The Role and Influence of Exclusively Online Degree Programs in Higher Education

Abstract

This study leverages national data and a quasi-experimental design to examine the influence of enrolling in an exclusively online degree program on students’ likelihood of completing their degree. We find that enrolling in an exclusively online degree program had a negative influence on students’ likelihood of completing their bachelor's degree or any degree when compared to their otherwise-similar peers who enrolled in at least some face-to-face courses. The negative relationship between exclusively online enrollment and students’ likelihood of bachelor's degree completion was relatively consistent among White, Black, Hispanic, Asian, low-income, and military students. Findings focused solely on those students enrolled in exclusively online degree programs revealed that the negative influence of exclusively online enrollment was exacerbated when the student attended a for-profit 4-year institution.

Keywords

online education exclusively online degree programs degree completion

Introduction

Colleges and universities have been criticized for low completion numbers, particularly among racially minoritized and historically underserved subgroups of students (e.g., Mfume, 2018). The completion gap between Black and White students has increased from 16 to 20 percentage points at public universities and 8 to 15 percentage points at public community colleges (National Center for Education Statistics, 2021). Completion rates among historically underserved students can vary considerably according to the type of institution attended. The 6-year graduation rate for Black and Hispanic students who began college in Fall 2012 was substantially higher at public or private nonprofit 4-year universities (between 42.9% and 57.1%) than for-profit 4-year institutions (between 14.2% and 28.1%) (Ortagus & Hughes, 2021). These differential completion patterns reveal the need to further explore strategic approaches designed to increase access and improve educational attainment for historically underserved subgroups of students.

Online education, particularly exclusively online degree programs, has been identified as a viable strategy to increase the number of students who graduate from college (Sener, 2012). More specifically, online education can accommodate time- or location-constrained students by offering learning and student services in an online setting that accommodates learners in ways that do not align with the rigid class times and office hours associated with residential, face-to-face education (Bouchey et al., 2021; Goodman et al., 2019; LeBlanc, 2013). The proportion of college students enrolling in online courses has grown considerably over the past two decades, increasing from 5.9% in 2000 to 42.9% in 2016 (Ortagus, 2017). Among college students who enroll in online courses, 24.4% enroll in exclusively online degree programs (authors’ calculations using National Postsecondary Student Aid Study data).

For colleges and universities, exclusively online degree programs may represent a way to reduce costs and enhance revenue in ways that make financial sense. More specifically, Deming et al. (2015) reported that institutions with a higher percentage of exclusively online students are able to leverage advances in online learning technology to be able to charge lower tuition prices, suggesting that exclusively online degree programs can “bend the cost curve” in higher education. Additional work has indicated that exclusively online degree programs can generate substantial increases in the net revenue of colleges and universities when the online courses are offered at larger enrollment levels (Cheslock et al., 2016; Morris, 2008). Despite the growing prevalence and financial promise of online education in higher education, the quality of online degree programs has been called into question in recent years.

A fundamental question facing colleges and universities is whether online degree programs have a democratizing effect by opening new doors to higher education or a diversionary effect by attracting students who may have performed better in a face-to-face learning environment. Although online education has the potential to remove barriers and increase access to higher education, previous studies have revealed mixed results regarding the effects of online enrollment on college students’ academic outcomes (Ortagus, 2020; Xu & Xu, 2020). However, prior research focuses primarily on comparing the short-term outcomes (e.g., exam scores and course grades) of online students to those of face-to-face students in the same type of class (Bowen et al., 2014; Figlio et al., 2013; Johnson & Mejia, 2014). This work is typically focused on a single class at a single institution and does not speak to the relationship between exclusively online degree programs and degree completion, which represents the most important outcome for students, administrators, and policymakers.

Past research on the effectiveness of online education typically compares the academic outcomes of students enrolled in an online course relative to those in the face-to-face version of the same course but, there may be important differences between students who enroll in an exclusively online degree program and students enrolling in a mixture of face-to-face and online coursework. Online courses often require stronger self-directed learning skills than in-person coursework, and students in entirely online degree programs may be more dependent on these skills than students also enrolled in face-to-face instruction (Allen & Seaman, 2014; Bambara et al., 2009; Guglielmino & Guglielmino, 2003; Xu & Xu, 2020). Additional work has shown that students underestimate the challenges of online learning when entering college (Bork & Rucks-Ahidiana, 2013). Performance reduction in online learning environments can impact certain populations of online learners, such as Black, Hispanic, and low-income students, at higher rates than their peers (Xu & Jaggars, 2014) due in part to the systemic disadvantages these populations encounter in schooling prior to postsecondary enrollment (Xu & Xu, 2020).

To examine the influence of enrolling in an exclusively online degree program on degree completion, we address the following research questions:

Research Question 1: To what extent does enrolling in an exclusively online degree program influence students’ likelihood of degree completion?

Research Question 2: Do results vary according to students’ demographic or background characteristics?

Research Question 3: Among exclusively online students, what is the relationship between enrollment at a specific institution type and students’ likelihood of degree completion?

Literature Review

Prior research on postsecondary online education has typically reported negative or mixed effects of online courses on students’ academic outcomes. Students in face-to-face instruction have earned higher course grades than their peers in online courses in several experimental studies (Alpert et al., 2016; Figlio et al., 2013; Joyce et al., 2015). While course grade is the most common measure of student performance, other research has highlighted subsequent outcomes of online learning, including course repetition, subject persistence, and college persistence. Many past studies on the effects of online course-taking focus on community college students (Hart et al., 2018; Huntington-Klein et al., 2017; Krieg & Henson, 2016; Shea & Bidjerano, 2018). Performance in online education varies by subpopulation, and previous work has shown that racially minoritized students and academically underprepared students may be at the greatest risk of performance decrement in online coursework (Xu & Xu, 2020).

The Impact of Online Education on Students’ Academic Outcomes

Numerous studies related to the provision of online education in higher education focus on comparing the academic outcomes between face-to-face and online students (Ortagus, 2018; Shea & Bidjerano, 2014; Xu & Jaggars, 2011, 2013, 2014). A meta-analysis of the empirical literature on the effectiveness of online education found no significant difference between the academic outcomes of exclusively online students and face-to-face students (Means et al., 2009), but additional studies have provided differing takeaways regarding the effect of online enrollment on the academic outcomes of college students (Ortagus, 2018; Xu & Jaggars, 2011, 2013). In general, prior work focusing on online education in higher education has identified a negative relationship between online enrollment and short-term, course-level outcomes (Xu & Jaggars, 2011, 2013) and a positive relationship between online enrollment and longer-term academic outcomes, such as completing an associate degree or transferring to a 4-year institution (Ortagus, 2018; Shea & Bidjerano, 2014).

Shea and Bidjerano (2014) used online enrollment data in 2004 and propensity score methods to find that students who enrolled in at least one online course during their 1st year of college were more likely to earn their associate degree than their peers who did not enroll in any online courses in the 1st year. Importantly, the authors did not disaggregate between students who enroll in a single online course and those who enroll in exclusively online degree programs. Additional research has leveraged quasi-experimental approaches and shown that enrolling in some, but not all, online courses had a positive influence on the likelihood of both completing an associate degree and transferring vertically from a 2- to 4-year institution (Ortagus, 2018, 2023).

Fischer et al. (2022) also found that students from a large public university in California who enrolled in an online course that was required for their major had higher rates of bachelor's degree completion and a slightly shorter time to degree. Contrary to prior work indicating a positive relationship between online enrollment and degree completion, Huntington-Klein et al. (2017) used regional and longitudinal variation in the number of high-speed Internet providers as a source of identifying variation, reporting that students at Washington State community colleges who took the online version of a comparable face-to-face course were less likely to complete their associate or bachelor's degree.

