Advertisement

Research in Higher Education

, Volume 60, Issue 4, pp 458–484 | Cite as

Can Transfer Guides Improve the Uptake of Major Prerequisites? Evidence from Ohio’s Transfer and Articulation Policy Reform

  • George SpencerEmail author
Article

Abstract

This study investigates the use of transfer guides to help students identify transferrable prerequisite credits for academic majors. Employing administrative data of students enrolled in Ohio community colleges, I examine the impact of these curricular roadmaps, called Transfer Assurance Guides (TAG), which were featured in reforms of the state's articulation policy. Leveraging variation in the availability of TAGs as a natural experiment, I estimate the impact of these guides on prerequisite course-taking in the fields of history, business, and economics. I also consider whether the effect varies for first-year students compared to returning students. I find that effects on prerequisite course-taking differed across academic majors, and the effects were lower for students enrolled in their first year compared to returning students. The findings suggest that transfer guides may affect student course-taking, but the impact is small and varies by the preparedness level of students.

Keywords

Community college Transfer Prerequisite coursework Articulation policy 

Introduction

Students who first enroll in community colleges have conventionally been expected to complete their lower-division coursework before transferring to a four-year college (Cohen et al. in The American community college (6th ed), Jossey-Bass, San Francisco, 2014). Although community colleges often structure transfer pathways to satisfy all general education and major prerequisite coursework—a threshold of approximately sixty credits, many students fall short of transferring with these requirements completed. Research shows that community college students who first enrolled in 2003 earned 34.7 credits, on average, before transferring to a four-year college (Simone in Transferability of postsecondary credit following student transfer or coenrollment. statistical analysis report. (NCES 2014-163), U.S. Department of Education, National Center for Education Statistics, Washington, DC, 2014). Thus, many students transfer from community colleges at different times in their academic progress and without all of the credits needed to fulfill their lower-division requirements.

Bachelor’s degree-seeking students may be disadvantaged if they transfer to 4-year institutions without completing all lower-division requirements. While 2-year college entrants have a lower probability of bachelor’s degree completion compared to those who first enroll at 4-year colleges (Alfonso 2006; Doyle 2009; Goldrick-Rab and Pfeffer 2009; Long and Kurlaender 2009; Sandy et al. 2006; Stephan et al. 2009), research shows that students who transfer with sixty credits are as likely to attain a bachelor’s degree as rising juniors at a 4-year college (Dietrich and Lichtenberger 2014; Melguizo and Dowd 2009; Melguizo et al. 2011; Monaghan and Attewell 2015). As such, the community college “penalty” may be diminished if students transfer successfully with a standard threshold of sixty credits from lower-division coursework.

Additionally, when students fail to identify and complete college-level courses deemed transferrable, they may lose credits in the transfer process or fail to have their credits meet all requirements (Monaghan and Attewell 2015). Credit loss may commonly occur when the receiving, 4-year institution does not accept the credits for community college coursework because the curricula does not meet the same level of rigor, content, or learning objectives. Rather than applying to meet the requirements for a specific major, the credits for community college courses may be accepted only as an elective instead (Cohen et al. 2014). Recent evidence from qualitative research suggests that transfer students may be most likely to lose credits for the proportion of credits that would apply for a specific major compared to elective credits (Hodara et al. 2016). Therefore, even if students are able to transfer the credits for the majority of their general education coursework, they may be required to retake pre-requisite courses, which would exacerbate costs and extend the time to degree.

This study focuses on the impact of a policy strategy intended to improve the uptake of transferrable, prerequisite coursework. In the fall semester of 2005, academic roadmaps called Transfer Assurance Guides (TAG) were introduced to improve the transfer function as a component of articulation policy reforms in Ohio. The new TAGs were used to communicate a sequence of prerequisite courses required for upper-division coursework in specific majors (Tafel 2010). The credits for courses identified in TAGs were also guaranteed to transfer across all state colleges and universities.

Using administrative data of all students enrolled at Ohio community colleges from 2004 though 2007, this study uses several strategies to identify the effect of TAG availability for students attending community colleges. To estimate the effect of TAG availability on students’ course-taking behavior, I exploit variation in the introduction of TAGs across Ohio’s public community colleges. Specifically, I use difference-in-differences to compare the course-taking outcomes of students enrolled before and after TAGs are introduced in 2005. I also use a difference-in-difference-in-differences approach to explore variation in the effect of TAG availability for returning students compared to those who first enroll when TAGs are introduced.

In what follows, I first review the relevant background literature and theoretical perspectives grounding this study. I subsequently discuss Ohio’s articulation reform pertaining to the Transfer Assurance Guides (TAGs) and I introduce my research questions. I then describe the data and empirical strategies used to answer the research questions and I detail the results. I conclude with a discussion of the findings.

Background and Context

Structural Barriers at Community Colleges

Community college students are often ill-equipped to successfully persist and complete a degree program. Rosenbaum, Deil-Amen, and Person (2006) found that many students attending community college lacked the knowledge and skills required to successfully navigate bureaucratic procedures, program requirements, and resources of support. Referencing Bourdieu and Wacquant (1992), the authors’ define this social know-how as information capital (Person et al. 2006; Rosenbaum et al. 2006). Research has well documented how lacking information is negatively associated with various postsecondary-related decisions such as whether to enroll in college, how to pay for college costs, and enrolling in credit-bearing coursework, among others (Avery and Kane 2004; Bettinger et al. 2012; Goldrick-Rab 2010). Because many community college students first matriculate without clear career goals, lacking information capital further complicates an understanding of the educational pathways needed to attain a credential (Gardenhire et al. 2006; Moore and Shulock 2011).

Considering the informational challenges facing many students, scholars have argued that curricular structure plays an incredibly important role in either facilitating or impeding student success (Bailey et al. 2015; Rosenbaum et al. 2006; Scott-Clayton 2011). Van Noy et al. (2016) have defined structure as the integration of policies and practices intended to support community college students in making informed decisions regarding their educational pursuits. At community colleges, the curricular structure provides students with considerable flexibility in order to facilitate transfer to a bachelor’s degree program as well as sub-baccalaureate degrees (e.g. associate degrees and certificates).

On one hand, curricular flexibility is a considerable strength of the community college system: by allowing students to explore numerous degree options—often for numerous major fields of study and with multiple course delivery methods—these institutions can accommodate different educational needs and personal circumstances. These accommodations are especially important since many community college students are likely to engage in discontinuous enrollment behaviors like temporarily interrupting their enrollment (i.e. stopping out), enrolling part-time, and taking remedial coursework (Goldrick-Rab 2010).

On the other hand, the organizational structure of community colleges may be too complex for students to navigate essential degree requirements, especially for academic programs. Scott-Clayton (2011) argues that students at community colleges would be best served by a program of study with a curriculum that is clearly prescribed and helps to prevent deviations from the designated pathway; however, most community colleges are found to fall short of providing this standard. Bailey et al. (2015) have described community colleges as offering a “cafeteria-style” model in which students enroll in a wide array of courses. Because academic programs are often less structured compared to occupational degree programs, many students lack a coherently structured curriculum when they first enroll (Van Noy et al. 2016). Given the plethora of choices, students may be inadvertently distracted away from optimal degree pathways (Jenkins and Cho 2013; Moore and Shulock 2011; Rosenbaum et al. 2006).

Structural challenges are exacerbated for bachelor’s degree-seeking students due to the complexities of transferring to 4-year colleges and universities. The transfer process is especially complicated because specific transfer requirements will vary across states, institutions, and major departments. Indeed, transferring requires navigating institutions with different curricular standards and course numbering systems. Thus, students who intend to transfer must first investigate and complete the requirements for the 4-year institution they aspire to attend and the department for their chosen program of study. Students risk losing credits in the transfer process if they do not adhere to the prescribed coursework of an existing agreement (Jenkins and Cho 2013; Roksa and Keith 2008). Yet, research shows that transfer requirements are often difficult to identify and complicated for students to understand (Fincher et al. 2014).

Because community colleges are often strained to meet students’ informational needs, many students are unprepared to transfer successfully. Research shows that counseling services are targeted primarily for new students during their orientation period (Karp 2013), counseling services are often difficult for students to find (Cox 2009; Deil-Amen and Rosenbaum 2003; Rosenbaum et al. 2006), and counselor-to-student ratios are commonly found to be in excess of 1 to 1000 (Bailey et al. 2015; Gallagher 2015; Grubb 2006). Because student advising is often poorly resourced and underutilized, students often acquire credits in programs that do not prepare them to transfer seamlessly to upper-division courses (Grubb 2006; Jenkins and Cho 2013). By failing to distinguish between transferrable and non-transferrable courses, students risk failing to acquire a coherent sequence of courses for an optimal program of study (Crosta 2014; Gardenhire et al. 2006; Goldrick-Rab 2010; Zeidenberg and Scott 2011).

