A Study on Effects of Financial Aid on Student Persistence in Dual Enrollment and Advanced Placement Participation

Abstract

Despite concerted efforts to increase participation in advanced placement (AP) and dual credit (DC) programs, their efficacy remains unexplored. Drawing upon St. John’s model as the conceptual framework, this study employed a discrete-time event history analysis to examine the interplay between forms of financial aid and persistence toward degree completion for students participating in DC and AP programs and enrolling in a large, multicampus, Midwestern, U.S. University. First-time, first-year baccalaureate degree-seeking students who began studies in Fall 2012 were tracked for 4 years. The findings suggest that many factors are significantly related to college success, including student demographics such as race (especially Latino identity), first-generation status, housing status, socioeconomic status, and dependency status; high school performance, AP/DC participation, and SAT or ACT scores; and financial aid, such as Pell and federal grant aid and institutional grant programs. Results suggest that receiving Pell and federal grant aid and institutional grant-in-aid consistently and significantly attenuated the risks of student departure. In relation to prematriculation college-level credits, AP participants were more likely to receive institutional grant programs, whereas DC participants were more likely to have student loans. These findings have implications regarding the efficacy of DC/AP programs in regard to their interplay with financial aid systems in affecting persistence outcomes.

Keywords

dual credit or advanced placement student persistence financial aid

Policy Context

There have been fundamental changes in the financing of higher education over the last few decades in the United States and internationally (Johnston & Marcucci, 2010). Of particular importance is the increasingly critical role of financial aid to mitigate substantially rising costs, especially for students from low-income and disadvantaged backgrounds (McPherson & Schapiro, 1998; St. John, Paulsen, & Carter, 2005). Although financial aid alone may not remove barriers to student success (Stinebrickner & Stinebrickner, 2003), research has shown that financial aid has significant positive effects on student persistence across different ethnic and income groups (Hofmann, Vargas, & Santos, 2008; St. John et al., 2005), and varying forms of aid packages have also been demonstrated to have differential effects on college persistence (DesJardins, 2003; Hossler, Ziskin, Kim, Cekic, & Jacob, 2009; St. John, Hu, & Tuttle, 2000).

Concomitantly, there has been a substantial increase in the availability of college-level courses for secondary-level students nationwide, including advanced placement (AP) and, what has been variously called concurrent enrollment, dual enrollment, or dual credit (DC) (Borden, Taylor, Park, & Seiler, 2013; Thomas, Marken, Gray, & Lewis, 2013). While all DC/AP programs operate at the junction between secondary and higher education systems, states and institutions design and implement them in various ways (Taylor & Pretlow, 2015a). The primary purpose of these postsecondary course options for high school students is to encourage college participation, facilitate a smooth transition between secondary and postsecondary education, and alleviate some of the cost of earning a degree. Research has revealed that students who take college-level courses while in high schools have more accurate expectations of how much time they need to devote to class participation and studying and are more engaged during the first year of college than those who did not take those courses (National Survey of Student Engagement, n.d.). It is estimated that 40 states have initiatives and strong support for DC and AP programs (Borden et al., 2013; Karp, Bailey, Hughes, & Fermin, 2005; Krueger, 2006; Swanson, 2008; Taylor, Borden, & Park, 2015), and participation in these programs has been increasing notably over the last decade. According to the data collected by the U.S. Department of Education though national surveys during the 2010–2011 school year, approximately 82% of public high schools offer courses for college credit, and high schools reported approximately 2 million enrollments in DC courses and 3.5 million enrollments in AP or international baccalaureate courses (Thomas et al., 2013). The College Board (2017) reports that the annual participation in AP has nearly doubled (from 645,000 in 2006 to over 1.1 million students in 2016) while that of low-income and minority students more than quadrupled (College Board, 2017). Most states pay for the cost of DC courses and AP exams for students from low socioeconomic families and disadvantaged backgrounds. Even for more affluent students, DC courses, for which students typically pay only a portion of regular tuition, reduce the cost of college while helping middle- and low-achieving students to meet high academic standards (Karp, Calcagno, Hughes, Jeong, & Bailey, 2007). In the same vein, college-credit examinations, exemplified by AP, accelerate learning opportunities and increase student aspirations in pursuit of a postsecondary degree. Geiser and Santelices (2004), controlling for student characteristics, found that students who took AP courses and exams were likely to outperform non-AP participants, and AP participation was positively related to degree attainment. More recently, researchers have turned their attention to the relationship between DC/AP participation and college learning outcomes (Godfrey, Wyatt, & Beard, 2016).

In an unpublished study that employed the same dataset as employed in this study, Rutkowski, Rutkowski, Plucker, Borden, and Prusinski (2011) reported that the average estimated family contribution among DC participants is much lower than that of AP enrollees, but empirical exploration of effects is remarkably limited, particularly in how participation in DC and AP programs interacts with various forms of aid to influence the persistence of students from low socioeconomic and otherwise disadvantaged backgrounds. Even though there is a significant support for DC/AP programs by practitioners and policy makers, research has not been able to keep pace with the rapid growth of DC and AP opportunities (Rutkowski et al., 2011). Such questions as whether DC and AP programs increase college enrollment, improve retention, and reduce tuition costs so as to accelerate pathways to college degrees are critical to educational policy at federal, state, and local levels. Consequently, the expanded role of high school students earning college-level credit warrants more scrutiny, particularly with regard to how this practice interacts with financial aid policies and promotes efficient academic progression and degree completion.

This article, accordingly, examines the effects of DC and AP policies and practices on student outcomes in association with financial aid packages within a large, statewide, multicampus institutions in the U.S. Midwest. This institution is located in one of 10 states that have mandated high schools to offer college credit opportunities to expand students’ access to postsecondary education and encourage college participation (Borden et al., 2013). Concomitant with this scope of DC and AP opportunities, the state has also strengthened support for college access and persistence for low-income and underrepresented minority students with a unique financial aid program known as the Twenty-First Century Scholars (Toutkoushian, Hossler, DesJardins, McCall, & Canche, 2013). Therefore, this study is an investigation of how participation in DC and AP programs interacts with financial aid policy and practice to promote college degree completion and minimize educational disparities across different socioeconomic backgrounds in a U.S. Midwestern state.

Research Purpose and Questions

This study focuses on the following research questions:

How are different types of financial aid distributed among DC and AP students related to their demographics (e.g., socioeconomic status [SES] and race or ethnicity)?

Are there differences in persistence between DC and AP students who receive financial aid and DC and AP students who do not?

How are different types of financial aid associated with postsecondary persistence for students who have participated in DC and AP? Does the association vary over time?

Conceptual Framework

This study builds upon previous research on the effects of DC and AP on college persistence in association with the role of financial aid. Its aim is to contribute to understanding ways in which policies and programs can enable a larger and more diverse student population to benefit from participating in DC/AP programs.