However, numerous researchers have shown that college students perform worse in individual online courses at a variety of institution types (Alpert et al., 2016; Figlio et al., 2013; Joyce et al., 2015), including for-profit institutions (Bettinger et al., 2017). Xu and Jaggars (2011, 2013) found that college students who enrolled in online courses were more likely to receive a lower grade and withdraw from their course when compared to their face-to-face peers. Additional researchers found that college students were less likely to earn an A or B or complete a given course when taking it online (Hart et al., 2018). Additional work examining students’ performance on college exit exams in Colombia found that students in online programs performed worse on their exit exams when compared to students in face-to-face programs (Cellini & Grueso, 2021).

Even though 24.4% of online students enroll in exclusively online degree programs (authors’ calculations using National Postsecondary Student Aid Study data), extant literature has yet to examine the influence of enrolling in exclusively online degree programs on students’ academic outcomes. In a qualitative study, Su and Waugh (2018) reported that the work requirements for an exclusively online degree program exceeded students’ expectations, providing further evidence that students may underestimate the academic challenges associated with online learning (Bork & Ruks-Ahidiana, 2013). Additional qualitative research revealed that student engagement and faculty-student interactions are strong predictors of academic success in exclusively online programs and should be a critical goal in developing any exclusively online degree program (Ortagus & Derreth, 2020).

Heterogeneous Impacts of Online Education

Prior literature has also examined the varying relationships between online education and academic outcomes based on students’ demographic characteristics or the discipline of the online course. Xu and Jaggars (2014) focused specifically on short-term performance gaps between face-to-face and online students, reporting that racially minoritized students, particularly Black students, were less likely to persist in online courses. While decreased course performance in online courses is present for all populations, Johnson and Meija (2014) found that racially minoritized students, male students, part-time students, and students with lower levels of academic achievement saw the largest decline in passing grades in online instruction compared to face-to-face instruction. These gaps are likely connected to the systemic disadvantages that low-income and racially minoritized students face prior to postsecondary enrollment, but more research is needed to understand the role of online education for these outcomes (Xu & Xu, 2020).

Past work has also demonstrated that the academic subject of the online class plays a role in outcomes for students (Hart et al., 2018; Xu & Jaggars, 2014). While the passing rates for online students was lower than the passing rate for face-to-face students in nearly every subject, Johnson and Mejia (2014) noted that students’ passing rates dropped between 18.5% and 19.4% for media and communications, engineering, and public and protective services courses that were offered online. Some of the performance decrement in online classes can be attributed to peer effects. Xu and Jaggars (2014) tracked how students’ online performance weakened more in classes where there were more at-risk online learners, regardless of their own identities.

Although the body of literature examining the effectiveness of online education in higher education is continually growing, little is known regarding the influence of enrolling in an exclusively online degree program or the extent to which exclusively online enrollment affects different subgroups of students in varying ways. This lack of evidence pertaining to the long-term impacts of exclusively online enrollment precludes administrators and policymakers from improving their understanding of the implications associated with increasing commitments to exclusively online degree programs.

Differential Completion Outcomes in Higher Education

Students’ background characteristics and type of institution attended can play an important role in whether they complete their degree. Regardless of medium of instruction, specific subgroups of students have been identified as less likely to graduate than their peers. Completion rates among college students are stratified by race/ethnicity and socioeconomic status, with Black, Hispanic, and low-income students completing college at significantly lower rates than their White, Asian, and higher-income peers (T. Bailey et al., 2015; Causey et al., 2022). Flores et al. (2017) found that achievement gaps among Black and Hispanic students can be attributed to circumstances occurring prior to beginning college, including lack of access to challenging coursework, secondary school racial composition, and familial immigration status.

In an analysis focused on bachelor's degree attainment, Nguyen et al. (2019) found that students from the bottom half of the income distribution had a 6-year degree attainment rate of 12.0%, whereas students in the top half of the distribution had a bachelor's degree attainment rate of 58.8%. While there is limited research on the outcomes of nontraditional students, existing data suggest that students who do not enter as 1st-year, full-time students directly after high school have lower bachelor's degree attainment (National Center for Education Statistics, 2022; Taniguchi & Kaufman, 2005). Veterans, one subgroup of nontraditional learners, as a whole have similar completion rates to other nontraditional groups, with 51.7% of enrolled veterans graduating with a certificate or degree (Ford & Vignare, 2015). However, the financial benefits earned from the federal GI Bill may support veteran persistence in unique ways (Barr, 2019).

College students attending different types of institutions also have varying outcomes. Some of the largest gaps in degree completion exist for students attending 4-year for-profit institutions (Deming et al., 2013). Lynch et al. (2010) found that only 22% of first-time full-time students at for-profit colleges earned a bachelor's degree in 6 years compared to 55% of students at nonprofit public schools and 65% of students at nonprofit private schools. Graduation rates are also lower for students at 2-year colleges. Moore and Shulock (2010) found that in the California community college system, only 31% of students had earned a certificate or degree or transferred to a 4-year institution within 6 years of enrolling. Black and Hispanic students in the cohort studied completed fewer credits on average and were less likely to earn a degree (Moore & Shulock, 2010). Past work has also found that increases in institutional selectivity and indicators of college quality are positively related to completion rates (Long, 2008; Melguizo, 2010).

Conceptual Framework

Exclusively online degree programs have the potential to increase access by removing traditional time and location constraints associated with face-to-face education. However, the benefits of exclusively online enrollment may not be distributed equally. Students who do not take a single course on campus may not have the same level of built-in, personalized, and consistent interactions with their faculty, which forces exclusively online students to rely disproportionately on self-directed learning. Previous research has reported that White students and individuals with higher levels of educational attainment fare better in learning environments that require self-directed learning when compared to Black students and individuals with lower levels of educational attainment, suggesting that the systemic disadvantages facing historically underserved students in face-to-face courses may be exacerbated in online environments (Xu & Xu, 2020).

Prior learning theories have outlined the importance of high-quality personal interactions in ways that center students’ active interactions with faculty and peers as a critical predictor of students’ sense of belonging and likelihood of academic success, particularly in the online learning environment (Anderson, 2008; Balaji & Chakrabarti, 2010). Anderson's (2008) “Theory of Online Learning” contends that any effective learning environment must prioritize multiple modalities of effective interactions between students, faculty, and the content in a given course. Online learning is thereby situated as a subset of learning in general, regardless of medium of instruction, in which faculty-student interactions and content-centered interactions outside of the traditional lecture structure are paramount in any high-quality course experience. Such interactions are made possible when the learning environment for a given course or degree program is learner-centered, knowledge-centered, assessment-centered, and community-centered.

Rovai’s (2003) composite persistence model, which combines earlier persistence models of Tinto (1975) and Bean and Metzner (1985) in order to explain postsecondary student persistence in online education, can also guide the logical rationale of this study. Previous research focused on the impact of online education in higher education has extended Rovai's composite persistence model to examine long-term academic outcomes, such as degree completion (e.g., Fischer et al., 2022; Shea & Bidjerano, 2014). As one example, Fischer et al. (2022) used Rovai's model to classify students’ academic and nonacademic characteristics as a way to offer a logical rationale for the selection of covariates, subgroup analyses, and additional empirical decisions.

Rovai’s (2003) composite model is divided into (a) students’ background characteristics prior to enrollment and (b) factors affecting students after enrollment, such as work or family responsibilities. This conceptual model synthesizes prior persistence models and considers relevant research pertaining to online students’ needs and learning styles to better explain persistence and degree completion among online students. By integrating key aspects of previous student persistence models and applying them to an online environment, Rovai informs the logical rationale of this study and provides a useful lens through which to interpret our findings. The academic and background characteristics explaining a student's likelihood of both exclusively online enrollment and subsequent degree completion are deemed as critical to better understanding whether exclusively online enrollment has democratizing or diversionary effects for student populations pursuing degree completion. The composite model described above, combined with previous literature, informs the logical rationale, selection of covariates (outlined in Table 1), and interpretation of the findings for the study.