Transfer Guides: Articulation Policy and Counseling Strategy

Transfer guides—also referred to as transfer maps—are used as an approach to facilitate the identification of transferrable courses. According to the Education Commission of the States, 35 states in 2010 offered transfer guides as an approach to facilitate statewide articulation (Smith 2010). Created by faculty within specific disciplines, transfer guides identify pathways that clearly define a sequence of lower-division coursework including general education courses and prerequisites for a given major (Jenkins and Cho 2013; Van Noy et al. 2016). These guides also provide community colleges with a systematic counseling strategy that counselors can use to help students select a program of study (Van Noy et al. 2016). Thus, students can progress more efficiently through coursework in which the credits are guaranteed to transfer across all state public institutions and will best prepare them academically to achieve their degree program goals.

Although transfer guides may be a promising strategy to improve structural barriers, there is little evidence of the effectiveness for this articulation approach. Using data from the Beginning Postsecondary Study (BPS: 03/09), LaSota and Zumeta (2015) found a positive association of transfer rates among students in states with statewide transfer guides; among these students, the authors’ found that the odds of transfer was 57% higher than students in states without transfer guides. Still, it is unclear whether this association is attributed to trends that have a confounding influence on the outcomes or other higher education policies that these states engage into proactively address transfer barriers. Nevertheless, there is a growing body of evidence from behavioral economics suggesting that students benefit from “nudges” that will direct students to complete requirements for admission, financial aid, and other complex higher education processes (Bettinger and Baker 2014; Bettinger et al. 2012; Castleman et al. 2014; Hoxby and Turner 2013). If utilized, transfer guides may similarly nudge students to acquire credits from transferrable coursework but more evidence is needed to explore the effectiveness of this approach on student trajectories.

Policy Context

Ohio is the home of a vast higher education system that includes 23 community colleges and technical colleges, 14 four-year colleges and universities, and 24 branch campuses. Because there is significant transfer activity between Ohio’s public institutions, state policymakers have engaged in a series of initiatives to improve the transfer of credits. Originally approved in 1990, the state’s articulation and transfer policy first introduced a transferrable core of general education coursework called the Ohio Transfer Module (OTM). Students who complete the courses comprising the OTM are ensured that approximately two-thirds of their required lower-division semester hours would transfer universally to all public institutions (Ohio Board of Regents 2010). However, subsequent evaluations by the Ohio Board of Regents revealed that students continued to lose credits for lower-division courses not included in the OTM and they needed more helpful advising to navigate the transfer process (Tafel 2010).

In 2003, the state general assembly in Ohio enacted reforms of its statewide articulation policy with House Bill 95. The policy updates sought to address persistent barriers to bachelor degree attainment by improving direction to transferable coursework. Central to the articulation reform efforts was the introduction a transfer map approach called Transfer Assurance Guides (TAGs): new academic roadmaps that identified a sequence of prerequisite coursework for specific majors in addition to general education course requirements.1 TAGs were designed with the intention to supplement the OTM so that students could now be better prepared for upper-division courses.

Coinciding with the introduction of TAGs, a statewide course equivalency system was also implemented to ensure the successful transfer of credits for all courses recommended in the new guides. The Ohio Board of Regents collaborated with more than 600 faculty members from the state’s 2- and 4-year colleges in order to establish the new course equivalency system (Kisker et al. 2011). Each discipline-specific faculty defined new TAG pathways across specific majors by identifying the learning standards and competencies for coursework that would be deemed universally transferrable (Tafel 2010).

The faculty panels then evaluated pre-existing courses at Ohio colleges to determine if they “matched” the new standards. In this matching process, courses submitted for review were determined to serve the purpose of course equivalency if they matched 70% of the new benchmarks. This matching process was facilitated by newly established Ohio Articulation Numbers (OAN), which identified the course-specific standards (Tafel 2010). Consider an introductory Psychology course at Community College X has an institutional course numbering system identifying this class as Psychology 101; the credits for this class will be deemed universally transferable if the content and outcomes standards are equivalent to the course created for the Psychology Transfer Assurance Guide called Intro/Fundamentals of Psychology, which has the Ohio Articulation Number OSS021. Though procedurally complex, the new course equivalency was an effective means to establish curricula consistency across state institutions while also maintaining institutional autonomy.

Although pathways were eventually developed for more than 40 disciplinary pathways, TAGs were first introduced in the fall semester of 2005 for 25 fields of study (Kisker et al. 2011; Ohio Board of Regents 2010).2 When the courses were approved to meet universal transferability, guides were made available for students to identify the TAG equivalent coursework offered by each respective institution (Tafel 2010). At this time, TAGs were used as an advising tool during sessions with counselors, admissions interviews, orientations, and made available on websites for specific institutions.

The Current Study

Considering the structural barriers that many community college students encounter, the availability of transfer guides may positively change transfer and degree completion behavior. To this end, transfer guides provide the clarity needed to simplify transfer requirements and provide a curricular structure with fewer course options (Bailey et al. 2015). According to Gross and Goldhaber (2009), a pivotal function of articulation approaches such as transfer guides is to direct students to coursework in which the credits will be accepted by 4-year institutions. Therefore, plausible changes to post-transfer outcomes are unlikely to occur unless students are first redirected to transferrable pathways.

This study considers whether the availability of TAGs is able to affect student enrollment in transferrable, prerequisite courses and when students are most likely to do so. Research suggests that early enrollment in prerequisite courses is particularly important. Jenkins and Cho (2013) found that, when students in their first year of enrollment concentrate in a program of study, they are more likely to attain a degree. But because many students enroll without clear academic goals or enroll part-time, they may be unprepared to take major prerequisites in the first year of enrollment. Even when students have selected a major program of study, they may be required to finish remedial and general education coursework before they take major prerequisites. Indeed, in an evaluation of California’s 2010 articulation policy, Baker (2016) found that students enrolled for more terms were most likely to enroll in transferrable coursework. Therefore, the effect of TAG availability may be greater for returning students compared to those in their first year of enrollment.

To investigate the possible effect of TAG availability and differences across students, this study answers the following research questions:

Research Question 1 Did the introduction of Transfer Assurance Guides (TAGs) increase the probability of taking prerequisite coursework, and does this TAG effect differ across academic disciplines?

Research Question 2 Does the TAG effect vary by the duration of students’ enrollment in community college?

Data and Empirical Strategies

Data

To answer the research questions, I used longitudinal, administrative data maintained by the Ohio Board of Regents (OBR). These data include the entire universe of students enrolled throughout the Ohio public, higher education system. The OBR collects demographic information from student applications such as their gender, race, and age as well as measures from financial aid data such as eligibility for the federal Pell Grant. At the time of entry, information pertaining to the students’ highest academic intentions is also derived from student questionnaires. Additional transcript data are organized by student, enrolled term, and institution attended, which nicely facilitate an investigation of enrollment and course-taking behavior across years.

I restrict the sample for this study to students who attended Ohio community colleges for their first, second, or third year of enrollment during the 2004–2005, 2005–2006, and 2006–2007 school years.3 I then created a panel dataset structured by each of the three respective academic terms. For example, in the 2004–2005 school year, a student who initially matriculated in fall of 2002 would be enrolled for her third year, a student who matriculated in the fall of 2003 would be in her second year, and a student who starts in the fall of 2004 would be enrolled as a first year student. As such, students who started at one community college but transferred to another would remain in the sample.

Given the study’s focus on transfer-intending students, the sample is also restricted by institutional sector. Prior to the establishment of the University System of Ohio in 2007, Ohio’s public institutions of higher education were organized in systems with distinct administrative and governance structures in addition to differences in sector and mission. Considering these differences across the higher education system, this study is focused on students enrolled in one of the six local community colleges, which were established by county governments, or one of the nine state community colleges established by the Ohio Board of Regents.4 The final dataset includes a pooled sample of 159,158 student observations from 15 of Ohio’s community colleges during the years of interest. The dataset is further conditioned to facilitate the analysis, as described below.

Empirical Strategy

Although the availability of TAGs was not randomly assigned, I leverage variation across institutions in the approval of transferrable prerequisites to estimate their impact. Before Ohio colleges could use the new course maps to advise students of transferrable course sequences, major prerequisites were first subject to approval as meeting the new TAG standards. In the fall semester of 2005, all public colleges in Ohio offered coursework approved for transfer in at least one of the 25 TAG disciplines that were introduced at the time; however, not every institution offered courses in all of the 25 fields of study. Indeed, many colleges offered prerequisite courses that were not approved until at least the 2007–2008 school year or later.5 Accordingly, the use of TAGs as an advising tool varied across disciplines and institutions for the first few years to correspond with course approval in a given discipline at each school.

