2020.405

30.

Hébert

(2001). A comparison of learning outcomes for dual‐enrollment mathematics students taught by high school teachers versus college faculty. Community College Review, 29(3), 22–38. https://doi.org/10.1177/009155210102900302

31.

Hemelt

S. W.

Schwartz

N. L.

Dynarski

S. M.

(2020). Dual-credit courses and the road to college: Experimental evidence from Tennessee. Journal of Policy Analysis and Management, 39(3), 686–719. http://doi.org/10.1002/pam.22180

32.

Hemelt

S. W.

Swiderski

(2022). College comes to high school: Participation and performance in Tennessee's innovative wave of dual‐credit courses. Educational Evaluation and Policy Analysis, 44(2), 313–341. https://doi.org/10.3102/01623737211052310

33.

Holian

Alberg

Strahl

J. D.

Burgette

Cramer

(2014). Online and distance learning in southwest Tennessee: Implementation and challenges (REL 2015-045). National Center for Education Evaluation and Regional Assistance. https://ies.ed.gov/ncee/rel/Products/Region/appalachia/Publication/3614

34.

Chan

H.-Y.

(2021). Does delivery location matter? A national study of the impact of dual enrollment on college readiness and early academic momentum. Teachers College Record, 123(4), 1–32. https://doi.org/10.1177/016146812112300401

35.

Karp

M. M.

(2012). “I don't know, I've never been to college!” Dual enrollment as a college readiness strategy. New Directions for Higher Education, 2012(158), 21–28. https://doi.org/10.1002/he.20011

36.

Karp

M. M.

Calcagno

J. C.

Hughes

K. L.

Jeong

D. W.

Bailey

T. R.

(2007). The postsecondary achievement of participants in dual enrollment: An analysis of student outcomes in two states. National Research Center for Career and Technical Education.

37.

Levi Martin

. (2000). What is field theory? American Journal of Sociology, 109(1), 1–49. https://doi.org/10.1086/375201

38.

Lile

J. R.

Ottusch

T. M.

Jones

Richards

L. N.

(2018). Understanding college‐student roles: Perspectives of participants in a high school/community college dual‐enrollment program. Community College Journal of Research and Practice, 42(2), 95–111. https://doi.org/10.1080/10668926.2016.1264899

39.

Liu

(2022). Happy together? The peer effects of dual enrollment students on community college student outcomes. American Educational Research Journal, 59(1), 3–37. https://doi.org/10.3102/00028312211008490

40.

Liu

V. Y. T.

Minaya

Zhang

(2020). High school dual enrollment in Florida: Effects on college outcomes by race/ethnicity and course modality. Community College Research Center. https://ccrc.tc.columbia.edu/publications/dual-enrollment-florida-race-ethnicity-course-modality.html

41.

Lochmiller

C. R.

Sugimoto

T. J.

Muller

P. A.

Mosier

G. G.

Williamson

S. E.

(2016). Dual enrollment courses in Kentucky: High school students’ participation and completion (REL 2016-137). National Center for Education Evaluation and Regional Assistance. https://ies.ed.gov/ncee/rel/Products/Region/appalachia/Publication/3745

42.

Mehl

Wyner

Barnett

E. A.

Fink

Jenkins

(2020). The dual enrollment playbook: A guide to equitable acceleration for students. Aspen Institute and Community College Research Center. https://ccrc.tc.columbia.edu/publications/dual-enrollment-playbook-equitable-acceleration.html

43.

Miller

Kosiewicz

Wang

E. L.

Marwah

E. V. P.

Delhommer

Daugherty

(2017). Dual credit education in Texas: Interim report. RAND Corporation. https://www.rand.org/pubs/research_reports/RR2043.html

44.

Minaya

(2021). Can dual enrollment algebra reduce racial/ethnic gaps in early STEM outcomes? Evidence from Florida. Community College Research Center. https://ccrc.tc.columbia.edu/publications/dual-enrollment-algebra-stem-outcomes.html

45.

Ortagus

J. C.

(2018). National evidence of the impact of first‐year online enrollment on postsecondary students’ long‐term academic outcomes. Research in Higher Education, 59(8), 1035–1058. https://doi.org/10.1007/s11162-018-9495-1

46.

Ran

F. X.

Sanders

(2020). Instruction quality or working condition? The effects of part‐time faculty on student academic outcomes in community college introductory courses. AERA Open, 6(1), 1–18. https://doi.org/10.1177/2332858420901495

47.

Ran

F. X.

(2019). Does contractual form matter? The impact of different types of non‐tenure‐track faculty on college students’ academic outcomes. Journal of Human Resources, 54(4), 1081–1120. https://doi.org/10.3368/jhr.54.4.0117.8505R

48.

Shivji

Wilson

(2019). Dual enrollment: Participation and characteristics (NCES 2019-176, Data Point). National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2019176

49.

Southern Association of Colleges and Schools Commission on Colleges (SACSCOC). (2018). Faculty credentials. https://sacscoc.org/app/uploads/2019/07/faculty-credentials.pdf

50.

Spence

(1973). Job market signaling. Quarterly Journal of Economics, 87(3), 355–374. https://doi.org/10.2307/1882010

51.

Speroni

(2011). Determinants of students’ success: The role of advanced placement and dual enrollment programs. National Center for Postsecondary Research. https://ccrc.tc.columbia.edu/publications/role-advanced-placement-dual-enrollment.html

52.

Taie

Lewis

(2020). Dual or concurrent enrollment in public schools in the United States. (NCES 2020-125, Data Point). National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2020125

53.

Taylor

J. L.

Yan

(2018). Exploring the outcomes of standards‐based concurrent enrollment and Advanced Placement in Arkansas. Educational Policy Analysis Archives, 26(123), 1–25. https://doi.org/10.14507/epaa.26.3647

54.

Texas Education Code Sec. 28.009. (2015). College credit program. https://statutes.capitol.texas.gov/Docs/ED/htm/ED.28.htm#28.009

55.

Texas Higher Education Coordinating Board (THECB). (2018). Overview: Dual enrollment. Author. https://reportcenter.highered.texas.gov/agency-publication/miscellaneous/dual-credit/

56.

Texas Higher Education Coordinating Board (THECB). (n.d.). Distance education modifications & requests. https://www.highered.texas.gov/our-work/supporting-our-institutions/academic-program-resources/distance-education-resources/distance-education-modifications-resources/

57.

Thomas

Marken

Gray

Lewis

(2013). Dual credit and exam‐based courses in U.S. public schools: 2010–11 (NCES 2013-001). National Center for Education Statistics. https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2013001

58.

Troutman

D. R.

Hendrix-Soto

Creusere

Mayer

(2018). The University of Texas System dual‐credit study: Dual credit and success in college. The University of Texas System, Institutional Research and Analysis. https://utsystem.edu/documents/docs/ut-system-reports/2018/dual-credit-and-success-college

59.

What Works Clearinghouse (WWC). (2017). Dual enrollment programs: Transition to college. Institute of Education Sciences. https://ies.ed.gov/ncee/wwc/EvidenceSnapshot/671

60.

Jaggars

S. S.

(2011). The effectiveness of distance education across Virginia's community colleges: Evidence from introductory college‐level math and English courses. Educational Evaluation and Policy Analysis, 33(3), 360–377. https://doi.org/10.3102/0162373711413814

61.

Solanki

Fink

(2021). College acceleration for all? Mapping racial gaps in Advanced Placement and dual enrollment participation. American Educational Research Journal, 58(5), 954–992. https://doi.org/10.3102/0002831221991138