Table 1

Descriptive Statistics by Modality and Degree Type

	Modality		Bachelor’s Degree		Associate Degree
	Online	Not Online	Completed	Did Not Complete	Completed	Did Not Complete
Student is female	.614	.549	.593	.530	.568	.553
	(.487)	(.498)	(.491)	(.499)	(.495)	(.497)
Student is Black	.213	.132	.086	.177	.115	.145
	(.409)	(.339)	(.280)	(.382)	(.319)	(.352)
Student is Hispanic	.134	.185	.123	.220	.205	.176
	(.341)	(.389)	(.328)	(.414)	(.404)	(.380)
Student is Asian	.022	.061	.079	.043	.041	.061
	(.147)	(.239)	(.270)	(.202)	(.199)	(.239)
Student is American Indian or Native American	.010	.008	.004	.012	.008	.009
	(.100)	(.091)	(.062)	(.108)	(.088)	(.093)
Student is White	.589	.571	.670	.505	.589	.569
	(.492)	(.495)	(.470)	(.500)	(.492)	(.495)
Student’s age at end of 2011	23.8	19.9	18.8	21.2	21.2	20.0
	(8.0)	(5.0)	(2.9)	(6.5)	(7.0)	(5.1)
Student was independent as of 2011–2012	.507	.159	.060	.282	.259	.178
	(.500)	(.365)	(.237)	(.450)	(.438)	(.382)
Student was low-income as of 2011–2012	.525	.318	.175	.450	.364	.332
	(.500)	(.466)	(.380)	(.497)	(.481)	(.471)
Student was first-generation as of 2011–2012	.765	.571	.382	.731	.683	.569
	(.424)	(.495)	(.486)	(.444)	(.466)	(.495)
Student was ever married as of 2011–2012	.178	.046	.017	.087	.102	.050
	(.382)	(.210)	(.129)	(.283)	(.303)	(.217)
Student had military service as of 2011–2012	.057	.019	.012	.029	.035	.020
	(.232)	(.136)	(.110)	(.169)	(.183)	(.139)
Student’s high school (HS) grade point average (GPA) 3.5–4.0	.150	.245	.387	.132	.161	.251
	(.357)	(.430)	(.487)	(.338)	(.368)	(.434)
Student’s HS GPA < 3.0	.360	.340	.185	.450	.361	.338
	(.480)	(.474)	(.388)	(.497)	(.480)	(.473)
Student’s concorded SAT/ACT (in 10s)	89.8	98.2	108.4	89.9	90.8	98.8
	(16.3)	(20.5)	(19.5)	(17.1)	(16.0)	(20.8)
Student attended full-time (FT) in 2011–2012	.607	.673	.838	.550	.650	.671
	(.489)	(.469)	(.369)	(.498)	(.477)	(.470)
Student attended part-time (PT) in 2011–2012	.266	.171	.020	.289	.175	.181
	(.442)	(.376)	(.141)	(.453)	(.380)	(.385)
Student worked FT in 2011–2012	.219	.105	.049	.161	.133	.112
	(.414)	(.306)	(.216)	(.367)	(.339)	(.315)
Student worked PT in 2011–2012	.216	.274	.310	.240	.314	.259
	(.411)	(.446)	(.463)	(.427)	(.464)	(.438)
Distance to institution (inst.) in 2011–2012 (in 100s of miles)	3.4	1.3	2.1	1.1	1.0	1.6
	(5.9)	(3.5)	(4.3)	(3.4)	(3.4)	(4.0)
Student inst. selective in 2011–2012	.152	.427	.746	.165	.111	.462
	(.359)	(.495)	(.435)	(.371)	(.315)	(.499)
Student inst. 2-year in 2011–2012	.557	.494	.161	.732	.823	.433
	(.497)	(.500)	(.367)	(.443)	(.382)	(.495)
Institution enrollment in 2011–2012 (in 100s of students)	52.1	16.1	18.5	20.1	21.1	19.1
	(97.8)	(16.8)	(22.1)	(42.5)	(46.1)	(33.0)
Student inst. public in 2011–2012	.560	.763	.646	.811	.822	.727
	(.497)	(.425)	(.478)	(.392)	(.382)	(.445)
Student inst. private in 2011–2012	.112	.192	.340	.078	.069	.209
	(.316)	(.394)	(.474)	(.268)	(.253)	(.406)
Student inst. for-profit in 2011–2012	.327	.045	.014	.112	.109	.064
	(.469)	(.208)	(.118)	(.315)	(.312)	(.245)
Student’s first program was 1 or 2 years	.691	.474	.145	.734	.824	.426
	(.462)	(.499)	(.352)	(.442)	(.381)	(.494)
Student transferred before degree completion	.323	.243	.157	.314	.179	.265
	(.468)	(.429)	(.364)	(.464)	(.384)	(.441)
Student expected bachelor's or higher	.805	.898	.992	.819	.843	.899
	(.396)	(.302)	(.089)	(.385)	(.364)	(.301)
First degree was entirely online	.217	.000	.005	.031	.024	.020
	(.412)	(.000)	(.073)	(.173)	(.153)	(.139)
Student completed a bachelor's degree	.182	.431			.254	.376
	(.386)	(.495)			(.436)	(.484)
Student completed an associate degree	.257	.163	.006	.076
	(.437)	(.369)	(.080)	(.265)
Number of observations	1,990	13,700	6,070	9,610	3,030	12,650

Note. Standard deviations in parentheses. The comparison group for students whose degree program was exclusively online as of 2017 is students who were not exclusively online in 2011, 2014, or 2017. Selective institutions are those that are moderately or very selective in BPS. Numbers of observations rounded to the nearest 10.

Background Information

The proportion of students in exclusively online degree programs has grown considerably over the past two decades, increasing from 2.5% in 2000 to 10.5% in 2016 (authors’ calculations using National Postsecondary Student Aid Study data). Among students in the Beginning Postsecondary Students Longitudinal Study for 2012–2017 (BPS:12/17) who were enrolled in an exclusively online degree program, 46.8% were in an associate degree program, 41.7% were in a bachelor's degree program, and 9.0% were in a certificate program as of 2017. However, we limit analyses, to be outlined below, to students who expected an associate or bachelor's degree.

In addition, 52.1% of students enrolled in exclusively online degree programs attended for-profit 4-year institutions, 18.6% attended public 2-year institutions, 12.4% attended private 4-year institutions, 10.4% attended public 4-year institutions, and 4.3% attended for-profit 2-year institutions. Most students in exclusively online degree programs (64.5%) attended open-access or minimally selective 4-year institutions, but 8.9% of exclusively online students attended moderately or very selective 4-year institutions. Regarding the academic majors of exclusively online students at 4-year institutions, 22.4% were in health-related majors, 21.1% of students were in business-related majors, and 16.3% of students were in STEM-related majors.

Data and Methods

This study draws national data from the BPS:12/17, which provides the most recent longitudinal tracking of a nationally representative sample of college students. Students included in BPS:12/17 enrolled initially at a college or university in 2011–2012 and participated in three rounds of data collection during their 1st, 3rd, and 6th years after beginning college. The sample size of BPS:12/17 is 22,500 college students. By using BPS:12/17 data, we are able to examine student characteristics, course-taking patterns (such as exclusively online enrollment), and academic outcomes over a 6-year period.

Key Variables

The primary outcome of this study is a given student's likelihood of degree completion for the pooled sample including all students and numerous subgroups of historically underserved students who have been found to be more likely to enroll in online degree programs or less likely to graduate from college than their peers. Degree completion is measured as associate degree completion, bachelor's degree completion, or any degree completion (an associate degree or higher). We examine subgroups according to race/ethnicity (White, Black, Hispanic, Asian), low-income status, veteran or military status, and financial independence.