Of the 25 TAG disciplines that were introduced in 2005, I found between-school variation in the respective fields of history, business, and economics to accommodate the analysis.6 For each of the three disciplines, Table 1 juxtaposes community colleges with courses approved for TAG usage in 2005 with colleges whose courses in similar disciplines were approved at a later date.7 As presented in the table, the comparisons across different colleges garner different analytic samples for examinations of each discipline-specific TAG: specifically, 159,158 for history, 95,019 for business, and 128,483 for economics. Although there are some contextual differences between community colleges across the state, the initial availability of TAGs was not strongly associated with institutional characteristics that are likely to have an effect on the outcomes.8 Exploiting variation in the TAG approval process, I employ two analytical approaches to estimate the effect of discipline-specific maps on prerequisite course-taking. First, I use a difference-in-differences (DD) approach to compare the effects of course-taking among students before and after the introduction of TAGs. In a first difference estimate, I compare students enrolled at institutions where guides for a specific TAG discipline were available for approved courses in the fall of 2005. Consider College X, which offers a sequence of introductory courses in history that were deemed universally transferrable in 2005: I estimate the impact of TAG course-maps that were concurrently introduced in 2005 on the probability that a student would take an approved history prerequisite by comparing students enrolled at College X before 2005 with those enrolled after.
Table 1

The timing of transfer assurance guide (TAG) availability at Ohio Community Colleges, by discipline.

Source Ohio Department of Higher Education, Transfer Assurance Guide Reporting System (https://www.ohiohighered.org/transfer/tag)

 

Business

Economics

History

Cincinnati State Technical and Community College

FALL ‘05*

 

DELAYED

Clark State Community College

DELAYED

FALL ‘05

FALL ‘05

Columbus State Community College

FALL ‘05

FALL ‘05

DELAYED

Cuyahoga Community College

FALL ‘05

FALL ‘05

FALL ‘05

Jefferson Community College

FALL ‘05*

DELAYED

DELAYED

Edison State Community College

DELAYED

FALL ‘05*

FALL ‘05

Lakeland Community College

DELAYED

DELAYED

DELAYED

Lorain County Community College

FALL ‘05

DELAYED

FALL ‘05

Northwest State Community College

 

FALL ‘05

DELAYED

Owens State Community College

FALL ‘05*

FALL ‘05

FALL ‘05

Rio Grande Community College

 

DELAYED*

FALL ‘05

Sinclair Community College

FALL ‘05*

FALL ‘05*

FALL ‘05

Southern State Community College

DELAYED

DELAYED

DELAYED

Terra State Community College

FALL ‘05

DELAYED

DELAYED

Washington State Community College

FALL ‘05

FALL ‘05

DELAYED

FALL’05 identifies colleges who initially offer courses approved for TAG usage in that discipline during the fall of 2005. DELAYED identifies with colleges whose courses in similar disciplines were approved in the fall of 2007 or later. The asterisk (*) denotes colleges that are excluded from the sample in the analysis because of data limitations

However, a first difference comparison only provides a naïve estimate of the policy impact. For example, labor market changes during the period of TAG introduction may shift student interests in taking courses for specific majors in high demand. Using a second difference, I can account for trends from other Ohio community colleges where the discipline-specific TAGs were not introduced until 2007 or later. For example, I compare the course-taking behavior in history among students at College X before and after 2005 with course-taking trends for the same discipline among students at College Y. The DD specification can be expressed as the following:
$$y_{isct} = \gamma_{0} + \gamma_{1} (TAG_{s} \times POST_{t} ) + \delta_{t} + \sigma_{cs} + X_{isct} + \varepsilon_{isct}$$
(1)
where \(y_{isct}\) represents the probability of taking a prerequisite course for a given disciple in school year t for individual i in enrollment year c attending college s.\(X_{isct}\) is a vector of time-invariant characteristics of students including race, age, state residency status, eligibility for the federal Pell Grant, and an indicator for whether the student took the ACT examination. In an extension of this basic DD specification, I also control for time-varying institutional covariates. Derived from the Integrated Postsecondary Education System (IPEDS), the institutional-level covariates include the total number of students enrolled, the ratio of full-time-equivalent staff to students, and expenditures dedicated to instruction, academic support, and student services.

Importantly, \(\delta_{t}\), represents a vector of school-year fixed effects and \(\sigma_{cs}\) represents a vector of college-by-enrollment year fixed effects. The college-by-enrollment year fixed effects control for unobserved differences between students that have been enrolled for the same duration of time at the same college. For example, a given community college may offer unique orientation services for first-year enrollees or advising for returning students that may be associated with course-taking. Using these fixed effects, the estimates are identified by comparing each combination of college and enrollment year to itself over time (i.e. first year students at College X in the 2004 school year, the 2005 school year, and so on).

TAG is a dichotomous variable coded 1 if a student is enrolled at a community college in which prerequisites were approved in 2005 for the Transfer Assurance Guide in a specific discipline (e.g. history, business, or economics) and it is coded 0 if the college offered similar courses that were not approved until after 2007.9POST is a dichotomous variable indicating whether a student is enrolled in the school years following the introduction of the new course maps. The interaction of these terms indicates the average difference in prerequisite course-taking among students at colleges offering TAGs after the new guides were introduced. Therefore, \(\gamma_{1}\) is my coefficient of interest. Here, I substitute the main effects of TAG and POST with fixed effects, which also account for unobserved effects of institutions and statewide shocks that affect all community college students yearly. Standard errors are clustered at the college level to allow unobserved components of the error term, \(\varepsilon_{isct}\), to correlate at that level.

I address the second research question by employing a difference-in-difference-in differences (DDD) strategy. To investigate whether the effect of TAG availability varies by enrollment year status, I estimate a DDD model as follows:
$$\begin{aligned} y_{icst} & = \beta_{0} + \beta_{1} \left( {TAG_{s} \times POST_{t} } \right) + \beta_{2} \left( {POST_{t} \times ENROLLYR3_{c} } \right) \\ & \quad + \beta_{3} \left( {TAG_{s} \times ENROLLYR3_{c} } \right) + \beta_{4} \left( {TAG_{s} \times POST_{t} \times ENROLLYR3_{c} } \right) \\ & \quad + \delta_{t} + \phi_{c} + \theta_{s} + X_{isct} + \varepsilon_{icst} \\ \end{aligned}$$
(2)

As presented in Eq. (1), TAG and POST as well as the school year fixed effects and controls are the same. For this analysis, I separately include vectors of school and enrollment year fixed effects, \(\theta_{s}\) and\(\phi_{c}\), in order to facilitate an estimation of interactions with enrollment years. Specifically, the analysis focuses on interactions of key terms with, \(ENROLLYR3_{c}\), one of the enrollment year dummies indicating whether a student was returning for the third year since her initial enrollment.10 For example, ENROLLYR3 would be coded as 1 in school year 2004 for an enrolled student who first matriculated in 2002.

Equation (2) builds on the difference-in-differences specification by adding new interaction terms. The coefficient for \(\beta_{2}\) represents the average difference in course-taking among untreated, returning third year students after 2005, and \(\beta_{3}\) represents the average difference in course-taking among third year students who were enrolled at TAG colleges. The coefficient of interest, \(\beta_{4}\), is associated with the three-way interaction representing the average change in prerequisite course-taking among treated, third year students following the introduction of a TAG at colleges where they were offered for the specific discipline.

Limitations and robustness checks I also specify a number of robustness checks for both the DD and DDD estimates. The difference-in-differences strategy is dependent on observing changes to course-taking behavior at community colleges across years, with a key assumption of identification relying on parallel trends in course-taking between the groups. Thus, any differences in course-taking between students enrolled at colleges offering TAGs in a given discipline and those that do not should be attributed only to the introduction of TAGs in 2005.

One way of accounting for differences in course-taking trends is to control for school-specific linear trends. To implement this strategy, I extend the sample to include observations in the 2003–2004 school year and then I generate a linear trend for each school in which the 2003 school year begins with 1, 2004 is 2, and so on. Because I do not have consistent data pertaining to history course-taking in 2003, I only incorporate linear trends to models estimating the effect of the TAG availability in economics and business. In addition, I also account for alternative explanations for estimates of the DD and DDD effects by excluding observations from the 2006 school year. Because the new articulation reforms were enacted in the previous year, students in the 2006 school year may have be compelled to purposefully enroll in colleges offering transferrable courses in their respective major. Therefore, selection may overstate the possible effect of TAG availability.