Policy Effects on DC/AP Participation and Finances

State policies have expanded participation in DC and AP programs in recent years (An & Taylor, 2015; Kim, Kirby, & Bragg, 2006). This increased participation has enriched high school students’ lives, diversified high school curriculum and course options, encouraged students’ college aspirations, and supported their academic performance as well as persistence to degree completion (Allen, 2010; An & Taylor, 2015; Boswell, 2001; Smith, 2007). In spite of the tremendous growth and popularity of DC and AP programs, however, research on their impact on students’ transition to and success in higher education has yielded mixed results. Several studies address issues of postsecondary equity and considered DC/AP opportunities as a viable pathway to college access and progress toward degree completion for low-income students (Allen & Dadgar, 2012; Taylor & Pretlow, 2015b). Some researchers (An, 2013a, 2013b; Hofmann, 2003; Hofmann et al., 2008; Klopfenstein, 2010; Speroni, 2011) found that DC participants are more likely than nonparticipants to graduate from high school, and Karp, Calcagno, Hughes, Jeong, and Bailey (2008) reported that DC programs are strongly and positively associated with postsecondary outcomes. Thus, as DC policies and enrollments expand, the gap in college degree attainment between students from different socioeconomic backgrounds is expected to decrease. Other studies have demonstrated that students from low socioeconomic backgrounds benefitted from DC to a greater extent than high socioeconomic peers (An, 2013a, 2013b; Karp et al., 2008). Conversely, Meade and Hofmann (2007) reported that DC programs seemed to favor high achievers from high socioeconomic backgrounds, which widened socioeconomic disparities in college degree attainment. Similarly, An (2013b) found that approximately 26% of DC enrollees from the bottom quartile of the income distribution attained college degrees compared with 59% of students from the top quartile.

While AP courses were initially offered in predominantly White, middle-class, suburban, and well-funded schools (Klopfenstein, 2010), various studies have demonstrated that they positively affect college academic performance of students from low socioeconomic backgrounds (College Board, 2011; Hargrove, Godin, & Dodd, 2008). To increase access to AP programs, many states provide fee reductions for low-income students and disadvantaged minorities. Nonetheless, Klopfenstein (2010) found that neither taking AP courses nor passing AP exams increased the likelihood of college graduation, while DC participation increased the probability of graduation by 25%. As both programs continue to grow, the degree attainment of students who participate will be a critical concern for policy makers and institutional leaders, especially in light of the mixed results from prior research. In brief, the popularity of AP/DC is partially grounded in this accumulating body of studies and has expanded dramatically, suggesting that evidence on the effects of DC/AP is critical. Moreover, especially in the case of low-income students, it is equally important to investigate how financial aid packages interact with AP/DC activity to influence postsecondary equity and student learning outcomes.

Research on the Impact of Financial Aid on Student Success

Much of the extensive research on student persistence and degree attainment in higher education (Bean, 1980, 1983, 1985; Braxton, 1999, 2000; Cabrera, Nora, & Castaneda, 1992, 1993; DesJardins, 2003; Pascarella & Terenzini, 1980; St. John, 1992, 2000; Tinto, 1975, 1982, 1987) was initiated by Tinto’s seminal student integration model. This model, which posits academic and social integration as the central components of student success in college, assumes that a wide range of individual attributes and prematriculation experiences (such as high school academic performance) impact students’ educational expectations and goal commitments when moving through the higher education system. Braxton and his coworkers (Braxton, 2000; Braxton, Hirschy, & McClendon, 2004; Braxton, Sullivan, & Johnson, 1997) have extensively reviewed research testing the propositions of Tinto’s model of student integration and refined its application to student persistence in different types of postsecondary institutions. Bean and Metzner (1985) identified the factors that influence student persistence for nontraditional students in higher education, including perceived financial barriers, working for long hours, a lack of encouragement, family responsibilities, and opportunity to transfer to other institutions. Based on the work of Tinto (1988, 1993) and Bean (1985, 1990), important internal factors were identified in relation to student persistence, including personal attributes, initial institutional commitments, and academic and social integration.

Cabrera et al. (1992, 1993) developed a model of student adjustment that merged Tinto’s and Bean’s models with complementary approaches and further validated the utility of modeling student persistence in higher education. Using cost–benefit analysis, they also examined effects of different types of aid on students’ perceived ability-to-pay for higher education costs as one factor in persistence. As central components, their model included preacademic performance, finances, academic and intellectual development, and academic and social integration. Subsequently, St. John and coworkers (St. John, Paulsen, & Starkey, 1996; St. John et al., 2005) conducted studies examining the role of financial aid in postsecondary access and college retention and developed a model for assessing the effects of financial aid on persistence decisions for freshmen. Building on prior work, they constructed a financial impact model related to student enrollment decisions and provided evidence that finance-related factors have intangible and indirect effects on students’ subsequent persistence decisions in college (Paulsen & St. John, 2002). In particular, St. John and coworkers have conducted extensive research on the effects of the target State’s seminal aid program for low-income students on student persistence (St. John, 1992; St. John et al., 2000, 2004; St. John, Hu, & Weber, 2001; St. John, Musoba, Simmons, & Chung, 2002). However, it should be noted that Toutkoushian et al. (2013) have argued that those electing to participate in the Twenty-First Century Scholar program could be assumed to be more motivated than other populations, resulting in the greater likelihood of their persisting regardless of the financial aid they receive, potentially leading to a spurious association between the Twenty-First Century Scholar program and persistence outcomes due to the effects of self-selection.

Empirical Model

Various models have been developed to explain student departure behaviors. Among the more widely cited are Tinto’s student integration model (1975, 1988, 1993), Bean’s model of attrition (1980, 1985, 1990), Cabrera et al.’s student adjustment model (1992, 1993), and St. John’s extension of research (St. John, 1992; St. John et al., 2000; St. John, Cabrera, Nora, & Asker, 2000) that has shown the following variables to be associated with student success, including student characteristics and precollege academic experiences. Emphasizing the importance of financial aid to student success, St. John’s model indicates that the adequacy of financial aid is associated with persistence, particularly for students from low socioeconomic backgrounds. Drawing upon St. John’s model, this study explores the effects of financial aid on persistence for students participating in AP/DC programs prior to matriculating at the two largest campuses of seven-campus degree-granting public institution in the Midwest. The variables of interest incorporated in this study included AP/DC participation status, student characteristics (i.e., gender, ethnicity, first-generation status, dependency status, and housing status), precollege academic background (i.e., SAT or ACT score, high school grade-point average and high school diploma type), socioeconomic background variables (i.e., the amount of financial needs to be needed), and the types and amounts of financial aid received (i.e., Pell and other federal aid grants; state aid grants; institutional and private aid; private, parent, and other loans; and federal subsidized and unsubsidized loans). All detailed variables are listed and described in Table 1. To examine the differentiated forms of financial aid on persistence of participants in the target programs, the empirical model controlled for students’ demographic characteristics, precollege academic background (including SAT/ACT, types of high school diploma, and high school grade point average), and financial needs.

Table 1.

Variables.