In addition, we examine whether enrolling in exclusively online degree programs at specific types of institutions (e.g., public 4-year, private nonprofit 4-year, for-profit 4-year, selective 4-year, broad-access 4-year, public 2-year, and for-profit 2-year) influences students’ likelihood of degree completion, particularly among the historically underserved subgroups of students described previously. Selective 4-year institutions are defined as those that are moderately or very selective in BPS:12/17, and broad-access 4-year institutions are defined as those that are either open-access or minimally selective. We are unable to include subgroup analyses for Native American students or private nonprofit 2-year institutions due to extremely low proportions of exclusively online students. As a robustness check, we ran alternative specifications for an exploratory outcome of “dropping out” of college (i.e., no longer enrolled and never graduated).

The treatment variable is enrollment in an exclusively online degree program at any degree-granting college or university. In particular, we considered students to be treated if they were enrolled in an exclusively online degree program as of 2017. Approximately, 8.8% of students in the BPS:12/17 sample were classified as treated. The BPS:12/17 sample represents the most recent nationally generalizable data available to allow researchers to link exclusively online enrollment to students’ completion outcomes. To be clear, our treatment variable is defined at the programmatic level, but a very small proportion (fewer than 1% in 2012) of students may seek to enroll solely in online courses throughout the entirety of their college experience despite not being enrolled in an exclusively online degree program. If students graduated or left college prior to 2017, we used their most recent degree program to determine treatment status. The comparison group included students who did not enroll in exclusively online degree programs at any point during college, as we excluded students who enrolled in an exclusively online degree program during an earlier year but switched to a face-to-face program. We excluded 4.2% of students in the BPS:12/17 sample for this reason, but we ran alternative specifications in which we included these students in the control group.

Analytic Strategy

When examining the influence of online enrollment in any nonexperimental study, selection bias is a primary concern. If the college students who enroll in exclusively online degree programs differ from students who do not in ways that influence their likelihood of degree completion, we would face challenges determining whether differences in outcomes across groups are due to enrolling in the exclusively online degree program rather than preexisting differences across student subgroups. In quasi-experimental studies, the counterfactual model can be used to allow treatment and control groups to be equivalent based on a host of pretreatment characteristics. We account for individuals’ conditional probability of receiving the treatment (enrolling in exclusively online degree programs) by using inverse probability of treatment weighting (IPTW). When random assignment is not possible, IPTW represents an effective approach to reduce selection bias and achieve better balance between nonequivalent groups of data (Austin & Stuart, 2015).

After calculating the inverse probability treatment weights, we use a weighting scheme to calculate the average treatment effect on the treated (ATT) to identify the effects on those who experienced the treatment. We ran alternative specifications in which we removed units with extreme propensity scores (below 1% or above 99% and below 5% or above 95%). These alternative specifications revealed consistent results relative to the preferred specifications, which are reported below, in statistical significance, direction, and magnitude of coefficients.

In Figures 1 and 2, we demonstrate the overlap between treated and untreated units within the overall sample and the race/ethnicity subsamples by including multiple graphs showing the probability densities for online and face-to-face students, displaying the estimated likelihood of choosing each modality, before and after inverse probability of treatment weighting. When investigating subgroups, including by students’ race/ethnicity and institution sector, we recalculated IPTWs for each subgroup by using students from a given subgroup or students who enrolled at institutions at the same 2- or 4-year level. For analyses by student subgroup, we estimate propensity scores capturing the likelihood of students enrolling exclusively online. For analyses by institution type, we limit analyses to students who enrolled exclusively online and capture online students’ propensity to enroll at different institution types (probability densities for all subgroups are available upon request). In addition, results from the model used to estimate propensity scores for exclusively online enrollment and enrollment among exclusively online students at public 4-year institutions are available in Supplemental Appendix Table 1 in the online version of the journal.

Figure 1.

Probability densities for exclusively online and students enrolled in some or all face-to-face courses (before IPTW).

Figure 2.

Probability densities for exclusively online and students enrolled in some or all face-to-face courses (after IPTW).

To estimate the influence of enrolling in exclusively online degree programs on students’ likelihood of degree completion, we use various linear probability models in light of the ease of interpretability of linear regression coefficients relative to log-odds or odds ratios from logit models. Following Abadie et al. (2017), we cluster standard errors by sector of students’ institutions attended upon entering college, which constituted the strata used for sampling institutions in the underlying National Postsecondary Student Aid Study from which the BPS:12/17 cohort was drawn. Additional analyses used logit models rather than linear probability models and found consistent results in statistical significance, direction, and magnitude of coefficients. The linear probability models for this study can be represented by the following equation:

y_{i} = β_{0} + β_{1} E O_{i} + β_{2} X_{i} + ε_{i},

where $y_{i}$ is an indicator equaling one or zero for whether individual i completes an associate or bachelor's degree. The variable $EO$ is a treatment indicator that equals one if an individual enrolled in an exclusively online degree program. $X_{i}$ represents a vector of demographic variables and information on the academic performance and circumstances of individual i. The error term is represented by $ε_{i}$ .

In both the IPTW models and linear probability models, we included the following covariates: sex, race/ethnicity, age, low-income status, first-\generation status, veteran status, marital status, dependency, employment status, attendance intensity, composite SAT scores, high school grade point average (GPA), distance from the target institution, enrollment size of the institution, institutional selectivity, transfer status, and academic major. Table 1 displays means and standard deviations for model covariates and outcomes for treated and control students. According to Table 1, students whose degree program was exclusively online as of 2017 were more likely to be older, financially independent, enrolled at for-profit institutions, and living farther from their institutions relative to their peers.

To produce nationally generalizable findings, we combined IPTWs and nationally generalizable sample weights into one weight by multiplication. Both IPTWs and sample weights are probability-type quantities, and previous research has identified the combination of both weights into one regression equation through multiplication as a valid empirical approach (Guo & Fraser, 2015). In a relevant example, DuGoff et al. (2014) compared four different methods to estimate treatment effects, finding that combining a propensity score method and survey weighting was the most effective approach when seeking to achieve unbiased estimates deemed generalizable to the target population of interest.

In addition, we employed a covariate balancing approach using standardized mean differences and variance ratios in alignment with recommendations by Shadish et al. (2008) and Rubin (2001). We calculated the standardized mean differences (Cohen’s d) before and after using IPTW using the following equation: $d = ({\bar{x}}_{t} - {\bar{x}}_{c}) / \sqrt{(s_{t}^{2} + s_{c}^{2}) / 2}$ . After applying the IPTWs, Cohen's d was close to zero. We tested the variance ratio before and after the IPTW approach and specified an acceptable range near one (4/5 < v < 5/4) after using the IPTW approach. Taken together, this balance check affirms that covariates were well-balanced after employing the IPTW approach and ensuring that the standardized mean differences and variance ratios typically fell within their recommended ranges (results for all additional balance checks for individual covariates can be found in Supplemental Appendix Tables 2–5 in the online version of the journal).

Results

In this section, we present findings on the influence of exclusively online enrollment on bachelor's degree completion before considering the relationship between exclusively online enrollment and the completion of any degree (either an associate degree or a bachelor's degree). We focus initially on the pooled sample and subgroups by race/ethnicity (Table 2) prior to examining additional subgroups by low-income status, veteran status, and financial independence (Table 3). The comparison group for Tables 2 and 3 only includes students who did not enroll in an exclusively online degree program during college. For Tables 4 and 5, we shift our focus to consider only the subsample of students who enrolled in exclusively online degree programs, examining the influence of exclusively online enrollment among different types of institutions offering bachelor's degrees (Table 4) and institutions offering associate degrees (Table 5). The comparison group for Tables 4 and 5 includes students who enrolled in an exclusively online degree program at a different type of institution that offers the same level of degrees. For example, the first column of Table 4 focuses on exclusively online students at public 4-year institutions, so the comparison group includes exclusively online 4-year students who did not attend a public 4-year college or university.