Despite these robustness checks, unobserved factors may still threaten the validity of estimates regarding the TAG effect. Mainly, efforts to establish bilateral articulation between institutions may affect the efficacy of the statewide effort featuring TAGs. Articulation agreements themselves are not new and it is certainly the case that community colleges in Ohio have continued to advance transfer pathways deriving from preexisting bilateral agreements colleges that affect student course taking behavior. On one hand, the difference-in-differences techniques employed in the study nicely account for any time-invariant differences between institutions such as preexisting transfer relationships between institutions. On the other hand, if new bilateral transfer agreements or other competing transfer efforts were introduced at some community colleges simultaneously during the period in which transfer guides are introduced, these unobserved changes may have affected students’ course-taking outcomes, which could not be disentangled from effects of the transfer assurance guides.

Results

Describing Characteristics and Trends of the Samples Across Years

In Table 2, I examine the descriptive characteristics of students who enrolled in Ohio community colleges. Among students enrolled for their first, second, and third years, I compare the demographic differences between those in the 2004 school year before TAGs were introduced and those in the 2005 and 2006 school years. The table presents pre- and post-differences for each combination of institutions offering TAGs in 2005 for the history, business, and economics disciplines and comparison group colleges offering TAGs after 2007.
Table 2

Descriptive statistics of first, second, and third year students enrolled in Ohio Community Colleges from 2004 to 2006.

Source: Ohio Board of Regents

 

History

Economics

Business

TAG delayed

TAG initially introduced

TAG delayed

TAG initially Introduced

TAG delayed

TAG initially introduced

2004

2005–06

2004

2005–06

2004

2005–06

2004

2005–06

2004

2005–06

2004

2005–06

Demographic characteristics

 Age: under 20

0.61

0.61

0.49

0.50

0.66

0.65

0.47

0.48

0.59

0.58

0.53

0.55

 Female

0.55

0.54

0.51

0.49

0.56

0.56

0.52

0.49

0.58

0.57

0.56

0.54

 Black

0.12

0.14

0.17

0.16

0.07

0.08

0.19

0.19

0.08

0.10

0.20

0.20

 Hispanic

0.02

0.02

0.03

0.03

0.03

0.04

0.03

0.03

0.01

0.02

0.03

0.03

 Other Minorities

0.02

0.02

0.02

0.02

0.01

0.02

0.02

0.02

0.01

0.02

0.03

0.03

 Pell eligible

0.12

0.30

0.12

0.28

0.13

0.29

0.12

0.28

0.13

0.31

0.13

0.30

 In state

0.96

0.97

0.98

0.98

0.98

0.98

0.97

0.97

0.99

0.99

0.98

0.99

 Took ACT

0.43

0.42

0.36

0.36

0.45

0.43

0.36

0.37

0.41

0.41

0.39

0.39

Initial major program interests

 Business

0.13

0.13

0.12

0.12

0.12

0.13

0.12

0.12

0.12

0.13

0.13

0.13

 Liberal arts

0.03

0.03

0.05

0.05

0.03

0.03

0.05

0.05

0.03

0.03

0.06

0.05

 Education

0.03

0.03

0.06

0.05

0.05

0.04

0.05

0.04

0.03

0.03

0.05

0.04

 General education

0.22

0.22

0.14

0.14

0.23

0.24

0.17

0.17

0.28

0.28

0.17

0.18

 Health

0.23

0.24

0.22

0.21

0.24

0.24

0.22

0.22

0.24

0.25

0.24

0.24

 Studio arts

0.02

0.02

0.03

0.04

0.02

0.02

0.03

0.03

0.03

0.03

0.03

0.03

 Undecided

0.06

0.05

0.10

0.09

0.07

0.07

0.07

0.06

0.06

0.04

0.10

0.09

 STEM

0.15

0.15

0.18

0.20

0.13

0.12

0.19

0.21

0.10

0.10

0.13

0.13

Initial academic intentions

 Bachelor’s Degree

0.35

0.34

0.36

0.35

0.43

0.39

0.32

0.32

0.36

0.30

0.41

0.43

 Certificate

0.06

0.06

0.05

0.04

0.06

0.07

0.05

0.04

0.06

0.07

0.04

0.04

 Associate degree

0.41

0.37

0.30

0.28

0.36

0.36

0.34

0.31

0.47

0.44

0.31

0.28

Prerequisite course taken (in the respective discipline)

 Proportion enrolled

0.16

0.14

0.05

0.05

0.08

0.09

0.02

0.02

0.09

0.10

0.04

0.04

 Sample size

17,582

36,592

35,577

69,407

9672

19,627

33,387

65,797

6889

13,866

25,092

49,172

The samples consist of students enrolled at Ohio community colleges for their first, second, or third years during the 2004–2006 school years. Each sample contains students from different grouping of community colleges according to whether the college initially offered courses approved for Transfer Assurance Guide usage (TAG Initially Introduced) or if approval for these courses was delayed to 2007 or after (TAG Delayed)

Across grouped institutions and years, approximately half of Ohio community college students during this period were female and 40% had taken the ACT exam. The proportion of students of color—including Black, Hispanic, and other minorities—ranges from 14 to 22%, the majority of students are state residents, and most are under the age of 21. Regarding student’s initial academic program, approximately a quarter of all students enrolled in a general education program but a similarly large proportion had interests in health related programs. Majors in STEM (Science, Technology, Engineering, and Mathematics) as well as business are also relatively popular, with approximately 12 to 15% of students enrolled in each. In contrast, only a small percentage of students (under 10%) had an interest in Liberal Arts disciplines, which include history and economics among other social science and humanities disciplines.

Table 2 shows that there are small, within-group differences across years on most demographic characteristics. Between years, there was a 1–2% point difference in the proportion of female students, students under the age of 20, students with bachelor’s degree intentions, and those who took the ACT. At colleges where TAGs were not introduced, there were also a higher proportion of students of color in the 2005–2006 school years, but this demographic shift was marginal. The most substantial change occurred in the proportion of students who were eligible for the federal Pell Grant. This substantial change over time was consistent across colleges that did and did not offer TAGs. Across the groups of colleges, the proportion of Pell-eligible students increased by 16 to 18% points. Because this demographic shift is constant across institutions, the difference-in-differences strategy can account for this statewide, secular trend.

There are also interesting trends in course-taking across years between the grouped institutions. Among colleges that did not introduce a history TAG in 2005, there was a 2-percentage point decline in the proportion of students taking a prerequisite course in this discipline. Also, the proportion of students taking prerequisites in economics and business was higher after 2005 at colleges that did not initially offer TAGs. In contrast, among the colleges where TAGs were offered, the proportion of students taking these courses is noticeably smaller and there was no observed change in course-taking after 2005. This suggests that either TAGs did not have a substantial impact among the general population of students, or alternatively, the magnitude of any possible effect was practically very small. Because the target population of TAGs is generally very narrow, it is reasonable to assume that only a small proportion of students would be affected. Indeed, this further compels the importance of investigating effects among specific groups.

Estimating the Effect of the Transfer Assurance Guide (TAG) Introduction

In Table 3, I present the results for estimates of the introduction to discipline-specific TAGs using difference-in-differences (DD). The outcomes capture whether students who were enrolled in colleges offering TAGs in a given discipline were more likely to enroll in a prerequisite course after 2005. Because the outcomes are dichotomous—coded 1 if the student enrolled in a prerequisite course for the discipline—the models are fitted using logistic regression. Coefficients are presented as log odds with corresponding odds ratios for ease of interpretation. Each model in the table presents estimates using alternative DD specifications. I begin with a model including only the main effects of TAG, POST, and their interaction with fixed effects for enrollment year. I then present the results using a model substituting the main effects for college and school year fixed effects. I conclude with the preferred specification as presented in Eq. (1), which includes college-by-enrollment year fixed effects, first without institution-level controls and then with these covariates included.
Table 3

Effect of TAG availability on taking academic major prerequisite (odds ratios)

 

(1)

(2)

(3)

(4)

Log odds

Odds ratio

Log odds

Odds ratio

Log odds

Odds ratio

Log odds

Odds ratio

A. History prerequisites

 TAG × POST

0.107*

1.113

0.107**

1.112

0.117**

1.124

0.111**

1.118

(0.062)

 

(0.044)

 

(0.046)

 

(0.047)

 

 POST

− 0.134***

0.874

      

(0.048)

       

 TAG

− 1.061*

0.346

      

(0.575)

       

 N

159,158

159,158

159,158

159,158

B. Economics prerequisites

 TAG × POST

− 0.109

0.896

− 0.091

0.913

− 0.092

0.912

0.046

1.047

(0.144)