Student characteristics	Definition
ID	Student ID
Gender	Gender reported by the institution (0=male, 1=female)
AP status	The participation status of AP program (1=participated in, 0=otherwise)
DC status	The participation status of DC program (1=participated in, 0=otherwise)
Ethnicity	Racial or ethnic group (White, African American, Hispanic/Latino, Asian/Pacific American, Nonresident Alien, and Others)
SAT/ACT equivalent	Students' scholastic achievement tests
High school diploma types	Four types of high school diploma, including general Indiana, core 40 regular, and core 40 with academic honors and others
High school GPA	Grades reported by the institution
Financial need amount (SES proxy)	The amount of money students needed for paying for cost of attendance
First generation	Whether students are first-generation status
Financial aid
Dependency status	Either as a financially “dependent” or “independent” of their parents or guardians
Pell grant and other federal aid	Sum of Pell grant and other federal aid received by students during the fiscal year
State aid	Sum of state aid received by students from the institution during the fiscal year
Institutional and private aid	Sum of Institutional and private aid received by students during the fiscal year
Federal subsidized and unsubsidized loan	Sum of federal subsidized and unsubsidized loan received by students during the fiscal year
Private, other, and parent loan	Sum of private, other, and parent loan during the fiscal year. Calculated based on student' residency status.
Dependent variables
Persistence (First stopout)	The first observation during which students were not reported as enrolled for the following year (the fall and spring semester)

Note. AP = advanced placement; DC= dual credit; GPA = grade point average; SES = socioeconomic status.

Methodology

Event history analysis has received attention in the field of social science and empirical research and is considered a valuable analytical tool to investigate event occurrence with longitudinal data (Singer & Willett, 2003). The method referred to as hazard modeling or survival analysis is a technique used to determine whether and how various factors are related to the occurrence of an event at different points in time. The approach involves two important concepts: censoring and the risk of event occurrence. Censoring is applied to situations in which an event of interest (e.g., graduation) is unknown at the end of the observation period, so some students might experience the event beyond the observation period and others might never experience the event (Singer & Willett, 2003). A major advantage of employing survival analysis is that it enables the researcher to distinguish censored from uncensored cases to estimate the duration of data and the effects over time (Chen & DeJardins, 2008). Therefore, censored cases could be removed from the base sample in each year to facilitate the analysis.

In this study, the discrete-time method was applied to estimate the effects of types of aid packages on persistence in relation to students’ participation status in DC and AP programs. DesJardins and coworkers (Chen & DesJardins, 2008; DesJardins, 2003; DesJardins, Ahlburg, & McCall, 1999) conceptualized degree attainment as a temporal process and argued that student financial aid could be considered as a time-dependent covariate in an event history model of student departure. The effects of aid can vary over time, and the different forms of aid students receive are also included as a time-dependent covariate. The outcome of interest in the study was defined as the conditional probability that a student would drop out at time j, given that the individual did not experience this event prior to time j. The hazard of student dropout is a logit function of two sets of predictors. The following equation denotes the general form of the model where h(t_ij) represents the hazard rate at a discrete point in time. The first coefficient-parameter set (αDs) represents the change of the baseline logit hazard for each academic year, and the second set (βXs) denotes the unit changes corresponding to the associated predictors, which may include time-varying or time-invariant variables.

Logit h (t_{i j}) = [α_{1} D_{1 i j} + α_{2} D_{2 i j} + α_{3} D_{3 i j} + \dots + α_{J} D_{J i j}] + [β_{1} X_{1 i j} + β_{2} X_{2 i j} + \dots + β_{P} X_{P i j}_{]}

(1)

Since this was a secondary study, the collection to access to (limited access to variables) and the sample may suffer the problem of self-selection bias. That is, the effect of AP/DC on the outcome variable (i.e., student persistence) may have been caused by the nonrandom participation in AP/DC groups, and unknown confounding variables may have produced the participation status and correlation to college outcomes. However, by employing event history methods, three procedures from another methodological perspective were implemented in this study to mitigate the selection bias problem. First, the event history model, rather than a cross-sectional analysis, was employed to analyze longitudinal data to ensure the casual inference between variables with distinct time ordering (Chen & DesJardins, 2010). Second, the potentially confounding variable, SES, which relates to both AP/DC status and student persistence, was included in the model as a covariate. Finally, a random component is introduced into the model to handle heterogeneity between students, which turns the model (Equation (1)) into a frailty model (Yashin, Vaupel, & Iachine, 1995).

Analysis

The analysis entailed four phases. First, descriptive statistics were examined to observe the interplay among students’ participation in DC/AP, financial aid, and persistence behaviors. Second, life-table and Nelson–Aalen estimation methods were used to estimate whether survival and hazard functions varied during the observation period of the study. Third, we added interaction terms referring to the different forms of financial aid students received and their DC/AP participation to examine whether the probability of student dropout behavior varied over time. Fourth and finally, a z-test is employed to determine whether those variables were statistically significant.

Data Source and Samples

Data for this study came from the target University’s university-wide student record system. For analytical purposes, data from the operational system are routinely extracted into a nonvolatile, read-only data warehouse. The population of interest in this study comprised first-time, first-year baccalaureate degree-seeking college students who began undergraduate studies in Fall 2012 at the university’s two large campuses, which enrolled a combined total of just over 50,000 students.

Data Construction

The dataset was structured in a person-period format tracking students for 4 years (from entry in 2012 to 2016). It is worth noting that the factor of student financial needs in the study was considered as a proxy for student SES. It is also a reverse variable, indicating that the more money students needed, the lower their SES. In addition, the academic year was employed as the temporal unit of the analysis in the model. We also constructed two binary measures to denote whether students participated in AP and DC programs. For instance, the variable for AP participation status is equal to one if student participated in AP program and zero otherwise. The outcome measure in this study refers to the risk of event occurrence (dropout) in each discrete time period, which is also the conditional probability that students will experience dropout in a particular time in college given they have not dropped out earlier. The censored cases refer to students who graduated with a certificate or associate degree and did not continue on for a baccalaureate degree during our observation period. Table 2 provides a narrative history of event occurrence over time. At the beginning of 2012, all 15,643 students were enrolled. During the first year, 2,617 students dropped out, leaving 13,026 students enrolled in the next year. By the end of the second year, 1,290 additional students had dropped out and 16 cases were censored. By the end of the third year, 678 additional students had dropped out and 214 cases were censored. That is, the largest number of students dropped out during the first 2 years, and based on the number of students remaining enrolled in each subsequent year, the risk of dropping out during our observation period became lower. On average, the survival function decreased from 0.83 to 0.70 in the third year. Accordingly, the cumulative hazard increased from 16.7% in the first year to 32.4% in the fourth year. As Singer and Willett (1993, 2003) recommended, logistic regression procedures (Equation (1)) were used to estimate baseline hazard models and effect coefficients via the software of HLM6.02 (Raudenbush, Bryk, & Congdon, 2005). Given the default of the statistics software, listwise deletion was executed in the student-level data if the participant had missing information in the person-period dataset. Therefore, the final sample had a total of 14,248 students for survival analysis in this study.