Table 2

The Influence of Exclusively Online Enrollment on Degree Completion by Race/Ethnicity

Panel A. All Students
	All Students
	BA	AA	Any
Exclusively online	−.083***	.003	−.084***
degree program	(.022)	(.007)	(.018)
Number of observations	13,610	2,070	15,680
R ²	.198	.046	.119
Panel B. Subgroups by Race/Ethnicity
	Black Students			Hispanic Students			Asian Students			White Students
	BA	AA	Any	BA	AA	Any	BA	AA	Any	BA	AA	Any
Exclusively online	−.086****	.057***	−.088***	−.056*	−.031	−.087****	−.217****	—	−.209****	−.081**	−.008	−.082***
degree program	(.016)	(.014)	(.023)	(.027)	(.036)	(.014)	(.017)	—	(.018)	(.033)	(.013)	(.020)
Number of observations	2,060	360	2,420	2,590	410	3,000	720	—	750	7,520	1,170	8,690
R ²	.260	.082	.143	.237	.218	.183	.428	—	.326	.199	.071	.122

Note. All models include inverse-probability-of-treatment weights estimated for each subsample and previously described covariates. The “BA” column is limited to students who expected a bachelor's degree or higher, the “AA” column is limited to students who expected specifically an associate degree, and the “Any” column is limited to students who expected an associate degree or higher. Results for AA completion among Asian students were excluded due to an extremely low number of treated students for this specification. Standard errors clustered on the sector of the first institution attended. Numbers of observations rounded to the nearest 10.

p < .05; ***p < .01; ****p < .001.

Table 3

The Influence of Exclusively Online Enrollment on Degree Completion by Low-Income Status, Veteran Status, and Financial Independence

	Lower-Income Students			Military Students			Independent Students
	BA	AA	Any	BA	AA	Any	BA	AA	Any
Exclusively online degree program	−.089***	.014*	−.093***	−.114**	.044	−.069	−.037	.019*	−.055***
	(.019)	(.006)	(.020)	(.039)	(.206)	(.048)	(.024)	(.010)	(.017)
Number of observations	4,860	1,190	6,040	250	70	320	2,680	990	3,660
R ²	.219	.047	.124	.308	.543	.228	.141	.057	.082

Notes. All models include inverse-probability-of-treatment weights estimated for each subsample and previously described covariates. The “BA” column is limited to students who expected a bachelor's degree or higher, the “AA” column is limited to students who expected specifically an associate degree, and the “Any” column is limited to students who expected an associate degree or higher. Standard errors clustered on the sector of the first institution attended. Regression models for Military Students exclude the covariate for Asian race. Numbers of observations rounded to the nearest 10.

p < .1; **p < .05; ***p < .01.

Table 4

The Influence of Institution Type on Bachelor's Degree Completion Among Exclusively Online Students

	Public 4-Year	Private 4-Year	For-Profit 4-Year	Selective 4-Year	Broad-Access 4-Year
Exclusively online student attended institution type	.012	.076	−.119**	.081*	−.054
Exclusively online student attended institution type	(.071)	(.059)	(.043)	(.039)	(.077)
Number of observations	770	770	770	770	770
R ²	.239	.176	.145	.218	.167

Note. All models include inverse-probability-of-treatment weights estimated for each subsample and previously described covariates. Samples are limited to students who were in exclusively online degree programs as of 2017, who pursued a bachelor's degree, and who expected a bachelor's degree or higher. The regression sample includes 440 exclusively online students at for-profit 4-year institutions, 170 students at public 4-year institutions, 160 students at private 4-year institutions, and fewer than 5 each at public 2-year and for-profit 2-year institutions. Standard errors clustered on the sector of the first institution attended. Numbers of observations rounded to the nearest 10.

p < .1; **p < .05.

Table 5

The Influence of Institution Type on Associate Degree Completion Among Exclusively Online Students

	Public 2-Year		For-Profit 2-Year		For-Profit 4-Year
	Expected Only AA	Expected AA or Higher	Expected Only AA	Expected AA or Higher	Expected Only AA	Expected AA or Higher
Exclusively online student attended institution type	−.033	−.021	−.005	.022	−.034	−.035**
Exclusively online student attended institution type	(.026)	(.023)	(.069)	(.033)	(.058)	(.011)
Number of observations	260	900	260	900	260	900
R2	.249	.059	.303	.229	.134	.048

Note. All models include inverse-probability-of-treatment weights estimated for each subsample and previously described covariates. Samples are limited to students who were in exclusively online degree programs as of 2017, who pursued an associate degree, and who expected specifically an associate degree (“Expected Only AA”) or an associate degree or higher (“Expected AA or Higher”). Standard errors clustered on the sector of the first institution attended. Regression models for For-Profit 2-Year exclude the covariates for Asian and American Indian or Alaska Native race and institutional selectivity. Numbers of observations rounded to the nearest 10.

p < .05.

Table 2 shows that enrollment in exclusively online degree programs had a negative influence on bachelor's degree completion across student subgroups by race and ethnicity. When compared to students who did not enroll in online degree programs, enrolling exclusively online had a negative influence (8.3 percentage points) on bachelor's degree completion for students who expected to earn a bachelor's degree or higher (first column of panel A of Table 2). The relationship between exclusively online enrollment and bachelor's degree completion was negative and statistically significant for Black students in the first column of panel B (8.6 percentage points) and Asian students in the seventh column of panel B (21.7 percentage points). The negative relationship between exclusively online enrollment and bachelor's degree completion was marginally statistically significant for Hispanic students in the fourth column of panel B (5.6 percentage points) and significant for White students in the 10th column of panel B (8.1 percentage points). When focusing on the influence of exclusively online enrollment and completion of any degree (e.g., associate or bachelor’s), findings were negative and statistically significant for all students and for each subgroup (20.9 percentage points for Asian students and between 8 and 9 percentage points for the pooled sample and all other subgroups).

Table 3 reports the influence of enrolling in exclusively online degree programs for the additional subgroups of students from lower-income households, students who had military service, and students who were financially independent. Enrollment in exclusively online degree programs was associated with a decrease of 8.9 percentage points in low-income students’ likelihood of bachelor's degree completion (first column) and a decrease of 11.4 percentage points in the likelihood of bachelor's degree completion among students with military service (fourth column). Exclusively online enrollment was unrelated to financially independent students’ likelihood of earning a bachelor's degree (10th column) but had a positive and marginally statistically significant influence (1.9 percentage points) on financially independent students’ likelihood of earning an associate degree. Exclusively online enrollment had a statistically significantly negative influence on completion of any degree for low-income students (9.3 percentage points) and financially independent students (5.5 percentage points).

In Table 4, only 4-year students who were enrolled in exclusively online degree programs as of 2017 are included in the results, comparing exclusively online students in a bachelor's degree program at a specific 4-year institution type to exclusively online students in a bachelor's degree program at all other 4-year institution types. For these analyses, we accounted for exclusively online students’ selection into different types of institutions and explored relationships between exclusively online enrollment at one institution type and exclusively online students’ likelihood of degree completion. Results in the third column of Table 4 show that exclusively online students who attended for-profit 4-year institutions were 11.9 percentage points less likely to complete bachelor's degrees compared to students at other types of 4-year institutions (third column of Table 4). However, exclusively online students who attended a selective 4-year institution had a marginally significant increase (8.1 percentage points) in their likelihood to complete a bachelor's degree when compared to exclusively online students in bachelor's degree programs at other less-selective 4-year institutions (fourth column of Table 4).