 

(0.146)

 

(0.140)

 

(0.084)

 

 POST

0.046

1.047

      

(0.139)

       

 TAG

− 1.351***

0.259

      

(0.461)

       

 N

128,483

128,483

128,483

128,483

C. Business prerequisites

 TAG × POST

− 0.238**

0.789

− 0.217**

0.805

− 0.221**

0.802

− 0.528**

0.590

(0.100)

 

(0.091)

 

(0.089)

 

(0.254)

 

 POST

0.092***

1.096

      

(0.017)

       

 TAG

− 0.842

0.431

      

(0.520)

       

 N

95,019

95,019

95,019

95,019

 Student controls

YES

YES

YES

YES

 Institution controls

   

YES

 Enrollment year FE

YES

YES

  

College FE

 

YES

  

School year FE

 

YES

YES

YES

 College × enrollment year FE

  

YES

YES

Reported are coefficients from logistic regressions. Coefficients are presented as log odds. Corresponding odds ratios are also presented for ease of interpretation. Student-level controls include race, age, state residency status, eligibility for the federal Pell Grant, and an indicator for whether the student took the ACT examination. Institutional-level covariates include the number of students enrolled, the ratio of FTE staff to students, and expenses for instruction, academic support, and student services. Standard errors in parentheses are clustered at the college level. TAG indicates whether a community college’s prerequisites were initially approved for the Transfer Assurance Guide. POST indicates the school years following the introduction of the new course maps. FE indicates fixed effects

*p < 0.1 **p < 0.05 ***p < 0.01

Panel A of Table 3 presents the estimated effect for students enrolled at colleges offering a TAG in history. In Model (1), the coefficient on the interaction suggests that—at colleges where a TAG in history was available—the odds of taking a prerequisite course in history was 11.3% higher, on average, among students who enrolled after 2005. This pattern is consistent across the other specifications, which show that there was a positive and statistically significant effect on course-taking in history. Model (3) suggests that, on average, there was a 12.4% increase in the odds of taking a prerequisite course in history, but after accounting for institution-level controls, the estimate suggests that there is an 11.8% increase in the odds for students enrolled at colleges offering a TAG in this discipline.

Estimates for the effect of TAG availability in other disciplines diverge with the findings for the history TAG. In Panel B of Table 3, I present the estimated effect for students enrolled at colleges offering a TAG in economics. Across specifications, the models show that the TAG effect was not statistically significant. In contrast, Panel C suggests that the introduction of business TAGs was associated with a decline in the odds of prerequisite course-taking in this field. Each model shows that the effect of TAG availability was negative and statistically significant. In Model (4), the coefficient on the interaction suggests that, on average, the odds of taking a prerequisite course in business was 41% lower among students who enrolled after 2005 at colleges offering a TAG in this discipline.

To address issues threatening the internal validity of these estimates, Table 4 presents estimates from robustness checks for all three discipline-specific TAGs. For parsimony, I only present the coefficients using my preferred specification including institution-level controls. Panel A presents estimates using a reduced sample that excludes observations in the 2006–2007 school year. While the effect of the history TAG availability is still significant, it is smaller in magnitude, suggesting that the odds of taking a prerequisite course only increased by 9.3%. Most notably, the effect of the business TAG is positive in direction but no longer statistically significant.
Table 4

Difference-in-differences robustness checks

 

History prerequisites

Economics prerequisites

Business prerequisites

Log odds

Odds ratio

Log odds

Odds ratio

Log odds

Odds ratio

A. 2006 school year excluded

TAG × POST

0.089***

1.093

0.002

1.002

0.229

1.257

 

(0.032)

 

(0.035)

 

(0.147)

 

N

107,019

86,643

63,720

B. School linear trends included (2003–2006 school years)

TAG × POST

− 0.195

0.823

− 0.168

0.877

 

 

(0.202)

 

(0.220)

 

N

  

171,835

127,171

Student controls

YES

YES

YES

Institution controls

YES

YES

YES

School year FE

YES

YES

YES

College × enrollment year FE

YES

YES

YES

Reported are coefficients from logistic regressions. Coefficients are presented as log odds. Corresponding odds ratios are also presented for ease of interpretation. Standard errors in parentheses are clustered at the college level. Student-level controls include race, age, state residency status, eligibility for the federal Pell Grant, and an indicator for whether the student took the ACT examination. Institutional-level covariates include the number of students enrolled, the ratio of FTE staff to students, and expenses for instruction, academic support, and student services. TAG indicates whether a community college’s prerequisites were initially approved for the Transfer Assurance Guide. POST indicates the school years following the introduction of the new course maps. FE indicates fixed effects

*p < 0.1 **p < 0.05 ***p < 0.01

In Panel B, I also produce estimates of the TAG effect using school linear trends to control for possible pre-existing differences between schools in course-taking. As observed in Table 2, there were small changes in course-taking among the comparison group colleges that were not similarly observed in the colleges offering TAGs. If pre-existing trends in course-taking explain these observed differences, my results may suffer from a possible violation of the parallel trends assumption. Because I have limited data of course-taking in history to accommodate this analysis, I only present estimates of this approach for the economics and business TAGs. The table shows that the null estimate for the TAG in economics is consistent. However, the negative estimate for TAG availability in business is no longer statistically significant once the linear trends are employed. Considering the results are not robust across specifications, the average difference in business course-taking for the full sample are likely unrelated to the introduction of business TAGs.

Estimating Differences in the TAG Effect by the Duration of Student Enrollment

Next, I explore whether the effect of TAG availability varies by the duration of a student’s enrollment in community college. Table 5 presents the results from models using the DDD specification introduced in Eq. (2) and institution-level controls. With dichotomous measures to indicate enrollment year status in a given year, this analysis focuses on differences in the TAG effect for students returning for their third year relative to those starting their first year. Here, \(\beta_{1}\) now represents the average difference in prerequisite course-taking among first year students at TAG colleges after 2005, and my coefficient of interest, \(\beta_{4}\), represents the impact specific to third year students. Because I am primarily concerned with the relative effect for returning students, my discussion is focused on estimates for the three-way interaction.
Table 5

Differential effects of TAG availability on taking a major prerequisite by enrollment duration

 

History prerequisites

Economics prerequisites

Business prerequisites

Log odds

Odds ratio

Log odds

Odds ratio

Log odds

Odds ratio

ENROLLYR3 × TAG × POST

0.103**

1.108

− 0.058

0.944

0.422***

1.524

 

(0.047)

 

(0.157)

 

(0.140)

 

TAG × POST

0.086*

1.090

0.063

1.066

− 0.630**

0.533

(0.046)

 

(0.104)

 

(0.256)

 

ENROLLYR3 × POST

0.001

1.001

0.121

1.128

− 0.072

0.930

(0.033)

 

(0.115)

 

(0.088)

 

ENROLLYR3 × TAG

0.058

1.060

0.718***

2.049

0.007

1.007

(0.127)

 

(0.179)

 

(0.121)

 

ENROLLYR3

− 0.119

0.888

− 0.304**

0.738

0.197***

1.217

(0.120)

 

(0.146)

 

(0.058)

 

ENROLLYR2

0.325***

1.384

0.406***

1.501

0.431***

1.539

(0.063)

 

(0.098)

 

(0.058)

 

Student controls

YES

YES

YES

Institution controls

YES

YES

YES

School year fixed effects

YES

YES

YES

College fixed effects

YES

YES

YES

N

159,158

128,483

95,019

Reported are coefficients from logistic regressions. Coefficients are presented as log odds. Corresponding odds ratios are also presented for ease of interpretation. Standard errors in parentheses are clustered at the college level. Student-level controls include race, age, state residency status, eligibility for the federal Pell Grant, and an indicator for whether the student took the ACT examination. Institutional-level covariates include the number of students enrolled, the ratio of FTE staff to students, and expenses for instruction, academic support, and student services. TAG indicates whether a community college’s prerequisites were initially approved for the Transfer Assurance Guide. POST indicates the school years following the introduction of the new course maps. ENROLLYR3 and ENROLLYR2 are enrollment year dummies indicating status as a student enrolled for the third year and the second year respectively

*p < 0.1 **p < 0.05 ***p < 0.01

The table shows that the effect of TAG availability does vary by enrollment year status for some disciplines. The first column suggests that among third year students at colleges offering a TAG after 2005, the odds of taking a prerequisite course in history was 10.8% higher on average. The odds were greater relative to first year students, who were also more likely to take a class in history after TAGs were introduced (OR = 1.090). Additionally, the table shows that estimates for the economics TAG did not vary by enrollment year, but there was a significant interaction for the TAG in business. Although the odds of taking a business course is negative among first year students, relative to these students, the odds of take a business course is 52% higher for third year students at colleges offering a business TAG.