Table 2.

Life-Table Describing Event Patterns During the Observation Period.

Year	Beginning total	Dropout	Censored cases	Survival function	Hazard function	Cumulative hazard function
1	15643	2617	22	0.833	0.167	0.167
2	13004	1290	16	0.750	0.099	0.266
3	11698	678	214	0.707	0.058	0.324

Results

After cases with missing values for the outcome variable were removed, the original dataset comprising the participants for this study was the 15,643 first-time, full-time, state-resident baccalaureate degree seekers who began college in Fall 2012 and completed a Free Application for Financial Student Aid. As indicated in Table 3, about 56% of the sample were female, and 76.9% of the students were White, 7.7% were Black/African American, 6.2% were Hispanics/Latino, 4.4% were multiracial, and 4.8% were other ethnic minority groups combined (Asian American and Native American or Pacific Islander). For analysis purposes, students were divided into four groups based on their AP/DC participation status in high school: both AP and DC, AP only, DC only, and neither. Students taking both AP and DC scored the highest in high school grade point average (GPA; 3.81) and SAT/ACT (1,274), followed by AP participants (3.75; 1,268), DC participants (3.54; 1,105), and those who participated in neither program (3.32; 1,047).

Table 3.

Descriptive Statistics of the Sample (Variables for LR Analysis).

Variable name	Year one range	Year one % or mean
Admit type: First time in college	0–1	100.00
Gender: Female	0–1	55.75
Race: White	0–1	76.89
Race: Black/African American	0–1	7.74
Race: Hispanic/Latino	0–1	6.21
First-generation status: First generation	0–1	27.42
HS diploma type: Core 40 academic	0–1	16.58
HS diploma type: Core 40 honors	0–1	62.38
HS diploma type: General	0–1	19.77
Housing status: On campus	0–1	39.64
AP status	0–1	19.20
DC status	0–1	28.46
Financial need amount (SES proxy)	0–46.52	12.34
HS GPA (4.0 GPA scale)	1.52–4.00	3.54
SAT or ACT equivalent (in 100s)	4.70–16.00	11.39
Pell grant and other federal aid offered^a (in 1,000s)	0–11.90	1.63
State aid offered^a (in 1,000s)	0–13.34	1.58
Institutional and private aid offered^a (in 1,000s)	0–46.81	3.66
Private, other, and parent loan offered^a (in 1,000s)	0–44.57	2.29
Federal subsidized and unsubsidized loan offered^a (in 1,000s)	0–10.50	3.68
Dependency status (Y1–Y4)Enrollment status (Y2–Y4)

^aTime-varying variable.

In terms of types of high school diploma and AP/DC participation, DC participants had the highest number of recipients of high school honors diplomas, followed by participants of neither program, AP participants, and both AP and DC participants. There are different patterns in the distribution of aid types among students from various backgrounds in relation to their AP/DC participation status (see Table 4). On average, institutional and private aid constituted the largest source of financial assistance for AP participants, followed by federal subsidized and unsubsidized loans and private loans. In addition, DC participants received the largest amount of aid in the form of federal subsidized and unsubsidized loans, followed by private loans and institutional and private aid. As a combined group, AP and DC participants received their highest amounts from the institution and private aid, followed by federal subsidized and unsubsidized loan, and then private, other, and parent loan. It should be noted that Pell grants and other federal aid constituted the lowest source of financial support for AP participants, DC participants, and both AP/DC participants. Descriptive statistics also revealed that the substantial students who participated in both AP and DC programs tend to be from advantaged socioeconomic families, followed by AP takers and DC takers. It should be noted that amount of federal aid for work study distributed to these groups was very small, so the model does not include this form in further analysis.

Table 4.

The Average of Aid Dollars Awarded Between 2012 and 2016.

AP status	DC status	Pell grant and other federal aid offered	State aid offered	Institutional and private aid offered	Federal subsidized and unsubsidized loan offered	Private loans offered	Financial need amount (SES proxy)
1	1	726	744	5,721	3,264	2,092	10,215
1	0	1,009	1,075	5,460	3,498	1,967	11,064
0	1	1,389	1,563	2,432	4,111	2,279	11,878
0	0	2,098	1,995	1,875	4,159	2,032	14,044

Note. AP/DC status 1 = Participated in; AP/DC status 0 = Otherwise; AP = advanced placement; DC= dual credit; SES = socioeconomic status.

The final step of the longitudinal analysis was to take types of aid as a time-dependent covariate and introduce interaction effects between aid types and students’ AP/DC participation status into the model for the inferential analysis so as to estimate the hazard of dropout over the 4-year period while controlling for other confounding factors. Accordingly, two models were specified and tested, the first including the proportional hazards model that included all exogenous variables in the model without interaction terms. Logistic regression analysis for proportional hazard models (Chi-square =2828.92; df= 21; p< .001) was conducted. As indicated in Table 5, controlling for other factors, the result revealed that ethnic group (Hispanic/Latino), first-generation status, housing status, AP participation, DC participation, high school GPA, SAT/ACT, SES, Pell grants and other federal aid, institutional and private aid, and dependency status were significantly related to student persistence behaviors. Moreover, gender did not significantly differentiate students with regard to likelihood of dropping out, while first-generation students were significantly more likely to drop out of college than nonfirst-generation students. There was no significant difference in dropout risks between minorities and Whites except for Hispanic/Latinos, who were more likely to drop out. Students from low socioeconomic families were more likely to be at risk for dropping out than their counterparts. Being dependent on family financial support was significantly associated with lower risk of dropping out compared with being an independent earner.

Table 5.

Analysis of Student Dropout With Logistic Regression.

	No interaction		Interaction
	Odds ratio	Standard error	Odds ratio	Standard error
Gender: Female	1.005	0.042	1.011	0.043
Race: White	0.923	0.072	0.925	0.072
Race: Black/African American	0.894	0.073	0.903	0.074
Race: Hispanic/Latino	0.813	0.083*	0.819	0.084*
First-generation status: First generation	1.238	0.046***	1.236	0.046***
HS diploma type: Core 40 academic	0.943	0.079	0.946	0.080
HS diploma type: Core 40 honors	0.904	0.066	0.885	0.067
Housing status: On campus	0.710	0.045***	0.706	0.045***
AP status	0.628	0.058***	0.569	0.128***
DC status	0.531	0.045***	0.532	0.110***
HS GPA (4.0 GPA scale)	0.359	0.069***	0.364	0.069***
SAT or ACT equivalent	0.955	0.018**	0.956	0.018*
Financial need amount (SES Proxy)	1.018	0.003***	1.017	0.003***
Pell grant and other federal aid offered	0.969	0.012*	0.956	0.014**
interaction with AP/DC group: AP			1.042	0.026
interaction with AP/DC group: DC			1.029	0.022
State aid offered	1.003	0.008	0.989	0.010
interaction with AP/DC group: AP			1.024	0.021
interaction with AP/DC group: DC			1.027	0.017
Institutional and private aid offered	0.951	0.007***	0.966	0.009***
interaction with AP/DC group: AP			0.971	0.013*
interaction with AP/DC group: DC			0.987	0.012
Private, other, and parent loan offered	0.997	0.003	1.001	0.005
interaction with AP/DC group: AP			1.002	0.007
interaction with AP/DC group: DC			0.985	0.007*
Federal subsidized and unsubsidized loan offered	1.003	0.008	1.002	0.011
interaction with AP/DC Group: AP			1.018	0.019
interaction with AP/DC Group: DC			0.992	0.017
Dependency status: Dependent	0.558	0.110***	0.557	0.110***
Year 3	0.638	0.045***	0.640	0.045***
Year 4	0.388	0.057***	0.388	0.056***

Note. N = 14,248.