Table 5 focuses solely on exclusively online students at institutions offering associate degrees. Each exclusively online student in Table 5 was enrolled in an associate degree program at their institution. The first, third, and fifth columns of Table 5 present results for students who expected to earn only an associate degree; and the second, fourth, and sixth columns report results for students who expected to earn an associate degree or higher, which may also include a bachelor's degree or higher. Results are negative but not statistically significant for students at public 2-year, for-profit 2-year, and for-profit 4-year institutions who expected to earn associate degrees. The relationship with associate degree completion is negative and statistically significant (3.5 percentage points) for students enrolled in associate degree programs at for-profit 4-year institutions and expected an associate degree or higher (sixth column of Table 5).

To explore whether exclusively online students were leaving college without a degree or merely taking longer to graduate, we ran alternative specifications in which the outcome was “dropping out” of college (i.e., no longer enrolled and never graduated) and found that exclusively online students in associate degree programs at either public or for-profit 2-year institutions were more likely to drop out of college than their peers (results for these exploratory analyses can be found in Supplemental Appendix Tables 6 and 7 in the online version of the journal). As mentioned previously, we ran an additional robustness check to include students who enrolled initially in an exclusively online degree program before switching to a face-to-face program. These students, who completed their coursework in a face-to-face program, were included in the control group for the purpose of this additional robustness check, and we found similar results whether we included or excluded them from analyses (see Supplemental Appendix Table 8 in the online version of the journal). Finally, we also explored whether the results pertaining to institutional types varied when comparing exclusively online students at a given institution type to students who did not enroll in online degree programs at the same institution type. Results for these exploratory analyses indicated that exclusively online students at all types of 4-year institutions had a lower likelihood of completing a bachelor's degree relative to their peers at the same institution type. At the 2-year level, the negative relationship between exclusively online enrollment and associate degree completion was concentrated among for-profit institutions (see Supplemental Appendix Table 9 in the online version of the journal).

Sensitivity Analysis

Unobserved pretreatment differences between exclusively online students and their peers represent a potential issue for our study. Our analyses would violate the strong ignorability assumption if unobserved pretreatment differences of students in the BPS:12/17 sample were independent of the estimated propensity scores. To directly address this potential issue, we conducted Rosenbaum's (2002) sensitivity analysis to determine whether the presence of hidden bias due to unobserved covariates would affect our findings.

The extent to which an unobserved covariate influences the odds of a given student enrolling in an exclusively online degree program is captured by the value of Γ. Insignificant and low values of Γ are at a greater risk for contamination due to hidden bias, but significant and high values of Γ would be less likely to be impacted by hidden bias. For example, if two college students with the same observed covariates have a Γ = 2, one of the students would need to be twice as likely to receive treatment due to an unobserved covariate (i.e., hidden bias) than the observed covariate of interest to cast doubt on the observed treatment effect. Our sensitivity analysis checked all outcomes and model specifications, revealing that it would be unlikely that an observed treatment effect associated with enrolling in an exclusively online degree program was caused by an unmeasured or hidden confounder.

Discussion

Online education has the potential to mitigate time- or location-based constraints and increase access to higher education, particularly exclusively online degree programs. Prior literature pertaining to the impact of online enrollment on college students’ academic outcomes has reported mixed results and typically focused on comparing short-term outcomes, such as exam scores and course grades, of students in the online version of a course relative to those in the face-to-face version (Xu & Xu, 2020). Additional work focused specifically on the impact of online education on long-term outcomes typically reveals a positive relationship between enrolling in at least one online course, but not necessarily exclusively online enrollment, and degree completion (Fischer et al., 2022; Ortagus, 2023; Shea & Bidjerano, 2014). However, Huntington-Klein et al. (2017), which focused solely on Washington State community college students, reported that students who took the online version of a comparable face-to-face course were less likely to complete their associate or bachelor's degree. Despite the burgeoning body of literature on postsecondary online education and considerable growth of exclusively online degree programs across all types of American colleges and universities, little is known regarding the long-term implications of exclusively online enrollment for students’ likelihood of degree completion.

Because online education can remove barriers to higher education for low-income students forced to work during college, adults, parents, veterans, and other independent students, many colleges and universities have increased their number of online offerings and exclusively online degree programs (Ortagus, 2017). Unfortunately, the potential benefits of online education are not distributed equally, as exclusively online students who do not take any coursework on campus may not have not have the same level of personalized and consistent interactions with their faculty and peers. This dynamic forces exclusively online students to rely disproportionately on self-directed learning, which may exacerbate the systemic inequities benefitting White and higher-achieving students to a further degree than their peers (Xu & Xu, 2020). Anderson's (2008) “Theory of Online Learning” outlines the importance of high levels of engagement for online students by describing the effective online learning environment as one that prioritizes multiple modalities of effective interactions between students, faculty, and course content.

In this study, we use nationally generalizable data and a quasi-experimental approach to show that students who enrolled exclusively in online degree programs were less likely to complete bachelor's degrees than their otherwise-similar peers. Importantly, the negative influence of exclusively online enrollment was relatively consistent across race/ethnicity subgroups and concentrated primarily among low-income and veteran students. Findings also revealed that the negative influence of enrolling exclusively in online degree programs was exacerbated when the student attended a for-profit 4-year institution. Exclusively online students at selective 4-year institutions had a marginally significant increase in their likelihood of completing a bachelor's degree relative to exclusively online students at other 4-year institutions, and financially independent students who enrolled in an exclusively online degree program had a marginally significant increase in their likelihood of completing an associate degree.

A disproportionate share of exclusively online students face time- or location-based constraints that can make them less likely to graduate from college—regardless of medium of instruction. This suggests that readers should exercise caution when interpreting our results, as some of the observed effects outlined in the present study may be due to selection. We sought to directly address this issue in our analyses by employing a quasi-experimental design to reduce selection bias and conducting a series of sensitivity checks to measure the presence of hidden bias from unobserved covariates. However, we encourage readers to interpret our findings as another piece of evidence suggesting that online education can have a negative influence on students’ academic outcomes, but that does not necessarily suggest that enrolling in an exclusively online degree program is the sole cause of a given student's decision to leave college without a degree.

Implications for Practice, Policy, and Future Research

In alignment with learning theories outlined by Xu and Xu (2020), we contend that systematic inequities associated with self-directed learning can explain why racially minoritized students and students attending institutions without adequate academic support and services struggle in exclusively online learning environments. From a policy perspective, our work has a clear link to critical issues surrounding both quality and accountability in higher education. Specifically, students enrolling in exclusively online degree programs are eligible for federal financial aid programs. Taxpayers are best served when financial aid funds support degree programs that enable students to complete their degrees, which has direct and positive implications for their likelihood to secure productive employment and make sustainable progress in repaying any loans.

Given that the likelihood of degree completion in exclusively online programs appears to differ across student and institution types, our findings can contribute to the design and targeting of accountability measures specific to online degree programs and inclusive of multiple institution types, including for-profit institutions, with the greatest need to ensure quality for these degree programs. Future research can explore whether exclusively online students in specific academic disciplines fare better (or worse) in online settings when compared to their peers in other disciplines. In addition, future research should consider the effects of varying levels of online enrollment, including exclusively online degree programs, on student debt, students’ labor market outcomes, and additional outcomes that may allow students to make evidence-based decisions when determining the benefits and burdens of online education.

Due to the institutional findings showing that exclusively online students at for-profit 4-year institutions appear to be less likely to earn their degree than exclusively online students at other 4-year institutions, several policy implications should be considered. First, policymakers should require transparent reporting of costs and revenues among exclusively online degree programs. Prior research has shown that for-profit degree programs invest considerably more toward advertising than institutional spending (Vazquez-Martinez & Hansen, 2020), which can have a negative impact on students’ academic outcomes (Cellini, 2021; Ortagus & Hughes, 2021). Second, policymakers should regulate the use of for-profit online program managers (OPMs) by nonprofit colleges and universities, as the economic model of for-profit online offerings appears to be misaligned with the critical need to increase degree attainment and narrow completion gaps facing historically underserved students. Finally, colleges and universities should increase their investment in wraparound services and targeted engagement for exclusively online students given the compelling body of literature showing the positive impact of these types of approaches in higher education, regardless of medium of instruction (e.g., Bouchey et al., 2021; Miller & Weiss, 2022).