Robustness checks of the DDD estimates are presented in Table 6. In Panels A and B, estimates of the TAG effect in history, economics, and business among third year students are consistent with the main results presented in Table 5. Therefore, estimates for the three-way interaction are robust to alternative specifications. This suggests that TAGs are more beneficial at improving prerequisite course-taking among returning students compared to those who enrolled for the first time, but the impact also varies across the TAG disciplines.11
Table 6

Difference-in-difference-in-differences robustness checks

 

History prerequisites

Economics prerequisites

Business prerequisites

Log odds

Odds ratio

Log odds

Odds ratio

Log odds

Odds ratio

A. 2006 school year excluded

 ENROLLYR3 × TAG × POST

0.171**

1.186

− 0.101

0.904

0.313*

1.367

(0.080)

 

(0.195)

 

(0.186)

 

 TAG × POST

0.049

1.051

0.028

1.029

0.183

1.201

(0.040)

 

(0.063)

 

(0.129)

 

 ENROLLYR3 × POST

0.004

1.004

0.084

1.088

− 0.073

0.930

(0.037)

 

(0.146)

 

(0.098)

 

 ENROLLYR3 × TAG

0.059

1.061

0.712***

2.038

0.002

1.002

(0.126)

 

(0.177)

 

(0.124)

 

 ENROLLYR3

− 0.121

0.886

− 0.304*

0.738

0.173***

1.189

(0.119)

 

(0.162)

 

(0.059)

 

 ENROLLYR2

0.314***

1.369

0.406***

1.500

0.390***

1.478

(0.064)

 

(0.124)

 

(0.048)

 

 N

107,019

86,643

63,720

B. School linear trends included (2003–2006 school years)

 ENROLLYR3 × TAG × POST

0.061

1.063

0.416**

1.516

 

(0.124)

 

(0.179)

 

 TAG × POST

− 0.224

0.800

− 0.243

0.784

 

(0.216)

 

(0.291)

 

 ENROLLYR3 × POST

− 0.015

0.985

− 0.126

0.881

 

(0.089)

 

(0.099)

 

 ENROLLYR3 × TAG

0.598***

1.818

0.015

1.015

 

(0.123)

 

(0.074)

 

 ENROLLYR3

− 0.163

0.850

0.256***

1.292

 

(0.101)

 

(0.046)

 

 ENROLLYR2

0.413***

1.510

0.455***

1.575

 

(0.092)

 

(0.060)

 

 N

  

171,835

127,171

 Student controls

YES

YES

YES

 Institution controls

YES

YES

YES

 School year fixed effects

YES

YES

YES

 College fixed effects

YES

YES

YES

Reported are coefficients from logistic regressions. Coefficients are presented as log odds. Standard errors in parentheses are clustered at the college level. TAG indicates whether a community college’s prerequisites were initially approved for the Transfer Assurance Guide. POST indicates the school years following the introduction of the new course maps. ENROLLYR3 and ENROLLYR2 are enrollment year dummies indicating status as a student enrolled for the third year and the second year respectively

*p < 0.1 **p < 0.05 ***p < 0.01

Discussion

Although many community college students seek bachelor’s degrees, they often experience challenges in acquiring the credits for coursework that will seamlessly transfer to four-year institutions. Community colleges offer multiple pathways to facilitate transfer but the curricular structure of these institutions may be far too complex for students to easily identify transferable coursework. The introduction of Transfer Assurance Guides (TAG) in Ohio was an attempt to address these structural challenges. This transfer guide strategy was intended to improve clarity of transfer requirements by specifying a sequence of courses that best prepare students for a given major. By supplementing the Ohio Transfer Module, TAGs were featured in efforts to establish articulation across all state public institutions for prerequisite coursework.

This study found that the introduction of TAGs had a varied effect on prerequisite course-taking across disciplines and student enrollment years, albeit the impact was small. I find that there was no difference in the effect of TAG availability for economics course-taking, but there was a small, positive effect in history course-taking. Among all students enrolled at colleges with approved courses in history, I found that, on average, the odds of taking a history prerequisite increased by an average of approximately 9–12% following the introduction of TAGs. However, compared to first year students, the odds of taking a history prerequisite was 10–17% higher for third year students. In contrast, effects on course-taking in business was mixed. I found that there was generally no difference in business course-taking following the introduction of TAGs, on average, but the odds for returning students was 37–52% higher for third year students relative to first-year students.

Widespread changes in course-taking across Ohio community colleges could partially explain heterogeneity in the TAG effect. Because I was unable to investigate the TAG effect for all 25 programs of study, perhaps TAGs affected student interests in ways that are unobserved in this study. At colleges introducing TAGs in multiple disciplines, students were enticed to consider several transfer options. Hence, the introduction of multiple transferable choices may have contributed to a re-sorting of students across classes and majors. For example, because business programs were among the more popular majors before the new reform efforts, students may have been less likely to take business prerequisites after TAGs were introduced if they were compelled to consider other transferable degree options instead.

In other words, new information communicated by TAGs may have prompted students to update their course-taking interests. As previously noted, TAGs feature information concerning major-specific prerequisites in addition to the other recommended coursework needed to fulfill general education requirements. For the TAG in business, the subjects for recommended courses include calculus, statistics, and microeconomics, which are commonly considered to be challenging for many students with lower mathematics achievement and self-efficacy. If students are more informed of these requirements, perhaps they are less likely to enroll in the major prerequisites if they are intimidated by the additional coursework requirements. Thus, TAGs may not only induce students who were reluctant to take any prerequisites before transferring, they likely also redirect students who would have otherwise taken courses in a different discipline.

Consistent with findings from Baker (2016), the TAGs also appear to affect returning students differently from first-year students. There are a number of factors that may explain variation across students’ duration of enrollment. First, it is important to note that some major prerequisites featured in TAGs may also fulfill the general education requirements for the Ohio Transfer Module, particularly those in history and economics since the OTM expects students to complete six semester hours in arts and humanities and an additional six in social sciences, among other core areas. This may explain why I find positive and statistically significant results for course-taking in history among returning students as well as those who are first-year students. In other words, compared to business, some courses for history and economics may serve the dual function of fulfilling the general education or major prerequisite requirements; however, economics may be a generally less popular option to fulfill the social sciences requirement compared to other disciplines.

Additionally, accumulating credits for major prerequisites would take longer if a student is enrolled part-time or has to take remedial coursework in her first enrolled semesters. Indeed, the descriptive characteristics of third-year students suggests that these students have completed 35 credits, on average, which amounts to the approximate number of general education course requirements.12 Compared to returning students, who have had more time to explore curricular offerings, first-year students may be unprepared to select prerequisite courses if they enroll in college without knowing what major they would like to pursue (Gardenhire et al. 2006). Therefore, transfer guides may only be effective for a very narrow population of community college students: those who have first completed their other required courses and have made a decision concerning their major program of study.

If students are generally unlikely to complete prerequisite requirements until their third year, they may be more inclined to transfer without first completing their lower-division coursework. However, community college students who matriculate to 4-year colleges without sixty credits may have a lower probability of earning a bachelor’s degree (Dietrich and Lichtenberger 2014; Melguizo and Dowd 2009; Melguizo et al. 2011; Monaghan and Attewell 2015). Without taking prerequisite courses before transferring, students may face barriers to be accepted into their academic department of choice with junior standing (Gross and Goldhaber 2009). In addition, students are also disadvantaged if they wait to concentrate in an academic major (Jenkins and Cho 2013).

Conclusion

Future research should consider the effectiveness of transfer guides in facilitating course decision-making. Because the current study could only leverage the availability of TAGs to estimate a potential impact, it is still unclear what effect this approach has for students who actively engage with using transfer guides or the number of students who do so. Qualitative research suggests that institutional context can influence how practitioners at 2-year colleges may interpret transfer policy and how this subsequently influences their advice to students (Chase 2014). Hence, the current findings may be driven, in part, by differences across institutions in the degree to which TAGs were publicized or used by advising staff to guide students. In addition, it is also important to further explore whether such guides are primarily helpful with supporting students to fulfill pre-existing major intentions or if students’ interests change according to the transferability of their initial degree program. Considering differences by discipline in the current study, it is important for future research to consider the role of transfer guides in influencing how students parse major coursework decisions.