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

The second regression referred to the model of time-varying effects, which included interaction terms. The interaction regression tested the interaction between aid types and student AP/DC participation in the model, aiming to estimate whether the aid effects on AP/DC participants varied over time and how they affected students’ persistence in their pursuit of college credentials. The results indicated that the interaction between institutional and private aid, private, other and parent loan and AP/DC participation was significant, whereas interactions between participation in AP/DC and state aid, private, other, parent loans, and federal subsidized and unsubsidized loans were not significant. To be specific, the interaction between institutional and private aid was significantly related to AP/DC group: AP, namely, AP participants with higher institutional and private aid were less likely to drop out than non-AP participants. Moreover, the interaction between private, other, and parent loans was significantly associated to participation in that DC participants were less likely to depart as they received higher private, other, and parent loans than non-DC participants. In addition, ethnic group (Hispanic/Latino), first-generation status, housing status, AP participation, DC participation, high school GPA, SAT/ACT, SES, and dependency status were significantly related to student persistence behaviors. The third and fourth were also significantly related student dropout behaviors that indicated that students were less likely to drop out in the latter years (Chi-square statistic = 2057.38; df= 31; p< .001; see Table 5).

In summary, the results from the proportional hazards model is consistent with the interaction analysis after controlling for all other factors, suggesting student characteristics (i.e., race, first-generation status, housing status, high school GPA, SAT/ACT, and SES), dependency status, Pell grants and other federal aid, and institutional and private aid were consistently significantly associated with the risks of dropping out. AP participation with more institutional and private aid had statistically significant effects on persistence for those who did not participate in the program, while DC participation with higher private, other, and parent loans was significantly related to the odds of dropping out compared with those counterparts. In particular, AP students who received substantial institutional and private aid were significantly more likely to persist, while DC students who received loans from any source were significantly more likely to persist.

Conclusions and Discussions

Increasing degree attainment to meet national completion goals has become the focus of many state-level initiatives in the United States to improve student success. The descriptive analyses provide information regarding the underlying patterns in aid distribution and dropout risk by AP/DC participation. Descriptive analyses showed that AP participants were more likely to receive higher institutional and private aid, while DC participants received higher federal subsidized and unsubsidized loans. As Hillman (2014) pointed out, the federal government’s policy has shifted away from the grant-based aid to a loan-based system. The findings revealed that AP participants rely increasingly on grant aid awarded based on criteria other than need, while DC participants rely increasingly on loans to fund their postsecondary education. Pell grants and other federal aid constituted the lowest source of financial support for either AP or DC participants, which may be attributed to the comparatively complex eligibility and application rules for Pell grants, resulting in the majority of students failing to successfully complete applications (Toutkoushian et al., 2013). Thus, the complexity of financial aid programs undermines their effectiveness and further influences student behaviors (Scott-C, 2015). Furthermore, the analyses of the longitudinal dataset reveal that changes in two types of aid (Pell grants and other federal aid, institutional and private aid) were found to have significant effects on student persistence for AP/DC participants, indicating the interactive relationships between the types of aid and the intention to remain enrolled for students from varying socioeconomic backgrounds.

In addition, Latino/Hispanic students were found to be less likely to complete college degrees than their counterparts, an educational attainment gap which researchers have in part attributed to their having insufficient financial resources to pay for postsecondary education (Zerquera & Gross, 2015). SES was found to have significant effects on college degree success of participants, which further confirmed a gap in college success for students from low socioeconomic backgrounds compared with their counterparts, especially for first-generation students (Pike & Kuh, 2005). Low SES further limited educational opportunities for low-income students working toward degree completion (Engberg, 2011). These disparities in postsecondary access strongly reinforce other findings related to educational opportunities across difference student population (Engberg, 2011; Scott-C, 2015). Student demographic backgrounds such as first-generation status and academic preparation were found to be related to college success. Information that helps institutional practitioners to recruit high-achieving and diverse students and streamline the student experience from first contact to graduation and beyond.

In summary, the results of differential aid effects consistently confirm findings of previous studies (Hossler et al., 2009; St. John, 1992; St. John et al., 2000) that Pell grants and other federal aid and institutional and private aid were found to have significant impact on student persistence. To examine whether aid effects vary over time, we estimated the interaction in such a way as to mitigate the selection and sorting effects of AP/DC participation. According to Rutkowski et al. (2011), AP/DC participation can be considered as another proxy for selection and sorting. Those who participated in AP (and especially those among them who also participate in DC) represent the most well-prepared students with the highest social and economic capital: They do best in college almost regardless of aid but, ironically, are the recipients of most of the institutional gift aid due its use to increase selectivity. The AP and DC recipients represent the most highly “college motivated” among these students, in addition to their higher levels of academic preparation and affluence. The effect of financial aid on persistence, thus, is expected to be larger for students without AP or DC participation. If the financial aid is used in such a way as to mitigate the selection and sorting effects of AP/DC participation, the interaction is observed. The results indicated that the interaction between Pell grants and other federal aid, IU and private aid was significant for AP/DC participation, whereas interactions between participation and state aid, private, other, parent loans, and federal subsidized and unsubsidized loans were not significant. AP participants were less likely to drop out than non-AP participants, and DC participants, who received higher financial assistance than non-DC participants, were less likely to depart. The effect of financial aid on students with or without AP/DC participation appears to counter expectations. That is, the prediction that aid would mitigate the selection and sorting effects of AP/DC participation was not supported by the results. Further investigation of the impact of financial aid on persistence for students with or without AP/DC participation should be conducted.

As financial aid policy has shifted its focus, affordability concerns have replaced improving postsecondary access for low-income families, and grant programs in the form of loans, merit aid, and tax credits are now widely used instead of need-based aid (Hossler et al., 2009; Hillman, 2014). These factors may exacerbate rather than ameliorate the effectiveness of aid programs for students from disadvantaged backgrounds. In addition, the findings with regard to the influence of ethnic group or race, first-generation status, academic preparation, and housing status indicate that student demographics are still important factors in relation to postsecondary success. This result emphasizes the need for institutional researchers and policy makers to reexamine or adjust institutional recruitment and retention programs to offer assistance for needy students. To sum up, the time-varying nature of student success revealed a substantial number of students leaving school in the first year, indicating the need for institutional programs and practices that support student persistence at the different time points (Chen & DesJardins, 2008). To be specific, students with different aid packages respond in different ways according to analyses based on the time-varying model. Results are mixed regarding the effects of different types of aid on learning outcomes for students with prior participation in AP/DC programs, and the impact of some important demographic parameters of persistence associated with students’ time to degree, suggesting continued challenges in understanding how differences in socioeconomic backgrounds between AP and DC participants influence opportunities for receiving grant-aid as opposed to loans and how they affect students’ pathways toward getting a college degree.