As noted previously, many colleges and universities may seek to expand their level of reliance on exclusively online degree programs due to financial motivations. For example, previous researchers found that public universities responded to decreases in state appropriations by increasing their online enrollments (Ortagus & Yang, 2018). Because online offerings are not subject to the same physical space limitations as face-to-face education, exclusively online degree programs can generate substantial tuition revenue by leveraging extremely large enrollment levels (Cheslock et al., 2016). However, the financial advantage of exclusively online degree programs, which is associated with offering high-enrollment courses at scale, may come at the expense of high-quality and student-centered learning experiences (Ortagus, 2020).

For those time- or location-constrained students who have no choice but to enroll in exclusively online degree programs, colleges and universities would benefit from widespread distribution and communication of established best practices for the development and delivery of high-quality online degree programs. Future accreditation efforts should evaluate clear, shared, and elevated standards for exclusively online degree programs and online education at large.

This study advances knowledge pertaining to exclusively online degree programs, but our findings are subject to several limitations. First, national data capturing online enrollment patterns at the student level are rather limited, which forces researchers to capture whether a student was enrolled in an exclusively online degree program as of 2011–2012, 2013–2014, and 2016–2017 and precludes more nuanced comparisons according to students’ online enrollment patterns. Second, BPS:12/17 offers limited information pertaining to local characteristics, such as the local unemployment rate or county-level economic considerations, that may affect a given student's likelihood of selecting into and completing an online degree program. Third, BPS:12/17 provides a 6-year window into students’ enrollment patterns and educational attainment, but researchers are unable to continue to follow students in the BPS:12/17 cohort to examine whether they completed a degree after the 2016–2017 academic year, which is especially worrisome for students in bachelor's degree programs enrolling on a part-time basis. We address this particular limitation by conducting balance checks on students’ likelihood to enroll on a part-time basis and find that part-time enrollment coefficients are balanced across the treatment and control groups.

Prior research has offered some actionable ways to enhance online course design in ways that can potentially improve students’ academic outcomes (Xu & Xu, 2020). In particular, students have seen increased achievement in courses that allow for students to interact with their instructor and other students (Bernard et al., 2009; Young, 2006). These types of online course interactions can be facilitated through synchronous design components, ongoing feedback, and purposeful discussion among students (Ortagus, 2020; Means et al., 2009). Despite a growing body of evidence regarding best practices in online course design, the same types of synchronous interactions and student engagement activities that improve the quality of an online course may also make the course less convenient for nontraditional students who cannot participate in traditional face-to-face offerings (Jaggars & Xu, 2016).

Supplemental Material

sj-pdf-1-aer-10.3102_00028312231222264 – Supplemental material for The Role and Influence of Exclusively Online Degree Programs in Higher Education

Supplemental material, sj-pdf-1-aer-10.3102_00028312231222264 for The Role and Influence of Exclusively Online Degree Programs in Higher Education by Justin C. Ortagus, Rodney Hughes and Hope Allchin in American Educational Research Journal

Footnotes

ORCID iDs

Justin C. Ortagus

Rodney Hughes

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a research grant from Arnold Ventures.

Justin C. Ortagus is an associate professor of higher education administration & policy and the director of the Institute of Higher Education at the University of Florida, 2705D Norman Hall, PO Box 117049, Gainesville, FL 32611, USA; e-mail: jortagus@coe.ufl.edu . His research examines how online education, community colleges, and various state policies affect the opportunities and outcomes of underserved students.

Rodney Hughes is an associate professor of higher education administration and the director of the Center for the Future of Land-Grant Education at West Virginia University. His research interests include higher education governance, institutional research, P–20 to workforce transitions, and education policy.

Hope Allchin is a PhD student in the higher education administration and policy program at the University of Florida. Her research interests include college access, admissions practices, and student activism on college campuses.

References

Abadie

Athey

Imbens

G. W.

Wooldridge

(2017). When should you adjust standard errors for clustering? (National Bureau of Economic Research Working Paper No. 24003). National Bureau of Economics Research. nber.org/papers/w24003

Allen

I. E.

Seaman

(2014). Grade change: Tracking online education in the United States. Babson Survey Research Group.

Alpert

W. T.

Couch

K. A.

Harmon

O. R.

(2016). A randomized assessment of online learning. American Economic Review, 106(5), 378–382.

Anderson

(2008). The theory and practice of online learning (2nd ed.). Athabasca University Press.

Austin

P. C.

Stuart

E. A.

(2015). Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Statistics in Medicine, 34(28), 3661–3679.

Bailey

Jaggars

S. S.

Jenkins

(2015). Redesigning America’s community colleges. Cambridge, MA: Harvard University Press.

Balaji

M. S.

Chakrabarti

(2010). Student interactions in online discussion forum: Empirical research from “media richness theory” perspective. Journal of Interactive Online Learning, 9(1), 1–22.

Bambara

C. S.

Harbour

C. P.

Davies

T. G.

Athey

(2009). Delicate engagement: The lived experience of community college students enrolled in high-risk online courses. Community College Review, 36(3), 219–238.

Barr

(2019). Fighting for education: Financial aid and degree attainment. Journal of Labor Economics, 37(2), 509–544.

10.

Bean

Metzner

(1985). A conceptual model of nontraditional undergraduate student attrition. Review of Educational Research, 55, 485–540.

11.

Bernard

R. M.

Abrami

P. C.

Borokhovski

Wade

C. A.

Tamim

R. M.

Surkes

M. A.

Bethel

E. C.

(2009). A meta-analysis of three types of interaction treatments in distance education. Review of Educational Research, 79(3), 1243–1289.

12.

Bettinger

E. P.

Fox

Loeb

Taylor

E. S.

(2017). Virtual classrooms: How online college courses affect student success. American Economic Review, 107(9), 2855–2875.

13.

Bork

R. H.

Rucks-Ahidiana

(2013). Role ambiguity in online courses: An analysis of student and instructor expectations (CCRC Working Paper No. 64). Columbia University, Teachers College, Community College Research Center.

14.

Bouchey

Gratz

Kurland

(2021). Remote student support during COVID-19: Perspectives of chief online officers in higher education. Online Learning, 25(1), 28–40.

15.

Bowen

W. G.

Chingos

M. M.

Lack

K. A.

Nygren

T. I.

(2014). Interactive learning online at public universities: Evidence from a six-campus randomized trial. Journal of Policy Analysis and Management, 33(1), 94–111.

16.

Causey

Pevitz

Ryu

Scheetz

& Shapiro

(Feb. 2022). Completing college: National and state report on six-year completion rates for fall 2015 beginning cohort (Signature Report 20). Herndon, VA: National Student Clearinghouse Research Center.

17.

Cellini

(2021). For-profit colleges in the United States: Insights from two decades of research (EdWorkingPaper: 21-398). Annenberg Institute at Brown University. https://doi.org/10.26300/bbna-vh38.

18.

Cellini

S. R.

Grueso

(2021). Student learning in online college programs. AERA Open, 7, 23328584211008105.

19.

Cheslock

Ortagus

Umbricht

Wymore

(2016). The cost of producing higher education: An exploration of theory, evidence, and institutional policy. In Smart

(Ed.), Higher education: Handbook of theory and research (Vol. 31). Springer.

20.

Deming

Goldin

Katz

(2013). For-profit colleges. The Future of Children, 23, 137–163.

21.

Deming

D. J.

Goldin

Katz

L. F.

Yuchtman

(2015). Can online learning bend the higher education cost curve? American Economic Review, 105(5), 496–501.

22.