Notably, Ohio has recently begun a new phase of articulation reforms featuring the guided pathways model advocated by researchers affiliated with the Community College Research Center at Teacher’s College (Jenkins et al. 2017). Although TAG pathways remain a feature of Ohio’s statewide articulation, the new reforms to improve the transfer function suggests that the past efforts may have been insufficient to address the range of needs for many community college students. For example, to my knowledge, the introduction of TAGs in Ohio was not accompanied by a robust change in funding for counseling resources or outreach. However, comprehensive advising should ideally supplement the use of such tools. Before a mapping approach can be used to support students in identifying coursework that is suitable for transfer, students also need help to identify their academic goals and to find pathways that will accelerate their accumulation of credits. In other words, providing information absent consistent guidance is not enough to nudge students in meaningful ways. As a new framework to help students in planning for transfer, the new guided pathways may better support students in the effort to complete programs according to career interests and goals they are first expected to first identify (Bailey et al. 2015).

Nevertheless, though few students were prompted by TAGs to enroll in prerequisite courses, the current study offers important insight for policy consideration. Whether a student was compelled by TAGs to change from her initial major of interest, or if she enrolled in a prerequisite course when she otherwise would not have, the student was likely advantaged by doing so. Because TAGs directed students to coursework that is universally transferrable, students may be able to preserve credits for their lower-division courses without the risk of spending additional time and money to retake courses. Students also benefit financially by taking prerequisite courses at a community college, which offer lower tuition costs compared to 4-year institutions, and those who completed the transferrable prerequisites may also matriculate to 4-year colleges better prepared for upper-division coursework. Although this study was unable to investigate the full extent of the TAG impact, perhaps the observed changes in course-taking signals an important shift in the way that community college students, particularly returning students, prepared for transfer. Thus, to ensure that students have improved post-transfer outcomes, articulation policies must ensure that students are first directed to the transferrable coursework. The TAG approach in particular may be an important feature of statewide articulation to achieve this effort, but given its relatively small impact, a more robust effort such as guided pathways may be required to ensure that students are adequately supported.

Footnotes

  1. 1.

    Examples of Transfer Assurance Guide Pathways are available on the website for the Ohio Department of Higher Education at https://www.ohiohighered.org/transfer/tag/definitions.

  2. 2.

    According to the University System of Ohio Transfer Assurance Guide Reporting System, TAGs were made available in the following 25 fields of study in the fall of 2005: Biology, Business, Chemistry (Organic), Chemistry (General Education), Communications, Dietetics, Economics, Education, English Literature, Engineering, Geography, Health, History, Journalism, Mathematics, Medical Laboratory, Music, Physics, Political Science, Public Relations, Psychology, Psychology, Studio Arts, Telecommunications, and Theatre.

  3. 3.

    Although I would ideally like to observe course-taking trends across multiple years to account for possible pre-existing differences, the consistency of course data across years varies by college and discipline because of changes to the numbering of courses from year-to-year as well as changes to the general course offerings. I only have consistent data for the same courses within the three disciplines of interest for this study from the 2004 to 2005, 2005 to 2006, and 2006 to 2007 school years.

  4. 4.

    Although I exclude students enrolled at technical colleges, which specialize in preparing students for career/technical education programs, I do not further restrict the sample by degree intentions. Because there are so many degree pathway options for community college students, it is often difficult to truly know whether or not they intend to transfer or not. Indeed, research shows that as many as 70–80% of community college students express the intention to attain a bachelor’s degree (Bailey et al. 2006; Horn and Skomsvold 2011). Yet, even among those who initially enroll without the intention to transfer, research finds that aspirations often “warm up” over time (Alexander et al. 2008; Rosenbaum et al. 2006). Therefore, the sample includes all students at the selected institutions despite their initial degree intentions.

  5. 5.

    I was able to determine the exact timing of TAG course approval using the Transfer Assurance Guide Reporting System available on the Ohio Department of Higher Education website (https://www.ohiohighered.org/transfer/tag). The system facilitates the matching of TAG prerequisite and introductory courses that are offered at Ohio colleges and universities with their corresponding Ohio Articulation Number, the date in which the equivalency became effective, and the specific major in which the course is articulated to transfer.

  6. 6.

    Variation in TAG course availability was also discovered in other disciplinary areas but I do not estimate the effect of TAGs introduced for these fields because of data limitations. For example, in some disciplines, such as the natural sciences, the proportion of students taking prerequisites from year to year was either inconsistent or incredibly rare. In addition, annual data for some courses within a given institution was not always available in the data provided by the Ohio Board of Regents.

  7. 7.

    Although courses that were not approved until after 2007 did not necessarily meet the 70% benchmark for the new statewide articulation standards, they are still intended to serve as major prerequisites that could be deemed transferable at specific institutions with a bilateral articulation agreement. In addition, while some colleges created new prerequisite courses that would meet the new standards, many simply modified the standards for pre-exiting courses (Tafel 2010).

  8. 8.

    Appendix Table 7 presents the mean summary statistics for the colleges grouped by TAG availability and p values from two-tailed t tests (See Table 1 for the specific colleges in each group). The table shows that there are no statistically significant differences at the .05 level for the presented institutional characteristics of colleges that offered TAGs for a given discipline in 2005 compared to the schools that delayed. In one exception, colleges that delayed a TAG in Business were rural institutions with lower student enrollment; however, because these characteristics also define a small percentage of the colleges offering business TAGs in 2005, the availability of TAGs for business—as with the other disciplines—appears to be orthogonal to institutional characteristics.

  9. 9.

    See Table 1 for a list of the institutions included in the sample for each discipline-specific TAG. As indicated by the asterisk, some colleges offering TAGs were excluded from the analytic sample because of data limitations as previously discussed.

  10. 10.

    Appendix 2 presents summary statistics from the pooled sample of third year returning students. The table shows that these students are more likely to be younger (under 20 years of age), White, and female. These students are also less likely to be eligible for the Pell Grant, which suggests that there are a lower proportion of low-income students and students of color who persist. In addition, a significant proportion of third year returning students first enrolled with the intention of aspiring to bachelor’s degree (40 %) and these students had accumulated 35 credit hours, on average. Although there is likely to be variation among third-year students regarding the their enrollment intensity, on average, these students are more advanced according to accumulated credits and completed requirements compared to first-year students

  11. 11.

    Additional analyses also examined the extent that the TAG effects may vary for students who are historically underserved compared to students who may be more advantaged. As such, I estimated the TAG effect for conditioned samples of low-income versus higher-income students, respectively, using eligibility for the Pell Grant as a threshold for income status. The results suggest that the effect are fairly similar across both groups and relatively consistent with the main results. For this reason, these findings are not presented but the results are available by request.

  12. 12.

    See “Appendix 1” for the full table of descriptive statistics for third year students.

Notes

Funding

Funding was provided by American Educational Research Association (Grant No. Minority Dissertation Fellowship).