Policy and Research Implications

The current study suggests that the ability of DC and AP programs to extend college access to a broader range of the socioeconomic spectrum can be further magnified by various sources of financial aid. The findings from this study point to several implications worth consideration by educational policy makers and practitioners. First, this study provides longitudinal data on the effects of the interaction of participation in AP/DC programs with financial aid on the risk of dropping out from college. In agreement with prior studies (Chen & DesJardins, 2008; Hillman, 2013), we observed a shift in student financial aid from grant to loans and from federal grant programs to institutional aid. However, after controlling for other factors, both Pell grants and other federal aid and institutional and private aid were significantly associated with lowering the risks of dropping out. As a result, as the cost of higher education continues to rise, an increasing number of students from financially disadvantaged families will depend heavily on federal and institutional grant aid programs to lessen the financial burden of attaining a college degree. As another way of alleviating costs, state policy makers and institutional practitioners should promote educational efficiency by continuing to expand AP/DC programs throughout the nation (Borden et al., 2013).

Second, the findings indicate that the impact of precollege academic preparation and experience, including AP/DC participation, high school GPA, and SAT/ACT scores, is significantly related to student success. To be specific, AP/DC participants were more likely to persist in college than their counterparts. Indeed, the intention of AP/DC programs is to accelerate progress toward a college degree, and this head start is especially needed by students from low socioeconomic backgrounds. This point highlights the significant role of informational, advising, and counseling services for high school students prior to college matriculation (Lin & Borden, 2016; Perna, 2010). Also, high schools do not have sole responsibility for these services. Colleges must work with high school counselors to ensure that baccalaureate-aspiring high school students know about feasible program options for continuing their education as well as provide support in overcoming obstacles and transitioning into college-level studies (Wang & Wickersham, 2014).

Third, the persistence of AP participants was more likely to be linked with institutional grant programs and that of DC participants to private, other, and parent loans. The analysis of the life-table indicated that the number of students leaving the school was the largest in the first year of college, and the estimated risk of dropping out was lower in the later academic years. The model showing the time-varying nature of student persistence may help institutions identify students’ risk of dropping out at any point, especially in the transition period, so they can design the mechanisms (i.e., early alert warning systems) to keep students on track to baccalaureate completion. Our results also provide evidence that different sources of aid are differentially related to the likelihood of student dropout risks, indicating that students’ needs respond to various types of aid in different ways. The overall likelihood test of the model for both noninteraction and interaction terms is significant, revealing that there is a significant difference between these various factors (student precollege experience, demographics, source of aid, and amount of dollars) and student departures. AP participants were likely to receive different kinds of aid. That is, AP participants were more likely to qualify for institutional grants while DC participants, whose socioeconomic backgrounds were generally lower, were more likely to qualify for loan programs, thereby incurring higher risk of graduating with indebtedness, which can be a demotivating factor. This difference further indicates that the socioeconomic gap in educational opportunities still remains. In sum, student participation in AP/DC programs has continued to grow and diversify over the past few decades, while educational pathways and trajectories are becoming more complex after taking finances into account. Understanding the role of student financial aid is a critical factor in examining the efficiency and effectiveness of DC and AP programs in the national quest for equity of educational opportunities in U.S. higher education (Anderson & Hearn, 1992). Further research is needed to investigate whether and how the effects of aid programs mediate or moderate student persistence policies that intentionally link the two warrant empirical examination. Given the rise of institutional aid and the decline of need-based aid, taking an inquiry-driven approach to such issues will help policy makers and institutional leaders identify the extent to which the effects of differentiated aid are truly equitable for students.

Limitations

Given that this is an exploratory study, care should be taken not to over-generalize the results. Several limitations deserve further discussion. First, we attempted to address the prior limitation suggested by Chen and DesJardins (2008) by using types and amounts of financial aid. Nonetheless, we tracked only 4 years of student data, which may limit the scope of the findings, suggesting the importance of continued efforts to capture the interplay between differentiated aid and DC/AP participation as these affect students’ persistence in pursuit of a college degree. In addition, the analysis was limited by the availability of data, so there is a lack of information pertaining to specific factors such as academic integration and environmental factors to account for more comprehensive picture. This study is also limited to a single institution, although it should be added that, given the size and academic and socioeconomic diversity of this institution, it is a reasonable proxy at least for large, Midwestern public universities, if not public, socioeconomically diverse institutions more generally. Finally, the data were from institutional records that included students’ nonresponses to some items. Accordingly, the pair-wise deletion method was employed to deal with missing data in our dataset. By doing so, we could overcome the issue of deleting too many cases, which would have resulted in samples too small for the analysis.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Ching-Hui Lin, PhD, assistant research fellow at National Sun Yat-sen University of Taiwan, served in Asia University and Indiana University. Her expertise falls into postsecondary equity and access, institutional research, and college student experiences, as well as quantitative method.

Victor M. H. Borden, PhD, is professor of Higher Education within the Department of Educational Leadership and Policy Studies at Indiana University Bloomington. He also directs the Project on Academic Success and the Carnegie Classification of Institutions of Higher Education through the Indiana University Center for Postsecondary Research.

Jyun-Hong Chen, PhD, assistant research fellow at National Sun Yat-sen University of Taiwan, serves in the Office for Institutional Research. His expertise includes statistical modeling, measurement and evaluation, and computer-based testing.

References

Allen

(2010). A comprehensive literature review and bibliography. New York, NY: CUNY Collaborative Programs.

Allen

Dadgar

(2012). Does dual enrollment increase students’ success in college? Evidence from a quasi‐experimental analysis of dual enrollment in New York City. New Directions for Higher Education, 2012(158), 11–19.

B. P.

(2013a). The influence of dual enrollment on academic performance and college readiness: Differences by socioeconomic status. Research in Higher Education, 54, 407–432.

B. P.

(2013b). The impact of dual enrollment on college degree attainment: Do low-SES students benefit? Educational Evaluation and Policy Analysis, 35, 57–75.

B. P.

Taylor

J. L.

(2015). Are dual enrollment students college ready? Evidence from the Wabash National Study of Liberal Arts Education. Education Policy Analysis Archives. Retrieved from http://epaa.asu.edu/ojs/article/view/1781/1624

Anderson

M. S.

Hearn

J. C.

(1992). Equity issues in higher education outcomes. In Becker

W. E.

Lewis

(Eds.), The economics of American higher education (pp. 301–334). Boston, MA: Kluwer Academic Publishers.