DuGoff

E. H.

Schuler

Stuart

E. A.

(2014). Generalizing observational study results: Applying propensity score methods to complex surveys. Health Services Research, 49(1), 284–303.

23.

Figlio

Rush

Yin

(2013). Is it live or is it internet? Experimental estimates of the effects of online instruction on student learning. Journal of Labor Economics, 31(4), 763–784.

24.

Fischer

Baker

Orona

G. A.

Warschauer

(2022). Increasing success in higher education: The relationships of online course taking with college completion and time-to-degree. Educational Evaluation and Policy Analysis, 44(3), 355–379.

25.

Flores

S. M.

Park

T. J.

Baker

D. J.

(2017). The racial college completion gap: Evidence from Texas. Journal of Higher Education, 88(6), 894–921.

26.

Ford

Vignare

(2015). The evolving military learner population: A review of the literature. Online Learning, 19(1), 7–30.

27.

Goodman

Melkers

Pallais

(2019). Can online delivery increase access to education? Journal of Labor Economics, 37(1), 1–34.

28.

Guglielmino

L. M.

Guglielmino

P. J.

(2003). Identifying learners who are ready for e-learning and supporting their success. In Piskurich

G. M.

(Ed.), Preparing learners for e-learning (pp. 18–33). Pfeiffer.

29.

Guo

Fraser

(2015). Propensity score analysis: Statistical methods and applications (2nd ed.). Thousand Oaks, CA: SAGE Publications.

30.

Hart

C. M.

Friedmann

Hill

(2018). Online course-taking and student outcomes in California community colleges. Education Finance and Policy, 13(1), 42–71.

31.

Huntington-Klein

Cowan

Goldhaber

(2017). Selection into online community college courses and their effects on persistence. Research in Higher Education, 58(3), 244–269.

32.

Jaggars

S. S.

(2016). How do online course design features influence student performance? Computers & Education, 95, 270–284.

33.

Johnson

H. P.

Mejia

M. C.

(2014). Online learning and student outcomes in California's community colleges. Public Policy Institute.

34.

Joyce

Crockett

Jaeger

D. A.

Altindag

O’Connell

S. D.

(2015). Does classroom time matter? Economics of Education Review, 46, 64–77.

35.

Krieg

J. M.

Henson

S. E.

(2016). The educational impact of online learning: How do university students perform in subsequent courses? Education Finance and Policy, 11(4), 426–448.

36.

LeBlanc

(2013). Disruptive technologies and higher education: Toward the next generation of delivery models. In Kelly

Carey

(Eds.), Stretching the higher education dollar: How innovation can improve access, equity, and affordability. Cambridge, MA: Harvard Education Press.

37.

Long

M. C.

(2008). College quality and early adult outcomes. Economics of Education Review, 27(5), 588–602.

38.

Lynch

Engle

Cruz

J. L.

(2010). Subprime opportunity: The unfulfilled promise of for-profit colleges and universities. Education Trust.

39.

Means

Toyama

Murphy

Bakia

Jones

(2009). Evaluation of evidence-based practices in online learning: A meta-analysis and review of online learning studies. Association for Learning Technology.

40.

Melguizo

(2010). Are students of color more likely to graduate from college if they attend more selective institutions? Evidence from a cohort of recipients and nonrecipients of the Gates Millennium Scholarship Program. Educational Evaluation and Policy Analysis, 32(2), 230–248.

41.

Mfume

T. B.

(2018). The college completion glass—half-full or half-empty? Exploring the value of postsecondary education. Rowman & Littlefield.

42.

Miller

Weiss

M. J.

(2022). Increasing community college graduation rates: A synthesis of findings on the ASAP model from six colleges across two states. Educational Evaluation and Policy Analysis, 44(2), 210–233.

43.

Moore

Shulock

(2010). Divided we fail: Improving completion and closing racial gaps in California's community colleges. Institute for Higher Education Leadership & Policy.

44.

Morris

(2008). Economies of scale and scope in e-learning. Studies in Higher Education, 33(3), 331–343.

45.

National Center for Education Statistics. (2021). Digest of education statistics. U.S. Department of Education, Institute of Education Sciences.

46.

National Center for Education Statistics. (2022). Postsecondary outcomes for nontraditional undergraduate students. U.S. Department of Education, Institute of Education Sciences.

47.

Nguyen

T. D.

Kramer

J. W.

Evans

B. J.

(2019). The effects of grant aid on student persistence and degree attainment: A systematic review and meta-analysis of the causal evidence. Review of Educational Research, 89(6), 831–874.

48.

Ortagus

J. C.

(2017). From the periphery to prominence: An examination of the changing profile of online students in American higher education. The Internet and Higher Education, 32, 47–57.

49.

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, 1035–1058.

50.

Ortagus

J. C.

(2020). What we know about the cost and quality of online education. Third Way.

51.

Ortagus

J. C.

(2023). The relationship between varying levels of online enrollment and degree completion. Educational Researcher, 52(3), 170–173.

52.

Ortagus

J. C.

Derreth

R. T.

(2020). “Like having a tiger by the tail”: A qualitative analysis of the provision of online education in higher education. Teachers College Record, 122(2), 1–32.

53.

Ortagus

J. C.

Hughes

(2021). Paying more for less? A new classification system to prioritize outcomes in higher education. Third Way.

54.

Ortagus

J. C.

Yang

(2018). An examination of the influence of decreases in state appropriations on online enrollment at public universities. Research in Higher Education, 59, 847–865.

55.

Rosenbaum

P. R.

(2002). Observational studies. In Observational studies (pp. 1–17). Springer.

56.

Rovai

A. P.

(2003). In search of higher persistence rates in distance education online programs. The Internet and Higher Education, 6, 1–16.

57.

Rubin

D. B.

(2001). Using propensity scores to help design observational studies: Application to the tobacco litigation. Health Services and Outcomes Research Methodology, 2, 169–188.

58.

Sener

(2012). The seven futures of American education: Improving learning and teaching in a screen-captured world. North Charleston, SC: CreateSpace.

59.

Shadish

W. R.

Clark

M. H.

Steiner

P. M.

(2008). Can nonrandomized experiments yield accurate answers? A randomized experiment comparing random and nonrandom assignments. Journal of the American Statistical Association, 103(484), 1334–1344.

60.

Shea

Bidjerano

(2014). Does online learning impede degree completion? A national study of community college students. Computers & Education, 75, 103–111.

61.

Shea

Bidjerano

(2018). Online course enrollment in community college and degree completion: The tipping point. International Review of Research in Open and Distributed Learning, 19(2).

62.

Waugh

M. L.

(2018). Online student persistence or attrition: Observations related to expectations, preferences, and outcomes. Journal of Interactive Online Learning, 16(1), 63–79.

63.

Taniguchi

Kaufman

(2005). Degree completion among nontraditional college students. Social Science Quarterly, 86(4), 912–927.

64.

Tinto

(1975). Dropout from higher education: a theoretical synthesis of recent research. Review of Educational Research, 45(1), 89–125.

65.

Vazquez-Martinez

Hansen

(2020). For-profit colleges drastically outspend competing institutions on advertising. Brookings. www.brookings.edu/blog/brown-center-chalkboard/2020/05/19/for-profit-colleges-advertising/.

66.

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.

67.

Jaggars

S. S.

(2013). The impact of online learning on students’ course outcomes: Evidence from a large community and technical college system. Economics of Education Review, 37, 46–57.

68.

Jaggars

S. S.

(2014). Performance gaps between online and face-to-face courses: Differences across types of students and academic subject areas. Journal of Higher Education, 85(5), 633–659.

69.

(2020). The ambivalence about distance learning in higher education: Challenges, opportunities, and policy implications. Higher Education: Handbook of Theory and Research, 35, 351–401.

70.

Young

(2006). Student views of effective online teaching in higher education. American Journal of Distance Education, 20(2), 65–77.