References

  1. Alexander, K., Bozick, R., & Entwisle, D. (2008). Warming up, cooling out, or holding steady? Persistence and change in educational expectations after high school. Sociology of Education, 81(4), 371–396.CrossRefGoogle Scholar
  2. Alfonso, M. (2006). The impact of community college attendance on baccalaureate attainment. Research in Higher Education, 47(8), 873–903.CrossRefGoogle Scholar
  3. Avery, C., & Kane, T. J. (2004). Student perceptions of college opportunities. The Boston COACH program. In C. Hoxby (Ed.), College choices: The economics of where to go, when to go, and how to pay for it (pp. 355–394). Chicago, IL: University of Chicago Press.Google Scholar
  4. Bailey, T., Jenkins, D., & Leinbach, T. (2006). Is student success labeled institutional failure? Student goals and graduation rates in the accountability debate at community colleges. New York, NY: Columbia University, Teachers College, Community College Research Center.Google Scholar
  5. Bailey, T., Jenkins, D., & Smith Jaggars, S. (2015). Redesigning America’s community colleges: A clearer path to student success. Cambridge, Massachusetts: Harvard University Press.CrossRefGoogle Scholar
  6. Baker, R. (2016). The effects of structured transfer pathways in community colleges. Educational Evaluation and Policy Analysis, 38(4), 626–646.CrossRefGoogle Scholar
  7. Bettinger, E. P., & Baker, R. B. (2014). The effects of student coaching: An evaluation of a randomized experiment in student advising. Educational Evaluation and Policy Analysis, 36(1), 3–19.CrossRefGoogle Scholar
  8. Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2012). The role of application assistance and information in college decisions: Results from the H&R block FAFSA experiment. The Quarterly Journal of Economics, 127(3), 1205–1242.CrossRefGoogle Scholar
  9. Bourdieu, P., & Wacquant, L. (1992). An invitation to reflexive sociology. Chicago: University of Chicago Press.Google Scholar
  10. Castleman, B. L., Page, L. C., & Schooley, K. (2014). The forgotten summer: Does the offer of college counseling after high school mitigate summer melt among college-intending, low-income high school graduates? Journal of Policy Analysis and Management, 33(2), 320–344.CrossRefGoogle Scholar
  11. Chase, M. M. (2014). Culture, politics, and policy interpretation: How practitioners make sense of a transfer policy in a 2-year college. Educational Policy, 30, 959–998.CrossRefGoogle Scholar
  12. Cohen, A. M., Brawer, F. B., & Kisker, C. B. (2014). The American community college (6th ed.). San Francisco: Jossey-Bass.Google Scholar
  13. Cox, R. D. (2009). The college fear factor: How students and professors misunderstand one another. Cambridge, MA: Harvard University Press.CrossRefGoogle Scholar
  14. Crosta, P. M. (2014). Intensity and attachment: How the chaotic enrollment patterns of community college students relate to educational outcomes. Community College Review, 42(2), 118–142.CrossRefGoogle Scholar
  15. Deil-Amen, R., & Rosenbaum, J. E. (2003). The social prerequisites of success: Can college structure reduce the need for social know-how? The ANNALS of the American Academy of Political and Social Science, 586(1), 120–143.CrossRefGoogle Scholar
  16. Dietrich, C. C., & Lichtenberger, E. J. (2014). Using propensity score matching to test the community college penalty assumption. The Review of Higher Education, 38(2), 193–219.CrossRefGoogle Scholar
  17. Doyle, W. R. (2009). The effect of community college enrollment on bachelor’s degree completion. Economics of Education Review, 28(2), 199–206.CrossRefGoogle Scholar
  18. Fincher, M., Sharp, L., Burks, J., Lyon, K., Parker, M., Ward, J., et al. (2014). Articulation visibility at two-year colleges. Community College Journal of Research and Practice, 38(7), 684–692.CrossRefGoogle Scholar
  19. Gallagher, R. P. (2015). National survey of college counseling centers 2014. Retrieved from http://www.collegecounseling.org/wp-content/uploads/NCCCS2014_v2.pdf.
  20. Gardenhire, A., Collado, H., & Ray, B. (2006). A whole’nother world: Students navigating community college. New York: MDRC.Google Scholar
  21. Goldrick-Rab, S. (2010). Challenges and opportunities for improving community college student success. Review of Educational Research, 80(3), 437–469.CrossRefGoogle Scholar
  22. Goldrick-Rab, S., & Pfeffer, F. T. (2009). Beyond access: Explaining socioeconomic differences in college transfer. Sociology of Education, 82(2), 101–125.CrossRefGoogle Scholar
  23. Gross, B., & Goldhaber, D. (2009). Community college transfer and articulation policies: Looking beneath the surface (Tech. Rep.). (CRPE Working Paper No. 2009_1R). Seattle, WA: University of Washington Bothell, Center on Reinventing Public Education.Google Scholar
  24. Grubb, W. N. (2006). ‘‘Like, what do I do now?’’ The dilemmas of guidance counseling. In T. Bailey & V. S. Morest (Eds.), Defending the community college equity agenda (pp. 195–222). Baltimore, MD: Johns Hopkins University Press.Google Scholar
  25. Hodara, M., Martinez-Wenzl, M., Stevens, D., & Mazzeo, C. (2016). Improving credit mobility for community college transfer students: Findings and recommendations from a 10-state study. Portland, OR: Education Northwest.Google Scholar
  26. Horn, P., & Skomsvold, P. (2011). Community college student outcomes: 1994-2009. Washington, DC: U.S. Department of Education, National Center for Education Statistics.Google Scholar
  27. Hoxby, C., & Turner, S. (2013). Expanding college opportunities for high-achieving, low income students. Stanford: Stanford Institute for Economic Policy Research.Google Scholar
  28. Jenkins, D., & Cho, S. (2013). Get with the program … and finish it: Building guided pathways to accelerate student completion. New Directions for Community Colleges, 164, 27–35.CrossRefGoogle Scholar
  29. Jenkins, D., Lahr, H., & Fink, J. (2017). Building blocks: Laying the groundwork for guided pathways reform in Ohio. New York, NY: Columbia University, Teachers College, Community College Research Center.Google Scholar
  30. Karp, M. M. (2013). Entering a program: Helping students make academic and career decisions. New York, NY: Columbia University, Teachers College, Community College Research Center.Google Scholar
  31. Kisker, C. B., Wagoner, R. L., & Cohen, A. M. (2011). Implementing statewide transfer & articulation reform: An analysis of transfer associate degrees in four states. Los Angeles, CA: Center for the Study of Community Colleges.Google Scholar
  32. LaSota, R. R., & Zumeta, W. (2015). What matters in increasing community college students’ upward transfer to the baccalaureate degree: Findings from the beginning postsecondary study 2003–2009. Research in Higher Education, 57(2), 152–189.CrossRefGoogle Scholar
  33. Long, B. T., & Kurlaender, M. (2009). Do community colleges provide a viable pathway to a baccalaureate degree? Educational Evaluation and Policy Analysis, 31(1), 30–53.CrossRefGoogle Scholar
  34. Melguizo, T., & Dowd, A. C. (2009). Baccalaureate success of transfers and rising 4-year college juniors. Teachers College Record, 111(1), 55–89.Google Scholar
  35. Melguizo, T., Kienzl, G. S., & Alfonso, M. (2011). Comparing the educational attainment of community college transfer students and four-year college rising juniors using propensity score matching methods. The Journal of Higher Education, 82(3), 265–291.CrossRefGoogle Scholar
  36. Monaghan, D. B., & Attewell, P. (2015). The community college route to the bachelor’s degree. Educational Evaluation and Policy Analysis, 37(1), 70–91.CrossRefGoogle Scholar
  37. Moore, C., & Shulock, N. (2011). Sense of direction: The importance of helping community college students select and enter a program of study. Sacramento, CA: Institute for Higher Education Leadership & Policy.Google Scholar
  38. Ohio Board of Regents. (2010). Ohio Transfer and Articulation Policy. Columbus: Ohio Board of Regents. Retrieved March 15, 2012, from http://regents.ohio.gov/transfer/policy/CreditTransferPolicy.pdf.
  39. Person, A. E., Rosenbaum, J. E., & Deil-Amen, R. (2006). Student planning and information problems in different college structures. Teachers College Record, 108(3), 374–396.CrossRefGoogle Scholar
  40. Roksa, J., & Keith, B. (2008). Credits, time, and attainment: Articulation policies and success after transfer. Educational Evaluation and Policy Analysis, 30(3), 236–254.CrossRefGoogle Scholar
  41. Rosenbaum, J. E., Deil-Amen, R., & Person, A. E. (2006). After admission. New York: Russell Sage Foundation.Google Scholar
  42. Sandy, J., Gonzalez, A., & Hilmer, M. J. (2006). Alternative paths to college completion: Effect of attending a 2-year school on the probability of completing a 4-year degree. Economics of Education Review, 25(5), 463–471.CrossRefGoogle Scholar
  43. Scott-Clayton, J. (2011). The shapeless river: Does a lack of structure inhibit students’ progress at community colleges?. New York, NY: Community College Research Center, Columbia University.Google Scholar
  44. Simone, S. A. (2014). Transferability of postsecondary credit following student transfer or coenrollment. statistical analysis report. (NCES 2014-163). Washington, DC: U.S. Department of Education, National Center for Education Statistics.Google Scholar
  45. Smith, M. (2010). Transfer and articulation policies. Denver, CO: Education Commission of the States.Google Scholar
  46. Stephan, J. L., Rosenbaum, J. E., & Person, A. E. (2009). Stratification in college entry and completion. Social Science Research, 38(3), 572–593.CrossRefGoogle Scholar
  47. Tafel, J. (2010). Bringing Down the Silos: A Primer on Credit Transfer and Student Mobility. Columbus: Ohio Board of Regents. Retrieved from http://regents.ohio.gov/transfer/documents/bringing-down-the-silos.pdf.
  48. Van Noy, M., Trimble, M., Jenkins, D., Barnett, E., & Wachen, J. (2016). Guided pathways to careers: Four dimensions of structure in community college career-technical programs. Community College Review, 44(4), 263–285.CrossRefGoogle Scholar
  49. Zeidenberg, M., & Scott, M. (2011). The content of their coursework: Understanding course-taking patterns at community colleges by clustering student transcripts. New York, NY: Columbia University, Teachers College, Community College Research Center.Google Scholar

Copyright information

© Springer Nature B.V. 2018

Authors and Affiliations

  1. 1.New York UniversityNew YorkUSA

Personalised recommendations