Bean

(1980). Dropouts and turnover: The synthesis and test of a causal model of student attrition. Research in Higher Education, 12(2), 155–187.

Bean

(1983). The application of a model of turnover in work organization. The Review of Higher Education, 6, 129–148.

Bean

(1985). Interaction effects based on class level in an explanatory model of college student dropout syndrome. American Educational Research Journal, 22(1), 35–64.

10.

Bean

(1990). Why students leave: Insights from research. In Hossler

Bean

(Eds.), The strategic of management of college enrollments. San Francisco, CA: Jossey-Bass.

11.

Bean

Metzner

B. S.

(1985). The estimations of a conceptual model of nontraditional undergraduate student attrition. Review in Higher Education, 27(1), 15–38.

12.

Boswell

(2001). State policy and postsecondary enrollment options: Creating seamless systems. New Direction for Community Colleges, 113, 7–14.

13.

Borden

Taylor

J. L.

Park

Seiler

(2013). The dual credit information project: State policy and quality assurance for college-level courses offered to high school students. Chicago, IL: The Higher Learning Commission.

14.

Braxton

(1999). Theory and elaboration and research development: Toward a fuller understanding of college student retention. Journal of College Student Retention: Research, Theory, and Practice, 1, 93–97.

15.

Braxton

J. M.

(2000). Reworking the student departure puzzle. Nashville, TN: Vanderbilt University Press.

16.

Braxton

J. M.

Bray

Berger

(2000). Faculty teaching skills and their influences on the college student departure process. Journal of College Student Development, 41, 215–227.

17.

Braxton

J. M.

Hirschy

A. S.

McClendon

S. A.

(2004). Understanding and reducing college student departure. San Francisco, CA: Jossey-Bass.

18.

Braxton

J. M.

Sullivan

A. S.

Johnson

R. M.

(1997). Appraising Tinto's theory of college student departure. Higher Education: Handbook of Theory and Research, 12, 107–164.

19.

Cabrera

A. F.

Nora

Castaneda

M. B.

(1992). The role of finances in the persistence process: A structural model. Research in Higher Education, 33(5), 571–593.

20.

Cabrera

A. F.

Nora

Castaneda

M. B.

(1993). College persistence: Structural equations modeling test of an integrated model of student retention. The Journal of Higher Education, 64(2), 123–139.

21.

Chen

DesJardins

S. L.

(2008). Exploring the effects of financial aid on the gap in student dropout risks by income level. Research in Higher Education, 49(1), 1–18.

22.

Chen

DesJardins

S. L.

(2010). Investigating the impact of financial aid on student dropout risks: Racial and ethnic differences. Journal of Higher Education, 81(2), 179–208.

23.

College Board. (2011). Low-SES students and college outcomes: The role of AP® fee reductions. Retrieved from https://research.collegeboard.org/sites/default/files/publications/2012/7/researchreport-2011-9-low-ses-college-outcomes-ap-fee-reductions.pdf

24.

College Board. (2017). States propel students to success in advanced placement. Retrieved from https://www.collegeboard.org/releases/2017/states-propel-students-success-advanced-placement

25.

DesJardins

S. L.

Ahlburg

D. A.

McCall

B. P.

(1999). An event history model of student departure. Economics of Education Review, 18(3), 375–390.

26.

DesJardins

S. L.

(2003). Even history methods: Conceptual issues and an application to student departure from college. In Smart

J. C.

(Ed.), Higher education: Handbook of theory and research ( Vol. 18, pp. 421–471). Dordrecht, The Netherlands: Kluwer.

27.

Engberg

M. E.

(2011). Pervasive inequality in the stratification of four-year college destinations. Paper presented at the annual conference of the Association for Study of Higher Education (ASHE), Charlotte, North Carolina.

28.

Geiser

Santelices

(2004). The role of advanced placement and honors courses in college admission. Center for Studies in Higher Education. Retrieved from http://escholarship.org/uc/item/3ftg8rz

29.

Godfrey

K. E.

Wyatt

J. N.

Beard

J. J.

(2016). Exploring college outcomes for low-income AP® exam takers with fee reductions (College Board Research Report No. 2016-2). New York, NY: College Board.

30.

Hargrove

Godin

Dodd

(2008). College outcomes comparisons by AP and Non-AP high school experiences (College Board Research Report No.2008-3). New York, NY: Author.

31.

Hillman

(2013). Economic diversity in elite higher education: Do no-loan programs impact pell enrollments? Journal of Higher Education, 84(6), 806–833.

32.

Hillman

(2014). College on credit: A multilevel analysis of student loan default. Review of Higher Education, 37(2), 169–195.

33.

Hofmann

(2003). College credit in high school: Increasing college attainment rates for underrepresented students. Change, 35(4), 42–48.

34.

Hofmann

Vargas

Santos

(2008). Blending high school and college: Rethinking the transition. New Directions for Higher Education, 14, 15–25.

35.

Hossler

Ziskin

Kim

Cekic

Jacob

P. K.

(2009). Student aid and its role in encouraging persistence. In Baum

McPherson

Steele

(Eds.), The effectiveness of student aid policies: What the research tells us (pp. 101–114). New York, NY: The College Board.

36.

Johnston

D. B.

Marcucci

P. N.

(2010). Financing higher education worldwide: Who pays? Who should pay? Baltimore, MD: Johns Hopkins University Press.

37.

Karp

M. M.

Bailey

T. R.

Hughes

K. L.

Fermin

B. J.

(2005). Update to state dual enrollments policies: Addressing access and quality. Washington, DC: U.S. Department of Education.

38.

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. New York, NY: Community College Research Center, Columbia University.

39.

Karp

M. M.

Calcagno

J. C.

Hughes

K. L.

Jeong

D. W.

Bailey

T. R.

(2008, February). Dual enrollment students in Florida and New York City: Postsecondary outcomes (CCRC Brief, pp. 1–6). New York, NY: Community College Research Center, Columbia University.

40.

Kim

Kirby

Bragg

D. D.

(2006, May). Dual credit: Then and now (In Brief, 1–4). Champaign, IL: University of Illinois at Urban-Champaign, Office of Community College research and leadership.

41.

Klopfenstein

(2010). The effects of AP participation on time to college graduation. Austin, TX: National Center for Postsecondary Research.

42.

Krueger

(2006, March). Dual enrollment: Policy issues confronting State policymakers (Policy Brief). Denver, CO: Education Commission of the States.

43.

Lin

C. H.

Borden

V. M.

(2016). Learning from U.S. research on the impact of financial aid on college student outcomes to advance institutional research in Taiwan. Contemporary Educational Research Quarterly, 24(4), 9–31.

44.

McPherson

M. S.

Shapiro

M. O.

(1998). The student aid game. Princeton, NJ: Princeton University Press.

45.

Meade

Hofmann

(2007/2007). CUNY college now: Extending the reach of dual enrollment. In Hofmann

Vargas

Venezia

Miller

(Eds.), Minding the gap: Why integrating high school with college makes sense and how to do it. Cambridge, MA: Harvard Education Press.

46.

National Survey of Student Engagement. (n.d.). Engagement insights: Survey findings on the quality of undergraduate education. Bloomington, IN: Indiana University Center for Postsecondary Research. Retrieved from http://nsse.indiana.edu/NSSE_2016_Results/pdf/NSSE_2016_Annual_Results.pdf

47.

Paulsen

St. John

(2002). Social class and college costs: Examining the financial nexus between college choice and persistence. The Journal of Higher Education, 73, 189–236.

48.

Pascarella

E. T.

Terenzini

(1980). Toward the validation of Tinto's model of college student attrition: A review of recent studies. Research in Higher Education, 12(3), 271–282.

49.

Perna

(2010). Understanding the working college student. Academe, 96, 30–33.

50.

Pike

Kuh

(2005). First- and second-generation college students: A comparison of their engagement and intellectual development. The Journal of Higher Education, 76(3), 276–300.

51.

Raudenbush

S. W.

Bryk

A. S.

Congdon

R. T.

(2005). HLM 6: Hierarchical linear and nonlinear modeling (Version 6.02) [Software]. Lincolnwood, IL: Scientific Software International, Inc.

52.

Rutkowski

Plucker

, Borden, V. M., & Prusinski

(2011). Not all credits are created equal: Differential impacts of dual and advanced placement credits on higher education outcomes. Report commissioned by the Office of the Vice President for University Regional Affairs, Planning, and Policy, Indiana University Bloomington, IN.

53.

Scott-C

(2015). The role of financial aid in promoting college access and success: Research evidence and proposals for reform. Journal of Student Financial Aid, 45(3), 7–22.

54.

Singer

J. D.

Willett

J. B.

(1993). It’s about time: Using discrete-time survival analysis to study duration and the timing of events. Journal of Educational and Behavioral Statistics, 18(2), 155–195.

55.

Singer

J. D.

Willett

J. B.

(2003). Applied longitudinal data analysis: Modeling change and event occurrence. New York, NY: Oxford University Press.

56.

Smith

(2007). Why expand dual-credit programs? Community College Journal of Research and Practice, 31, 371–387. doi:10.1080/10668920600932884

57.

Speroni

(2011). Determinants of students’ success: The role of Advanced Placement and dual enrollments programs. New York, NY: National Center for Postsecondary Research.

58.

St John

E. P.

(1992). Workable models for institutional research on the impact of student financial aid. The Journal of Student Financial Aid, 22(3), 13–26.

59.

St John

E. P.

(2000). The impact of student aid on recruitment and retention: What the research indicates. New Directions for Student Services, 2000(89), 61–75.

60.

St. John

E. P.

Tuttle

(2000). Persistence by undergraduates in an urban public university: Understanding the effects of financial aid. The Journal of Student Financial Aid, 30(2), 23–37.

61.

St. John

E. P.

Paulsen

M. B.

Carter

D. F.

(2005). Diversity, college costs, and postsecondary opportunity: An examination of the financial nexus between college choice and persistence for African Americans and Whites. The Journal of Higher Education, 76(5), 545–569.

62.

St. John

E. P.

Paulsen

M. B.

Starkey

J. B.

(1996). The nexus between college choices and persistence. Research in Higher Education, 37(2), 175–220.

63.

St. John

E. P.

Cabrera

A. F.

Nora

Asker

E. H.

(2000). How can finance research inform the reconceptualization of persistence models? In Braxton

J. M.

(Ed.), Reworking the student departure puzzle. Nashville, TN: Vanderbilt University Press.

64.

St. John

Weber

(2001). State policy and the affordability of public higher education: The influence of state grants on persistence in Indiana. Research in Higher Education, 42, 401–428.

65.

St. John

Musoba

Simmons

Chung

(2002). Meeting the access challenge: Indiana’s Twenty-first Century Scholars Program (New Agenda Series, Vol. 4). Indianapolis, IN: Lumina Foundation for Education.

66.

St. John

Musoba

Simmons

Chung

Schmit

Peng

(2004). Meeting the access challenge: An examination of Indiana’s twenty-first century scholars program. Research in Higher Education, 45, 829–873.

67.

Stinebrickner

(2003). Understanding educational outcomes of students from low-income families: Evidence from a liberal arts college with a full-tuition subsidy. The Journal of Human Resources, 38, 591–617.

68.

Swanson

J. L.

(2008). An analysis of the impact of high school dual enrollment course participation on postsecondary academic success, persistence, and degree completion. Paper presented at the meeting of the National Association for Gifted Children, Tampa, FL.

69.

Taylor

J. L.

Borden

Park

(2015). State dual credit policy: A national perspective. In L

Taylor Pretlow

(Eds.), Dual enrollment policies, pathways, and perspectives. New Directions for Community Colleges. San Francisco, CA: Jossey-Bass.

70.

Taylor

J. L.

Pretlow

(2015a). Editors’ notes. New Directions for Community Colleges, 1–7. Retrieved from https://dx-doi-org.web.bisu.edu.cn/10.1002/cc.20127

71.

Taylor

J. L.

Pretlow

(Eds.). (2015b). Dual enrollment policies, pathways, and perspectives. New Directions for Community Colleges. San Francisco, CA: Jossey-Bass.

72.

Tinto

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

73.

Tinto

(1982). Limits of theory and practice in student attrition. The Journal of Higher Education, 53(6), 687–700.

74.

Tinto

(1987). Leaving college: Rethinking the causes and cures of student attrition (1st ed.). Chicago, IL: University of Chicago Press.

75.

Tinto

(1988). Stages of student departure: Reflections on the longitudinal character of student leaving. The Journal of Higher Education, 59(4), 438–455.

76.

Tinto

(1993). Leaving College: Rethinking the causes and cures of student attrition (2nd ed.). Chicago, IL: University of Chicago Press.

77.

Thomas

Marken

Gray

Lewis

(2013). Dual credit and exam-based courses in U.S. public high schools: 2010–11 (NCES 2013-001, U.S. Department of Education). Washington, DC: National Center for Education Statistics.

78.

Toutkoushian

R. K.

Hossler

DesJardins

S. L.

McCall

B. P.

Canche

M. G.

(2013). Effect of TFCS on student aspirations and outcomes (pp. 1–52). Paper presented at the Association for Education Finance and Policy, New Orleans, LA.

79.

Wang

Wickersham

(2014). Postsecondary co-enrollment and baccalaureate completion: A look at both beginning 4-year college students and baccalaureate aspirants beginning at community colleges. Research in Higher Education, 55(2), 166–195.

80.

Yashin

A. I.

Vaupel

J. W.

Iachine

I. A.

(1995). Correlated individual frailty: An advantageous approach to survival analysis of bivariate data. Mathematical Population Studies, 5, 145–159.

81.

Zerquera

Gross

J. P. K.

(2015). Context matters: A critical consideration of Latino student success outcomes within different institutional contexts. Journal of Hispanic Higher Education, 16(3), 209–231.